From jan.nuyts at ua.ac.be Thu Sep 2 19:49:21 2010 From: jan.nuyts at ua.ac.be (jan nuyts) Date: Thu, 2 Sep 2010 21:49:21 +0200 Subject: Int'l Conference on Grammaticalization and (Inter)Subjectification - Update Message-ID: All information (program, abstracts, registration info, etc.) regarding the International Conference on Grammaticalization and (Inter)Subjectification, to be held in Brussels on November 11-13, 2010, is now available at the conference website at http://webh01.ua.ac.be/gramis/conference/conference.html From tono at ualberta.ca Fri Sep 3 19:02:10 2010 From: tono at ualberta.ca (Tsuyoshi Ono) Date: Fri, 3 Sep 2010 13:02:10 -0600 Subject: Position in Chinese Lingusitics and Pedagogy Message-ID: Assistant Professor - Assistant Professor in Chinese Linguistics and Pedagogy Department of East Asian Studies, University of Alberta Application deadline: 11 October 2010 The Department of East Asian Studies in the Faculty of Arts at the University of Alberta invites applications for a tenure track position at the rank of Assistant Professor in Chinese Linguistics and Pedagogy, which may include any or several of the following areas; Chinese Linguistics, Chinese Applied Linguistics, Chinese Second Language Acquisition, Chinese Language Pedagogy. Applicants should have, or be close to completing, a Ph.D. in an appropriate discipline, and demonstrate outstanding potential for a research career. They should also demonstrate native or near-native fluency in spoken and written Mandarin and English. Responsibilities will include teaching in both undergraduate and graduate student programs, and maintaining an active research program. Experience in teaching Chinese at the college/university level in North America, interest in instructional technology and experience in coordinating a Chinese language program are important assets. The teaching load is four courses per year. Salary is competitive, and will be commensurate with qualifications and experience. Applicants should send a letter of application together with a CV and writing sample, and, if available, a teaching dossier and evaluations of teaching performance to: Professor Yoshi Ono, Chair . Chinese Linguistics Search Committee Department of East Asian Studies 331 Pembina Hall, University of Alberta Edmonton, AB, T6G 2H8, Canada. Applicants must also arrange for three letters of reference to be sent to the Chair. Consideration of applications will begin October 11, 2010 and continue until the position is filled. The effective date of employment will be July 1, 2011. The Department of East Asian Studies (http://www.eastasianstudies.ualberta.ca/) has recently undergone a significant rebuilding process with the addition of several important positions, and efforts are continuing to further expand the Department.s offerings and expertise. The filling of this position is an important part of this process. Inquiries may be directed to Professor Ono at tono at ualberta.ca, or Professor Mikael Adolphson, Acting Chair at madolphs at ualberta.ca. Established in 1908 as a board-governed, public institution, the University of Alberta has earned the reputation of being one of the best universities in Canada based on our strengths in teaching, research, and services. The University of Alberta serves over 36,000 students in more than 200 undergraduate programs and 170 graduate programs (www.ualberta.ca/). The Faculty of Arts is the oldest and most diverse faculty on campus, and one of the largest research and teaching centres in western Canada (www.arts.ualberta.ca). The University.s main campus is located in Edmonton, the vibrant, cosmopolitan capital of the province of Alberta. The Edmonton metropolitan area is the sixth largest in the country with a population of approximately one million (www.edmonton.ca) with excellent health care facilities and recreation opportunities. Edmonton is located only a few hours drive from Banff and Jasper National Parks, which offer skiing in winter and excellent hiking and sightseeing in summer. All qualified candidates are encouraged to apply; however, Canadians and permanent residents will be given priority. The University of Alberta hires on the basis of merit. We are committed to the principle of equity in employment. We welcome diversity and encourage applications from all qualified women and men, including persons with disabilities, members of visible minorities, and Aboriginal persons. From edith at uwm.edu Fri Sep 3 20:26:38 2010 From: edith at uwm.edu (Edith A Moravcsik) Date: Fri, 3 Sep 2010 15:26:38 -0500 Subject: Call for papers re endangered languages Message-ID:   CALL FOR PAPERS   LANGUAGE DEATH, ENDANGERMENT, DOCUMENTATION AND REVITALIZATION   26th UWM Linguistics Symposium University of Wisconsin-Milwaukee Milwaukee, WI, USA October 20-22, 2011   ORGANIZING COMMITTEE Fred Eckman, Elena Mihas, Edith Moravcsik, Sally Noonan, Hamid Ouali, Bernard Perley, Gabriel Rei-Doval , Bozena Tieszen, Kathleen Wheatley   DESCRIPTION In a globalized world where hundreds of languages are expected to become extinct in the 21st century, it is highly relevant to analyze the viability and continuity of threatened languages. The purpose of this symposium is to discuss this impending loss to humankind from a multidisciplinary perspective.   We invite contributions for the assessment of this process from Linguistics, Sociology, Psychology, Anthropology, Education, and related fields. Equally welcome is the participation of practitioners in language revitalization efforts. We wish to combine theoretical and practical perspectives for the analysis of the linguistic and social processes involved in language death, endangerment, documentation and revitalization. Possible topics include the following: - The genetic and areal distribution of endangered   languages - Structural characteristics of endangered     languages - Cultural characteristics of endangered language   communities - Causes of language endangerment - Documentation of endangered languages - Language revitalization programs and practices - Academic ethics and advocacy in language   endangerment   SUBMISSION OF ABSTRACTS (a)    Length:         The abstract may be up to one page of text           plus up to half a page containing possible        examples, charts, and references.   (b)    Format:        The abstract should include the title of the          paper and the text of the abstract but not         the author’s name or affiliation. The e-mail        message to which it is attached should list       the title, the author’s name, and the       author’s affiliation. Abstracts will be         evaluated anonymously.   Please send the message to reidoval at uwm.edu   SUBMISSION DEADLINE : FEBRUARY 1 st , 2011 Authors will be notified on their acceptance status by April 30 th , 2011.   CONFERENCE WEBSITE: https://www4.uwm.edu/letsci/conferences/linguistics2011 or search for UWM Linguistics Symposium From jdavis at ccny.cuny.edu Fri Sep 3 21:30:16 2010 From: jdavis at ccny.cuny.edu (Joseph Davis) Date: Fri, 3 Sep 2010 17:30:16 -0400 Subject: Chi-square Message-ID: A colleague sent to me the June 28 posting by Yuri Tambovtsev below, to which I offer a belated reply that may be useful. The Use of Chi-square by Yuri Tambovtsev Adam Kilgariff wrote that it is not possible to use Chi-square in corpus linguistics. I do not think it is true. One can use Chi-square in linguistics in all cases under the condition that one keeps to the principle of commensurability. That is here, if two samples are equal. I have counted the occurrence of labial consonants in the equal samples of 10000 speech sounds of different Estonian and Russian authors. For instance, in the text of the Estonian writer Aarne Biin «Moetleja» and Enn Vetemaa «Neitsist Suendinud» labial consonants occur 896 and 962 times. Could we say that statistically it is the same? So, we put forward the null hypothesis under the 5% level of significance and one degree of freedom. The theoretical threshold value for Chi-square is 3.841. The actual Chi-square value should be less than 3.841 to state that the occurrence of labial consonant in these two samples is the same. We calculated the Chi-square between 896 and 962. It is 2.344. Thus, it is less than 3.841. So, the two text samples enter the same general sample or in other words it is statistically the same. I wonder if my reasoning is correct. [End of quotation from June 28 posting by Yuri Tambovtsev] The main requirement for the use of the chi-square test of significance is that the observations (data points, tokens) in the sample of some population be statistically independent. That is, there should be no statistical relation between one observation and another in the data set. It should not be possible, given the occurrence of one observation, to predict the next observation or any other observation. In my experience, such independence among observations typically is not a property of connected discourse. Rather, the occurrence of one observation typically raises the probability of the same type of observation occurring next or later in the discourse, no doubt because connected discourse is typically coherent, not random. For instance, if a text in English concerns largely the topic of ‘peace,’ then there will likely be many instances in the text of the labial [p], due to the frequency of the word ‘peace’ and related words (‘peaceful,’ ‘pacify,’ ‘peacenik,’ etc.). By contrast, if another text is about ‘health,’ then it will have a disproportionately high frequency of [h], relative to [p]. Consequently, given any occurrence of a labial in the first text, there will be a somewhat elevated probability for occurrence of a labial next or soon; versus the possibility of predicting another [h] in the second text. This is statistical dependence, not independence. As a result, chi-square is not appropriate as a test of significance; it will likely give an inaccurate measure of the degree to which the sample of labials is representative of the larger population of discourse from which the sample was drawn. (In this case, I suppose we can only imagine a hypothetical population of “English” discourse from which our text was in some idealistic sense “drawn”­-another reason the use of a statistical test of significance may be inappropriate: a text is not in any real sense a sample from a population.) I have a chapter from several years ago that addresses this problem in relation to somewhat different analytical concerns. The reference is: Joseph Davis, 2002, “Rethinking the place of statistics in Columbia School analysis,” in Wallis Reid, Ricardo Otheguy, and Nancy Stern (eds.), Signal, meaning, and message: Perspectives on sign-based linguistics (pp. 65-90). Amsterdam/Philadelphia: John Benjamins. Joseph Davis, Ph.D. Associate Professor School of Education, NAC 6207 The City College New York, NY 10031 From tgivon at uoregon.edu Fri Sep 3 21:59:19 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Fri, 3 Sep 2010 15:59:19 -0600 Subject: Alexandre Kimenyi, RIP In-Reply-To: Message-ID: ALEXANDRE KIMENYI It is with sad heart that I pass on, belatedly, the news of the sad departure of Prof. Alexandre Kimenyi. Out in the boonies, news travel slow, so this is 2 months after the sad facts but it is still a shock. Alexandre was a gentle soul who lighted my early years in linguistics, and his untimely passing brings back faded old memories. I have last seen him ca. 10 years ago when he came to an Africanist conference at Oregon & hung out with one of my last students there, Boniface Kawasha. Otherwise, I had lost touch with Alexandre almost the minute he finished his dissertation. So this takes me back to ancient days at UCLA, when we were all just beginning to learn what the heck we were doing. For me, Alexandre was part of that very slow learning process. Tho I was technically his dissertation adviser, Ed Keenan was the real inspiration for that marvelous work. In fact, it took me a few years before I realized why and how that work was so important. In those days, I was still experimenting (another detour...) with descriptions that had zero formal components. Only after moving to Oregon (1981) and supervising Noel Rude's work on Nez Perce did I realize how important Alexandre's work had been and how central GR's were to the whole mechanism of grammar. Larry Hyman may recall my less-than-clement comments on a beautiful paper he and Annie Hawkinson published in SAL. True, I did accept it without modifications, but under loud protest. I thought it was 'too formal'. I guess I have always been a slow learner. It is always a sad occasion to see a person much younger than oneself depart early, prematurely (well, is there really a timely departure?). But from the obit Larry Hyman furnished, it appears Alexandre has left behind a large family and many friends, students and colleagues who will surely cherish the memory of this gentle, caring, beautiful man. The Ruanda Genocide was, of course a big part of the story. But Alexandre had great difficulties with visas, permits and just plain survival long before that officially-designated horror. I watched him scurry around helplessly in those pre-genocide days, trying to save as many of his numerous relations. His (and our) utter helplessness about Ruanda, and indeed Africa and what still goes on there, is one of the reason I have never gone back to that beautiful horror-plagued continent. Hard to keep watching. But Alexandre didn't have a choice, so he struggled to do as much as could be done from afar. For those of you who work with native peoples of this continent, it is not that difficult to just close your eyes and almost see how just as much of this horror was happening right here a century and a half ago. I live among some of the survivors, indeed on their land. They are thriving by all official measures, but the scars are still there. Will we ever learn? May Alexandre's gentle soul rest in peace. TG ============================================ From yutamb at mail.ru Sat Sep 4 19:04:28 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Sun, 5 Sep 2010 02:04:28 +0700 Subject: Phono-typological distances received by Tambovtsev Message-ID: Dear Funknet colleagues, We have computed the phono-typological distances between Old Russian and the other Slavonic and Baltic languages. The ordered series of the TMB coefficient is as follows: Old Russian - Lithuanian (2.84); Old Russian - Latvian (3,65); Old Russian - Russian (4,71); Old Russian - Ukrainian (5,20); Belorussian - 6,42; Slovenian - 8,60; Czech - 10,29; Bulgarian - 11,08; Macedonian - 13,92; Slovak - 14,20; Serbo-Croatian - 15,31; Sorbian (Luzhits) - 20,16; Polish - 30,54. Therefore, the distribution of the elements of the speech sound chains makes us conclude that Baltic (Lithuanian and Latvian) sound more similarly to the Old Russian language, but not to Russian, Belorussian or Ukrainian as one may expect. Can you share your opinion and write me to yutamb at mail.ru Yuri Tambovtsev, Novosibirsk Ped.University, Russia. From sepkit at utu.fi Mon Sep 6 04:57:16 2010 From: sepkit at utu.fi (=?iso-8859-1?B?IlNlcHBvIEtpdHRpbOQi?=) Date: Mon, 6 Sep 2010 07:57:16 +0300 Subject: FYI: Workshop on role complexes (Zurich, Switzerland) Message-ID: (apologies for multiple postings) Role complexes: (new) approaches to defining semantic roles Since one of the important functions of many communicative acts is to make clear ‘who is doing what to whom’ when portraying a particular state of affairs, it is hardly surprising that semantic roles, thematic roles or thematic relations constitute the topic of countless studies in linguistics and are also always discussed, either explicitly or implicitly, in reference grammars. Numerous studies have dealt with agents and patients within and across languages since the 1970s, and there are several comparatively recent studies that address other roles as well (e.g., Stolz et al 2006 for comitatives, Næss 2007 for transitivity in general, Kittilä 2008 for recipients and goals, Zúñiga & Kittilä 2010 for benefactives among numerous others). Among the many interesting characteristics of accounts of semantic roles, it is noteworthy that semantic relationships between predicates and their arguments are treated in different ways. On the one hand, the explicit formal distinction made in natural languages between agents and patients is typically reflected in their analytical status: the volitionally acting instigator of an event (agent) and the inactive, thoroughly affected target of the event (patient) are invariably regarded as two separate roles. On the other hand, different kinds of beneficiaries (e.g., the first person in John tossed me a salad and John mowed the lawn instead of me) are usually considered instances of one and the same role despite their different meanings. Similarly, different subtypes of agents have tended to be treated as different roles while different kinds of experiencers have not. Against this background, the goal of this workshop is to explore approaches to the notion of semantic role in terms of ROLE COMPLEXES, i.e., of clusters of several related sub-roles that might be distinguished by some constructions in certain languages but are otherwise subsumed under a general umbrella notion. For example, different instances of goals differ according to the exact nature of motion (e.g., he threw the ball to the box / behind the box / on the box). The basic definition of the goal role remains unchanged: we are dealing with an endpoint of motion in all cases. Nevertheless, the potential differences between the roles are thus determined by features not typically considered in studies of semantic roles; features usually used for distinguishing between semantic roles, such as instigation, volitionality and affectedness (cf. e.g.. Næss 2007), can explain neither the semantic nor the formal differences between these three subtypes of goals, or the different codi ng of goals and beneficiaries. We welcome all abstracts dealing with role complexes within and across languages. Possible topics for papers include (but, as always, are not restricted to) the following: - When should we speak of distinct roles, and when are two slightly different (potentially differently coded) roles rather manifestations of one basic role? Are, e.g., inanimate goals and animate goals manifestations of a single role or should they rather be treated separately? - What consequences does role synonymy have for our understanding of semantic roles? What are the features that any adequate theory of semantic roles should consider, what is the ‘correct’ number of semantic roles, etc.? - How should we treat partial formal mismatches between roles? - How do we deal with semantically/pragmatically determined differences in the coding of roles (e.g., marking conditioned by definiteness, referentiality, specificity, topicality, focality)? - Corpus-based studies of role synonymy: What determines the use of different (yet semantically similar) manifestations of a role in actual language use? - How do we best treat the diachronic development of multifunctional coding devices (syncretisms, polysemies, homonymies, etc.)? - Formal manifestation of semantic role synonymy: case marking, verbal marking, lexical differences, etc. - Role synonymy of core and peripheral roles: Are there any differences, is synonymy more common for one of these? Organizers of the workshop Fernando Zúñiga (Zurich) and Seppo Kittilä (Helsinki) Venue University of Zurich, Zurich, Switzerland Dates April 4-5, 2011 Abstract submission Please send your (maximally) 500-word abstract (excluding data and references) to both fernando.zuniga at spw.uzh.ch and kittila at mappi.helsinki.fi no later than November 14, 2010. The letters of acceptance will be sent by December 12, 2010. Abstracts must be anonymous, but the body of the e-mail should include the following information: Name of the author(s) Title of the paper Affiliation(s) E-mail In case you have any questions about the workshop, please don’t hesitate to contact us. We are looking forward to welcoming you all to Zurich. Fernando and Seppo From n.m.stukker at hum.leidenuniv.nl Tue Sep 7 11:47:36 2010 From: n.m.stukker at hum.leidenuniv.nl (Stukker, N.M.) Date: Tue, 7 Sep 2010 13:47:36 +0200 Subject: Call for papers: Stylistics across disciplines Message-ID: Stylistics across disciplines University of Leiden, The Netherlands June 16-17, 2011 Confirmed keynote speakers: Prof. dr. Barbara Dancygier, University of British Columbia, Vancouver (Canada) Prof. dr. Arie Verhagen, Leiden University (The Netherlands) Stylistics is a field of study that is growing and developing fast. Its central concern is the way cognitive and communicative effects are achieved by means of linguistic choices. It therefore encompasses literary studies and linguistics as well as discourse studies. In spite of the shared, overarching definition of what it is, the field of study of Stylistics is highly fragmented. It mainly takes place within the boundaries of the various, more traditional, domains of study, e.g. literary analysis, rhetoric, (critical) discourse analysis, applied linguistics, etc. As a result, a comprehensive understanding of the wide variety of interests and foci of attention in stylistic studies, as well as exchange of knowledge between these research domains, is developing relatively slowly. In recent years, successful attempts have been made to take an integrative, cross-disciplinary perspective on Stylistics, focusing on the shared research object: language use. An example is the expanding body of studies associated with the International Poetics and Linguistics Association (PALA). Especially fruitful has proven to be the developing area of 'cognitive poetics', a field closely allied with the theoretical framework of cognitive linguistics, which includes attention for contextual factors and the inherently 'subjective' basis of language in linguistic analysis. This Stylistics across disciplines conference links up with these developments and intends to offer a platform for exchange of ideas and to stimulate fruitful collaboration among linguists, literary scholars and discourse scholars studying 'style'. We invite participants from all relevant fields to participate in the Stylistics across disciplines conference to discuss the opportunities and problems regarding the development of stylistics as a coherent and methodologically sound research discipline. We welcome papers on (but not limited to) the subject of: * Possibilities and limitations of an interdisciplinary perspective: what can literary scholars learn from the way style is studied in linguistics or rhetoric, and vice versa? * Opportunities and problems of a 'linguistic stylistics' * Methodological issues: qualitative (interpretive analysis) or quantitative methods (digital humanities, corpus stylistics) and different research methods (corpus analysis, experimental effect studies) in relation to various research contexts * Development of theoretical notions and analytical tools especially suited for stylistic analysis * Context-sensitivity of stylistic patterns and analysis: how does stylistic choice interact with contextual factors such as institution, genre characteristics, etc.? * Language specificity and culture specificity of stylistic phenomena and analysis Please submit your abstract (in Word or PDF format, containing the title of your paper, author's name(s) and affiliation(s), max. 500 words) to stylistics at hum.leidenuniv.nl. The deadline for abstract submission is December 15, 2010. Notification of acceptance will be by February 1, 2011. Organizing committee: Suzanne Fagel Maarten van Leeuwen Ninke Stukker stylistics at hum.leidenuniv.nl Scientific committee: Jaap Goedegebuure (literary studies) Ton van Haaften (language and communication) Jaap de Jong (rhetoric) Arie Verhagen (linguistics) The Stylistics across disciplines conference is organized by researchers from the NWO research project Stylistics of Dutch (Leiden University 2007-2012); website: www.stylistics.leidenuniv.nl From Freek.VanDeVelde at arts.kuleuven.be Tue Sep 7 13:17:26 2010 From: Freek.VanDeVelde at arts.kuleuven.be (Freek Van de Velde) Date: Tue, 7 Sep 2010 15:17:26 +0200 Subject: workshop proposal 'exaptation' at ICHL 20 Message-ID: Preliminary call for papers for a workshop at ICHL 20 (Osaka, 25-30 July 2011) Subject: Exaptation Convenors: Muriel Norde (University of Groningen) & Freek Van de Velde (University of Leuven) Contact: freek.vandevelde at arts.kuleuven.be Deadline: 13 september 2010 (no specific title or abstract needed at this point. Just let us know whether you are interested in participating). Although some morphological changes seem to obey general tendencies, as formulated for instance by Kuryłowicz or Mańczak (see Hock 1986, ch.10) or Van Loon (2005), most of these tendencies can just as easily be violated. Diachronic morphology is largely idiosyncratic (Joseph 1998). Morphological paradigms appear to be ripped up at random in order to establish "local generalizations" (Joseph 1992). One particular way in which unpredictable changes come about is through 'bricolage' with junk morphology, which goes under the name of exaptation (Lass 1990, 1997: 316ff.). Exaptation is a concept that was first used in evolutionary biology, to refer to co-optation of a certain trait for a new function. A typical example is the use of feathers, originally serving a thermo-regulatory function, for flight. In linguistics, exaptation is defined as follows: "Say a language has a grammatical distinction of some sort, coded by means of morphology. Then say this distinction is jettisoned, PRIOR TO the loss of the morphological material that codes it. This morphology is now, functionally speaking, junk; and there are three things that can in principle be done with it: (i) it can be dumped entirely; (ii) it can be kept as marginal garbage or nonfunctional/nonexpressive residue (suppletion, 'irregularity'); (iii) it can be kept, but instead of being relegated as in (ii), it can be used for something else, perhaps just as systematic. (...) Option (iii) is linguistic exaptation." (Lass 1990: 81-82) Lass originally understood exaptation in a rather narrow sense. First, the term exaptation was reserved for changes affecting functionless (or 'junk') morphology. Second, in order to qualify as exaptation, the new function of a morpheme needed to be entirely novel. In Lass's own words: "Exaptation then is the opportunistic co-optation of a feature whose origin is unrelated or only marginally related to its later use. In other words (loosely) a 'conceptual novelty' or 'invention'." Both criteria have been criticized. With regard to the first criterion, Vincent (1995: 435), Giacalone Ramat (1998), Smith (2006) and Willis (ms.) pointed out difficulties with regard to the notion of junk. And indeed, Lass later stretched his notion of exaptation, admitting that linguistic exaptation - just like biological exaptation - could also affect non-junk morphology (see Lass 1997: 318), to the effect that the old and the new function may co-exist. Doubt has also been raised with regard to the second criterion, the novelty of the new function, which is central to the notion of exaptation according to Lass (1990: 82) (see also Norde 2001: 244, 2009: 117 and Traugott 2004). Some scholars have argued against the purported novelty of the function after exaptation (Vincent 1995: 436; Giacalone Ramat 1998, Hopper & Traugott 2003: 135-136). If this criterion is jettisoned, we arrive at a fairly broad definition of exaptation, like for instance in Booij (2010: 211), who defines it as "[t]he re-use of morphological markers". Such a broad conception of exaptation is in line with the notion in evolutionary biology, where neither of the two criteria is decisive for the application of the term to shifts in function, but the question then arises whether this does not make the concept vacuous (see De Cuypere 2005). Despite these criticisms, exaptation has been used as a convenient label for morphological changes that at first sight seem to proceed unpredictably, e.g. by running counter to grammaticalization clines (see Norde 2009: 115-118). It has been applied to various cases of morphological change, discussed in Lass (1990), Norde (2002), Fudeman (2004), Van de Velde (2005, 2006), Narrog (2007), Booij (2010, ms.), Willis (ms.) among others. In this workshop, we aim to come to terms with exaptation. Apart from specific case studies drawing on original data, we welcome papers that address the following issues: (1) Do we need exaptation in diachronic morphology, or does it reduce to more traditional mechanisms such as reanalysis and analogy, as e.g. De Cuypere (2005) argues? (2) Does exaptation only apply to morphology (Heine 2003: 173), or is it relevant to syntactic change as well, as Brinton & Stein (1995) have argued? (3) Does exaptation presuppose irregularity and unpredictability? If so, does this entail that exaptation is language-specific (as argued by Heine 2003: 173), and that cross-linguistics generalizations are not possible? See, however, Narrog (2007) for evidence to the contrary. (4) Does exaptation happen primarily in cases of 'system disruption', such as typological word order change or deflection (see Norde 2002: 49, 60, 61)? (5) How should we define the concept of 'novelty', and is it a useful criterion for a change to be qualified as exaptation? Currently, there seem to be different views in the literature on what is exactly understood by a 'new' function. Does this mean (a) an entirely new category in the grammar, (b) a function unrelated to the morpheme's old function, or (c) a different though perhaps not totally unrelated function from the old function? (6) Is exaptation infrequent (Heine 2003:174, Traugott 2004) and non-recurrent (as argued by Heine 2003: 172)? Or can one morpheme undergo several successive stages of exaptation (as argued by Giacalone Ramat 1998: 110-111 with regard to the -sk- suffix and by Van de Velde 2006 with regard to the Germanic adjective inflection)? (7) What is the relation between exaptation and grammaticalization? Do they refer to fundamentally different kinds of changes (Vincent 1995), is exaptation a final stage of grammaticalization (Greenberg 1991, Traugott 2004), or are exaptation and grammaticalization just two different labels for the same type of change? After all, both processes involve reanalysis (Narrog 2007), both processes can come about through pragmatic strengthening (see Croft 2000: 126-130). Furthermore, if the old and new function of the exaptatum co-exist (see above) and if the new function is related to the old one, then exaptation involves 'layering' and 'persistence', respectively (see Van de Velde 2006: 61-62), which are also key features of grammaticalization (see Hopper 1991). (8) What is the relation between exaptation and degrammaticalization? Does exaptation always entail some sort of 'degrammaticalization' (as argued by Heine 2003 and arguably Narrog 2007: 9, 18), or does exaptation often, but not always, go together with degrammaticalization (Norde 2009: 118)? (9) Is exaptation the same thing as what Greenberg (1991) understands by 'regrammaticalization' and as what Croft (2000) understands by 'hypoanalysis', or are there significant differences between these concepts? And what is the overlap with related concept such as 'functional renewal' (Brinton & Stein 1995)? References Booij, G. 2010 (to appear). Construction morphology. Oxford: Oxford University Press. Booij, G. manuscript. Recycling morphology: Case endings as markers of Dutch constructions. . Brinton, L. & D. Stein. 1995. Functional renewal. In: H. Andersen (ed.), Historical Linguistics 1993. Amsterdam: Benjamins. 33-47. Croft, W. 2000. Explaining language change. An evolutionary approach. Harlow : Longman. De Cuypere, L. 2005. Exploring exaptation in language change. Folia Linguistica Historica 26: 13-26. Fudeman, K. 2004. Adjectival agreement vs. adverbal inflection in Balanta. Lingua 114: 105-23. Giacalone Ramat, A. 1998. Testing the boundaries of grammaticalization. In: A. Giacalone Ramat & P.J. Hopper (eds.), The limits of grammaticalization. Amsterdam: Benjamins. 227-270. Greenberg, J.H. 1991. The last stages of grammatical elements: Contractive and expansive desemanticization. In: E.C. Traugott & B. Heine (eds.), Approaches to grammaticalization. Amsterdam: Benjamins. 301-314. Heine, B. 2003. On degrammaticalization. In: B.J. Blake & K. Burridge (eds.), Historical linguistics 2001. Amsterdam: Benjamins. 163-179. Hock, H.H. 1986. Principles of historical linguistics. Berlin: de Gruyter. Hopper, P.J. 1991. On some principles of grammaticalization. In: E.C. Traugott & B. Heine (eds.), Approaches to grammaticalization. Amsterdam: Benjamins. 17-35. Hopper, P.J. & E.C. Traugott. 2003. Grammaticalization. 2nd edn. Cambridge: Cambridge University Press. Joseph, B.D. 1992. Diachronic explanation: Putting the speaker back into the picture. In: G.W. Davis & G.K. Iverson (eds.), Explanations in historical linguistics. Amsterdam: Benjamins. 123-144. Joseph, B.D. 1998. Diachronic morphology. In: A. Spencer & A.M. Zwicky (eds.), Handbook of morphology. Oxford: Blackell. 351-373. Lass, R. 1990. How to do things with junk: Exaptation in language evolution. Journal of Linguistics 26: 79-102. Lass, R. 1997. Historical linguistics and language change. Cambridge: Cambridge University Press. Narrog, H. 2007. Exaptation, grammaticalization, and reanalysis. California Linguistic Notes 32 (1). . Norde, M. 2001. Deflexion as a counterdirectional factor in grammatical change. Language Sciences 23: 231-264. Norde, M. 2002. The final stages of grammaticalization: Affixhood and beyond. In: I. Wischer & G. Diewald (eds.), New reflections on grammaticalization. Amsterdam: Benjamins. 45-81. Norde, M. 2009. Degrammaticalization. Oxford: Oxford University Press. Smith, J.C. 2006. How to do things without junk: the refunctionalization of a pronominal subsystem between Latin and Romance. In: J.-P.Y. Montreuil (ed.), New perspectives on Romance linguistics. Volume II: Phonetics, phonology and dialectology. Amsterdam: Benjamins. 183-205. Traugott, E.C. 2004. Exaptation and grammaticalization. In: M. Akimoto (ed.), Linguistic studies based on corpora. Tokyo: Hituzi Syobo. 133-156. Van de Velde, F. 2005. Exaptatie en subjectificatie in de Nederlandse adverbiale morfologie [Exaptation and subjectification in Dutch adverbial morphology]. Handelingen der Koninklijke Zuid-Nederlandse Maatschappij voor Taal- en Letterkunde en Geschiedenis 58: 105-124. Van de Velde, F. 2006. Herhaalde exaptatie. Een diachrone analyse van de Germaanse adjectiefflexie [Iterative exaptation. A diachronic analysis of the Germanic adjectival inflection]. In: M. Hüning, A. Verhagen, U. Vogl & T. van der Wouden (eds.), Nederlands tussen Duits en Engels. Leiden: Stichting Neerlandistiek Leiden. 47-69. Van Loon, J. 2005. Principles of historical morphology. Heidelberg: Universitätsverlag Winter. Vincent, N. 1995. Exaptation and grammaticalization. In: H. Andersen (ed.), Historical linguistics 1993. Amsterdam: Benjamins. 433-445. Willis, D. Manuscript. Degrammaticalization and obsolescent morphology: Evidence from Slavonic. < http://people.pwf.cam.ac.uk/dwew2/willis_degramm_berlin.pdf>. Freek Van de Velde Postdoctoral research fellow Research Foundation Flanders (FWO), University of Leuven Fac. of Arts, Dept. of Linguistics Blijde Inkomststraat 21, P.O. Box 3308 BE-3000 Leuven Tel. 0032 16 32 47 81 Fax 0032 16 32 47 67 http://wwwling.arts.kuleuven.be/nedling/fvandevelde/ From bischoff.st at gmail.com Wed Sep 8 12:46:55 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Wed, 8 Sep 2010 08:46:55 -0400 Subject: analysis: unhappiness Message-ID: Hi all, I had an interesting exchange with a few generative syntacticians/morphologists (former classmates of mine) regarding an analysis of "unhappiness". Two things that they said surprised me a bit, they are the following: (1) un- (negation, 'not') only attaches to adjectives (now this clearly isn't the case, a simple cursory view of the etymology in the OED provides a number of examples of un- with nouns and verbs...though to significantly lesser degrees...in addition works on English morphology contain examples as well) (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis such as [un-[happy-ness]] is impossible (due to (1) above according to my former colleagues). My questions are the following: (1) Is there a good/well grounded reason to believe un- "only" attaches to adjectives? (2) What would be the consensus on an analysis of "unhappiness" that most linguists would agree on? Thanks, Shannon From amnfn at well.com Wed Sep 8 13:05:11 2010 From: amnfn at well.com (A. Katz) Date: Wed, 8 Sep 2010 06:05:11 -0700 Subject: analysis: unhappiness In-Reply-To: Message-ID: Many of the functionalist linguists I know take the opposite view that derviational morphology is not really active in English, and that people just memorize the lexemes "unhappy", "unhappiness", and "happy"/"happiness" separately. The idea is that while speakers recognize that these words are related, they do not rederive them every time, and they learned them from others, rather than putting them together by themselves. I personally think derivational morphology is a little more active than that, but there you have your two theoretical extremes, the one you related about your former colleagues and the one I related about my former colleagues. Best, --Aya On Wed, 8 Sep 2010, s.t. bischoff wrote: > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon > > From dan at daneverett.org Wed Sep 8 13:07:26 2010 From: dan at daneverett.org (Daniel Everett) Date: Wed, 8 Sep 2010 09:07:26 -0400 Subject: analysis: unhappiness Message-ID: Shannon, There is a long literature on this under the heading of 'bracketing paradoxes'. One of the best articles came out in the early 90s in Language, by Andy Spencer. Though just about everyone and their dog was writing on it back then. The generative analysis is what leads to the 'paradoxes', which are either a discovery or an error depending on your perspective. -- Dan On Sep 8, 2010, at 8:46 AM, s.t. bischoff wrote: > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon From amber at cs.toronto.edu Wed Sep 8 13:38:09 2010 From: amber at cs.toronto.edu (L. Amber Wilcox-O'Hearn) Date: Wed, 8 Sep 2010 09:38:09 -0400 Subject: analysis: unhappiness Message-ID: On Wed, Sep 8, 2010 at 6:46 AM, s.t. bischoff wrote: > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) Maybe what they are saying is not that there is no un- attaching to verbs, there obviously are many examples. But in those cases, un- is not negation, it is a reverse action - 'untie' does not mean 'not tie'. \ L. Amber Wilcox-O'Hearn * http://www.cs.toronto.edu/~amber/ / -\ Graduate student * Computational Linguistics Research Group /- --\ Department of Computer Science * University of Toronto /-- From jrubba at calpoly.edu Wed Sep 8 15:26:29 2010 From: jrubba at calpoly.edu (Johanna Rubba) Date: Wed, 8 Sep 2010 08:26:29 -0700 Subject: analysis: unhappiness In-Reply-To: <20100908133810.00EE92B01BD@apps0.cs.toronto.edu> Message-ID: One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. Dr. Johanna Rubba, Ph. D. Professor, Linguistics Linguistics Minor Advisor English Dept. Cal Poly State University San Luis Obispo San Luis Obispo, CA 93407 Ofc. tel. : 805-756-2184 Dept. tel.: 805-756-2596 Dept. fax: 805-756-6374 E-mail: jrubba at calpoly.edu URL: http://cla.calpoly.edu/~jrubba From dubartell at edinboro.edu Wed Sep 8 16:41:52 2010 From: dubartell at edinboro.edu (DUBARTELL) Date: Wed, 8 Sep 2010 12:41:52 -0400 Subject: cc: Re: analysis: unhappiness Message-ID: Hello All, I think this issue is not perhaps one of analysis or intuition, but one of K-12 teaching/learning. Many teachers provide exercises over the school years that reinforce prefixation with less lesson time given to the suffixes. My students can easily explain the meanings of pre-, anti-, and ex-, but are often unable to explain the meaning of -ness, -tion-, and -ence, for example, as readily. I know from teaching student teachers that many of their grammar lesson plans focus primarily on prefixes. (In addition, some students do not realize that suffixes appear in dictionaries). I must also wonder if the stress difficulties are related to the possible infrequency of such lessons in the K-12 curricula. Very generally, many college students are better at those grammatical exercises with which they may have at least some familiarity from their K-12 experience. Regards, Deborah DuBartell On Wednesday, September 08, 2010 11:26 AM, Johanna Rubba wrote: > >Date: Wed, 8 Sep 2010 08:26:29 -0700 >From: Johanna Rubba >To: funknet at mailman.rice.edu >cc: >Subject: Re: [FUNKNET] analysis: unhappiness > >One thing that consistently occurs in my intro linguistics classes is >that at least half of my students do not analyze complex words the >way a linguist would -- many would analyze "unhappiness" as "un" + >"happiness." They make such analyses over and over. It makes one >wonder, of course, about how much native-speaker intuition is in >agreement with some linguistic analyses. I can *feel* that the >analysis is [[un-happy]-ness], but, apparently, large numbers of >native speakers cannot. > >Another thing I often find is that many students cannot locate either >primary or (especially) secondary stress in words. This is very >bizarre, considering that they produce the stresses correctly and >hear them correctly in others' speech. So many are unsuccessful at >this that I have stopped requiring them to find stress in words on >tests. I give them tricks like singing the word and monitoring for >the highest-pitched syllable, but the tricks don't work. That many >students can't be tone-deaf. > >Dr. Johanna Rubba, Ph. D. >Professor, Linguistics >Linguistics Minor Advisor >English Dept. >Cal Poly State University San Luis Obispo >San Luis Obispo, CA 93407 >Ofc. Tel. : 805-756-2184 >Dept. Tel.: 805-756-2596 >Dept. fax: 805-756-6374 >E-mail: jrubba at calpoly.edu >URL: http://cla.calpoly.edu/~jrubba > > > Deborah DuBartell, Ph.D. Professor of Linguistics Department of English 295 Meadville St. Centennial Hall 238 Edinboro University of Pennsylvania Edinboro, PA 16444 tele: 814.732.2269 web: users.edinboro.edu/DuBartell From djh514 at york.ac.uk Wed Sep 8 17:18:49 2010 From: djh514 at york.ac.uk (Damien Hall) Date: Wed, 8 Sep 2010 18:18:49 +0100 Subject: Audit of School-Leavers' Grammatical Knowledge In-Reply-To: Message-ID: Inspired by the 'unhappiness' thread and the subsequent reflections on what grammar / grammatical terms students do and do not know, I'm reposting here a link to something that I saw on the LINGUIST List this morning, in case others did not see it. Dick Hudson would appreciate it if colleagues not in the UK would administer an apparently simple questionnaire to their beginning-level Linguistics classes, to gauge the amount of grammatical terminology they know when they enter University (ie after their high-school education). Further details are on LINGUIST, here: http://linguistlist.org/pubs/sums/query-details.cfm?submissionid=2646814 Damien -- Damien Hall University of York Department of Language and Linguistic Science Heslington YORK YO10 5DD UK Tel. (office) +44 (0)1904 432665 (mobile) +44 (0)771 853 5634 Fax +44 (0)1904 432673 http://www.york.ac.uk/res/aiseb http://www.york.ac.uk/depts/lang/people/pages/hall.htm DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm From lesleyne at msu.edu Wed Sep 8 19:02:50 2010 From: lesleyne at msu.edu (Diane Frances Lesley-Neuman) Date: Wed, 8 Sep 2010 15:02:50 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID:  You are getting into the Lexical Phonology of English debate over strata.  To begin with consult: April Mcmahon (1999) Lexical Strata and the History of English Hans Giegerich  Lexical Strata in English The old literature was debated throughout the 1980's: Paul Kiparsky, K.P. Mohanan (with Morris Halle), Jerzy Rubach and Geert Booij come to mind.  Toni Borowsky as well, although her analysis has come under question.  i also think that Ricardo Bermudez-Otero, an English lexical phonology and stratal OT specialist at Manchester somewhere has a critique of the various models or can guide you further.  In general this is the object of study of the English phonology faculty at Edinburgh where Mcmahon and Giegerich are and I think an another specialist person to weigh in on the issue would be Patrick Honeybone, a specialist in the diachronic phonology of English also Edinburgh-based.   ______________________________ Diane Lesley-Neuman Linguistics Program Wells A-614 Michigan State University East Lansing, MI 48824 Quoting "s.t. bischoff" : > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the  following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to  believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon > From Lise.Menn at Colorado.EDU Thu Sep 9 00:16:15 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Wed, 8 Sep 2010 18:16:15 -0600 Subject: analysis: unhappiness In-Reply-To: Message-ID: So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > One thing that consistently occurs in my intro linguistics classes > is that at least half of my students do not analyze complex words > the way a linguist would -- many would analyze "unhappiness" as "un" > + "happiness." They make such analyses over and over. It makes one > wonder, of course, about how much native-speaker intuition is in > agreement with some linguistic analyses. I can *feel* that the > analysis is [[un-happy]-ness], but, apparently, large numbers of > native speakers cannot. > > Another thing I often find is that many students cannot locate > either primary or (especially) secondary stress in words. This is > very bizarre, considering that they produce the stresses correctly > and hear them correctly in others' speech. So many are unsuccessful > at this that I have stopped requiring them to find stress in words > on tests. I give them tricks like singing the word and monitoring > for the highest-pitched syllable, but the tricks don't work. That > many students can't be tone-deaf. > > Dr. Johanna Rubba, Ph. D. > Professor, Linguistics > Linguistics Minor Advisor > English Dept. > Cal Poly State University San Luis Obispo > San Luis Obispo, CA 93407 > Ofc. tel. : 805-756-2184 > Dept. tel.: 805-756-2596 > Dept. fax: 805-756-6374 > E-mail: jrubba at calpoly.edu > URL: http://cla.calpoly.edu/~jrubba > > > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 http://spot.colorado.edu/~menn/index.html Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Fellow, Linguistic Society of America Campus Mail Address: UCB 594, Institute for Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From tgivon at uoregon.edu Thu Sep 9 01:03:48 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Wed, 8 Sep 2010 19:03:48 -0600 Subject: analysis: unhappiness In-Reply-To: <9DE8281B-3822-43CF-8001-C70B10E98441@colorado.edu> Message-ID: Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG ============== Lise Menn wrote: > So we see an important phenomenon: Tacit knowledge really IS tacit, > and 'intuitions' are very poor guides to what our minds are doing when > we are using the patterns of our language as speakers/hearers. > Introspection cannot replace observation of actual usage and > psycholinguistic experiments; it can only act as a suggestion of where > to dig. After all, we can't figure out vision or digestion by > thinking about how they feel, although we certainly have to account > for subjective feelings of contrast and indigestion. The same is true > for language, mutatis mutandis. > > On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > >> One thing that consistently occurs in my intro linguistics classes is >> that at least half of my students do not analyze complex words the >> way a linguist would -- many would analyze "unhappiness" as "un" + >> "happiness." They make such analyses over and over. It makes one >> wonder, of course, about how much native-speaker intuition is in >> agreement with some linguistic analyses. I can *feel* that the >> analysis is [[un-happy]-ness], but, apparently, large numbers of >> native speakers cannot. >> >> Another thing I often find is that many students cannot locate either >> primary or (especially) secondary stress in words. This is very >> bizarre, considering that they produce the stresses correctly and >> hear them correctly in others' speech. So many are unsuccessful at >> this that I have stopped requiring them to find stress in words on >> tests. I give them tricks like singing the word and monitoring for >> the highest-pitched syllable, but the tricks don't work. That many >> students can't be tone-deaf. >> >> Dr. Johanna Rubba, Ph. D. >> Professor, Linguistics >> Linguistics Minor Advisor >> English Dept. >> Cal Poly State University San Luis Obispo >> San Luis Obispo, CA 93407 >> Ofc. tel. : 805-756-2184 >> Dept. tel.: 805-756-2596 >> Dept. fax: 805-756-6374 >> E-mail: jrubba at calpoly.edu >> URL: http://cla.calpoly.edu/~jrubba >> >> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > http://spot.colorado.edu/~menn/index.html > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > Fellow, Linguistic Society of America > > Campus Mail Address: > UCB 594, Institute for Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From jrubba at calpoly.edu Thu Sep 9 01:20:17 2010 From: jrubba at calpoly.edu (Johanna Rubba) Date: Wed, 8 Sep 2010 18:20:17 -0700 Subject: analysis: unhappiness In-Reply-To: <4C883274.3020503@uoregon.edu> Message-ID: The reason I make much of my students' disagreements with linguists' analyses is that, at least as I recall from my graduate education, native-speaker intuitions were routinely used in defense of analyses of sentence structure. I agree completely with Lise that introspection alone is not sufficient for defending an analysis. I never meant to suggest that my students store complex words disconnected from their component parts. Such a notion would be bizarre for a practitioner of Cognitive Grammar. Plus, the behavior I described doesn't contradict network connections; it challenges the usefulness of NS intuitions, which would, in turn, challenge analyses defended based on same. It's odd that, in some cases, students' tacit knowledge can be brought to consciousness, and in other cases not. It seems clear that context (including priming) affects a language-user's analysis. Certainly, many expressions that sound odd out of context sound perfectly fine in a suitable context. Jo On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG ============== Lise Menn wrote: > So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. > > On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > >> One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. >> >> Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. >> >> Dr. Johanna Rubba, Ph. D. >> Professor, Linguistics >> Linguistics Minor Advisor >> English Dept. >> Cal Poly State University San Luis Obispo >> San Luis Obispo, CA 93407 >> Ofc. tel. : 805-756-2184 >> Dept. tel.: 805-756-2596 >> Dept. fax: 805-756-6374 >> E-mail: jrubba at calpoly.edu >> URL: http://cla.calpoly.edu/~jrubba >> >> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > http://spot.colorado.edu/~menn/index.html > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > Fellow, Linguistic Society of America > > Campus Mail Address: > UCB 594, Institute for Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > Dr. Johanna Rubba, Professor, Linguistics Linguistics Minor Advisor English Department California Polytechnic State University, San Luis Obispo E-mail: jrubba at calpoly.edu Tel.: 805.756.2184 Dept. Ofc. Tel.: 805.756.2596 Dept. Fax: 805.756.6374 URL: http://www.cla.calpoly.edu/~jrubba From Lise.Menn at Colorado.EDU Thu Sep 9 02:31:31 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Wed, 8 Sep 2010 20:31:31 -0600 Subject: analysis: unhappiness In-Reply-To: <4C883274.3020503@uoregon.edu> Message-ID: if you-all will forgive me for a little advance self-promotion, my new book, Psycholinguistics: Introduction and Applications, will be available from Plural Publishing by the end of the year. It's intended to be completely accessible for anyone (linguist, psychologist, speech-language pathologist, language teacher...) who needs to be able to think about brain and language, and what Tom says about experimental techniques and word storage is one of the things I explore, starting from analyses of slips of the tongue. Also a chapter each on the brain, reading, language development, aphasia, and second language learning. You can see the full table of contents on the web page for the book: http://www.pluralpublishing.com/publication_psl.htm Best regards to all, Lise On Sep 8, 2010, at 7:03 PM, Tom Givon wrote: > > Right on, Lise. And further, there is a well-known experimental > technique called "semantic priming" that is admirably well suited > for investigating whether when a language used hears "unhappiness", > "happy" and "happiness" are activated ('come to mind'). This > technique will probably not answer the question of the differential > bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to > answer questions of directionality (does "unhappy" prime "happy" > stronger than vice versa?). But it does tends to suggest that we > don't store complex words in total disconnect from their parts, at > least not as frequent adult users. And that phonological similarity > (shared parts of words) has semantic consequences. Cheers, TG > > ============== > > > Lise Menn wrote: >> So we see an important phenomenon: Tacit knowledge really IS tacit, >> and 'intuitions' are very poor guides to what our minds are doing >> when we are using the patterns of our language as speakers/ >> hearers. Introspection cannot replace observation of actual usage >> and psycholinguistic experiments; it can only act as a suggestion >> of where to dig. After all, we can't figure out vision or >> digestion by thinking about how they feel, although we certainly >> have to account for subjective feelings of contrast and >> indigestion. The same is true for language, mutatis mutandis. >> >> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >> >>> One thing that consistently occurs in my intro linguistics classes >>> is that at least half of my students do not analyze complex words >>> the way a linguist would -- many would analyze "unhappiness" as >>> "un" + "happiness." They make such analyses over and over. It >>> makes one wonder, of course, about how much native-speaker >>> intuition is in agreement with some linguistic analyses. I can >>> *feel* that the analysis is [[un-happy]-ness], but, apparently, >>> large numbers of native speakers cannot. >>> >>> Another thing I often find is that many students cannot locate >>> either primary or (especially) secondary stress in words. This is >>> very bizarre, considering that they produce the stresses correctly >>> and hear them correctly in others' speech. So many are >>> unsuccessful at this that I have stopped requiring them to find >>> stress in words on tests. I give them tricks like singing the word >>> and monitoring for the highest-pitched syllable, but the tricks >>> don't work. That many students can't be tone-deaf. >>> >>> Dr. Johanna Rubba, Ph. D. >>> Professor, Linguistics >>> Linguistics Minor Advisor >>> English Dept. >>> Cal Poly State University San Luis Obispo >>> San Luis Obispo, CA 93407 >>> Ofc. tel. : 805-756-2184 >>> Dept. tel.: 805-756-2596 >>> Dept. fax: 805-756-6374 >>> E-mail: jrubba at calpoly.edu >>> URL: http://cla.calpoly.edu/~jrubba >>> >>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> http://spot.colorado.edu/~menn/index.html >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> Fellow, Linguistic Society of America >> >> Campus Mail Address: >> UCB 594, Institute for Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From dick at ling.ucl.ac.uk Thu Sep 9 07:51:19 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Thu, 9 Sep 2010 08:51:19 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: This discussion about the role of native-speaker intuition treats all native speakers and all levels of language equally. And yet we all agree with Joanna when she says "I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. " Why? Certainly not because we've got psycholinguistic evidence on this particular word. Isn't it something to do with maturity and training for the speakers, and meaningfulness for the levels? Think of that classic 1979 experiment by the Gleitmans that found massive differences in sensitivity to language structure with both age and education, so that /eat house bird/ is interpreted as 'a house-bird who is very eat' by students with a PhD (but not in linguistics) but as 'everybody is eating up their pet birds' by clerical staff. In this case, as in the other examples they tried, the more educated were right (by our standards) and the less educated were wrong. They also found big differences in reliability from level to level, with semantic judgements easiest and most reliable and phonological judgements least reliable, and syntax in between. That's presumably because ordinary speakers spend most of their time grappling with meaning ('Look after the sense and the sounds will look after themselves', as someone said to Alice in Wonderland). We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. Best wishes, Dick Hudson Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 09/09/2010 02:20, Johanna Rubba wrote: > The reason I make much of my students' disagreements with linguists' analyses is that, at least as I recall from my graduate education, native-speaker intuitions were routinely used in defense of analyses of sentence structure. I agree completely with Lise that introspection alone is not sufficient for defending an analysis. > > I never meant to suggest that my students store complex words disconnected from their component parts. Such a notion would be bizarre for a practitioner of Cognitive Grammar. Plus, the behavior I described doesn't contradict network connections; it challenges the usefulness of NS intuitions, which would, in turn, challenge analyses defended based on same. It's odd that, in some cases, students' tacit knowledge can be brought to consciousness, and in other cases not. > > It seems clear that context (including priming) affects a language-user's analysis. Certainly, many expressions that sound odd out of context sound perfectly fine in a suitable context. > > Jo > > On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: > > > Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG > > ============== > > > Lise Menn wrote: >> So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. >> >> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >> >>> One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. >>> >>> Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. >>> >>> Dr. Johanna Rubba, Ph. D. >>> Professor, Linguistics >>> Linguistics Minor Advisor >>> English Dept. >>> Cal Poly State University San Luis Obispo >>> San Luis Obispo, CA 93407 >>> Ofc. tel. : 805-756-2184 >>> Dept. tel.: 805-756-2596 >>> Dept. fax: 805-756-6374 >>> E-mail: jrubba at calpoly.edu >>> URL: http://cla.calpoly.edu/~jrubba >>> >>> >>> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> http://spot.colorado.edu/~menn/index.html >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> Fellow, Linguistic Society of America >> >> Campus Mail Address: >> UCB 594, Institute for Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> > > Dr. Johanna Rubba, Professor, Linguistics > Linguistics Minor Advisor > English Department > California Polytechnic State University, San Luis Obispo > E-mail: jrubba at calpoly.edu > Tel.: 805.756.2184 > Dept. Ofc. Tel.: 805.756.2596 > Dept. Fax: 805.756.6374 > URL: http://www.cla.calpoly.edu/~jrubba > > > From dan at daneverett.org Thu Sep 9 10:39:42 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 06:39:42 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8891F7.2070608@ling.ucl.ac.uk> Message-ID: Dick, You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. -- Dan > > We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. > > Best wishes, Dick Hudson > From dick at ling.ucl.ac.uk Thu Sep 9 12:16:50 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Thu, 9 Sep 2010 13:16:50 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 09/09/2010 11:39, Daniel Everett wrote: > Dick, > > You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". > > Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. > > It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. > > The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. > > -- Dan > >> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >> >> Best wishes, Dick Hudson >> > > From dan at daneverett.org Thu Sep 9 12:21:30 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 08:21:30 -0400 Subject: analysis: unhappiness In-Reply-To: <4C88D032.1060406@ling.ucl.ac.uk> Message-ID: Dick, I just am not so sure I share the faith in linguistic judgements across the board. Now *that* is something to test experimentally! Dan On 9 Sep 2010, at 08:16, Richard Hudson wrote: > Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. > > But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 09/09/2010 11:39, Daniel Everett wrote: >> Dick, >> >> You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". >> >> Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. >> >> It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. >> >> The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. >> >> -- Dan >> >>> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >>> >>> Best wishes, Dick Hudson >>> >> >> > From dan at daneverett.org Thu Sep 9 12:23:00 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 08:23:00 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: I mean, 'judgements by linguists'. -- Dan On 9 Sep 2010, at 08:21, Daniel Everett wrote: > Dick, > > I just am not so sure I share the faith in linguistic judgements across the board. > > Now *that* is something to test experimentally! > > Dan > > > > On 9 Sep 2010, at 08:16, Richard Hudson wrote: > >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. >>> >>> -- Dan >>> >>>> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> > > From amnfn at well.com Thu Sep 9 12:26:02 2010 From: amnfn at well.com (A. Katz) Date: Thu, 9 Sep 2010 05:26:02 -0700 Subject: analysis: unhappiness In-Reply-To: <4C8891F7.2070608@ling.ucl.ac.uk> Message-ID: Richard, What you just pointed out -- that speaker sensitivity to language structure varies from individual to individual and can be affected by experience and training -- goes toward an even more fundamental point: that language structure exists separate and apart from how individual speakers process it. --Aya On Thu, 9 Sep 2010, Richard Hudson wrote: > This discussion about the role of native-speaker intuition treats all native > speakers and all levels of language equally. And yet we all agree with Joanna > when she says "I can *feel* that the analysis is [[un-happy]-ness], but, > apparently, large numbers of native speakers cannot. " Why? Certainly not > because we've got psycholinguistic evidence on this particular word. Isn't it > something to do with maturity and training for the speakers, and > meaningfulness for the levels? > > Think of that classic 1979 experiment by the Gleitmans that found massive > differences in sensitivity to language structure with both age and education, > so that /eat house bird/ is interpreted as 'a house-bird who is very eat' by > students with a PhD (but not in linguistics) but as 'everybody is eating up > their pet birds' by clerical staff. In this case, as in the other examples > they tried, the more educated were right (by our standards) and the less > educated were wrong. They also found big differences in reliability from > level to level, with semantic judgements easiest and most reliable and > phonological judgements least reliable, and syntax in between. That's > presumably because ordinary speakers spend most of their time grappling with > meaning ('Look after the sense and the sounds will look after themselves', as > someone said to Alice in Wonderland). > > We linguists can add a further layer of explanation to the judgements, but > some judgements do seem to be more reliable than others. And if we have to > wait for psycholinguistic evidence for every detailed analysis we make, our > whole discipline will immediately grind to a halt. Like it or not, native > speaker judgements are what put us linguists ahead of the rest in handling > fine detail. Imagine writing the Cambridge Grammar of the English Language > (or the OED) without using native speaker judgements. > > Best wishes, Dick Hudson > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 09/09/2010 02:20, Johanna Rubba wrote: >> The reason I make much of my students' disagreements with linguists' >> analyses is that, at least as I recall from my graduate education, >> native-speaker intuitions were routinely used in defense of analyses of >> sentence structure. I agree completely with Lise that introspection alone >> is not sufficient for defending an analysis. >> >> I never meant to suggest that my students store complex words disconnected >> from their component parts. Such a notion would be bizarre for a >> practitioner of Cognitive Grammar. Plus, the behavior I described doesn't >> contradict network connections; it challenges the usefulness of NS >> intuitions, which would, in turn, challenge analyses defended based on >> same. It's odd that, in some cases, students' tacit knowledge can be >> brought to consciousness, and in other cases not. >> >> It seems clear that context (including priming) affects a language-user's >> analysis. Certainly, many expressions that sound odd out of context sound >> perfectly fine in a suitable context. >> >> Jo >> >> On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: >> >> >> Right on, Lise. And further, there is a well-known experimental technique >> called "semantic priming" that is admirably well suited for investigating >> whether when a language used hears "unhappiness", "happy" and "happiness" >> are activated ('come to mind'). This technique will probably not answer the >> question of the differential bracketing (un[happiness] vs. [unhappy]ness). >> And it is too rough to answer questions of directionality (does "unhappy" >> prime "happy" stronger than vice versa?). But it does tends to suggest that >> we don't store complex words in total disconnect from their parts, at least >> not as frequent adult users. And that phonological similarity (shared parts >> of words) has semantic consequences. Cheers, TG >> >> ============== >> >> >> Lise Menn wrote: >>> So we see an important phenomenon: Tacit knowledge really IS tacit, and >>> 'intuitions' are very poor guides to what our minds are doing when we are >>> using the patterns of our language as speakers/hearers. Introspection >>> cannot replace observation of actual usage and psycholinguistic >>> experiments; it can only act as a suggestion of where to dig. After all, >>> we can't figure out vision or digestion by thinking about how they feel, >>> although we certainly have to account for subjective feelings of contrast >>> and indigestion. The same is true for language, mutatis mutandis. >>> >>> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >>> >>>> One thing that consistently occurs in my intro linguistics classes is >>>> that at least half of my students do not analyze complex words the way a >>>> linguist would -- many would analyze "unhappiness" as "un" + "happiness." >>>> They make such analyses over and over. It makes one wonder, of course, >>>> about how much native-speaker intuition is in agreement with some >>>> linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], >>>> but, apparently, large numbers of native speakers cannot. >>>> >>>> Another thing I often find is that many students cannot locate either >>>> primary or (especially) secondary stress in words. This is very bizarre, >>>> considering that they produce the stresses correctly and hear them >>>> correctly in others' speech. So many are unsuccessful at this that I have >>>> stopped requiring them to find stress in words on tests. I give them >>>> tricks like singing the word and monitoring for the highest-pitched >>>> syllable, but the tricks don't work. That many students can't be >>>> tone-deaf. >>>> >>>> Dr. Johanna Rubba, Ph. D. >>>> Professor, Linguistics >>>> Linguistics Minor Advisor >>>> English Dept. >>>> Cal Poly State University San Luis Obispo >>>> San Luis Obispo, CA 93407 >>>> Ofc. tel. : 805-756-2184 >>>> Dept. tel.: 805-756-2596 >>>> Dept. fax: 805-756-6374 >>>> E-mail: jrubba at calpoly.edu >>>> URL: http://cla.calpoly.edu/~jrubba >>>> >>>> >>>> >>> Lise Menn Home Office: 303-444-4274 >>> 1625 Mariposa Ave Fax: 303-413-0017 >>> Boulder CO 80302 >>> http://spot.colorado.edu/~menn/index.html >>> Professor Emerita of Linguistics >>> Fellow, Institute of Cognitive Science >>> University of Colorado >>> >>> Secretary, AAAS Section Z [Linguistics] >>> Fellow, Linguistic Society of America >>> >>> Campus Mail Address: >>> UCB 594, Institute for Cognitive Science >>> >>> Campus Physical Address: >>> CINC 234 >>> 1777 Exposition Ave, Boulder >>> >>> >>> >> >> Dr. Johanna Rubba, Professor, Linguistics >> Linguistics Minor Advisor >> English Department >> California Polytechnic State University, San Luis Obispo >> E-mail: jrubba at calpoly.edu >> Tel.: 805.756.2184 >> Dept. Ofc. Tel.: 805.756.2596 >> Dept. Fax: 805.756.6374 >> URL: http://www.cla.calpoly.edu/~jrubba >> >> >> > > From bischoff.st at gmail.com Thu Sep 9 12:53:48 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Thu, 9 Sep 2010 08:53:48 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Thanks to all for the replies. One thing that strikes me as a result of these responses is how my conception of linguistic concepts, objects, and processes, if you will, is so surprisingly skewed and limited. I say surprisingly because while I have been trained as a generativist I have aspired to look at language from various perspectives and find that many of those are more appealing various reasons. Yet, at some basic level, generative linguistics seems to be my point of reference. I recall as a graduate student one instructor noting that "they could no longer think of language in any other way [than generative]"...that when they saw a sentence they "saw trees." Curious... On Wed, Sep 8, 2010 at 4:35 PM, George Lakoff wrote: > Think about "undo" and "untangle." > > In addition, morphology may not reflect semantics. One might have > un+happiness in surface morphology and [un+happy] ness for semantics, which > is perfectly natural. > > George > > > On Wed, Sep 8, 2010 at 5:46 AM, s.t. bischoff wrote: > >> Hi all, >> >> I had an interesting exchange with a few generative >> syntacticians/morphologists (former classmates of mine) regarding an >> analysis of "unhappiness". Two things that they said surprised me a bit, >> they are the following: >> >> (1) un- (negation, 'not') only attaches to adjectives (now this clearly >> isn't the case, a simple cursory view of the etymology in the OED provides >> a >> number of examples of un- with nouns and verbs...though to significantly >> lesser degrees...in addition works on English morphology contain examples >> as >> well) >> >> (2) the analysis of unhappiness can only be [[un-happy]-ness]...an >> analysis >> such as [un-[happy-ness]] is impossible (due to (1) above according to my >> former colleagues). >> >> My questions are the following: >> >> (1) Is there a good/well grounded reason to believe un- "only" attaches >> to >> adjectives? >> >> (2) What would be the consensus on an analysis of "unhappiness" that most >> linguists would agree on? >> >> Thanks, >> Shannon >> > > From dryer at buffalo.edu Thu Sep 9 14:00:47 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Thu, 9 Sep 2010 10:00:47 -0400 Subject: analysis: unhappiness Message-ID: Two comments. First (elaborating perhaps on Dick Hudson's comment), I think there is an important distinction between low-level linguistic intuitions (like whether a word or sentence is well-formed or what it means) and higher-level intuitions (like what the structure of a word or sentence is). One can take the position that we need to account for the former (while recognizing that they are not always reliable) but not the latter. Second, the tension here is not only between evidence from speaker intuitions versus evidence from psycholinguistic experiments. There is also a tension between deciding on the correct analysis on the basis of a priori simplicity arguments versus deciding on the correct analysis on the basis of psycholinguistic evidence (see Derwing 1973). The bracketing paradox that Dan referred to that arises with the word (semantics argues for [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] (the comparative suffix can only be attached to adjectives containing one or two syllables) is only a paradox if one assumes that speakers adopt the simplest analysis. For example, if speakers adopt a more complex rule for either of these (e.g. perhaps the rule for attaching -er can apply exceptionally to trisyllabic words beginning with un-), then the bracketing paradox disappears. Matthew On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > Thanks Dan. I'm sure you're right, and I'd be the first to agree that > conscious judgements are only one kind of evidence that we need to take > into account. I admire Carson Schutze's work (which I reviewed in fact), > and of course I've been aware of complaints about judgements by people > like Labov for decades. > > But you're missing my main point, which is that all judgements aren't > equally reliable. If you want to know how /unhappiness/ is structured, > ask a linguist, not a five-year old. And one of the by-products of > education may be increased sensitivity to syntax - which is one of the > many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > On 09/09/2010 11:39, Daniel Everett wrote: > > Dick, > > > > You raise an important issue here about > methodology. I believe that intuitions are a fine way to generate > hypotheses and even to test them - to a degree. But while it might not have > been feasible for Huddleston, Pullum, and the other contributors to the > Cambridge Grammar to conduct experiments on every point of the grammar, > experiments could have only made the grammar better. The use of intuitions, > corpora, and standard psycholinguistic experimentation (indeed, Standard > Social Science Methodology) is vital for taking the field forward and for > providing the best support for different analyses. Ted Gibson and Ev > Fedorenko have written a very useful new paper on this, showing serious > shortcomings with intuitions as the sole source of evidence, in their > paper: "The need for quantitative methods in syntax and semantics > research".> > > Carson Schutze and Wayne Cowart, among others, > have also written convincingly on this.> > > It is one reason that a team from Stanford, MIT > (Brain and Cognitive Science), and researchers from Brazil are beginning a > third round of experimental work among the Pirahas, since my own work on > the syntax was, like almost every other field researcher's, based on native > speaker intuitions and corpora.> > > The discussion of methodologies reminds me of > the initial reactions to Greenberg's work on classifying the languages of > the Americas. His methods were strongly (and justifiably) criticized. > However, I always thought that his methods were a great way of generating > hypotheses, so long as they were ultimately put to the test of standard > historical linguistics methods. And the same seems true for use of > native-speaker intuitions.> > > -- Dan > > > >> We linguists can add a further layer of > explanation to the judgements, but some judgements do seem to be more > reliable than others. And if we have to wait for psycholinguistic evidence > for every detailed analysis we make, our whole discipline will immediately > grind to a halt. Like it or not, native speaker judgements are what put us > linguists ahead of the rest in handling fine detail. Imagine writing the > Cambridge Grammar of the English Language (or the OED) without using native > speaker judgements.>> > >> Best wishes, Dick Hudson > >> > > > > > > > > > From amnfn at well.com Thu Sep 9 14:27:08 2010 From: amnfn at well.com (A. Katz) Date: Thu, 9 Sep 2010 07:27:08 -0700 Subject: analysis: unhappiness In-Reply-To: <7866.1284040847@buffalo.edu> Message-ID: The use of "unhappier" is idiosyncratic and not universal among English speakers. Rather than assuming that a person who says "unhappier" has a very complicated rule about trisyllabic words beginning in "un", would it not make more sense to see this as an indication that for this particular speaker, the derivation of "unhappy" is opague? That is, the word is treated as an indivisible whole. On the other hand, a speaker who says "more unhappy" is probably aware of the derivation of unhappy and takes it into account when applying the rule for comparatives. Best, --Aya On Thu, 9 Sep 2010, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there is an > important distinction between low-level linguistic intuitions (like whether a > word or sentence is well-formed or what it means) and higher-level intuitions > (like what the structure of a word or sentence is). One can take the position > that we need to account for the former (while recognizing that they are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker intuitions > versus evidence from psycholinguistic experiments. There is also a tension > between deciding on the correct analysis on the basis of a priori simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing paradox that Dan > referred to that arises with the word (semantics argues for > [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] (the > comparative suffix can only be attached to adjectives containing one or two > syllables) is only a paradox if one assumes that speakers adopt the simplest > analysis. For example, if speakers adopt a more complex rule for either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to take >> into account. I admire Carson Schutze's work (which I reviewed in fact), >> and of course I've been aware of complaints about judgements by people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable. If you want to know how /unhappiness/ is structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might not have >> been feasible for Huddleston, Pullum, and the other contributors to the >> Cambridge Grammar to conduct experiments on every point of the grammar, >> experiments could have only made the grammar better. The use of intuitions, >> corpora, and standard psycholinguistic experimentation (indeed, Standard >> Social Science Methodology) is vital for taking the field forward and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are beginning a >> third round of experimental work among the Pirahas, since my own work on >> the syntax was, like almost every other field researcher's, based on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of generating >> hypotheses, so long as they were ultimately put to the test of standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic evidence >> for every detailed analysis we make, our whole discipline will immediately >> grind to a halt. Like it or not, native speaker judgements are what put us >> linguists ahead of the rest in handling fine detail. Imagine writing the >> Cambridge Grammar of the English Language (or the OED) without using native >> speaker judgements.>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> >> >> >> >> > > From dryer at buffalo.edu Thu Sep 9 14:42:16 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Thu, 9 Sep 2010 10:42:16 -0400 Subject: analysis: unhappiness Message-ID: Aya proposes a possible alternative hypothesis, but my point is that whether this alternative is correct or not is not a question of whether it "makes more sense", but an empirical question that can really only be decided on the basis of psycholinguistic evidence. Matthew On Thu 09/09/10 10:27 AM , "A. Katz" amnfn at well.com sent: > The use of "unhappier" is idiosyncratic and not universal among > English speakers. Rather than assuming that a person who says "unhappier" > has a very complicated rule about trisyllabic words beginning in "un", > would it not make more sense to see this as an indication that for this particular > speaker, the derivation of "unhappy" is opague? That is, the word > is treated as an indivisible whole. > > On the other hand, a speaker who says "more unhappy" is probably > aware of the derivation of unhappy and takes it into account when applying the rule > for comparatives. > > Best, > > --Aya > > > On Thu, 9 Sep 2010, Matthew S. Dryer wrote: > > > > > Two comments. > > > > First (elaborating perhaps on Dick Hudson's > comment), I think there is an> important distinction between low-level > linguistic intuitions (like whether a> word or sentence is well-formed or what it > means) and higher-level intuitions> (like what the structure of a word or sentence > is). One can take the position> that we need to account for the former (while > recognizing that they are not> always reliable) but not the latter. > > > > Second, the tension here is not only between > evidence from speaker intuitions> versus evidence from psycholinguistic > experiments. There is also a tension> between deciding on the correct analysis on the > basis of a priori simplicity> arguments versus deciding on the correct > analysis on the basis of> psycholinguistic evidence (see Derwing 1973). > The bracketing paradox that Dan> referred to that arises with the word > (semantics argues for> [[un+happi] + er], morphology and phonology > argues for [un + [happi+er]] (the> comparative suffix can only be attached to > adjectives containing one or two> syllables) is only a paradox if one assumes that > speakers adopt the simplest> analysis. For example, if speakers adopt a more > complex rule for either of these> (e.g. perhaps the rule for attaching -er can > apply exceptionally to trisyllabic> words beginning with un-), then the bracketing > paradox disappears.> > > Matthew > > > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling > .ucl.ac.uk sent:>> Thanks Dan. I'm sure you're right, and I'd > be the first to agree that>> conscious judgements are only one kind of > evidence that we need to take>> into account. I admire Carson Schutze's work > (which I reviewed in fact),>> and of course I've been aware of complaints > about judgements by people>> like Labov for decades. > >> > >> But you're missing my main point, which is > that all judgements aren't>> equally reliable. If you want to know how > /unhappiness/ is structured,>> ask a linguist, not a five-year old. And one > of the by-products of>> education may be increased sensitivity to > syntax - which is one of the>> many reasons why linguists need to pay more > attention to education.>> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>> On 09/09/2010 11:39, Daniel Everett > wrote:>>> Dick, > >>> > >>> You raise an important issue here > about>> methodology. I believe that intuitions are a > fine way to generate>> hypotheses and even to test them - to a > degree. But while it might not have>> been feasible for Huddleston, Pullum, and > the other contributors to the>> Cambridge Grammar to conduct experiments on > every point of the grammar,>> experiments could have only made the grammar > better. The use of intuitions,>> corpora, and standard psycholinguistic > experimentation (indeed, Standard>> Social Science Methodology) is vital for > taking the field forward and for>> providing the best support for different > analyses. Ted Gibson and Ev>> Fedorenko have written a very useful new > paper on this, showing serious>> shortcomings with intuitions as the sole > source of evidence, in their>> paper: "The need for quantitative > methods in syntax and semantics>> research".> > >>> Carson Schutze and Wayne Cowart, among > others,>> have also written convincingly on > this.>>>> It is one reason that a team from > Stanford, MIT>> (Brain and Cognitive Science), and > researchers from Brazil are beginning a>> third round of experimental work among the > Pirahas, since my own work on>> the syntax was, like almost every other > field researcher's, based on native>> speaker intuitions and > corpora.>>>> The discussion of methodologies reminds > me of>> the initial reactions to Greenberg's work on > classifying the languages of>> the Americas. His methods were strongly (and > justifiably) criticized.>> However, I always thought that his methods > were a great way of generating>> hypotheses, so long as they were ultimately > put to the test of standard>> historical linguistics methods. And the same > seems true for use of>> native-speaker intuitions.> > >>> -- Dan > >>> > >>>> We linguists can add a further layer > of>> explanation to the judgements, but some > judgements do seem to be more>> reliable than others. And if we have to wait > for psycholinguistic evidence>> for every detailed analysis we make, our > whole discipline will immediately>> grind to a halt. Like it or not, native > speaker judgements are what put us>> linguists ahead of the rest in handling fine > detail. Imagine writing the>> Cambridge Grammar of the English Language > (or the OED) without using native>> speaker judgements.>> > >>>> Best wishes, Dick Hudson > >>>> > >>> > >>> > >> > >> > >> > >> > >> > > > > > > > > > From khildeb at siue.edu Thu Sep 9 14:58:22 2010 From: khildeb at siue.edu (Kristine Hildebrandt) Date: Thu, 9 Sep 2010 09:58:22 -0500 Subject: Call For Papers: Special Issue of Himalayan Linguistics Message-ID: Himalayan Linguistics 10.1 (June 2011) Guest Editors: Yogendra Yadava, Karen Grunow-Harsta, Kristine Hildebrandt, Stephen Watters Himalayan Linguistics, a free peer-reviewed web journal and archive devoted to the study of the languages of the Himalayas, is now accepting submissions to a special issue in memory of our late colleagues, HL Associate Editors Michael (Mickey) Noonan and David Watters. Articles on all languages of the Himalayan region are welcome, as are those that significantly draw on work by Noonan or Watters on Himalayan languages. Deadline for submissions: 15 October 2010 Address inquiries to Guest Editor Yogendra Yadava (ypyadava at gmail.com) or HL Editor Carol Genetti (cgenetti at linguistics.ucsb.edu) http://www.linguistics.ucsb.edu/HimalayanLinguistics/ -- Kristine A. Hildebrandt Department of English Language & Literature Southern Illinois University Edwardsville Edwardsville, IL 62026 U.S.A. 618-650-3380 (office) khildeb at siue.edu http://www.siue.edu/~khildeb From Jean-Christophe.Verstraete at arts.kuleuven.be Thu Sep 9 18:54:23 2010 From: Jean-Christophe.Verstraete at arts.kuleuven.be (Jean-Christophe Verstraete) Date: Thu, 9 Sep 2010 20:54:23 +0200 Subject: First call: Greenberg Award 2011 Message-ID: The Joseph Greenberg Award 2011 The Association for Linguistic Typology's Joseph Greenberg Award recognizes and honours the best piece of typological research embodied in a doctoral dissertation or equivalent in 2009-2010. Theses are eligible if they were accepted by a university between 1 January 2009 and 31 December 2010. The award will consist of payment of travel, per diem expenses and registration fee to attend the ALT IX Conference, to be held in Hong Kong, July 21-24, 2011, and to present a synopsis or element of the prize-winning work as a plenary lecture at that meeting. The Joseph Greenberg Award was named to remember Joseph Greenberg's (1915-2001) fundamental contributions to typology and the interest he showed in encouraging young researchers. Between 1998 and 2006, it was known as the "ALT Junior Award". To be eligible, those submitting their dissertation must be members of the Association for Linguistic Typology (ALT). They are asked to submit their manuscript by email in pdf format, with all non-standard fonts in Unicode, to the chair of the jury, to arrive no later than January 31st 2011. If this proves technically difficult, the candidate is asked to discuss the problem with the chair. A jury, consisting of about ten ALT members, will be appointed by ALT's president, appropriate to the work submitted. The chair will be: Isabelle Bril Laboratoire LACITO 7, rue Guy Môquet (bât. D) 94801 Villejuif Cedex France From sclancy at uchicago.edu Thu Sep 9 19:10:15 2010 From: sclancy at uchicago.edu (Steven Clancy) Date: Thu, 9 Sep 2010 14:10:15 -0500 Subject: SCLC-2010 Conference Program In-Reply-To: <4C827393.1060608@sheffield.ac.uk> Message-ID: THE TENTH ANNUAL CONFERENCE OF THE SLAVIC COGNITIVE LINGUISTICS ASSOCIATION (SCLC-2010) October 9-11, 2010 Brown University (Providence, Rhode Island, USA) Hosted by the Department of Slavic Languages, Brown University (Providence, Rhode Island) with support from the Office of International Affairs, the Colver Lectureship Fund from the Office of the Dean of the Faculty, the Center for Language Studies, the Department of Cognitive, Linguistics and Psychologial Sciences, and John Benjamins Publishing Company. THE TENTH ANNUAL CONFERENCE OF THE SLAVIC COGNITIVE LINGUISTICS ASSOCIATION (SCLC-2010) October 9-11, 2010 The Slavic Cognitive Linguistics Association (SCLA) announces the program for the 2010 annual conference. The conference will be held on the campus of Brown University (Providence, Rhode Island) on Saturday, October 9 through Monday, October 11, 2010. SCLC-2010 Keynote Speakers Cognitively Plausible Parsing Eugene Charniak Brown University, USA The role of rational discounting: what to say, what not to say Adele E. Goldberg Princeton University, USA Access, Activation, and Overlap: Focusing on the Differential Ronald W. Langacker University of California, San Diego, USA MAIN SESSIONS (Saturday, Sunday, and Monday) Each presentation for the main sessions will be given 20 minutes and will be followed by a 10-minute discussion period. SCHEDULE View the Conference Schedule in PDF Format. CONFERENCE INFORMATION Membership in the SCLA is free, please see "Joining SCLA" for more information. Attendance at SCLC-2010 is open and all interested individuals are encouraged to attend. Information on Conference Fees and Registration are provided below. Brown University is located in Providence, Rhode Island and is accessible from Boston Logan International Airport (BOS, 55 miles away) or T.F. Green Airport (PVD) in Providence. Transportation from Airports: If you arrive at Boston Logan International Airport (BOS), you can get to Providence by bus (Peter Pan Bus, currently $36 round trip) from Boston Logan to Kennedy Plaza (in downtown Providence, tell the driver you are dropping off at Kennedy Plaza when you board), then by taxi to the Radisson Hotel in Providence (1.6 miles to the hotel) or by subway to South Station and by bus or train to Providence, then by taxi to the Radisson Hotel in Providence. Taxi from Logan to Providence would cost more than $100 and is not recommended. Ground transportation from Boston: further information. If you arrive directly at Providence T.F. Green International Airport (PVD) To and from PVD to downtown area: further information. Taxi from Providence Airport to Providence will cost around $30 one way. Ground transportation from Providence is much simpler. But international travelers may find more flights flying into Boston. Conference Hotel: Radisson Hotel Providence Harbor 220 India Street, Providence, RI 02903, Reservations: 1-800-395-7046 US/Canada Toll-free Telephone: (401) 272-5577 Fax: (401) 272-0251 Email: RHI_PROV at radisson.com Participants should mention the Slavic Cognitive Linguistics Association Conference to reserve a room at the special conference rate ($79USD plus 13% tax). Special rates are valid from Thursday 10/7 to Monday 10/11 (fewer rooms are available on Thursday and Monday). Additional people per room is $10 per person. Reservations at the conference rate must be made by 5:00pm local time, September 8, 2010. Please make your reservations early to ensure a room at the conference hotel. Conference Fees: Registration Fee: Regular participants $60USD Graduate student participants $40USD Conference dinner: $50USD Please make your checks payable to "Brown University" and send them to: Masako Fidler SCLC Organizer 20 Manning Walk, Box E Department of Slavic Languages Brown University Providence, RI02912 If you wish to preregister by wiring money for conference fees, please contact Gisela Belton for details. If you will need to register on-site, please contactSteven Clancy to confirm your participation. We hope you will be able to join us for SCLC-2010. Please forward this call for papers to your colleagues and graduate students who may be interested in presenting or attending. Sincerely, Steven Clancy President, SCLA Tore Nesset Vice-President, SCLA Masako Fidler Conference Organizer and Host Brown University on behalf of the SCLA officers and the 2010 SCLA organizing committee From jaejung.song at stonebow.otago.ac.nz Thu Sep 9 22:46:38 2010 From: jaejung.song at stonebow.otago.ac.nz (Jae Jung Song) Date: Fri, 10 Sep 2010 10:46:38 +1200 Subject: Oxford Handbook of Linguistic Typology Message-ID: Dear Funknetters, May I draw your attention to the imminent publication of the Oxford Handbook of Linguistic Typology? The table of contents appears below. If you would like (to ask your university/institution library) to pre-order a copy, you can go to: http://www.oup.com/us/catalog/general/subject/Linguistics/SyntaxMorphology/?view=usa&ci=9780199281251 With best wishes, Jae Jung Song ----------------------------------------------------------------- The Oxford Handbook of Linguistic Typology Edited by Jae Jung Song November 2010, 776 pages, 978-0-19-928125-1, hardback, £ 85.00 Oxford University Press Part I. Setting the Stage - Jae Jung Song Part II. Foundations: History, Theory and Method 1. The (Early) History of Linguistic Typology - Paolo Ramat 2. The Pioneers of Linguistic Typology: From Gabelentz to Greenberg - Giorgio Graffi 3. Linguistic Typology and the Study of Language - Michael Daniel 4. Explaining Language Universals - Edith Moravcsik 5. The Problem of Cross-Linguistic Identification - Leon Stassen 6. Language Sampling - Dik Bakker Part III. Theoretical Dimensions of Linguistic Typology 7. Markedness: Iconicity, Economy and Frequency - Joan Bybee 8. Competing Motivations - John Haiman 9. Categories and Prototypes - Johan van der Auwera and Volker Gast 10. Implicational Hierarchies - Greville Corbett 11. Processing Efficiency and Complexity in Typological Patterns - John Hawkins 12. Language Universals and Linguistic Knowledge - Sonia Cristofaro Part IV. Empirical Dimensions of Linguistic Typology 13. Word Order Typology - Jae Jung Song 14. Word Classes - Walter Bisang 15. Case-Marking Typology - Beatrice Primus 16. Person Marking - Anna Siewierska 17. Transitivity Typology - Seppo Kittilä 18. Voice Typology - Leonid Kulikov 19. Grammatical Relation Typology - Balthasar Bickel 20.Typology of Tense, Aspect and Modality Systems - Ferdinand de Haan 21. Syntactic Typology - Lindsay Whaley 22. Morphological Typology - Dunstan Brown 23. Semantic Typology - Nicholas Evans 24. Typology of Phonological Systems - Ian Maddieson Part V. Linguistic Typology in a Wider Context 25. Linguistic Typology and Historical Linguistics - Kenneth Shields 26. Linguistic Typology and Language Contact - Maria Koptjevskaja-Tamm 27. Linguistic Typology and First Language Acquisition - Melissa Bowerman 28. Linguistic Typology and Second Language Acquisition - Fred Eckman 29. Linguistic Typology and Language Documentation - Patience Epps 30. Linguistic Typology and Formal Grammar - Maria Polinsky References Author Index Language Index Subject Index From Lise.Menn at Colorado.EDU Thu Sep 9 23:26:13 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Thu, 9 Sep 2010 17:26:13 -0600 Subject: analysis: unhappiness In-Reply-To: <7866.1284040847@buffalo.edu> Message-ID: I wish we had better terminology for keeping track of whether, at a given time, we are talking about the patterns that are 'out there' in the language and might possibly be apprehended (subconsciously) by a speaker, and when we are talking about the patterns that a particular speaker actually does apprehend, as indicated by experiments, from simple 'wug tests' up to brain wave and eye-gaze studies. And for distinguishing among the degrees of pattern apprehension that a person may have, from vague preferences detectable in reaction times or other behavior all the way up through clear metalinguistic insights. Dick Hudson's note reminding us of the Gleitman and Gleitman study is right on target. Since we don't in fact have such an agreed-on terminology, we have to be quite careful in making clear what we are referring to when we talk about 'the correct analysis' of a form like 'unhappiness'. We know, but tend to forget - and tend to forget to tell our students! - that it's an empirical question as to whether the formal simplicity and coherence of description of forms 'out there' (e.g. lovely abstract morphophonemics) is any kind of approximation to the way knowledge of the same forms is organized in a particular person's head. If we remember that a very large proportion of what we know about our language is 'out there' when we are infants and has to be internalized through experience with the language (even if you believe in innate 'core language'), the variation in internal knowledge from one person to another is more understandable. We especially need to consider (and try to test) the possibility that since the brain can make multiple cross-connections, multiple patterns are involved simultaneously in morphological and syntactic analyses. I suggest that that's the case with 'unhappiness' - and the linguistic analyses that I know about are not good at handling that kind of idea. Lise On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there > is an > important distinction between low-level linguistic intuitions (like > whether a > word or sentence is well-formed or what it means) and higher-level > intuitions > (like what the structure of a word or sentence is). One can take > the position > that we need to account for the former (while recognizing that they > are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker > intuitions > versus evidence from psycholinguistic experiments. There is also a > tension > between deciding on the correct analysis on the basis of a priori > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing > paradox that Dan > referred to that arises with the word (semantics argues > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one > or two > syllables) is only a paradox if one assumes that speakers adopt the > simplest > analysis. For example, if speakers adopt a more complex rule for > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to >> take >> into account. I admire Carson Schutze's work (which I reviewed in >> fact), >> and of course I've been aware of complaints about judgements by >> people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable. If you want to know how /unhappiness/ is >> structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of >> the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might >> not have >> been feasible for Huddleston, Pullum, and the other contributors to >> the >> Cambridge Grammar to conduct experiments on every point of the >> grammar, >> experiments could have only made the grammar better. The use of >> intuitions, >> corpora, and standard psycholinguistic experimentation (indeed, >> Standard >> Social Science Methodology) is vital for taking the field forward >> and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing >> serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are >> beginning a >> third round of experimental work among the Pirahas, since my own >> work on >> the syntax was, like almost every other field researcher's, based >> on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the >> languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of >> generating >> hypotheses, so long as they were ultimately put to the test of >> standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic >> evidence >> for every detailed analysis we make, our whole discipline will >> immediately >> grind to a halt. Like it or not, native speaker judgements are what >> put us >> linguists ahead of the rest in handling fine detail. Imagine >> writing the >> Cambridge Grammar of the English Language (or the OED) without >> using native >> speaker judgements.>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> >> >> >> >> > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From edith at uwm.edu Thu Sep 9 23:55:17 2010 From: edith at uwm.edu (Edith A Moravcsik) Date: Thu, 9 Sep 2010 18:55:17 -0500 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: It seems to me that it is important to construct grammars that are based simply on the form-meaning correspondences of the language 'out there', as Lise said, apart from psycholinguistic data, and that are constrained by the usual requirements of scientific descriptions, such as generality and simplicity. Such grammars may not be psycholinguistically real at all nor do they try to be; but they provide a baseline in reference to which we can then assess people's actual ways of learning, processing, and analyzing structures. In the absence of such basic descriptions, it would be hard to know what psycholinguistic data we should be "surprised at" - i.e., what it is that needs to be explained about acquisition, processing, and people's intuitive analyses of structures. Edith Moravcsik From: "Lise Menn" To: dryer at buffalo.edu, "Funknet" Cc: "Richard Hudson" Sent: Thursday, September 9, 2010 6:26:13 PM Subject: Re: [FUNKNET] analysis: unhappiness I wish we had better terminology for keeping track of whether, at a   given time, we are talking about the patterns that are 'out there' in   the language and might possibly be apprehended (subconsciously) by a   speaker, and when we are talking about the patterns that a particular   speaker actually does apprehend, as indicated by experiments, from   simple 'wug tests' up to brain wave and eye-gaze studies.  And for   distinguishing among the degrees of pattern apprehension that a person   may have, from vague preferences detectable in reaction times or other   behavior all the way up through clear metalinguistic insights.  Dick   Hudson's note reminding us of the Gleitman and Gleitman study is right   on target.         Since we don't in fact have such an agreed-on terminology, we have to   be quite careful in making clear what we are referring to when we talk   about 'the correct analysis' of a form like 'unhappiness'.  We know,   but tend to forget - and tend to forget to tell our students! - that   it's an empirical question as to whether the formal simplicity and   coherence of description of forms 'out there' (e.g. lovely abstract   morphophonemics) is any kind of approximation to the way knowledge of   the same forms is organized in a particular person's head.  If we   remember that a very large proportion of what we know about our   language is 'out there' when we are infants and has to be internalized   through experience with the language (even if you believe in innate   'core language'), the variation in internal knowledge from one person   to another is more understandable.                  We especially need to consider (and try to test) the possibility that   since the brain can make multiple cross-connections, multiple patterns are   involved simultaneously in morphological and syntactic analyses. I suggest that   that's the case with 'unhappiness' - and the linguistic analyses that   I know about are not good at handling that kind of idea.         Lise On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there   > is an > important distinction between low-level linguistic intuitions (like   > whether a > word or sentence is well-formed or what it means) and higher-level   > intuitions > (like what the structure of a word or sentence is).  One can take   > the position > that we need to account for the former (while recognizing that they   > are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker   > intuitions > versus evidence from psycholinguistic experiments.  There is also a   > tension > between deciding on the correct analysis on the basis of a priori   > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973).  The bracketing   > paradox that Dan > referred to that arises with the word (semantics argues   > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one   > or two > syllables) is only a paradox if one assumes that speakers adopt the   > simplest > analysis.  For example, if speakers adopt a more complex rule for   > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to   > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10  8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to   >> take >> into account. I admire Carson Schutze's work (which I reviewed in   >> fact), >> and of course I've been aware of complaints about judgements by   >> people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable.  If you want to know how /unhappiness/ is   >> structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of   >> the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes,  Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might   >> not have >> been feasible for Huddleston, Pullum, and the other contributors to   >> the >> Cambridge Grammar to conduct experiments on every point of the   >> grammar, >> experiments could have only made the grammar better. The use of   >> intuitions, >> corpora, and standard psycholinguistic experimentation (indeed,   >> Standard >> Social Science Methodology)  is vital for taking the field forward   >> and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing   >> serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are   >> beginning a >> third round of experimental work among the Pirahas, since my own   >> work on >> the syntax was, like almost every other field researcher's, based   >> on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the   >> languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of   >> generating >> hypotheses, so long as they were ultimately put to the test of   >> standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic   >> evidence >> for every detailed analysis we make, our whole discipline will   >> immediately >> grind to a halt. Like it or not, native speaker judgements are what   >> put us >> linguists ahead of the rest in handling fine detail. Imagine   >> writing the >> Cambridge Grammar of the English Language (or the OED) without   >> using native >> speaker judgements.>> >>>> Best wishes,  Dick Hudson >>>> >>> >>> >> >> >> >> >> > Lise Menn                      Home Office: 303-444-4274 1625 Mariposa Ave       Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of  Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder -- Professor Emerita of Linguistics Department of Linguistics University of Wisconsin-Milwaukee Milwaukee, WI 53201-0413 USA From dcyr at yorku.ca Fri Sep 10 02:38:59 2010 From: dcyr at yorku.ca (Danielle E. Cyr) Date: Thu, 9 Sep 2010 22:38:59 -0400 Subject: analysis: unhappiness In-Reply-To: <960039811.375031.1284076517806.JavaMail.root@mail03.pantherlink.uwm.edu> Message-ID: We might also want to take into account 1) the historical time span extension of different people's grammatical knowledge and 2) the fact that grammaticalization occurs progressively over generations so that not all people of a same generation have the same grammatical analysis in mind when they think about morphological analysis. By 1) I mean that the more one knows about the history/evolution of a language's morphology, spread over the longer a period of time, the finest morphological analysis one will be able to make. For instance, when I present my students with the French adverb MAINTENANT 'now' at first they can't parse it into separate morphemes. After some minutes and a little bit of coaxing they come too see MAIN+TEMANT 'in hand' + 'holding', and finally MAIN+TEN+ANT 'in hand'+'hold'+ '-ing'. From that moment on, these students will integrate that precise morphological parsing. It will become part of their "internal historical grammar" and they will not really be able to go back to the "feeling" that MAINTENANT is only one morpheme. Now when I present them with the adverb AUJOURD'HUI 'today' and we go back in time to AD+ILLU(M)+DIURN(UM)+DE+HO(C)+DIE(M), the evolution of which spans over more than 2000 years, it totally transforms their inner historical grammar. They become different speakers of French from who they were before. By 2) I mean what is made totally explicit in Hopper and Traugott Grammaticalization (1993) and Marchello-Nizia (2006) Grammaticalisation et changement linguistique, Bruxelles, De Boeck). Speakers of a same language community do not all have the exact same grammar in mind simply because language change occurs constantly and progressively among different social groups, classes and generations. So it will always be impossible to get everyone to produce morphological parsing in the same exact way. Language, and grammar with it, are realities in constant flux. Conceiving that there is a unique stable grammar "out there" is "une vue de l'esprit". It helps us to think about the flux with a sensation of being on solid ground. Perhaps just like believing there is a God out there helps some of us to cope with impermanence. Cordialement, Danielle Cyr From haspelmath at eva.mpg.de Fri Sep 10 11:18:34 2010 From: haspelmath at eva.mpg.de (Martin Haspelmath) Date: Fri, 10 Sep 2010 13:18:34 +0200 Subject: analysis: phenomenological vs. cognitive In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. I suggested two terms for these kinds of patterns in a 2004 paper (reference below): – phenomenological description (for what is "out there", or "grammars that are based simply on the form-meaning correspondences", in Edith Moravcsik's terms) – cognitive description (for what a speaker has in their head) In the paper my main claim is that we don't really need cognitive description in order to explain the patterns of languages in functional terms (just as Darwin didn't need full descriptions of genomes to come up with functional explanations of the phenomenological properties of species). In generative linguistics, of course, one needs a unique "analysis" for each structure, because explanation and description/analysis are the same enterprise (just two different aspects of it), whereas in functional linguistics, description and explanation are separate. Thus, the fact that the best cognitive description is somewhat elusive doesn't matter to this approach. Martin Reference: Haspelmath, Martin. 2004. Does linguistic explanation presuppose linguistic description? Studies in Language 28. 554-579. (doi:10.1075/sl.28.3.06has) -- Martin Haspelmath (haspelmath at eva.mpg.de) Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 From A.Foolen at let.ru.nl Fri Sep 10 11:21:05 2010 From: A.Foolen at let.ru.nl (Ad Foolen) Date: Fri, 10 Sep 2010 13:21:05 +0200 Subject: analysis: unhappiness In-Reply-To: <1284086339.4c899a434ddd8@mymail.yorku.ca> Message-ID: When the historical dimension is taken into consideration, as Danielle Cyr proposes, it might be worthwhile to re-read Saussure's Cours, in particular his methodological reflections in the Appendices A, B, and C, following part 3 (in my 1968 Payot edition page 251-260). Saussure distinguishes between 'analyse subjective' and 'analyse objective'. The first is made by the native speaker, the second one by the linguist ('le grammairien'). The subjective analysis is the one that plays a role in language change via 'formations analogiques'. Analyses that come to the surface in psycholinguistic experiments as discussed in the present exchange did not play a role in Saussure's methodology. Two examples from Appendix A (Cours, p. 251): "L'analyse objective voit quatre sous-unités dans amabas (am-a-ba-s); les Latins coupaient ama-ba-s; il est même probable qu'ils regardaient -bas comme un tout flexionell opposé au radical. Dans les mots français 'entier' (lat. in-teger 'intact'), 'enfant' (lat. infans 'qui ne parle pas') 'enceinte' (lat. in-cincta 'sans ceinture'), l'historien dégagera un préfixe commun 'en-', identique au 'in-'privatif du latin; l'analyse subjective des sujets parlants l'ignore totalement." In Saussure's view, both perspectives have their value, but "en dernier resort celle [l'analyse] des sujets importe seule, car elle est fondé directement sur les faits de langue." (Cours, p. 252). Ad Foolen -------------------------------------------------- From: "Danielle E. Cyr" Sent: Friday, September 10, 2010 4:38 AM To: "Edith A Moravcsik" Cc: "Lise Menn" ; "Richard Hudson" ; ; "Funknet" Subject: Re: [FUNKNET] analysis: unhappiness > We might also want to take into account 1) the historical time span > extension of > different people's grammatical knowledge and 2) the fact that > grammaticalization > occurs progressively over generations so that not all people of a same > generation have the same grammatical analysis in mind when they think > about > morphological analysis. > > By 1) I mean that the more one knows about the history/evolution of a > language's > morphology, spread over the longer a period of time, the finest > morphological > analysis one will be able to make. For instance, when I present my > students > with the French adverb MAINTENANT 'now' at first they can't parse it into > separate morphemes. After some minutes and a little bit of coaxing they > come > too see > MAIN+TEMANT 'in hand' + 'holding', and finally MAIN+TEN+ANT 'in > hand'+'hold'+ > '-ing'. From that moment on, these students will integrate that precise > morphological parsing. It will become part of their "internal historical > grammar" and they will not really be able to go back to the "feeling" that > MAINTENANT is only one morpheme. > > Now when I present them with the adverb AUJOURD'HUI 'today' and we go back > in > time to AD+ILLU(M)+DIURN(UM)+DE+HO(C)+DIE(M), the evolution of which spans > over > more than 2000 years, it totally transforms their inner historical > grammar. They > become different speakers of French from who they were before. > > By 2) I mean what is made totally explicit in Hopper and Traugott > Grammaticalization (1993) and Marchello-Nizia (2006) Grammaticalisation et > changement linguistique, Bruxelles, De Boeck). Speakers of a same language > community do not all have the exact same grammar in mind simply because > language change occurs constantly and progressively among different social > groups, classes and generations. So it will always be impossible to get > everyone to produce morphological parsing in the same exact way. > > Language, and grammar with it, are realities in constant flux. Conceiving > that > there is a unique stable grammar "out there" is "une vue de l'esprit". It > helps > us to think about the flux with a sensation of being on solid ground. > Perhaps > just like believing there is a God out there helps some of us to cope with > impermanence. > > Cordialement, > Danielle Cyr > From egibson at MIT.EDU Fri Sep 10 13:03:20 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Fri, 10 Sep 2010 09:03:20 -0400 Subject: analysis: unhappiness Message-ID: Dear Dan, Dick: I would like to clarify some points that Dan Everett makes, in response to Dick Hudson. Ev Fedorenko and I have written a couple of papers recently (Gibson & Fedorenko, 2010, in press, see references and links below) on what we think are weak methodological standards in syntax and semantics research over the past many years. The issue that we address is the prevalent method in syntax and semantics research, which involves obtaining a judgment of the acceptability of a sentence / meaning pair, typically by just the author of the paper, sometimes with feedback from colleagues. As we address in our papers, this methodology does not allow proper testing of scientific hypotheses because of (a) the small number of experimental participants (typically one); (b) the small number of experimental stimuli (typically one); (c) cognitive biases on the part of the researcher and participants; and (d) the effect of the preceding context (e.g., other constructions the researcher may have been recently considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and several others cited in Gibson & Fedorenko, in press; for similar points, but with not as strong a conclusion as ours). Three issues need to be separated here: (1) the use of intuitive judgments as a dependent measure in a language experiment; (2) potential cognitive biases on the part of experimental subjects and experimenters in language experiments; and (3) the need for obtaining quantitative evidence, whatever the dependent measure might be. The paper that Ev and I wrote addresses the last two issues, but does not go into depth on the first issue (the use of intuitions as a dependent measure in language experiments). Regarding this issue, we don't think that there is anything wrong with gathering intuitive judgments as a dependent measure, as long as the task is clear to the experimental participants. In the longer paper (Gibson & Fedorenko, in press) we respond to some arguments that have been given in support of continuing to use the traditional non-quantitative method in syntax / semantics research. One recent defense of the traditional method comes from Phillips (2008), who argues that no harm has come from the non-quantitative approach in syntax research thus far. Phillips argues that there are no cases in the literature where an incorrect intuitive judgment has become the basis for a widely accepted generalization or an important theoretical claim. He therefore concludes that there is no reason to adopt more rigorous data collection standards. We challenge Philips’ conclusion by presenting three cases from the literature where a faulty intuition has led to incorrect generalizations and mistaken theorizing, plausibly due to cognitive biases on the part of the researchers. A second argument that is sometimes presented for the continued use of the traditional non-quantitative method is that it would be too inefficient to evaluate every syntactic / semantic hypothesis or phenomenon quantitatively. For example, Culicover & Jackendoff (2010) make this argument explicitly in their response to Gibson & Fedorenko (2010): “It would cripple linguistic investigation if it were required that all judgments of ambiguity and grammaticality be subject to statistically rigorous experiments on naive subjects, especially when investigating languages whose speakers are hard to access” (Culicover & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point earlier in the discussion here.) Whereas we agree that in circumstances where gathering data is difficult, some evidence is better than no evidence, we do not agree that research would be slowed with respect to languages where experimental participants are easy to access, such as English. In contrast, we think that the opposite is true: the field’s progress is probably slowed by not doing quantitative research. Suppose that a typical syntax / semantics paper that lacks quantitative evidence includes judgments for 50 or more sentences / meaning pairs, corresponding to 50 or more empirical claims. Even if most of the judgments from such a paper are correct or are on the right track, the problem is in knowing which judgments are correct. For example, suppose that 90% of the judgments from an arbitrary paper are correct (which is probably a high estimate). (Colin Phillips and some of his former students / postdocs have commented to us that, in their experience, quantitative acceptability judgment studies almost always validate the claim(s) in the literature. This is not our experience, however. Most experiments that we have run which attempt to test some syntactic / semantic hypothesis in the literature end up providing us with a pattern of data that had not been known before the experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in press; Patel et al., 2009; Scontras & Gibson, submitted).) This means that in a paper with 50 empirical claims 45/50 are correct. But which 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over two million different theories. By quantitatively evaluating the empirical claims, we reduce the uncertainty a great deal. To make progress, it is better to have theoretical claims supported by solid quantitative evidence, so that even if the interpretation of the data changes over time as new evidence becomes available – as is often the case in any field of science – the empirical pattern can be used as a basis for further theorizing. Furthermore, it is no longer expensive to run behavioral experiments, at least in English and other widely spoken languages. There now exists a marketplace interface – Amazon.com’s Mechanical Turk – which can be used for collecting behavioral data over the internet quickly and inexpensively. The cost of using an interface like this is minimal, and the time that it takes for the results to be returned is short. For example, currently on Mechanical Turk, a survey of approximately 50 items will be answered by 50 or more participants within a couple of hours, at a cost of approximately $1 per participant. Thus a survey can be completed within a day, at a cost of less than $50. (The hard work of designing the experiment, and constructing controlled materials remains of course.) Sorry to be so verbose. But I think that these methodological points are very important. Best wishes, Ted Gibson Gibson, E. & Fedorenko, E. (In press). The need for quantitative methods in syntax and semantics research. Language and Cognitive Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko InPress LCP.pdf Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in linguistics research. Trends in Cognitive Science, 14, 233-234. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko 2010 TICS.pdf > Dick, > > You raise an important issue here about methodology. I believe that > intuitions are a fine way to generate hypotheses and even to test > them - to a degree. But while it might not have been feasible for > Huddleston, Pullum, and the other contributors to the Cambridge > Grammar to conduct experiments on every point of the grammar, > experiments could have only made the grammar better. The use of > intuitions, corpora, and standard psycholinguistic experimentation > (indeed, Standard Social Science Methodology) is vital for taking > the field forward and for providing the best support for different > analyses. Ted Gibson and Ev Fedorenko have written a very useful new > paper on this, showing serious shortcomings with intuitions as the > sole source of evidence, in their paper: "The need for quantitative > methods in syntax and semantics research". > > Carson Schutze and Wayne Cowart, among others, have also written > convincingly on this. > > It is one reason that a team from Stanford, MIT (Brain and Cognitive > Science), and researchers from Brazil are beginning a third round of > experimental work among the Pirahas, since my own work on the syntax > was, like almost every other field researcher's, based on native > speaker intuitions and corpora. > > The discussion of methodologies reminds me of the initial reactions > to Greenberg's work on classifying the languages of the Americas. > His methods were strongly (and justifiably) criticized. However, I > always thought that his methods were a great way of generating > hypotheses, so long as they were ultimately put to the test of > standard historical linguistics methods. And the same seems true for > use of native-speaker intuitions. > > -- Dan >> We linguists can add a further layer of explanation to the >> judgements, but some judgements do seem to be more reliable than >> others. And if we have to wait for psycholinguistic evidence for >> every detailed analysis we make, our whole discipline will >> immediately grind to a halt. Like it or not, native speaker >> judgements are what put us linguists ahead of the rest in handling >> fine detail. Imagine writing the Cambridge Grammar of the English >> Language (or the OED) without using native speaker judgements. >> >> Best wishes, Dick Hudson From amnfn at well.com Fri Sep 10 13:12:03 2010 From: amnfn at well.com (A. Katz) Date: Fri, 10 Sep 2010 06:12:03 -0700 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: Lise, I like very much the terminology that you used here: "what is out there in a language" versus the way it is "apprehended by a speaker." The problem is that too many of our colleagues don't agree that there is such a distinction-- or that language could ever possibly be out there! They think language exists only in the mind of speakers, and they look for uniformity of processing where none exists. That language is an abstraction and that each speaker has to crack the code individually and alone is not universally accepted among linguists. Best, --Aya On Thu, 9 Sep 2010, Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a given > time, we are talking about the patterns that are 'out there' in the language > and might possibly be apprehended (subconsciously) by a speaker, and when we > are talking about the patterns that a particular speaker actually does > apprehend, as indicated by experiments, from simple 'wug tests' up to brain > wave and eye-gaze studies. And for distinguishing among the degrees of > pattern apprehension that a person may have, from vague preferences > detectable in reaction times or other behavior all the way up through clear > metalinguistic insights. Dick Hudson's note reminding us of the Gleitman and > Gleitman study is right on target. > > Since we don't in fact have such an agreed-on terminology, we have to > be quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'. We know, but tend to > forget - and tend to forget to tell our students! - that it's an empirical > question as to whether the formal simplicity and coherence of description of > forms 'out there' (e.g. lovely abstract morphophonemics) is any kind of > approximation to the way knowledge of the same forms is organized in a > particular person's head. If we remember that a very large proportion of > what we know about our language is 'out there' when we are infants and has to > be internalized through experience with the language (even if you believe in > innate 'core language'), the variation in internal knowledge from one person > to another is more understandable. > We especially need to consider (and try to test) the > possibility that since > the brain can make multiple cross-connections, multiple patterns are involved > simultaneously in morphological and syntactic analyses. I suggest that that's > the case with 'unhappiness' - and the linguistic analyses that I know about > are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> >> Two comments. >> >> First (elaborating perhaps on Dick Hudson's comment), I think there is an >> important distinction between low-level linguistic intuitions (like whether >> a >> word or sentence is well-formed or what it means) and higher-level >> intuitions >> (like what the structure of a word or sentence is). One can take the >> position >> that we need to account for the former (while recognizing that they are not >> always reliable) but not the latter. >> >> Second, the tension here is not only between evidence from speaker >> intuitions >> versus evidence from psycholinguistic experiments. There is also a tension >> between deciding on the correct analysis on the basis of a priori >> simplicity >> arguments versus deciding on the correct analysis on the basis of >> psycholinguistic evidence (see Derwing 1973). The bracketing paradox that >> Dan >> referred to that arises with the word (semantics argues for >> [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] >> (the >> comparative suffix can only be attached to adjectives containing one or two >> syllables) is only a paradox if one assumes that speakers adopt the >> simplest >> analysis. For example, if speakers adopt a more complex rule for either of >> these >> (e.g. perhaps the rule for attaching -er can apply exceptionally to >> trisyllabic >> words beginning with un-), then the bracketing paradox disappears. >> >> Matthew >> >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>> conscious judgements are only one kind of evidence that we need to take >>> into account. I admire Carson Schutze's work (which I reviewed in fact), >>> and of course I've been aware of complaints about judgements by people >>> like Labov for decades. >>> >>> But you're missing my main point, which is that all judgements aren't >>> equally reliable. If you want to know how /unhappiness/ is structured, >>> ask a linguist, not a five-year old. And one of the by-products of >>> education may be increased sensitivity to syntax - which is one of the >>> many reasons why linguists need to pay more attention to education. >>> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> On 09/09/2010 11:39, Daniel Everett wrote: >>>> Dick, >>>> >>>> You raise an important issue here about >>> methodology. I believe that intuitions are a fine way to generate >>> hypotheses and even to test them - to a degree. But while it might not >>> have >>> been feasible for Huddleston, Pullum, and the other contributors to the >>> Cambridge Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, >>> corpora, and standard psycholinguistic experimentation (indeed, Standard >>> Social Science Methodology) is vital for taking the field forward and for >>> providing the best support for different analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new paper on this, showing serious >>> shortcomings with intuitions as the sole source of evidence, in their >>> paper: "The need for quantitative methods in syntax and semantics >>> research".> >>>> Carson Schutze and Wayne Cowart, among others, >>> have also written convincingly on this.> >>>> It is one reason that a team from Stanford, MIT >>> (Brain and Cognitive Science), and researchers from Brazil are beginning a >>> third round of experimental work among the Pirahas, since my own work on >>> the syntax was, like almost every other field researcher's, based on >>> native >>> speaker intuitions and corpora.> >>>> The discussion of methodologies reminds me of >>> the initial reactions to Greenberg's work on classifying the languages of >>> the Americas. His methods were strongly (and justifiably) criticized. >>> However, I always thought that his methods were a great way of generating >>> hypotheses, so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions.> >>>> -- Dan >>>> >>>>> We linguists can add a further layer of >>> explanation to the judgements, but some judgements do seem to be more >>> reliable than others. And if we have to wait for psycholinguistic evidence >>> for every detailed analysis we make, our whole discipline will immediately >>> grind to a halt. Like it or not, native speaker judgements are what put us >>> linguists ahead of the rest in handling fine detail. Imagine writing the >>> Cambridge Grammar of the English Language (or the OED) without using >>> native >>> speaker judgements.>> >>>>> Best wishes, Dick Hudson >>>>> >>>> >>>> >>> >>> >>> >>> >>> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From tpayne at uoregon.edu Fri Sep 10 14:16:56 2010 From: tpayne at uoregon.edu (Thomas E. Payne) Date: Fri, 10 Sep 2010 07:16:56 -0700 Subject: "Relative clauses" with no relativized element Message-ID: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne From eitkonen at utu.fi Fri Sep 10 14:01:38 2010 From: eitkonen at utu.fi (Esa Itkonen) Date: Fri, 10 Sep 2010 17:01:38 +0300 Subject: analysis: phenomenological vs. cognitive In-Reply-To: <4C8A140A.8000804@eva.mpg.de> Message-ID: Dear Friends: These questions were given what I consider the definitive answer in the context of the 'psychological reality' debate in the mid and late 70's. But there is no harm in reinventing (or renaming) the wheel. Esa Homepage: http://users.utu.fi/eitkonen ----- Original Message ----- From: Martin Haspelmath Date: Friday, September 10, 2010 2:19 pm Subject: Re: [FUNKNET] analysis: phenomenological vs. cognitive To: Funknet > Lise Menn wrote: > > I wish we had better terminology for keeping track of whether, at a > > > given time, we are talking about the patterns that are 'out there' > in > > the language and might possibly be apprehended (subconsciously) by > a > > speaker, and when we are talking about the patterns that a > particular > > speaker actually does apprehend, as indicated by experiments, from > > > simple 'wug tests' up to brain wave and eye-gaze studies. > I suggested two terms for these kinds of patterns in a 2004 paper > (reference below): > > – phenomenological description (for what is "out there", or "grammars > > that are based simply on the form-meaning correspondences", in Edith > > Moravcsik's terms) > > – cognitive description (for what a speaker has in their head) > > In the paper my main claim is that we don't really need cognitive > description in order to explain the patterns of languages in > functional > terms (just as Darwin didn't need full descriptions of genomes to > come > up with functional explanations of the phenomenological properties of > > species). > > In generative linguistics, of course, one needs a unique "analysis" > for > each structure, because explanation and description/analysis are the > > same enterprise (just two different aspects of it), whereas in > functional linguistics, description and explanation are separate. > Thus, > the fact that the best cognitive description is somewhat elusive > doesn't > matter to this approach. > > Martin > > > Reference: > Haspelmath, Martin. 2004. Does linguistic explanation presuppose > linguistic description? Studies in Language 28. 554-579. > (doi:10.1075/sl.28.3.06has) > > -- > Martin Haspelmath (haspelmath at eva.mpg.de) > Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz > 6 > D-04103 Leipzig > Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 > > > > > > From jscheibm at odu.edu Fri Sep 10 15:15:22 2010 From: jscheibm at odu.edu (Scheibman, Joanne) Date: Fri, 10 Sep 2010 11:15:22 -0400 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Hi Tom, Quirk et al. analyze these as appositive clauses (in 17.26 and elsewhere in that chapter). The head of the NP is a general abstract noun, e.g. "fact", "idea", "claim", "belief", (or "condition"), and "that" has no role in the subordinate clause, as you mentioned. Joanne on 9/10/10 10:16 AM, Thomas E. Payne wrote: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne -- BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach CanIt if this mail (ID 358169235) is spam: Spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=s Not spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=n Forget vote: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=f ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS From rchen at csusb.edu Fri Sep 10 15:42:43 2010 From: rchen at csusb.edu (Rong Chen) Date: Fri, 10 Sep 2010 23:42:43 +0800 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: To add to Joanne's comments: There are basically three ways to distinguish an appositive clause (AC) from a relative clause (RC). 1) An AC can only be led by *that* while an RC can be led by other pronouns. 2) The AC and the noun it modifies display an equative relationship--one can say X (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't (except, perhaps, when the relative clause is sentential). 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun is always part of the clause in an RC. Rong Chen -----Original Message----- From: funknet-bounces at mailman.rice.edu [mailto:funknet-bounces at mailman.rice.edu] On Behalf Of Scheibman, Joanne Sent: Friday, September 10, 2010 11:15 PM To: Thomas E. Payne; FUNKNET Subject: Re: [FUNKNET] "Relative clauses" with no relativized element Hi Tom, Quirk et al. analyze these as appositive clauses (in 17.26 and elsewhere in that chapter). The head of the NP is a general abstract noun, e.g. "fact", "idea", "claim", "belief", (or "condition"), and "that" has no role in the subordinate clause, as you mentioned. Joanne on 9/10/10 10:16 AM, Thomas E. Payne wrote: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne -- BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach CanIt if this mail (ID 358169235) is spam: Spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=s Not spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=n Forget vote: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=f ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS From carol.moder at okstate.edu Fri Sep 10 15:52:37 2010 From: carol.moder at okstate.edu (Moder, Carol) Date: Fri, 10 Sep 2010 10:52:37 -0500 Subject: Job advertisement Message-ID: Dear Funknetters: Please distribute the following position description to interested colleagues and doctoral students completing their degrees. Our linguistics program has a cognitive-functional orientation and we seek a new colleague with this perspective. Carol Moder Professor & Head Department of English 205 Morrill Hall Stillwater, OK 74078 (405)744-6140 Assistant Professor -Linguistics with specialization in phonetics or phonology Department of English Oklahoma State University Tenure-track position. Ph.D. in Linguistics, or related area with a specialization in phonetics or phonology. Ability to teach a broad range of undergraduate and graduate linguistics courses in support of specializations in linguistics and Teaching English as a Second Language. 3-2 teaching load beginning August 2011. Strong research agenda and demonstrated excellence in teaching required. Salary competitive and commensurate with experience. OSU offers the BA, MA, and the PhD in English with an M.A. option in Teaching English as a Second Language and M.A. and Ph.D. specializations in Linguistics.. For further information on the department see our webpage at http://english.okstate.edu. To ensure full consideration, applications must be received by November 3, 2010 However, we will continue to accept and consider applications until the position has been filled. Send letter of application, cv, writing sample, and dossier, including three letters of recommendation and transcript to Carol Moder, Head, English Department, Oklahoma State University, 205 Morrill Hall, Stillwater, OK 74078-4069. Filling of this position is contingent upon availability of funding. Oklahoma State University is an AA/EEO/E-Verify employer committed to diversity. OSU-Stillwater is a tobacco-free campus. From Arie.Verhagen at hum.LeidenUniv.nl Fri Sep 10 16:21:10 2010 From: Arie.Verhagen at hum.LeidenUniv.nl (Arie Verhagen) Date: Fri, 10 Sep 2010 18:21:10 +0200 Subject: "Relative clauses" with no relativized ele Message-ID: And as another addition: the clauses that can only be introduced by *that* (with no role to play in the subordinate clause) may be seen as (subtypes of) complement clauses, expressing a proposition with the noun functioning as Complement Taking Predicate (CTP), expressing a propostional attitude, epistemic/evaluative stance, etc. (following analyses by Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. constructions like "The claim is that X" (traditionally analysed as subject clauses), "I claim that X", "I put forward the claim that X", in which the relationship between the verb or noun and the that-clause is comparable to the one in "The claim that X". --Arie Verhagen ---------------- Message from Rong Chen 10 Sep 2010, 23:42 Subject: Re: [FUNKNET] "Relative clauses" with no relativi > To add to Joanne's comments: > > There are basically three ways to distinguish an appositive clause > (AC) from a relative clause (RC). > > 1) An AC can only be led by *that* while an RC can be led by other > pronouns. > > 2) The AC and the noun it modifies display an equative relationship--one can say X > (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't > (except, perhaps, when the relative clause is sentential). > 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun > is always part of the clause in an RC. > > Rong Chen > From shinja_hwang at gial.edu Fri Sep 10 17:03:47 2010 From: shinja_hwang at gial.edu (Hwang, Shinja) Date: Fri, 10 Sep 2010 17:03:47 +0000 Subject: "Relative clauses" with no relativized ele In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: I also prefer to refer to them as complement clauses. They differ from the relative clauses since the embedded clause is complete in itself, not lacking anything. These clauses are in an appositive relation with the head noun like 'fact, claim, evidence' and in English they can be linked by 'that', but not by 'which', unlike the relative clause. Compare the examples below: The evidence [that(*which) he stole the book] is clearly shown to us. The evidence [that/which they used 0 in the trial] was very powerful. In some languages there is no syntactic difference (such as 'that' vs. 'which' in English), and the criteria may need to resort to some other clues such as whether there is a missing element in the embedded clause which is the head noun. Shin Ja Hwang -----Original Message----- From: funknet-bounces at mailman.rice.edu [mailto:funknet-bounces at mailman.rice.edu] On Behalf Of Arie Verhagen Sent: Friday, September 10, 2010 11:21 AM To: 'FUNKNET' Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele And as another addition: the clauses that can only be introduced by *that* (with no role to play in the subordinate clause) may be seen as (subtypes of) complement clauses, expressing a proposition with the noun functioning as Complement Taking Predicate (CTP), expressing a propostional attitude, epistemic/evaluative stance, etc. (following analyses by Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. constructions like "The claim is that X" (traditionally analysed as subject clauses), "I claim that X", "I put forward the claim that X", in which the relationship between the verb or noun and the that-clause is comparable to the one in "The claim that X". --Arie Verhagen ---------------- Message from Rong Chen 10 Sep 2010, 23:42 Subject: Re: [FUNKNET] "Relative clauses" with no relativi > To add to Joanne's comments: > > There are basically three ways to distinguish an appositive clause > (AC) from a relative clause (RC). > > 1) An AC can only be led by *that* while an RC can be led by other > pronouns. > > 2) The AC and the noun it modifies display an equative relationship--one can say X > (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't > (except, perhaps, when the relative clause is sentential). > 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun > is always part of the clause in an RC. > > Rong Chen > From kemmer at rice.edu Fri Sep 10 17:16:09 2010 From: kemmer at rice.edu (Suzanne Kemmer) Date: Fri, 10 Sep 2010 12:16:09 -0500 Subject: "Relative clauses" with no relativized element In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: In Generative Syntax these clauses were viewed as complement clauses with an NP head, distinct from relative clauses but having some parallels with them. I think it was Joan Bresnan that brought out the parallels and distinctions, maybe in her doctoral dissertation . As I recall (but my remembrance may be faulty), Bresnan named the THAT element a COMP for complementizer. The term 'appositive' isn't very good because in traditional grammar that is reserved for an 'UNrestrictive' relation of a noun and its complement--an incidental description of a head N's referent rather than a specification of which referent ("the tree, a live oak, survived another 100 years or so"). In Cognitive Grammar nouns like claim, statement, idea, realization, belief etc. are in almost all cases nominalizations of 'viewing predicates' (verbs like claim, believe, etc.) that introduce on-stage predications 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the viewing predicates are space builders in Fauconnier's mental spaces terminology) For the nominalizations of these predicates, the semantics of the nouns intrinsically has an "e-site" or elaboration site that allows for spelling out the content of the viewed predicate in the form of a complement clause. The e-site inherent to the semantics of the nouns is parallel to the e-site inherent to the semantics of the corresponding verbs. There are a few cases I can think of of nouns that have 'viewing predicate' e-sites but don't have corresponding verbs . For example the noun _view_ "The view that global climate change is anthropogenic is widely held by scientists" ( ' X views that (proposition)' is not possible, only 'X views Y as ...' , with a restriction to equative or descriptive propositions). Also _idea_----the verb has to be changed to something like 'believe' to make a corresponding full predicate. I view (!) these nouns as semantically parallel in interesting ways to picture nouns. The conceptualizer (viewer) in both cases can designate the noun in a possessive phrase, but after that the syntax diverges. --Suzanne On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > And as another addition: the clauses that can only be introduced by > *that* (with no role to > play in the subordinate clause) may be seen as (subtypes of) > complement clauses, > expressing a proposition with the noun functioning as Complement > Taking Predicate (CTP), > expressing a propostional attitude, epistemic/evaluative stance, > etc. (following analyses by > Thompson, Diessel, Langacker, myself, and others), i.e. not > relatives. Cf. constructions like > "The claim is that X" (traditionally analysed as subject clauses), > "I claim that X", "I put forward > the claim that X", in which the relationship between the verb or > noun and the that-clause is > comparable to the one in "The claim that X". > > --Arie Verhagen > > ---------------- > Message from Rong Chen > 10 Sep 2010, 23:42 > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > >> To add to Joanne's comments: >> >> There are basically three ways to distinguish an appositive clause >> (AC) from a relative clause (RC). >> >> 1) An AC can only be led by *that* while an RC can be led by other >> pronouns. >> >> 2) The AC and the noun it modifies display an equative >> relationship--one can say X >> (denoted by the noun) is Y (presented by the appositive)--while an >> RC often doesn't >> (except, perhaps, when the relative clause is sentential). > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but >> a relative pronoun >> is always part of the clause in an RC. >> >> Rong Chen >> > > From kuzar at research.haifa.ac.il Fri Sep 10 17:42:24 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Fri, 10 Sep 2010 20:42:24 +0300 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: A thorough discussion of the head nouns and their relation with their complement clauses may be found in Hans-Joerg Schmid's book on Shell Nouns (this is his term for the head nouns). Ron Kuzar --------- On Fri, Sep 10, 2010 at 8:16 PM, Suzanne Kemmer wrote: > In Generative Syntax these clauses were viewed as complement clauses with > an NP head, distinct > from relative clauses but having some parallels with them. I think it was > Joan Bresnan that > brought out the parallels and distinctions, maybe in her doctoral > dissertation . As I recall (but > my remembrance may be faulty), Bresnan named the > THAT element a COMP for complementizer. > > The term 'appositive' isn't very good because in traditional grammar > that is reserved for an 'UNrestrictive' relation of a noun and its > complement--an incidental description of > a head N's referent rather than a specification of which referent ("the > tree, a live oak, survived another 100 years or so"). > > In Cognitive Grammar nouns like claim, statement, idea, realization, > belief etc. are in almost all cases nominalizations of 'viewing predicates' > (verbs like claim, believe, etc.) that introduce on-stage predications > 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the > viewing > predicates are space builders in Fauconnier's mental spaces terminology) > > For the nominalizations of these predicates, the semantics of the nouns > intrinsically has an "e-site" or elaboration site > that allows for spelling out the content of the viewed predicate in the > form of a complement clause. The e-site > inherent to the semantics of the nouns is parallel to the e-site inherent > to the semantics of the corresponding verbs. > > There are a few cases I can think of of nouns that have 'viewing > predicate' e-sites but don't have corresponding verbs . > For example > the noun _view_ "The view that global climate change is anthropogenic is > widely held by scientists" > ( ' X views that (proposition)' is not possible, only 'X views Y as ...' , > with a restriction to equative or descriptive propositions). > Also _idea_----the verb has to be changed to something like 'believe' to > make a corresponding full predicate. > > I view (!) these nouns as semantically parallel in interesting ways to > picture nouns. The conceptualizer (viewer) in > both cases can designate the noun in a possessive phrase, but after that > the syntax diverges. > > --Suzanne > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > And as another addition: the clauses that can only be introduced by *that* >> (with no role to >> play in the subordinate clause) may be seen as (subtypes of) complement >> clauses, >> expressing a proposition with the noun functioning as Complement Taking >> Predicate (CTP), >> expressing a propostional attitude, epistemic/evaluative stance, etc. >> (following analyses by >> Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. >> constructions like >> "The claim is that X" (traditionally analysed as subject clauses), "I >> claim that X", "I put forward >> the claim that X", in which the relationship between the verb or noun and >> the that-clause is >> comparable to the one in "The claim that X". >> >> --Arie Verhagen >> >> ---------------- >> Message from Rong Chen >> 10 Sep 2010, 23:42 >> >> Subject: Re: [FUNKNET] "Relative clauses" with no relativi >> >> To add to Joanne's comments: >>> >>> There are basically three ways to distinguish an appositive clause >>> (AC) from a relative clause (RC). >>> >>> 1) An AC can only be led by *that* while an RC can be led by other >>> pronouns. >>> >>> 2) The AC and the noun it modifies display an equative relationship--one >>> can say X >>> (denoted by the noun) is Y (presented by the appositive)--while an RC >>> often doesn't >>> (except, perhaps, when the relative clause is sentential). >>> >> >> 3)--which Tom noted--*that* is not part of the clause in an AC; but a >>> relative pronoun >>> is always part of the clause in an RC. >>> >>> Rong Chen >>> >>> >> >> > -- =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From tgivon at uoregon.edu Fri Sep 10 17:43:34 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Fri, 10 Sep 2010 11:43:34 -0600 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this construction described and analyzed as the product of nominalization of clauses with verbs that take verbal complements ('know', 'think', 'say' etc. The preceding section (6.6.3. "Complex NP's arising through nominalization", p. 287) deals more generally with nominalizations. The term "noun complements" was used in syntax classes at UCLA in the mid 1960s, so certainly Joan Bresnan did not invent it. Best, TG =========== Thomas E. Payne wrote: > Can anyone help me name the following structure in English, and maybe point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > > From giuliana.fiorentino at unimol.it Fri Sep 10 17:53:15 2010 From: giuliana.fiorentino at unimol.it (Giuliana Fiorentino) Date: Fri, 10 Sep 2010 19:53:15 +0200 Subject: "Relative clauses" with no relativized element Message-ID: Hi Tom, clauses like: The importance of being Earnest the fact of being late the fact that you are late the idea that world is round etcetera are not relative clauses but can be considered among syntactic strategies in order to nominalise events after a generic noun (working as a classifier for nominalised events). Giuliana ----- Original Message ----- From: Thomas E. Payne To: FUNKNET Sent: Friday, September 10, 2010 4:16 PM Subject: [FUNKNET] "Relative clauses" with no relativized element Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne From eitan.eg at gmail.com Fri Sep 10 17:54:06 2010 From: eitan.eg at gmail.com (E.G.) Date: Fri, 10 Sep 2010 20:54:06 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: Hi all, I'd agree with Arie Verhagen. But there's a way that cross-linguistic comparison can help what seems to be a purely theoretical question based on a single language. The problem here is that English uses the same element to mark regular relatives and these "appositional" relatives. But if at least one language encodes them by different means, then there's at least a good case for seeing them as distinct functions. It's basically the same principle that's used to decide whether to put a meaning on a semantic map. So here are two languages that I know that encode them differently. In Modern Hebrew, these clauses can be encoded as a dedicated complement clause (ki), which differs from the relative clause marker (Se-), e.g. ha-hoda'a Se-kibalnu the-announcment rel-we_got "The announcement that we got." ha-hoda'a ki hitbatel ha-mifgaS the-message CMP was_cancelled the-meeting "The announcement that the meeting was cancelled." In Coptic, these clauses are marked by ce-, which marks complement clauses, *inter alia*, but not relative clauses: ph-mewi ce- (complement clause) 'the-thought that (we are angry)' ph-mewi ete- (relative clause) 'the thought that (we used to think)' This seems to be a pretty clear indication that these are complement clauses rather than relatives. Even if one doesn't like the notion of nouns taking complement clauses (and why not? nominalizations in some languages can take accusative modifiers as well as genitives), it still probably isn't incidental that the nominalizations are of verbs that take complement clauses when finite. As usual, the perspective in Talmy Givón's *Syntax* (vol. 2) is worth looking at. Best, Eitan On 10 September 2010 19:21, Arie Verhagen wrote: > And as another addition: the clauses that can only be introduced by *that* > (with no role to > play in the subordinate clause) may be seen as (subtypes of) complement > clauses, > expressing a proposition with the noun functioning as Complement Taking > Predicate (CTP), > expressing a propostional attitude, epistemic/evaluative stance, etc. > (following analyses by > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. > constructions like > "The claim is that X" (traditionally analysed as subject clauses), "I claim > that X", "I put forward > the claim that X", in which the relationship between the verb or noun and > the that-clause is > comparable to the one in "The claim that X". > > --Arie Verhagen > > ---------------- > Message from Rong Chen > 10 Sep 2010, 23:42 > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > To add to Joanne's comments: > > > > There are basically three ways to distinguish an appositive clause > > (AC) from a relative clause (RC). > > > > 1) An AC can only be led by *that* while an RC can be led by other > > pronouns. > > > > 2) The AC and the noun it modifies display an equative relationship--one > can say X > > (denoted by the noun) is Y (presented by the appositive)--while an RC > often doesn't > > (except, perhaps, when the relative clause is sentential). > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > relative pronoun > > is always part of the clause in an RC. > > > > Rong Chen > > > > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From kemmer at rice.edu Fri Sep 10 17:55:30 2010 From: kemmer at rice.edu (Suzanne Kemmer) Date: Fri, 10 Sep 2010 12:55:30 -0500 Subject: "Relative clauses" with no relativized element In-Reply-To: <4C8A6E46.3040301@uoregon.edu> Message-ID: Talmy, yes Joan B did not invent noun complements nor the term for them; but my recollection, in passing, was that she named 'that' in such structures as "complementizer" (and I also recall that she referred to 'that' relativizers with the same term, while recognizing other differences between the two structures.) I may be wrong on that, but it's a different recollection, claim, or whatever, than the one you refer to. Not all the head nouns are nominalizations, but most are. S. On Sep 10, 2010, at 12:43 PM, Tom Givon wrote: > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this construction described and analyzed as the product of nominalization of clauses with verbs that take verbal complements ('know', 'think', 'say' etc. The preceding section (6.6.3. "Complex NP's arising through nominalization", p. 287) deals more generally with nominalizations. The term "noun complements" was used in syntax classes at UCLA in the mid 1960s, so certainly Joan Bresnan did not invent it. Best, TG > > =========== > > Thomas E. Payne wrote: >> Can anyone help me name the following structure in English, and maybe point >> me to some references? I do not find reference to this in the Cambridge >> Grammar of the English Language or any other of my English grammar books. >> But then, maybe I just don't know where to look. >> >> Here are two examples from a play: >> >> His protestations of devotion in the trial scene are, in our opinion, >> genuine, as is his confession [that his affair with the Countess is >> platonic]. >> >> The bracketed clause seems to modify "confession", though there is no >> position for a confession in the clause itself. >> >> . . . forced hither with an impious black design [to have my innocence and >> youth become the sacrifice of brutal violence]. >> >> Here the bracketed non-finite clause seems to modify "design." >> >> These are not all that rare. I'm reminded of examples like: >> >> "The claim [that my client is a murderer] is totally false." >> >> Are these relative clauses? If so what kind? Thanks for any help. >> >> Tom Payne >> >> > > From eitan.eg at gmail.com Fri Sep 10 17:56:23 2010 From: eitan.eg at gmail.com (E.G.) Date: Fri, 10 Sep 2010 20:56:23 +0300 Subject: "Relative clauses" with no relativized element In-Reply-To: <036386D3933D4CB99389FFB2576ADBE8@giuliana> Message-ID: Jespersen and his nexus-substantives should be mentioned (Philosophy of Grammar, 1924). Also in his MEG and Analytic Syntax one could find interesting discussions. Eitan On 10 September 2010 20:53, Giuliana Fiorentino < giuliana.fiorentino at unimol.it> wrote: > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From dick at ling.ucl.ac.uk Fri Sep 10 17:59:12 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Fri, 10 Sep 2010 18:59:12 +0100 Subject: analysis: unhappiness In-Reply-To: <676152F6-8654-4BCD-BC30-B39E807C07B7@mit.edu> Message-ID: Dear Ted, Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 14:03, Ted Gibson wrote: > Dear Dan, Dick: > > I would like to clarify some points that Dan Everett makes, in > response to Dick Hudson. > > Ev Fedorenko and I have written a couple of papers recently (Gibson & > Fedorenko, 2010, in press, see references and links below) on what we > think are weak methodological standards in syntax and semantics > research over the past many years. The issue that we address is the > prevalent method in syntax and semantics research, which involves > obtaining a judgment of the acceptability of a sentence / meaning > pair, typically by just the author of the paper, sometimes with > feedback from colleagues. As we address in our papers, this > methodology does not allow proper testing of scientific hypotheses > because of (a) the small number of experimental participants > (typically one); (b) the small number of experimental stimuli > (typically one); (c) cognitive biases on the part of the researcher > and participants; and (d) the effect of the preceding context (e.g., > other constructions the researcher may have been recently > considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > several others cited in Gibson & Fedorenko, in press; for similar > points, but with not as strong a conclusion as ours). > > Three issues need to be separated here: (1) the use of intuitive > judgments as a dependent measure in a language experiment; (2) > potential cognitive biases on the part of experimental subjects and > experimenters in language experiments; and (3) the need for obtaining > quantitative evidence, whatever the dependent measure might be. The > paper that Ev and I wrote addresses the last two issues, but does not > go into depth on the first issue (the use of intuitions as a dependent > measure in language experiments). Regarding this issue, we don't think > that there is anything wrong with gathering intuitive judgments as a > dependent measure, as long as the task is clear to the experimental > participants. > > In the longer paper (Gibson & Fedorenko, in press) we respond to some > arguments that have been given in support of continuing to use the > traditional non-quantitative method in syntax / semantics research. > One recent defense of the traditional method comes from Phillips > (2008), who argues that no harm has come from the non-quantitative > approach in syntax research thus far. Phillips argues that there are > no cases in the literature where an incorrect intuitive judgment has > become the basis for a widely accepted generalization or an important > theoretical claim. He therefore concludes that there is no reason to > adopt more rigorous data collection standards. We challenge Philips’ > conclusion by presenting three cases from the literature where a > faulty intuition has led to incorrect generalizations and mistaken > theorizing, plausibly due to cognitive biases on the part of the > researchers. > > A second argument that is sometimes presented for the continued use of > the traditional non-quantitative method is that it would be too > inefficient to evaluate every syntactic / semantic hypothesis or > phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > make this argument explicitly in their response to Gibson & Fedorenko > (2010): “It would cripple linguistic investigation if it were required > that all judgments of ambiguity and grammaticality be subject to > statistically rigorous experiments on naive subjects, especially when > investigating languages whose speakers are hard to access” (Culicover > & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > earlier in the discussion here.) Whereas we agree that in > circumstances where gathering data is difficult, some evidence is > better than no evidence, we do not agree that research would be slowed > with respect to languages where experimental participants are easy to > access, such as English. In contrast, we think that the opposite is > true: the field’s progress is probably slowed by not doing > quantitative research. > Suppose that a typical syntax / semantics paper that lacks > quantitative evidence includes judgments for 50 or more sentences / > meaning pairs, corresponding to 50 or more empirical claims. Even if > most of the judgments from such a paper are correct or are on the > right track, the problem is in knowing which judgments are correct. > For example, suppose that 90% of the judgments from an arbitrary paper > are correct (which is probably a high estimate). (Colin Phillips and > some of his former students / postdocs have commented to us that, in > their experience, quantitative acceptability judgment studies almost > always validate the claim(s) in the literature. This is not our > experience, however. Most experiments that we have run which attempt > to test some syntactic / semantic hypothesis in the literature end up > providing us with a pattern of data that had not been known before the > experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > that in a paper with 50 empirical claims 45/50 are correct. But which > 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over > two million different theories. By quantitatively evaluating the > empirical claims, we reduce the uncertainty a great deal. To make > progress, it is better to have theoretical claims supported by solid > quantitative evidence, so that even if the interpretation of the data > changes over time as new evidence becomes available – as is often the > case in any field of science – the empirical pattern can be used as a > basis for further theorizing. > > Furthermore, it is no longer expensive to run behavioral experiments, > at least in English and other widely spoken languages. There now > exists a marketplace interface – Amazon.com’s Mechanical Turk – which > can be used for collecting behavioral data over the internet quickly > and inexpensively. The cost of using an interface like this is > minimal, and the time that it takes for the results to be returned is > short. For example, currently on Mechanical Turk, a survey of > approximately 50 items will be answered by 50 or more participants > within a couple of hours, at a cost of approximately $1 per > participant. Thus a survey can be completed within a day, at a cost of > less than $50. (The hard work of designing the experiment, and > constructing controlled materials remains of course.) > > Sorry to be so verbose. But I think that these methodological points > are very important. > > Best wishes, > > Ted Gibson > > Gibson, E. & Fedorenko, E. (In press). The need for quantitative > methods in syntax and semantics research. Language and Cognitive > Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > & Fedorenko InPress LCP.pdf > > Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > linguistics research. Trends in Cognitive Science, 14, 233-234. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > 2010 TICS.pdf > > > > >> Dick, >> >> You raise an important issue here about methodology. I believe that >> intuitions are a fine way to generate hypotheses and even to test >> them - to a degree. But while it might not have been feasible for >> Huddleston, Pullum, and the other contributors to the Cambridge >> Grammar to conduct experiments on every point of the grammar, >> experiments could have only made the grammar better. The use of >> intuitions, corpora, and standard psycholinguistic experimentation >> (indeed, Standard Social Science Methodology) is vital for taking the >> field forward and for providing the best support for different >> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >> paper on this, showing serious shortcomings with intuitions as the >> sole source of evidence, in their paper: "The need for quantitative >> methods in syntax and semantics research". >> >> Carson Schutze and Wayne Cowart, among others, have also written >> convincingly on this. >> >> It is one reason that a team from Stanford, MIT (Brain and Cognitive >> Science), and researchers from Brazil are beginning a third round of >> experimental work among the Pirahas, since my own work on the syntax >> was, like almost every other field researcher's, based on native >> speaker intuitions and corpora. >> >> The discussion of methodologies reminds me of the initial reactions >> to Greenberg's work on classifying the languages of the Americas. His >> methods were strongly (and justifiably) criticized. However, I always >> thought that his methods were a great way of generating hypotheses, >> so long as they were ultimately put to the test of standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions. >> >> -- Dan > > > >>> We linguists can add a further layer of explanation to the >>> judgements, but some judgements do seem to be more reliable than >>> others. And if we have to wait for psycholinguistic evidence for >>> every detailed analysis we make, our whole discipline will >>> immediately grind to a halt. Like it or not, native speaker >>> judgements are what put us linguists ahead of the rest in handling >>> fine detail. Imagine writing the Cambridge Grammar of the English >>> Language (or the OED) without using native speaker judgements. >>> >>> Best wishes, Dick Hudson > > > From egibson at MIT.EDU Fri Sep 10 18:30:16 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Fri, 10 Sep 2010 14:30:16 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8A71F0.4050507@ling.ucl.ac.uk> Message-ID: Dear Dick: Perhaps we are talking at cross purposes. I don't understand what is confusing about what Ev Fedorenko and I are claiming. All we are saying is that if you have some testable claim involving a general hypothesis about a language, then you need to get quantitative data from unbiased sources to evaluate that claim. If you are interested in English past tense morphology, then depending on the claims that you might want to investigate, there are lots of ways to get relevant quantitative evidence. Corpus data will probably be useful. For very low frequency words, you can run experiments to test behavior with respect to such words. Your example of the past tense of "bid" is a fine such example. You can run an experiment like the one you suggested to find out what people think the past tense is. If you then found that 20/50 people responded "bidded" and 30/50 respond "bid", that is a lot of useful information. As you suggest in your discussion, this result wouldn't answer the question of how past tense is stored in each individual. This result would be ambiguous among several possible explanations. One possibility is that the probability distribution that is being discovered reflects different dialects, such that 2/5 of the population has one past tense, and 3/5 has another. Another possibility is that each person has a similar probability distribution in their heads, such that 2/5 of the time I respond one way, and 3/5 of the time I respond another. Further experiments would be necessary to answer between these and other possible theories (e.g., with repeated trials from the same person, carefully planned so that the participants don't notice that they are being asked multiple times). Without the quantitative evidence in the first place, there is no way to answer these kinds of questions. Regarding the past tense of "go", this would be useful as a baseline in an experiment involving the less frequent ones. So, yes, it would useful to gather quantitative evidence in such a case also, as baselines with respect to the more interesting cases for theories. The bottom line: if you have a generalization about a language that you wish to evaluate (such that you hypothesize that it is true across the speakers of the language), then you need quantitative evidence from multiple individuals, using an unbiased data collection method, to evaluate such a claim. The point about Mechanical Turk is that it is really *easy* to do this now, at least for languages like English. Best wishes, Ted Gibson & Ev Fedorenko On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying > that I shouldn't claim, for example, that the past tense of GO is > "went" without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic > behaviour (language 'out there', E-language) and trying to explain > it to ourselves in terms of a language system (language 'in here', I- > language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language. > > Given that view of language development, I don't see how > quantitative data will help. Let's take a real uncertainty, such as > the past tense of BID. If I want to say I did it, do I say "I > bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > people to oblige on Mechanical Turk, and 20 of them give "bidded" > and 30 "bid". So what? Does that mean that the correct answer is > "bidded"? Surely not. How is it better than my judgement? I agree > you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a > rare word, so in a sense quantitative data are irrelevant even > there. What would solve the problem of subjectivity, of course, > would be a machine for probing the bit of my mind (or even brain) > that holds BID and its details; but I suspect that even that > wouldn't move us much further forward than my original "don't know". > (Incidentally I write as a fan of quantitative sociolinguistics, so > I do accept that quantitative data are relevant to linguistic > analysis in some areas, where the I-language phenomenon is frequent > enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or > individual. The problem isn't unique to linguistics of course; it's > the same throughout the social sciences. But what's special about > linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the > differences between individuals really matter. I don't see that > we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate > reports of individual I-language. A rotten situation for a science, > but I don't see how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: >> Dear Dan, Dick: >> >> I would like to clarify some points that Dan Everett makes, in >> response to Dick Hudson. >> >> Ev Fedorenko and I have written a couple of papers recently (Gibson & >> Fedorenko, 2010, in press, see references and links below) on what we >> think are weak methodological standards in syntax and semantics >> research over the past many years. The issue that we address is the >> prevalent method in syntax and semantics research, which involves >> obtaining a judgment of the acceptability of a sentence / meaning >> pair, typically by just the author of the paper, sometimes with >> feedback from colleagues. As we address in our papers, this >> methodology does not allow proper testing of scientific hypotheses >> because of (a) the small number of experimental participants >> (typically one); (b) the small number of experimental stimuli >> (typically one); (c) cognitive biases on the part of the researcher >> and participants; and (d) the effect of the preceding context (e.g., >> other constructions the researcher may have been recently >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >> several others cited in Gibson & Fedorenko, in press; for similar >> points, but with not as strong a conclusion as ours). >> >> Three issues need to be separated here: (1) the use of intuitive >> judgments as a dependent measure in a language experiment; (2) >> potential cognitive biases on the part of experimental subjects and >> experimenters in language experiments; and (3) the need for obtaining >> quantitative evidence, whatever the dependent measure might be. The >> paper that Ev and I wrote addresses the last two issues, but does not >> go into depth on the first issue (the use of intuitions as a >> dependent >> measure in language experiments). Regarding this issue, we don't >> think >> that there is anything wrong with gathering intuitive judgments as a >> dependent measure, as long as the task is clear to the experimental >> participants. >> >> In the longer paper (Gibson & Fedorenko, in press) we respond to some >> arguments that have been given in support of continuing to use the >> traditional non-quantitative method in syntax / semantics research. >> One recent defense of the traditional method comes from Phillips >> (2008), who argues that no harm has come from the non-quantitative >> approach in syntax research thus far. Phillips argues that there are >> no cases in the literature where an incorrect intuitive judgment has >> become the basis for a widely accepted generalization or an important >> theoretical claim. He therefore concludes that there is no reason to >> adopt more rigorous data collection standards. We challenge Philips’ >> conclusion by presenting three cases from the literature where a >> faulty intuition has led to incorrect generalizations and mistaken >> theorizing, plausibly due to cognitive biases on the part of the >> researchers. >> >> A second argument that is sometimes presented for the continued use >> of >> the traditional non-quantitative method is that it would be too >> inefficient to evaluate every syntactic / semantic hypothesis or >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >> make this argument explicitly in their response to Gibson & Fedorenko >> (2010): “It would cripple linguistic investigation if it were >> required >> that all judgments of ambiguity and grammaticality be subject to >> statistically rigorous experiments on naive subjects, especially when >> investigating languages whose speakers are hard to access” (Culicover >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >> earlier in the discussion here.) Whereas we agree that in >> circumstances where gathering data is difficult, some evidence is >> better than no evidence, we do not agree that research would be >> slowed >> with respect to languages where experimental participants are easy to >> access, such as English. In contrast, we think that the opposite is >> true: the field’s progress is probably slowed by not doing >> quantitative research. >> Suppose that a typical syntax / semantics paper that lacks >> quantitative evidence includes judgments for 50 or more sentences / >> meaning pairs, corresponding to 50 or more empirical claims. Even if >> most of the judgments from such a paper are correct or are on the >> right track, the problem is in knowing which judgments are correct. >> For example, suppose that 90% of the judgments from an arbitrary >> paper >> are correct (which is probably a high estimate). (Colin Phillips and >> some of his former students / postdocs have commented to us that, in >> their experience, quantitative acceptability judgment studies almost >> always validate the claim(s) in the literature. This is not our >> experience, however. Most experiments that we have run which attempt >> to test some syntactic / semantic hypothesis in the literature end up >> providing us with a pattern of data that had not been known before >> the >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >> that in a paper with 50 empirical claims 45/50 are correct. But which >> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over >> two million different theories. By quantitatively evaluating the >> empirical claims, we reduce the uncertainty a great deal. To make >> progress, it is better to have theoretical claims supported by solid >> quantitative evidence, so that even if the interpretation of the data >> changes over time as new evidence becomes available – as is often the >> case in any field of science – the empirical pattern can be used as a >> basis for further theorizing. >> >> Furthermore, it is no longer expensive to run behavioral experiments, >> at least in English and other widely spoken languages. There now >> exists a marketplace interface – Amazon.com’s Mechanical Turk – which >> can be used for collecting behavioral data over the internet quickly >> and inexpensively. The cost of using an interface like this is >> minimal, and the time that it takes for the results to be returned is >> short. For example, currently on Mechanical Turk, a survey of >> approximately 50 items will be answered by 50 or more participants >> within a couple of hours, at a cost of approximately $1 per >> participant. Thus a survey can be completed within a day, at a cost >> of >> less than $50. (The hard work of designing the experiment, and >> constructing controlled materials remains of course.) >> >> Sorry to be so verbose. But I think that these methodological points >> are very important. >> >> Best wishes, >> >> Ted Gibson >> >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >> methods in syntax and semantics research. Language and Cognitive >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >> & Fedorenko InPress LCP.pdf >> >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >> linguistics research. Trends in Cognitive Science, 14, 233-234. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >> Fedorenko >> 2010 TICS.pdf >> >> >> >> >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that >>> intuitions are a fine way to generate hypotheses and even to test >>> them - to a degree. But while it might not have been feasible for >>> Huddleston, Pullum, and the other contributors to the Cambridge >>> Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, corpora, and standard psycholinguistic experimentation >>> (indeed, Standard Social Science Methodology) is vital for taking >>> the >>> field forward and for providing the best support for different >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>> paper on this, showing serious shortcomings with intuitions as the >>> sole source of evidence, in their paper: "The need for quantitative >>> methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written >>> convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>> Science), and researchers from Brazil are beginning a third round of >>> experimental work among the Pirahas, since my own work on the syntax >>> was, like almost every other field researcher's, based on native >>> speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions >>> to Greenberg's work on classifying the languages of the Americas. >>> His >>> methods were strongly (and justifiably) criticized. However, I >>> always >>> thought that his methods were a great way of generating hypotheses, >>> so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions. >>> >>> -- Dan >> >> >> >>>> We linguists can add a further layer of explanation to the >>>> judgements, but some judgements do seem to be more reliable than >>>> others. And if we have to wait for psycholinguistic evidence for >>>> every detailed analysis we make, our whole discipline will >>>> immediately grind to a halt. Like it or not, native speaker >>>> judgements are what put us linguists ahead of the rest in handling >>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>> Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >> >> >> From dryer at buffalo.edu Fri Sep 10 18:51:45 2010 From: dryer at buffalo.edu (dryer at buffalo.edu) Date: Fri, 10 Sep 2010 14:51:45 -0400 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: The following sentence of Lise's "Since we don't in fact have such an agreed-on terminology, we have to be quite careful in making clear what we are referring to when we talk about 'the correct analysis' of a form like 'unhappiness'" suggests that there are two senses in which an analysis can be "the correct analysis", one in terms of what is in people's heads, the other in terms of what is "out there". There are a vast number of patterns "out there". The only distinction, on my view, amongst this vast number of patterns, is between those that reflect something inside people's heads and those that don't. But if that is the case, then there is no coherent sense in which one can talk of "the correct analysis" of what is "out there", except in terms of what is in people's heads, and thus no second sense of "the correct analysis". The patterns that don't correspond to things in people's heads fall into (at least) two categories. There are those that are akin to constellations of stars and, as with constellations, there is no reality to these patterns, except in the minds of linguists. And there are those patterns which are the fossil remains of what was in the heads of speakers of an earlier stage of the language but which no longer are. These latter patterns are real, and they are relevant to exlaining why the language is now the way it is, but they are not relevant, I think many would agree, as to what is the "correct analysis" of the language today. For this reason, I claim that the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads. Again, I recommend the work of Bruce Derwing for lengthy discussion of these issues. Matthew --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. And for > distinguishing among the degrees of pattern apprehension that a person > may have, from vague preferences detectable in reaction times or other > behavior all the way up through clear metalinguistic insights. Dick > Hudson's note reminding us of the Gleitman and Gleitman study is right > on target. > > Since we don't in fact have such an agreed-on terminology, we have to > be quite careful in making clear what we are referring to when we talk > about 'the correct analysis' of a form like 'unhappiness'. We know, but > tend to forget - and tend to forget to tell our students! - that it's an > empirical question as to whether the formal simplicity and coherence of > description of forms 'out there' (e.g. lovely abstract morphophonemics) > is any kind of approximation to the way knowledge of the same forms is > organized in a particular person's head. If we remember that a very > large proportion of what we know about our language is 'out there' when > we are infants and has to be internalized through experience with the > language (even if you believe in innate 'core language'), the variation > in internal knowledge from one person to another is more understandable. > > We especially need to consider (and try to test) the possibility that > since > the brain can make multiple cross-connections, multiple patterns are > involved > simultaneously in morphological and syntactic analyses. I suggest that > that's the case with 'unhappiness' - and the linguistic analyses that I > know about are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> >> Two comments. >> >> First (elaborating perhaps on Dick Hudson's comment), I think there >> is an >> important distinction between low-level linguistic intuitions (like >> whether a >> word or sentence is well-formed or what it means) and higher-level >> intuitions >> (like what the structure of a word or sentence is). One can take >> the position >> that we need to account for the former (while recognizing that they >> are not >> always reliable) but not the latter. >> >> Second, the tension here is not only between evidence from speaker >> intuitions >> versus evidence from psycholinguistic experiments. There is also a >> tension >> between deciding on the correct analysis on the basis of a priori >> simplicity >> arguments versus deciding on the correct analysis on the basis of >> psycholinguistic evidence (see Derwing 1973). The bracketing >> paradox that Dan >> referred to that arises with the word (semantics argues >> for >> [[un+happi] + er], morphology and phonology argues for [un + [happi >> +er]] (the >> comparative suffix can only be attached to adjectives containing one >> or two >> syllables) is only a paradox if one assumes that speakers adopt the >> simplest >> analysis. For example, if speakers adopt a more complex rule for >> either of these >> (e.g. perhaps the rule for attaching -er can apply exceptionally to >> trisyllabic >> words beginning with un-), then the bracketing paradox disappears. >> >> Matthew >> >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>> conscious judgements are only one kind of evidence that we need to >>> take >>> into account. I admire Carson Schutze's work (which I reviewed in >>> fact), >>> and of course I've been aware of complaints about judgements by >>> people >>> like Labov for decades. >>> >>> But you're missing my main point, which is that all judgements aren't >>> equally reliable. If you want to know how /unhappiness/ is >>> structured, >>> ask a linguist, not a five-year old. And one of the by-products of >>> education may be increased sensitivity to syntax - which is one of >>> the >>> many reasons why linguists need to pay more attention to education. >>> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> On 09/09/2010 11:39, Daniel Everett wrote: >>>> Dick, >>>> >>>> You raise an important issue here about >>> methodology. I believe that intuitions are a fine way to generate >>> hypotheses and even to test them - to a degree. But while it might >>> not have >>> been feasible for Huddleston, Pullum, and the other contributors to >>> the >>> Cambridge Grammar to conduct experiments on every point of the >>> grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, >>> corpora, and standard psycholinguistic experimentation (indeed, >>> Standard >>> Social Science Methodology) is vital for taking the field forward >>> and for >>> providing the best support for different analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new paper on this, showing >>> serious >>> shortcomings with intuitions as the sole source of evidence, in their >>> paper: "The need for quantitative methods in syntax and semantics >>> research".> >>>> Carson Schutze and Wayne Cowart, among others, >>> have also written convincingly on this.> >>>> It is one reason that a team from Stanford, MIT >>> (Brain and Cognitive Science), and researchers from Brazil are >>> beginning a >>> third round of experimental work among the Pirahas, since my own >>> work on >>> the syntax was, like almost every other field researcher's, based >>> on native >>> speaker intuitions and corpora.> >>>> The discussion of methodologies reminds me of >>> the initial reactions to Greenberg's work on classifying the >>> languages of >>> the Americas. His methods were strongly (and justifiably) criticized. >>> However, I always thought that his methods were a great way of >>> generating >>> hypotheses, so long as they were ultimately put to the test of >>> standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions.> >>>> -- Dan >>>> >>>>> We linguists can add a further layer of >>> explanation to the judgements, but some judgements do seem to be more >>> reliable than others. And if we have to wait for psycholinguistic >>> evidence >>> for every detailed analysis we make, our whole discipline will >>> immediately >>> grind to a halt. Like it or not, native speaker judgements are what >>> put us >>> linguists ahead of the rest in handling fine detail. Imagine >>> writing the >>> Cambridge Grammar of the English Language (or the OED) without >>> using native >>> speaker judgements.>> >>>>> Best wishes, Dick Hudson >>>>> >>>> >>>> >>> >>> >>> >>> >>> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > From amnfn at well.com Fri Sep 10 19:09:07 2010 From: amnfn at well.com (A. Katz) Date: Fri, 10 Sep 2010 12:09:07 -0700 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284130304@cast-dryerm2.caset.buffalo.edu> Message-ID: Matthew, Thanks for stating that, because I was almost beginning to imagine that there was no essential disagreement, and that all of us agree that there is more -- and less -- to language than what is found in people's heads. Your position is the one I am familiar with from the functionalist point of view, and I was beginning to feel that it was underrepresented on Funknet. Those of us who disagree with your stated position -- but are very familiar with it -- are interested not just in psycholinguistics and how people process language -- but also in the communicative function of language as a system whereby information is transferred. Just as you and I may not be aware of the way our emails are encoded and then decoded by the computers that help us send emails back and forth, speakers may be compeltely unaware of what language does in order to transmit information. After speakers have finished sending forth their linguistic output, it matters not at all how they arrived at this output. Language processing is separate from language in the same way that data processing is separate from data. Best, --Aya On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have to be > quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the correct > analysis", one in terms of what is in people's heads, the other in terms of > what is "out there". > > There are a vast number of patterns "out there". The only distinction, on my > view, amongst this vast number of patterns, is between those that reflect > something inside people's heads and those that don't. But if that is the > case, then there is no coherent sense in which one can talk of "the correct > analysis" of what is "out there", except in terms of what is in people's > heads, and thus no second sense of "the correct analysis". The patterns that > don't correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of stars and, as > with constellations, there is no reality to these patterns, except in the > minds of linguists. And there are those patterns which are the fossil > remains of what was in the heads of speakers of an earlier stage of the > language but which no longer are. These latter patterns are real, and they > are relevant to exlaining why the language is now the way it is, but they are > not relevant, I think many would agree, as to what is the "correct analysis" > of the language today. > > For this reason, I claim that the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion of these > issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > >> I wish we had better terminology for keeping track of whether, at a >> given time, we are talking about the patterns that are 'out there' in >> the language and might possibly be apprehended (subconsciously) by a >> speaker, and when we are talking about the patterns that a particular >> speaker actually does apprehend, as indicated by experiments, from >> simple 'wug tests' up to brain wave and eye-gaze studies. And for >> distinguishing among the degrees of pattern apprehension that a person >> may have, from vague preferences detectable in reaction times or other >> behavior all the way up through clear metalinguistic insights. Dick >> Hudson's note reminding us of the Gleitman and Gleitman study is right >> on target. >> >> Since we don't in fact have such an agreed-on terminology, we have to >> be quite careful in making clear what we are referring to when we talk >> about 'the correct analysis' of a form like 'unhappiness'. We know, but >> tend to forget - and tend to forget to tell our students! - that it's an >> empirical question as to whether the formal simplicity and coherence of >> description of forms 'out there' (e.g. lovely abstract morphophonemics) >> is any kind of approximation to the way knowledge of the same forms is >> organized in a particular person's head. If we remember that a very >> large proportion of what we know about our language is 'out there' when >> we are infants and has to be internalized through experience with the >> language (even if you believe in innate 'core language'), the variation >> in internal knowledge from one person to another is more understandable. >> We especially need to consider (and try to test) the >> possibility that >> since >> the brain can make multiple cross-connections, multiple patterns are >> involved >> simultaneously in morphological and syntactic analyses. I suggest that >> that's the case with 'unhappiness' - and the linguistic analyses that I >> know about are not good at handling that kind of idea. >> >> Lise >> >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >> >>> >>> Two comments. >>> >>> First (elaborating perhaps on Dick Hudson's comment), I think there >>> is an >>> important distinction between low-level linguistic intuitions (like >>> whether a >>> word or sentence is well-formed or what it means) and higher-level >>> intuitions >>> (like what the structure of a word or sentence is). One can take >>> the position >>> that we need to account for the former (while recognizing that they >>> are not >>> always reliable) but not the latter. >>> >>> Second, the tension here is not only between evidence from speaker >>> intuitions >>> versus evidence from psycholinguistic experiments. There is also a >>> tension >>> between deciding on the correct analysis on the basis of a priori >>> simplicity >>> arguments versus deciding on the correct analysis on the basis of >>> psycholinguistic evidence (see Derwing 1973). The bracketing >>> paradox that Dan >>> referred to that arises with the word (semantics argues >>> for >>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>> +er]] (the >>> comparative suffix can only be attached to adjectives containing one >>> or two >>> syllables) is only a paradox if one assumes that speakers adopt the >>> simplest >>> analysis. For example, if speakers adopt a more complex rule for >>> either of these >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>> trisyllabic >>> words beginning with un-), then the bracketing paradox disappears. >>> >>> Matthew >>> >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>> conscious judgements are only one kind of evidence that we need to >>>> take >>>> into account. I admire Carson Schutze's work (which I reviewed in >>>> fact), >>>> and of course I've been aware of complaints about judgements by >>>> people >>>> like Labov for decades. >>>> >>>> But you're missing my main point, which is that all judgements aren't >>>> equally reliable. If you want to know how /unhappiness/ is >>>> structured, >>>> ask a linguist, not a five-year old. And one of the by-products of >>>> education may be increased sensitivity to syntax - which is one of >>>> the >>>> many reasons why linguists need to pay more attention to education. >>>> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>> Dick, >>>>> >>>>> You raise an important issue here about >>>> methodology. I believe that intuitions are a fine way to generate >>>> hypotheses and even to test them - to a degree. But while it might >>>> not have >>>> been feasible for Huddleston, Pullum, and the other contributors to >>>> the >>>> Cambridge Grammar to conduct experiments on every point of the >>>> grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, >>>> corpora, and standard psycholinguistic experimentation (indeed, >>>> Standard >>>> Social Science Methodology) is vital for taking the field forward >>>> and for >>>> providing the best support for different analyses. Ted Gibson and Ev >>>> Fedorenko have written a very useful new paper on this, showing >>>> serious >>>> shortcomings with intuitions as the sole source of evidence, in their >>>> paper: "The need for quantitative methods in syntax and semantics >>>> research".> >>>>> Carson Schutze and Wayne Cowart, among others, >>>> have also written convincingly on this.> >>>>> It is one reason that a team from Stanford, MIT >>>> (Brain and Cognitive Science), and researchers from Brazil are >>>> beginning a >>>> third round of experimental work among the Pirahas, since my own >>>> work on >>>> the syntax was, like almost every other field researcher's, based >>>> on native >>>> speaker intuitions and corpora.> >>>>> The discussion of methodologies reminds me of >>>> the initial reactions to Greenberg's work on classifying the >>>> languages of >>>> the Americas. His methods were strongly (and justifiably) criticized. >>>> However, I always thought that his methods were a great way of >>>> generating >>>> hypotheses, so long as they were ultimately put to the test of >>>> standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions.> >>>>> -- Dan >>>>> >>>>>> We linguists can add a further layer of >>>> explanation to the judgements, but some judgements do seem to be more >>>> reliable than others. And if we have to wait for psycholinguistic >>>> evidence >>>> for every detailed analysis we make, our whole discipline will >>>> immediately >>>> grind to a halt. Like it or not, native speaker judgements are what >>>> put us >>>> linguists ahead of the rest in handling fine detail. Imagine >>>> writing the >>>> Cambridge Grammar of the English Language (or the OED) without >>>> using native >>>> speaker judgements.>> >>>>>> Best wishes, Dick Hudson >>>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> >>>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> >> Campus Mail Address: >> UCB 594, Institute of Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> >> > > > > > From kuzar at research.haifa.ac.il Fri Sep 10 20:26:18 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Fri, 10 Sep 2010 23:26:18 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: Message-ID: The Modern Hebrew data supplied by Eitan are incomplete. Hebrew distinguishes between locution (say, hear, think, etc.) and situation (action, event, state, etc.). What Eitan describes is only true with regard to nouns (and clauses) expressing locution. 'Announcement' is indeed such a noun. Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', etc., whose denotatum is a situation, cannot be followed by ki, but only by Se-, e.g.: margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) annoys me the-situation that-all went 'I am upset about the situation that all have gone' On the other hand, the relative Se- may be replaced by the more elegant and classical aSer, while the Se- of situation clauses may not. Sorry about the invented example. I am overseas now. All this has been described (with corpus data) in: Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit [Hebrew Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. The article is somewhat outdated and contains some inaccuracies I would formulate differently today, but the basic distinction is valid in my opinion. Best, Ron Kuzar --------------- On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > Hi all, > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > comparison can help what seems to be a purely theoretical question based on > a single language. The problem here is that English uses the same element > to > mark regular relatives and these "appositional" relatives. But if at least > one language encodes them by different means, then there's at least a good > case for seeing them as distinct functions. It's basically the same > principle that's used to decide whether to put a meaning on a semantic map. > So here are two languages that I know that encode them differently. > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > clause (ki), which differs from the relative clause marker (Se-), e.g. > > ha-hoda'a Se-kibalnu > the-announcment rel-we_got > "The announcement that we got." > > ha-hoda'a ki hitbatel ha-mifgaS > the-message CMP was_cancelled the-meeting > "The announcement that the meeting was cancelled." > > In Coptic, these clauses are marked by ce-, which marks complement clauses, > *inter alia*, but not relative clauses: > > ph-mewi ce- (complement clause) > 'the-thought that (we are angry)' > > ph-mewi ete- (relative clause) > 'the thought that (we used to think)' > > This seems to be a pretty clear indication that these are complement > clauses > rather than relatives. Even if one doesn't like the notion of nouns taking > complement clauses (and why not? nominalizations in some languages can take > accusative modifiers as well as genitives), it still probably isn't > incidental that the nominalizations are of verbs that take complement > clauses when finite. > > As usual, the perspective in Talmy Givón's *Syntax* (vol. 2) is worth > looking at. > > Best, > Eitan > > > On 10 September 2010 19:21, Arie Verhagen > wrote: > > > And as another addition: the clauses that can only be introduced by > *that* > > (with no role to > > play in the subordinate clause) may be seen as (subtypes of) complement > > clauses, > > expressing a proposition with the noun functioning as Complement Taking > > Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > > constructions like > > "The claim is that X" (traditionally analysed as subject clauses), "I > claim > > that X", "I put forward > > the claim that X", in which the relationship between the verb or noun and > > the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > To add to Joanne's comments: > > > > > > There are basically three ways to distinguish an appositive clause > > > (AC) from a relative clause (RC). > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > pronouns. > > > > > > 2) The AC and the noun it modifies display an equative > relationship--one > > can say X > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > often doesn't > > > (except, perhaps, when the relative clause is sentential). > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > relative pronoun > > > is always part of the clause in an RC. > > > > > > Rong Chen > > > > > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > -- =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From macw at cmu.edu Fri Sep 10 22:23:54 2010 From: macw at cmu.edu (Brian MacWhinney) Date: Fri, 10 Sep 2010 18:23:54 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8A71F0.4050507@ling.ucl.ac.uk> Message-ID: Dick and Ted, I agree with parts of what each of you are saying. Which means that I also disagree with other parts. In practice,, Gibson and Fedorenko, in press, (which I downloaded and scanned) deals with no more than two or three constructions. They mention the fact that people don't have problems with sentences such as "Susan muttered him the news" despite claims that verbs such as "mutter" cannot take the double object construction. They also note that the claims from Jackendoff and Culicover about the differences between the two sentences below are not supported by results from the Mechanical Turk: 1. Peter was trying to remember who carried what. 2. Peter was trying to remember who carried what when. These are interesting facts. If these sentences are supposed to be different and people judge them to be similarly grammatical, then theories based on the supposed differences should be reexamined. There are big chunks of syntactic theory resting on shaky judgments about complex sentences of this type. Getting some of this straight would be a big win, I would say, particularly if linguists would pay attention to the results. But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications. But, Dick, you then move on to questioning data on bid-bidded. Here we have a case of true variation in the population. I would love to know its distribution. As a "fan of quantitative sociolinguistics" shouldn't you too? My take on this is that constructions are not created equal. The three types mentioned here are probably just a start on an inventory of evidentiary types. We need to correctly pair up appropriate methods with each of the types. And we to make sure that people pay attention to the results, once they are in --Brian MacWhinney On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. > > Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: >> Dear Dan, Dick: >> >> I would like to clarify some points that Dan Everett makes, in >> response to Dick Hudson. >> >> Ev Fedorenko and I have written a couple of papers recently (Gibson & >> Fedorenko, 2010, in press, see references and links below) on what we >> think are weak methodological standards in syntax and semantics >> research over the past many years. The issue that we address is the >> prevalent method in syntax and semantics research, which involves >> obtaining a judgment of the acceptability of a sentence / meaning >> pair, typically by just the author of the paper, sometimes with >> feedback from colleagues. As we address in our papers, this >> methodology does not allow proper testing of scientific hypotheses >> because of (a) the small number of experimental participants >> (typically one); (b) the small number of experimental stimuli >> (typically one); (c) cognitive biases on the part of the researcher >> and participants; and (d) the effect of the preceding context (e.g., >> other constructions the researcher may have been recently >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >> several others cited in Gibson & Fedorenko, in press; for similar >> points, but with not as strong a conclusion as ours). >> >> Three issues need to be separated here: (1) the use of intuitive >> judgments as a dependent measure in a language experiment; (2) >> potential cognitive biases on the part of experimental subjects and >> experimenters in language experiments; and (3) the need for obtaining >> quantitative evidence, whatever the dependent measure might be. The >> paper that Ev and I wrote addresses the last two issues, but does not >> go into depth on the first issue (the use of intuitions as a dependent >> measure in language experiments). Regarding this issue, we don't think >> that there is anything wrong with gathering intuitive judgments as a >> dependent measure, as long as the task is clear to the experimental >> participants. >> >> In the longer paper (Gibson & Fedorenko, in press) we respond to some >> arguments that have been given in support of continuing to use the >> traditional non-quantitative method in syntax / semantics research. >> One recent defense of the traditional method comes from Phillips >> (2008), who argues that no harm has come from the non-quantitative >> approach in syntax research thus far. Phillips argues that there are >> no cases in the literature where an incorrect intuitive judgment has >> become the basis for a widely accepted generalization or an important >> theoretical claim. He therefore concludes that there is no reason to >> adopt more rigorous data collection standards. We challenge Philips’ >> conclusion by presenting three cases from the literature where a >> faulty intuition has led to incorrect generalizations and mistaken >> theorizing, plausibly due to cognitive biases on the part of the >> researchers. >> >> A second argument that is sometimes presented for the continued use of >> the traditional non-quantitative method is that it would be too >> inefficient to evaluate every syntactic / semantic hypothesis or >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >> make this argument explicitly in their response to Gibson & Fedorenko >> (2010): “It would cripple linguistic investigation if it were required >> that all judgments of ambiguity and grammaticality be subject to >> statistically rigorous experiments on naive subjects, especially when >> investigating languages whose speakers are hard to access” (Culicover >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >> earlier in the discussion here.) Whereas we agree that in >> circumstances where gathering data is difficult, some evidence is >> better than no evidence, we do not agree that research would be slowed >> with respect to languages where experimental participants are easy to >> access, such as English. In contrast, we think that the opposite is >> true: the field’s progress is probably slowed by not doing >> quantitative research. >> Suppose that a typical syntax / semantics paper that lacks >> quantitative evidence includes judgments for 50 or more sentences / >> meaning pairs, corresponding to 50 or more empirical claims. Even if >> most of the judgments from such a paper are correct or are on the >> right track, the problem is in knowing which judgments are correct. >> For example, suppose that 90% of the judgments from an arbitrary paper >> are correct (which is probably a high estimate). (Colin Phillips and >> some of his former students / postdocs have commented to us that, in >> their experience, quantitative acceptability judgment studies almost >> always validate the claim(s) in the literature. This is not our >> experience, however. Most experiments that we have run which attempt >> to test some syntactic / semantic hypothesis in the literature end up >> providing us with a pattern of data that had not been known before the >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >> that in a paper with 50 empirical claims 45/50 are correct. But which >> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over >> two million different theories. By quantitatively evaluating the >> empirical claims, we reduce the uncertainty a great deal. To make >> progress, it is better to have theoretical claims supported by solid >> quantitative evidence, so that even if the interpretation of the data >> changes over time as new evidence becomes available – as is often the >> case in any field of science – the empirical pattern can be used as a >> basis for further theorizing. >> >> Furthermore, it is no longer expensive to run behavioral experiments, >> at least in English and other widely spoken languages. There now >> exists a marketplace interface – Amazon.com’s Mechanical Turk – which >> can be used for collecting behavioral data over the internet quickly >> and inexpensively. The cost of using an interface like this is >> minimal, and the time that it takes for the results to be returned is >> short. For example, currently on Mechanical Turk, a survey of >> approximately 50 items will be answered by 50 or more participants >> within a couple of hours, at a cost of approximately $1 per >> participant. Thus a survey can be completed within a day, at a cost of >> less than $50. (The hard work of designing the experiment, and >> constructing controlled materials remains of course.) >> >> Sorry to be so verbose. But I think that these methodological points >> are very important. >> >> Best wishes, >> >> Ted Gibson >> >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >> methods in syntax and semantics research. Language and Cognitive >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >> & Fedorenko InPress LCP.pdf >> >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >> linguistics research. Trends in Cognitive Science, 14, 233-234. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >> 2010 TICS.pdf >> >> >> >> >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that >>> intuitions are a fine way to generate hypotheses and even to test >>> them - to a degree. But while it might not have been feasible for >>> Huddleston, Pullum, and the other contributors to the Cambridge >>> Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, corpora, and standard psycholinguistic experimentation >>> (indeed, Standard Social Science Methodology) is vital for taking the >>> field forward and for providing the best support for different >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>> paper on this, showing serious shortcomings with intuitions as the >>> sole source of evidence, in their paper: "The need for quantitative >>> methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written >>> convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>> Science), and researchers from Brazil are beginning a third round of >>> experimental work among the Pirahas, since my own work on the syntax >>> was, like almost every other field researcher's, based on native >>> speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions >>> to Greenberg's work on classifying the languages of the Americas. His >>> methods were strongly (and justifiably) criticized. However, I always >>> thought that his methods were a great way of generating hypotheses, >>> so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions. >>> >>> -- Dan >> >> >> >>>> We linguists can add a further layer of explanation to the >>>> judgements, but some judgements do seem to be more reliable than >>>> others. And if we have to wait for psycholinguistic evidence for >>>> every detailed analysis we make, our whole discipline will >>>> immediately grind to a halt. Like it or not, native speaker >>>> judgements are what put us linguists ahead of the rest in handling >>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>> Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >> >> >> > From dryer at buffalo.edu Fri Sep 10 22:33:49 2010 From: dryer at buffalo.edu (dryer at buffalo.edu) Date: Fri, 10 Sep 2010 18:33:49 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Aya, I actually agree with everything you say here. Personally, I am MORE interested in the communicative function of language than I am in psycholinguistics and how people process language. But none of that is relevant, I believe, to the very specific question of what it means for an analysis to be correct. While one might conclude from what I said that one ought to do psycholinguistics, that is not my intention. Rather, my conclusion is that since I myself prefer not to do psycholinguistics, I cannot really claim that the analyses I come up with are "the correct" ones. And if it is really important to someone that they identify "correct" analyses, then they ought to be doing psycholinguistics, since there is no coherent notion of correct analysis outside of what is inside of people's heads. Matthew --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" wrote: > Matthew, > > Thanks for stating that, because I was almost beginning to imagine that > there was no essential disagreement, and that all of us agree that there > is more -- and less -- to language than what is found in people's heads. > > Your position is the one I am familiar with from the functionalist point > of view, and I was beginning to feel that it was underrepresented on > Funknet. > > Those of us who disagree with your stated position -- but are very > familiar with it -- are interested not just in psycholinguistics and how > people process language -- but also in the communicative function of > language as a system whereby information is transferred. Just as you and > I may not be aware of the way our emails are encoded and then decoded by > the computers that help us send emails back and forth, speakers may be > compeltely unaware of what language does in order to transmit information. > > After speakers have finished sending forth their linguistic output, it > matters not at all how they arrived at this output. Language processing > is separate from language in the same way that data processing is > separate from data. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > >> >> The following sentence of Lise's >> >> "Since we don't in fact have such an agreed-on terminology, we have to >> be quite careful in making clear what we are referring to when we talk >> about 'the correct analysis' of a form like 'unhappiness'" >> >> suggests that there are two senses in which an analysis can be "the >> correct analysis", one in terms of what is in people's heads, the other >> in terms of what is "out there". >> >> There are a vast number of patterns "out there". The only distinction, >> on my view, amongst this vast number of patterns, is between those that >> reflect something inside people's heads and those that don't. But if >> that is the case, then there is no coherent sense in which one can talk >> of "the correct analysis" of what is "out there", except in terms of >> what is in people's heads, and thus no second sense of "the correct >> analysis". The patterns that don't correspond to things in people's >> heads fall into (at least) two categories. There are those that are >> akin to constellations of stars and, as with constellations, there is >> no reality to these patterns, except in the minds of linguists. And >> there are those patterns which are the fossil remains of what was in >> the heads of speakers of an earlier stage of the language but which no >> longer are. These latter patterns are real, and they are relevant to >> exlaining why the language is now the way it is, but they are not >> relevant, I think many would agree, as to what is the "correct analysis" >> of the language today. >> >> For this reason, I claim that the only sense in which an analysis can be >> "the correct analysis" is in terms of what is inside of people's heads. >> >> Again, I recommend the work of Bruce Derwing for lengthy discussion of >> these issues. >> >> Matthew >> >> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn >> wrote: >> >>> I wish we had better terminology for keeping track of whether, at a >>> given time, we are talking about the patterns that are 'out there' in >>> the language and might possibly be apprehended (subconsciously) by a >>> speaker, and when we are talking about the patterns that a particular >>> speaker actually does apprehend, as indicated by experiments, from >>> simple 'wug tests' up to brain wave and eye-gaze studies. And for >>> distinguishing among the degrees of pattern apprehension that a person >>> may have, from vague preferences detectable in reaction times or other >>> behavior all the way up through clear metalinguistic insights. Dick >>> Hudson's note reminding us of the Gleitman and Gleitman study is right >>> on target. >>> >>> Since we don't in fact have such an agreed-on terminology, we have to >>> be quite careful in making clear what we are referring to when we talk >>> about 'the correct analysis' of a form like 'unhappiness'. We know, >>> but tend to forget - and tend to forget to tell our students! - that >>> it's an empirical question as to whether the formal simplicity and >>> coherence of description of forms 'out there' (e.g. lovely abstract >>> morphophonemics) is any kind of approximation to the way knowledge of >>> the same forms is organized in a particular person's head. If we >>> remember that a very large proportion of what we know about our >>> language is 'out there' when we are infants and has to be internalized >>> through experience with the language (even if you believe in innate >>> 'core language'), the variation in internal knowledge from one person >>> to another is more understandable. We especially need to consider (and >>> try to test) the >>> possibility that >>> since >>> the brain can make multiple cross-connections, multiple patterns are >>> involved >>> simultaneously in morphological and syntactic analyses. I suggest that >>> that's the case with 'unhappiness' - and the linguistic analyses that I >>> know about are not good at handling that kind of idea. >>> >>> Lise >>> >>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >>> >>>> >>>> Two comments. >>>> >>>> First (elaborating perhaps on Dick Hudson's comment), I think there >>>> is an >>>> important distinction between low-level linguistic intuitions (like >>>> whether a >>>> word or sentence is well-formed or what it means) and higher-level >>>> intuitions >>>> (like what the structure of a word or sentence is). One can take >>>> the position >>>> that we need to account for the former (while recognizing that they >>>> are not >>>> always reliable) but not the latter. >>>> >>>> Second, the tension here is not only between evidence from speaker >>>> intuitions >>>> versus evidence from psycholinguistic experiments. There is also a >>>> tension >>>> between deciding on the correct analysis on the basis of a priori >>>> simplicity >>>> arguments versus deciding on the correct analysis on the basis of >>>> psycholinguistic evidence (see Derwing 1973). The bracketing >>>> paradox that Dan >>>> referred to that arises with the word (semantics argues >>>> for >>>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>>> +er]] (the >>>> comparative suffix can only be attached to adjectives containing one >>>> or two >>>> syllables) is only a paradox if one assumes that speakers adopt the >>>> simplest >>>> analysis. For example, if speakers adopt a more complex rule for >>>> either of these >>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>>> trisyllabic >>>> words beginning with un-), then the bracketing paradox disappears. >>>> >>>> Matthew >>>> >>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>>> conscious judgements are only one kind of evidence that we need to >>>>> take >>>>> into account. I admire Carson Schutze's work (which I reviewed in >>>>> fact), >>>>> and of course I've been aware of complaints about judgements by >>>>> people >>>>> like Labov for decades. >>>>> >>>>> But you're missing my main point, which is that all judgements aren't >>>>> equally reliable. If you want to know how /unhappiness/ is >>>>> structured, >>>>> ask a linguist, not a five-year old. And one of the by-products of >>>>> education may be increased sensitivity to syntax - which is one of >>>>> the >>>>> many reasons why linguists need to pay more attention to education. >>>>> >>>>> Best wishes, Dick >>>>> >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about >>>>> methodology. I believe that intuitions are a fine way to generate >>>>> hypotheses and even to test them - to a degree. But while it might >>>>> not have >>>>> been feasible for Huddleston, Pullum, and the other contributors to >>>>> the >>>>> Cambridge Grammar to conduct experiments on every point of the >>>>> grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, >>>>> corpora, and standard psycholinguistic experimentation (indeed, >>>>> Standard >>>>> Social Science Methodology) is vital for taking the field forward >>>>> and for >>>>> providing the best support for different analyses. Ted Gibson and Ev >>>>> Fedorenko have written a very useful new paper on this, showing >>>>> serious >>>>> shortcomings with intuitions as the sole source of evidence, in their >>>>> paper: "The need for quantitative methods in syntax and semantics >>>>> research".> >>>>>> Carson Schutze and Wayne Cowart, among others, >>>>> have also written convincingly on this.> >>>>>> It is one reason that a team from Stanford, MIT >>>>> (Brain and Cognitive Science), and researchers from Brazil are >>>>> beginning a >>>>> third round of experimental work among the Pirahas, since my own >>>>> work on >>>>> the syntax was, like almost every other field researcher's, based >>>>> on native >>>>> speaker intuitions and corpora.> >>>>>> The discussion of methodologies reminds me of >>>>> the initial reactions to Greenberg's work on classifying the >>>>> languages of >>>>> the Americas. His methods were strongly (and justifiably) criticized. >>>>> However, I always thought that his methods were a great way of >>>>> generating >>>>> hypotheses, so long as they were ultimately put to the test of >>>>> standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions.> >>>>>> -- Dan >>>>>> >>>>>>> We linguists can add a further layer of >>>>> explanation to the judgements, but some judgements do seem to be more >>>>> reliable than others. And if we have to wait for psycholinguistic >>>>> evidence >>>>> for every detailed analysis we make, our whole discipline will >>>>> immediately >>>>> grind to a halt. Like it or not, native speaker judgements are what >>>>> put us >>>>> linguists ahead of the rest in handling fine detail. Imagine >>>>> writing the >>>>> Cambridge Grammar of the English Language (or the OED) without >>>>> using native >>>>> speaker judgements.>> >>>>>>> Best wishes, Dick Hudson >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>> >>> Lise Menn Home Office: 303-444-4274 >>> 1625 Mariposa Ave Fax: 303-413-0017 >>> Boulder CO 80302 >>> >>> Professor Emerita of Linguistics >>> Fellow, Institute of Cognitive Science >>> University of Colorado >>> >>> Secretary, AAAS Section Z [Linguistics] >>> >>> Campus Mail Address: >>> UCB 594, Institute of Cognitive Science >>> >>> Campus Physical Address: >>> CINC 234 >>> 1777 Exposition Ave, Boulder >>> >>> >>> >>> >> >> >> >> >> > > From dick at ling.ucl.ac.uk Fri Sep 10 23:20:15 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 00:20:15 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Brian, What a helpful message! I think you're right: we need a typology of cases, each needing a different range of methods, ranging from the linguist's own judgements for really easy cases to more complicated quantitative methods for more complicated ones. The trouble with our discipline is that for any community of N speakers, and a language consisting of M 'items' (however you may choose to define 'community' and 'item'), we have N*M datapoints that, in principle, all need to be validated somehow. We might reduce the number by focusing on one speaker, but then you can't use data from other speakers as evidence for that speaker's language; or we might try to construct a 'typical' speaker, but we don't know how to do that; or we might reduce the size of the community by trying to find a 'dialect' (but dialects don't really exist); or we might ignore most of the linguistic items and focus on, say, the modal verbs - but then we miss their links to all the other items. It's different for psycholinguists because they're only interested in general processes, for which linguistic items are just evidence, not the thing under investigation; but for us linguists, the fine detail is everything because we're the people who explore the connections between items. So I look forward to the day when your typology of cases will guide us through a range of different methods to the appropriate ones for any given item. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 23:23, Brian MacWhinney wrote: > Dick and Ted, > I agree with parts of what each of you are saying. Which means that I also disagree with other parts. In practice,, Gibson and Fedorenko, in press, (which I downloaded and scanned) deals with no more than two or three constructions. They mention the fact that people don't have problems with sentences such as "Susan muttered him the news" despite claims that verbs such as "mutter" cannot take the double object construction. They also note that the claims from Jackendoff and Culicover about the differences between the two sentences below are not supported by results from the Mechanical Turk: > 1. Peter was trying to remember who carried what. > 2. Peter was trying to remember who carried what when. > These are interesting facts. If these sentences are supposed to be different and people judge them to be similarly grammatical, then theories based on the supposed differences should be reexamined. There are big chunks of syntactic theory resting on shaky judgments about complex sentences of this type. Getting some of this straight would be a big win, I would say, particularly if linguists would pay attention to the results. > But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications. > But, Dick, you then move on to questioning data on bid-bidded. Here we have a case of true variation in the population. I would love to know its distribution. As a "fan of quantitative sociolinguistics" shouldn't you too? > My take on this is that constructions are not created equal. The three types mentioned here are probably just a start on an inventory of evidentiary types. We need to correctly pair up appropriate methods with each of the types. And we to make sure that people pay attention to the results, once they are in > > --Brian MacWhinney > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> Dear Ted, >> Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? >> >> Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. >> >> Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) >> >> It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. >> >> Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 14:03, Ted Gibson wrote: >>> Dear Dan, Dick: >>> >>> I would like to clarify some points that Dan Everett makes, in >>> response to Dick Hudson. >>> >>> Ev Fedorenko and I have written a couple of papers recently (Gibson& >>> Fedorenko, 2010, in press, see references and links below) on what we >>> think are weak methodological standards in syntax and semantics >>> research over the past many years. The issue that we address is the >>> prevalent method in syntax and semantics research, which involves >>> obtaining a judgment of the acceptability of a sentence / meaning >>> pair, typically by just the author of the paper, sometimes with >>> feedback from colleagues. As we address in our papers, this >>> methodology does not allow proper testing of scientific hypotheses >>> because of (a) the small number of experimental participants >>> (typically one); (b) the small number of experimental stimuli >>> (typically one); (c) cognitive biases on the part of the researcher >>> and participants; and (d) the effect of the preceding context (e.g., >>> other constructions the researcher may have been recently >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>> several others cited in Gibson& Fedorenko, in press; for similar >>> points, but with not as strong a conclusion as ours). >>> >>> Three issues need to be separated here: (1) the use of intuitive >>> judgments as a dependent measure in a language experiment; (2) >>> potential cognitive biases on the part of experimental subjects and >>> experimenters in language experiments; and (3) the need for obtaining >>> quantitative evidence, whatever the dependent measure might be. The >>> paper that Ev and I wrote addresses the last two issues, but does not >>> go into depth on the first issue (the use of intuitions as a dependent >>> measure in language experiments). Regarding this issue, we don't think >>> that there is anything wrong with gathering intuitive judgments as a >>> dependent measure, as long as the task is clear to the experimental >>> participants. >>> >>> In the longer paper (Gibson& Fedorenko, in press) we respond to some >>> arguments that have been given in support of continuing to use the >>> traditional non-quantitative method in syntax / semantics research. >>> One recent defense of the traditional method comes from Phillips >>> (2008), who argues that no harm has come from the non-quantitative >>> approach in syntax research thus far. Phillips argues that there are >>> no cases in the literature where an incorrect intuitive judgment has >>> become the basis for a widely accepted generalization or an important >>> theoretical claim. He therefore concludes that there is no reason to >>> adopt more rigorous data collection standards. We challenge Philips’ >>> conclusion by presenting three cases from the literature where a >>> faulty intuition has led to incorrect generalizations and mistaken >>> theorizing, plausibly due to cognitive biases on the part of the >>> researchers. >>> >>> A second argument that is sometimes presented for the continued use of >>> the traditional non-quantitative method is that it would be too >>> inefficient to evaluate every syntactic / semantic hypothesis or >>> phenomenon quantitatively. For example, Culicover& Jackendoff (2010) >>> make this argument explicitly in their response to Gibson& Fedorenko >>> (2010): “It would cripple linguistic investigation if it were required >>> that all judgments of ambiguity and grammaticality be subject to >>> statistically rigorous experiments on naive subjects, especially when >>> investigating languages whose speakers are hard to access” (Culicover >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>> earlier in the discussion here.) Whereas we agree that in >>> circumstances where gathering data is difficult, some evidence is >>> better than no evidence, we do not agree that research would be slowed >>> with respect to languages where experimental participants are easy to >>> access, such as English. In contrast, we think that the opposite is >>> true: the field’s progress is probably slowed by not doing >>> quantitative research. >>> Suppose that a typical syntax / semantics paper that lacks >>> quantitative evidence includes judgments for 50 or more sentences / >>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>> most of the judgments from such a paper are correct or are on the >>> right track, the problem is in knowing which judgments are correct. >>> For example, suppose that 90% of the judgments from an arbitrary paper >>> are correct (which is probably a high estimate). (Colin Phillips and >>> some of his former students / postdocs have commented to us that, in >>> their experience, quantitative acceptability judgment studies almost >>> always validate the claim(s) in the literature. This is not our >>> experience, however. Most experiments that we have run which attempt >>> to test some syntactic / semantic hypothesis in the literature end up >>> providing us with a pattern of data that had not been known before the >>> experiment (e.g., Breen et al., in press; Fedorenko& Gibson, in >>> press; Patel et al., 2009; Scontras& Gibson, submitted).) This means >>> that in a paper with 50 empirical claims 45/50 are correct. But which >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over >>> two million different theories. By quantitatively evaluating the >>> empirical claims, we reduce the uncertainty a great deal. To make >>> progress, it is better to have theoretical claims supported by solid >>> quantitative evidence, so that even if the interpretation of the data >>> changes over time as new evidence becomes available – as is often the >>> case in any field of science – the empirical pattern can be used as a >>> basis for further theorizing. >>> >>> Furthermore, it is no longer expensive to run behavioral experiments, >>> at least in English and other widely spoken languages. There now >>> exists a marketplace interface – Amazon.com’s Mechanical Turk – which >>> can be used for collecting behavioral data over the internet quickly >>> and inexpensively. The cost of using an interface like this is >>> minimal, and the time that it takes for the results to be returned is >>> short. For example, currently on Mechanical Turk, a survey of >>> approximately 50 items will be answered by 50 or more participants >>> within a couple of hours, at a cost of approximately $1 per >>> participant. Thus a survey can be completed within a day, at a cost of >>> less than $50. (The hard work of designing the experiment, and >>> constructing controlled materials remains of course.) >>> >>> Sorry to be so verbose. But I think that these methodological points >>> are very important. >>> >>> Best wishes, >>> >>> Ted Gibson >>> >>> Gibson, E.& Fedorenko, E. (In press). The need for quantitative >>> methods in syntax and semantics research. Language and Cognitive >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>> & Fedorenko InPress LCP.pdf >>> >>> Gibson, E.& Fedorenko, E. (2010). Weak quantitative standards in >>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson& Fedorenko >>> 2010 TICS.pdf >>> >>> >>> >>> >>>> Dick, >>>> >>>> You raise an important issue here about methodology. I believe that >>>> intuitions are a fine way to generate hypotheses and even to test >>>> them - to a degree. But while it might not have been feasible for >>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>> Grammar to conduct experiments on every point of the grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, corpora, and standard psycholinguistic experimentation >>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>> field forward and for providing the best support for different >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>> paper on this, showing serious shortcomings with intuitions as the >>>> sole source of evidence, in their paper: "The need for quantitative >>>> methods in syntax and semantics research". >>>> >>>> Carson Schutze and Wayne Cowart, among others, have also written >>>> convincingly on this. >>>> >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>> Science), and researchers from Brazil are beginning a third round of >>>> experimental work among the Pirahas, since my own work on the syntax >>>> was, like almost every other field researcher's, based on native >>>> speaker intuitions and corpora. >>>> >>>> The discussion of methodologies reminds me of the initial reactions >>>> to Greenberg's work on classifying the languages of the Americas. His >>>> methods were strongly (and justifiably) criticized. However, I always >>>> thought that his methods were a great way of generating hypotheses, >>>> so long as they were ultimately put to the test of standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions. >>>> >>>> -- Dan >>> >>> >>>>> We linguists can add a further layer of explanation to the >>>>> judgements, but some judgements do seem to be more reliable than >>>>> others. And if we have to wait for psycholinguistic evidence for >>>>> every detailed analysis we make, our whole discipline will >>>>> immediately grind to a halt. Like it or not, native speaker >>>>> judgements are what put us linguists ahead of the rest in handling >>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>> Language (or the OED) without using native speaker judgements. >>>>> >>>>> Best wishes, Dick Hudson >>> >>> > > From dick at ling.ucl.ac.uk Fri Sep 10 23:40:06 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 00:40:06 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Ted and Ev, Yes, I understand your view, but I think it's a psycholinguist's view. Your goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality. And (as you know) I admire the way you pursue that goal. But my goal, as a linguist, is different. I want to explore the structure of a language so that I can understand how all the bits fit together. Like you, I'm aiming to model cognition, but my focus is on items and structures, and I start from the assumption that these can and do vary across speakers. However, having said all that I do agree with you that linguists should all get used to collecting and using quantitative data; and with the help of Brian MacWhinney's typology we'd know what methods to use when. And I do agree with your points about bid/bidded: in cases like that, quantitative data would be at least a very good starting point for a proper investigation. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 19:30, Ted Gibson wrote: > Dear Dick: > > Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are > saying is that if you have some testable claim involving a general > hypothesis about a language, then you need to get quantitative data > from unbiased sources to evaluate that claim. If you are interested in > English past tense morphology, then depending on the claims that you > might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very > low frequency words, you can run experiments to test behavior with > respect to such words. > > Your example of the past tense of "bid" is a fine such example. You > can run an experiment like the one you suggested to find out what > people think the past tense is. If you then found that 20/50 people > responded "bidded" and 30/50 respond "bid", that is a lot of useful > information. As you suggest in your discussion, this result wouldn't > answer the question of how past tense is stored in each individual. > This result would be ambiguous among several possible explanations. > One possibility is that the probability distribution that is being > discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution > in their heads, such that 2/5 of the time I respond one way, and 3/5 > of the time I respond another. Further experiments would be necessary > to answer between these and other possible theories (e.g., with > repeated trials from the same person, carefully planned so that the > participants don't notice that they are being asked multiple times). > Without the quantitative evidence in the first place, there is no way > to answer these kinds of questions. > > Regarding the past tense of "go", this would be useful as a baseline > in an experiment involving the less frequent ones. So, yes, it would > useful to gather quantitative evidence in such a case also, as > baselines with respect to the more interesting cases for theories. > > The bottom line: if you have a generalization about a language that > you wish to evaluate (such that you hypothesize that it is true across > the speakers of the language), then you need quantitative evidence > from multiple individuals, using an unbiased data collection method, > to evaluate such a claim. The point about Mechanical Turk is that it > is really *easy* to do this now, at least for languages like English. > > Best wishes, > > Ted Gibson & Ev Fedorenko > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> Dear Ted, >> Thanks for the very interesting comment, but are you REALLY saying >> that I shouldn't claim, for example, that the past tense of GO is >> "went" without first cross-checking with 50 native speakers? >> >> Isn't there a danger of missing the point that we all, as native >> speakers, spend our whole lives scanning other people's linguistic >> behaviour (language 'out there', E-language) and trying to explain it >> to ourselves in terms of a language system (language 'in here', >> I-language)? So every judgement we make is based on thousands or >> millions of observed exemplars, and reflects a unique experience of >> E-language filtered through a unique I-language. >> >> Given that view of language development, I don't see how quantitative >> data will help. Let's take a real uncertainty, such as the past tense >> of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My >> judgement: I don't know. Ok, you get 50 people to oblige on >> Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? >> Does that mean that the correct answer is "bidded"? Surely not. How >> is it better than my judgement? I agree you could record my speech >> and find how often I use each alternative; but the reason I don't >> know is precisely because it's a rare word, so in a sense >> quantitative data are irrelevant even there. What would solve the >> problem of subjectivity, of course, would be a machine for probing >> the bit of my mind (or even brain) that holds BID and its details; >> but I suspect that even that wouldn't move us much further forward >> than my original "don't know". (Incidentally I write as a fan of >> quantitative sociolinguistics, so I do accept that quantitative data >> are relevant to linguistic analysis in some areas, where the >> I-language phenomenon is frequent enough to produce usable data.) >> >> It seems to me that this discussion raises the really fundamental >> question of what kind of thing we think language is: social or >> individual. The problem isn't unique to linguistics of course; it's >> the same throughout the social sciences. But what's special about >> linguistics is that we deal in very fine details of culture (e.g. >> details of how a particular word is used or pronounced) so the >> differences between individuals really matter. I don't see that we're >> ever going to have anything better than judgements to go on, so what >> we need is a way to ensure that judgements are accurate reports of >> individual I-language. A rotten situation for a science, but I don't >> see how it can get better. >> >> Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 14:03, Ted Gibson wrote: >>> Dear Dan, Dick: >>> >>> I would like to clarify some points that Dan Everett makes, in >>> response to Dick Hudson. >>> >>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>> Fedorenko, 2010, in press, see references and links below) on what we >>> think are weak methodological standards in syntax and semantics >>> research over the past many years. The issue that we address is the >>> prevalent method in syntax and semantics research, which involves >>> obtaining a judgment of the acceptability of a sentence / meaning >>> pair, typically by just the author of the paper, sometimes with >>> feedback from colleagues. As we address in our papers, this >>> methodology does not allow proper testing of scientific hypotheses >>> because of (a) the small number of experimental participants >>> (typically one); (b) the small number of experimental stimuli >>> (typically one); (c) cognitive biases on the part of the researcher >>> and participants; and (d) the effect of the preceding context (e.g., >>> other constructions the researcher may have been recently >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>> several others cited in Gibson & Fedorenko, in press; for similar >>> points, but with not as strong a conclusion as ours). >>> >>> Three issues need to be separated here: (1) the use of intuitive >>> judgments as a dependent measure in a language experiment; (2) >>> potential cognitive biases on the part of experimental subjects and >>> experimenters in language experiments; and (3) the need for obtaining >>> quantitative evidence, whatever the dependent measure might be. The >>> paper that Ev and I wrote addresses the last two issues, but does not >>> go into depth on the first issue (the use of intuitions as a dependent >>> measure in language experiments). Regarding this issue, we don't think >>> that there is anything wrong with gathering intuitive judgments as a >>> dependent measure, as long as the task is clear to the experimental >>> participants. >>> >>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>> arguments that have been given in support of continuing to use the >>> traditional non-quantitative method in syntax / semantics research. >>> One recent defense of the traditional method comes from Phillips >>> (2008), who argues that no harm has come from the non-quantitative >>> approach in syntax research thus far. Phillips argues that there are >>> no cases in the literature where an incorrect intuitive judgment has >>> become the basis for a widely accepted generalization or an important >>> theoretical claim. He therefore concludes that there is no reason to >>> adopt more rigorous data collection standards. We challenge Philips’ >>> conclusion by presenting three cases from the literature where a >>> faulty intuition has led to incorrect generalizations and mistaken >>> theorizing, plausibly due to cognitive biases on the part of the >>> researchers. >>> >>> A second argument that is sometimes presented for the continued use of >>> the traditional non-quantitative method is that it would be too >>> inefficient to evaluate every syntactic / semantic hypothesis or >>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>> make this argument explicitly in their response to Gibson & Fedorenko >>> (2010): “It would cripple linguistic investigation if it were required >>> that all judgments of ambiguity and grammaticality be subject to >>> statistically rigorous experiments on naive subjects, especially when >>> investigating languages whose speakers are hard to access” (Culicover >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>> earlier in the discussion here.) Whereas we agree that in >>> circumstances where gathering data is difficult, some evidence is >>> better than no evidence, we do not agree that research would be slowed >>> with respect to languages where experimental participants are easy to >>> access, such as English. In contrast, we think that the opposite is >>> true: the field’s progress is probably slowed by not doing >>> quantitative research. >>> Suppose that a typical syntax / semantics paper that lacks >>> quantitative evidence includes judgments for 50 or more sentences / >>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>> most of the judgments from such a paper are correct or are on the >>> right track, the problem is in knowing which judgments are correct. >>> For example, suppose that 90% of the judgments from an arbitrary paper >>> are correct (which is probably a high estimate). (Colin Phillips and >>> some of his former students / postdocs have commented to us that, in >>> their experience, quantitative acceptability judgment studies almost >>> always validate the claim(s) in the literature. This is not our >>> experience, however. Most experiments that we have run which attempt >>> to test some syntactic / semantic hypothesis in the literature end up >>> providing us with a pattern of data that had not been known before the >>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>> that in a paper with 50 empirical claims 45/50 are correct. But which >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over >>> two million different theories. By quantitatively evaluating the >>> empirical claims, we reduce the uncertainty a great deal. To make >>> progress, it is better to have theoretical claims supported by solid >>> quantitative evidence, so that even if the interpretation of the data >>> changes over time as new evidence becomes available – as is often the >>> case in any field of science – the empirical pattern can be used as a >>> basis for further theorizing. >>> >>> Furthermore, it is no longer expensive to run behavioral experiments, >>> at least in English and other widely spoken languages. There now >>> exists a marketplace interface – Amazon.com’s Mechanical Turk – which >>> can be used for collecting behavioral data over the internet quickly >>> and inexpensively. The cost of using an interface like this is >>> minimal, and the time that it takes for the results to be returned is >>> short. For example, currently on Mechanical Turk, a survey of >>> approximately 50 items will be answered by 50 or more participants >>> within a couple of hours, at a cost of approximately $1 per >>> participant. Thus a survey can be completed within a day, at a cost of >>> less than $50. (The hard work of designing the experiment, and >>> constructing controlled materials remains of course.) >>> >>> Sorry to be so verbose. But I think that these methodological points >>> are very important. >>> >>> Best wishes, >>> >>> Ted Gibson >>> >>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>> methods in syntax and semantics research. Language and Cognitive >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>> & Fedorenko InPress LCP.pdf >>> >>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>> 2010 TICS.pdf >>> >>> >>> >>> >>>> Dick, >>>> >>>> You raise an important issue here about methodology. I believe that >>>> intuitions are a fine way to generate hypotheses and even to test >>>> them - to a degree. But while it might not have been feasible for >>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>> Grammar to conduct experiments on every point of the grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, corpora, and standard psycholinguistic experimentation >>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>> field forward and for providing the best support for different >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>> paper on this, showing serious shortcomings with intuitions as the >>>> sole source of evidence, in their paper: "The need for quantitative >>>> methods in syntax and semantics research". >>>> >>>> Carson Schutze and Wayne Cowart, among others, have also written >>>> convincingly on this. >>>> >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>> Science), and researchers from Brazil are beginning a third round of >>>> experimental work among the Pirahas, since my own work on the syntax >>>> was, like almost every other field researcher's, based on native >>>> speaker intuitions and corpora. >>>> >>>> The discussion of methodologies reminds me of the initial reactions >>>> to Greenberg's work on classifying the languages of the Americas. His >>>> methods were strongly (and justifiably) criticized. However, I always >>>> thought that his methods were a great way of generating hypotheses, >>>> so long as they were ultimately put to the test of standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions. >>>> >>>> -- Dan >>> >>> >>> >>>>> We linguists can add a further layer of explanation to the >>>>> judgements, but some judgements do seem to be more reliable than >>>>> others. And if we have to wait for psycholinguistic evidence for >>>>> every detailed analysis we make, our whole discipline will >>>>> immediately grind to a halt. Like it or not, native speaker >>>>> judgements are what put us linguists ahead of the rest in handling >>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>> Language (or the OED) without using native speaker judgements. >>>>> >>>>> Best wishes, Dick Hudson >>> >>> >>> > > > From Lise.Menn at Colorado.EDU Sat Sep 11 00:40:47 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Fri, 10 Sep 2010 18:40:47 -0600 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284130304@cast-dryerm2.caset.buffalo.edu> Message-ID: Matt, I have to disagree with you on the validity of describing what's 'out there' (what Dick Hudson says is his interest, in his contribution of 5:40:06 PM MDT today). We DO have to account for it in order to understand how 'the language in speakers' heads' gets into those heads in the first place. In more detail: Each of us is immersed from (before) birth in a sampling of utterances (and if we are literate, eventually also written forms of the language). In order to understand how we really create our internal representations of our language, we have to know (or be able to estimate) something about the data our brains get as input. There are at least better and worse descriptions of the patterns in those data, and certainly there are wrong ones, though in many cases - for example in the 'unhappiness' case - there are probably conflicting right ones, rather than any single correct one. (OT offers some help in thinking about this.) To take a concrete example, in order to account for the still- unstable changes in English pronominal case marking in compound NP objects of prepositions from a system based on syntactic case (He gave the cookies to Mary and me) to a system apparently based partly on whether the pronoun is next to the governing preposition (He gave the cookies to Mary and I/ to me and Mary), you first have to do an analysis of usage and figure out what the pattern is. And usage is not in our heads (although it's the result of what's in our heads), it's 'out there'. Even fossils and obscure patterns contribute to the redundancy of the language, making it more learnable and and helping to create the resonances used by great poets and orators. (I admit to having oversimplified in speaking as if there were always one 'correct' analysis of the patterns 'out there' that might be (subconsciously) discoverable by speakers. That's not true.) And because not all speakers are equally sensitive to language patterns - again, the Gleitman and Gleitman book is a terrific example - it's also an oversimplification to talk about 'what is in speaker's heads' as if the same thing is in everyone's head. (K.P. Mohanan has also published on this.) At the lexical level, Danielle Cyr's examples (September 9, 2010 8:38:59 PM MDT) further remind us that what's inside each person's head changes over time. So we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover. It's not - it's more like a complex mosaic that does not fit together perfectly. Lise On Sep 10, 2010, at 12:51 PM, dryer at buffalo.edu wrote: > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have > to be quite careful in making clear what we are referring to when > we talk about 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the > correct analysis", one in terms of what is in people's heads, the > other in terms of what is "out there". > > There are a vast number of patterns "out there". The only > distinction, on my view, amongst this vast number of patterns, is > between those that reflect something inside people's heads and those > that don't. But if that is the case, then there is no coherent > sense in which one can talk of "the correct analysis" of what is > "out there", except in terms of what is in people's heads, and thus > no second sense of "the correct analysis". The patterns that don't > correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of > stars and, as with constellations, there is no reality to these > patterns, except in the minds of linguists. And there are those > patterns which are the fossil remains of what was in the heads of > speakers of an earlier stage of the language but which no longer > are. These latter patterns are real, and they are relevant to > exlaining why the language is now the way it is, but they are not > relevant, I think many would agree, as to what is the "correct > analysis" of the language today. > > For this reason, I claim that the only sense in which an analysis > can be "the correct analysis" is in terms of what is inside of > people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion > of these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > >> I wish we had better terminology for keeping track of whether, at a >> given time, we are talking about the patterns that are 'out there' in >> the language and might possibly be apprehended (subconsciously) by a >> speaker, and when we are talking about the patterns that a particular >> speaker actually does apprehend, as indicated by experiments, from >> simple 'wug tests' up to brain wave and eye-gaze studies. And for >> distinguishing among the degrees of pattern apprehension that a >> person >> may have, from vague preferences detectable in reaction times or >> other >> behavior all the way up through clear metalinguistic insights. Dick >> Hudson's note reminding us of the Gleitman and Gleitman study is >> right >> on target. >> >> Since we don't in fact have such an agreed-on terminology, we have >> to >> be quite careful in making clear what we are referring to when we >> talk >> about 'the correct analysis' of a form like 'unhappiness'. We >> know, but >> tend to forget - and tend to forget to tell our students! - that >> it's an >> empirical question as to whether the formal simplicity and >> coherence of >> description of forms 'out there' (e.g. lovely abstract >> morphophonemics) >> is any kind of approximation to the way knowledge of the same >> forms is >> organized in a particular person's head. If we remember that a very >> large proportion of what we know about our language is 'out there' >> when >> we are infants and has to be internalized through experience with >> the >> language (even if you believe in innate 'core language'), the >> variation >> in internal knowledge from one person to another is more >> understandable. >> >> We especially need to consider (and try to test) the possibility >> that >> since >> the brain can make multiple cross-connections, multiple patterns are >> involved >> simultaneously in morphological and syntactic analyses. I suggest >> that >> that's the case with 'unhappiness' - and the linguistic analyses >> that I >> know about are not good at handling that kind of idea. >> >> Lise >> >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >> >>> >>> Two comments. >>> >>> First (elaborating perhaps on Dick Hudson's comment), I think there >>> is an >>> important distinction between low-level linguistic intuitions (like >>> whether a >>> word or sentence is well-formed or what it means) and higher-level >>> intuitions >>> (like what the structure of a word or sentence is). One can take >>> the position >>> that we need to account for the former (while recognizing that they >>> are not >>> always reliable) but not the latter. >>> >>> Second, the tension here is not only between evidence from speaker >>> intuitions >>> versus evidence from psycholinguistic experiments. There is also a >>> tension >>> between deciding on the correct analysis on the basis of a priori >>> simplicity >>> arguments versus deciding on the correct analysis on the basis of >>> psycholinguistic evidence (see Derwing 1973). The bracketing >>> paradox that Dan >>> referred to that arises with the word (semantics argues >>> for >>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>> +er]] (the >>> comparative suffix can only be attached to adjectives containing one >>> or two >>> syllables) is only a paradox if one assumes that speakers adopt the >>> simplest >>> analysis. For example, if speakers adopt a more complex rule for >>> either of these >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>> trisyllabic >>> words beginning with un-), then the bracketing paradox disappears. >>> >>> Matthew >>> >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree >>>> that >>>> conscious judgements are only one kind of evidence that we need to >>>> take >>>> into account. I admire Carson Schutze's work (which I reviewed in >>>> fact), >>>> and of course I've been aware of complaints about judgements by >>>> people >>>> like Labov for decades. >>>> >>>> But you're missing my main point, which is that all judgements >>>> aren't >>>> equally reliable. If you want to know how /unhappiness/ is >>>> structured, >>>> ask a linguist, not a five-year old. And one of the by-products of >>>> education may be increased sensitivity to syntax - which is one of >>>> the >>>> many reasons why linguists need to pay more attention to education. >>>> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>> Dick, >>>>> >>>>> You raise an important issue here about >>>> methodology. I believe that intuitions are a fine way to generate >>>> hypotheses and even to test them - to a degree. But while it might >>>> not have >>>> been feasible for Huddleston, Pullum, and the other contributors to >>>> the >>>> Cambridge Grammar to conduct experiments on every point of the >>>> grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, >>>> corpora, and standard psycholinguistic experimentation (indeed, >>>> Standard >>>> Social Science Methodology) is vital for taking the field forward >>>> and for >>>> providing the best support for different analyses. Ted Gibson and >>>> Ev >>>> Fedorenko have written a very useful new paper on this, showing >>>> serious >>>> shortcomings with intuitions as the sole source of evidence, in >>>> their >>>> paper: "The need for quantitative methods in syntax and semantics >>>> research".> >>>>> Carson Schutze and Wayne Cowart, among others, >>>> have also written convincingly on this.> >>>>> It is one reason that a team from Stanford, MIT >>>> (Brain and Cognitive Science), and researchers from Brazil are >>>> beginning a >>>> third round of experimental work among the Pirahas, since my own >>>> work on >>>> the syntax was, like almost every other field researcher's, based >>>> on native >>>> speaker intuitions and corpora.> >>>>> The discussion of methodologies reminds me of >>>> the initial reactions to Greenberg's work on classifying the >>>> languages of >>>> the Americas. His methods were strongly (and justifiably) >>>> criticized. >>>> However, I always thought that his methods were a great way of >>>> generating >>>> hypotheses, so long as they were ultimately put to the test of >>>> standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions.> >>>>> -- Dan >>>>> >>>>>> We linguists can add a further layer of >>>> explanation to the judgements, but some judgements do seem to be >>>> more >>>> reliable than others. And if we have to wait for psycholinguistic >>>> evidence >>>> for every detailed analysis we make, our whole discipline will >>>> immediately >>>> grind to a halt. Like it or not, native speaker judgements are what >>>> put us >>>> linguists ahead of the rest in handling fine detail. Imagine >>>> writing the >>>> Cambridge Grammar of the English Language (or the OED) without >>>> using native >>>> speaker judgements.>> >>>>>> Best wishes, Dick Hudson >>>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> >>>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> >> Campus Mail Address: >> UCB 594, Institute of Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> >> > > > > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From Lise.Menn at Colorado.EDU Sat Sep 11 01:03:41 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Fri, 10 Sep 2010 19:03:41 -0600 Subject: analysis: unhappiness In-Reply-To: <4C8AC1D6.3010203@ling.ucl.ac.uk> Message-ID: Dick, I think that what you say is true, but it is only (I hope) a temporary description of the state of the art of psycholinguistics. At least some of us would like our methods to become sensitive enough to individual differences so that we can look at how 'the general processes and principles' interact with the level of an individual person's knowledge of particular constructions, to find out how much each person knows of the patterns 'out there' in the language. Some experimental methods are almost at that point already; they can distinguish degrees of mastery of particular constructions of a language among groups of second-language learners. Have a look at Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. 2002. Overhearing a language during childhood. Psychological Science 13.3, 238-243. Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on to childhood language memory. Cognition, 86(3), B53-B64. Lise On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's > view. Your goal is to find general processes and principles that > apply uniformly across individuals, so you have to use methods to > check for generality. And (as you know) I admire the way you pursue > that goal. But my goal, as a linguist, is different. I want to > explore the structure of a language so that I can understand how all > the bits fit together. Like you, I'm aiming to model cognition, but > my focus is on items and structures, and I start from the assumption > that these can and do vary across speakers. > > However, having said all that I do agree with you that linguists > should all get used to collecting and using quantitative data; and > with the help of Brian MacWhinney's typology we'd know what methods > to use when. And I do agree with your points about bid/bidded: in > cases like that, quantitative data would be at least a very good > starting point for a proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: >> Dear Dick: >> >> Perhaps we are talking at cross purposes. I don't understand what >> is confusing about what Ev Fedorenko and I are claiming. All we are >> saying is that if you have some testable claim involving a general >> hypothesis about a language, then you need to get quantitative data >> from unbiased sources to evaluate that claim. If you are interested >> in English past tense morphology, then depending on the claims that >> you might want to investigate, there are lots of ways to get >> relevant quantitative evidence. Corpus data will probably be >> useful. For very low frequency words, you can run experiments to >> test behavior with respect to such words. >> >> Your example of the past tense of "bid" is a fine such example. You >> can run an experiment like the one you suggested to find out what >> people think the past tense is. If you then found that 20/50 people >> responded "bidded" and 30/50 respond "bid", that is a lot of useful >> information. As you suggest in your discussion, this result >> wouldn't answer the question of how past tense is stored in each >> individual. This result would be ambiguous among several possible >> explanations. One possibility is that the probability distribution >> that is being discovered reflects different dialects, such that 2/5 >> of the population has one past tense, and 3/5 has another. Another >> possibility is that each person has a similar probability >> distribution in their heads, such that 2/5 of the time I respond >> one way, and 3/5 of the time I respond another. Further experiments >> would be necessary to answer between these and other possible >> theories (e.g., with repeated trials from the same person, >> carefully planned so that the participants don't notice that they >> are being asked multiple times). Without the quantitative evidence >> in the first place, there is no way to answer these kinds of >> questions. >> >> Regarding the past tense of "go", this would be useful as a >> baseline in an experiment involving the less frequent ones. So, >> yes, it would useful to gather quantitative evidence in such a case >> also, as baselines with respect to the more interesting cases for >> theories. >> >> The bottom line: if you have a generalization about a language that >> you wish to evaluate (such that you hypothesize that it is true >> across the speakers of the language), then you need quantitative >> evidence from multiple individuals, using an unbiased data >> collection method, to evaluate such a claim. The point about >> Mechanical Turk is that it is really *easy* to do this now, at >> least for languages like English. >> >> Best wishes, >> >> Ted Gibson & Ev Fedorenko >> >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >> >>> Dear Ted, >>> Thanks for the very interesting comment, but are you REALLY saying >>> that I shouldn't claim, for example, that the past tense of GO is >>> "went" without first cross-checking with 50 native speakers? >>> >>> Isn't there a danger of missing the point that we all, as native >>> speakers, spend our whole lives scanning other people's linguistic >>> behaviour (language 'out there', E-language) and trying to explain >>> it to ourselves in terms of a language system (language 'in here', >>> I-language)? So every judgement we make is based on thousands or >>> millions of observed exemplars, and reflects a unique experience >>> of E-language filtered through a unique I-language. >>> >>> Given that view of language development, I don't see how >>> quantitative data will help. Let's take a real uncertainty, such >>> as the past tense of BID. If I want to say I did it, do I say "I >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>> and 30 "bid". So what? Does that mean that the correct answer is >>> "bidded"? Surely not. How is it better than my judgement? I agree >>> you could record my speech and find how often I use each >>> alternative; but the reason I don't know is precisely because it's >>> a rare word, so in a sense quantitative data are irrelevant even >>> there. What would solve the problem of subjectivity, of course, >>> would be a machine for probing the bit of my mind (or even brain) >>> that holds BID and its details; but I suspect that even that >>> wouldn't move us much further forward than my original "don't >>> know". (Incidentally I write as a fan of quantitative >>> sociolinguistics, so I do accept that quantitative data are >>> relevant to linguistic analysis in some areas, where the I- >>> language phenomenon is frequent enough to produce usable data.) >>> >>> It seems to me that this discussion raises the really fundamental >>> question of what kind of thing we think language is: social or >>> individual. The problem isn't unique to linguistics of course; >>> it's the same throughout the social sciences. But what's special >>> about linguistics is that we deal in very fine details of culture >>> (e.g. details of how a particular word is used or pronounced) so >>> the differences between individuals really matter. I don't see >>> that we're ever going to have anything better than judgements to >>> go on, so what we need is a way to ensure that judgements are >>> accurate reports of individual I-language. A rotten situation for >>> a science, but I don't see how it can get better. >>> >>> Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> >>> On 10/09/2010 14:03, Ted Gibson wrote: >>>> Dear Dan, Dick: >>>> >>>> I would like to clarify some points that Dan Everett makes, in >>>> response to Dick Hudson. >>>> >>>> Ev Fedorenko and I have written a couple of papers recently >>>> (Gibson & >>>> Fedorenko, 2010, in press, see references and links below) on >>>> what we >>>> think are weak methodological standards in syntax and semantics >>>> research over the past many years. The issue that we address is the >>>> prevalent method in syntax and semantics research, which involves >>>> obtaining a judgment of the acceptability of a sentence / meaning >>>> pair, typically by just the author of the paper, sometimes with >>>> feedback from colleagues. As we address in our papers, this >>>> methodology does not allow proper testing of scientific hypotheses >>>> because of (a) the small number of experimental participants >>>> (typically one); (b) the small number of experimental stimuli >>>> (typically one); (c) cognitive biases on the part of the researcher >>>> and participants; and (d) the effect of the preceding context >>>> (e.g., >>>> other constructions the researcher may have been recently >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>> several others cited in Gibson & Fedorenko, in press; for similar >>>> points, but with not as strong a conclusion as ours). >>>> >>>> Three issues need to be separated here: (1) the use of intuitive >>>> judgments as a dependent measure in a language experiment; (2) >>>> potential cognitive biases on the part of experimental subjects and >>>> experimenters in language experiments; and (3) the need for >>>> obtaining >>>> quantitative evidence, whatever the dependent measure might be. The >>>> paper that Ev and I wrote addresses the last two issues, but does >>>> not >>>> go into depth on the first issue (the use of intuitions as a >>>> dependent >>>> measure in language experiments). Regarding this issue, we don't >>>> think >>>> that there is anything wrong with gathering intuitive judgments >>>> as a >>>> dependent measure, as long as the task is clear to the experimental >>>> participants. >>>> >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to >>>> some >>>> arguments that have been given in support of continuing to use the >>>> traditional non-quantitative method in syntax / semantics research. >>>> One recent defense of the traditional method comes from Phillips >>>> (2008), who argues that no harm has come from the non-quantitative >>>> approach in syntax research thus far. Phillips argues that there >>>> are >>>> no cases in the literature where an incorrect intuitive judgment >>>> has >>>> become the basis for a widely accepted generalization or an >>>> important >>>> theoretical claim. He therefore concludes that there is no reason >>>> to >>>> adopt more rigorous data collection standards. We challenge >>>> Philips’ >>>> conclusion by presenting three cases from the literature where a >>>> faulty intuition has led to incorrect generalizations and mistaken >>>> theorizing, plausibly due to cognitive biases on the part of the >>>> researchers. >>>> >>>> A second argument that is sometimes presented for the continued >>>> use of >>>> the traditional non-quantitative method is that it would be too >>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>> phenomenon quantitatively. For example, Culicover & Jackendoff >>>> (2010) >>>> make this argument explicitly in their response to Gibson & >>>> Fedorenko >>>> (2010): “It would cripple linguistic investigation if it were >>>> required >>>> that all judgments of ambiguity and grammaticality be subject to >>>> statistically rigorous experiments on naive subjects, especially >>>> when >>>> investigating languages whose speakers are hard to >>>> access” (Culicover >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>> earlier in the discussion here.) Whereas we agree that in >>>> circumstances where gathering data is difficult, some evidence is >>>> better than no evidence, we do not agree that research would be >>>> slowed >>>> with respect to languages where experimental participants are >>>> easy to >>>> access, such as English. In contrast, we think that the opposite is >>>> true: the field’s progress is probably slowed by not doing >>>> quantitative research. >>>> Suppose that a typical syntax / semantics paper that lacks >>>> quantitative evidence includes judgments for 50 or more sentences / >>>> meaning pairs, corresponding to 50 or more empirical claims. Even >>>> if >>>> most of the judgments from such a paper are correct or are on the >>>> right track, the problem is in knowing which judgments are correct. >>>> For example, suppose that 90% of the judgments from an arbitrary >>>> paper >>>> are correct (which is probably a high estimate). (Colin Phillips >>>> and >>>> some of his former students / postdocs have commented to us that, >>>> in >>>> their experience, quantitative acceptability judgment studies >>>> almost >>>> always validate the claim(s) in the literature. This is not our >>>> experience, however. Most experiments that we have run which >>>> attempt >>>> to test some syntactic / semantic hypothesis in the literature >>>> end up >>>> providing us with a pattern of data that had not been known >>>> before the >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This >>>> means >>>> that in a paper with 50 empirical claims 45/50 are correct. But >>>> which >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s >>>> over >>>> two million different theories. By quantitatively evaluating the >>>> empirical claims, we reduce the uncertainty a great deal. To make >>>> progress, it is better to have theoretical claims supported by >>>> solid >>>> quantitative evidence, so that even if the interpretation of the >>>> data >>>> changes over time as new evidence becomes available – as is often >>>> the >>>> case in any field of science – the empirical pattern can be used >>>> as a >>>> basis for further theorizing. >>>> >>>> Furthermore, it is no longer expensive to run behavioral >>>> experiments, >>>> at least in English and other widely spoken languages. There now >>>> exists a marketplace interface – Amazon.com’s Mechanical Turk – >>>> which >>>> can be used for collecting behavioral data over the internet >>>> quickly >>>> and inexpensively. The cost of using an interface like this is >>>> minimal, and the time that it takes for the results to be >>>> returned is >>>> short. For example, currently on Mechanical Turk, a survey of >>>> approximately 50 items will be answered by 50 or more participants >>>> within a couple of hours, at a cost of approximately $1 per >>>> participant. Thus a survey can be completed within a day, at a >>>> cost of >>>> less than $50. (The hard work of designing the experiment, and >>>> constructing controlled materials remains of course.) >>>> >>>> Sorry to be so verbose. But I think that these methodological >>>> points >>>> are very important. >>>> >>>> Best wishes, >>>> >>>> Ted Gibson >>>> >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>> methods in syntax and semantics research. Language and Cognitive >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>> & Fedorenko InPress LCP.pdf >>>> >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >>>> Fedorenko >>>> 2010 TICS.pdf >>>> >>>> >>>> >>>> >>>>> Dick, >>>>> >>>>> You raise an important issue here about methodology. I believe >>>>> that >>>>> intuitions are a fine way to generate hypotheses and even to test >>>>> them - to a degree. But while it might not have been feasible for >>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>> Grammar to conduct experiments on every point of the grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>> (indeed, Standard Social Science Methodology) is vital for >>>>> taking the >>>>> field forward and for providing the best support for different >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful >>>>> new >>>>> paper on this, showing serious shortcomings with intuitions as the >>>>> sole source of evidence, in their paper: "The need for >>>>> quantitative >>>>> methods in syntax and semantics research". >>>>> >>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>> convincingly on this. >>>>> >>>>> It is one reason that a team from Stanford, MIT (Brain and >>>>> Cognitive >>>>> Science), and researchers from Brazil are beginning a third >>>>> round of >>>>> experimental work among the Pirahas, since my own work on the >>>>> syntax >>>>> was, like almost every other field researcher's, based on native >>>>> speaker intuitions and corpora. >>>>> >>>>> The discussion of methodologies reminds me of the initial >>>>> reactions >>>>> to Greenberg's work on classifying the languages of the >>>>> Americas. His >>>>> methods were strongly (and justifiably) criticized. However, I >>>>> always >>>>> thought that his methods were a great way of generating >>>>> hypotheses, >>>>> so long as they were ultimately put to the test of standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions. >>>>> >>>>> -- Dan >>>> >>>> >>>> >>>>>> We linguists can add a further layer of explanation to the >>>>>> judgements, but some judgements do seem to be more reliable than >>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>> every detailed analysis we make, our whole discipline will >>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>> judgements are what put us linguists ahead of the rest in >>>>>> handling >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>> Language (or the OED) without using native speaker judgements. >>>>>> >>>>>> Best wishes, Dick Hudson >>>> >>>> >>>> >> >> >> Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From dan at daneverett.org Sat Sep 11 01:05:18 2010 From: dan at daneverett.org (Daniel Everett) Date: Fri, 10 Sep 2010 21:05:18 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8AC1D6.3010203@ling.ucl.ac.uk> Message-ID: I think that Brian and Dick make excellent points. There are very good grammars written that could be improved by psycholinguistic experimentation and more quantitative approaches. But large sections of those grammars aren't going to change one bit (go-went) with quantitative tests and such tests would be completely counterproductive given the shortness of life and the vastness of the field linguist's tasks. Part of the problem is that linguistics is not simply a subdiscipline of psychology. Linguistics has its own objectives and those only occasionally overlap with psychology. The same for methods. On another note, I don't buy the 'in my head' 'out of my head' distinction either (that Matt seems to be urging upon us). We study different things and have different reasons for being satisfied with the results we achieve. I believe that we linguists are often complacent and fail to apply better methods. But of course that applies to all disciplines. In the meantime, checking corpora, collecting data as a result of careful interviews with native speakers, and the other aspects of the field linguist's task are vital parts of the linguist's task and much of this won't be improved by quantitative methods as we currently understand them. Maybe sometime. Dan P.S. In my original reference to Ted and Ev's paper, I said that they showed the danger of using intuitions. What I meant to say of using intuitions as standardly used by linguists. They convinced me that there is a lot to learn from quantitative methods. On 10 Sep 2010, at 19:40, Richard Hudson wrote: > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's view. Your goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality. And (as you know) I admire the way you pursue that goal. But my goal, as a linguist, is different. I want to explore the structure of a language so that I can understand how all the bits fit together. Like you, I'm aiming to model cognition, but my focus is on items and structures, and I start from the assumption that these can and do vary across speakers. > > However, having said all that I do agree with you that linguists should all get used to collecting and using quantitative data; and with the help of Brian MacWhinney's typology we'd know what methods to use when. And I do agree with your points about bid/bidded: in cases like that, quantitative data would be at least a very good starting point for a proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: >> Dear Dick: >> >> Perhaps we are talking at cross purposes. I don't understand what is confusing about what Ev Fedorenko and I are claiming. All we are saying is that if you have some testable claim involving a general hypothesis about a language, then you need to get quantitative data from unbiased sources to evaluate that claim. If you are interested in English past tense morphology, then depending on the claims that you might want to investigate, there are lots of ways to get relevant quantitative evidence. Corpus data will probably be useful. For very low frequency words, you can run experiments to test behavior with respect to such words. >> >> Your example of the past tense of "bid" is a fine such example. You can run an experiment like the one you suggested to find out what people think the past tense is. If you then found that 20/50 people responded "bidded" and 30/50 respond "bid", that is a lot of useful information. As you suggest in your discussion, this result wouldn't answer the question of how past tense is stored in each individual. This result would be ambiguous among several possible explanations. One possibility is that the probability distribution that is being discovered reflects different dialects, such that 2/5 of the population has one past tense, and 3/5 has another. Another possibility is that each person has a similar probability distribution in their heads, such that 2/5 of the time I respond one way, and 3/5 of the time I respond another. Further experiments would be necessary to answer between these and other possible theories (e.g., with repeated trials from the same person, carefully planned so that the participants don't notice that they are being asked multiple times). Without the quantitative evidence in the first place, there is no way to answer these kinds of questions. >> >> Regarding the past tense of "go", this would be useful as a baseline in an experiment involving the less frequent ones. So, yes, it would useful to gather quantitative evidence in such a case also, as baselines with respect to the more interesting cases for theories. >> >> The bottom line: if you have a generalization about a language that you wish to evaluate (such that you hypothesize that it is true across the speakers of the language), then you need quantitative evidence from multiple individuals, using an unbiased data collection method, to evaluate such a claim. The point about Mechanical Turk is that it is really *easy* to do this now, at least for languages like English. >> >> Best wishes, >> >> Ted Gibson & Ev Fedorenko >> >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >> >>> Dear Ted, >>> Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? >>> >>> Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. >>> >>> Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) >>> >>> It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. >>> >>> Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> >>> On 10/09/2010 14:03, Ted Gibson wrote: >>>> Dear Dan, Dick: >>>> >>>> I would like to clarify some points that Dan Everett makes, in >>>> response to Dick Hudson. >>>> >>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>>> Fedorenko, 2010, in press, see references and links below) on what we >>>> think are weak methodological standards in syntax and semantics >>>> research over the past many years. The issue that we address is the >>>> prevalent method in syntax and semantics research, which involves >>>> obtaining a judgment of the acceptability of a sentence / meaning >>>> pair, typically by just the author of the paper, sometimes with >>>> feedback from colleagues. As we address in our papers, this >>>> methodology does not allow proper testing of scientific hypotheses >>>> because of (a) the small number of experimental participants >>>> (typically one); (b) the small number of experimental stimuli >>>> (typically one); (c) cognitive biases on the part of the researcher >>>> and participants; and (d) the effect of the preceding context (e.g., >>>> other constructions the researcher may have been recently >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>> several others cited in Gibson & Fedorenko, in press; for similar >>>> points, but with not as strong a conclusion as ours). >>>> >>>> Three issues need to be separated here: (1) the use of intuitive >>>> judgments as a dependent measure in a language experiment; (2) >>>> potential cognitive biases on the part of experimental subjects and >>>> experimenters in language experiments; and (3) the need for obtaining >>>> quantitative evidence, whatever the dependent measure might be. The >>>> paper that Ev and I wrote addresses the last two issues, but does not >>>> go into depth on the first issue (the use of intuitions as a dependent >>>> measure in language experiments). Regarding this issue, we don't think >>>> that there is anything wrong with gathering intuitive judgments as a >>>> dependent measure, as long as the task is clear to the experimental >>>> participants. >>>> >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>>> arguments that have been given in support of continuing to use the >>>> traditional non-quantitative method in syntax / semantics research. >>>> One recent defense of the traditional method comes from Phillips >>>> (2008), who argues that no harm has come from the non-quantitative >>>> approach in syntax research thus far. Phillips argues that there are >>>> no cases in the literature where an incorrect intuitive judgment has >>>> become the basis for a widely accepted generalization or an important >>>> theoretical claim. He therefore concludes that there is no reason to >>>> adopt more rigorous data collection standards. We challenge Philips’ >>>> conclusion by presenting three cases from the literature where a >>>> faulty intuition has led to incorrect generalizations and mistaken >>>> theorizing, plausibly due to cognitive biases on the part of the >>>> researchers. >>>> >>>> A second argument that is sometimes presented for the continued use of >>>> the traditional non-quantitative method is that it would be too >>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>>> make this argument explicitly in their response to Gibson & Fedorenko >>>> (2010): “It would cripple linguistic investigation if it were required >>>> that all judgments of ambiguity and grammaticality be subject to >>>> statistically rigorous experiments on naive subjects, especially when >>>> investigating languages whose speakers are hard to access” (Culicover >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>> earlier in the discussion here.) Whereas we agree that in >>>> circumstances where gathering data is difficult, some evidence is >>>> better than no evidence, we do not agree that research would be slowed >>>> with respect to languages where experimental participants are easy to >>>> access, such as English. In contrast, we think that the opposite is >>>> true: the field’s progress is probably slowed by not doing >>>> quantitative research. >>>> Suppose that a typical syntax / semantics paper that lacks >>>> quantitative evidence includes judgments for 50 or more sentences / >>>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>>> most of the judgments from such a paper are correct or are on the >>>> right track, the problem is in knowing which judgments are correct. >>>> For example, suppose that 90% of the judgments from an arbitrary paper >>>> are correct (which is probably a high estimate). (Colin Phillips and >>>> some of his former students / postdocs have commented to us that, in >>>> their experience, quantitative acceptability judgment studies almost >>>> always validate the claim(s) in the literature. This is not our >>>> experience, however. Most experiments that we have run which attempt >>>> to test some syntactic / semantic hypothesis in the literature end up >>>> providing us with a pattern of data that had not been known before the >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>>> that in a paper with 50 empirical claims 45/50 are correct. But which >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s over >>>> two million different theories. By quantitatively evaluating the >>>> empirical claims, we reduce the uncertainty a great deal. To make >>>> progress, it is better to have theoretical claims supported by solid >>>> quantitative evidence, so that even if the interpretation of the data >>>> changes over time as new evidence becomes available – as is often the >>>> case in any field of science – the empirical pattern can be used as a >>>> basis for further theorizing. >>>> >>>> Furthermore, it is no longer expensive to run behavioral experiments, >>>> at least in English and other widely spoken languages. There now >>>> exists a marketplace interface – Amazon.com’s Mechanical Turk – which >>>> can be used for collecting behavioral data over the internet quickly >>>> and inexpensively. The cost of using an interface like this is >>>> minimal, and the time that it takes for the results to be returned is >>>> short. For example, currently on Mechanical Turk, a survey of >>>> approximately 50 items will be answered by 50 or more participants >>>> within a couple of hours, at a cost of approximately $1 per >>>> participant. Thus a survey can be completed within a day, at a cost of >>>> less than $50. (The hard work of designing the experiment, and >>>> constructing controlled materials remains of course.) >>>> >>>> Sorry to be so verbose. But I think that these methodological points >>>> are very important. >>>> >>>> Best wishes, >>>> >>>> Ted Gibson >>>> >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>> methods in syntax and semantics research. Language and Cognitive >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>> & Fedorenko InPress LCP.pdf >>>> >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>>> 2010 TICS.pdf >>>> >>>> >>>> >>>> >>>>> Dick, >>>>> >>>>> You raise an important issue here about methodology. I believe that >>>>> intuitions are a fine way to generate hypotheses and even to test >>>>> them - to a degree. But while it might not have been feasible for >>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>> Grammar to conduct experiments on every point of the grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>>> field forward and for providing the best support for different >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>>> paper on this, showing serious shortcomings with intuitions as the >>>>> sole source of evidence, in their paper: "The need for quantitative >>>>> methods in syntax and semantics research". >>>>> >>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>> convincingly on this. >>>>> >>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>>> Science), and researchers from Brazil are beginning a third round of >>>>> experimental work among the Pirahas, since my own work on the syntax >>>>> was, like almost every other field researcher's, based on native >>>>> speaker intuitions and corpora. >>>>> >>>>> The discussion of methodologies reminds me of the initial reactions >>>>> to Greenberg's work on classifying the languages of the Americas. His >>>>> methods were strongly (and justifiably) criticized. However, I always >>>>> thought that his methods were a great way of generating hypotheses, >>>>> so long as they were ultimately put to the test of standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions. >>>>> >>>>> -- Dan >>>> >>>> >>>> >>>>>> We linguists can add a further layer of explanation to the >>>>>> judgements, but some judgements do seem to be more reliable than >>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>> every detailed analysis we make, our whole discipline will >>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>> judgements are what put us linguists ahead of the rest in handling >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>> Language (or the OED) without using native speaker judgements. >>>>>> >>>>>> Best wishes, Dick Hudson >>>> >>>> >>>> >> >> >> > From phdebrab at yahoo.co.uk Sat Sep 11 10:03:35 2010 From: phdebrab at yahoo.co.uk (Philippe De Brabanter) Date: Sat, 11 Sep 2010 10:03:35 +0000 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Dear all, this is just to say that the Cambridge Grammar of the English Language, like several other grammars I'm aware of, does mention the sorts of constructions Tom is wondering about (esp. pp. 964-67). They're treated as noun complements, whereas relative clauses usually function as modifiers of nouns. H&P give a useful list of nouns licensing these complements (a list which confirms Suzanne Kemmerer's point that these nouns do not always have a verbal counterpart taking a content clause as its complement — H&P suggest that the most frequent of those licensing nouns is fact). They also point out that content clauses can also sometimes function as supplements (i.e. appositives), as in I'm inclined to favour your first suggestion, that we shelve the proposal until after the election. This confirms Suzanne's suggestion that we shouldn't say that the clausal noun complements are appositives. One last interesting point. On p. 967, H&P show that the licensor may sometimes be more than just a noun, with certain constructions like have + licensing NP or existential there + be facilitating (or being conditions for) the clausal noun complement: The present system has the disadvantage that it is inordinately complicated. vs. ? The disadvantage that it is inordinately complicated has been overlooked. Probably an example like This principle may ground some optimism that the account can be usefully pursued. (M. Sainsbury 2002: "Reference and anaphora", Mind & Language) also derives its acceptability from a construction rather than from just optimism. Best, Philippe De Brabanter Paris 4 - Sorbonne ________________________________ From: E.G. To: funknet at mailman.rice.edu Sent: Fri, 10 September, 2010 19:56:23 Subject: Re: [FUNKNET] "Relative clauses" with no relativized element Jespersen and his nexus-substantives should be mentioned (Philosophy of Grammar, 1924). Also in his MEG and Analytic Syntax one could find interesting discussions. Eitan On 10 September 2010 20:53, Giuliana Fiorentino < giuliana.fiorentino at unimol.it> wrote: > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From dick at ling.ucl.ac.uk Sat Sep 11 11:03:56 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 12:03:56 +0100 Subject: analysis: unhappiness In-Reply-To: <10638E94-DC8F-4181-8135-CE74F045079E@colorado.edu> Message-ID: Dear Lise, Many thanks for these references, which I've just looked at. As you say, they do offer hope that we can apply the methods of psycholinguists to the questions of linguists (e.g. how do individuals categorise Korean or Spanish consonants?), and at the same time build a clearer understanding of how these individuals' I-language is related to their E-language (i.e. the language they've heard). That is encouraging, not least because it breaks down what I see as a gulf between linguists and psycholinguists. But even better, it complements very nicely some work in sociolinguistics which I think you'd enjoy reading. It's by a young Scottish sociolinguist called Jennifer Smith, who did a very careful quantitative study (with two colleagues) of 24 3-year olds in a small fishing town in the north of Scotland. She recorded each of them with their mother, and then analysed two sociolinguistic variables that she'd also analysed in adult speech: a phonological variable (pronunciation of the /au/ vowel in "cow", and a morphosyntactic one, the use of -s on a verb with a plural subject, e.g. "My trousers is falling doon." In both cases usage is variable, so the analysis produces a percentage score for each speaker (e.g. 5% of words with the /au/ vowel have a monophthong). She then compared the children's scores with those of their mothers, and found an astonishingly close match for the phonological variable but no match at all for the morphosyntactic one. The reference is Jennifer Smith, Mercedes Durham & Liane Fortune, 2007. 'Mam, my trousers is fa'in doon!' Community, caregiver and child in the acquisition of variation in a Scottish dialect. (Language, Variation and Change 19. 63-99) Once again we have evidence that output can be closely linked to input, we have a nice quantitative method, and we see an example of the 'global' analysis of language that we should all be striving for: one which embraces both E-language (directly observable to the learner as well as to the linguist) and I-language (only indirectly accessible to both), and which tries to both describe and explain the relation between the two. Neither E-language nor I-language is the 'real' language - they're both part of it. And this global enterprise needs all the methods we can muster. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 11/09/2010 02:03, Lise Menn wrote: > Dick, I think that what you say is true, but it is only (I hope) a > temporary description of the state of the art of psycholinguistics. > At least some of us would like our methods to become sensitive enough > to individual differences so that we can look at how 'the general > processes and principles' interact with the level of an individual > person's knowledge of particular constructions, to find out how much > each person knows of the patterns 'out there' in the language. Some > experimental methods are almost at that point already; they can > distinguish degrees of mastery of particular constructions of a > language among groups of second-language learners. > Have a look at > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > 2002. Overhearing a language during childhood. /Psychological Science/ > 13.3, 238-243. > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on > to childhood language memory. /Cognition, 86/(3), B53-B64. > > Lise > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > >> Dear Ted and Ev, >> Yes, I understand your view, but I think it's a psycholinguist's >> view. Your goal is to find general processes and principles that >> apply uniformly across individuals, so you have to use methods to >> check for generality. And (as you know) I admire the way you pursue >> that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how all >> the bits fit together. Like you, I'm aiming to model cognition, but >> my focus is on items and structures, and I start from the assumption >> that these can and do vary across speakers. >> >> However, having said all that I do agree with you that linguists >> should all get used to collecting and using quantitative data; and >> with the help of Brian MacWhinney's typology we'd know what methods >> to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good >> starting point for a proper investigation. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> >> On 10/09/2010 19:30, Ted Gibson wrote: >>> Dear Dick: >>> >>> Perhaps we are talking at cross purposes. I don't understand what is >>> confusing about what Ev Fedorenko and I are claiming. All we are >>> saying is that if you have some testable claim involving a general >>> hypothesis about a language, then you need to get quantitative data >>> from unbiased sources to evaluate that claim. If you are interested >>> in English past tense morphology, then depending on the claims that >>> you might want to investigate, there are lots of ways to get >>> relevant quantitative evidence. Corpus data will probably be useful. >>> For very low frequency words, you can run experiments to test >>> behavior with respect to such words. >>> >>> Your example of the past tense of "bid" is a fine such example. You >>> can run an experiment like the one you suggested to find out what >>> people think the past tense is. If you then found that 20/50 people >>> responded "bidded" and 30/50 respond "bid", that is a lot of useful >>> information. As you suggest in your discussion, this result wouldn't >>> answer the question of how past tense is stored in each individual. >>> This result would be ambiguous among several possible explanations. >>> One possibility is that the probability distribution that is being >>> discovered reflects different dialects, such that 2/5 of the >>> population has one past tense, and 3/5 has another. Another >>> possibility is that each person has a similar probability >>> distribution in their heads, such that 2/5 of the time I respond one >>> way, and 3/5 of the time I respond another. Further experiments >>> would be necessary to answer between these and other possible >>> theories (e.g., with repeated trials from the same person, carefully >>> planned so that the participants don't notice that they are being >>> asked multiple times). Without the quantitative evidence in the >>> first place, there is no way to answer these kinds of questions. >>> >>> Regarding the past tense of "go", this would be useful as a baseline >>> in an experiment involving the less frequent ones. So, yes, it would >>> useful to gather quantitative evidence in such a case also, as >>> baselines with respect to the more interesting cases for theories. >>> >>> The bottom line: if you have a generalization about a language that >>> you wish to evaluate (such that you hypothesize that it is true >>> across the speakers of the language), then you need quantitative >>> evidence from multiple individuals, using an unbiased data >>> collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at least >>> for languages like English. >>> >>> Best wishes, >>> >>> Ted Gibson & Ev Fedorenko >>> >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >>> >>>> Dear Ted, >>>> Thanks for the very interesting comment, but are you REALLY saying >>>> that I shouldn't claim, for example, that the past tense of GO is >>>> "went" without first cross-checking with 50 native speakers? >>>> >>>> Isn't there a danger of missing the point that we all, as native >>>> speakers, spend our whole lives scanning other people's linguistic >>>> behaviour (language 'out there', E-language) and trying to explain >>>> it to ourselves in terms of a language system (language 'in here', >>>> I-language)? So every judgement we make is based on thousands or >>>> millions of observed exemplars, and reflects a unique experience of >>>> E-language filtered through a unique I-language. >>>> >>>> Given that view of language development, I don't see how >>>> quantitative data will help. Let's take a real uncertainty, such as >>>> the past tense of BID. If I want to say I did it, do I say "I >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>>> and 30 "bid". So what? Does that mean that the correct answer is >>>> "bidded"? Surely not. How is it better than my judgement? I agree >>>> you could record my speech and find how often I use each >>>> alternative; but the reason I don't know is precisely because it's >>>> a rare word, so in a sense quantitative data are irrelevant even >>>> there. What would solve the problem of subjectivity, of course, >>>> would be a machine for probing the bit of my mind (or even brain) >>>> that holds BID and its details; but I suspect that even that >>>> wouldn't move us much further forward than my original "don't >>>> know". (Incidentally I write as a fan of quantitative >>>> sociolinguistics, so I do accept that quantitative data are >>>> relevant to linguistic analysis in some areas, where the I-language >>>> phenomenon is frequent enough to produce usable data.) >>>> >>>> It seems to me that this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual. The problem isn't unique to linguistics of course; it's >>>> the same throughout the social sciences. But what's special about >>>> linguistics is that we deal in very fine details of culture (e.g. >>>> details of how a particular word is used or pronounced) so the >>>> differences between individuals really matter. I don't see that >>>> we're ever going to have anything better than judgements to go on, >>>> so what we need is a way to ensure that judgements are accurate >>>> reports of individual I-language. A rotten situation for a science, >>>> but I don't see how it can get better. >>>> >>>> Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> >>>> >>>> On 10/09/2010 14:03, Ted Gibson wrote: >>>>> Dear Dan, Dick: >>>>> >>>>> I would like to clarify some points that Dan Everett makes, in >>>>> response to Dick Hudson. >>>>> >>>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>>>> Fedorenko, 2010, in press, see references and links below) on what we >>>>> think are weak methodological standards in syntax and semantics >>>>> research over the past many years. The issue that we address is the >>>>> prevalent method in syntax and semantics research, which involves >>>>> obtaining a judgment of the acceptability of a sentence / meaning >>>>> pair, typically by just the author of the paper, sometimes with >>>>> feedback from colleagues. As we address in our papers, this >>>>> methodology does not allow proper testing of scientific hypotheses >>>>> because of (a) the small number of experimental participants >>>>> (typically one); (b) the small number of experimental stimuli >>>>> (typically one); (c) cognitive biases on the part of the researcher >>>>> and participants; and (d) the effect of the preceding context (e.g., >>>>> other constructions the researcher may have been recently >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>>> several others cited in Gibson & Fedorenko, in press; for similar >>>>> points, but with not as strong a conclusion as ours). >>>>> >>>>> Three issues need to be separated here: (1) the use of intuitive >>>>> judgments as a dependent measure in a language experiment; (2) >>>>> potential cognitive biases on the part of experimental subjects and >>>>> experimenters in language experiments; and (3) the need for obtaining >>>>> quantitative evidence, whatever the dependent measure might be. The >>>>> paper that Ev and I wrote addresses the last two issues, but does not >>>>> go into depth on the first issue (the use of intuitions as a dependent >>>>> measure in language experiments). Regarding this issue, we don't think >>>>> that there is anything wrong with gathering intuitive judgments as a >>>>> dependent measure, as long as the task is clear to the experimental >>>>> participants. >>>>> >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>>>> arguments that have been given in support of continuing to use the >>>>> traditional non-quantitative method in syntax / semantics research. >>>>> One recent defense of the traditional method comes from Phillips >>>>> (2008), who argues that no harm has come from the non-quantitative >>>>> approach in syntax research thus far. Phillips argues that there are >>>>> no cases in the literature where an incorrect intuitive judgment has >>>>> become the basis for a widely accepted generalization or an important >>>>> theoretical claim. He therefore concludes that there is no reason to >>>>> adopt more rigorous data collection standards. We challenge Philips' >>>>> conclusion by presenting three cases from the literature where a >>>>> faulty intuition has led to incorrect generalizations and mistaken >>>>> theorizing, plausibly due to cognitive biases on the part of the >>>>> researchers. >>>>> >>>>> A second argument that is sometimes presented for the continued use of >>>>> the traditional non-quantitative method is that it would be too >>>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>>>> make this argument explicitly in their response to Gibson & Fedorenko >>>>> (2010): "It would cripple linguistic investigation if it were required >>>>> that all judgments of ambiguity and grammaticality be subject to >>>>> statistically rigorous experiments on naive subjects, especially when >>>>> investigating languages whose speakers are hard to access" (Culicover >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>>> earlier in the discussion here.) Whereas we agree that in >>>>> circumstances where gathering data is difficult, some evidence is >>>>> better than no evidence, we do not agree that research would be slowed >>>>> with respect to languages where experimental participants are easy to >>>>> access, such as English. In contrast, we think that the opposite is >>>>> true: the field's progress is probably slowed by not doing >>>>> quantitative research. >>>>> Suppose that a typical syntax / semantics paper that lacks >>>>> quantitative evidence includes judgments for 50 or more sentences / >>>>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>>>> most of the judgments from such a paper are correct or are on the >>>>> right track, the problem is in knowing which judgments are correct. >>>>> For example, suppose that 90% of the judgments from an arbitrary paper >>>>> are correct (which is probably a high estimate). (Colin Phillips and >>>>> some of his former students / postdocs have commented to us that, in >>>>> their experience, quantitative acceptability judgment studies almost >>>>> always validate the claim(s) in the literature. This is not our >>>>> experience, however. Most experiments that we have run which attempt >>>>> to test some syntactic / semantic hypothesis in the literature end up >>>>> providing us with a pattern of data that had not been known before the >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>>>> that in a paper with 50 empirical claims 45/50 are correct. But which >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That's over >>>>> two million different theories. By quantitatively evaluating the >>>>> empirical claims, we reduce the uncertainty a great deal. To make >>>>> progress, it is better to have theoretical claims supported by solid >>>>> quantitative evidence, so that even if the interpretation of the data >>>>> changes over time as new evidence becomes available -- as is often the >>>>> case in any field of science -- the empirical pattern can be used as a >>>>> basis for further theorizing. >>>>> >>>>> Furthermore, it is no longer expensive to run behavioral experiments, >>>>> at least in English and other widely spoken languages. There now >>>>> exists a marketplace interface -- Amazon.com's Mechanical Turk -- >>>>> which >>>>> can be used for collecting behavioral data over the internet quickly >>>>> and inexpensively. The cost of using an interface like this is >>>>> minimal, and the time that it takes for the results to be returned is >>>>> short. For example, currently on Mechanical Turk, a survey of >>>>> approximately 50 items will be answered by 50 or more participants >>>>> within a couple of hours, at a cost of approximately $1 per >>>>> participant. Thus a survey can be completed within a day, at a cost of >>>>> less than $50. (The hard work of designing the experiment, and >>>>> constructing controlled materials remains of course.) >>>>> >>>>> Sorry to be so verbose. But I think that these methodological points >>>>> are very important. >>>>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson >>>>> >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>>> methods in syntax and semantics research. Language and Cognitive >>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>>> & Fedorenko InPress LCP.pdf >>>>> >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>>>> 2010 TICS.pdf >>>>> >>>>> >>>>> >>>>> >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about methodology. I believe that >>>>>> intuitions are a fine way to generate hypotheses and even to test >>>>>> them - to a degree. But while it might not have been feasible for >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>>> Grammar to conduct experiments on every point of the grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>>>> field forward and for providing the best support for different >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>>>> paper on this, showing serious shortcomings with intuitions as the >>>>>> sole source of evidence, in their paper: "The need for quantitative >>>>>> methods in syntax and semantics research". >>>>>> >>>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>>> convincingly on this. >>>>>> >>>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>>>> Science), and researchers from Brazil are beginning a third round of >>>>>> experimental work among the Pirahas, since my own work on the syntax >>>>>> was, like almost every other field researcher's, based on native >>>>>> speaker intuitions and corpora. >>>>>> >>>>>> The discussion of methodologies reminds me of the initial reactions >>>>>> to Greenberg's work on classifying the languages of the Americas. His >>>>>> methods were strongly (and justifiably) criticized. However, I always >>>>>> thought that his methods were a great way of generating hypotheses, >>>>>> so long as they were ultimately put to the test of standard >>>>>> historical linguistics methods. And the same seems true for use of >>>>>> native-speaker intuitions. >>>>>> >>>>>> -- Dan >>>>> >>>>> >>>>> >>>>>>> We linguists can add a further layer of explanation to the >>>>>>> judgements, but some judgements do seem to be more reliable than >>>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>>> every detailed analysis we make, our whole discipline will >>>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>>> judgements are what put us linguists ahead of the rest in handling >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>>> Language (or the OED) without using native speaker judgements. >>>>>>> >>>>>>> Best wishes, Dick Hudson >>>>> >>>>> >>>>> >>> >>> >>> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From amnfn at well.com Sat Sep 11 13:00:08 2010 From: amnfn at well.com (A. Katz) Date: Sat, 11 Sep 2010 06:00:08 -0700 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284143629@cast-dryerm2.caset.buffalo.edu> Message-ID: Matt, We appear to be fairly close in our approach, but I would have to add that sometimes there isn't a unique "correct" analysis, because the language allows equally for several different ones, and how any particular speaker analyzes a phrase out of context says more about how their individual brain is wired and less about the language. As an example, take the slogan the Coca-Cola company is currently using: "Open happiness". I first saw it on a cocktail napkin in flight. Reading the English slogan, my first analysis was that "open" was an adjective modifying the noun happiness, as opposed to say "closed happiness." That seemed weird, so I considered a few other possiblities. Maybe "open" is a verb in the imperative, and "happiness" is a proper noun in the vocative, as in "Open Sesame!" Then again, it could be that "Happiness" was just a proper noun in objective case: as in "open America (to tourism)." Then I read the French translation on the napkin: "Ouvrez du bonheur." "Oh! So this means "open some happiness"!" I said to myself. English lexemes are underspecified for category, which is why we need little words like "some" to disamnbiguate. That's how the language works. But... all those different analyses could have been correct, given the proper context, and experimenting even with a large population as to which one they thought of first would tell you less about the language and more about the people. The only analysis that seems a bit doubtful is the one suggested by the translation. Best, --Aya On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > Aya, > > I actually agree with everything you say here. Personally, I am MORE > interested in the communicative function of language than I am in > psycholinguistics and how people process language. > > But none of that is relevant, I believe, to the very specific question of > what it means for an analysis to be correct. While one might conclude from > what I said that one ought to do psycholinguistics, that is not my intention. > Rather, my conclusion is that since I myself prefer not to do > psycholinguistics, I cannot really claim that the analyses I come up with are > "the correct" ones. And if it is really important to someone that they > identify "correct" analyses, then they ought to be doing psycholinguistics, > since there is no coherent notion of correct analysis outside of what is > inside of people's heads. > > Matthew > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > wrote: > >> Matthew, >> >> Thanks for stating that, because I was almost beginning to imagine that >> there was no essential disagreement, and that all of us agree that there >> is more -- and less -- to language than what is found in people's heads. >> >> Your position is the one I am familiar with from the functionalist point >> of view, and I was beginning to feel that it was underrepresented on >> Funknet. >> >> Those of us who disagree with your stated position -- but are very >> familiar with it -- are interested not just in psycholinguistics and how >> people process language -- but also in the communicative function of >> language as a system whereby information is transferred. Just as you and >> I may not be aware of the way our emails are encoded and then decoded by >> the computers that help us send emails back and forth, speakers may be >> compeltely unaware of what language does in order to transmit information. >> >> After speakers have finished sending forth their linguistic output, it >> matters not at all how they arrived at this output. Language processing >> is separate from language in the same way that data processing is >> separate from data. >> >> Best, >> >> --Aya >> >> >> On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: >> >>> >>> The following sentence of Lise's >>> >>> "Since we don't in fact have such an agreed-on terminology, we have to >>> be quite careful in making clear what we are referring to when we talk >>> about 'the correct analysis' of a form like 'unhappiness'" >>> >>> suggests that there are two senses in which an analysis can be "the >>> correct analysis", one in terms of what is in people's heads, the other >>> in terms of what is "out there". >>> >>> There are a vast number of patterns "out there". The only distinction, >>> on my view, amongst this vast number of patterns, is between those that >>> reflect something inside people's heads and those that don't. But if >>> that is the case, then there is no coherent sense in which one can talk >>> of "the correct analysis" of what is "out there", except in terms of >>> what is in people's heads, and thus no second sense of "the correct >>> analysis". The patterns that don't correspond to things in people's >>> heads fall into (at least) two categories. There are those that are >>> akin to constellations of stars and, as with constellations, there is >>> no reality to these patterns, except in the minds of linguists. And >>> there are those patterns which are the fossil remains of what was in >>> the heads of speakers of an earlier stage of the language but which no >>> longer are. These latter patterns are real, and they are relevant to >>> exlaining why the language is now the way it is, but they are not >>> relevant, I think many would agree, as to what is the "correct analysis" >>> of the language today. >>> >>> For this reason, I claim that the only sense in which an analysis can be >>> "the correct analysis" is in terms of what is inside of people's heads. >>> >>> Again, I recommend the work of Bruce Derwing for lengthy discussion of >>> these issues. >>> >>> Matthew >>> >>> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn >>> wrote: >>> >>>> I wish we had better terminology for keeping track of whether, at a >>>> given time, we are talking about the patterns that are 'out there' in >>>> the language and might possibly be apprehended (subconsciously) by a >>>> speaker, and when we are talking about the patterns that a particular >>>> speaker actually does apprehend, as indicated by experiments, from >>>> simple 'wug tests' up to brain wave and eye-gaze studies. And for >>>> distinguishing among the degrees of pattern apprehension that a person >>>> may have, from vague preferences detectable in reaction times or other >>>> behavior all the way up through clear metalinguistic insights. Dick >>>> Hudson's note reminding us of the Gleitman and Gleitman study is right >>>> on target. >>>> >>>> Since we don't in fact have such an agreed-on terminology, we have to >>>> be quite careful in making clear what we are referring to when we talk >>>> about 'the correct analysis' of a form like 'unhappiness'. We know, >>>> but tend to forget - and tend to forget to tell our students! - that >>>> it's an empirical question as to whether the formal simplicity and >>>> coherence of description of forms 'out there' (e.g. lovely abstract >>>> morphophonemics) is any kind of approximation to the way knowledge of >>>> the same forms is organized in a particular person's head. If we >>>> remember that a very large proportion of what we know about our >>>> language is 'out there' when we are infants and has to be internalized >>>> through experience with the language (even if you believe in innate >>>> 'core language'), the variation in internal knowledge from one person >>>> to another is more understandable. We especially need to consider (and >>>> try to test) the >>>> possibility that >>>> since >>>> the brain can make multiple cross-connections, multiple patterns are >>>> involved >>>> simultaneously in morphological and syntactic analyses. I suggest that >>>> that's the case with 'unhappiness' - and the linguistic analyses that I >>>> know about are not good at handling that kind of idea. >>>> >>>> Lise >>>> >>>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >>>> >>>>> >>>>> Two comments. >>>>> >>>>> First (elaborating perhaps on Dick Hudson's comment), I think there >>>>> is an >>>>> important distinction between low-level linguistic intuitions (like >>>>> whether a >>>>> word or sentence is well-formed or what it means) and higher-level >>>>> intuitions >>>>> (like what the structure of a word or sentence is). One can take >>>>> the position >>>>> that we need to account for the former (while recognizing that they >>>>> are not >>>>> always reliable) but not the latter. >>>>> >>>>> Second, the tension here is not only between evidence from speaker >>>>> intuitions >>>>> versus evidence from psycholinguistic experiments. There is also a >>>>> tension >>>>> between deciding on the correct analysis on the basis of a priori >>>>> simplicity >>>>> arguments versus deciding on the correct analysis on the basis of >>>>> psycholinguistic evidence (see Derwing 1973). The bracketing >>>>> paradox that Dan >>>>> referred to that arises with the word (semantics argues >>>>> for >>>>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>>>> +er]] (the >>>>> comparative suffix can only be attached to adjectives containing one >>>>> or two >>>>> syllables) is only a paradox if one assumes that speakers adopt the >>>>> simplest >>>>> analysis. For example, if speakers adopt a more complex rule for >>>>> either of these >>>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>>>> trisyllabic >>>>> words beginning with un-), then the bracketing paradox disappears. >>>>> >>>>> Matthew >>>>> >>>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>>>> conscious judgements are only one kind of evidence that we need to >>>>>> take >>>>>> into account. I admire Carson Schutze's work (which I reviewed in >>>>>> fact), >>>>>> and of course I've been aware of complaints about judgements by >>>>>> people >>>>>> like Labov for decades. >>>>>> >>>>>> But you're missing my main point, which is that all judgements aren't >>>>>> equally reliable. If you want to know how /unhappiness/ is >>>>>> structured, >>>>>> ask a linguist, not a five-year old. And one of the by-products of >>>>>> education may be increased sensitivity to syntax - which is one of >>>>>> the >>>>>> many reasons why linguists need to pay more attention to education. >>>>>> >>>>>> Best wishes, Dick >>>>>> >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>>>> Dick, >>>>>>> >>>>>>> You raise an important issue here about >>>>>> methodology. I believe that intuitions are a fine way to generate >>>>>> hypotheses and even to test them - to a degree. But while it might >>>>>> not have >>>>>> been feasible for Huddleston, Pullum, and the other contributors to >>>>>> the >>>>>> Cambridge Grammar to conduct experiments on every point of the >>>>>> grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, >>>>>> corpora, and standard psycholinguistic experimentation (indeed, >>>>>> Standard >>>>>> Social Science Methodology) is vital for taking the field forward >>>>>> and for >>>>>> providing the best support for different analyses. Ted Gibson and Ev >>>>>> Fedorenko have written a very useful new paper on this, showing >>>>>> serious >>>>>> shortcomings with intuitions as the sole source of evidence, in their >>>>>> paper: "The need for quantitative methods in syntax and semantics >>>>>> research".> >>>>>>> Carson Schutze and Wayne Cowart, among others, >>>>>> have also written convincingly on this.> >>>>>>> It is one reason that a team from Stanford, MIT >>>>>> (Brain and Cognitive Science), and researchers from Brazil are >>>>>> beginning a >>>>>> third round of experimental work among the Pirahas, since my own >>>>>> work on >>>>>> the syntax was, like almost every other field researcher's, based >>>>>> on native >>>>>> speaker intuitions and corpora.> >>>>>>> The discussion of methodologies reminds me of >>>>>> the initial reactions to Greenberg's work on classifying the >>>>>> languages of >>>>>> the Americas. His methods were strongly (and justifiably) criticized. >>>>>> However, I always thought that his methods were a great way of >>>>>> generating >>>>>> hypotheses, so long as they were ultimately put to the test of >>>>>> standard >>>>>> historical linguistics methods. And the same seems true for use of >>>>>> native-speaker intuitions.> >>>>>>> -- Dan >>>>>>> >>>>>>>> We linguists can add a further layer of >>>>>> explanation to the judgements, but some judgements do seem to be more >>>>>> reliable than others. And if we have to wait for psycholinguistic >>>>>> evidence >>>>>> for every detailed analysis we make, our whole discipline will >>>>>> immediately >>>>>> grind to a halt. Like it or not, native speaker judgements are what >>>>>> put us >>>>>> linguists ahead of the rest in handling fine detail. Imagine >>>>>> writing the >>>>>> Cambridge Grammar of the English Language (or the OED) without >>>>>> using native >>>>>> speaker judgements.>> >>>>>>>> Best wishes, Dick Hudson >>>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >>>> Lise Menn Home Office: 303-444-4274 >>>> 1625 Mariposa Ave Fax: 303-413-0017 >>>> Boulder CO 80302 >>>> >>>> Professor Emerita of Linguistics >>>> Fellow, Institute of Cognitive Science >>>> University of Colorado >>>> >>>> Secretary, AAAS Section Z [Linguistics] >>>> >>>> Campus Mail Address: >>>> UCB 594, Institute of Cognitive Science >>>> >>>> Campus Physical Address: >>>> CINC 234 >>>> 1777 Exposition Ave, Boulder >>>> >>>> >>>> >>>> >>> >>> >>> >>> >>> >> >> > > > > > From cbutler at ntlworld.com Sat Sep 11 11:17:29 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Sat, 11 Sep 2010 12:17:29 +0100 Subject: FUNKNET] analysis: unhappiness Message-ID: Dick's comment that "this discussion raises the really fundamental question of what kind of thing we think language is: social or individual" is, it seems to me, an important one, particularly for those of us who are committed functionalists. My own view is that a truly functional model of language would be one which aims to account for how human beings communicate using language, or in other words tries to answer the question which was posed by Simon Dik a long time ago now, but which was not tackled head-on in his own work: "How does the natural language user work?' In trying to answer this question we need to accept that language is BOTH social AND individual, and we need to explore both aspects to get as complete a picture as possible of how we communicate using language. We need to know BOTH how people create and respond to meanings and express those meanings in forms during social interaction AND the mechanisms which operate in the brains of individuals in order to make such interaction possible. Both are important parts of the answer to the question 'How do we communicate using language?', though this particular thread of the Funknet discussion has concentrated more on the second aspect, and so will I. This doesn't mean that all the work linguists have done on "exploring the structure of a language so that I can understand how all the bits fit together" and "exploring the connections between items", as Dick puts it, is useless - far from it. After all, the hypotheses that psycholinguists test are based on ideas about what languages are like. But it does mean, in my view, that ultimately we need to get evidence that the constructs and analyses we propose are ones that are at least consistent with what we know of the processes which go on when we use language. So I am with Matthew when he says that for him, "the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads". Of course, this doesn't imply that linguists should just give up their jobs until such time as we know everything there is to know about language processing. But it does mean that we need to collaborate with psycholinguists, psychologists and neurologists, as has also been pointed out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more with sociolinguists and sociologists, so that we can get a better handle on the sociocultural aspects of how we communicate.] And it also means that psycholinguists, for their part, need whenever possible to follow up tightly controlled lab experiments with studies under more naturalistic conditions, to avoid the criticism that what happens in artifical lab situations may not happen in natural communicative conditions. I also agree with Dick when he says that "the differences between individuals really matter", and with Lise when she points out that "we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover". However, there are surely processing mechanisms which are common to all language users by virtue of the evolution of the language faculty and which constitute the "general processes" which Dick says psycholinguists are interested in. On the issue of quantitative methodology, I'm sympathetic in general to Ted and Ev's views, though it does seem sensible to prioritise cases in terms of a hierarchy such as Brian suggests. One thing this means is that we should be giving our university students of linguistics (and some of our linguistics lecturers!) courses in quantitative aspects of linguistics that introduce them to the use of at least some of the basic statistical methods in language study, and I'm sure this is indeed going on in some enlightened places. To those who suspect this can't be done with maths-shy students who don't initially see the need for it, I offer my own experience, over quite a long period, of teaching such courses to people with little or no prior experience in quantitative techniques. For some years in the 1990s, I taught such courses to all linguistics students in an institution where we had many mature students who had come into university level studies with non-standard qualifications, and were not well equipped for courses of this kind by their previous experience. I'm glad to say that teaching the subject from their own perspective as language students rather than that of the statistician, and explaining the reasons for doing things in particular ways rather than just presenting formulae, paid off in the end, so that most students were able to appreciate the relevance of these courses and to turn in very creditable projects showing an understanding of research design and competence in the use of a range of basic statistical techniques. And I still find that bright graduate students respond well to similar courses which incorporate some of the rather more advanced techniques needed for many real research projects in various areas of linguistics. But I may well be out of date with what is now already happening in our fine institutions of higher education! Chris Butler From amnfn at well.com Sat Sep 11 14:18:29 2010 From: amnfn at well.com (A. Katz) Date: Sat, 11 Sep 2010 07:18:29 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <24693E13135D4D6AAD5A098FAA952480@OwnerPC> Message-ID: The problem is that once we achieve the psycholinguistic goal and see what is happening in each person's head, and we see that conflicting analyses are the norm, rather than the exception, among normal language speakers, then we will realize that the way language works to transmit information is despite individual differences and not because of uniform processing strategies. Even when all are agreed as to the meaning of an utterance, they do not process it the same way. Which means that processing is seocndary to information transmission. --Aya On Sat, 11 Sep 2010, Chris Butler wrote: > Dick's comment that "this discussion raises the really fundamental question of what kind of thing we think language is: social or individual" is, it seems to me, an important one, particularly for those of us who are committed functionalists. My own view is that a truly functional model of language would be one which aims to account for how human beings communicate using language, or in other words tries to answer the question which was posed by Simon Dik a long time ago now, but which was not tackled head-on in his own work: "How does the natural language user work?' In trying to answer this question we need to accept that language is BOTH social AND individual, and we need to explore both aspects to get as complete a picture as possible of how we communicate using language. We need to know BOTH how people create and respond to meanings and express those meanings in forms during social interaction AND the mechanisms which operate in the brains of individuals in order to make such interaction possible. Both are important parts of the answer to the question 'How do we communicate using language?', though this particular thread of the Funknet discussion has concentrated more on the second aspect, and so will I. > > This doesn't mean that all the work linguists have done on "exploring the structure of a language so that I can understand how all the bits fit together" and "exploring the connections between items", as Dick puts it, is useless - far from it. After all, the hypotheses that psycholinguists test are based on ideas about what languages are like. But it does mean, in my view, that ultimately we need to get evidence that the constructs and analyses we propose are ones that are at least consistent with what we know of the processes which go on when we use language. So I am with Matthew when he says that for him, "the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads". Of course, this doesn't imply that linguists should just give up their jobs until such time as we know everything there is to know about language processing. But it does mean that we need to collaborate with psycholinguists, psychologists and neurologists, as has also been pointed out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more with sociolinguists and sociologists, so that we can get a better handle on the sociocultural aspects of how we communicate.] And it also means that psycholinguists, for their part, need whenever possible to follow up tightly controlled lab experiments with studies under more naturalistic conditions, to avoid the criticism that what happens in artifical lab situations may not happen in natural communicative conditions. > > I also agree with Dick when he says that "the differences between individuals really matter", and with Lise when she points out that "we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover". However, there are surely processing mechanisms which are common to all language users by virtue of the evolution of the language faculty and which constitute the "general processes" which Dick says psycholinguists are interested in. > > On the issue of quantitative methodology, I'm sympathetic in general to Ted and Ev's views, though it does seem sensible to prioritise cases in terms of a hierarchy such as Brian suggests. One thing this means is that we should be giving our university students of linguistics (and some of our linguistics lecturers!) courses in quantitative aspects of linguistics that introduce them to the use of at least some of the basic statistical methods in language study, and I'm sure this is indeed going on in some enlightened places. To those who suspect this can't be done with maths-shy students who don't initially see the need for it, I offer my own experience, over quite a long period, of teaching such courses to people with little or no prior experience in quantitative techniques. For some years in the 1990s, I taught such courses to all linguistics students in an institution where we had many mature students who had come into university level studies with non-standard qualifications, and were not well equipped for courses of this kind by their previous experience. I'm glad to say that teaching the subject from their own perspective as language students rather than that of the statistician, and explaining the reasons for doing things in particular ways rather than just presenting formulae, paid off in the end, so that most students were able to appreciate the relevance of these courses and to turn in very creditable projects showing an understanding of research design and competence in the use of a range of basic statistical techniques. And I still find that bright graduate students respond well to similar courses which incorporate some of the rather more advanced techniques needed for many real research projects in various areas of linguistics. But I may well be out of date with what is now already happening in our fine institutions of higher education! > > Chris Butler > > From eitan.eg at gmail.com Sat Sep 11 14:54:58 2010 From: eitan.eg at gmail.com (E.G.) Date: Sat, 11 Sep 2010 17:54:58 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: Message-ID: Hi all, I'd like to thank Ron for pointing out the incompleteness of what I wrote, and for the reference to his article. I hope I didn't give the impression I was trying to give a complete description of noun complementation in Modern Hebrew in an email. What I *was* trying to say is that in some languages, unlike English, some nominalizations (of utterance and cognition verbs, as Ron points out, but also of some perception verbs too) can occur with a construction that is explicitly and unmistakably marked as a complement clause. Moreover, the nominalizations that take these explicit complement clauses are related to verbs that can take the same type of complement clause. In such languages, then, it's pretty clear that these instances involve complement clauses. It doesn't mean that other noun complementation strategies don't exist for other types of nouns. However, my main point was more general, albeit poorly expressed. It's that we can turn to cross-linguistic comparison in order to try to reach generalizations about how languages encode meaning. These generalizations are useful, because they can be used to ask "why" questions. For example, it's not really possible to ask *why* Hebrew has two distinct strategies for noun complementation, how *why* English has one. That's because it could always be otherwise, and language change can alter the picture (and has!). However, if we find that in languages that have two strategies, one is limited to nominalizations of PCU verbs, then we have the beginnings of a hierarchy that is amenable to functional explanation. Anyway, it seems that Thomas Payne's question has turned up a pretty general consensus that these constructions are complement clauses. Best wishes, Eitan On 10 September 2010 23:26, Ron Kuzar wrote: > The Modern Hebrew data supplied by Eitan are incomplete. > Hebrew distinguishes between locution (say, hear, think, etc.) and > situation (action, event, state, etc.). > What Eitan describes is only true with regard to nouns (and clauses) > expressing locution. 'Announcement' is indeed such a noun. > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > etc., whose denotatum is a situation, cannot be followed by ki, but only > by Se-, e.g.: > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > annoys me the-situation that-all went > 'I am upset about the situation that all have gone' > > On the other hand, the relative Se- may be replaced by the more > elegant and classical aSer, while the Se- of situation clauses may not. > Sorry about the invented example. I am overseas now. > All this has been described (with corpus data) in: > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit > [Hebrew > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > The article is somewhat outdated and contains some inaccuracies I would > formulate differently today, but the basic distinction is valid in my > opinion. > Best, > Ron Kuzar > --------------- > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > Hi all, > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > comparison can help what seems to be a purely theoretical question based > on > > a single language. The problem here is that English uses the same element > > to > > mark regular relatives and these "appositional" relatives. But if at > least > > one language encodes them by different means, then there's at least a > good > > case for seeing them as distinct functions. It's basically the same > > principle that's used to decide whether to put a meaning on a semantic > map. > > So here are two languages that I know that encode them differently. > > > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > ha-hoda'a Se-kibalnu > > the-announcment rel-we_got > > "The announcement that we got." > > > > ha-hoda'a ki hitbatel ha-mifgaS > > the-message CMP was_cancelled the-meeting > > "The announcement that the meeting was cancelled." > > > > In Coptic, these clauses are marked by ce-, which marks complement > clauses, > > *inter alia*, but not relative clauses: > > > > ph-mewi ce- (complement clause) > > 'the-thought that (we are angry)' > > > > ph-mewi ete- (relative clause) > > 'the thought that (we used to think)' > > > > This seems to be a pretty clear indication that these are complement > > clauses > > rather than relatives. Even if one doesn't like the notion of nouns > taking > > complement clauses (and why not? nominalizations in some languages can > take > > accusative modifiers as well as genitives), it still probably isn't > > incidental that the nominalizations are of verbs that take complement > > clauses when finite. > > > > As usual, the perspective in Talmy Givón's *Syntax* (vol. 2) is worth > > looking at. > > > > Best, > > Eitan > > > > > > On 10 September 2010 19:21, Arie Verhagen > > wrote: > > > > > And as another addition: the clauses that can only be introduced by > > *that* > > > (with no role to > > > play in the subordinate clause) may be seen as (subtypes of) complement > > > clauses, > > > expressing a proposition with the noun functioning as Complement Taking > > > Predicate (CTP), > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > (following analyses by > > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > > Cf. > > > constructions like > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > claim > > > that X", "I put forward > > > the claim that X", in which the relationship between the verb or noun > and > > > the that-clause is > > > comparable to the one in "The claim that X". > > > > > > --Arie Verhagen > > > > > > ---------------- > > > Message from Rong Chen > > > 10 Sep 2010, 23:42 > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > To add to Joanne's comments: > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > (AC) from a relative clause (RC). > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > pronouns. > > > > > > > > 2) The AC and the noun it modifies display an equative > > relationship--one > > > can say X > > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > > often doesn't > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > > relative pronoun > > > > is always part of the clause in an RC. > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > -- > > Eitan Grossman > > Martin Buber Society of Fellows > > Hebrew University of Jerusalem > > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From khildeb at siue.edu Sat Sep 11 15:19:05 2010 From: khildeb at siue.edu (Kristine Hildebrandt) Date: Sat, 11 Sep 2010 10:19:05 -0500 Subject: Job Advertisement Message-ID: Dear Funknetters: Please distribute this job advertisement to interested colleagues and doctoral students completing their degrees. *HIRING UNIT*: *DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE* ** *TITLE/RANK*: Assistant Professor of English-Linguistics *DESCRIPTION OF DUTIES*: The Department of English Language and Literature invites applications for a tenure-track position in general linguistics, with secondary specialization in applied linguistics. The candidate will teach courses in the MA TESL program, along with undergraduate courses in linguistics, composition (ESL and regular), and some general education courses. Academic year: 3/3 load. *TERMS OF APPOINTMENT*: Academic, tenure-track beginning August 16, 2011, 100% appointment. *SOURCE OF FUNDS*: State *SALARY RANGE*: commensurate with training and experience *QUALIFICATIONS REQUIRED: *A Ph.D. in Linguistics, Applied Linguistics, or related field required. If Ph.D. is not completed by the beginning of the contract period, appointment will be at the rank of Instructor until all degree requirements are fulfilled. A record of ESL and/or TESL experience is desirable. *CLOSING DATE FOR APPLICATIONS*: Position open until filled; completed applications postmarked by November 15, 2010 will have priority. Possible interviews at LSA in January 2011. * ** **SEND COVER LETTER, VITA, UNOFFICIAL TRANSCRIPT, STATEMENT OF TEACHING PHILOSOPHY AND RESEARCH AGENDA, AND THREE LETTERS OF RECOMMENDATION TO:** * Linguistics Search Committee Southern Illinois University Edwardsville Department of English Language & Literature Campus Box 1431 Edwardsville, IL 62026-1431 NOTE: Electronic applications will not be accepted for this position. SIUE is a state university-benefits under state sponsored plans may not be available to holders of F1 or J1 visas. Applicants may be subject to a background check prior to an offer of employment. SIUE is an affirmative action and equal opportunity employer. The SIUE ANNUAL SECURITY REPORT is available on-line at: http://admin.siu.edu/studentrightto/. The report contains safety and security information and crime statistics for the past three (3) calendar years. This report is published in compliance with Federal law, entitled *the “Jeanne Clery Disclosure of Campus Security Policy and Campus Crime Statistics Act.” * You may also access this report through the SIUE Home Page: http://www.siue.edu under *Ready References, Quick Links or Publications/Reports*. For those without computer access, a paper copy of the report may be obtained from the Office of the Vice Chancellor for Administration, Rendleman Hall, Room 2228. -- *Kristine A. Hildebrandt* *Assistant Professor, Department of English Language & Literature Southern Illinois University Edwardsville* *Box 1431 Edwardsville, IL 62026 U.S.A. 618-650-3380 (office)* *khildeb at siue.edu http://www.siue.edu/~khildeb* From egibson at MIT.EDU Sat Sep 11 15:45:27 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Sat, 11 Sep 2010 11:45:27 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Brian, Dick, Dan et al: Thanks for the discussion. Here are a few responses: 1. Brian: "But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications." The answer that we give to this question in Gibson & Fedorenko (in press) is as follows (the final paragraph in the paper): "Finally, a question that is often put to us is whether it is necessary to evaluate every empirical claim quantitatively. A major problem with the fields of syntax and semantics is that many papers include no quantitative evidence in support of their research hypotheses. Because conducting experiments is now so easy to do with the advent of Amazon.com’s Mechanical Turk, we recommend gathering quantitative evidence for all empirical claims. However, it would clearly be a vast improvement to the field for all research papers to include at least some quantitative evidence evaluating their research hypotheses." Another possible answer to this question is: the more important some observation is, the better your evidence should be. If the observation is a key reason for some important theoretical claim, then there should be solid quantitative data supporting that observation. In practice, once a linguist starts gathering quantitative data, s/he will realize (a) how easy it is to do; and (b) how beneficial the methods are, with the consequence that these researchers will probably do most or all of their work quantitatively in the future. 2. Dick (by the way, thank you for the kind responses, and your positive tone): "Your [the psycholinguists'] goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality." in contrast to "my focus is on items and structures, and I start from the assumption that these can and do vary across speakers." Many cognitive psychologists / cognitive scientists (all the ones I know at MIT for example) are interested in both cognitive generalizations across people and ways in which people differ cognitively. In fact, some methods (e.g., the individual differences approach where co-variation of various behaviors / characteristics is examined across individuals) have been specifically developed to study differences among individuals. Both kinds of data are important for understanding human cognition, including language. This applies to language research directly: generalizations across people are important, but so are individual differences. In either case, quantitative data are necessary to evaluate research questions and test hypotheses. On a related note, it is a mistake to characterize researchers with a background in "psychology" or cognitive science as being interested in "processing", and researchers with a background in "linguistics" as being interested in "knowledge" or "representation / structure". Both psychologists and linguists should be interested in *both* representation and processing (and learning, for that matter). We wrote a little about this confusion in Gibson & Fedorenko (in press), which we include at the end of the message. This leads to something that Dan said: 3. Dan says: "linguistics is not simply a subdiscipline of psychology" Both linguistics and psychology are big fields. We assume Dan is referring to cognitive psychology / cognitive science here. (Of course, there are sub-fields of psychology - e.g., personality psychology or abnormal psychology - which are somewhat distinct from linguistics, but those sub-fields are also distinct from cognitive psychology.) It is true that historically linguistics is not treated as a subfield of cognitive psychology / cognitive science. However, key research questions in linguistics (i.e., the form of the knowledge structures and algorithms underlying human language) are indeed a subset of those investigated by cognitive psychologists / cognitive scientists. We think that the biggest factor separating linguistics from psychology is the methods used to explore the research questions, rather than the research questions themselves. Consequently, we would like to continue to see tighter connections among the fields of psychology / cognitive science, linguistics, as well as other fields like anthropology and computer science. Thanks to all for the interesting discussions. Ted & Ev We have encountered a claim that the reason for different kinds of methods being used across the different fields of language study (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the research questions are different across these fields, and some methods may be better suited to ask some questions than others. Although the latter is likely true, the premise – that the research questions are different across the fields – is false. The typical claim is that researchers in the field of linguistics are investigating linguistic representations, and researchers in the fields of psycho-/neuro- linguistics are investigating the computations that take place as language is understood or produced. However, many researchers in the fields of psycho-/neuro-linguistics are also interested in the nature of the linguistic representations (at all levels; e.g., phonological representations, lexical representations, syntactic representations, etc.) [1]. By the same token, many researchers in the field of linguistics are interested in the computations that take place in the course of online comprehension or production. However, inferences – drawn from any dependent measure – about either the linguistic representations or computations are always indirect. And these inferences are no more indirect in reading times or event-related potentials, etc., than in acceptability judgments: across all dependent measures we take some observable (e.g., a participant’s rating on an acceptability judgment task or the time it took a participant to read a sentence) and we try to infer something about the underlying cognitive representations / processes. More generally, methods in cognitive science are often used to jointly learn about representations and computations, because inferences about representations can inform questions about the computations, and vice versa. For example, certain data structures can make a computation more or less difficult to perform, and certain representations may require assumptions about the algorithms being used. In our opinion then, the distinction between the fields of linguistics and psycho-/neuro-linguistics is purely along the lines of the kind of data that are used as evidence for or against theoretical hypotheses: typically non-quantitative data in linguistics vs. typically quantitative data in psycho-/neuro-linguistics. Given the superficial nature of this distinction, we think that there should be one field of language study where a wide range of dependent measures is used to investigate linguistic representations and computations. [1] In fact, some methods in cognitive science and cognitive neuroscience were specifically developed to get at representational questions (e.g., lexical / syntactic priming methods, neural adaptation or multi-voxel pattern analyses in functional MRI). On Sep 10, 2010, at 9:05 PM, Daniel Everett wrote: > I think that Brian and Dick make excellent points. There are very > good grammars written that could be improved by psycholinguistic > experimentation and more quantitative approaches. But large sections > of those grammars aren't going to change one bit (go-went) with > quantitative tests and such tests would be completely > counterproductive given the shortness of life and the vastness of > the field linguist's tasks. > > Part of the problem is that linguistics is not simply a > subdiscipline of psychology. Linguistics has its own objectives and > those only occasionally overlap with psychology. The same for methods. > > On another note, I don't buy the 'in my head' 'out of my head' > distinction either (that Matt seems to be urging upon us). We study > different things and have different reasons for being satisfied with > the results we achieve. > > I believe that we linguists are often complacent and fail to apply > better methods. But of course that applies to all disciplines. > > In the meantime, checking corpora, collecting data as a result of > careful interviews with native speakers, and the other aspects of > the field linguist's task are vital parts of the linguist's task and > much of this won't be improved by quantitative methods as we > currently understand them. Maybe sometime. > > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that > they showed the danger of using intuitions. What I meant to say of > using intuitions as standardly used by linguists. They convinced me > that there is a lot to learn from quantitative methods. > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > >> Dear Ted and Ev, >> Yes, I understand your view, but I think it's a psycholinguist's >> view. Your goal is to find general processes and principles that >> apply uniformly across individuals, so you have to use methods to >> check for generality. And (as you know) I admire the way you pursue >> that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how >> all the bits fit together. Like you, I'm aiming to model cognition, >> but my focus is on items and structures, and I start from the >> assumption that these can and do vary across speakers. >> >> However, having said all that I do agree with you that linguists >> should all get used to collecting and using quantitative data; and >> with the help of Brian MacWhinney's typology we'd know what methods >> to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good >> starting point for a proper investigation. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 19:30, Ted Gibson wrote: >>> Dear Dick: >>> >>> Perhaps we are talking at cross purposes. I don't understand what >>> is confusing about what Ev Fedorenko and I are claiming. All we >>> are saying is that if you have some testable claim involving a >>> general hypothesis about a language, then you need to get >>> quantitative data from unbiased sources to evaluate that claim. If >>> you are interested in English past tense morphology, then >>> depending on the claims that you might want to investigate, there >>> are lots of ways to get relevant quantitative evidence. Corpus >>> data will probably be useful. For very low frequency words, you >>> can run experiments to test behavior with respect to such words. >>> >>> Your example of the past tense of "bid" is a fine such example. >>> You can run an experiment like the one you suggested to find out >>> what people think the past tense is. If you then found that 20/50 >>> people responded "bidded" and 30/50 respond "bid", that is a lot >>> of useful information. As you suggest in your discussion, this >>> result wouldn't answer the question of how past tense is stored in >>> each individual. This result would be ambiguous among several >>> possible explanations. One possibility is that the probability >>> distribution that is being discovered reflects different dialects, >>> such that 2/5 of the population has one past tense, and 3/5 has >>> another. Another possibility is that each person has a similar >>> probability distribution in their heads, such that 2/5 of the time >>> I respond one way, and 3/5 of the time I respond another. Further >>> experiments would be necessary to answer between these and other >>> possible theories (e.g., with repeated trials from the same >>> person, carefully planned so that the participants don't notice >>> that they are being asked multiple times). Without the >>> quantitative evidence in the first place, there is no way to >>> answer these kinds of questions. >>> >>> Regarding the past tense of "go", this would be useful as a >>> baseline in an experiment involving the less frequent ones. So, >>> yes, it would useful to gather quantitative evidence in such a >>> case also, as baselines with respect to the more interesting cases >>> for theories. >>> >>> The bottom line: if you have a generalization about a language >>> that you wish to evaluate (such that you hypothesize that it is >>> true across the speakers of the language), then you need >>> quantitative evidence from multiple individuals, using an unbiased >>> data collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at >>> least for languages like English. >>> >>> Best wishes, >>> >>> Ted Gibson & Ev Fedorenko >>> >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >>> >>>> Dear Ted, >>>> Thanks for the very interesting comment, but are you REALLY >>>> saying that I shouldn't claim, for example, that the past tense >>>> of GO is "went" without first cross-checking with 50 native >>>> speakers? >>>> >>>> Isn't there a danger of missing the point that we all, as native >>>> speakers, spend our whole lives scanning other people's >>>> linguistic behaviour (language 'out there', E-language) and >>>> trying to explain it to ourselves in terms of a language system >>>> (language 'in here', I-language)? So every judgement we make is >>>> based on thousands or millions of observed exemplars, and >>>> reflects a unique experience of E-language filtered through a >>>> unique I-language. >>>> >>>> Given that view of language development, I don't see how >>>> quantitative data will help. Let's take a real uncertainty, such >>>> as the past tense of BID. If I want to say I did it, do I say "I >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>>> and 30 "bid". So what? Does that mean that the correct answer is >>>> "bidded"? Surely not. How is it better than my judgement? I agree >>>> you could record my speech and find how often I use each >>>> alternative; but the reason I don't know is precisely because >>>> it's a rare word, so in a sense quantitative data are irrelevant >>>> even there. What would solve the problem of subjectivity, of >>>> course, would be a machine for probing the bit of my mind (or >>>> even brain) that holds BID and its details; but I suspect that >>>> even that wouldn't move us much further forward than my original >>>> "don't know". (Incidentally I write as a fan of quantitative >>>> sociolinguistics, so I do accept that quantitative data are >>>> relevant to linguistic analysis in some areas, where the I- >>>> language phenomenon is frequent enough to produce usable data.) >>>> >>>> It seems to me that this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual. The problem isn't unique to linguistics of course; >>>> it's the same throughout the social sciences. But what's special >>>> about linguistics is that we deal in very fine details of culture >>>> (e.g. details of how a particular word is used or pronounced) so >>>> the differences between individuals really matter. I don't see >>>> that we're ever going to have anything better than judgements to >>>> go on, so what we need is a way to ensure that judgements are >>>> accurate reports of individual I-language. A rotten situation for >>>> a science, but I don't see how it can get better. >>>> >>>> Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> >>>> On 10/09/2010 14:03, Ted Gibson wrote: >>>>> Dear Dan, Dick: >>>>> >>>>> I would like to clarify some points that Dan Everett makes, in >>>>> response to Dick Hudson. >>>>> >>>>> Ev Fedorenko and I have written a couple of papers recently >>>>> (Gibson & >>>>> Fedorenko, 2010, in press, see references and links below) on >>>>> what we >>>>> think are weak methodological standards in syntax and semantics >>>>> research over the past many years. The issue that we address is >>>>> the >>>>> prevalent method in syntax and semantics research, which involves >>>>> obtaining a judgment of the acceptability of a sentence / meaning >>>>> pair, typically by just the author of the paper, sometimes with >>>>> feedback from colleagues. As we address in our papers, this >>>>> methodology does not allow proper testing of scientific hypotheses >>>>> because of (a) the small number of experimental participants >>>>> (typically one); (b) the small number of experimental stimuli >>>>> (typically one); (c) cognitive biases on the part of the >>>>> researcher >>>>> and participants; and (d) the effect of the preceding context >>>>> (e.g., >>>>> other constructions the researcher may have been recently >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>>> several others cited in Gibson & Fedorenko, in press; for similar >>>>> points, but with not as strong a conclusion as ours). >>>>> >>>>> Three issues need to be separated here: (1) the use of intuitive >>>>> judgments as a dependent measure in a language experiment; (2) >>>>> potential cognitive biases on the part of experimental subjects >>>>> and >>>>> experimenters in language experiments; and (3) the need for >>>>> obtaining >>>>> quantitative evidence, whatever the dependent measure might be. >>>>> The >>>>> paper that Ev and I wrote addresses the last two issues, but >>>>> does not >>>>> go into depth on the first issue (the use of intuitions as a >>>>> dependent >>>>> measure in language experiments). Regarding this issue, we don't >>>>> think >>>>> that there is anything wrong with gathering intuitive judgments >>>>> as a >>>>> dependent measure, as long as the task is clear to the >>>>> experimental >>>>> participants. >>>>> >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to >>>>> some >>>>> arguments that have been given in support of continuing to use the >>>>> traditional non-quantitative method in syntax / semantics >>>>> research. >>>>> One recent defense of the traditional method comes from Phillips >>>>> (2008), who argues that no harm has come from the non-quantitative >>>>> approach in syntax research thus far. Phillips argues that there >>>>> are >>>>> no cases in the literature where an incorrect intuitive judgment >>>>> has >>>>> become the basis for a widely accepted generalization or an >>>>> important >>>>> theoretical claim. He therefore concludes that there is no >>>>> reason to >>>>> adopt more rigorous data collection standards. We challenge >>>>> Philips’ >>>>> conclusion by presenting three cases from the literature where a >>>>> faulty intuition has led to incorrect generalizations and mistaken >>>>> theorizing, plausibly due to cognitive biases on the part of the >>>>> researchers. >>>>> >>>>> A second argument that is sometimes presented for the continued >>>>> use of >>>>> the traditional non-quantitative method is that it would be too >>>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff >>>>> (2010) >>>>> make this argument explicitly in their response to Gibson & >>>>> Fedorenko >>>>> (2010): “It would cripple linguistic investigation if it were >>>>> required >>>>> that all judgments of ambiguity and grammaticality be subject to >>>>> statistically rigorous experiments on naive subjects, especially >>>>> when >>>>> investigating languages whose speakers are hard to >>>>> access” (Culicover >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>>> earlier in the discussion here.) Whereas we agree that in >>>>> circumstances where gathering data is difficult, some evidence is >>>>> better than no evidence, we do not agree that research would be >>>>> slowed >>>>> with respect to languages where experimental participants are >>>>> easy to >>>>> access, such as English. In contrast, we think that the opposite >>>>> is >>>>> true: the field’s progress is probably slowed by not doing >>>>> quantitative research. >>>>> Suppose that a typical syntax / semantics paper that lacks >>>>> quantitative evidence includes judgments for 50 or more >>>>> sentences / >>>>> meaning pairs, corresponding to 50 or more empirical claims. >>>>> Even if >>>>> most of the judgments from such a paper are correct or are on the >>>>> right track, the problem is in knowing which judgments are >>>>> correct. >>>>> For example, suppose that 90% of the judgments from an arbitrary >>>>> paper >>>>> are correct (which is probably a high estimate). (Colin Phillips >>>>> and >>>>> some of his former students / postdocs have commented to us >>>>> that, in >>>>> their experience, quantitative acceptability judgment studies >>>>> almost >>>>> always validate the claim(s) in the literature. This is not our >>>>> experience, however. Most experiments that we have run which >>>>> attempt >>>>> to test some syntactic / semantic hypothesis in the literature >>>>> end up >>>>> providing us with a pattern of data that had not been known >>>>> before the >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This >>>>> means >>>>> that in a paper with 50 empirical claims 45/50 are correct. But >>>>> which >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That’s >>>>> over >>>>> two million different theories. By quantitatively evaluating the >>>>> empirical claims, we reduce the uncertainty a great deal. To make >>>>> progress, it is better to have theoretical claims supported by >>>>> solid >>>>> quantitative evidence, so that even if the interpretation of the >>>>> data >>>>> changes over time as new evidence becomes available – as is >>>>> often the >>>>> case in any field of science – the empirical pattern can be used >>>>> as a >>>>> basis for further theorizing. >>>>> >>>>> Furthermore, it is no longer expensive to run behavioral >>>>> experiments, >>>>> at least in English and other widely spoken languages. There now >>>>> exists a marketplace interface – Amazon.com’s Mechanical Turk – >>>>> which >>>>> can be used for collecting behavioral data over the internet >>>>> quickly >>>>> and inexpensively. The cost of using an interface like this is >>>>> minimal, and the time that it takes for the results to be >>>>> returned is >>>>> short. For example, currently on Mechanical Turk, a survey of >>>>> approximately 50 items will be answered by 50 or more participants >>>>> within a couple of hours, at a cost of approximately $1 per >>>>> participant. Thus a survey can be completed within a day, at a >>>>> cost of >>>>> less than $50. (The hard work of designing the experiment, and >>>>> constructing controlled materials remains of course.) >>>>> >>>>> Sorry to be so verbose. But I think that these methodological >>>>> points >>>>> are very important. >>>>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson >>>>> >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>>> methods in syntax and semantics research. Language and Cognitive >>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>>> & Fedorenko InPress LCP.pdf >>>>> >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >>>>> Fedorenko >>>>> 2010 TICS.pdf >>>>> >>>>> >>>>> >>>>> >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about methodology. I believe >>>>>> that >>>>>> intuitions are a fine way to generate hypotheses and even to test >>>>>> them - to a degree. But while it might not have been feasible for >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>>> Grammar to conduct experiments on every point of the grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, corpora, and standard psycholinguistic >>>>>> experimentation >>>>>> (indeed, Standard Social Science Methodology) is vital for >>>>>> taking the >>>>>> field forward and for providing the best support for different >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very >>>>>> useful new >>>>>> paper on this, showing serious shortcomings with intuitions as >>>>>> the >>>>>> sole source of evidence, in their paper: "The need for >>>>>> quantitative >>>>>> methods in syntax and semantics research". >>>>>> >>>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>>> convincingly on this. >>>>>> >>>>>> It is one reason that a team from Stanford, MIT (Brain and >>>>>> Cognitive >>>>>> Science), and researchers from Brazil are beginning a third >>>>>> round of >>>>>> experimental work among the Pirahas, since my own work on the >>>>>> syntax >>>>>> was, like almost every other field researcher's, based on native >>>>>> speaker intuitions and corpora. >>>>>> >>>>>> The discussion of methodologies reminds me of the initial >>>>>> reactions >>>>>> to Greenberg's work on classifying the languages of the >>>>>> Americas. His >>>>>> methods were strongly (and justifiably) criticized. However, I >>>>>> always >>>>>> thought that his methods were a great way of generating >>>>>> hypotheses, >>>>>> so long as they were ultimately put to the test of standard >>>>>> historical linguistics methods. And the same seems true for use >>>>>> of >>>>>> native-speaker intuitions. >>>>>> >>>>>> -- Dan >>>>> >>>>> >>>>> >>>>>>> We linguists can add a further layer of explanation to the >>>>>>> judgements, but some judgements do seem to be more reliable than >>>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>>> every detailed analysis we make, our whole discipline will >>>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>>> judgements are what put us linguists ahead of the rest in >>>>>>> handling >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the >>>>>>> English >>>>>>> Language (or the OED) without using native speaker judgements. >>>>>>> >>>>>>> Best wishes, Dick Hudson >>>>> >>>>> >>>>> >>> >>> >>> >> > From dan at daneverett.org Sat Sep 11 16:53:23 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 12:53:23 -0400 Subject: analysis: unhappiness In-Reply-To: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9@MIT.EDU> Message-ID: Ted, Let me clarify this: > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > Both linguistics and psychology are big fields. We assume Dan is referring to cognitive psychology / cognitive science here. (Of course, there are sub-fields of psychology - e.g., personality psychology or abnormal psychology - which are somewhat distinct from linguistics, but those sub-fields are also distinct from cognitive psychology.) It is true that historically linguistics is not treated as a subfield of cognitive psychology / cognitive science. However, key research questions in linguistics (i.e., the form of the knowledge structures and algorithms underlying human language) are indeed a subset of those investigated by cognitive psychologists / cognitive scientists. We think that the biggest factor separating linguistics from psychology is the methods used to explore the research questions, rather than the research questions themselves. Consequently, we would like to continue to see tighter connections among the fields of psychology / cognitive science, linguistics, as well as other fields like anthropology and computer science. Correct, I meant cognitive psychology, not, say, psychoanalysis. There are definitely overlapping concerns. But my main concern about language is less about representations and more about the cultural and sociological values that lead to sentences and expressions in the corpus, rather than the mind. I used to think that my main interest was representations in the mind. But I find the psychology less interesting than the anthropology these days. But this is not an excuse to avoid quantitative methods. I believe that you and Ev, and others, have made a convincing case for quantitative methods. Quantitative methods in field research on syntax and semantics is vital. -- dan From dryer at buffalo.edu Sat Sep 11 20:25:54 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Sat, 11 Sep 2010 16:25:54 -0400 Subject: analysis: unhappiness Message-ID: Lise, First, let me concede the obvious: what is inside people's heads varies from individual to individual and, for a given individual, from one time to another. And that whether a given pattern is internalized by a given speaker at a given time is not a black and white issue, but often a matter of degree. But while that makes the story more complicated, it seems largely orthogonal to the fundamental issue we are discussing. It's not clear to me at this point whether the issue we are disagreeing about is substantive or terminological, whether we are meaning different things when we talk about patterns in the data. For example, the frequency with which different sorts of sentences occur in the data a child is exposed to has profound effects on what the child internalizes as rules and hence is an important part of explanations for why languages are the way they are. In so far as such relative frequencies in the data are "patterns in the data", we need to know what those patterns are in order to explain why languages are the way they are. However, that is not what I was talking about when I referred to patterns in the data. I was intending the sorts of patterns that distinguish two analyses of the same set of data, and relative frequency in the data does not play a role in distinguishing what two analyses of the same set of data say. For example, when we compare a more abstract analysis of some phonological phenomenon that posits some rule based on a perceived pattern in the data with a more concrete analysis of the same phenomenon that denies the generalization covered by the rule in the more abstract analysis, and ask which is "the correct" analysis, my claim is that the only sense of that question is whether the generalization corresponds to something inside speakers' heads (admitting, again, that this can vary from speaker to speaker, from time to time, and is a matter of degree). There is no second sense in which the question can be interpreted as a question about which analysis is correct as far as what is "out there" is concerned. As far as I can see, everything you say in your response to me is consistent with this claim of mine. And when you say " I have to disagree with you on the validity of describing what's 'out there'", your wording implies that I have said that describing what's "other there" is not a valid activity. But, to the contrary, since I am not a psycholinguist, that is ALL that I do. I am currently (co-)writing a grammatical description of a language, and that is describing what's "out there". However, when confronted with two analyses which are consistent with the data, which differ in that one describes the data in terms of a rule that the other does not, I do not believe that one of the two analyses must be the "correct" one in terms of what is "out there". At most, one of them may be more correct in terms of what is inside speakers' heads (and perhaps one is correct for some speakers and the other for other speakers), but since we are not doing psycholinguistic research on those speakers, I can't tell. But we have enough work to do describing the language that I needn't worry about which analysis is correct. Some old notions from Chomsky are useful here. We have enough to do trying to achieve observational adequacy in describing the language we are working on that I really don't worry about descriptive adequacy. When we have competing analyses that we need to decide on, they normally differ in observational adequacy, so we look for data that will tell us which is correct in terms of observational adequacy. Now it's true that our description is full of generalizations based on patterns in the data; after all, a listing of our raw data would achieve observational adequacy. But in doing so, we are not claiming that the generalizations we are making are necessarily real. I suspect some of them aren't. (And by "real", I mean psychologically real, since my whole point in this discussion is that the only sense in which generalizations can or cannot be real is psychological.) So why don't we just provide a listing of our raw data? Because we want to communicate to others what the language is like and a description that describes it in terms of patterns in the data better serves that purpose, whether or not the patterns are real. Matthew On Fri 09/10/10 8:40 PM , Lise Menn Lise.Menn at Colorado.EDU sent: > Matt, I have to disagree with you on the validity of describing what's > 'out there' (what Dick Hudson says is his interest, in his > contribution of 5:40:06 PM MDT today). We DO have to account for it > in order to understand how 'the language in speakers' heads' gets > into those heads in the first place. In more detail: Each of us is > immersed from (before) birth in a sampling of utterances (and if we > are literate, eventually also written forms of the language). In > order to understand how we really create our internal representations > of our language, we have to know (or be able to estimate) something > about the data our brains get as input. There are at least better and > worse descriptions of the patterns in those data, and certainly there > are wrong ones, though in many cases - for example in the > 'unhappiness' case - there are probably conflicting right ones, > rather than any single correct one. (OT offers some help in thinking > about this.) > To take a concrete example, in order to account for the > still-unstable changes in English pronominal case marking in compound > NP objects of prepositions from a system based on syntactic case (He > gave the cookies to Mary and me) to a system apparently based partly > on whether the pronoun is next to the governing preposition (He gave > the cookies to Mary and I/ to me and Mary), you first have to do an > analysis of usage and figure out what the pattern is. And usage is > not in our heads (although it's the result of what's in our heads), > it's 'out there' . > Even fossils and obscure patterns contribute to the redundancy of > the language, making it more learnable and and helping to create the > resonances used by great poets and orators. (I admit to having > oversimplified in speaking as if there were always one 'correct' > analysis of the patterns 'out there' that might be (subconsciously) > discoverable by speakers. That's not true.) And because not all > speakers are equally sensitive to language patterns - again, the > Gleitman and Gleitman book is a terrific example - it's also an > oversimplification to talk about 'what is in speaker's heads' as if > the same thing is in everyone's head. (K.P. Mohanan has also > published on this.) At the lexical level, Danielle Cyr's examples ( > September 9, 2010 8:38:59 PM MDT) further remind us that what's > inside each person's head changes over time. So we must also be > careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover. It's not - > it's more like a complex mosaic that does not fit together perfectly. > Lise > On Sep 10, 2010, at 12:51 PM, wrote: > > The following sentence of Lise's > "Since we don't in fact have such an agreed-on terminology, we have > to be quite careful in making clear what we are referring to when we > talk about 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the > correct analysis", one in terms of what is in people's heads, the > other in terms of what is "out there". > > There are a vast number of patterns "out there". The only > distinction, on my view, amongst this vast number of patterns, is > between those that reflect something inside people's heads and those > that don't. But if that is the case, then there is no coherent sense > in which one can talk of "the correct analysis" of what is "out > there", except in terms of what is in people's heads, and thus no > second sense of "the correct analysis". The patterns that don't > correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of stars > and, as with constellations, there is no reality to these patterns, > except in the minds of linguists. And there are those patterns which > are the fossil remains of what was in the heads of speakers of an > earlier stage of the language but which no longer are. These latter > patterns are real, and they are relevant to exlaining why the > language is now the way it is, but they are not relevant, I think > many would agree, as to what is the "correct analysis" of the > language today. > For this reason, I claim that the only sense in which an analysis can > be "the correct analysis" is in terms of what is inside of people's > heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion > of these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn wrote: > > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' > in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. And for > distinguishing among the degrees of pattern apprehension that a > person > may have, from vague preferences detectable in reaction times or > other > behavior all the way up through clear metalinguistic insights. Dick > Hudson's note reminding us of the Gleitman and Gleitman study is > right > on target. > Since we don't in fact have such an agreed-on terminology, we have > to > be quite careful in making clear what we are referring to when we > talk > about 'the correct analysis' of a form like 'unhappiness'. We know, > but > tend to forget - and tend to forget to tell our students! - that > it's an > empirical question as to whether the formal simplicity and > coherence of > description of forms 'out there' (e.g. lovely abstract > morphophonemics) > is any kind of approximation to the way knowledge of the same forms > is > organized in a particular person's head. If we remember that a very > large proportion of what we know about our language is 'out there' > when > we are infants and has to be internalized through experience with > the > language (even if you believe in innate 'core language'), the > variation > in internal knowledge from one person to another is more > understandable. > We especially need to consider (and try to test) the possibility > that > since > the brain can make multiple cross-connections, multiple patterns are > involved > simultaneously in morphological and syntactic analyses. I suggest > that > that's the case with 'unhappiness' - and the linguistic analyses > that I > know about are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there > is an > important distinction between low-level linguistic intuitions (like > whether a > word or sentence is well-formed or what it means) and higher-level > intuitions > (like what the structure of a word or sentence is). One can take > the position > that we need to account for the former (while recognizing that they > are not > always reliable) but not the latter. > Second, the tension here is not only between evidence from speaker > intuitions > versus evidence from psycholinguistic experiments. There is also a > tension > between deciding on the correct analysis on the basis of a priori > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing > paradox that Dan > referred to that arises with the word (semantics argues > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one > or two > syllables) is only a paradox if one assumes that speakers adopt the > simplest > analysis. For example, if speakers adopt a more complex rule for > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson sent: > Thanks Dan. I'm sure you're right, and I'd be the first to agree > that > conscious judgements are only one kind of evidence that we need to > take > into account. I admire Carson Schutze's work (which I reviewed in > fact), > and of course I've been aware of complaints about judgements by > people > like Labov for decades. > > But you're missing my main point, which is that all judgements aren't > equally reliable. If you want to know how /unhappiness/ is > structured, > ask a linguist, not a five-year old. And one of the by-products of > education may be increased sensitivity to syntax - which is one of > the > many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm [4] > On 09/09/2010 11:39, Daniel Everett wrote: > Dick, > > You raise an important issue here about > methodology. I believe that intuitions are a fine way to generate > hypotheses and even to test them - to a degree. But while it might > not have > been feasible for Huddleston, Pullum, and the other contributors to > the > Cambridge Grammar to conduct experiments on every point of the > grammar, > experiments could have only made the grammar better. The use of > intuitions, > corpora, and standard psycholinguistic experimentation (indeed, > Standard > Social Science Methodology) is vital for taking the field forward > and for > providing the best support for different analyses. Ted Gibson and Ev > Fedorenko have written a very useful new paper on this, showing > serious > shortcomings with intuitions as the sole source of evidence, in > their > paper: "The need for quantitative methods in syntax and semantics > research".> > Carson Schutze and Wayne Cowart, among others, > have also written convincingly on this.> > It is one reason that a team from Stanford, MIT > (Brain and Cognitive Science), and researchers from Brazil are > beginning a > third round of experimental work among the Pirahas, since my own > work on > the syntax was, like almost every other field researcher's, based > on native > speaker intuitions and corpora.> > The discussion of methodologies reminds me of > the initial reactions to Greenberg's work on classifying the > languages of > the Americas. His methods were strongly (and justifiably) criticized. > However, I always thought that his methods were a great way of > generating > hypotheses, so long as they were ultimately put to the test of > standard > historical linguistics methods. And the same seems true for use of > native-speaker intuitions.> > -- Dan > > We linguists can add a further layer of > explanation to the judgements, but some judgements do seem to be more > reliable than others. And if we have to wait for psycholinguistic > evidence > for every detailed analysis we make, our whole discipline will > immediately > grind to a halt. Like it or not, native speaker judgements are what > put us > linguists ahead of the rest in handling fine detail. Imagine > writing the > Cambridge Grammar of the English Language (or the OED) without > using native > speaker judgements.>> > Best wishes, Dick Hudson > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > Lise Menn Home Office: 303-444-42741625 > Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 > Professor Emerita of Linguistics Fellow, Institute of Cognitive > Science University of Colorado > Secretary, AAAS Section Z [Linguistics] > Campus Mail Address:UCB 594, Institute of Cognitive Science > Campus Physical Address:CINC 234 1777 Exposition Ave, Boulder > > > Links: > ------ > [4] http://www.phon.ucl.ac.uk/home/dick/home.htm > > From dick at ling.ucl.ac.uk Sat Sep 11 21:14:16 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 22:14:16 +0100 Subject: analysis: unhappiness In-Reply-To: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9@MIT.EDU> Message-ID: Dear Ted, Thanks for this. Sorry I misrepresented your goals; I'm afraid I was stereotyping you. But then, I was also stereotyping myself. It looks as though we both share the same global goal that I sketched in my earlier message: a comprehensive model for language which covers both structure and processing, and both behaviour and cognition; and which allows individual variation in all four. Hooray! All we need is the wisdom to 'think global and act local', as they keep telling us. I'm afraid it's all too easy to get side-tracked into something narrow and easier. Best wishes, Dick > 2. Dick (by the way, thank you for the kind responses, and your > positive tone): > > "Your [the psycholinguists'] goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality." > > in contrast to "my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers." > > Many cognitive psychologists / cognitive scientists (all the ones I > know at MIT for example) are interested in both cognitive > generalizations across people and ways in which people differ > cognitively. In fact, some methods (e.g., the individual differences > approach where co-variation of various behaviors / characteristics is > examined across individuals) have been specifically developed to study > differences among individuals. Both kinds of data are important for > understanding human cognition, including language. This applies to > language research directly: generalizations across people are > important, but so are individual differences. In either case, > quantitative data are necessary to evaluate research questions and > test hypotheses. > > On a related note, it is a mistake to characterize researchers with a > background in "psychology" or cognitive science as being interested in > "processing", and researchers with a background in "linguistics" as > being interested in "knowledge" or "representation / structure". Both > psychologists and linguists should be interested in *both* > representation and processing (and learning, for that matter). We > wrote a little about this confusion in Gibson & Fedorenko (in press), > which we include at the end of the message. > From dryer at buffalo.edu Sun Sep 12 01:26:25 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Sat, 11 Sep 2010 21:26:25 -0400 Subject: analysis: unhappiness Message-ID: Dan etc, There have unfortunately been two sub-threads with the same subject heading "Re: [FUNKNET] analysis: unhappiness". Even I have found the Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have pursued the other one only because it wasn't clear that others hadn't misunderstood what I was trying to say. Unfortunately, the following comment from Dan seems to illustrate further that I haven't made myself clear: "On another note, I don't buy the 'in my head' 'out of my head' distinction either (that Matt seems to be urging upon us)." BUT, it was Lise who urged that distinction on us. The whole point of my emails has been to deny such a distinction, to argue that the only reality is the "in the head" one. In fact, the gradual convergence of thinking in the Gibson-etc subthread seems to reflect the idea that despite apparent differences, there is a common underlying goal. Matthew On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: > I think that Brian and Dick make excellent points. There are very good > grammars written that could be improved by psycholinguistic experimentation > and more quantitative approaches. But large sections of those grammars > aren't going to change one bit (go-went) with quantitative tests and such > tests would be completely counterproductive given the shortness of life and > the vastness of the field linguist's tasks. > Part of the problem is that linguistics is not simply a subdiscipline of > psychology. Linguistics has its own objectives and those only occasionally > overlap with psychology. The same for methods. > On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us). We study different things > and have different reasons for being satisfied with the results we > achieve. > I believe that we linguists are often complacent and fail to apply better > methods. But of course that applies to all disciplines. > In the meantime, checking corpora, collecting data as a result of careful > interviews with native speakers, and the other aspects of the field > linguist's task are vital parts of the linguist's task and much of this > won't be improved by quantitative methods as we currently understand them. > Maybe sometime. > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that they > showed the danger of using intuitions. What I meant to say of using > intuitions as standardly used by linguists. They convinced me that there is > a lot to learn from quantitative methods. > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a > psycholinguist's view. Your goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality. And (as you know) I admire the way you > pursue that goal. But my goal, as a linguist, is different. I want to > explore the structure of a language so that I can understand how all the > bits fit together. Like you, I'm aiming to model cognition, but my focus is > on items and structures, and I start from the assumption that these can and > do vary across speakers.> > > However, having said all that I do agree with > you that linguists should all get used to collecting and using quantitative > data; and with the help of Brian MacWhinney's typology we'd know what > methods to use when. And I do agree with your points about bid/bidded: in > cases like that, quantitative data would be at least a very good starting > point for a proper investigation.> > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> > > On 10/09/2010 19:30, Ted Gibson > wrote:>> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I > don't understand what is confusing about what Ev Fedorenko and I are > claiming. All we are saying is that if you have some testable claim > involving a general hypothesis about a language, then you need to get > quantitative data from unbiased sources to evaluate that claim. If you are > interested in English past tense morphology, then depending on the claims > that you might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very low > frequency words, you can run experiments to test behavior with respect to > such words.>> > >> Your example of the past tense of > "bid" is a fine such example. You can run an experiment like the > one you suggested to find out what people think the past tense is. If you > then found that 20/50 people responded "bidded" and 30/50 respond > "bid", that is a lot of useful information. As you suggest in > your discussion, this result wouldn't answer the question of how past tense > is stored in each individual. This result would be ambiguous among several > possible explanations. One possibility is that the probability distribution > that is being discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another possibility is > that each person has a similar probability distribution in their heads, > such that 2/5 of the time I respond one way, and 3/5 of the time I respond > another. Further experiments would be necessary to answer between these and > other possible theories (e.g., with repeated trials from the same person, > carefully planned so that the participants don't notice that they are being > asked multiple times). Without the quantitative evidence in the first > place, there is no way to answer these kinds of questions.>> > >> Regarding the past tense of "go", > this would be useful as a baseline in an experiment involving the less > frequent ones. So, yes, it would useful to gather quantitative evidence in > such a case also, as baselines with respect to the more interesting cases > for theories.>> > >> The bottom line: if you have a > generalization about a language that you wish to evaluate (such that you > hypothesize that it is true across the speakers of the language), then you > need quantitative evidence from multiple individuals, using an unbiased > data collection method, to evaluate such a claim. The point about > Mechanical Turk is that it is really *easy* to do this now, at least for > languages like English.>> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson > wrote:>> > >>> Dear Ted, > >>> Thanks for the very interesting comment, > but are you REALLY saying that I shouldn't claim, for example, that the > past tense of GO is "went" without first cross-checking with 50 > native speakers?>>> > >>> Isn't there a danger of missing the > point that we all, as native speakers, spend our whole lives scanning other > people's linguistic behaviour (language 'out there', E-language) and trying > to explain it to ourselves in terms of a language system (language 'in > here', I-language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language.>>> > >>> Given that view of language development, > I don't see how quantitative data will help. Let's take a real uncertainty, > such as the past tense of BID. If I want to say I did it, do I say "I > bidded" or "I bid"? My judgement: I don't know. Ok, you get > 50 people to oblige on Mechanical Turk, and 20 of them give > "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my > judgement? I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for > probing the bit of my mind (or even brain) that holds BID and its details; > but I suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of > quantitative sociolinguistics, so I do accept that quantitative data are > relevant to linguistic analysis in some areas, where the I-language > phenomenon is frequent enough to produce usable data.)>>> > >>> It seems to me that this discussion > raises the really fundamental question of what kind of thing we think > language is: social or individual. The problem isn't unique to linguistics > of course; it's the same throughout the social sciences. But what's special > about linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the differences > between individuals really matter. I don't see that we're ever going to > have anything better than judgements to go on, so what we need is a way to > ensure that judgements are accurate reports of individual I-language. A > rotten situation for a science, but I don't see how it can get > better.>>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> > >>> On 10/09/2010 14:03, Ted Gibson > wrote:>>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points > that Dan Everett makes, in>>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a > couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see > references and links below) on what we>>>> think are weak methodological > standards in syntax and semantics>>>> research over the past many years. > The issue that we address is the>>>> prevalent method in syntax and > semantics research, which involves>>>> obtaining a judgment of the > acceptability of a sentence / meaning>>>> pair, typically by just the author > of the paper, sometimes with>>>> feedback from colleagues. As we > address in our papers, this>>>> methodology does not allow proper > testing of scientific hypotheses>>>> because of (a) the small number of > experimental participants>>>> (typically one); (b) the small > number of experimental stimuli>>>> (typically one); (c) cognitive > biases on the part of the researcher>>>> and participants; and (d) the effect > of the preceding context (e.g.,>>>> other constructions the researcher > may have been recently>>>> considering). (As Dan said, see > Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & > Fedorenko, in press; for similar>>>> points, but with not as strong a > conclusion as ours).>>>> > >>>> Three issues need to be separated > here: (1) the use of intuitive>>>> judgments as a dependent measure in > a language experiment; (2)>>>> potential cognitive biases on the > part of experimental subjects and>>>> experimenters in language > experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the > dependent measure might be. The>>>> paper that Ev and I wrote addresses > the last two issues, but does not>>>> go into depth on the first issue > (the use of intuitions as a dependent>>>> measure in language experiments). > Regarding this issue, we don't think>>>> that there is anything wrong with > gathering intuitive judgments as a>>>> dependent measure, as long as the > task is clear to the experimental>>>> participants. > >>>> > >>>> In the longer paper (Gibson & > Fedorenko, in press) we respond to some>>>> arguments that have been given in > support of continuing to use the>>>> traditional non-quantitative method > in syntax / semantics research.>>>> One recent defense of the > traditional method comes from Phillips>>>> (2008), who argues that no harm has > come from the non-quantitative>>>> approach in syntax research thus > far. Phillips argues that there are>>>> no cases in the literature where an > incorrect intuitive judgment has>>>> become the basis for a widely > accepted generalization or an important>>>> theoretical claim. He therefore > concludes that there is no reason to>>>> adopt more rigorous data collection > standards. We challenge Philipsâ��>>>> conclusion by presenting three cases > from the literature where a>>>> faulty intuition has led to > incorrect generalizations and mistaken>>>> theorizing, plausibly due to > cognitive biases on the part of the>>>> researchers. > >>>> > >>>> A second argument that is sometimes > presented for the continued use of>>>> the traditional non-quantitative > method is that it would be too>>>> inefficient to evaluate every > syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For > example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in > their response to Gibson & Fedorenko>>>> (2010): â��It would cripple > linguistic investigation if it were required>>>> that all judgments of ambiguity and > grammaticality be subject to>>>> statistically rigorous experiments > on naive subjects, especially when>>>> investigating languages whose > speakers are hard to accessâ�� (Culicover>>>> & Jackendoff, 2010, p. 234). > (Dick Hudson makes a similar point>>>> earlier in the discussion here.) > Whereas we agree that in>>>> circumstances where gathering data > is difficult, some evidence is>>>> better than no evidence, we do not > agree that research would be slowed>>>> with respect to languages where > experimental participants are easy to>>>> access, such as English. In > contrast, we think that the opposite is>>>> true: the fieldâ��s progress > is probably slowed by not doing>>>> quantitative research. > >>>> Suppose that a typical syntax / > semantics paper that lacks>>>> quantitative evidence includes > judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 > or more empirical claims. Even if>>>> most of the judgments from such a > paper are correct or are on the>>>> right track, the problem is in > knowing which judgments are correct.>>>> For example, suppose that 90% of the > judgments from an arbitrary paper>>>> are correct (which is probably a > high estimate). (Colin Phillips and>>>> some of his former students / > postdocs have commented to us that, in>>>> their experience, quantitative > acceptability judgment studies almost>>>> always validate the claim(s) in the > literature. This is not our>>>> experience, however. Most > experiments that we have run which attempt>>>> to test some syntactic / semantic > hypothesis in the literature end up>>>> providing us with a pattern of data > that had not been known before the>>>> experiment (e.g., Breen et al., in > press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras > & Gibson, submitted).) This means>>>> that in a paper with 50 empirical > claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to > choose 45 items from 50. Thatâ��s over>>>> two million different theories. By > quantitatively evaluating the>>>> empirical claims, we reduce the > uncertainty a great deal. To make>>>> progress, it is better to have > theoretical claims supported by solid>>>> quantitative evidence, so that even > if the interpretation of the data>>>> changes over time as new evidence > becomes available â�� as is often the>>>> case in any field of science > â�� the empirical pattern can be used as a>>>> basis for further > theorizing.>>>> > >>>> Furthermore, it is no longer > expensive to run behavioral experiments,>>>> at least in English and other widely > spoken languages. There now>>>> exists a marketplace interface > â�� Amazon.comâ��s Mechanical Turk â�� which>>>> can be used for collecting > behavioral data over the internet quickly>>>> and inexpensively. The cost of using > an interface like this is>>>> minimal, and the time that it takes > for the results to be returned is>>>> short. For example, currently on > Mechanical Turk, a survey of>>>> approximately 50 items will be > answered by 50 or more participants>>>> within a couple of hours, at a cost > of approximately $1 per>>>> participant. Thus a survey can be > completed within a day, at a cost of>>>> less than $50. (The hard work of > designing the experiment, and>>>> constructing controlled materials > remains of course.)>>>> > >>>> Sorry to be so verbose. But I think > that these methodological points>>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In > press). The need for quantitative>>>> methods in syntax and semantics > research. Language and Cognitive>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress > LCP.pdf>>>> > >>>> Gibson, E. & Fedorenko, E. > (2010). Weak quantitative standards in>>>> linguistics research. Trends in > Cognitive Science, 14, 233-234.>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue > here about methodology. I believe that>>>>> intuitions are a fine way to > generate hypotheses and even to test>>>>> them - to a degree. But while it > might not have been feasible for>>>>> Huddleston, Pullum, and the > other contributors to the Cambridge>>>>> Grammar to conduct experiments > on every point of the grammar,>>>>> experiments could have only made > the grammar better. The use of>>>>> intuitions, corpora, and > standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science > Methodology) is vital for taking the>>>>> field forward and for providing > the best support for different>>>>> analyses. Ted Gibson and Ev > Fedorenko have written a very useful new>>>>> paper on this, showing serious > shortcomings with intuitions as the>>>>> sole source of evidence, in > their paper: "The need for quantitative>>>>> methods in syntax and semantics > research".>>>>> > >>>>> Carson Schutze and Wayne Cowart, > among others, have also written>>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team > from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from > Brazil are beginning a third round of>>>>> experimental work among the > Pirahas, since my own work on the syntax>>>>> was, like almost every other > field researcher's, based on native>>>>> speaker intuitions and > corpora.>>>>> > >>>>> The discussion of methodologies > reminds me of the initial reactions>>>>> to Greenberg's work on > classifying the languages of the Americas. His>>>>> methods were strongly (and > justifiably) criticized. However, I always>>>>> thought that his methods were a > great way of generating hypotheses,>>>>> so long as they were ultimately > put to the test of standard>>>>> historical linguistics methods. > And the same seems true for use of>>>>> native-speaker > intuitions.>>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a > further layer of explanation to the>>>>>> judgements, but some > judgements do seem to be more reliable than>>>>>> others. And if we have to > wait for psycholinguistic evidence for>>>>>> every detailed analysis we > make, our whole discipline will>>>>>> immediately grind to a halt. > Like it or not, native speaker>>>>>> judgements are what put us > linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing > the Cambridge Grammar of the English>>>>>> Language (or the OED) > without using native speaker judgements.>>>>>> > >>>>>> Best wishes, Dick > Hudson>>>> > >>>> > >>>> > >> > >> > >> > > > > > > > From dan at daneverett.org Sun Sep 12 01:32:59 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 21:32:59 -0400 Subject: analysis: unhappiness In-Reply-To: <12429.1284254785@buffalo.edu> Message-ID: Sorry, Matt, for being careless in my attribution of that distinction. I do believe that there is such a distinction, however, and that it is very important. If you are after careful and deep understanding of the nature of representations, then the 'in the head' focus is important. If you are more interested (always, occasionally, etc) in how discourses, sentences, words, phrases and so on are underwritten by/structured by or merely interact with cultural values, then there is a sense in which the focus is on language outside the head, in the community. Even cognition works this way, I think. For example, I know somethings about trees. But even greater knowledge, knowledge I can access, about trees is found in my culture. In studying knowledge, I might want to know what people know individually, or I might want to know about the knowledge of cultures, some of which will not be known/mastered by any one member of the culture, and how the cultural knowledge affects/is accessed by, individuals. Values are other things that can have both a psychological and a cultural existence. Dan On 11 Sep 2010, at 21:26, Matthew S. Dryer wrote: > > Dan etc, > > There have unfortunately been two sub-threads with the same subject heading "Re: > [FUNKNET] analysis: unhappiness". Even I have found the > Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have > pursued the other one only because it wasn't clear that others hadn't > misunderstood what I was trying to say. > > Unfortunately, the following comment from Dan seems to illustrate further that I > haven't made myself clear: > > "On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us)." > > BUT, it was Lise who urged that distinction on us. The whole point of my emails > has been to deny such a distinction, to argue that the only reality is the "in > the head" one. In fact, the gradual convergence of thinking in the Gibson-etc > subthread seems to reflect the idea that despite apparent differences, there is a > common underlying goal. > > Matthew > > On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: >> I think that Brian and Dick make excellent points. There are very good >> grammars written that could be improved by psycholinguistic experimentation >> and more quantitative approaches. But large sections of those grammars >> aren't going to change one bit (go-went) with quantitative tests and such >> tests would be completely counterproductive given the shortness of life and >> the vastness of the field linguist's tasks. >> Part of the problem is that linguistics is not simply a subdiscipline of >> psychology. Linguistics has its own objectives and those only occasionally >> overlap with psychology. The same for methods. >> On another note, I don't buy the 'in my head' 'out of my head' distinction >> either (that Matt seems to be urging upon us). We study different things >> and have different reasons for being satisfied with the results we >> achieve. >> I believe that we linguists are often complacent and fail to apply better >> methods. But of course that applies to all disciplines. >> In the meantime, checking corpora, collecting data as a result of careful >> interviews with native speakers, and the other aspects of the field >> linguist's task are vital parts of the linguist's task and much of this >> won't be improved by quantitative methods as we currently understand them. >> Maybe sometime. >> Dan >> >> P.S. In my original reference to Ted and Ev's paper, I said that they >> showed the danger of using intuitions. What I meant to say of using >> intuitions as standardly used by linguists. They convinced me that there is >> a lot to learn from quantitative methods. >> On 10 Sep 2010, at 19:40, Richard Hudson wrote: >> >>> Dear Ted and Ev, >>> Yes, I understand your view, but I think it's a >> psycholinguist's view. Your goal is to find general processes and >> principles that apply uniformly across individuals, so you have to use >> methods to check for generality. And (as you know) I admire the way you >> pursue that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how all the >> bits fit together. Like you, I'm aiming to model cognition, but my focus is >> on items and structures, and I start from the assumption that these can and >> do vary across speakers.> >>> However, having said all that I do agree with >> you that linguists should all get used to collecting and using quantitative >> data; and with the help of Brian MacWhinney's typology we'd know what >> methods to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good starting >> point for a proper investigation.> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> >>> On 10/09/2010 19:30, Ted Gibson >> wrote:>> Dear Dick: >>>> >>>> Perhaps we are talking at cross purposes. I >> don't understand what is confusing about what Ev Fedorenko and I are >> claiming. All we are saying is that if you have some testable claim >> involving a general hypothesis about a language, then you need to get >> quantitative data from unbiased sources to evaluate that claim. If you are >> interested in English past tense morphology, then depending on the claims >> that you might want to investigate, there are lots of ways to get relevant >> quantitative evidence. Corpus data will probably be useful. For very low >> frequency words, you can run experiments to test behavior with respect to >> such words.>> >>>> Your example of the past tense of >> "bid" is a fine such example. You can run an experiment like the >> one you suggested to find out what people think the past tense is. If you >> then found that 20/50 people responded "bidded" and 30/50 respond >> "bid", that is a lot of useful information. As you suggest in >> your discussion, this result wouldn't answer the question of how past tense >> is stored in each individual. This result would be ambiguous among several >> possible explanations. One possibility is that the probability distribution >> that is being discovered reflects different dialects, such that 2/5 of the >> population has one past tense, and 3/5 has another. Another possibility is >> that each person has a similar probability distribution in their heads, >> such that 2/5 of the time I respond one way, and 3/5 of the time I respond >> another. Further experiments would be necessary to answer between these and >> other possible theories (e.g., with repeated trials from the same person, >> carefully planned so that the participants don't notice that they are being >> asked multiple times). Without the quantitative evidence in the first >> place, there is no way to answer these kinds of questions.>> >>>> Regarding the past tense of "go", >> this would be useful as a baseline in an experiment involving the less >> frequent ones. So, yes, it would useful to gather quantitative evidence in >> such a case also, as baselines with respect to the more interesting cases >> for theories.>> >>>> The bottom line: if you have a >> generalization about a language that you wish to evaluate (such that you >> hypothesize that it is true across the speakers of the language), then you >> need quantitative evidence from multiple individuals, using an unbiased >> data collection method, to evaluate such a claim. The point about >> Mechanical Turk is that it is really *easy* to do this now, at least for >> languages like English.>> >>>> Best wishes, >>>> >>>> Ted Gibson & Ev Fedorenko >>>> >>>> On Sep 10, 2010, at 1:59 PM, Richard Hudson >> wrote:>> >>>>> Dear Ted, >>>>> Thanks for the very interesting comment, >> but are you REALLY saying that I shouldn't claim, for example, that the >> past tense of GO is "went" without first cross-checking with 50 >> native speakers?>>> >>>>> Isn't there a danger of missing the >> point that we all, as native speakers, spend our whole lives scanning other >> people's linguistic behaviour (language 'out there', E-language) and trying >> to explain it to ourselves in terms of a language system (language 'in >> here', I-language)? So every judgement we make is based on thousands or >> millions of observed exemplars, and reflects a unique experience of >> E-language filtered through a unique I-language.>>> >>>>> Given that view of language development, >> I don't see how quantitative data will help. Let's take a real uncertainty, >> such as the past tense of BID. If I want to say I did it, do I say "I >> bidded" or "I bid"? My judgement: I don't know. Ok, you get >> 50 people to oblige on Mechanical Turk, and 20 of them give >> "bidded" and 30 "bid". So what? Does that mean that the >> correct answer is "bidded"? Surely not. How is it better than my >> judgement? I agree you could record my speech and find how often I use each >> alternative; but the reason I don't know is precisely because it's a rare >> word, so in a sense quantitative data are irrelevant even there. What would >> solve the problem of subjectivity, of course, would be a machine for >> probing the bit of my mind (or even brain) that holds BID and its details; >> but I suspect that even that wouldn't move us much further forward than my >> original "don't know". (Incidentally I write as a fan of >> quantitative sociolinguistics, so I do accept that quantitative data are >> relevant to linguistic analysis in some areas, where the I-language >> phenomenon is frequent enough to produce usable data.)>>> >>>>> It seems to me that this discussion >> raises the really fundamental question of what kind of thing we think >> language is: social or individual. The problem isn't unique to linguistics >> of course; it's the same throughout the social sciences. But what's special >> about linguistics is that we deal in very fine details of culture (e.g. >> details of how a particular word is used or pronounced) so the differences >> between individuals really matter. I don't see that we're ever going to >> have anything better than judgements to go on, so what we need is a way to >> ensure that judgements are accurate reports of individual I-language. A >> rotten situation for a science, but I don't see how it can get >> better.>>> >>>>> Dick >>>>> >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> >>>>> On 10/09/2010 14:03, Ted Gibson >> wrote:>>>> Dear Dan, Dick: >>>>>> >>>>>> I would like to clarify some points >> that Dan Everett makes, in>>>> response to Dick Hudson. >>>>>> >>>>>> Ev Fedorenko and I have written a >> couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see >> references and links below) on what we>>>> think are weak methodological >> standards in syntax and semantics>>>> research over the past many years. >> The issue that we address is the>>>> prevalent method in syntax and >> semantics research, which involves>>>> obtaining a judgment of the >> acceptability of a sentence / meaning>>>> pair, typically by just the author >> of the paper, sometimes with>>>> feedback from colleagues. As we >> address in our papers, this>>>> methodology does not allow proper >> testing of scientific hypotheses>>>> because of (a) the small number of >> experimental participants>>>> (typically one); (b) the small >> number of experimental stimuli>>>> (typically one); (c) cognitive >> biases on the part of the researcher>>>> and participants; and (d) the effect >> of the preceding context (e.g.,>>>> other constructions the researcher >> may have been recently>>>> considering). (As Dan said, see >> Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & >> Fedorenko, in press; for similar>>>> points, but with not as strong a >> conclusion as ours).>>>> >>>>>> Three issues need to be separated >> here: (1) the use of intuitive>>>> judgments as a dependent measure in >> a language experiment; (2)>>>> potential cognitive biases on the >> part of experimental subjects and>>>> experimenters in language >> experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the >> dependent measure might be. The>>>> paper that Ev and I wrote addresses >> the last two issues, but does not>>>> go into depth on the first issue >> (the use of intuitions as a dependent>>>> measure in language experiments). >> Regarding this issue, we don't think>>>> that there is anything wrong with >> gathering intuitive judgments as a>>>> dependent measure, as long as the >> task is clear to the experimental>>>> participants. >>>>>> >>>>>> In the longer paper (Gibson & >> Fedorenko, in press) we respond to some>>>> arguments that have been given in >> support of continuing to use the>>>> traditional non-quantitative method >> in syntax / semantics research.>>>> One recent defense of the >> traditional method comes from Phillips>>>> (2008), who argues that no harm has >> come from the non-quantitative>>>> approach in syntax research thus >> far. Phillips argues that there are>>>> no cases in the literature where an >> incorrect intuitive judgment has>>>> become the basis for a widely >> accepted generalization or an important>>>> theoretical claim. He therefore >> concludes that there is no reason to>>>> adopt more rigorous data collection >> standards. We challenge Philipsâ��>>>> conclusion by presenting three cases >> from the literature where a>>>> faulty intuition has led to >> incorrect generalizations and mistaken>>>> theorizing, plausibly due to >> cognitive biases on the part of the>>>> researchers. >>>>>> >>>>>> A second argument that is sometimes >> presented for the continued use of>>>> the traditional non-quantitative >> method is that it would be too>>>> inefficient to evaluate every >> syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For >> example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in >> their response to Gibson & Fedorenko>>>> (2010): â��It would cripple >> linguistic investigation if it were required>>>> that all judgments of > ambiguity and >> grammaticality be subject to>>>> statistically rigorous experiments >> on naive subjects, especially when>>>> investigating languages whose >> speakers are hard to accessâ�� (Culicover>>>> & Jackendoff, 2010, p. 234). >> (Dick Hudson makes a similar point>>>> earlier in the discussion here.) >> Whereas we agree that in>>>> circumstances where gathering data >> is difficult, some evidence is>>>> better than no evidence, we do not >> agree that research would be slowed>>>> with respect to languages where >> experimental participants are easy to>>>> access, such as English. In >> contrast, we think that the opposite is>>>> true: the fieldâ��s progress >> is probably slowed by not doing>>>> quantitative research. >>>>>> Suppose that a typical syntax / >> semantics paper that lacks>>>> quantitative evidence includes >> judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 >> or more empirical claims. Even if>>>> most of the judgments from such a >> paper are correct or are on the>>>> right track, the problem is in >> knowing which judgments are correct.>>>> For example, suppose that 90% of the >> judgments from an arbitrary paper>>>> are correct (which is probably a >> high estimate). (Colin Phillips and>>>> some of his former students / >> postdocs have commented to us that, in>>>> their experience, quantitative >> acceptability judgment studies almost>>>> always validate the claim(s) in the >> literature. This is not our>>>> experience, however. Most >> experiments that we have run which attempt>>>> to test some syntactic / semantic >> hypothesis in the literature end up>>>> providing us with a pattern of data >> that had not been known before the>>>> experiment (e.g., Breen et al., in >> press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras >> & Gibson, submitted).) This means>>>> that in a paper with 50 empirical >> claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to >> choose 45 items from 50. Thatâ��s over>>>> two million different theories. By >> quantitatively evaluating the>>>> empirical claims, we reduce the >> uncertainty a great deal. To make>>>> progress, it is better to have >> theoretical claims supported by solid>>>> quantitative evidence, so that even >> if the interpretation of the data>>>> changes over time as new evidence >> becomes available â�� as is often the>>>> case in any field of science >> â�� the empirical pattern can be used as a>>>> basis for further >> theorizing.>>>> >>>>>> Furthermore, it is no longer >> expensive to run behavioral experiments,>>>> at least in English and other widely >> spoken languages. There now>>>> exists a marketplace interface >> â�� Amazon.comâ��s Mechanical Turk â�� which>>>> can be used for collecting >> behavioral data over the internet quickly>>>> and inexpensively. The cost of using >> an interface like this is>>>> minimal, and the time that it takes >> for the results to be returned is>>>> short. For example, currently on >> Mechanical Turk, a survey of>>>> approximately 50 items will be >> answered by 50 or more participants>>>> within a couple of hours, at a cost >> of approximately $1 per>>>> participant. Thus a survey can be >> completed within a day, at a cost of>>>> less than $50. (The hard work of >> designing the experiment, and>>>> constructing controlled materials >> remains of course.)>>>> >>>>>> Sorry to be so verbose. But I think >> that these methodological points>>>> are very important. >>>>>> >>>>>> Best wishes, >>>>>> >>>>>> Ted Gibson >>>>>> >>>>>> Gibson, E. & Fedorenko, E. (In >> press). The need for quantitative>>>> methods in syntax and semantics >> research. Language and Cognitive>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress >> LCP.pdf>>>> >>>>>> Gibson, E. & Fedorenko, E. >> (2010). Weak quantitative standards in>>>> linguistics research. Trends in >> Cognitive Science, 14, 233-234.>>>> > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 > TICS.pdf >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Dick, >>>>>>> >>>>>>> You raise an important issue >> here about methodology. I believe that>>>>> intuitions are a fine way to >> generate hypotheses and even to test>>>>> them - to a degree. But while it >> might not have been feasible for>>>>> Huddleston, Pullum, and the >> other contributors to the Cambridge>>>>> Grammar to conduct experiments >> on every point of the grammar,>>>>> experiments could have only made >> the grammar better. The use of>>>>> intuitions, corpora, and >> standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science >> Methodology) is vital for taking the>>>>> field forward and for providing >> the best support for different>>>>> analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new>>>>> paper on this, showing serious >> shortcomings with intuitions as the>>>>> sole source of evidence, in >> their paper: "The need for quantitative>>>>> methods in syntax and semantics >> research".>>>>> >>>>>>> Carson Schutze and Wayne Cowart, >> among others, have also written>>>>> convincingly on this. >>>>>>> >>>>>>> It is one reason that a team >> from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from >> Brazil are beginning a third round of>>>>> experimental work among the >> Pirahas, since my own work on the syntax>>>>> was, like almost every other >> field researcher's, based on native>>>>> speaker intuitions and >> corpora.>>>>> >>>>>>> The discussion of methodologies >> reminds me of the initial reactions>>>>> to Greenberg's work on >> classifying the languages of the Americas. His>>>>> methods were strongly (and >> justifiably) criticized. However, I always>>>>> thought that his methods were a >> great way of generating hypotheses,>>>>> so long as they were ultimately >> put to the test of standard>>>>> historical linguistics methods. >> And the same seems true for use of>>>>> native-speaker >> intuitions.>>>>> >>>>>>> -- Dan >>>>>> >>>>>> >>>>>> >>>>>>>> We linguists can add a >> further layer of explanation to the>>>>>> judgements, but some >> judgements do seem to be more reliable than>>>>>> others. And if we have to >> wait for psycholinguistic evidence for>>>>>> every detailed analysis we >> make, our whole discipline will>>>>>> immediately grind to a halt. >> Like it or not, native speaker>>>>>> judgements are what put us >> linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing >> the Cambridge Grammar of the English>>>>>> Language (or the OED) >> without using native speaker judgements.>>>>>> >>>>>>>> Best wishes, Dick >> Hudson>>>> >>>>>> >>>>>> >>>> >>>> >>>> >>> >> >> >> >> >> > > From dan at daneverett.org Sun Sep 12 01:35:54 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 21:35:54 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: To clarify, I actually *do* buy the distinction, not as a false dichotomy, not as something that every linguist/psychologist should be concerned with, but in the way I just outlined. Dan On 11 Sep 2010, at 21:32, Daniel Everett wrote: > Sorry, Matt, for being careless in my attribution of that distinction. I do believe that there is such a distinction, however, and that it is very important. If you are after careful and deep understanding of the nature of representations, then the 'in the head' focus is important. If you are more interested (always, occasionally, etc) in how discourses, sentences, words, phrases and so on are underwritten by/structured by or merely interact with cultural values, then there is a sense in which the focus is on language outside the head, in the community. > > Even cognition works this way, I think. For example, I know somethings about trees. But even greater knowledge, knowledge I can access, about trees is found in my culture. In studying knowledge, I might want to know what people know individually, or I might want to know about the knowledge of cultures, some of which will not be known/mastered by any one member of the culture, and how the cultural knowledge affects/is accessed by, individuals. Values are other things that can have both a psychological and a cultural existence. > > Dan > > > > > On 11 Sep 2010, at 21:26, Matthew S. Dryer wrote: > >> >> Dan etc, >> >> There have unfortunately been two sub-threads with the same subject heading "Re: >> [FUNKNET] analysis: unhappiness". Even I have found the >> Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have >> pursued the other one only because it wasn't clear that others hadn't >> misunderstood what I was trying to say. >> >> Unfortunately, the following comment from Dan seems to illustrate further that I >> haven't made myself clear: >> >> "On another note, I don't buy the 'in my head' 'out of my head' distinction >> either (that Matt seems to be urging upon us)." >> >> BUT, it was Lise who urged that distinction on us. The whole point of my emails >> has been to deny such a distinction, to argue that the only reality is the "in >> the head" one. In fact, the gradual convergence of thinking in the Gibson-etc >> subthread seems to reflect the idea that despite apparent differences, there is a >> common underlying goal. >> >> Matthew >> >> On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: >>> I think that Brian and Dick make excellent points. There are very good >>> grammars written that could be improved by psycholinguistic experimentation >>> and more quantitative approaches. But large sections of those grammars >>> aren't going to change one bit (go-went) with quantitative tests and such >>> tests would be completely counterproductive given the shortness of life and >>> the vastness of the field linguist's tasks. >>> Part of the problem is that linguistics is not simply a subdiscipline of >>> psychology. Linguistics has its own objectives and those only occasionally >>> overlap with psychology. The same for methods. >>> On another note, I don't buy the 'in my head' 'out of my head' distinction >>> either (that Matt seems to be urging upon us). We study different things >>> and have different reasons for being satisfied with the results we >>> achieve. >>> I believe that we linguists are often complacent and fail to apply better >>> methods. But of course that applies to all disciplines. >>> In the meantime, checking corpora, collecting data as a result of careful >>> interviews with native speakers, and the other aspects of the field >>> linguist's task are vital parts of the linguist's task and much of this >>> won't be improved by quantitative methods as we currently understand them. >>> Maybe sometime. >>> Dan >>> >>> P.S. In my original reference to Ted and Ev's paper, I said that they >>> showed the danger of using intuitions. What I meant to say of using >>> intuitions as standardly used by linguists. They convinced me that there is >>> a lot to learn from quantitative methods. >>> On 10 Sep 2010, at 19:40, Richard Hudson wrote: >>> >>>> Dear Ted and Ev, >>>> Yes, I understand your view, but I think it's a >>> psycholinguist's view. Your goal is to find general processes and >>> principles that apply uniformly across individuals, so you have to use >>> methods to check for generality. And (as you know) I admire the way you >>> pursue that goal. But my goal, as a linguist, is different. I want to >>> explore the structure of a language so that I can understand how all the >>> bits fit together. Like you, I'm aiming to model cognition, but my focus is >>> on items and structures, and I start from the assumption that these can and >>> do vary across speakers.> >>>> However, having said all that I do agree with >>> you that linguists should all get used to collecting and using quantitative >>> data; and with the help of Brian MacWhinney's typology we'd know what >>> methods to use when. And I do agree with your points about bid/bidded: in >>> cases like that, quantitative data would be at least a very good starting >>> point for a proper investigation.> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> >>>> On 10/09/2010 19:30, Ted Gibson >>> wrote:>> Dear Dick: >>>>> >>>>> Perhaps we are talking at cross purposes. I >>> don't understand what is confusing about what Ev Fedorenko and I are >>> claiming. All we are saying is that if you have some testable claim >>> involving a general hypothesis about a language, then you need to get >>> quantitative data from unbiased sources to evaluate that claim. If you are >>> interested in English past tense morphology, then depending on the claims >>> that you might want to investigate, there are lots of ways to get relevant >>> quantitative evidence. Corpus data will probably be useful. For very low >>> frequency words, you can run experiments to test behavior with respect to >>> such words.>> >>>>> Your example of the past tense of >>> "bid" is a fine such example. You can run an experiment like the >>> one you suggested to find out what people think the past tense is. If you >>> then found that 20/50 people responded "bidded" and 30/50 respond >>> "bid", that is a lot of useful information. As you suggest in >>> your discussion, this result wouldn't answer the question of how past tense >>> is stored in each individual. This result would be ambiguous among several >>> possible explanations. One possibility is that the probability distribution >>> that is being discovered reflects different dialects, such that 2/5 of the >>> population has one past tense, and 3/5 has another. Another possibility is >>> that each person has a similar probability distribution in their heads, >>> such that 2/5 of the time I respond one way, and 3/5 of the time I respond >>> another. Further experiments would be necessary to answer between these and >>> other possible theories (e.g., with repeated trials from the same person, >>> carefully planned so that the participants don't notice that they are being >>> asked multiple times). Without the quantitative evidence in the first >>> place, there is no way to answer these kinds of questions.>> >>>>> Regarding the past tense of "go", >>> this would be useful as a baseline in an experiment involving the less >>> frequent ones. So, yes, it would useful to gather quantitative evidence in >>> such a case also, as baselines with respect to the more interesting cases >>> for theories.>> >>>>> The bottom line: if you have a >>> generalization about a language that you wish to evaluate (such that you >>> hypothesize that it is true across the speakers of the language), then you >>> need quantitative evidence from multiple individuals, using an unbiased >>> data collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at least for >>> languages like English.>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson & Ev Fedorenko >>>>> >>>>> On Sep 10, 2010, at 1:59 PM, Richard Hudson >>> wrote:>> >>>>>> Dear Ted, >>>>>> Thanks for the very interesting comment, >>> but are you REALLY saying that I shouldn't claim, for example, that the >>> past tense of GO is "went" without first cross-checking with 50 >>> native speakers?>>> >>>>>> Isn't there a danger of missing the >>> point that we all, as native speakers, spend our whole lives scanning other >>> people's linguistic behaviour (language 'out there', E-language) and trying >>> to explain it to ourselves in terms of a language system (language 'in >>> here', I-language)? So every judgement we make is based on thousands or >>> millions of observed exemplars, and reflects a unique experience of >>> E-language filtered through a unique I-language.>>> >>>>>> Given that view of language development, >>> I don't see how quantitative data will help. Let's take a real uncertainty, >>> such as the past tense of BID. If I want to say I did it, do I say "I >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get >>> 50 people to oblige on Mechanical Turk, and 20 of them give >>> "bidded" and 30 "bid". So what? Does that mean that the >>> correct answer is "bidded"? Surely not. How is it better than my >>> judgement? I agree you could record my speech and find how often I use each >>> alternative; but the reason I don't know is precisely because it's a rare >>> word, so in a sense quantitative data are irrelevant even there. What would >>> solve the problem of subjectivity, of course, would be a machine for >>> probing the bit of my mind (or even brain) that holds BID and its details; >>> but I suspect that even that wouldn't move us much further forward than my >>> original "don't know". (Incidentally I write as a fan of >>> quantitative sociolinguistics, so I do accept that quantitative data are >>> relevant to linguistic analysis in some areas, where the I-language >>> phenomenon is frequent enough to produce usable data.)>>> >>>>>> It seems to me that this discussion >>> raises the really fundamental question of what kind of thing we think >>> language is: social or individual. The problem isn't unique to linguistics >>> of course; it's the same throughout the social sciences. But what's special >>> about linguistics is that we deal in very fine details of culture (e.g. >>> details of how a particular word is used or pronounced) so the differences >>> between individuals really matter. I don't see that we're ever going to >>> have anything better than judgements to go on, so what we need is a way to >>> ensure that judgements are accurate reports of individual I-language. A >>> rotten situation for a science, but I don't see how it can get >>> better.>>> >>>>>> Dick >>>>>> >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> >>>>>> On 10/09/2010 14:03, Ted Gibson >>> wrote:>>>> Dear Dan, Dick: >>>>>>> >>>>>>> I would like to clarify some points >>> that Dan Everett makes, in>>>> response to Dick Hudson. >>>>>>> >>>>>>> Ev Fedorenko and I have written a >>> couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see >>> references and links below) on what we>>>> think are weak methodological >>> standards in syntax and semantics>>>> research over the past many years. >>> The issue that we address is the>>>> prevalent method in syntax and >>> semantics research, which involves>>>> obtaining a judgment of the >>> acceptability of a sentence / meaning>>>> pair, typically by just the author >>> of the paper, sometimes with>>>> feedback from colleagues. As we >>> address in our papers, this>>>> methodology does not allow proper >>> testing of scientific hypotheses>>>> because of (a) the small number of >>> experimental participants>>>> (typically one); (b) the small >>> number of experimental stimuli>>>> (typically one); (c) cognitive >>> biases on the part of the researcher>>>> and participants; and (d) the effect >>> of the preceding context (e.g.,>>>> other constructions the researcher >>> may have been recently>>>> considering). (As Dan said, see >>> Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & >>> Fedorenko, in press; for similar>>>> points, but with not as strong a >>> conclusion as ours).>>>> >>>>>>> Three issues need to be separated >>> here: (1) the use of intuitive>>>> judgments as a dependent measure in >>> a language experiment; (2)>>>> potential cognitive biases on the >>> part of experimental subjects and>>>> experimenters in language >>> experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the >>> dependent measure might be. The>>>> paper that Ev and I wrote addresses >>> the last two issues, but does not>>>> go into depth on the first issue >>> (the use of intuitions as a dependent>>>> measure in language experiments). >>> Regarding this issue, we don't think>>>> that there is anything wrong with >>> gathering intuitive judgments as a>>>> dependent measure, as long as the >>> task is clear to the experimental>>>> participants. >>>>>>> >>>>>>> In the longer paper (Gibson & >>> Fedorenko, in press) we respond to some>>>> arguments that have been given in >>> support of continuing to use the>>>> traditional non-quantitative method >>> in syntax / semantics research.>>>> One recent defense of the >>> traditional method comes from Phillips>>>> (2008), who argues that no harm has >>> come from the non-quantitative>>>> approach in syntax research thus >>> far. Phillips argues that there are>>>> no cases in the literature where an >>> incorrect intuitive judgment has>>>> become the basis for a widely >>> accepted generalization or an important>>>> theoretical claim. He therefore >>> concludes that there is no reason to>>>> adopt more rigorous data collection >>> standards. We challenge Philipsâ��>>>> conclusion by presenting three cases >>> from the literature where a>>>> faulty intuition has led to >>> incorrect generalizations and mistaken>>>> theorizing, plausibly due to >>> cognitive biases on the part of the>>>> researchers. >>>>>>> >>>>>>> A second argument that is sometimes >>> presented for the continued use of>>>> the traditional non-quantitative >>> method is that it would be too>>>> inefficient to evaluate every >>> syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For >>> example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in >>> their response to Gibson & Fedorenko>>>> (2010): â��It would cripple >>> linguistic investigation if it were required>>>> that all judgments of >> ambiguity and >>> grammaticality be subject to>>>> statistically rigorous experiments >>> on naive subjects, especially when>>>> investigating languages whose >>> speakers are hard to accessâ�� (Culicover>>>> & Jackendoff, 2010, p. 234). >>> (Dick Hudson makes a similar point>>>> earlier in the discussion here.) >>> Whereas we agree that in>>>> circumstances where gathering data >>> is difficult, some evidence is>>>> better than no evidence, we do not >>> agree that research would be slowed>>>> with respect to languages where >>> experimental participants are easy to>>>> access, such as English. In >>> contrast, we think that the opposite is>>>> true: the fieldâ��s progress >>> is probably slowed by not doing>>>> quantitative research. >>>>>>> Suppose that a typical syntax / >>> semantics paper that lacks>>>> quantitative evidence includes >>> judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 >>> or more empirical claims. Even if>>>> most of the judgments from such a >>> paper are correct or are on the>>>> right track, the problem is in >>> knowing which judgments are correct.>>>> For example, suppose that 90% of the >>> judgments from an arbitrary paper>>>> are correct (which is probably a >>> high estimate). (Colin Phillips and>>>> some of his former students / >>> postdocs have commented to us that, in>>>> their experience, quantitative >>> acceptability judgment studies almost>>>> always validate the claim(s) in the >>> literature. This is not our>>>> experience, however. Most >>> experiments that we have run which attempt>>>> to test some syntactic / semantic >>> hypothesis in the literature end up>>>> providing us with a pattern of data >>> that had not been known before the>>>> experiment (e.g., Breen et al., in >>> press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras >>> & Gibson, submitted).) This means>>>> that in a paper with 50 empirical >>> claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to >>> choose 45 items from 50. Thatâ��s over>>>> two million different theories. By >>> quantitatively evaluating the>>>> empirical claims, we reduce the >>> uncertainty a great deal. To make>>>> progress, it is better to have >>> theoretical claims supported by solid>>>> quantitative evidence, so that even >>> if the interpretation of the data>>>> changes over time as new evidence >>> becomes available â�� as is often the>>>> case in any field of science >>> â�� the empirical pattern can be used as a>>>> basis for further >>> theorizing.>>>> >>>>>>> Furthermore, it is no longer >>> expensive to run behavioral experiments,>>>> at least in English and other widely >>> spoken languages. There now>>>> exists a marketplace interface >>> â�� Amazon.comâ��s Mechanical Turk â�� which>>>> can be used for collecting >>> behavioral data over the internet quickly>>>> and inexpensively. The cost of using >>> an interface like this is>>>> minimal, and the time that it takes >>> for the results to be returned is>>>> short. For example, currently on >>> Mechanical Turk, a survey of>>>> approximately 50 items will be >>> answered by 50 or more participants>>>> within a couple of hours, at a cost >>> of approximately $1 per>>>> participant. Thus a survey can be >>> completed within a day, at a cost of>>>> less than $50. (The hard work of >>> designing the experiment, and>>>> constructing controlled materials >>> remains of course.)>>>> >>>>>>> Sorry to be so verbose. But I think >>> that these methodological points>>>> are very important. >>>>>>> >>>>>>> Best wishes, >>>>>>> >>>>>>> Ted Gibson >>>>>>> >>>>>>> Gibson, E. & Fedorenko, E. (In >>> press). The need for quantitative>>>> methods in syntax and semantics >>> research. Language and Cognitive>>>> Processes. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress >>> LCP.pdf>>>> >>>>>>> Gibson, E. & Fedorenko, E. >>> (2010). Weak quantitative standards in>>>> linguistics research. Trends in >>> Cognitive Science, 14, 233-234.>>>> >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 >> TICS.pdf >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Dick, >>>>>>>> >>>>>>>> You raise an important issue >>> here about methodology. I believe that>>>>> intuitions are a fine way to >>> generate hypotheses and even to test>>>>> them - to a degree. But while it >>> might not have been feasible for>>>>> Huddleston, Pullum, and the >>> other contributors to the Cambridge>>>>> Grammar to conduct experiments >>> on every point of the grammar,>>>>> experiments could have only made >>> the grammar better. The use of>>>>> intuitions, corpora, and >>> standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science >>> Methodology) is vital for taking the>>>>> field forward and for providing >>> the best support for different>>>>> analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new>>>>> paper on this, showing serious >>> shortcomings with intuitions as the>>>>> sole source of evidence, in >>> their paper: "The need for quantitative>>>>> methods in syntax and semantics >>> research".>>>>> >>>>>>>> Carson Schutze and Wayne Cowart, >>> among others, have also written>>>>> convincingly on this. >>>>>>>> >>>>>>>> It is one reason that a team >>> from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from >>> Brazil are beginning a third round of>>>>> experimental work among the >>> Pirahas, since my own work on the syntax>>>>> was, like almost every other >>> field researcher's, based on native>>>>> speaker intuitions and >>> corpora.>>>>> >>>>>>>> The discussion of methodologies >>> reminds me of the initial reactions>>>>> to Greenberg's work on >>> classifying the languages of the Americas. His>>>>> methods were strongly (and >>> justifiably) criticized. However, I always>>>>> thought that his methods were a >>> great way of generating hypotheses,>>>>> so long as they were ultimately >>> put to the test of standard>>>>> historical linguistics methods. >>> And the same seems true for use of>>>>> native-speaker >>> intuitions.>>>>> >>>>>>>> -- Dan >>>>>>> >>>>>>> >>>>>>> >>>>>>>>> We linguists can add a >>> further layer of explanation to the>>>>>> judgements, but some >>> judgements do seem to be more reliable than>>>>>> others. And if we have to >>> wait for psycholinguistic evidence for>>>>>> every detailed analysis we >>> make, our whole discipline will>>>>>> immediately grind to a halt. >>> Like it or not, native speaker>>>>>> judgements are what put us >>> linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing >>> the Cambridge Grammar of the English>>>>>> Language (or the OED) >>> without using native speaker judgements.>>>>>> >>>>>>>>> Best wishes, Dick >>> Hudson>>>> >>>>>>> >>>>>>> >>>>> >>>>> >>>>> >>>> >>> >>> >>> >>> >>> >> >> > > From bischoff.st at gmail.com Sun Sep 12 02:30:20 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Sat, 11 Sep 2010 22:30:20 -0400 Subject: FUNKNET Digest, Vol 84, Issue 10 In-Reply-To: Message-ID: Ted & Ev write... ____________ We have encountered a claim that the reason for different kinds of methods being used across the different fields of language study (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the research questions are different across these fields, and some methods may be better suited to ask some questions than others. Although the latter is likely true, the premise ? that the research questions are different across the fields ? is false... In our opinion then, the distinction between the fields of linguistics and psycho-/neuro-linguistics is purely along the lines of the kind of data that are used as evidence for or against theoretical hypotheses: typically non-quantitative data in linguistics vs. typically quantitative data in psycho-/neuro-linguistics. Given the superficial nature of this distinction, we think that there should be one field of language study where a wide range of dependent measures is used to investigate linguistic representations and computations. _____________ I would be curious to know how many on this listserv agree with the above...it seems to me that the assertion rests on a very narrow definition of "linguistics"...if language is viewed as a meso-oject and linguistics thus as a meso-science (perhaps it is not), it doesn't seem that the assertion necessarily holds...I'm not convinced that all linguistic inquiry is about computations and algorithms (despite my research being primarily in formal and computational linguistics)...nor am I convinced that the root of all linguistic inquiry is grounded in representations and computations (though I would be curious to know what "representations" refers to here as I could be misreading it)...in addition it seems that the object of study in neuro-linguistics and phsycholinguistics is the mind (possibly) and the brain (certainly) but for many linguists that doesn't seem to be the case and I do think this leads to different reseach questions...but then again I'm the one that started all this by asking the naive question about "unhappiness" in the first place. Cheers, Shannon On Sat, Sep 11, 2010 at 1:00 PM, wrote: > Send FUNKNET mailing list submissions to > funknet at mailman.rice.edu > > To subscribe or unsubscribe via the World Wide Web, visit > https://mailman.rice.edu/mailman/listinfo/funknet > or, via email, send a message with subject or body 'help' to > funknet-request at mailman.rice.edu > > You can reach the person managing the list at > funknet-owner at mailman.rice.edu > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of FUNKNET digest..." > > > Today's Topics: > > 1. Re: "Relative clauses" with no relativized element > (Suzanne Kemmer) > 2. Re: "Relative clauses" with no relativized element (Ron Kuzar) > 3. Re: "Relative clauses" with no relativized element (Tom Givon) > 4. Re: "Relative clauses" with no relativized element > (Giuliana Fiorentino) > 5. Re: "Relative clauses" with no relativized ele (E.G.) > 6. Re: "Relative clauses" with no relativized element > (Suzanne Kemmer) > 7. Re: "Relative clauses" with no relativized element (E.G.) > 8. Re: analysis: unhappiness (Richard Hudson) > 9. Re: analysis: unhappiness (Ted Gibson) > 10. Re: analysis: unhappiness (dryer at buffalo.edu) > 11. Re: analysis: unhappiness (A. Katz) > 12. Re: "Relative clauses" with no relativized ele (Ron Kuzar) > 13. Re: analysis: unhappiness (Brian MacWhinney) > 14. Re: analysis: unhappiness (dryer at buffalo.edu) > 15. Re: analysis: unhappiness (Richard Hudson) > 16. Re: analysis: unhappiness (Richard Hudson) > 17. Re: analysis: unhappiness (Lise Menn) > 18. Re: analysis: unhappiness (Lise Menn) > 19. Re: analysis: unhappiness (Daniel Everett) > 20. Re: "Relative clauses" with no relativized element > (Philippe De Brabanter) > 21. Re: analysis: unhappiness (Richard Hudson) > 22. Re: analysis: unhappiness (A. Katz) > 23. Re: FUNKNET] analysis: unhappiness (Chris Butler) > 24. Re: FUNKNET] analysis: unhappiness (A. Katz) > 25. Re: "Relative clauses" with no relativized ele (E.G.) > 26. Job Advertisement (Kristine Hildebrandt) > 27. Re: analysis: unhappiness (Ted Gibson) > 28. Re: analysis: unhappiness (Daniel Everett) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 10 Sep 2010 12:16:09 -0500 > From: Suzanne Kemmer > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: FUNKNET > Message-ID: > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > In Generative Syntax these clauses were viewed as complement clauses > with an NP head, distinct > from relative clauses but having some parallels with them. I think it > was Joan Bresnan that > brought out the parallels and distinctions, maybe in her doctoral > dissertation . As I recall (but > my remembrance may be faulty), Bresnan named the > THAT element a COMP for complementizer. > > The term 'appositive' isn't very good because in traditional grammar > that is reserved for an 'UNrestrictive' relation of a noun and its > complement--an incidental description of > a head N's referent rather than a specification of which referent > ("the tree, a live oak, survived another 100 years or so"). > > In Cognitive Grammar nouns like claim, statement, idea, realization, > belief etc. are in almost all cases nominalizations of 'viewing > predicates' > (verbs like claim, believe, etc.) that introduce on-stage > predications 'viewed' by a conceptualizer (the person doing the > claiming, etc.). (the viewing > predicates are space builders in Fauconnier's mental spaces terminology) > > For the nominalizations of these predicates, the semantics of the > nouns intrinsically has an "e-site" or elaboration site > that allows for spelling out the content of the viewed predicate in > the form of a complement clause. The e-site > inherent to the semantics of the nouns is parallel to the e-site > inherent to the semantics of the corresponding verbs. > > There are a few cases I can think of of nouns that have 'viewing > predicate' e-sites but don't have corresponding verbs . > For example > the noun _view_ "The view that global climate change is > anthropogenic is widely held by scientists" > ( ' X views that (proposition)' is not possible, only 'X views Y > as ...' , with a restriction to equative or descriptive propositions). > Also _idea_----the verb has to be changed to something like 'believe' > to make a corresponding full predicate. > > I view (!) these nouns as semantically parallel in interesting ways to > picture nouns. The conceptualizer (viewer) in > both cases can designate the noun in a possessive phrase, but after > that the syntax diverges. > > --Suzanne > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > > And as another addition: the clauses that can only be introduced by > > *that* (with no role to > > play in the subordinate clause) may be seen as (subtypes of) > > complement clauses, > > expressing a proposition with the noun functioning as Complement > > Taking Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, > > etc. (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not > > relatives. Cf. constructions like > > "The claim is that X" (traditionally analysed as subject clauses), > > "I claim that X", "I put forward > > the claim that X", in which the relationship between the verb or > > noun and the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > >> To add to Joanne's comments: > >> > >> There are basically three ways to distinguish an appositive clause > >> (AC) from a relative clause (RC). > >> > >> 1) An AC can only be led by *that* while an RC can be led by other > >> pronouns. > >> > >> 2) The AC and the noun it modifies display an equative > >> relationship--one can say X > >> (denoted by the noun) is Y (presented by the appositive)--while an > >> RC often doesn't > >> (except, perhaps, when the relative clause is sentential). > > > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but > >> a relative pronoun > >> is always part of the clause in an RC. > >> > >> Rong Chen > >> > > > > > > > > ------------------------------ > > Message: 2 > Date: Fri, 10 Sep 2010 20:42:24 +0300 > From: Ron Kuzar > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: FUNKNET > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > A thorough discussion of the head nouns and their relation with their > complement clauses may be found in Hans-Joerg Schmid's book on Shell > Nouns (this is his term for the head nouns). > Ron Kuzar > --------- > On Fri, Sep 10, 2010 at 8:16 PM, Suzanne Kemmer wrote: > > > In Generative Syntax these clauses were viewed as complement clauses with > > an NP head, distinct > > from relative clauses but having some parallels with them. I think it > was > > Joan Bresnan that > > brought out the parallels and distinctions, maybe in her doctoral > > dissertation . As I recall (but > > my remembrance may be faulty), Bresnan named the > > THAT element a COMP for complementizer. > > > > The term 'appositive' isn't very good because in traditional grammar > > that is reserved for an 'UNrestrictive' relation of a noun and its > > complement--an incidental description of > > a head N's referent rather than a specification of which referent ("the > > tree, a live oak, survived another 100 years or so"). > > > > In Cognitive Grammar nouns like claim, statement, idea, realization, > > belief etc. are in almost all cases nominalizations of 'viewing > predicates' > > (verbs like claim, believe, etc.) that introduce on-stage predications > > 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the > > viewing > > predicates are space builders in Fauconnier's mental spaces terminology) > > > > For the nominalizations of these predicates, the semantics of the nouns > > intrinsically has an "e-site" or elaboration site > > that allows for spelling out the content of the viewed predicate in the > > form of a complement clause. The e-site > > inherent to the semantics of the nouns is parallel to the e-site inherent > > to the semantics of the corresponding verbs. > > > > There are a few cases I can think of of nouns that have 'viewing > > predicate' e-sites but don't have corresponding verbs . > > For example > > the noun _view_ "The view that global climate change is anthropogenic > is > > widely held by scientists" > > ( ' X views that (proposition)' is not possible, only 'X views Y as ...' > , > > with a restriction to equative or descriptive propositions). > > Also _idea_----the verb has to be changed to something like 'believe' to > > make a corresponding full predicate. > > > > I view (!) these nouns as semantically parallel in interesting ways to > > picture nouns. The conceptualizer (viewer) in > > both cases can designate the noun in a possessive phrase, but after that > > the syntax diverges. > > > > --Suzanne > > > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > > > And as another addition: the clauses that can only be introduced by > *that* > >> (with no role to > >> play in the subordinate clause) may be seen as (subtypes of) complement > >> clauses, > >> expressing a proposition with the noun functioning as Complement Taking > >> Predicate (CTP), > >> expressing a propostional attitude, epistemic/evaluative stance, etc. > >> (following analyses by > >> Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > >> constructions like > >> "The claim is that X" (traditionally analysed as subject clauses), "I > >> claim that X", "I put forward > >> the claim that X", in which the relationship between the verb or noun > and > >> the that-clause is > >> comparable to the one in "The claim that X". > >> > >> --Arie Verhagen > >> > >> ---------------- > >> Message from Rong Chen > >> 10 Sep 2010, 23:42 > >> > >> Subject: Re: [FUNKNET] "Relative clauses" with no relativi > >> > >> To add to Joanne's comments: > >>> > >>> There are basically three ways to distinguish an appositive clause > >>> (AC) from a relative clause (RC). > >>> > >>> 1) An AC can only be led by *that* while an RC can be led by other > >>> pronouns. > >>> > >>> 2) The AC and the noun it modifies display an equative > relationship--one > >>> can say X > >>> (denoted by the noun) is Y (presented by the appositive)--while an RC > >>> often doesn't > >>> (except, perhaps, when the relative clause is sentential). > >>> > >> > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but a > >>> relative pronoun > >>> is always part of the clause in an RC. > >>> > >>> Rong Chen > >>> > >>> > >> > >> > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > ------------------------------ > > Message: 3 > Date: Fri, 10 Sep 2010 11:43:34 -0600 > From: Tom Givon > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: "Thomas E. Payne" > Cc: FUNKNET > Message-ID: <4C8A6E46.3040301 at uoregon.edu> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), > vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this > construction described and analyzed as the product of nominalization of > clauses with verbs that take verbal complements ('know', 'think', 'say' > etc. The preceding section (6.6.3. "Complex NP's arising through > nominalization", p. 287) deals more generally with nominalizations. The > term "noun complements" was used in syntax classes at UCLA in the mid > 1960s, so certainly Joan Bresnan did not invent it. Best, TG > > =========== > > Thomas E. Payne wrote: > > Can anyone help me name the following structure in English, and maybe > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > > > ------------------------------ > > Message: 4 > Date: Fri, 10 Sep 2010 19:53:15 +0200 > From: "Giuliana Fiorentino" > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: "Thomas E. Payne" , "FUNKNET" > > Message-ID: <036386D3933D4CB99389FFB2576ADBE8 at giuliana> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > > > ------------------------------ > > Message: 5 > Date: Fri, 10 Sep 2010 20:54:06 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: Arie Verhagen , > funknet at mailman.rice.edu > Message-ID: > > > > Content-Type: text/plain; charset=ISO-8859-1 > > Hi all, > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > comparison can help what seems to be a purely theoretical question based on > a single language. The problem here is that English uses the same element > to > mark regular relatives and these "appositional" relatives. But if at least > one language encodes them by different means, then there's at least a good > case for seeing them as distinct functions. It's basically the same > principle that's used to decide whether to put a meaning on a semantic map. > So here are two languages that I know that encode them differently. > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > clause (ki), which differs from the relative clause marker (Se-), e.g. > > ha-hoda'a Se-kibalnu > the-announcment rel-we_got > "The announcement that we got." > > ha-hoda'a ki hitbatel ha-mifgaS > the-message CMP was_cancelled the-meeting > "The announcement that the meeting was cancelled." > > In Coptic, these clauses are marked by ce-, which marks complement clauses, > *inter alia*, but not relative clauses: > > ph-mewi ce- (complement clause) > 'the-thought that (we are angry)' > > ph-mewi ete- (relative clause) > 'the thought that (we used to think)' > > This seems to be a pretty clear indication that these are complement > clauses > rather than relatives. Even if one doesn't like the notion of nouns taking > complement clauses (and why not? nominalizations in some languages can take > accusative modifiers as well as genitives), it still probably isn't > incidental that the nominalizations are of verbs that take complement > clauses when finite. > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > looking at. > > Best, > Eitan > > > On 10 September 2010 19:21, Arie Verhagen > wrote: > > > And as another addition: the clauses that can only be introduced by > *that* > > (with no role to > > play in the subordinate clause) may be seen as (subtypes of) complement > > clauses, > > expressing a proposition with the noun functioning as Complement Taking > > Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > > constructions like > > "The claim is that X" (traditionally analysed as subject clauses), "I > claim > > that X", "I put forward > > the claim that X", in which the relationship between the verb or noun and > > the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > To add to Joanne's comments: > > > > > > There are basically three ways to distinguish an appositive clause > > > (AC) from a relative clause (RC). > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > pronouns. > > > > > > 2) The AC and the noun it modifies display an equative > relationship--one > > can say X > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > often doesn't > > > (except, perhaps, when the relative clause is sentential). > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > relative pronoun > > > is always part of the clause in an RC. > > > > > > Rong Chen > > > > > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 6 > Date: Fri, 10 Sep 2010 12:55:30 -0500 > From: Suzanne Kemmer > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: Funknet > Message-ID: > Content-Type: text/plain; charset=us-ascii > > > > Talmy, yes Joan B did not invent noun complements nor > the term for them; but my recollection, in passing, was that she > named 'that' in such structures as "complementizer" (and I also > recall that she referred to 'that' relativizers with the same term, > while recognizing other differences between the two structures.) > > I may be wrong on that, but it's a different recollection, claim, or > whatever, > than the one you refer to. > > Not all the head nouns are nominalizations, but most are. > S. > > On Sep 10, 2010, at 12:43 PM, Tom Givon wrote: > > > > > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. > I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this > construction described and analyzed as the product of nominalization of > clauses with verbs that take verbal complements ('know', 'think', 'say' etc. > The preceding section (6.6.3. "Complex NP's arising through nominalization", > p. 287) deals more generally with nominalizations. The term "noun > complements" was used in syntax classes at UCLA in the mid 1960s, so > certainly Joan Bresnan did not invent it. Best, TG > > > > =========== > > > > Thomas E. Payne wrote: > >> Can anyone help me name the following structure in English, and maybe > point > >> me to some references? I do not find reference to this in the Cambridge > >> Grammar of the English Language or any other of my English grammar > books. > >> But then, maybe I just don't know where to look. > >> > >> Here are two examples from a play: > >> > >> His protestations of devotion in the trial scene are, in our opinion, > >> genuine, as is his confession [that his affair with the Countess is > >> platonic]. > >> > >> The bracketed clause seems to modify "confession", though there is no > >> position for a confession in the clause itself. > >> > >> . . . forced hither with an impious black design [to have my innocence > and > >> youth become the sacrifice of brutal violence]. > >> > >> Here the bracketed non-finite clause seems to modify "design." > >> > >> These are not all that rare. I'm reminded of examples like: > >> > >> "The claim [that my client is a murderer] is totally false." > >> > >> Are these relative clauses? If so what kind? Thanks for any help. > >> > >> Tom Payne > >> > >> > > > > > > > > ------------------------------ > > Message: 7 > Date: Fri, 10 Sep 2010 20:56:23 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: funknet at mailman.rice.edu > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Jespersen and his nexus-substantives should be mentioned (Philosophy of > Grammar, 1924). Also in his MEG and Analytic Syntax one could find > interesting discussions. > > Eitan > > > On 10 September 2010 20:53, Giuliana Fiorentino < > giuliana.fiorentino at unimol.it> wrote: > > > Hi Tom, > > clauses like: > > > > The importance of being Earnest > > the fact of being late > > the fact that you are late > > the idea that world is round > > etcetera > > > > are not relative clauses but can be considered among syntactic strategies > > in order to nominalise events after a generic noun (working as a > classifier > > for nominalised events). > > > > Giuliana > > > > ----- Original Message ----- > > From: Thomas E. Payne > > To: FUNKNET > > Sent: Friday, September 10, 2010 4:16 PM > > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > > > > Can anyone help me name the following structure in English, and maybe > > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar > books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 8 > Date: Fri, 10 Sep 2010 18:59:12 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Message-ID: <4C8A71F0.4050507 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" > without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic > behaviour (language 'out there', E-language) and trying to explain it to > ourselves in terms of a language system (language 'in here', > I-language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language. > > Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical > Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean > that the correct answer is "bidded"? Surely not. How is it better than > my judgement? I agree you could record my speech and find how often I > use each alternative; but the reason I don't know is precisely because > it's a rare word, so in a sense quantitative data are irrelevant even > there. What would solve the problem of subjectivity, of course, would be > a machine for probing the bit of my mind (or even brain) that holds BID > and its details; but I suspect that even that wouldn't move us much > further forward than my original "don't know". (Incidentally I write as > a fan of quantitative sociolinguistics, so I do accept that quantitative > data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or > individual. The problem isn't unique to linguistics of course; it's the > same throughout the social sciences. But what's special about > linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the > differences between individuals really matter. I don't see that we're > ever going to have anything better than judgements to go on, so what we > need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see > how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: > > Dear Dan, Dick: > > > > I would like to clarify some points that Dan Everett makes, in > > response to Dick Hudson. > > > > Ev Fedorenko and I have written a couple of papers recently (Gibson & > > Fedorenko, 2010, in press, see references and links below) on what we > > think are weak methodological standards in syntax and semantics > > research over the past many years. The issue that we address is the > > prevalent method in syntax and semantics research, which involves > > obtaining a judgment of the acceptability of a sentence / meaning > > pair, typically by just the author of the paper, sometimes with > > feedback from colleagues. As we address in our papers, this > > methodology does not allow proper testing of scientific hypotheses > > because of (a) the small number of experimental participants > > (typically one); (b) the small number of experimental stimuli > > (typically one); (c) cognitive biases on the part of the researcher > > and participants; and (d) the effect of the preceding context (e.g., > > other constructions the researcher may have been recently > > considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > > several others cited in Gibson & Fedorenko, in press; for similar > > points, but with not as strong a conclusion as ours). > > > > Three issues need to be separated here: (1) the use of intuitive > > judgments as a dependent measure in a language experiment; (2) > > potential cognitive biases on the part of experimental subjects and > > experimenters in language experiments; and (3) the need for obtaining > > quantitative evidence, whatever the dependent measure might be. The > > paper that Ev and I wrote addresses the last two issues, but does not > > go into depth on the first issue (the use of intuitions as a dependent > > measure in language experiments). Regarding this issue, we don't think > > that there is anything wrong with gathering intuitive judgments as a > > dependent measure, as long as the task is clear to the experimental > > participants. > > > > In the longer paper (Gibson & Fedorenko, in press) we respond to some > > arguments that have been given in support of continuing to use the > > traditional non-quantitative method in syntax / semantics research. > > One recent defense of the traditional method comes from Phillips > > (2008), who argues that no harm has come from the non-quantitative > > approach in syntax research thus far. Phillips argues that there are > > no cases in the literature where an incorrect intuitive judgment has > > become the basis for a widely accepted generalization or an important > > theoretical claim. He therefore concludes that there is no reason to > > adopt more rigorous data collection standards. We challenge Philips? > > conclusion by presenting three cases from the literature where a > > faulty intuition has led to incorrect generalizations and mistaken > > theorizing, plausibly due to cognitive biases on the part of the > > researchers. > > > > A second argument that is sometimes presented for the continued use of > > the traditional non-quantitative method is that it would be too > > inefficient to evaluate every syntactic / semantic hypothesis or > > phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > > make this argument explicitly in their response to Gibson & Fedorenko > > (2010): ?It would cripple linguistic investigation if it were required > > that all judgments of ambiguity and grammaticality be subject to > > statistically rigorous experiments on naive subjects, especially when > > investigating languages whose speakers are hard to access? (Culicover > > & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > > earlier in the discussion here.) Whereas we agree that in > > circumstances where gathering data is difficult, some evidence is > > better than no evidence, we do not agree that research would be slowed > > with respect to languages where experimental participants are easy to > > access, such as English. In contrast, we think that the opposite is > > true: the field?s progress is probably slowed by not doing > > quantitative research. > > Suppose that a typical syntax / semantics paper that lacks > > quantitative evidence includes judgments for 50 or more sentences / > > meaning pairs, corresponding to 50 or more empirical claims. Even if > > most of the judgments from such a paper are correct or are on the > > right track, the problem is in knowing which judgments are correct. > > For example, suppose that 90% of the judgments from an arbitrary paper > > are correct (which is probably a high estimate). (Colin Phillips and > > some of his former students / postdocs have commented to us that, in > > their experience, quantitative acceptability judgment studies almost > > always validate the claim(s) in the literature. This is not our > > experience, however. Most experiments that we have run which attempt > > to test some syntactic / semantic hypothesis in the literature end up > > providing us with a pattern of data that had not been known before the > > experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > > press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > > that in a paper with 50 empirical claims 45/50 are correct. But which > > 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > > two million different theories. By quantitatively evaluating the > > empirical claims, we reduce the uncertainty a great deal. To make > > progress, it is better to have theoretical claims supported by solid > > quantitative evidence, so that even if the interpretation of the data > > changes over time as new evidence becomes available ? as is often the > > case in any field of science ? the empirical pattern can be used as a > > basis for further theorizing. > > > > Furthermore, it is no longer expensive to run behavioral experiments, > > at least in English and other widely spoken languages. There now > > exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > > can be used for collecting behavioral data over the internet quickly > > and inexpensively. The cost of using an interface like this is > > minimal, and the time that it takes for the results to be returned is > > short. For example, currently on Mechanical Turk, a survey of > > approximately 50 items will be answered by 50 or more participants > > within a couple of hours, at a cost of approximately $1 per > > participant. Thus a survey can be completed within a day, at a cost of > > less than $50. (The hard work of designing the experiment, and > > constructing controlled materials remains of course.) > > > > Sorry to be so verbose. But I think that these methodological points > > are very important. > > > > Best wishes, > > > > Ted Gibson > > > > Gibson, E. & Fedorenko, E. (In press). The need for quantitative > > methods in syntax and semantics research. Language and Cognitive > > Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > > & Fedorenko InPress LCP.pdf > > > > Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > > linguistics research. Trends in Cognitive Science, 14, 233-234. > > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > > 2010 TICS.pdf > > > > > > > > > >> Dick, > >> > >> You raise an important issue here about methodology. I believe that > >> intuitions are a fine way to generate hypotheses and even to test > >> them - to a degree. But while it might not have been feasible for > >> Huddleston, Pullum, and the other contributors to the Cambridge > >> Grammar to conduct experiments on every point of the grammar, > >> experiments could have only made the grammar better. The use of > >> intuitions, corpora, and standard psycholinguistic experimentation > >> (indeed, Standard Social Science Methodology) is vital for taking the > >> field forward and for providing the best support for different > >> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >> paper on this, showing serious shortcomings with intuitions as the > >> sole source of evidence, in their paper: "The need for quantitative > >> methods in syntax and semantics research". > >> > >> Carson Schutze and Wayne Cowart, among others, have also written > >> convincingly on this. > >> > >> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >> Science), and researchers from Brazil are beginning a third round of > >> experimental work among the Pirahas, since my own work on the syntax > >> was, like almost every other field researcher's, based on native > >> speaker intuitions and corpora. > >> > >> The discussion of methodologies reminds me of the initial reactions > >> to Greenberg's work on classifying the languages of the Americas. His > >> methods were strongly (and justifiably) criticized. However, I always > >> thought that his methods were a great way of generating hypotheses, > >> so long as they were ultimately put to the test of standard > >> historical linguistics methods. And the same seems true for use of > >> native-speaker intuitions. > >> > >> -- Dan > > > > > > > >>> We linguists can add a further layer of explanation to the > >>> judgements, but some judgements do seem to be more reliable than > >>> others. And if we have to wait for psycholinguistic evidence for > >>> every detailed analysis we make, our whole discipline will > >>> immediately grind to a halt. Like it or not, native speaker > >>> judgements are what put us linguists ahead of the rest in handling > >>> fine detail. Imagine writing the Cambridge Grammar of the English > >>> Language (or the OED) without using native speaker judgements. > >>> > >>> Best wishes, Dick Hudson > > > > > > > > > ------------------------------ > > Message: 9 > Date: Fri, 10 Sep 2010 14:30:16 -0400 > From: Ted Gibson > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Evelina Fedorenko , funknet > > Message-ID: > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dear Dick: > > Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are > saying is that if you have some testable claim involving a general > hypothesis about a language, then you need to get quantitative data > from unbiased sources to evaluate that claim. If you are interested > in English past tense morphology, then depending on the claims that > you might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very > low frequency words, you can run experiments to test behavior with > respect to such words. > > Your example of the past tense of "bid" is a fine such example. You > can run an experiment like the one you suggested to find out what > people think the past tense is. If you then found that 20/50 people > responded "bidded" and 30/50 respond "bid", that is a lot of useful > information. As you suggest in your discussion, this result wouldn't > answer the question of how past tense is stored in each individual. > This result would be ambiguous among several possible explanations. > One possibility is that the probability distribution that is being > discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution > in their heads, such that 2/5 of the time I respond one way, and 3/5 > of the time I respond another. Further experiments would be necessary > to answer between these and other possible theories (e.g., with > repeated trials from the same person, carefully planned so that the > participants don't notice that they are being asked multiple times). > Without the quantitative evidence in the first place, there is no way > to answer these kinds of questions. > > Regarding the past tense of "go", this would be useful as a baseline > in an experiment involving the less frequent ones. So, yes, it would > useful to gather quantitative evidence in such a case also, as > baselines with respect to the more interesting cases for theories. > > The bottom line: if you have a generalization about a language that > you wish to evaluate (such that you hypothesize that it is true across > the speakers of the language), then you need quantitative evidence > from multiple individuals, using an unbiased data collection method, > to evaluate such a claim. The point about Mechanical Turk is that it > is really *easy* to do this now, at least for languages like English. > > Best wishes, > > Ted Gibson & Ev Fedorenko > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > Dear Ted, > > Thanks for the very interesting comment, but are you REALLY saying > > that I shouldn't claim, for example, that the past tense of GO is > > "went" without first cross-checking with 50 native speakers? > > > > Isn't there a danger of missing the point that we all, as native > > speakers, spend our whole lives scanning other people's linguistic > > behaviour (language 'out there', E-language) and trying to explain > > it to ourselves in terms of a language system (language 'in here', I- > > language)? So every judgement we make is based on thousands or > > millions of observed exemplars, and reflects a unique experience of > > E-language filtered through a unique I-language. > > > > Given that view of language development, I don't see how > > quantitative data will help. Let's take a real uncertainty, such as > > the past tense of BID. If I want to say I did it, do I say "I > > bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > > people to oblige on Mechanical Turk, and 20 of them give "bidded" > > and 30 "bid". So what? Does that mean that the correct answer is > > "bidded"? Surely not. How is it better than my judgement? I agree > > you could record my speech and find how often I use each > > alternative; but the reason I don't know is precisely because it's a > > rare word, so in a sense quantitative data are irrelevant even > > there. What would solve the problem of subjectivity, of course, > > would be a machine for probing the bit of my mind (or even brain) > > that holds BID and its details; but I suspect that even that > > wouldn't move us much further forward than my original "don't know". > > (Incidentally I write as a fan of quantitative sociolinguistics, so > > I do accept that quantitative data are relevant to linguistic > > analysis in some areas, where the I-language phenomenon is frequent > > enough to produce usable data.) > > > > It seems to me that this discussion raises the really fundamental > > question of what kind of thing we think language is: social or > > individual. The problem isn't unique to linguistics of course; it's > > the same throughout the social sciences. But what's special about > > linguistics is that we deal in very fine details of culture (e.g. > > details of how a particular word is used or pronounced) so the > > differences between individuals really matter. I don't see that > > we're ever going to have anything better than judgements to go on, > > so what we need is a way to ensure that judgements are accurate > > reports of individual I-language. A rotten situation for a science, > > but I don't see how it can get better. > > > > Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 14:03, Ted Gibson wrote: > >> Dear Dan, Dick: > >> > >> I would like to clarify some points that Dan Everett makes, in > >> response to Dick Hudson. > >> > >> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >> Fedorenko, 2010, in press, see references and links below) on what we > >> think are weak methodological standards in syntax and semantics > >> research over the past many years. The issue that we address is the > >> prevalent method in syntax and semantics research, which involves > >> obtaining a judgment of the acceptability of a sentence / meaning > >> pair, typically by just the author of the paper, sometimes with > >> feedback from colleagues. As we address in our papers, this > >> methodology does not allow proper testing of scientific hypotheses > >> because of (a) the small number of experimental participants > >> (typically one); (b) the small number of experimental stimuli > >> (typically one); (c) cognitive biases on the part of the researcher > >> and participants; and (d) the effect of the preceding context (e.g., > >> other constructions the researcher may have been recently > >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >> several others cited in Gibson & Fedorenko, in press; for similar > >> points, but with not as strong a conclusion as ours). > >> > >> Three issues need to be separated here: (1) the use of intuitive > >> judgments as a dependent measure in a language experiment; (2) > >> potential cognitive biases on the part of experimental subjects and > >> experimenters in language experiments; and (3) the need for obtaining > >> quantitative evidence, whatever the dependent measure might be. The > >> paper that Ev and I wrote addresses the last two issues, but does not > >> go into depth on the first issue (the use of intuitions as a > >> dependent > >> measure in language experiments). Regarding this issue, we don't > >> think > >> that there is anything wrong with gathering intuitive judgments as a > >> dependent measure, as long as the task is clear to the experimental > >> participants. > >> > >> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >> arguments that have been given in support of continuing to use the > >> traditional non-quantitative method in syntax / semantics research. > >> One recent defense of the traditional method comes from Phillips > >> (2008), who argues that no harm has come from the non-quantitative > >> approach in syntax research thus far. Phillips argues that there are > >> no cases in the literature where an incorrect intuitive judgment has > >> become the basis for a widely accepted generalization or an important > >> theoretical claim. He therefore concludes that there is no reason to > >> adopt more rigorous data collection standards. We challenge Philips? > >> conclusion by presenting three cases from the literature where a > >> faulty intuition has led to incorrect generalizations and mistaken > >> theorizing, plausibly due to cognitive biases on the part of the > >> researchers. > >> > >> A second argument that is sometimes presented for the continued use > >> of > >> the traditional non-quantitative method is that it would be too > >> inefficient to evaluate every syntactic / semantic hypothesis or > >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >> make this argument explicitly in their response to Gibson & Fedorenko > >> (2010): ?It would cripple linguistic investigation if it were > >> required > >> that all judgments of ambiguity and grammaticality be subject to > >> statistically rigorous experiments on naive subjects, especially when > >> investigating languages whose speakers are hard to access? (Culicover > >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >> earlier in the discussion here.) Whereas we agree that in > >> circumstances where gathering data is difficult, some evidence is > >> better than no evidence, we do not agree that research would be > >> slowed > >> with respect to languages where experimental participants are easy to > >> access, such as English. In contrast, we think that the opposite is > >> true: the field?s progress is probably slowed by not doing > >> quantitative research. > >> Suppose that a typical syntax / semantics paper that lacks > >> quantitative evidence includes judgments for 50 or more sentences / > >> meaning pairs, corresponding to 50 or more empirical claims. Even if > >> most of the judgments from such a paper are correct or are on the > >> right track, the problem is in knowing which judgments are correct. > >> For example, suppose that 90% of the judgments from an arbitrary > >> paper > >> are correct (which is probably a high estimate). (Colin Phillips and > >> some of his former students / postdocs have commented to us that, in > >> their experience, quantitative acceptability judgment studies almost > >> always validate the claim(s) in the literature. This is not our > >> experience, however. Most experiments that we have run which attempt > >> to test some syntactic / semantic hypothesis in the literature end up > >> providing us with a pattern of data that had not been known before > >> the > >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >> that in a paper with 50 empirical claims 45/50 are correct. But which > >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >> two million different theories. By quantitatively evaluating the > >> empirical claims, we reduce the uncertainty a great deal. To make > >> progress, it is better to have theoretical claims supported by solid > >> quantitative evidence, so that even if the interpretation of the data > >> changes over time as new evidence becomes available ? as is often the > >> case in any field of science ? the empirical pattern can be used as a > >> basis for further theorizing. > >> > >> Furthermore, it is no longer expensive to run behavioral experiments, > >> at least in English and other widely spoken languages. There now > >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >> can be used for collecting behavioral data over the internet quickly > >> and inexpensively. The cost of using an interface like this is > >> minimal, and the time that it takes for the results to be returned is > >> short. For example, currently on Mechanical Turk, a survey of > >> approximately 50 items will be answered by 50 or more participants > >> within a couple of hours, at a cost of approximately $1 per > >> participant. Thus a survey can be completed within a day, at a cost > >> of > >> less than $50. (The hard work of designing the experiment, and > >> constructing controlled materials remains of course.) > >> > >> Sorry to be so verbose. But I think that these methodological points > >> are very important. > >> > >> Best wishes, > >> > >> Ted Gibson > >> > >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >> methods in syntax and semantics research. Language and Cognitive > >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >> & Fedorenko InPress LCP.pdf > >> > >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >> linguistics research. Trends in Cognitive Science, 14, 233-234. > >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >> Fedorenko > >> 2010 TICS.pdf > >> > >> > >> > >> > >>> Dick, > >>> > >>> You raise an important issue here about methodology. I believe that > >>> intuitions are a fine way to generate hypotheses and even to test > >>> them - to a degree. But while it might not have been feasible for > >>> Huddleston, Pullum, and the other contributors to the Cambridge > >>> Grammar to conduct experiments on every point of the grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, corpora, and standard psycholinguistic experimentation > >>> (indeed, Standard Social Science Methodology) is vital for taking > >>> the > >>> field forward and for providing the best support for different > >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>> paper on this, showing serious shortcomings with intuitions as the > >>> sole source of evidence, in their paper: "The need for quantitative > >>> methods in syntax and semantics research". > >>> > >>> Carson Schutze and Wayne Cowart, among others, have also written > >>> convincingly on this. > >>> > >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>> Science), and researchers from Brazil are beginning a third round of > >>> experimental work among the Pirahas, since my own work on the syntax > >>> was, like almost every other field researcher's, based on native > >>> speaker intuitions and corpora. > >>> > >>> The discussion of methodologies reminds me of the initial reactions > >>> to Greenberg's work on classifying the languages of the Americas. > >>> His > >>> methods were strongly (and justifiably) criticized. However, I > >>> always > >>> thought that his methods were a great way of generating hypotheses, > >>> so long as they were ultimately put to the test of standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions. > >>> > >>> -- Dan > >> > >> > >> > >>>> We linguists can add a further layer of explanation to the > >>>> judgements, but some judgements do seem to be more reliable than > >>>> others. And if we have to wait for psycholinguistic evidence for > >>>> every detailed analysis we make, our whole discipline will > >>>> immediately grind to a halt. Like it or not, native speaker > >>>> judgements are what put us linguists ahead of the rest in handling > >>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>> Language (or the OED) without using native speaker judgements. > >>>> > >>>> Best wishes, Dick Hudson > >> > >> > >> > > > > ------------------------------ > > Message: 10 > Date: Fri, 10 Sep 2010 14:51:45 -0400 > From: dryer at buffalo.edu > Subject: Re: [FUNKNET] analysis: unhappiness > To: Lise Menn , Funknet > > Cc: Richard Hudson > Message-ID: <2147483647.1284130304 at cast-dryerm2.caset.buffalo.edu> > Content-Type: text/plain; charset=us-ascii; format=flowed > > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have to be > quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the correct > analysis", one in terms of what is in people's heads, the other in terms of > what is "out there". > > There are a vast number of patterns "out there". The only distinction, on > my view, amongst this vast number of patterns, is between those that > reflect something inside people's heads and those that don't. But if that > is the case, then there is no coherent sense in which one can talk of "the > correct analysis" of what is "out there", except in terms of what is in > people's heads, and thus no second sense of "the correct analysis". The > patterns that don't correspond to things in people's heads fall into (at > least) two categories. There are those that are akin to constellations of > stars and, as with constellations, there is no reality to these patterns, > except in the minds of linguists. And there are those patterns which are > the fossil remains of what was in the heads of speakers of an earlier stage > of the language but which no longer are. These latter patterns are real, > and they are relevant to exlaining why the language is now the way it is, > but they are not relevant, I think many would agree, as to what is the > "correct analysis" of the language today. > > For this reason, I claim that the only sense in which an analysis can be > "the correct analysis" is in terms of what is inside of people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion of > these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > > > I wish we had better terminology for keeping track of whether, at a > > given time, we are talking about the patterns that are 'out there' in > > the language and might possibly be apprehended (subconsciously) by a > > speaker, and when we are talking about the patterns that a particular > > speaker actually does apprehend, as indicated by experiments, from > > simple 'wug tests' up to brain wave and eye-gaze studies. And for > > distinguishing among the degrees of pattern apprehension that a person > > may have, from vague preferences detectable in reaction times or other > > behavior all the way up through clear metalinguistic insights. Dick > > Hudson's note reminding us of the Gleitman and Gleitman study is right > > on target. > > > > Since we don't in fact have such an agreed-on terminology, we have > to > > be quite careful in making clear what we are referring to when we talk > > about 'the correct analysis' of a form like 'unhappiness'. We know, but > > tend to forget - and tend to forget to tell our students! - that it's an > > empirical question as to whether the formal simplicity and coherence of > > description of forms 'out there' (e.g. lovely abstract morphophonemics) > > is any kind of approximation to the way knowledge of the same forms is > > organized in a particular person's head. If we remember that a very > > large proportion of what we know about our language is 'out there' when > > we are infants and has to be internalized through experience with the > > language (even if you believe in innate 'core language'), the variation > > in internal knowledge from one person to another is more understandable. > > > > We especially need to consider (and try to test) the possibility > that > > since > > the brain can make multiple cross-connections, multiple patterns are > > involved > > simultaneously in morphological and syntactic analyses. I suggest that > > that's the case with 'unhappiness' - and the linguistic analyses that I > > know about are not good at handling that kind of idea. > > > > Lise > > > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > > >> > >> Two comments. > >> > >> First (elaborating perhaps on Dick Hudson's comment), I think there > >> is an > >> important distinction between low-level linguistic intuitions (like > >> whether a > >> word or sentence is well-formed or what it means) and higher-level > >> intuitions > >> (like what the structure of a word or sentence is). One can take > >> the position > >> that we need to account for the former (while recognizing that they > >> are not > >> always reliable) but not the latter. > >> > >> Second, the tension here is not only between evidence from speaker > >> intuitions > >> versus evidence from psycholinguistic experiments. There is also a > >> tension > >> between deciding on the correct analysis on the basis of a priori > >> simplicity > >> arguments versus deciding on the correct analysis on the basis of > >> psycholinguistic evidence (see Derwing 1973). The bracketing > >> paradox that Dan > >> referred to that arises with the word (semantics argues > >> for > >> [[un+happi] + er], morphology and phonology argues for [un + [happi > >> +er]] (the > >> comparative suffix can only be attached to adjectives containing one > >> or two > >> syllables) is only a paradox if one assumes that speakers adopt the > >> simplest > >> analysis. For example, if speakers adopt a more complex rule for > >> either of these > >> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >> trisyllabic > >> words beginning with un-), then the bracketing paradox disappears. > >> > >> Matthew > >> > >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>> conscious judgements are only one kind of evidence that we need to > >>> take > >>> into account. I admire Carson Schutze's work (which I reviewed in > >>> fact), > >>> and of course I've been aware of complaints about judgements by > >>> people > >>> like Labov for decades. > >>> > >>> But you're missing my main point, which is that all judgements aren't > >>> equally reliable. If you want to know how /unhappiness/ is > >>> structured, > >>> ask a linguist, not a five-year old. And one of the by-products of > >>> education may be increased sensitivity to syntax - which is one of > >>> the > >>> many reasons why linguists need to pay more attention to education. > >>> > >>> Best wishes, Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>> Dick, > >>>> > >>>> You raise an important issue here about > >>> methodology. I believe that intuitions are a fine way to generate > >>> hypotheses and even to test them - to a degree. But while it might > >>> not have > >>> been feasible for Huddleston, Pullum, and the other contributors to > >>> the > >>> Cambridge Grammar to conduct experiments on every point of the > >>> grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, > >>> corpora, and standard psycholinguistic experimentation (indeed, > >>> Standard > >>> Social Science Methodology) is vital for taking the field forward > >>> and for > >>> providing the best support for different analyses. Ted Gibson and Ev > >>> Fedorenko have written a very useful new paper on this, showing > >>> serious > >>> shortcomings with intuitions as the sole source of evidence, in their > >>> paper: "The need for quantitative methods in syntax and semantics > >>> research".> > >>>> Carson Schutze and Wayne Cowart, among others, > >>> have also written convincingly on this.> > >>>> It is one reason that a team from Stanford, MIT > >>> (Brain and Cognitive Science), and researchers from Brazil are > >>> beginning a > >>> third round of experimental work among the Pirahas, since my own > >>> work on > >>> the syntax was, like almost every other field researcher's, based > >>> on native > >>> speaker intuitions and corpora.> > >>>> The discussion of methodologies reminds me of > >>> the initial reactions to Greenberg's work on classifying the > >>> languages of > >>> the Americas. His methods were strongly (and justifiably) criticized. > >>> However, I always thought that his methods were a great way of > >>> generating > >>> hypotheses, so long as they were ultimately put to the test of > >>> standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions.> > >>>> -- Dan > >>>> > >>>>> We linguists can add a further layer of > >>> explanation to the judgements, but some judgements do seem to be more > >>> reliable than others. And if we have to wait for psycholinguistic > >>> evidence > >>> for every detailed analysis we make, our whole discipline will > >>> immediately > >>> grind to a halt. Like it or not, native speaker judgements are what > >>> put us > >>> linguists ahead of the rest in handling fine detail. Imagine > >>> writing the > >>> Cambridge Grammar of the English Language (or the OED) without > >>> using native > >>> speaker judgements.>> > >>>>> Best wishes, Dick Hudson > >>>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> > >> > > > > Lise Menn Home Office: 303-444-4274 > > 1625 Mariposa Ave Fax: 303-413-0017 > > Boulder CO 80302 > > > > Professor Emerita of Linguistics > > Fellow, Institute of Cognitive Science > > University of Colorado > > > > Secretary, AAAS Section Z [Linguistics] > > > > Campus Mail Address: > > UCB 594, Institute of Cognitive Science > > > > Campus Physical Address: > > CINC 234 > > 1777 Exposition Ave, Boulder > > > > > > > > > > > > > > > ------------------------------ > > Message: 11 > Date: Fri, 10 Sep 2010 12:09:07 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > Matthew, > > Thanks for stating that, because I was almost beginning to imagine that > there was no essential disagreement, and that all of us agree that there > is more -- and less -- to language than what is found in people's heads. > > Your position is the one I am familiar with from the functionalist point > of view, and I was beginning to feel that it was underrepresented on > Funknet. > > Those of us who disagree with your stated position -- but are very > familiar with it -- are interested not just in psycholinguistics and how > people process language -- but also in the communicative function of > language as a system whereby information is transferred. Just as you and I > may not be aware of the way our emails are encoded and then decoded by the > computers that help us send emails back and forth, speakers may be > compeltely unaware of what language does in order to transmit information. > > After speakers have finished sending forth their linguistic output, it > matters not at all how they arrived at this output. Language processing > is separate from language in the same way that data processing is separate > from data. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > > > The following sentence of Lise's > > > > "Since we don't in fact have such an agreed-on terminology, we have to > be > > quite careful in making clear what we are referring to when we talk > about > > 'the correct analysis' of a form like 'unhappiness'" > > > > suggests that there are two senses in which an analysis can be "the > correct > > analysis", one in terms of what is in people's heads, the other in terms > of > > what is "out there". > > > > There are a vast number of patterns "out there". The only distinction, > on my > > view, amongst this vast number of patterns, is between those that reflect > > something inside people's heads and those that don't. But if that is the > > case, then there is no coherent sense in which one can talk of "the > correct > > analysis" of what is "out there", except in terms of what is in people's > > heads, and thus no second sense of "the correct analysis". The patterns > that > > don't correspond to things in people's heads fall into (at least) two > > categories. There are those that are akin to constellations of stars > and, as > > with constellations, there is no reality to these patterns, except in the > > minds of linguists. And there are those patterns which are the fossil > > remains of what was in the heads of speakers of an earlier stage of the > > language but which no longer are. These latter patterns are real, and > they > > are relevant to exlaining why the language is now the way it is, but they > are > > not relevant, I think many would agree, as to what is the "correct > analysis" > > of the language today. > > > > For this reason, I claim that the only sense in which an analysis can be > "the > > correct analysis" is in terms of what is inside of people's heads. > > > > Again, I recommend the work of Bruce Derwing for lengthy discussion of > these > > issues. > > > > Matthew > > > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > > wrote: > > > >> I wish we had better terminology for keeping track of whether, at a > >> given time, we are talking about the patterns that are 'out there' in > >> the language and might possibly be apprehended (subconsciously) by a > >> speaker, and when we are talking about the patterns that a particular > >> speaker actually does apprehend, as indicated by experiments, from > >> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >> distinguishing among the degrees of pattern apprehension that a person > >> may have, from vague preferences detectable in reaction times or other > >> behavior all the way up through clear metalinguistic insights. Dick > >> Hudson's note reminding us of the Gleitman and Gleitman study is right > >> on target. > >> > >> Since we don't in fact have such an agreed-on terminology, we have > to > >> be quite careful in making clear what we are referring to when we talk > >> about 'the correct analysis' of a form like 'unhappiness'. We know, > but > >> tend to forget - and tend to forget to tell our students! - that it's > an > >> empirical question as to whether the formal simplicity and coherence of > >> description of forms 'out there' (e.g. lovely abstract morphophonemics) > >> is any kind of approximation to the way knowledge of the same forms is > >> organized in a particular person's head. If we remember that a very > >> large proportion of what we know about our language is 'out there' when > >> we are infants and has to be internalized through experience with the > >> language (even if you believe in innate 'core language'), the variation > >> in internal knowledge from one person to another is more > understandable. > >> We especially need to consider (and try to test) the > >> possibility that > >> since > >> the brain can make multiple cross-connections, multiple patterns are > >> involved > >> simultaneously in morphological and syntactic analyses. I suggest that > >> that's the case with 'unhappiness' - and the linguistic analyses that I > >> know about are not good at handling that kind of idea. > >> > >> Lise > >> > >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> > >>> > >>> Two comments. > >>> > >>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>> is an > >>> important distinction between low-level linguistic intuitions (like > >>> whether a > >>> word or sentence is well-formed or what it means) and higher-level > >>> intuitions > >>> (like what the structure of a word or sentence is). One can take > >>> the position > >>> that we need to account for the former (while recognizing that they > >>> are not > >>> always reliable) but not the latter. > >>> > >>> Second, the tension here is not only between evidence from speaker > >>> intuitions > >>> versus evidence from psycholinguistic experiments. There is also a > >>> tension > >>> between deciding on the correct analysis on the basis of a priori > >>> simplicity > >>> arguments versus deciding on the correct analysis on the basis of > >>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>> paradox that Dan > >>> referred to that arises with the word (semantics argues > >>> for > >>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>> +er]] (the > >>> comparative suffix can only be attached to adjectives containing one > >>> or two > >>> syllables) is only a paradox if one assumes that speakers adopt the > >>> simplest > >>> analysis. For example, if speakers adopt a more complex rule for > >>> either of these > >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>> trisyllabic > >>> words beginning with un-), then the bracketing paradox disappears. > >>> > >>> Matthew > >>> > >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>>> conscious judgements are only one kind of evidence that we need to > >>>> take > >>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>> fact), > >>>> and of course I've been aware of complaints about judgements by > >>>> people > >>>> like Labov for decades. > >>>> > >>>> But you're missing my main point, which is that all judgements aren't > >>>> equally reliable. If you want to know how /unhappiness/ is > >>>> structured, > >>>> ask a linguist, not a five-year old. And one of the by-products of > >>>> education may be increased sensitivity to syntax - which is one of > >>>> the > >>>> many reasons why linguists need to pay more attention to education. > >>>> > >>>> Best wishes, Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about > >>>> methodology. I believe that intuitions are a fine way to generate > >>>> hypotheses and even to test them - to a degree. But while it might > >>>> not have > >>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>> the > >>>> Cambridge Grammar to conduct experiments on every point of the > >>>> grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, > >>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>> Standard > >>>> Social Science Methodology) is vital for taking the field forward > >>>> and for > >>>> providing the best support for different analyses. Ted Gibson and Ev > >>>> Fedorenko have written a very useful new paper on this, showing > >>>> serious > >>>> shortcomings with intuitions as the sole source of evidence, in their > >>>> paper: "The need for quantitative methods in syntax and semantics > >>>> research".> > >>>>> Carson Schutze and Wayne Cowart, among others, > >>>> have also written convincingly on this.> > >>>>> It is one reason that a team from Stanford, MIT > >>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>> beginning a > >>>> third round of experimental work among the Pirahas, since my own > >>>> work on > >>>> the syntax was, like almost every other field researcher's, based > >>>> on native > >>>> speaker intuitions and corpora.> > >>>>> The discussion of methodologies reminds me of > >>>> the initial reactions to Greenberg's work on classifying the > >>>> languages of > >>>> the Americas. His methods were strongly (and justifiably) criticized. > >>>> However, I always thought that his methods were a great way of > >>>> generating > >>>> hypotheses, so long as they were ultimately put to the test of > >>>> standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions.> > >>>>> -- Dan > >>>>> > >>>>>> We linguists can add a further layer of > >>>> explanation to the judgements, but some judgements do seem to be more > >>>> reliable than others. And if we have to wait for psycholinguistic > >>>> evidence > >>>> for every detailed analysis we make, our whole discipline will > >>>> immediately > >>>> grind to a halt. Like it or not, native speaker judgements are what > >>>> put us > >>>> linguists ahead of the rest in handling fine detail. Imagine > >>>> writing the > >>>> Cambridge Grammar of the English Language (or the OED) without > >>>> using native > >>>> speaker judgements.>> > >>>>>> Best wishes, Dick Hudson > >>>>>> > >>>>> > >>>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>> > >> > >> Lise Menn Home Office: 303-444-4274 > >> 1625 Mariposa Ave Fax: 303-413-0017 > >> Boulder CO 80302 > >> > >> Professor Emerita of Linguistics > >> Fellow, Institute of Cognitive Science > >> University of Colorado > >> > >> Secretary, AAAS Section Z [Linguistics] > >> > >> Campus Mail Address: > >> UCB 594, Institute of Cognitive Science > >> > >> Campus Physical Address: > >> CINC 234 > >> 1777 Exposition Ave, Boulder > >> > >> > >> > >> > > > > > > > > > > > > > ------------------------------ > > Message: 12 > Date: Fri, 10 Sep 2010 23:26:18 +0300 > From: Ron Kuzar > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: FUNKNET > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > The Modern Hebrew data supplied by Eitan are incomplete. > Hebrew distinguishes between locution (say, hear, think, etc.) and > situation (action, event, state, etc.). > What Eitan describes is only true with regard to nouns (and clauses) > expressing locution. 'Announcement' is indeed such a noun. > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > etc., whose denotatum is a situation, cannot be followed by ki, but only > by Se-, e.g.: > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > annoys me the-situation that-all went > 'I am upset about the situation that all have gone' > > On the other hand, the relative Se- may be replaced by the more > elegant and classical aSer, while the Se- of situation clauses may not. > Sorry about the invented example. I am overseas now. > All this has been described (with corpus data) in: > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit > [Hebrew > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > The article is somewhat outdated and contains some inaccuracies I would > formulate differently today, but the basic distinction is valid in my > opinion. > Best, > Ron Kuzar > --------------- > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > Hi all, > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > comparison can help what seems to be a purely theoretical question based > on > > a single language. The problem here is that English uses the same element > > to > > mark regular relatives and these "appositional" relatives. But if at > least > > one language encodes them by different means, then there's at least a > good > > case for seeing them as distinct functions. It's basically the same > > principle that's used to decide whether to put a meaning on a semantic > map. > > So here are two languages that I know that encode them differently. > > > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > ha-hoda'a Se-kibalnu > > the-announcment rel-we_got > > "The announcement that we got." > > > > ha-hoda'a ki hitbatel ha-mifgaS > > the-message CMP was_cancelled the-meeting > > "The announcement that the meeting was cancelled." > > > > In Coptic, these clauses are marked by ce-, which marks complement > clauses, > > *inter alia*, but not relative clauses: > > > > ph-mewi ce- (complement clause) > > 'the-thought that (we are angry)' > > > > ph-mewi ete- (relative clause) > > 'the thought that (we used to think)' > > > > This seems to be a pretty clear indication that these are complement > > clauses > > rather than relatives. Even if one doesn't like the notion of nouns > taking > > complement clauses (and why not? nominalizations in some languages can > take > > accusative modifiers as well as genitives), it still probably isn't > > incidental that the nominalizations are of verbs that take complement > > clauses when finite. > > > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > > looking at. > > > > Best, > > Eitan > > > > > > On 10 September 2010 19:21, Arie Verhagen > > wrote: > > > > > And as another addition: the clauses that can only be introduced by > > *that* > > > (with no role to > > > play in the subordinate clause) may be seen as (subtypes of) complement > > > clauses, > > > expressing a proposition with the noun functioning as Complement Taking > > > Predicate (CTP), > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > (following analyses by > > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > > Cf. > > > constructions like > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > claim > > > that X", "I put forward > > > the claim that X", in which the relationship between the verb or noun > and > > > the that-clause is > > > comparable to the one in "The claim that X". > > > > > > --Arie Verhagen > > > > > > ---------------- > > > Message from Rong Chen > > > 10 Sep 2010, 23:42 > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > To add to Joanne's comments: > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > (AC) from a relative clause (RC). > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > pronouns. > > > > > > > > 2) The AC and the noun it modifies display an equative > > relationship--one > > > can say X > > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > > often doesn't > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > > relative pronoun > > > > is always part of the clause in an RC. > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > -- > > Eitan Grossman > > Martin Buber Society of Fellows > > Hebrew University of Jerusalem > > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > ------------------------------ > > Message: 13 > Date: Fri, 10 Sep 2010 18:23:54 -0400 > From: Brian MacWhinney > Subject: Re: [FUNKNET] analysis: unhappiness > To: Funknet > Message-ID: > Content-Type: text/plain; charset=windows-1252 > > Dick and Ted, > I agree with parts of what each of you are saying. Which means that I > also disagree with other parts. In practice,, Gibson and Fedorenko, in > press, (which I downloaded and scanned) deals with no more than two or > three constructions. They mention the fact that people don't have problems > with sentences such as "Susan muttered him the news" despite claims that > verbs such as "mutter" cannot take the double object construction. They > also note that the claims from Jackendoff and Culicover about the > differences between the two sentences below are not supported by results > from the Mechanical Turk: > 1. Peter was trying to remember who carried what. > 2. Peter was trying to remember who carried what when. > These are interesting facts. If these sentences are supposed to be > different and people judge them to be similarly grammatical, then theories > based on the supposed differences should be reexamined. There are big > chunks of syntactic theory resting on shaky judgments about complex > sentences of this type. Getting some of this straight would be a big win, I > would say, particularly if linguists would pay attention to the results. > But I understand Dick's worry about how far Gibson and Fedorenko are > trying to push this. Neither their email nor their paper sets clear limits > on what we should be testing and we certainly don't want to waste time > checking out go-goed-went. So, Gibson and Fedorenko owe us those > clarifications. > But, Dick, you then move on to questioning data on bid-bidded. Here we > have a case of true variation in the population. I would love to know its > distribution. As a "fan of quantitative sociolinguistics" shouldn't you > too? > My take on this is that constructions are not created equal. The three > types mentioned here are probably just a start on an inventory of > evidentiary types. We need to correctly pair up appropriate methods with > each of the types. And we to make sure that people pay attention to the > results, once they are in > > --Brian MacWhinney > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > Dear Ted, > > Thanks for the very interesting comment, but are you REALLY saying that I > shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > > > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > > > > Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quantit > ative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > > > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > > > > Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 14:03, Ted Gibson wrote: > >> Dear Dan, Dick: > >> > >> I would like to clarify some points that Dan Everett makes, in > >> response to Dick Hudson. > >> > >> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >> Fedorenko, 2010, in press, see references and links below) on what we > >> think are weak methodological standards in syntax and semantics > >> research over the past many years. The issue that we address is the > >> prevalent method in syntax and semantics research, which involves > >> obtaining a judgment of the acceptability of a sentence / meaning > >> pair, typically by just the author of the paper, sometimes with > >> feedback from colleagues. As we address in our papers, this > >> methodology does not allow proper testing of scientific hypotheses > >> because of (a) the small number of experimental participants > >> (typically one); (b) the small number of experimental stimuli > >> (typically one); (c) cognitive biases on the part of the researcher > >> and participants; and (d) the effect of the preceding context (e.g., > >> other constructions the researcher may have been recently > >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >> several others cited in Gibson & Fedorenko, in press; for similar > >> points, but with not as strong a conclusion as ours). > >> > >> Three issues need to be separated here: (1) the use of intuitive > >> judgments as a dependent measure in a language experiment; (2) > >> potential cognitive biases on the part of experimental subjects and > >> experimenters in language experiments; and (3) the need for obtaining > >> quantitative evidence, whatever the dependent measure might be. The > >> paper that Ev and I wrote addresses the last two issues, but does not > >> go into depth on the first issue (the use of intuitions as a dependent > >> measure in language experiments). Regarding this issue, we don't think > >> that there is anything wrong with gathering intuitive judgments as a > >> dependent measure, as long as the task is clear to the experimental > >> participants. > >> > >> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >> arguments that have been given in support of continuing to use the > >> traditional non-quantitative method in syntax / semantics research. > >> One recent defense of the traditional method comes from Phillips > >> (2008), who argues that no harm has come from the non-quantitative > >> approach in syntax research thus far. Phillips argues that there are > >> no cases in the literature where an incorrect intuitive judgment has > >> become the basis for a widely accepted generalization or an important > >> theoretical claim. He therefore concludes that there is no reason to > >> adopt more rigorous data collection standards. We challenge Philips? > >> conclusion by presenting three cases from the literature where a > >> faulty intuition has led to incorrect generalizations and mistaken > >> theorizing, plausibly due to cognitive biases on the part of the > >> researchers. > >> > >> A second argument that is sometimes presented for the continued use of > >> the traditional non-quantitative method is that it would be too > >> inefficient to evaluate every syntactic / semantic hypothesis or > >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >> make this argument explicitly in their response to Gibson & Fedorenko > >> (2010): ?It would cripple linguistic investigation if it were required > >> that all judgments of ambiguity and grammaticality be subject to > >> statistically rigorous experiments on naive subjects, especially when > >> investigating languages whose speakers are hard to access? (Culicover > >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >> earlier in the discussion here.) Whereas we agree that in > >> circumstances where gathering data is difficult, some evidence is > >> better than no evidence, we do not agree that research would be slowed > >> with respect to languages where experimental participants are easy to > >> access, such as English. In contrast, we think that the opposite is > >> true: the field?s progress is probably slowed by not doing > >> quantitative research. > >> Suppose that a typical syntax / semantics paper that lacks > >> quantitative evidence includes judgments for 50 or more sentences / > >> meaning pairs, corresponding to 50 or more empirical claims. Even if > >> most of the judgments from such a paper are correct or are on the > >> right track, the problem is in knowing which judgments are correct. > >> For example, suppose that 90% of the judgments from an arbitrary paper > >> are correct (which is probably a high estimate). (Colin Phillips and > >> some of his former students / postdocs have commented to us that, in > >> their experience, quantitative acceptability judgment studies almost > >> always validate the claim(s) in the literature. This is not our > >> experience, however. Most experiments that we have run which attempt > >> to test some syntactic / semantic hypothesis in the literature end up > >> providing us with a pattern of data that had not been known before the > >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >> that in a paper with 50 empirical claims 45/50 are correct. But which > >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >> two million different theories. By quantitatively evaluating the > >> empirical claims, we reduce the uncertainty a great deal. To make > >> progress, it is better to have theoretical claims supported by solid > >> quantitative evidence, so that even if the interpretation of the data > >> changes over time as new evidence becomes available ? as is often the > >> case in any field of science ? the empirical pattern can be used as a > >> basis for further theorizing. > >> > >> Furthermore, it is no longer expensive to run behavioral experiments, > >> at least in English and other widely spoken languages. There now > >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >> can be used for collecting behavioral data over the internet quickly > >> and inexpensively. The cost of using an interface like this is > >> minimal, and the time that it takes for the results to be returned is > >> short. For example, currently on Mechanical Turk, a survey of > >> approximately 50 items will be answered by 50 or more participants > >> within a couple of hours, at a cost of approximately $1 per > >> participant. Thus a survey can be completed within a day, at a cost of > >> less than $50. (The hard work of designing the experiment, and > >> constructing controlled materials remains of course.) > >> > >> Sorry to be so verbose. But I think that these methodological points > >> are very important. > >> > >> Best wishes, > >> > >> Ted Gibson > >> > >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >> methods in syntax and semantics research. Language and Cognitive > >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >> & Fedorenko InPress LCP.pdf > >> > >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >> linguistics research. Trends in Cognitive Science, 14, 233-234. > >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > >> 2010 TICS.pdf > >> > >> > >> > >> > >>> Dick, > >>> > >>> You raise an important issue here about methodology. I believe that > >>> intuitions are a fine way to generate hypotheses and even to test > >>> them - to a degree. But while it might not have been feasible for > >>> Huddleston, Pullum, and the other contributors to the Cambridge > >>> Grammar to conduct experiments on every point of the grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, corpora, and standard psycholinguistic experimentation > >>> (indeed, Standard Social Science Methodology) is vital for taking the > >>> field forward and for providing the best support for different > >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>> paper on this, showing serious shortcomings with intuitions as the > >>> sole source of evidence, in their paper: "The need for quantitative > >>> methods in syntax and semantics research". > >>> > >>> Carson Schutze and Wayne Cowart, among others, have also written > >>> convincingly on this. > >>> > >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>> Science), and researchers from Brazil are beginning a third round of > >>> experimental work among the Pirahas, since my own work on the syntax > >>> was, like almost every other field researcher's, based on native > >>> speaker intuitions and corpora. > >>> > >>> The discussion of methodologies reminds me of the initial reactions > >>> to Greenberg's work on classifying the languages of the Americas. His > >>> methods were strongly (and justifiably) criticized. However, I always > >>> thought that his methods were a great way of generating hypotheses, > >>> so long as they were ultimately put to the test of standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions. > >>> > >>> -- Dan > >> > >> > >> > >>>> We linguists can add a further layer of explanation to the > >>>> judgements, but some judgements do seem to be more reliable than > >>>> others. And if we have to wait for psycholinguistic evidence for > >>>> every detailed analysis we make, our whole discipline will > >>>> immediately grind to a halt. Like it or not, native speaker > >>>> judgements are what put us linguists ahead of the rest in handling > >>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>> Language (or the OED) without using native speaker judgements. > >>>> > >>>> Best wishes, Dick Hudson > >> > >> > >> > > > > > > ------------------------------ > > Message: 14 > Date: Fri, 10 Sep 2010 18:33:49 -0400 > From: dryer at buffalo.edu > Subject: Re: [FUNKNET] analysis: unhappiness > To: "A. Katz" > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: <2147483647.1284143629 at cast-dryerm2.caset.buffalo.edu> > Content-Type: text/plain; charset=us-ascii; format=flowed > > > Aya, > > I actually agree with everything you say here. Personally, I am MORE > interested in the communicative function of language than I am in > psycholinguistics and how people process language. > > But none of that is relevant, I believe, to the very specific question of > what it means for an analysis to be correct. While one might conclude from > what I said that one ought to do psycholinguistics, that is not my > intention. Rather, my conclusion is that since I myself prefer not to do > psycholinguistics, I cannot really claim that the analyses I come up with > are "the correct" ones. And if it is really important to someone that they > identify "correct" analyses, then they ought to be doing psycholinguistics, > since there is no coherent notion of correct analysis outside of what is > inside of people's heads. > > Matthew > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > wrote: > > > Matthew, > > > > Thanks for stating that, because I was almost beginning to imagine that > > there was no essential disagreement, and that all of us agree that there > > is more -- and less -- to language than what is found in people's heads. > > > > Your position is the one I am familiar with from the functionalist point > > of view, and I was beginning to feel that it was underrepresented on > > Funknet. > > > > Those of us who disagree with your stated position -- but are very > > familiar with it -- are interested not just in psycholinguistics and how > > people process language -- but also in the communicative function of > > language as a system whereby information is transferred. Just as you and > > I may not be aware of the way our emails are encoded and then decoded by > > the computers that help us send emails back and forth, speakers may be > > compeltely unaware of what language does in order to transmit > information. > > > > After speakers have finished sending forth their linguistic output, it > > matters not at all how they arrived at this output. Language processing > > is separate from language in the same way that data processing is > > separate from data. > > > > Best, > > > > --Aya > > > > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > >> > >> The following sentence of Lise's > >> > >> "Since we don't in fact have such an agreed-on terminology, we have to > >> be quite careful in making clear what we are referring to when we talk > >> about 'the correct analysis' of a form like 'unhappiness'" > >> > >> suggests that there are two senses in which an analysis can be "the > >> correct analysis", one in terms of what is in people's heads, the other > >> in terms of what is "out there". > >> > >> There are a vast number of patterns "out there". The only distinction, > >> on my view, amongst this vast number of patterns, is between those that > >> reflect something inside people's heads and those that don't. But if > >> that is the case, then there is no coherent sense in which one can talk > >> of "the correct analysis" of what is "out there", except in terms of > >> what is in people's heads, and thus no second sense of "the correct > >> analysis". The patterns that don't correspond to things in people's > >> heads fall into (at least) two categories. There are those that are > >> akin to constellations of stars and, as with constellations, there is > >> no reality to these patterns, except in the minds of linguists. And > >> there are those patterns which are the fossil remains of what was in > >> the heads of speakers of an earlier stage of the language but which no > >> longer are. These latter patterns are real, and they are relevant to > >> exlaining why the language is now the way it is, but they are not > >> relevant, I think many would agree, as to what is the "correct analysis" > >> of the language today. > >> > >> For this reason, I claim that the only sense in which an analysis can be > >> "the correct analysis" is in terms of what is inside of people's heads. > >> > >> Again, I recommend the work of Bruce Derwing for lengthy discussion of > >> these issues. > >> > >> Matthew > >> > >> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > >> wrote: > >> > >>> I wish we had better terminology for keeping track of whether, at a > >>> given time, we are talking about the patterns that are 'out there' in > >>> the language and might possibly be apprehended (subconsciously) by a > >>> speaker, and when we are talking about the patterns that a particular > >>> speaker actually does apprehend, as indicated by experiments, from > >>> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >>> distinguishing among the degrees of pattern apprehension that a person > >>> may have, from vague preferences detectable in reaction times or other > >>> behavior all the way up through clear metalinguistic insights. Dick > >>> Hudson's note reminding us of the Gleitman and Gleitman study is right > >>> on target. > >>> > >>> Since we don't in fact have such an agreed-on terminology, we have > to > >>> be quite careful in making clear what we are referring to when we talk > >>> about 'the correct analysis' of a form like 'unhappiness'. We know, > >>> but tend to forget - and tend to forget to tell our students! - that > >>> it's an empirical question as to whether the formal simplicity and > >>> coherence of description of forms 'out there' (e.g. lovely abstract > >>> morphophonemics) is any kind of approximation to the way knowledge of > >>> the same forms is organized in a particular person's head. If we > >>> remember that a very large proportion of what we know about our > >>> language is 'out there' when we are infants and has to be internalized > >>> through experience with the language (even if you believe in innate > >>> 'core language'), the variation in internal knowledge from one person > >>> to another is more understandable. We especially need to consider (and > >>> try to test) the > >>> possibility that > >>> since > >>> the brain can make multiple cross-connections, multiple patterns are > >>> involved > >>> simultaneously in morphological and syntactic analyses. I suggest that > >>> that's the case with 'unhappiness' - and the linguistic analyses that > I > >>> know about are not good at handling that kind of idea. > >>> > >>> Lise > >>> > >>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >>> > >>>> > >>>> Two comments. > >>>> > >>>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>>> is an > >>>> important distinction between low-level linguistic intuitions (like > >>>> whether a > >>>> word or sentence is well-formed or what it means) and higher-level > >>>> intuitions > >>>> (like what the structure of a word or sentence is). One can take > >>>> the position > >>>> that we need to account for the former (while recognizing that they > >>>> are not > >>>> always reliable) but not the latter. > >>>> > >>>> Second, the tension here is not only between evidence from speaker > >>>> intuitions > >>>> versus evidence from psycholinguistic experiments. There is also a > >>>> tension > >>>> between deciding on the correct analysis on the basis of a priori > >>>> simplicity > >>>> arguments versus deciding on the correct analysis on the basis of > >>>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>>> paradox that Dan > >>>> referred to that arises with the word (semantics argues > >>>> for > >>>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>>> +er]] (the > >>>> comparative suffix can only be attached to adjectives containing one > >>>> or two > >>>> syllables) is only a paradox if one assumes that speakers adopt the > >>>> simplest > >>>> analysis. For example, if speakers adopt a more complex rule for > >>>> either of these > >>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>>> trisyllabic > >>>> words beginning with un-), then the bracketing paradox disappears. > >>>> > >>>> Matthew > >>>> > >>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>>>> conscious judgements are only one kind of evidence that we need to > >>>>> take > >>>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>>> fact), > >>>>> and of course I've been aware of complaints about judgements by > >>>>> people > >>>>> like Labov for decades. > >>>>> > >>>>> But you're missing my main point, which is that all judgements aren't > >>>>> equally reliable. If you want to know how /unhappiness/ is > >>>>> structured, > >>>>> ask a linguist, not a five-year old. And one of the by-products of > >>>>> education may be increased sensitivity to syntax - which is one of > >>>>> the > >>>>> many reasons why linguists need to pay more attention to education. > >>>>> > >>>>> Best wishes, Dick > >>>>> > >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about > >>>>> methodology. I believe that intuitions are a fine way to generate > >>>>> hypotheses and even to test them - to a degree. But while it might > >>>>> not have > >>>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>>> the > >>>>> Cambridge Grammar to conduct experiments on every point of the > >>>>> grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, > >>>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>>> Standard > >>>>> Social Science Methodology) is vital for taking the field forward > >>>>> and for > >>>>> providing the best support for different analyses. Ted Gibson and Ev > >>>>> Fedorenko have written a very useful new paper on this, showing > >>>>> serious > >>>>> shortcomings with intuitions as the sole source of evidence, in their > >>>>> paper: "The need for quantitative methods in syntax and semantics > >>>>> research".> > >>>>>> Carson Schutze and Wayne Cowart, among others, > >>>>> have also written convincingly on this.> > >>>>>> It is one reason that a team from Stanford, MIT > >>>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>>> beginning a > >>>>> third round of experimental work among the Pirahas, since my own > >>>>> work on > >>>>> the syntax was, like almost every other field researcher's, based > >>>>> on native > >>>>> speaker intuitions and corpora.> > >>>>>> The discussion of methodologies reminds me of > >>>>> the initial reactions to Greenberg's work on classifying the > >>>>> languages of > >>>>> the Americas. His methods were strongly (and justifiably) criticized. > >>>>> However, I always thought that his methods were a great way of > >>>>> generating > >>>>> hypotheses, so long as they were ultimately put to the test of > >>>>> standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions.> > >>>>>> -- Dan > >>>>>> > >>>>>>> We linguists can add a further layer of > >>>>> explanation to the judgements, but some judgements do seem to be more > >>>>> reliable than others. And if we have to wait for psycholinguistic > >>>>> evidence > >>>>> for every detailed analysis we make, our whole discipline will > >>>>> immediately > >>>>> grind to a halt. Like it or not, native speaker judgements are what > >>>>> put us > >>>>> linguists ahead of the rest in handling fine detail. Imagine > >>>>> writing the > >>>>> Cambridge Grammar of the English Language (or the OED) without > >>>>> using native > >>>>> speaker judgements.>> > >>>>>>> Best wishes, Dick Hudson > >>>>>>> > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>> > >>> > >>> Lise Menn Home Office: 303-444-4274 > >>> 1625 Mariposa Ave Fax: 303-413-0017 > >>> Boulder CO 80302 > >>> > >>> Professor Emerita of Linguistics > >>> Fellow, Institute of Cognitive Science > >>> University of Colorado > >>> > >>> Secretary, AAAS Section Z [Linguistics] > >>> > >>> Campus Mail Address: > >>> UCB 594, Institute of Cognitive Science > >>> > >>> Campus Physical Address: > >>> CINC 234 > >>> 1777 Exposition Ave, Boulder > >>> > >>> > >>> > >>> > >> > >> > >> > >> > >> > > > > > > > > > > > ------------------------------ > > Message: 15 > Date: Sat, 11 Sep 2010 00:20:15 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet at mailman.rice.edu > Message-ID: <4C8ABD2F.8070902 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Brian, > What a helpful message! I think you're right: we need a typology of > cases, each needing a different range of methods, ranging from the > linguist's own judgements for really easy cases to more complicated > quantitative methods for more complicated ones. > > The trouble with our discipline is that for any community of N speakers, > and a language consisting of M 'items' (however you may choose to define > 'community' and 'item'), we have N*M datapoints that, in principle, all > need to be validated somehow. We might reduce the number by focusing on > one speaker, but then you can't use data from other speakers as evidence > for that speaker's language; or we might try to construct a 'typical' > speaker, but we don't know how to do that; or we might reduce the size > of the community by trying to find a 'dialect' (but dialects don't > really exist); or we might ignore most of the linguistic items and focus > on, say, the modal verbs - but then we miss their links to all the other > items. > > It's different for psycholinguists because they're only interested in > general processes, for which linguistic items are just evidence, not the > thing under investigation; but for us linguists, the fine detail is > everything because we're the people who explore the connections between > items. > > So I look forward to the day when your typology of cases will guide us > through a range of different methods to the appropriate ones for any > given item. > > Best wishes, Dick > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 23:23, Brian MacWhinney wrote: > > Dick and Ted, > > I agree with parts of what each of you are saying. Which means that > I also disagree with other parts. In practice,, Gibson and Fedorenko, in > press, (which I downloaded and scanned) deals with no more than two or > three constructions. They mention the fact that people don't have problems > with sentences such as "Susan muttered him the news" despite claims that > verbs such as "mutter" cannot take the double object construction. They > also note that the claims from Jackendoff and Culicover about the > differences between the two sentences below are not supported by results > from the Mechanical Turk: > > 1. Peter was trying to remember who carried what. > > 2. Peter was trying to remember who carried what when. > > These are interesting facts. If these sentences are supposed to be > different and people judge them to be similarly grammatical, then theories > based on the supposed differences should be reexamined. There are big > chunks of syntactic theory resting on shaky judgments about complex > sentences of this type. Getting some of this straight would be a big win, I > would say, particularly if linguists would pay attention to the results. > > But I understand Dick's worry about how far Gibson and Fedorenko > are trying to push this. Neither their email nor their paper sets clear > limits on what we should be testing and we certainly don't want to waste > time checking out go-goed-went. So, Gibson and Fedorenko owe us those > clarifications. > > But, Dick, you then move on to questioning data on bid-bidded. Here > we have a case of true variation in the population. I would love to know > its distribution. As a "fan of quantitative sociolinguistics" shouldn't you > too? > > My take on this is that constructions are not created equal. The > three types mentioned here are probably just a start on an inventory of > evidentiary types. We need to correctly pair up appropriate methods with > each of the types. And we to make sure that people pay attention to the > results, once they are in > > > > --Brian MacWhinney > > > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > >> Dear Ted, > >> Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > >> > >> Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > >> > >> Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quanti > tative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > >> > >> It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > >> > >> Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 14:03, Ted Gibson wrote: > >>> Dear Dan, Dick: > >>> > >>> I would like to clarify some points that Dan Everett makes, in > >>> response to Dick Hudson. > >>> > >>> Ev Fedorenko and I have written a couple of papers recently (Gibson& > >>> Fedorenko, 2010, in press, see references and links below) on what we > >>> think are weak methodological standards in syntax and semantics > >>> research over the past many years. The issue that we address is the > >>> prevalent method in syntax and semantics research, which involves > >>> obtaining a judgment of the acceptability of a sentence / meaning > >>> pair, typically by just the author of the paper, sometimes with > >>> feedback from colleagues. As we address in our papers, this > >>> methodology does not allow proper testing of scientific hypotheses > >>> because of (a) the small number of experimental participants > >>> (typically one); (b) the small number of experimental stimuli > >>> (typically one); (c) cognitive biases on the part of the researcher > >>> and participants; and (d) the effect of the preceding context (e.g., > >>> other constructions the researcher may have been recently > >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>> several others cited in Gibson& Fedorenko, in press; for similar > >>> points, but with not as strong a conclusion as ours). > >>> > >>> Three issues need to be separated here: (1) the use of intuitive > >>> judgments as a dependent measure in a language experiment; (2) > >>> potential cognitive biases on the part of experimental subjects and > >>> experimenters in language experiments; and (3) the need for obtaining > >>> quantitative evidence, whatever the dependent measure might be. The > >>> paper that Ev and I wrote addresses the last two issues, but does not > >>> go into depth on the first issue (the use of intuitions as a dependent > >>> measure in language experiments). Regarding this issue, we don't think > >>> that there is anything wrong with gathering intuitive judgments as a > >>> dependent measure, as long as the task is clear to the experimental > >>> participants. > >>> > >>> In the longer paper (Gibson& Fedorenko, in press) we respond to some > >>> arguments that have been given in support of continuing to use the > >>> traditional non-quantitative method in syntax / semantics research. > >>> One recent defense of the traditional method comes from Phillips > >>> (2008), who argues that no harm has come from the non-quantitative > >>> approach in syntax research thus far. Phillips argues that there are > >>> no cases in the literature where an incorrect intuitive judgment has > >>> become the basis for a widely accepted generalization or an important > >>> theoretical claim. He therefore concludes that there is no reason to > >>> adopt more rigorous data collection standards. We challenge Philips? > >>> conclusion by presenting three cases from the literature where a > >>> faulty intuition has led to incorrect generalizations and mistaken > >>> theorizing, plausibly due to cognitive biases on the part of the > >>> researchers. > >>> > >>> A second argument that is sometimes presented for the continued use of > >>> the traditional non-quantitative method is that it would be too > >>> inefficient to evaluate every syntactic / semantic hypothesis or > >>> phenomenon quantitatively. For example, Culicover& Jackendoff (2010) > >>> make this argument explicitly in their response to Gibson& Fedorenko > >>> (2010): ?It would cripple linguistic investigation if it were required > >>> that all judgments of ambiguity and grammaticality be subject to > >>> statistically rigorous experiments on naive subjects, especially when > >>> investigating languages whose speakers are hard to access? (Culicover > >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>> earlier in the discussion here.) Whereas we agree that in > >>> circumstances where gathering data is difficult, some evidence is > >>> better than no evidence, we do not agree that research would be slowed > >>> with respect to languages where experimental participants are easy to > >>> access, such as English. In contrast, we think that the opposite is > >>> true: the field?s progress is probably slowed by not doing > >>> quantitative research. > >>> Suppose that a typical syntax / semantics paper that lacks > >>> quantitative evidence includes judgments for 50 or more sentences / > >>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>> most of the judgments from such a paper are correct or are on the > >>> right track, the problem is in knowing which judgments are correct. > >>> For example, suppose that 90% of the judgments from an arbitrary paper > >>> are correct (which is probably a high estimate). (Colin Phillips and > >>> some of his former students / postdocs have commented to us that, in > >>> their experience, quantitative acceptability judgment studies almost > >>> always validate the claim(s) in the literature. This is not our > >>> experience, however. Most experiments that we have run which attempt > >>> to test some syntactic / semantic hypothesis in the literature end up > >>> providing us with a pattern of data that had not been known before the > >>> experiment (e.g., Breen et al., in press; Fedorenko& Gibson, in > >>> press; Patel et al., 2009; Scontras& Gibson, submitted).) This means > >>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>> two million different theories. By quantitatively evaluating the > >>> empirical claims, we reduce the uncertainty a great deal. To make > >>> progress, it is better to have theoretical claims supported by solid > >>> quantitative evidence, so that even if the interpretation of the data > >>> changes over time as new evidence becomes available ? as is often the > >>> case in any field of science ? the empirical pattern can be used as a > >>> basis for further theorizing. > >>> > >>> Furthermore, it is no longer expensive to run behavioral experiments, > >>> at least in English and other widely spoken languages. There now > >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>> can be used for collecting behavioral data over the internet quickly > >>> and inexpensively. The cost of using an interface like this is > >>> minimal, and the time that it takes for the results to be returned is > >>> short. For example, currently on Mechanical Turk, a survey of > >>> approximately 50 items will be answered by 50 or more participants > >>> within a couple of hours, at a cost of approximately $1 per > >>> participant. Thus a survey can be completed within a day, at a cost of > >>> less than $50. (The hard work of designing the experiment, and > >>> constructing controlled materials remains of course.) > >>> > >>> Sorry to be so verbose. But I think that these methodological points > >>> are very important. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson > >>> > >>> Gibson, E.& Fedorenko, E. (In press). The need for quantitative > >>> methods in syntax and semantics research. Language and Cognitive > >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>> & Fedorenko InPress LCP.pdf > >>> > >>> Gibson, E.& Fedorenko, E. (2010). Weak quantitative standards in > >>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson& Fedorenko > >>> 2010 TICS.pdf > >>> > >>> > >>> > >>> > >>>> Dick, > >>>> > >>>> You raise an important issue here about methodology. I believe that > >>>> intuitions are a fine way to generate hypotheses and even to test > >>>> them - to a degree. But while it might not have been feasible for > >>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>> Grammar to conduct experiments on every point of the grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>> field forward and for providing the best support for different > >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>> paper on this, showing serious shortcomings with intuitions as the > >>>> sole source of evidence, in their paper: "The need for quantitative > >>>> methods in syntax and semantics research". > >>>> > >>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>> convincingly on this. > >>>> > >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>> Science), and researchers from Brazil are beginning a third round of > >>>> experimental work among the Pirahas, since my own work on the syntax > >>>> was, like almost every other field researcher's, based on native > >>>> speaker intuitions and corpora. > >>>> > >>>> The discussion of methodologies reminds me of the initial reactions > >>>> to Greenberg's work on classifying the languages of the Americas. His > >>>> methods were strongly (and justifiably) criticized. However, I always > >>>> thought that his methods were a great way of generating hypotheses, > >>>> so long as they were ultimately put to the test of standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions. > >>>> > >>>> -- Dan > >>> > >>> > >>>>> We linguists can add a further layer of explanation to the > >>>>> judgements, but some judgements do seem to be more reliable than > >>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>> every detailed analysis we make, our whole discipline will > >>>>> immediately grind to a halt. Like it or not, native speaker > >>>>> judgements are what put us linguists ahead of the rest in handling > >>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>> Language (or the OED) without using native speaker judgements. > >>>>> > >>>>> Best wishes, Dick Hudson > >>> > >>> > > > > > > > > ------------------------------ > > Message: 16 > Date: Sat, 11 Sep 2010 00:40:06 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: Ted Gibson > Cc: Richard Hudson , Evelina Fedorenko > , funknet > Message-ID: <4C8AC1D6.3010203 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's view. > Your goal is to find general processes and principles that apply > uniformly across individuals, so you have to use methods to check for > generality. And (as you know) I admire the way you pursue that goal. But > my goal, as a linguist, is different. I want to explore the structure of > a language so that I can understand how all the bits fit together. Like > you, I'm aiming to model cognition, but my focus is on items and > structures, and I start from the assumption that these can and do vary > across speakers. > > However, having said all that I do agree with you that linguists should > all get used to collecting and using quantitative data; and with the > help of Brian MacWhinney's typology we'd know what methods to use when. > And I do agree with your points about bid/bidded: in cases like that, > quantitative data would be at least a very good starting point for a > proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: > > Dear Dick: > > > > Perhaps we are talking at cross purposes. I don't understand what is > > confusing about what Ev Fedorenko and I are claiming. All we are > > saying is that if you have some testable claim involving a general > > hypothesis about a language, then you need to get quantitative data > > from unbiased sources to evaluate that claim. If you are interested in > > English past tense morphology, then depending on the claims that you > > might want to investigate, there are lots of ways to get relevant > > quantitative evidence. Corpus data will probably be useful. For very > > low frequency words, you can run experiments to test behavior with > > respect to such words. > > > > Your example of the past tense of "bid" is a fine such example. You > > can run an experiment like the one you suggested to find out what > > people think the past tense is. If you then found that 20/50 people > > responded "bidded" and 30/50 respond "bid", that is a lot of useful > > information. As you suggest in your discussion, this result wouldn't > > answer the question of how past tense is stored in each individual. > > This result would be ambiguous among several possible explanations. > > One possibility is that the probability distribution that is being > > discovered reflects different dialects, such that 2/5 of the > > population has one past tense, and 3/5 has another. Another > > possibility is that each person has a similar probability distribution > > in their heads, such that 2/5 of the time I respond one way, and 3/5 > > of the time I respond another. Further experiments would be necessary > > to answer between these and other possible theories (e.g., with > > repeated trials from the same person, carefully planned so that the > > participants don't notice that they are being asked multiple times). > > Without the quantitative evidence in the first place, there is no way > > to answer these kinds of questions. > > > > Regarding the past tense of "go", this would be useful as a baseline > > in an experiment involving the less frequent ones. So, yes, it would > > useful to gather quantitative evidence in such a case also, as > > baselines with respect to the more interesting cases for theories. > > > > The bottom line: if you have a generalization about a language that > > you wish to evaluate (such that you hypothesize that it is true across > > the speakers of the language), then you need quantitative evidence > > from multiple individuals, using an unbiased data collection method, > > to evaluate such a claim. The point about Mechanical Turk is that it > > is really *easy* to do this now, at least for languages like English. > > > > Best wishes, > > > > Ted Gibson & Ev Fedorenko > > > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > >> Dear Ted, > >> Thanks for the very interesting comment, but are you REALLY saying > >> that I shouldn't claim, for example, that the past tense of GO is > >> "went" without first cross-checking with 50 native speakers? > >> > >> Isn't there a danger of missing the point that we all, as native > >> speakers, spend our whole lives scanning other people's linguistic > >> behaviour (language 'out there', E-language) and trying to explain it > >> to ourselves in terms of a language system (language 'in here', > >> I-language)? So every judgement we make is based on thousands or > >> millions of observed exemplars, and reflects a unique experience of > >> E-language filtered through a unique I-language. > >> > >> Given that view of language development, I don't see how quantitative > >> data will help. Let's take a real uncertainty, such as the past tense > >> of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > >> judgement: I don't know. Ok, you get 50 people to oblige on > >> Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? > >> Does that mean that the correct answer is "bidded"? Surely not. How > >> is it better than my judgement? I agree you could record my speech > >> and find how often I use each alternative; but the reason I don't > >> know is precisely because it's a rare word, so in a sense > >> quantitative data are irrelevant even there. What would solve the > >> problem of subjectivity, of course, would be a machine for probing > >> the bit of my mind (or even brain) that holds BID and its details; > >> but I suspect that even that wouldn't move us much further forward > >> than my original "don't know". (Incidentally I write as a fan of > >> quantitative sociolinguistics, so I do accept that quantitative data > >> are relevant to linguistic analysis in some areas, where the > >> I-language phenomenon is frequent enough to produce usable data.) > >> > >> It seems to me that this discussion raises the really fundamental > >> question of what kind of thing we think language is: social or > >> individual. The problem isn't unique to linguistics of course; it's > >> the same throughout the social sciences. But what's special about > >> linguistics is that we deal in very fine details of culture (e.g. > >> details of how a particular word is used or pronounced) so the > >> differences between individuals really matter. I don't see that we're > >> ever going to have anything better than judgements to go on, so what > >> we need is a way to ensure that judgements are accurate reports of > >> individual I-language. A rotten situation for a science, but I don't > >> see how it can get better. > >> > >> Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 14:03, Ted Gibson wrote: > >>> Dear Dan, Dick: > >>> > >>> I would like to clarify some points that Dan Everett makes, in > >>> response to Dick Hudson. > >>> > >>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>> Fedorenko, 2010, in press, see references and links below) on what we > >>> think are weak methodological standards in syntax and semantics > >>> research over the past many years. The issue that we address is the > >>> prevalent method in syntax and semantics research, which involves > >>> obtaining a judgment of the acceptability of a sentence / meaning > >>> pair, typically by just the author of the paper, sometimes with > >>> feedback from colleagues. As we address in our papers, this > >>> methodology does not allow proper testing of scientific hypotheses > >>> because of (a) the small number of experimental participants > >>> (typically one); (b) the small number of experimental stimuli > >>> (typically one); (c) cognitive biases on the part of the researcher > >>> and participants; and (d) the effect of the preceding context (e.g., > >>> other constructions the researcher may have been recently > >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>> several others cited in Gibson & Fedorenko, in press; for similar > >>> points, but with not as strong a conclusion as ours). > >>> > >>> Three issues need to be separated here: (1) the use of intuitive > >>> judgments as a dependent measure in a language experiment; (2) > >>> potential cognitive biases on the part of experimental subjects and > >>> experimenters in language experiments; and (3) the need for obtaining > >>> quantitative evidence, whatever the dependent measure might be. The > >>> paper that Ev and I wrote addresses the last two issues, but does not > >>> go into depth on the first issue (the use of intuitions as a dependent > >>> measure in language experiments). Regarding this issue, we don't think > >>> that there is anything wrong with gathering intuitive judgments as a > >>> dependent measure, as long as the task is clear to the experimental > >>> participants. > >>> > >>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>> arguments that have been given in support of continuing to use the > >>> traditional non-quantitative method in syntax / semantics research. > >>> One recent defense of the traditional method comes from Phillips > >>> (2008), who argues that no harm has come from the non-quantitative > >>> approach in syntax research thus far. Phillips argues that there are > >>> no cases in the literature where an incorrect intuitive judgment has > >>> become the basis for a widely accepted generalization or an important > >>> theoretical claim. He therefore concludes that there is no reason to > >>> adopt more rigorous data collection standards. We challenge Philips? > >>> conclusion by presenting three cases from the literature where a > >>> faulty intuition has led to incorrect generalizations and mistaken > >>> theorizing, plausibly due to cognitive biases on the part of the > >>> researchers. > >>> > >>> A second argument that is sometimes presented for the continued use of > >>> the traditional non-quantitative method is that it would be too > >>> inefficient to evaluate every syntactic / semantic hypothesis or > >>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>> make this argument explicitly in their response to Gibson & Fedorenko > >>> (2010): ?It would cripple linguistic investigation if it were required > >>> that all judgments of ambiguity and grammaticality be subject to > >>> statistically rigorous experiments on naive subjects, especially when > >>> investigating languages whose speakers are hard to access? (Culicover > >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>> earlier in the discussion here.) Whereas we agree that in > >>> circumstances where gathering data is difficult, some evidence is > >>> better than no evidence, we do not agree that research would be slowed > >>> with respect to languages where experimental participants are easy to > >>> access, such as English. In contrast, we think that the opposite is > >>> true: the field?s progress is probably slowed by not doing > >>> quantitative research. > >>> Suppose that a typical syntax / semantics paper that lacks > >>> quantitative evidence includes judgments for 50 or more sentences / > >>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>> most of the judgments from such a paper are correct or are on the > >>> right track, the problem is in knowing which judgments are correct. > >>> For example, suppose that 90% of the judgments from an arbitrary paper > >>> are correct (which is probably a high estimate). (Colin Phillips and > >>> some of his former students / postdocs have commented to us that, in > >>> their experience, quantitative acceptability judgment studies almost > >>> always validate the claim(s) in the literature. This is not our > >>> experience, however. Most experiments that we have run which attempt > >>> to test some syntactic / semantic hypothesis in the literature end up > >>> providing us with a pattern of data that had not been known before the > >>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>> two million different theories. By quantitatively evaluating the > >>> empirical claims, we reduce the uncertainty a great deal. To make > >>> progress, it is better to have theoretical claims supported by solid > >>> quantitative evidence, so that even if the interpretation of the data > >>> changes over time as new evidence becomes available ? as is often the > >>> case in any field of science ? the empirical pattern can be used as a > >>> basis for further theorizing. > >>> > >>> Furthermore, it is no longer expensive to run behavioral experiments, > >>> at least in English and other widely spoken languages. There now > >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>> can be used for collecting behavioral data over the internet quickly > >>> and inexpensively. The cost of using an interface like this is > >>> minimal, and the time that it takes for the results to be returned is > >>> short. For example, currently on Mechanical Turk, a survey of > >>> approximately 50 items will be answered by 50 or more participants > >>> within a couple of hours, at a cost of approximately $1 per > >>> participant. Thus a survey can be completed within a day, at a cost of > >>> less than $50. (The hard work of designing the experiment, and > >>> constructing controlled materials remains of course.) > >>> > >>> Sorry to be so verbose. But I think that these methodological points > >>> are very important. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson > >>> > >>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>> methods in syntax and semantics research. Language and Cognitive > >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>> & Fedorenko InPress LCP.pdf > >>> > >>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > >>> 2010 TICS.pdf > >>> > >>> > >>> > >>> > >>>> Dick, > >>>> > >>>> You raise an important issue here about methodology. I believe that > >>>> intuitions are a fine way to generate hypotheses and even to test > >>>> them - to a degree. But while it might not have been feasible for > >>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>> Grammar to conduct experiments on every point of the grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>> field forward and for providing the best support for different > >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>> paper on this, showing serious shortcomings with intuitions as the > >>>> sole source of evidence, in their paper: "The need for quantitative > >>>> methods in syntax and semantics research". > >>>> > >>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>> convincingly on this. > >>>> > >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>> Science), and researchers from Brazil are beginning a third round of > >>>> experimental work among the Pirahas, since my own work on the syntax > >>>> was, like almost every other field researcher's, based on native > >>>> speaker intuitions and corpora. > >>>> > >>>> The discussion of methodologies reminds me of the initial reactions > >>>> to Greenberg's work on classifying the languages of the Americas. His > >>>> methods were strongly (and justifiably) criticized. However, I always > >>>> thought that his methods were a great way of generating hypotheses, > >>>> so long as they were ultimately put to the test of standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions. > >>>> > >>>> -- Dan > >>> > >>> > >>> > >>>>> We linguists can add a further layer of explanation to the > >>>>> judgements, but some judgements do seem to be more reliable than > >>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>> every detailed analysis we make, our whole discipline will > >>>>> immediately grind to a halt. Like it or not, native speaker > >>>>> judgements are what put us linguists ahead of the rest in handling > >>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>> Language (or the OED) without using native speaker judgements. > >>>>> > >>>>> Best wishes, Dick Hudson > >>> > >>> > >>> > > > > > > > > > ------------------------------ > > Message: 17 > Date: Fri, 10 Sep 2010 18:40:47 -0600 > From: Lise Menn > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Richard Hudson , Funknet > > Message-ID: <91F0DEA7-53BE-405B-8087-88B7F1672447 at Colorado.EDU> > Content-Type: text/plain; charset=US-ASCII; format=flowed; > delsp=yes > > Matt, I have to disagree with you on the validity of describing what's > 'out there' (what Dick Hudson says is his interest, in his > contribution of 5:40:06 PM MDT today). We DO have to account for it > in order to understand how 'the language in speakers' heads' gets into > those heads in the first place. > In more detail: Each of us is immersed from (before) birth in a > sampling of utterances (and if we are literate, eventually also > written forms of the language). In order to understand how we really > create our internal representations of our language, we have to know > (or be able to estimate) something about the data our brains get as > input. There are at least better and worse descriptions of the > patterns in those data, and certainly there are wrong ones, though in > many cases - for example in the 'unhappiness' case - there are > probably conflicting right ones, rather than any single correct one. > (OT offers some help in thinking about this.) > > To take a concrete example, in order to account for the still- > unstable changes in English pronominal case marking in compound NP > objects of prepositions from a system based on syntactic case (He gave > the cookies to Mary and me) to a system apparently based partly on > whether the pronoun is next to the governing preposition (He gave the > cookies to Mary and I/ to me and Mary), you first have to do an > analysis of usage and figure out what the pattern is. And usage is > not in our heads (although it's the result of what's in our heads), > it's 'out there'. > > Even fossils and obscure patterns contribute to the redundancy of > the > language, making it more learnable and and helping to create the > resonances used by great poets and orators. (I admit to having > oversimplified in speaking as if there were always one 'correct' > analysis of the patterns 'out there' that might be (subconsciously) > discoverable by speakers. That's not true.) And because not all > speakers are equally sensitive to language patterns - again, the > Gleitman and Gleitman book is a terrific example - it's also an > oversimplification to talk about 'what is in speaker's heads' as if > the same thing is in everyone's head. (K.P. Mohanan has also published > on this.) At the lexical level, Danielle Cyr's examples (September 9, > 2010 8:38:59 PM MDT) further remind us that what's inside each > person's head changes over time. So we must also be careful not to > idealize "what's in people's heads" as if it were a single coherent > construct that we are trying to discover. It's not - it's more like a > complex mosaic that does not fit together perfectly. > > Lise > > On Sep 10, 2010, at 12:51 PM, dryer at buffalo.edu wrote: > > > > > The following sentence of Lise's > > > > "Since we don't in fact have such an agreed-on terminology, we have > > to be quite careful in making clear what we are referring to when > > we talk about 'the correct analysis' of a form like 'unhappiness'" > > > > suggests that there are two senses in which an analysis can be "the > > correct analysis", one in terms of what is in people's heads, the > > other in terms of what is "out there". > > > > There are a vast number of patterns "out there". The only > > distinction, on my view, amongst this vast number of patterns, is > > between those that reflect something inside people's heads and those > > that don't. But if that is the case, then there is no coherent > > sense in which one can talk of "the correct analysis" of what is > > "out there", except in terms of what is in people's heads, and thus > > no second sense of "the correct analysis". The patterns that don't > > correspond to things in people's heads fall into (at least) two > > categories. There are those that are akin to constellations of > > stars and, as with constellations, there is no reality to these > > patterns, except in the minds of linguists. And there are those > > patterns which are the fossil remains of what was in the heads of > > speakers of an earlier stage of the language but which no longer > > are. These latter patterns are real, and they are relevant to > > exlaining why the language is now the way it is, but they are not > > relevant, I think many would agree, as to what is the "correct > > analysis" of the language today. > > > > For this reason, I claim that the only sense in which an analysis > > can be "the correct analysis" is in terms of what is inside of > > people's heads. > > > > Again, I recommend the work of Bruce Derwing for lengthy discussion > > of these issues. > > > > Matthew > > > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > > > wrote: > > > >> I wish we had better terminology for keeping track of whether, at a > >> given time, we are talking about the patterns that are 'out there' in > >> the language and might possibly be apprehended (subconsciously) by a > >> speaker, and when we are talking about the patterns that a particular > >> speaker actually does apprehend, as indicated by experiments, from > >> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >> distinguishing among the degrees of pattern apprehension that a > >> person > >> may have, from vague preferences detectable in reaction times or > >> other > >> behavior all the way up through clear metalinguistic insights. Dick > >> Hudson's note reminding us of the Gleitman and Gleitman study is > >> right > >> on target. > >> > >> Since we don't in fact have such an agreed-on terminology, we have > >> to > >> be quite careful in making clear what we are referring to when we > >> talk > >> about 'the correct analysis' of a form like 'unhappiness'. We > >> know, but > >> tend to forget - and tend to forget to tell our students! - that > >> it's an > >> empirical question as to whether the formal simplicity and > >> coherence of > >> description of forms 'out there' (e.g. lovely abstract > >> morphophonemics) > >> is any kind of approximation to the way knowledge of the same > >> forms is > >> organized in a particular person's head. If we remember that a very > >> large proportion of what we know about our language is 'out there' > >> when > >> we are infants and has to be internalized through experience with > >> the > >> language (even if you believe in innate 'core language'), the > >> variation > >> in internal knowledge from one person to another is more > >> understandable. > >> > >> We especially need to consider (and try to test) the possibility > >> that > >> since > >> the brain can make multiple cross-connections, multiple patterns are > >> involved > >> simultaneously in morphological and syntactic analyses. I suggest > >> that > >> that's the case with 'unhappiness' - and the linguistic analyses > >> that I > >> know about are not good at handling that kind of idea. > >> > >> Lise > >> > >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> > >>> > >>> Two comments. > >>> > >>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>> is an > >>> important distinction between low-level linguistic intuitions (like > >>> whether a > >>> word or sentence is well-formed or what it means) and higher-level > >>> intuitions > >>> (like what the structure of a word or sentence is). One can take > >>> the position > >>> that we need to account for the former (while recognizing that they > >>> are not > >>> always reliable) but not the latter. > >>> > >>> Second, the tension here is not only between evidence from speaker > >>> intuitions > >>> versus evidence from psycholinguistic experiments. There is also a > >>> tension > >>> between deciding on the correct analysis on the basis of a priori > >>> simplicity > >>> arguments versus deciding on the correct analysis on the basis of > >>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>> paradox that Dan > >>> referred to that arises with the word (semantics argues > >>> for > >>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>> +er]] (the > >>> comparative suffix can only be attached to adjectives containing one > >>> or two > >>> syllables) is only a paradox if one assumes that speakers adopt the > >>> simplest > >>> analysis. For example, if speakers adopt a more complex rule for > >>> either of these > >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>> trisyllabic > >>> words beginning with un-), then the bracketing paradox disappears. > >>> > >>> Matthew > >>> > >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree > >>>> that > >>>> conscious judgements are only one kind of evidence that we need to > >>>> take > >>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>> fact), > >>>> and of course I've been aware of complaints about judgements by > >>>> people > >>>> like Labov for decades. > >>>> > >>>> But you're missing my main point, which is that all judgements > >>>> aren't > >>>> equally reliable. If you want to know how /unhappiness/ is > >>>> structured, > >>>> ask a linguist, not a five-year old. And one of the by-products of > >>>> education may be increased sensitivity to syntax - which is one of > >>>> the > >>>> many reasons why linguists need to pay more attention to education. > >>>> > >>>> Best wishes, Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about > >>>> methodology. I believe that intuitions are a fine way to generate > >>>> hypotheses and even to test them - to a degree. But while it might > >>>> not have > >>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>> the > >>>> Cambridge Grammar to conduct experiments on every point of the > >>>> grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, > >>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>> Standard > >>>> Social Science Methodology) is vital for taking the field forward > >>>> and for > >>>> providing the best support for different analyses. Ted Gibson and > >>>> Ev > >>>> Fedorenko have written a very useful new paper on this, showing > >>>> serious > >>>> shortcomings with intuitions as the sole source of evidence, in > >>>> their > >>>> paper: "The need for quantitative methods in syntax and semantics > >>>> research".> > >>>>> Carson Schutze and Wayne Cowart, among others, > >>>> have also written convincingly on this.> > >>>>> It is one reason that a team from Stanford, MIT > >>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>> beginning a > >>>> third round of experimental work among the Pirahas, since my own > >>>> work on > >>>> the syntax was, like almost every other field researcher's, based > >>>> on native > >>>> speaker intuitions and corpora.> > >>>>> The discussion of methodologies reminds me of > >>>> the initial reactions to Greenberg's work on classifying the > >>>> languages of > >>>> the Americas. His methods were strongly (and justifiably) > >>>> criticized. > >>>> However, I always thought that his methods were a great way of > >>>> generating > >>>> hypotheses, so long as they were ultimately put to the test of > >>>> standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions.> > >>>>> -- Dan > >>>>> > >>>>>> We linguists can add a further layer of > >>>> explanation to the judgements, but some judgements do seem to be > >>>> more > >>>> reliable than others. And if we have to wait for psycholinguistic > >>>> evidence > >>>> for every detailed analysis we make, our whole discipline will > >>>> immediately > >>>> grind to a halt. Like it or not, native speaker judgements are what > >>>> put us > >>>> linguists ahead of the rest in handling fine detail. Imagine > >>>> writing the > >>>> Cambridge Grammar of the English Language (or the OED) without > >>>> using native > >>>> speaker judgements.>> > >>>>>> Best wishes, Dick Hudson > >>>>>> > >>>>> > >>>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>> > >> > >> Lise Menn Home Office: 303-444-4274 > >> 1625 Mariposa Ave Fax: 303-413-0017 > >> Boulder CO 80302 > >> > >> Professor Emerita of Linguistics > >> Fellow, Institute of Cognitive Science > >> University of Colorado > >> > >> Secretary, AAAS Section Z [Linguistics] > >> > >> Campus Mail Address: > >> UCB 594, Institute of Cognitive Science > >> > >> Campus Physical Address: > >> CINC 234 > >> 1777 Exposition Ave, Boulder > >> > >> > >> > >> > > > > > > > > > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > > > ------------------------------ > > Message: 18 > Date: Fri, 10 Sep 2010 19:03:41 -0600 > From: Lise Menn > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Ted Gibson , Richard Hudson > , Evelina Fedorenko >, > funknet > Message-ID: <10638E94-DC8F-4181-8135-CE74F045079E at colorado.edu> > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dick, I think that what you say is true, but it is only (I hope) a > temporary description of the state of the art of psycholinguistics. > At least some of us would like our methods to become sensitive enough > to individual differences so that we can look at how 'the general > processes and principles' interact with the level of an individual > person's knowledge of particular constructions, to find out how much > each person knows of the patterns 'out there' in the language. Some > experimental methods are almost at that point already; they can > distinguish degrees of mastery of particular constructions of a > language among groups of second-language learners. > Have a look at > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > 2002. Overhearing a language during childhood. Psychological Science > 13.3, 238-243. > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding > on to childhood language memory. Cognition, 86(3), B53-B64. > > Lise > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a psycholinguist's > > view. Your goal is to find general processes and principles that > > apply uniformly across individuals, so you have to use methods to > > check for generality. And (as you know) I admire the way you pursue > > that goal. But my goal, as a linguist, is different. I want to > > explore the structure of a language so that I can understand how all > > the bits fit together. Like you, I'm aiming to model cognition, but > > my focus is on items and structures, and I start from the assumption > > that these can and do vary across speakers. > > > > However, having said all that I do agree with you that linguists > > should all get used to collecting and using quantitative data; and > > with the help of Brian MacWhinney's typology we'd know what methods > > to use when. And I do agree with your points about bid/bidded: in > > cases like that, quantitative data would be at least a very good > > starting point for a proper investigation. > > > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 19:30, Ted Gibson wrote: > >> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I don't understand what > >> is confusing about what Ev Fedorenko and I are claiming. All we are > >> saying is that if you have some testable claim involving a general > >> hypothesis about a language, then you need to get quantitative data > >> from unbiased sources to evaluate that claim. If you are interested > >> in English past tense morphology, then depending on the claims that > >> you might want to investigate, there are lots of ways to get > >> relevant quantitative evidence. Corpus data will probably be > >> useful. For very low frequency words, you can run experiments to > >> test behavior with respect to such words. > >> > >> Your example of the past tense of "bid" is a fine such example. You > >> can run an experiment like the one you suggested to find out what > >> people think the past tense is. If you then found that 20/50 people > >> responded "bidded" and 30/50 respond "bid", that is a lot of useful > >> information. As you suggest in your discussion, this result > >> wouldn't answer the question of how past tense is stored in each > >> individual. This result would be ambiguous among several possible > >> explanations. One possibility is that the probability distribution > >> that is being discovered reflects different dialects, such that 2/5 > >> of the population has one past tense, and 3/5 has another. Another > >> possibility is that each person has a similar probability > >> distribution in their heads, such that 2/5 of the time I respond > >> one way, and 3/5 of the time I respond another. Further experiments > >> would be necessary to answer between these and other possible > >> theories (e.g., with repeated trials from the same person, > >> carefully planned so that the participants don't notice that they > >> are being asked multiple times). Without the quantitative evidence > >> in the first place, there is no way to answer these kinds of > >> questions. > >> > >> Regarding the past tense of "go", this would be useful as a > >> baseline in an experiment involving the less frequent ones. So, > >> yes, it would useful to gather quantitative evidence in such a case > >> also, as baselines with respect to the more interesting cases for > >> theories. > >> > >> The bottom line: if you have a generalization about a language that > >> you wish to evaluate (such that you hypothesize that it is true > >> across the speakers of the language), then you need quantitative > >> evidence from multiple individuals, using an unbiased data > >> collection method, to evaluate such a claim. The point about > >> Mechanical Turk is that it is really *easy* to do this now, at > >> least for languages like English. > >> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> > >>> Dear Ted, > >>> Thanks for the very interesting comment, but are you REALLY saying > >>> that I shouldn't claim, for example, that the past tense of GO is > >>> "went" without first cross-checking with 50 native speakers? > >>> > >>> Isn't there a danger of missing the point that we all, as native > >>> speakers, spend our whole lives scanning other people's linguistic > >>> behaviour (language 'out there', E-language) and trying to explain > >>> it to ourselves in terms of a language system (language 'in here', > >>> I-language)? So every judgement we make is based on thousands or > >>> millions of observed exemplars, and reflects a unique experience > >>> of E-language filtered through a unique I-language. > >>> > >>> Given that view of language development, I don't see how > >>> quantitative data will help. Let's take a real uncertainty, such > >>> as the past tense of BID. If I want to say I did it, do I say "I > >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>> and 30 "bid". So what? Does that mean that the correct answer is > >>> "bidded"? Surely not. How is it better than my judgement? I agree > >>> you could record my speech and find how often I use each > >>> alternative; but the reason I don't know is precisely because it's > >>> a rare word, so in a sense quantitative data are irrelevant even > >>> there. What would solve the problem of subjectivity, of course, > >>> would be a machine for probing the bit of my mind (or even brain) > >>> that holds BID and its details; but I suspect that even that > >>> wouldn't move us much further forward than my original "don't > >>> know". (Incidentally I write as a fan of quantitative > >>> sociolinguistics, so I do accept that quantitative data are > >>> relevant to linguistic analysis in some areas, where the I- > >>> language phenomenon is frequent enough to produce usable data.) > >>> > >>> It seems to me that this discussion raises the really fundamental > >>> question of what kind of thing we think language is: social or > >>> individual. The problem isn't unique to linguistics of course; > >>> it's the same throughout the social sciences. But what's special > >>> about linguistics is that we deal in very fine details of culture > >>> (e.g. details of how a particular word is used or pronounced) so > >>> the differences between individuals really matter. I don't see > >>> that we're ever going to have anything better than judgements to > >>> go on, so what we need is a way to ensure that judgements are > >>> accurate reports of individual I-language. A rotten situation for > >>> a science, but I don't see how it can get better. > >>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> > >>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points that Dan Everett makes, in > >>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a couple of papers recently > >>>> (Gibson & > >>>> Fedorenko, 2010, in press, see references and links below) on > >>>> what we > >>>> think are weak methodological standards in syntax and semantics > >>>> research over the past many years. The issue that we address is the > >>>> prevalent method in syntax and semantics research, which involves > >>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>> pair, typically by just the author of the paper, sometimes with > >>>> feedback from colleagues. As we address in our papers, this > >>>> methodology does not allow proper testing of scientific hypotheses > >>>> because of (a) the small number of experimental participants > >>>> (typically one); (b) the small number of experimental stimuli > >>>> (typically one); (c) cognitive biases on the part of the researcher > >>>> and participants; and (d) the effect of the preceding context > >>>> (e.g., > >>>> other constructions the researcher may have been recently > >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>> points, but with not as strong a conclusion as ours). > >>>> > >>>> Three issues need to be separated here: (1) the use of intuitive > >>>> judgments as a dependent measure in a language experiment; (2) > >>>> potential cognitive biases on the part of experimental subjects and > >>>> experimenters in language experiments; and (3) the need for > >>>> obtaining > >>>> quantitative evidence, whatever the dependent measure might be. The > >>>> paper that Ev and I wrote addresses the last two issues, but does > >>>> not > >>>> go into depth on the first issue (the use of intuitions as a > >>>> dependent > >>>> measure in language experiments). Regarding this issue, we don't > >>>> think > >>>> that there is anything wrong with gathering intuitive judgments > >>>> as a > >>>> dependent measure, as long as the task is clear to the experimental > >>>> participants. > >>>> > >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to > >>>> some > >>>> arguments that have been given in support of continuing to use the > >>>> traditional non-quantitative method in syntax / semantics research. > >>>> One recent defense of the traditional method comes from Phillips > >>>> (2008), who argues that no harm has come from the non-quantitative > >>>> approach in syntax research thus far. Phillips argues that there > >>>> are > >>>> no cases in the literature where an incorrect intuitive judgment > >>>> has > >>>> become the basis for a widely accepted generalization or an > >>>> important > >>>> theoretical claim. He therefore concludes that there is no reason > >>>> to > >>>> adopt more rigorous data collection standards. We challenge > >>>> Philips? > >>>> conclusion by presenting three cases from the literature where a > >>>> faulty intuition has led to incorrect generalizations and mistaken > >>>> theorizing, plausibly due to cognitive biases on the part of the > >>>> researchers. > >>>> > >>>> A second argument that is sometimes presented for the continued > >>>> use of > >>>> the traditional non-quantitative method is that it would be too > >>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>> phenomenon quantitatively. For example, Culicover & Jackendoff > >>>> (2010) > >>>> make this argument explicitly in their response to Gibson & > >>>> Fedorenko > >>>> (2010): ?It would cripple linguistic investigation if it were > >>>> required > >>>> that all judgments of ambiguity and grammaticality be subject to > >>>> statistically rigorous experiments on naive subjects, especially > >>>> when > >>>> investigating languages whose speakers are hard to > >>>> access? (Culicover > >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>> earlier in the discussion here.) Whereas we agree that in > >>>> circumstances where gathering data is difficult, some evidence is > >>>> better than no evidence, we do not agree that research would be > >>>> slowed > >>>> with respect to languages where experimental participants are > >>>> easy to > >>>> access, such as English. In contrast, we think that the opposite is > >>>> true: the field?s progress is probably slowed by not doing > >>>> quantitative research. > >>>> Suppose that a typical syntax / semantics paper that lacks > >>>> quantitative evidence includes judgments for 50 or more sentences / > >>>> meaning pairs, corresponding to 50 or more empirical claims. Even > >>>> if > >>>> most of the judgments from such a paper are correct or are on the > >>>> right track, the problem is in knowing which judgments are correct. > >>>> For example, suppose that 90% of the judgments from an arbitrary > >>>> paper > >>>> are correct (which is probably a high estimate). (Colin Phillips > >>>> and > >>>> some of his former students / postdocs have commented to us that, > >>>> in > >>>> their experience, quantitative acceptability judgment studies > >>>> almost > >>>> always validate the claim(s) in the literature. This is not our > >>>> experience, however. Most experiments that we have run which > >>>> attempt > >>>> to test some syntactic / semantic hypothesis in the literature > >>>> end up > >>>> providing us with a pattern of data that had not been known > >>>> before the > >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This > >>>> means > >>>> that in a paper with 50 empirical claims 45/50 are correct. But > >>>> which > >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s > >>>> over > >>>> two million different theories. By quantitatively evaluating the > >>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>> progress, it is better to have theoretical claims supported by > >>>> solid > >>>> quantitative evidence, so that even if the interpretation of the > >>>> data > >>>> changes over time as new evidence becomes available ? as is often > >>>> the > >>>> case in any field of science ? the empirical pattern can be used > >>>> as a > >>>> basis for further theorizing. > >>>> > >>>> Furthermore, it is no longer expensive to run behavioral > >>>> experiments, > >>>> at least in English and other widely spoken languages. There now > >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? > >>>> which > >>>> can be used for collecting behavioral data over the internet > >>>> quickly > >>>> and inexpensively. The cost of using an interface like this is > >>>> minimal, and the time that it takes for the results to be > >>>> returned is > >>>> short. For example, currently on Mechanical Turk, a survey of > >>>> approximately 50 items will be answered by 50 or more participants > >>>> within a couple of hours, at a cost of approximately $1 per > >>>> participant. Thus a survey can be completed within a day, at a > >>>> cost of > >>>> less than $50. (The hard work of designing the experiment, and > >>>> constructing controlled materials remains of course.) > >>>> > >>>> Sorry to be so verbose. But I think that these methodological > >>>> points > >>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>> methods in syntax and semantics research. Language and Cognitive > >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>> & Fedorenko InPress LCP.pdf > >>>> > >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >>>> Fedorenko > >>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about methodology. I believe > >>>>> that > >>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>> them - to a degree. But while it might not have been feasible for > >>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>> Grammar to conduct experiments on every point of the grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>> (indeed, Standard Social Science Methodology) is vital for > >>>>> taking the > >>>>> field forward and for providing the best support for different > >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful > >>>>> new > >>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>> sole source of evidence, in their paper: "The need for > >>>>> quantitative > >>>>> methods in syntax and semantics research". > >>>>> > >>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team from Stanford, MIT (Brain and > >>>>> Cognitive > >>>>> Science), and researchers from Brazil are beginning a third > >>>>> round of > >>>>> experimental work among the Pirahas, since my own work on the > >>>>> syntax > >>>>> was, like almost every other field researcher's, based on native > >>>>> speaker intuitions and corpora. > >>>>> > >>>>> The discussion of methodologies reminds me of the initial > >>>>> reactions > >>>>> to Greenberg's work on classifying the languages of the > >>>>> Americas. His > >>>>> methods were strongly (and justifiably) criticized. However, I > >>>>> always > >>>>> thought that his methods were a great way of generating > >>>>> hypotheses, > >>>>> so long as they were ultimately put to the test of standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions. > >>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a further layer of explanation to the > >>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>> every detailed analysis we make, our whole discipline will > >>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>> judgements are what put us linguists ahead of the rest in > >>>>>> handling > >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>> Language (or the OED) without using native speaker judgements. > >>>>>> > >>>>>> Best wishes, Dick Hudson > >>>> > >>>> > >>>> > >> > >> > >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > > > ------------------------------ > > Message: 19 > Date: Fri, 10 Sep 2010 21:05:18 -0400 > From: Daniel Everett > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Ted Gibson , Richard Hudson > , Daniel Everett >, > Evelina Fedorenko , funknet > > Message-ID: > Content-Type: text/plain; charset=windows-1252 > > I think that Brian and Dick make excellent points. There are very good > grammars written that could be improved by psycholinguistic experimentation > and more quantitative approaches. But large sections of those grammars > aren't going to change one bit (go-went) with quantitative tests and such > tests would be completely counterproductive given the shortness of life and > the vastness of the field linguist's tasks. > > Part of the problem is that linguistics is not simply a subdiscipline of > psychology. Linguistics has its own objectives and those only occasionally > overlap with psychology. The same for methods. > > On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us). We study different things and > have different reasons for being satisfied with the results we achieve. > > I believe that we linguists are often complacent and fail to apply better > methods. But of course that applies to all disciplines. > > In the meantime, checking corpora, collecting data as a result of careful > interviews with native speakers, and the other aspects of the field > linguist's task are vital parts of the linguist's task and much of this > won't be improved by quantitative methods as we currently understand them. > Maybe sometime. > > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that they > showed the danger of using intuitions. What I meant to say of using > intuitions as standardly used by linguists. They convinced me that there is > a lot to learn from quantitative methods. > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a psycholinguist's view. > Your goal is to find general processes and principles that apply uniformly > across individuals, so you have to use methods to check for generality. And > (as you know) I admire the way you pursue that goal. But my goal, as a > linguist, is different. I want to explore the structure of a language so > that I can understand how all the bits fit together. Like you, I'm aiming to > model cognition, but my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers. > > > > However, having said all that I do agree with you that linguists should > all get used to collecting and using quantitative data; and with the help of > Brian MacWhinney's typology we'd know what methods to use when. And I do > agree with your points about bid/bidded: in cases like that, quantitative > data would be at least a very good starting point for a proper > investigation. > > > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 19:30, Ted Gibson wrote: > >> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are saying is > that if you have some testable claim involving a general hypothesis about a > language, then you need to get quantitative data from unbiased sources to > evaluate that claim. If you are interested in English past tense morphology, > then depending on the claims that you might want to investigate, there are > lots of ways to get relevant quantitative evidence. Corpus data will > probably be useful. For very low frequency words, you can run experiments to > test behavior with respect to such words. > >> > >> Your example of the past tense of "bid" is a fine such example. You can > run an experiment like the one you suggested to find out what people think > the past tense is. If you then found that 20/50 people responded "bidded" > and 30/50 respond "bid", that is a lot of useful information. As you suggest > in your discussion, this result wouldn't answer the question of how past > tense is stored in each individual. This result would be ambiguous among > several possible explanations. One possibility is that the probability > distribution that is being discovered reflects different dialects, such that > 2/5 of the population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution in > their heads, such that 2/5 of the time I respond one way, and 3/5 of the > time I respond another. Further experiments would be necessary to answer > between these and other possible theories (e.g., with repeated trials from > the same person, carefully pl > anned so that the participants don't notice that they are being asked > multiple times). Without the quantitative evidence in the first place, there > is no way to answer these kinds of questions. > >> > >> Regarding the past tense of "go", this would be useful as a baseline in > an experiment involving the less frequent ones. So, yes, it would useful to > gather quantitative evidence in such a case also, as baselines with respect > to the more interesting cases for theories. > >> > >> The bottom line: if you have a generalization about a language that you > wish to evaluate (such that you hypothesize that it is true across the > speakers of the language), then you need quantitative evidence from multiple > individuals, using an unbiased data collection method, to evaluate such a > claim. The point about Mechanical Turk is that it is really *easy* to do > this now, at least for languages like English. > >> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> > >>> Dear Ted, > >>> Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > >>> > >>> Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > >>> > >>> Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quant > itative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > >>> > >>> It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > >>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> > >>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points that Dan Everett makes, in > >>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>>> Fedorenko, 2010, in press, see references and links below) on what we > >>>> think are weak methodological standards in syntax and semantics > >>>> research over the past many years. The issue that we address is the > >>>> prevalent method in syntax and semantics research, which involves > >>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>> pair, typically by just the author of the paper, sometimes with > >>>> feedback from colleagues. As we address in our papers, this > >>>> methodology does not allow proper testing of scientific hypotheses > >>>> because of (a) the small number of experimental participants > >>>> (typically one); (b) the small number of experimental stimuli > >>>> (typically one); (c) cognitive biases on the part of the researcher > >>>> and participants; and (d) the effect of the preceding context (e.g., > >>>> other constructions the researcher may have been recently > >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>> points, but with not as strong a conclusion as ours). > >>>> > >>>> Three issues need to be separated here: (1) the use of intuitive > >>>> judgments as a dependent measure in a language experiment; (2) > >>>> potential cognitive biases on the part of experimental subjects and > >>>> experimenters in language experiments; and (3) the need for obtaining > >>>> quantitative evidence, whatever the dependent measure might be. The > >>>> paper that Ev and I wrote addresses the last two issues, but does not > >>>> go into depth on the first issue (the use of intuitions as a dependent > >>>> measure in language experiments). Regarding this issue, we don't think > >>>> that there is anything wrong with gathering intuitive judgments as a > >>>> dependent measure, as long as the task is clear to the experimental > >>>> participants. > >>>> > >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>>> arguments that have been given in support of continuing to use the > >>>> traditional non-quantitative method in syntax / semantics research. > >>>> One recent defense of the traditional method comes from Phillips > >>>> (2008), who argues that no harm has come from the non-quantitative > >>>> approach in syntax research thus far. Phillips argues that there are > >>>> no cases in the literature where an incorrect intuitive judgment has > >>>> become the basis for a widely accepted generalization or an important > >>>> theoretical claim. He therefore concludes that there is no reason to > >>>> adopt more rigorous data collection standards. We challenge Philips? > >>>> conclusion by presenting three cases from the literature where a > >>>> faulty intuition has led to incorrect generalizations and mistaken > >>>> theorizing, plausibly due to cognitive biases on the part of the > >>>> researchers. > >>>> > >>>> A second argument that is sometimes presented for the continued use of > >>>> the traditional non-quantitative method is that it would be too > >>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>>> make this argument explicitly in their response to Gibson & Fedorenko > >>>> (2010): ?It would cripple linguistic investigation if it were required > >>>> that all judgments of ambiguity and grammaticality be subject to > >>>> statistically rigorous experiments on naive subjects, especially when > >>>> investigating languages whose speakers are hard to access? (Culicover > >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>> earlier in the discussion here.) Whereas we agree that in > >>>> circumstances where gathering data is difficult, some evidence is > >>>> better than no evidence, we do not agree that research would be slowed > >>>> with respect to languages where experimental participants are easy to > >>>> access, such as English. In contrast, we think that the opposite is > >>>> true: the field?s progress is probably slowed by not doing > >>>> quantitative research. > >>>> Suppose that a typical syntax / semantics paper that lacks > >>>> quantitative evidence includes judgments for 50 or more sentences / > >>>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>>> most of the judgments from such a paper are correct or are on the > >>>> right track, the problem is in knowing which judgments are correct. > >>>> For example, suppose that 90% of the judgments from an arbitrary paper > >>>> are correct (which is probably a high estimate). (Colin Phillips and > >>>> some of his former students / postdocs have commented to us that, in > >>>> their experience, quantitative acceptability judgment studies almost > >>>> always validate the claim(s) in the literature. This is not our > >>>> experience, however. Most experiments that we have run which attempt > >>>> to test some syntactic / semantic hypothesis in the literature end up > >>>> providing us with a pattern of data that had not been known before the > >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>>> two million different theories. By quantitatively evaluating the > >>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>> progress, it is better to have theoretical claims supported by solid > >>>> quantitative evidence, so that even if the interpretation of the data > >>>> changes over time as new evidence becomes available ? as is often the > >>>> case in any field of science ? the empirical pattern can be used as a > >>>> basis for further theorizing. > >>>> > >>>> Furthermore, it is no longer expensive to run behavioral experiments, > >>>> at least in English and other widely spoken languages. There now > >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>>> can be used for collecting behavioral data over the internet quickly > >>>> and inexpensively. The cost of using an interface like this is > >>>> minimal, and the time that it takes for the results to be returned is > >>>> short. For example, currently on Mechanical Turk, a survey of > >>>> approximately 50 items will be answered by 50 or more participants > >>>> within a couple of hours, at a cost of approximately $1 per > >>>> participant. Thus a survey can be completed within a day, at a cost of > >>>> less than $50. (The hard work of designing the experiment, and > >>>> constructing controlled materials remains of course.) > >>>> > >>>> Sorry to be so verbose. But I think that these methodological points > >>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>> methods in syntax and semantics research. Language and Cognitive > >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>> & Fedorenko InPress LCP.pdf > >>>> > >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > Fedorenko > >>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about methodology. I believe that > >>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>> them - to a degree. But while it might not have been feasible for > >>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>> Grammar to conduct experiments on every point of the grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>>> field forward and for providing the best support for different > >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>> sole source of evidence, in their paper: "The need for quantitative > >>>>> methods in syntax and semantics research". > >>>>> > >>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>>> Science), and researchers from Brazil are beginning a third round of > >>>>> experimental work among the Pirahas, since my own work on the syntax > >>>>> was, like almost every other field researcher's, based on native > >>>>> speaker intuitions and corpora. > >>>>> > >>>>> The discussion of methodologies reminds me of the initial reactions > >>>>> to Greenberg's work on classifying the languages of the Americas. His > >>>>> methods were strongly (and justifiably) criticized. However, I always > >>>>> thought that his methods were a great way of generating hypotheses, > >>>>> so long as they were ultimately put to the test of standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions. > >>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a further layer of explanation to the > >>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>> every detailed analysis we make, our whole discipline will > >>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>> judgements are what put us linguists ahead of the rest in handling > >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>> Language (or the OED) without using native speaker judgements. > >>>>>> > >>>>>> Best wishes, Dick Hudson > >>>> > >>>> > >>>> > >> > >> > >> > > > > > > ------------------------------ > > Message: 20 > Date: Sat, 11 Sep 2010 10:03:35 +0000 (GMT) > From: Philippe De Brabanter > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: funknet at mailman.rice.edu > Message-ID: <650247.87750.qm at web25504.mail.ukl.yahoo.com> > Content-Type: text/plain; charset=utf-8 > > Dear all, > > this is just to say that the Cambridge Grammar of the English Language, > like > several other grammars I'm aware of, does mention the sorts of > constructions Tom > is wondering about (esp. pp. 964-67). They're treated as noun complements, > whereas relative clauses usually function as modifiers of nouns. > H&P give a useful list of nouns licensing these complements (a list which > confirms Suzanne Kemmerer's point that these nouns do not always have a > verbal > counterpart taking a content clause as its complement ? H&P suggest that > the > most frequent of those licensing nouns is fact). > They also point out that content clauses can also sometimes function as > supplements (i.e. appositives), as in > > I'm inclined to favour your first suggestion, that we shelve the proposal > until > after the election. > > This confirms Suzanne's suggestion that we shouldn't say that the clausal > noun > complements are appositives. > > One last interesting point. On p. 967, H&P show that the licensor may > sometimes > be more than just a noun, with certain constructions like have + licensing > NP or > existential there + be facilitating (or being conditions for) the clausal > noun > complement: > > The present system has the disadvantage that it is inordinately > complicated. > vs. ? The disadvantage that it is inordinately complicated has been > overlooked. > > Probably an example like > > This principle may ground some optimism that the account can be usefully > pursued. (M. Sainsbury 2002: "Reference and anaphora", Mind & Language) > > also derives its acceptability from a construction rather than from just > optimism. > > Best, > > Philippe De Brabanter > Paris 4 - Sorbonne > > > > ________________________________ > From: E.G. > To: funknet at mailman.rice.edu > Sent: Fri, 10 September, 2010 19:56:23 > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > > Jespersen and his nexus-substantives should be mentioned (Philosophy of > Grammar, 1924). Also in his MEG and Analytic Syntax one could find > interesting discussions. > > Eitan > > > On 10 September 2010 20:53, Giuliana Fiorentino < > giuliana.fiorentino at unimol.it> wrote: > > > Hi Tom, > > clauses like: > > > > The importance of being Earnest > > the fact of being late > > the fact that you are late > > the idea that world is round > > etcetera > > > > are not relative clauses but can be considered among syntactic strategies > > in order to nominalise events after a generic noun (working as a > classifier > > for nominalised events). > > > > Giuliana > > > > ----- Original Message ----- > > From: Thomas E. Payne > > To: FUNKNET > > Sent: Friday, September 10, 2010 4:16 PM > > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > > > > Can anyone help me name the following structure in English, and maybe > > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar > books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > > > > ------------------------------ > > Message: 21 > Date: Sat, 11 Sep 2010 12:03:56 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Cc: Jennifer Smith > Message-ID: <4C8B621C.6090609 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Dear Lise, > Many thanks for these references, which I've just looked at. As you say, > they do offer hope that we can apply the methods of psycholinguists to > the questions of linguists (e.g. how do individuals categorise Korean or > Spanish consonants?), and at the same time build a clearer understanding > of how these individuals' I-language is related to their E-language > (i.e. the language they've heard). That is encouraging, not least > because it breaks down what I see as a gulf between linguists and > psycholinguists. > > But even better, it complements very nicely some work in > sociolinguistics which I think you'd enjoy reading. It's by a young > Scottish sociolinguist called Jennifer Smith, who did a very careful > quantitative study (with two colleagues) of 24 3-year olds in a small > fishing town in the north of Scotland. She recorded each of them with > their mother, and then analysed two sociolinguistic variables that she'd > also analysed in adult speech: a phonological variable (pronunciation of > the /au/ vowel in "cow", and a morphosyntactic one, the use of -s on a > verb with a plural subject, e.g. "My trousers is falling doon." In both > cases usage is variable, so the analysis produces a percentage score for > each speaker (e.g. 5% of words with the /au/ vowel have a monophthong). > She then compared the children's scores with those of their mothers, and > found an astonishingly close match for the phonological variable but no > match at all for the morphosyntactic one. The reference is > > Jennifer Smith, Mercedes Durham & Liane Fortune, 2007. 'Mam, my > trousers is fa'in doon!' Community, caregiver and child in the > acquisition of variation in a Scottish dialect. (Language, Variation > and Change 19. 63-99) > > Once again we have evidence that output can be closely linked to input, > we have a nice quantitative method, and we see an example of the > 'global' analysis of language that we should all be striving for: one > which embraces both E-language (directly observable to the learner as > well as to the linguist) and I-language (only indirectly accessible to > both), and which tries to both describe and explain the relation between > the two. Neither E-language nor I-language is the 'real' language - > they're both part of it. And this global enterprise needs all the > methods we can muster. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 11/09/2010 02:03, Lise Menn wrote: > > Dick, I think that what you say is true, but it is only (I hope) a > > temporary description of the state of the art of psycholinguistics. > > At least some of us would like our methods to become sensitive enough > > to individual differences so that we can look at how 'the general > > processes and principles' interact with the level of an individual > > person's knowledge of particular constructions, to find out how much > > each person knows of the patterns 'out there' in the language. Some > > experimental methods are almost at that point already; they can > > distinguish degrees of mastery of particular constructions of a > > language among groups of second-language learners. > > Have a look at > > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > > 2002. Overhearing a language during childhood. /Psychological Science/ > > 13.3, 238-243. > > > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on > > to childhood language memory. /Cognition, 86/(3), B53-B64. > > > > Lise > > > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > > > >> Dear Ted and Ev, > >> Yes, I understand your view, but I think it's a psycholinguist's > >> view. Your goal is to find general processes and principles that > >> apply uniformly across individuals, so you have to use methods to > >> check for generality. And (as you know) I admire the way you pursue > >> that goal. But my goal, as a linguist, is different. I want to > >> explore the structure of a language so that I can understand how all > >> the bits fit together. Like you, I'm aiming to model cognition, but > >> my focus is on items and structures, and I start from the assumption > >> that these can and do vary across speakers. > >> > >> However, having said all that I do agree with you that linguists > >> should all get used to collecting and using quantitative data; and > >> with the help of Brian MacWhinney's typology we'd know what methods > >> to use when. And I do agree with your points about bid/bidded: in > >> cases like that, quantitative data would be at least a very good > >> starting point for a proper investigation. > >> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> > >> On 10/09/2010 19:30, Ted Gibson wrote: > >>> Dear Dick: > >>> > >>> Perhaps we are talking at cross purposes. I don't understand what is > >>> confusing about what Ev Fedorenko and I are claiming. All we are > >>> saying is that if you have some testable claim involving a general > >>> hypothesis about a language, then you need to get quantitative data > >>> from unbiased sources to evaluate that claim. If you are interested > >>> in English past tense morphology, then depending on the claims that > >>> you might want to investigate, there are lots of ways to get > >>> relevant quantitative evidence. Corpus data will probably be useful. > >>> For very low frequency words, you can run experiments to test > >>> behavior with respect to such words. > >>> > >>> Your example of the past tense of "bid" is a fine such example. You > >>> can run an experiment like the one you suggested to find out what > >>> people think the past tense is. If you then found that 20/50 people > >>> responded "bidded" and 30/50 respond "bid", that is a lot of useful > >>> information. As you suggest in your discussion, this result wouldn't > >>> answer the question of how past tense is stored in each individual. > >>> This result would be ambiguous among several possible explanations. > >>> One possibility is that the probability distribution that is being > >>> discovered reflects different dialects, such that 2/5 of the > >>> population has one past tense, and 3/5 has another. Another > >>> possibility is that each person has a similar probability > >>> distribution in their heads, such that 2/5 of the time I respond one > >>> way, and 3/5 of the time I respond another. Further experiments > >>> would be necessary to answer between these and other possible > >>> theories (e.g., with repeated trials from the same person, carefully > >>> planned so that the participants don't notice that they are being > >>> asked multiple times). Without the quantitative evidence in the > >>> first place, there is no way to answer these kinds of questions. > >>> > >>> Regarding the past tense of "go", this would be useful as a baseline > >>> in an experiment involving the less frequent ones. So, yes, it would > >>> useful to gather quantitative evidence in such a case also, as > >>> baselines with respect to the more interesting cases for theories. > >>> > >>> The bottom line: if you have a generalization about a language that > >>> you wish to evaluate (such that you hypothesize that it is true > >>> across the speakers of the language), then you need quantitative > >>> evidence from multiple individuals, using an unbiased data > >>> collection method, to evaluate such a claim. The point about > >>> Mechanical Turk is that it is really *easy* to do this now, at least > >>> for languages like English. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson & Ev Fedorenko > >>> > >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >>> > >>>> Dear Ted, > >>>> Thanks for the very interesting comment, but are you REALLY saying > >>>> that I shouldn't claim, for example, that the past tense of GO is > >>>> "went" without first cross-checking with 50 native speakers? > >>>> > >>>> Isn't there a danger of missing the point that we all, as native > >>>> speakers, spend our whole lives scanning other people's linguistic > >>>> behaviour (language 'out there', E-language) and trying to explain > >>>> it to ourselves in terms of a language system (language 'in here', > >>>> I-language)? So every judgement we make is based on thousands or > >>>> millions of observed exemplars, and reflects a unique experience of > >>>> E-language filtered through a unique I-language. > >>>> > >>>> Given that view of language development, I don't see how > >>>> quantitative data will help. Let's take a real uncertainty, such as > >>>> the past tense of BID. If I want to say I did it, do I say "I > >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>>> and 30 "bid". So what? Does that mean that the correct answer is > >>>> "bidded"? Surely not. How is it better than my judgement? I agree > >>>> you could record my speech and find how often I use each > >>>> alternative; but the reason I don't know is precisely because it's > >>>> a rare word, so in a sense quantitative data are irrelevant even > >>>> there. What would solve the problem of subjectivity, of course, > >>>> would be a machine for probing the bit of my mind (or even brain) > >>>> that holds BID and its details; but I suspect that even that > >>>> wouldn't move us much further forward than my original "don't > >>>> know". (Incidentally I write as a fan of quantitative > >>>> sociolinguistics, so I do accept that quantitative data are > >>>> relevant to linguistic analysis in some areas, where the I-language > >>>> phenomenon is frequent enough to produce usable data.) > >>>> > >>>> It seems to me that this discussion raises the really fundamental > >>>> question of what kind of thing we think language is: social or > >>>> individual. The problem isn't unique to linguistics of course; it's > >>>> the same throughout the social sciences. But what's special about > >>>> linguistics is that we deal in very fine details of culture (e.g. > >>>> details of how a particular word is used or pronounced) so the > >>>> differences between individuals really matter. I don't see that > >>>> we're ever going to have anything better than judgements to go on, > >>>> so what we need is a way to ensure that judgements are accurate > >>>> reports of individual I-language. A rotten situation for a science, > >>>> but I don't see how it can get better. > >>>> > >>>> Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> > >>>> > >>>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>>> Dear Dan, Dick: > >>>>> > >>>>> I would like to clarify some points that Dan Everett makes, in > >>>>> response to Dick Hudson. > >>>>> > >>>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>>>> Fedorenko, 2010, in press, see references and links below) on what we > >>>>> think are weak methodological standards in syntax and semantics > >>>>> research over the past many years. The issue that we address is the > >>>>> prevalent method in syntax and semantics research, which involves > >>>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>>> pair, typically by just the author of the paper, sometimes with > >>>>> feedback from colleagues. As we address in our papers, this > >>>>> methodology does not allow proper testing of scientific hypotheses > >>>>> because of (a) the small number of experimental participants > >>>>> (typically one); (b) the small number of experimental stimuli > >>>>> (typically one); (c) cognitive biases on the part of the researcher > >>>>> and participants; and (d) the effect of the preceding context (e.g., > >>>>> other constructions the researcher may have been recently > >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>>> points, but with not as strong a conclusion as ours). > >>>>> > >>>>> Three issues need to be separated here: (1) the use of intuitive > >>>>> judgments as a dependent measure in a language experiment; (2) > >>>>> potential cognitive biases on the part of experimental subjects and > >>>>> experimenters in language experiments; and (3) the need for obtaining > >>>>> quantitative evidence, whatever the dependent measure might be. The > >>>>> paper that Ev and I wrote addresses the last two issues, but does not > >>>>> go into depth on the first issue (the use of intuitions as a > dependent > >>>>> measure in language experiments). Regarding this issue, we don't > think > >>>>> that there is anything wrong with gathering intuitive judgments as a > >>>>> dependent measure, as long as the task is clear to the experimental > >>>>> participants. > >>>>> > >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>>>> arguments that have been given in support of continuing to use the > >>>>> traditional non-quantitative method in syntax / semantics research. > >>>>> One recent defense of the traditional method comes from Phillips > >>>>> (2008), who argues that no harm has come from the non-quantitative > >>>>> approach in syntax research thus far. Phillips argues that there are > >>>>> no cases in the literature where an incorrect intuitive judgment has > >>>>> become the basis for a widely accepted generalization or an important > >>>>> theoretical claim. He therefore concludes that there is no reason to > >>>>> adopt more rigorous data collection standards. We challenge Philips' > >>>>> conclusion by presenting three cases from the literature where a > >>>>> faulty intuition has led to incorrect generalizations and mistaken > >>>>> theorizing, plausibly due to cognitive biases on the part of the > >>>>> researchers. > >>>>> > >>>>> A second argument that is sometimes presented for the continued use > of > >>>>> the traditional non-quantitative method is that it would be too > >>>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>>>> make this argument explicitly in their response to Gibson & Fedorenko > >>>>> (2010): "It would cripple linguistic investigation if it were > required > >>>>> that all judgments of ambiguity and grammaticality be subject to > >>>>> statistically rigorous experiments on naive subjects, especially when > >>>>> investigating languages whose speakers are hard to access" (Culicover > >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>>> earlier in the discussion here.) Whereas we agree that in > >>>>> circumstances where gathering data is difficult, some evidence is > >>>>> better than no evidence, we do not agree that research would be > slowed > >>>>> with respect to languages where experimental participants are easy to > >>>>> access, such as English. In contrast, we think that the opposite is > >>>>> true: the field's progress is probably slowed by not doing > >>>>> quantitative research. > >>>>> Suppose that a typical syntax / semantics paper that lacks > >>>>> quantitative evidence includes judgments for 50 or more sentences / > >>>>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>>>> most of the judgments from such a paper are correct or are on the > >>>>> right track, the problem is in knowing which judgments are correct. > >>>>> For example, suppose that 90% of the judgments from an arbitrary > paper > >>>>> are correct (which is probably a high estimate). (Colin Phillips and > >>>>> some of his former students / postdocs have commented to us that, in > >>>>> their experience, quantitative acceptability judgment studies almost > >>>>> always validate the claim(s) in the literature. This is not our > >>>>> experience, however. Most experiments that we have run which attempt > >>>>> to test some syntactic / semantic hypothesis in the literature end up > >>>>> providing us with a pattern of data that had not been known before > the > >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>>>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That's over > >>>>> two million different theories. By quantitatively evaluating the > >>>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>>> progress, it is better to have theoretical claims supported by solid > >>>>> quantitative evidence, so that even if the interpretation of the data > >>>>> changes over time as new evidence becomes available -- as is often > the > >>>>> case in any field of science -- the empirical pattern can be used as > a > >>>>> basis for further theorizing. > >>>>> > >>>>> Furthermore, it is no longer expensive to run behavioral experiments, > >>>>> at least in English and other widely spoken languages. There now > >>>>> exists a marketplace interface -- Amazon.com's Mechanical Turk -- > >>>>> which > >>>>> can be used for collecting behavioral data over the internet quickly > >>>>> and inexpensively. The cost of using an interface like this is > >>>>> minimal, and the time that it takes for the results to be returned is > >>>>> short. For example, currently on Mechanical Turk, a survey of > >>>>> approximately 50 items will be answered by 50 or more participants > >>>>> within a couple of hours, at a cost of approximately $1 per > >>>>> participant. Thus a survey can be completed within a day, at a cost > of > >>>>> less than $50. (The hard work of designing the experiment, and > >>>>> constructing controlled materials remains of course.) > >>>>> > >>>>> Sorry to be so verbose. But I think that these methodological points > >>>>> are very important. > >>>>> > >>>>> Best wishes, > >>>>> > >>>>> Ted Gibson > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>>> methods in syntax and semantics research. Language and Cognitive > >>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>>> & Fedorenko InPress LCP.pdf > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > Fedorenko > >>>>> 2010 TICS.pdf > >>>>> > >>>>> > >>>>> > >>>>> > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about methodology. I believe that > >>>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>>> them - to a degree. But while it might not have been feasible for > >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>>> Grammar to conduct experiments on every point of the grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>>> (indeed, Standard Social Science Methodology) is vital for taking > the > >>>>>> field forward and for providing the best support for different > >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>>> sole source of evidence, in their paper: "The need for quantitative > >>>>>> methods in syntax and semantics research". > >>>>>> > >>>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>>> convincingly on this. > >>>>>> > >>>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>>>> Science), and researchers from Brazil are beginning a third round of > >>>>>> experimental work among the Pirahas, since my own work on the syntax > >>>>>> was, like almost every other field researcher's, based on native > >>>>>> speaker intuitions and corpora. > >>>>>> > >>>>>> The discussion of methodologies reminds me of the initial reactions > >>>>>> to Greenberg's work on classifying the languages of the Americas. > His > >>>>>> methods were strongly (and justifiably) criticized. However, I > always > >>>>>> thought that his methods were a great way of generating hypotheses, > >>>>>> so long as they were ultimately put to the test of standard > >>>>>> historical linguistics methods. And the same seems true for use of > >>>>>> native-speaker intuitions. > >>>>>> > >>>>>> -- Dan > >>>>> > >>>>> > >>>>> > >>>>>>> We linguists can add a further layer of explanation to the > >>>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>>> every detailed analysis we make, our whole discipline will > >>>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>>> judgements are what put us linguists ahead of the rest in handling > >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>>> Language (or the OED) without using native speaker judgements. > >>>>>>> > >>>>>>> Best wishes, Dick Hudson > >>>>> > >>>>> > >>>>> > >>> > >>> > >>> > > > > Lise Menn Home Office: 303-444-4274 > > 1625 Mariposa Ave Fax: 303-413-0017 > > Boulder CO 80302 > > > > Professor Emerita of Linguistics > > Fellow, Institute of Cognitive Science > > University of Colorado > > > > Secretary, AAAS Section Z [Linguistics] > > > > Campus Mail Address: > > UCB 594, Institute of Cognitive Science > > > > Campus Physical Address: > > CINC 234 > > 1777 Exposition Ave, Boulder > > > > > > > > > ------------------------------ > > Message: 22 > Date: Sat, 11 Sep 2010 06:00:08 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > Matt, > > We appear to be fairly close in our approach, but I would have to add that > sometimes there isn't a unique "correct" analysis, because the language > allows equally for several different ones, and how any particular speaker > analyzes a phrase out of context says more about how their individual > brain is wired and less about the language. > > As an example, take the slogan the Coca-Cola company is currently using: > > "Open happiness". > > I first saw it on a cocktail napkin in flight. Reading the English slogan, > my first analysis was that "open" was an adjective modifying the noun > happiness, as opposed to say "closed happiness." > > That seemed weird, so I considered a few other possiblities. Maybe "open" > is a verb in the imperative, and "happiness" is a proper noun in the > vocative, as in "Open Sesame!" > > Then again, it could be that "Happiness" was just a proper noun in > objective case: as in "open America (to tourism)." > > Then I read the French translation on the napkin: "Ouvrez du bonheur." > > "Oh! So this means "open some happiness"!" I said to myself. > > English lexemes are underspecified for category, which is why we need > little words like "some" to disamnbiguate. That's how the language works. > But... all those different analyses could have been correct, given the > proper context, and experimenting even with a large population as to which > one they thought of first would tell you less about the language and more > about the people. > > The only analysis that seems a bit doubtful is the one suggested by the > translation. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > > > Aya, > > > > I actually agree with everything you say here. Personally, I am MORE > > interested in the communicative function of language than I am in > > psycholinguistics and how people process language. > > > > But none of that is relevant, I believe, to the very specific question of > > what it means for an analysis to be correct. While one might conclude > from > > what I said that one ought to do psycholinguistics, that is not my > intention. > > Rather, my conclusion is that since I myself prefer not to do > > psycholinguistics, I cannot really claim that the analyses I come up with > are > > "the correct" ones. And if it is really important to someone that they > > identify "correct" analyses, then they ought to be doing > psycholinguistics, > > since there is no coherent notion of correct analysis outside of what is > > inside of people's heads. > > > > Matthew > > > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > > > wrote: > > > >> Matthew, > >> > >> Thanks for stating that, because I was almost beginning to imagine that > >> there was no essential disagreement, and that all of us agree that there > >> is more -- and less -- to language than what is found in people's > heads. > >> > >> Your position is the one I am familiar with from the functionalist point > >> of view, and I was beginning to feel that it was underrepresented on > >> Funknet. > >> > >> Those of us who disagree with your stated position -- but are very > >> familiar with it -- are interested not just in psycholinguistics and how > >> people process language -- but also in the communicative function of > >> language as a system whereby information is transferred. Just as you and > >> I may not be aware of the way our emails are encoded and then decoded by > >> the computers that help us send emails back and forth, speakers may be > >> compeltely unaware of what language does in order to transmit > information. > >> > >> After speakers have finished sending forth their linguistic output, it > >> matters not at all how they arrived at this output. Language processing > >> is separate from language in the same way that data processing is > >> separate from data. > >> > >> Best, > >> > >> --Aya > >> > >> > >> On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > >> > >>> > >>> The following sentence of Lise's > >>> > >>> "Since we don't in fact have such an agreed-on terminology, we have to > >>> be quite careful in making clear what we are referring to when we talk > >>> about 'the correct analysis' of a form like 'unhappiness'" > >>> > >>> suggests that there are two senses in which an analysis can be "the > >>> correct analysis", one in terms of what is in people's heads, the > other > >>> in terms of what is "out there". > >>> > >>> There are a vast number of patterns "out there". The only distinction, > >>> on my view, amongst this vast number of patterns, is between those > that > >>> reflect something inside people's heads and those that don't. But if > >>> that is the case, then there is no coherent sense in which one can > talk > >>> of "the correct analysis" of what is "out there", except in terms of > >>> what is in people's heads, and thus no second sense of "the correct > >>> analysis". The patterns that don't correspond to things in people's > >>> heads fall into (at least) two categories. There are those that are > >>> akin to constellations of stars and, as with constellations, there is > >>> no reality to these patterns, except in the minds of linguists. And > >>> there are those patterns which are the fossil remains of what was in > >>> the heads of speakers of an earlier stage of the language but which no > >>> longer are. These latter patterns are real, and they are relevant to > >>> exlaining why the language is now the way it is, but they are not > >>> relevant, I think many would agree, as to what is the "correct > analysis" > >>> of the language today. > >>> > >>> For this reason, I claim that the only sense in which an analysis can > be > >>> "the correct analysis" is in terms of what is inside of people's > heads. > >>> > >>> Again, I recommend the work of Bruce Derwing for lengthy discussion of > >>> these issues. > >>> > >>> Matthew > >>> > >>> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > >>> wrote: > >>> > >>>> I wish we had better terminology for keeping track of whether, at a > >>>> given time, we are talking about the patterns that are 'out there' in > >>>> the language and might possibly be apprehended (subconsciously) by a > >>>> speaker, and when we are talking about the patterns that a particular > >>>> speaker actually does apprehend, as indicated by experiments, from > >>>> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >>>> distinguishing among the degrees of pattern apprehension that a person > >>>> may have, from vague preferences detectable in reaction times or other > >>>> behavior all the way up through clear metalinguistic insights. Dick > >>>> Hudson's note reminding us of the Gleitman and Gleitman study is right > >>>> on target. > >>>> > >>>> Since we don't in fact have such an agreed-on terminology, we have > to > >>>> be quite careful in making clear what we are referring to when we talk > >>>> about 'the correct analysis' of a form like 'unhappiness'. We know, > >>>> but tend to forget - and tend to forget to tell our students! - that > >>>> it's an empirical question as to whether the formal simplicity and > >>>> coherence of description of forms 'out there' (e.g. lovely abstract > >>>> morphophonemics) is any kind of approximation to the way knowledge of > >>>> the same forms is organized in a particular person's head. If we > >>>> remember that a very large proportion of what we know about our > >>>> language is 'out there' when we are infants and has to be internalized > >>>> through experience with the language (even if you believe in innate > >>>> 'core language'), the variation in internal knowledge from one person > >>>> to another is more understandable. We especially need to consider (and > >>>> try to test) the > >>>> possibility that > >>>> since > >>>> the brain can make multiple cross-connections, multiple patterns are > >>>> involved > >>>> simultaneously in morphological and syntactic analyses. I suggest that > >>>> that's the case with 'unhappiness' - and the linguistic analyses that > I > >>>> know about are not good at handling that kind of idea. > >>>> > >>>> Lise > >>>> > >>>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >>>> > >>>>> > >>>>> Two comments. > >>>>> > >>>>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>>>> is an > >>>>> important distinction between low-level linguistic intuitions (like > >>>>> whether a > >>>>> word or sentence is well-formed or what it means) and higher-level > >>>>> intuitions > >>>>> (like what the structure of a word or sentence is). One can take > >>>>> the position > >>>>> that we need to account for the former (while recognizing that they > >>>>> are not > >>>>> always reliable) but not the latter. > >>>>> > >>>>> Second, the tension here is not only between evidence from speaker > >>>>> intuitions > >>>>> versus evidence from psycholinguistic experiments. There is also a > >>>>> tension > >>>>> between deciding on the correct analysis on the basis of a priori > >>>>> simplicity > >>>>> arguments versus deciding on the correct analysis on the basis of > >>>>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>>>> paradox that Dan > >>>>> referred to that arises with the word (semantics argues > >>>>> for > >>>>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>>>> +er]] (the > >>>>> comparative suffix can only be attached to adjectives containing one > >>>>> or two > >>>>> syllables) is only a paradox if one assumes that speakers adopt the > >>>>> simplest > >>>>> analysis. For example, if speakers adopt a more complex rule for > >>>>> either of these > >>>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>>>> trisyllabic > >>>>> words beginning with un-), then the bracketing paradox disappears. > >>>>> > >>>>> Matthew > >>>>> > >>>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree > that > >>>>>> conscious judgements are only one kind of evidence that we need to > >>>>>> take > >>>>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>>>> fact), > >>>>>> and of course I've been aware of complaints about judgements by > >>>>>> people > >>>>>> like Labov for decades. > >>>>>> > >>>>>> But you're missing my main point, which is that all judgements > aren't > >>>>>> equally reliable. If you want to know how /unhappiness/ is > >>>>>> structured, > >>>>>> ask a linguist, not a five-year old. And one of the by-products of > >>>>>> education may be increased sensitivity to syntax - which is one of > >>>>>> the > >>>>>> many reasons why linguists need to pay more attention to education. > >>>>>> > >>>>>> Best wishes, Dick > >>>>>> > >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>>>> Dick, > >>>>>>> > >>>>>>> You raise an important issue here about > >>>>>> methodology. I believe that intuitions are a fine way to generate > >>>>>> hypotheses and even to test them - to a degree. But while it might > >>>>>> not have > >>>>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>>>> the > >>>>>> Cambridge Grammar to conduct experiments on every point of the > >>>>>> grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, > >>>>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>>>> Standard > >>>>>> Social Science Methodology) is vital for taking the field forward > >>>>>> and for > >>>>>> providing the best support for different analyses. Ted Gibson and Ev > >>>>>> Fedorenko have written a very useful new paper on this, showing > >>>>>> serious > >>>>>> shortcomings with intuitions as the sole source of evidence, in > their > >>>>>> paper: "The need for quantitative methods in syntax and semantics > >>>>>> research".> > >>>>>>> Carson Schutze and Wayne Cowart, among others, > >>>>>> have also written convincingly on this.> > >>>>>>> It is one reason that a team from Stanford, MIT > >>>>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>>>> beginning a > >>>>>> third round of experimental work among the Pirahas, since my own > >>>>>> work on > >>>>>> the syntax was, like almost every other field researcher's, based > >>>>>> on native > >>>>>> speaker intuitions and corpora.> > >>>>>>> The discussion of methodologies reminds me of > >>>>>> the initial reactions to Greenberg's work on classifying the > >>>>>> languages of > >>>>>> the Americas. His methods were strongly (and justifiably) > criticized. > >>>>>> However, I always thought that his methods were a great way of > >>>>>> generating > >>>>>> hypotheses, so long as they were ultimately put to the test of > >>>>>> standard > >>>>>> historical linguistics methods. And the same seems true for use of > >>>>>> native-speaker intuitions.> > >>>>>>> -- Dan > >>>>>>> > >>>>>>>> We linguists can add a further layer of > >>>>>> explanation to the judgements, but some judgements do seem to be > more > >>>>>> reliable than others. And if we have to wait for psycholinguistic > >>>>>> evidence > >>>>>> for every detailed analysis we make, our whole discipline will > >>>>>> immediately > >>>>>> grind to a halt. Like it or not, native speaker judgements are what > >>>>>> put us > >>>>>> linguists ahead of the rest in handling fine detail. Imagine > >>>>>> writing the > >>>>>> Cambridge Grammar of the English Language (or the OED) without > >>>>>> using native > >>>>>> speaker judgements.>> > >>>>>>>> Best wishes, Dick Hudson > >>>>>>>> > >>>>>>> > >>>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>> > >>>> > >>>> Lise Menn Home Office: 303-444-4274 > >>>> 1625 Mariposa Ave Fax: 303-413-0017 > >>>> Boulder CO 80302 > >>>> > >>>> Professor Emerita of Linguistics > >>>> Fellow, Institute of Cognitive Science > >>>> University of Colorado > >>>> > >>>> Secretary, AAAS Section Z [Linguistics] > >>>> > >>>> Campus Mail Address: > >>>> UCB 594, Institute of Cognitive Science > >>>> > >>>> Campus Physical Address: > >>>> CINC 234 > >>>> 1777 Exposition Ave, Boulder > >>>> > >>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> > >> > >> > > > > > > > > > > > > > ------------------------------ > > Message: 23 > Date: Sat, 11 Sep 2010 12:17:29 +0100 > From: "Chris Butler" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > To: "FUNKNET" > Message-ID: <24693E13135D4D6AAD5A098FAA952480 at OwnerPC> > Content-Type: text/plain; charset="iso-8859-1" > > Dick's comment that "this discussion raises the really fundamental question > of what kind of thing we think language is: social or individual" is, it > seems to me, an important one, particularly for those of us who are > committed functionalists. My own view is that a truly functional model of > language would be one which aims to account for how human beings communicate > using language, or in other words tries to answer the question which was > posed by Simon Dik a long time ago now, but which was not tackled head-on in > his own work: "How does the natural language user work?' In trying to answer > this question we need to accept that language is BOTH social AND individual, > and we need to explore both aspects to get as complete a picture as possible > of how we communicate using language. We need to know BOTH how people create > and respond to meanings and express those meanings in forms during social > interaction AND the mechanisms which operate in the brains of individuals in > order to make > such interaction possible. Both are important parts of the answer to the > question 'How do we communicate using language?', though this particular > thread of the Funknet discussion has concentrated more on the second aspect, > and so will I. > > This doesn't mean that all the work linguists have done on "exploring the > structure of a language so that I can understand how all the bits fit > together" and "exploring the connections between items", as Dick puts it, is > useless - far from it. After all, the hypotheses that psycholinguists test > are based on ideas about what languages are like. But it does mean, in my > view, that ultimately we need to get evidence that the constructs and > analyses we propose are ones that are at least consistent with what we know > of the processes which go on when we use language. So I am with Matthew > when he says that for him, "the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads". Of > course, this doesn't imply that linguists should just give up their jobs > until such time as we know everything there is to know about language > processing. But it does mean that we need to collaborate with > psycholinguists, psychologists and neurologists, > as has also been pointed out by linguists such as Ray Jackendoff, Asif > Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more > with sociolinguists and sociologists, so that we can get a better handle on > the sociocultural aspects of how we communicate.] And it also means that > psycholinguists, for their part, need whenever possible to follow up tightly > controlled lab experiments with studies under more naturalistic conditions, > to avoid the criticism that what happens in artifical lab situations may not > happen in natural communicative conditions. > > I also agree with Dick when he says that "the differences between > individuals really matter", and with Lise when she points out that "we must > also be careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover". However, there > are surely processing mechanisms which are common to all language users by > virtue of the evolution of the language faculty and which constitute the > "general processes" which Dick says psycholinguists are interested in. > > On the issue of quantitative methodology, I'm sympathetic in general to Ted > and Ev's views, though it does seem sensible to prioritise cases in terms of > a hierarchy such as Brian suggests. One thing this means is that we should > be giving our university students of linguistics (and some of our > linguistics lecturers!) courses in quantitative aspects of linguistics that > introduce them to the use of at least some of the basic statistical methods > in language study, and I'm sure this is indeed going on in some enlightened > places. To those who suspect this can't be done with maths-shy students who > don't initially see the need for it, I offer my own experience, over quite a > long period, of teaching such courses to people with little or no prior > experience in quantitative techniques. For some years in the 1990s, I taught > such courses to all linguistics students in an institution where we had many > mature students who had come into university level studies with non-standard > qualificatio > ns, and were not well equipped for courses of this kind by their previous > experience. I'm glad to say that teaching the subject from their own > perspective as language students rather than that of the statistician, and > explaining the reasons for doing things in particular ways rather than just > presenting formulae, paid off in the end, so that most students were able to > appreciate the relevance of these courses and to turn in very creditable > projects showing an understanding of research design and competence in the > use of a range of basic statistical techniques. And I still find that bright > graduate students respond well to similar courses which incorporate some of > the rather more advanced techniques needed for many real research projects > in various areas of linguistics. But I may well be out of date with what is > now already happening in our fine institutions of higher education! > > Chris Butler > > > ------------------------------ > > Message: 24 > Date: Sat, 11 Sep 2010 07:18:29 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > To: Chris Butler > Cc: FUNKNET > Message-ID: > Content-Type: text/plain; charset="us-ascii"; Format="flowed" > > The problem is that once we achieve the psycholinguistic goal and see what > is happening in each person's head, and we see that conflicting analyses > are the norm, rather than the exception, among normal language speakers, > then we will realize that the way language works to transmit information > is despite individual differences and not because of uniform processing > strategies. > > Even when all are agreed as to the meaning of an utterance, they do not > process it the same way. Which means that processing is seocndary to > information transmission. > > --Aya > > > > > On Sat, 11 Sep 2010, Chris Butler wrote: > > > Dick's comment that "this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual" > is, it seems to me, an important one, particularly for those of us who are > committed functionalists. My own view is that a truly functional model of > language would be one which aims to account for how human beings communicate > using language, or in other words tries to answer the question which was > posed by Simon Dik a long time ago now, but which was not tackled head-on in > his own work: "How does the natural language user work?' In trying to answer > this question we need to accept that language is BOTH social AND individual, > and we need to explore both aspects to get as complete a picture as possible > of how we communicate using language. We need to know BOTH how people create > and respond to meanings and express those meanings in forms during social > interaction AND the mechanisms which operate in the brains of individuals in > order to ma > ke such interaction possible. Both are important parts of the answer to > the question 'How do we communicate using language?', though this particular > thread of the Funknet discussion has concentrated more on the second aspect, > and so will I. > > > > This doesn't mean that all the work linguists have done on "exploring the > structure of a language so that I can understand how all the bits fit > together" and "exploring the connections between items", as Dick puts it, is > useless - far from it. After all, the hypotheses that psycholinguists test > are based on ideas about what languages are like. But it does mean, in my > view, that ultimately we need to get evidence that the constructs and > analyses we propose are ones that are at least consistent with what we know > of the processes which go on when we use language. So I am with Matthew > when he says that for him, "the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads". Of > course, this doesn't imply that linguists should just give up their jobs > until such time as we know everything there is to know about language > processing. But it does mean that we need to collaborate with > psycholinguists, psychologists and neurologists > , as has also been pointed out by linguists such as Ray Jackendoff, Asif > Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more > with sociolinguists and sociologists, so that we can get a better handle on > the sociocultural aspects of how we communicate.] And it also means that > psycholinguists, for their part, need whenever possible to follow up tightly > controlled lab experiments with studies under more naturalistic conditions, > to avoid the criticism that what happens in artifical lab situations may not > happen in natural communicative conditions. > > > > I also agree with Dick when he says that "the differences between > individuals really matter", and with Lise when she points out that "we must > also be careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover". However, there > are surely processing mechanisms which are common to all language users by > virtue of the evolution of the language faculty and which constitute the > "general processes" which Dick says psycholinguists are interested in. > > > > On the issue of quantitative methodology, I'm sympathetic in general to > Ted and Ev's views, though it does seem sensible to prioritise cases in > terms of a hierarchy such as Brian suggests. One thing this means is that we > should be giving our university students of linguistics (and some of our > linguistics lecturers!) courses in quantitative aspects of linguistics that > introduce them to the use of at least some of the basic statistical methods > in language study, and I'm sure this is indeed going on in some enlightened > places. To those who suspect this can't be done with maths-shy students who > don't initially see the need for it, I offer my own experience, over quite a > long period, of teaching such courses to people with little or no prior > experience in quantitative techniques. For some years in the 1990s, I taught > such courses to all linguistics students in an institution where we had many > mature students who had come into university level studies with non-standard > qualificat > ions, and were not well equipped for courses of this kind by their > previous experience. I'm glad to say that teaching the subject from their > own perspective as language students rather than that of the statistician, > and explaining the reasons for doing things in particular ways rather than > just presenting formulae, paid off in the end, so that most students were > able to appreciate the relevance of these courses and to turn in very > creditable projects showing an understanding of research design and > competence in the use of a range of basic statistical techniques. And I > still find that bright graduate students respond well to similar courses > which incorporate some of the rather more advanced techniques needed for > many real research projects in various areas of linguistics. But I may well > be out of date with what is now already happening in our fine institutions > of higher education! > > > > Chris Butler > > > > > > ------------------------------ > > Message: 25 > Date: Sat, 11 Sep 2010 17:54:58 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: funknet at mailman.rice.edu > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Hi all, > > I'd like to thank Ron for pointing out the incompleteness of what I wrote, > and for the reference to his article. I hope I didn't give the impression I > was trying to give a complete description of noun complementation in Modern > Hebrew in an email. > > What I *was* trying to say is that in some languages, unlike English, some > nominalizations (of utterance and cognition verbs, as Ron points out, but > also of some perception verbs too) can occur with a construction that is > explicitly and unmistakably marked as a complement clause. Moreover, the > nominalizations that take these explicit complement clauses are related to > verbs that can take the same type of complement clause. In such languages, > then, it's pretty clear that these instances involve complement clauses. It > doesn't mean that other noun complementation strategies don't exist for > other types of nouns. > > However, my main point was more general, albeit poorly expressed. It's that > we can turn to cross-linguistic comparison in order to try to reach > generalizations about how languages encode meaning. These generalizations > are useful, because they can be used to ask "why" questions. For example, > it's not really possible to ask *why* Hebrew has two distinct strategies > for > noun complementation, how *why* English has one. That's because it could > always be otherwise, and language change can alter the picture (and has!). > However, if we find that in languages that have two strategies, one is > limited to nominalizations of PCU verbs, then we have the beginnings of a > hierarchy that is amenable to functional explanation. > > Anyway, it seems that Thomas Payne's question has turned up a pretty > general > consensus that these constructions are complement clauses. > > Best wishes, > Eitan > > > > > > > On 10 September 2010 23:26, Ron Kuzar wrote: > > > The Modern Hebrew data supplied by Eitan are incomplete. > > Hebrew distinguishes between locution (say, hear, think, etc.) and > > situation (action, event, state, etc.). > > What Eitan describes is only true with regard to nouns (and clauses) > > expressing locution. 'Announcement' is indeed such a noun. > > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > > etc., whose denotatum is a situation, cannot be followed by ki, but only > > by Se-, e.g.: > > > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > > annoys me the-situation that-all went > > 'I am upset about the situation that all have gone' > > > > On the other hand, the relative Se- may be replaced by the more > > elegant and classical aSer, while the Se- of situation clauses may not. > > Sorry about the invented example. I am overseas now. > > All this has been described (with corpus data) in: > > > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut > Ivrit > > [Hebrew > > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > > > The article is somewhat outdated and contains some inaccuracies I would > > formulate differently today, but the basic distinction is valid in my > > opinion. > > Best, > > Ron Kuzar > > --------------- > > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > > > Hi all, > > > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > > comparison can help what seems to be a purely theoretical question > based > > on > > > a single language. The problem here is that English uses the same > element > > > to > > > mark regular relatives and these "appositional" relatives. But if at > > least > > > one language encodes them by different means, then there's at least a > > good > > > case for seeing them as distinct functions. It's basically the same > > > principle that's used to decide whether to put a meaning on a semantic > > map. > > > So here are two languages that I know that encode them differently. > > > > > > In Modern Hebrew, these clauses can be encoded as a dedicated > complement > > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > > > ha-hoda'a Se-kibalnu > > > the-announcment rel-we_got > > > "The announcement that we got." > > > > > > ha-hoda'a ki hitbatel ha-mifgaS > > > the-message CMP was_cancelled the-meeting > > > "The announcement that the meeting was cancelled." > > > > > > In Coptic, these clauses are marked by ce-, which marks complement > > clauses, > > > *inter alia*, but not relative clauses: > > > > > > ph-mewi ce- (complement clause) > > > 'the-thought that (we are angry)' > > > > > > ph-mewi ete- (relative clause) > > > 'the thought that (we used to think)' > > > > > > This seems to be a pretty clear indication that these are complement > > > clauses > > > rather than relatives. Even if one doesn't like the notion of nouns > > taking > > > complement clauses (and why not? nominalizations in some languages can > > take > > > accusative modifiers as well as genitives), it still probably isn't > > > incidental that the nominalizations are of verbs that take complement > > > clauses when finite. > > > > > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > > > looking at. > > > > > > Best, > > > Eitan > > > > > > > > > On 10 September 2010 19:21, Arie Verhagen > > > wrote: > > > > > > > And as another addition: the clauses that can only be introduced by > > > *that* > > > > (with no role to > > > > play in the subordinate clause) may be seen as (subtypes of) > complement > > > > clauses, > > > > expressing a proposition with the noun functioning as Complement > Taking > > > > Predicate (CTP), > > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > > (following analyses by > > > > Thompson, Diessel, Langacker, myself, and others), i.e. not > relatives. > > > Cf. > > > > constructions like > > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > > claim > > > > that X", "I put forward > > > > the claim that X", in which the relationship between the verb or noun > > and > > > > the that-clause is > > > > comparable to the one in "The claim that X". > > > > > > > > --Arie Verhagen > > > > > > > > ---------------- > > > > Message from Rong Chen > > > > 10 Sep 2010, 23:42 > > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > > > To add to Joanne's comments: > > > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > > (AC) from a relative clause (RC). > > > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > > pronouns. > > > > > > > > > > 2) The AC and the noun it modifies display an equative > > > relationship--one > > > > can say X > > > > > (denoted by the noun) is Y (presented by the appositive)--while an > RC > > > > often doesn't > > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but > a > > > > relative pronoun > > > > > is always part of the clause in an RC. > > > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > > > > > > > -- > > > Eitan Grossman > > > Martin Buber Society of Fellows > > > Hebrew University of Jerusalem > > > > > > > > > > > -- > > =============================================== > > Dr. Ron Kuzar > > Address: Department of English Language and Literature > > University of Haifa > > IL-31905 Haifa, Israel > > Office: +972-4-824-9826, Fax: +972-4-824-9711 > > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > > Home fax: 153-77-481-9676 (only from Israel) > > Email: kuzar at research.haifa.ac.il > > Homepage: http://research.haifa.ac.il/~kuzar > > =============================================== > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 26 > Date: Sat, 11 Sep 2010 10:19:05 -0500 > From: Kristine Hildebrandt > Subject: [FUNKNET] Job Advertisement > To: funknet at mailman.rice.edu > Message-ID: > > > > Content-Type: text/plain; charset=windows-1252 > > Dear Funknetters: > > > Please distribute this job advertisement to interested colleagues and > doctoral students completing their degrees. > > > *HIRING UNIT*: *DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE* > > ** > > *TITLE/RANK*: Assistant Professor of English-Linguistics > > > *DESCRIPTION OF DUTIES*: The Department of English Language and > Literature > invites applications for a tenure-track position in general linguistics, > with secondary specialization in applied linguistics. The candidate will > teach courses in the MA TESL program, along with undergraduate courses in > linguistics, composition (ESL and regular), and some general education > courses. Academic year: 3/3 load. > > *TERMS OF APPOINTMENT*: Academic, tenure-track beginning August 16, 2011, > 100% appointment. > > *SOURCE OF FUNDS*: State > > > *SALARY RANGE*: commensurate with training and experience > > > *QUALIFICATIONS REQUIRED: *A Ph.D. in Linguistics, Applied Linguistics, or > related field required. If Ph.D. is not completed by the beginning of the > contract period, appointment will be at the rank of Instructor until all > degree requirements are fulfilled. A record of ESL and/or TESL experience > is desirable. > > > *CLOSING DATE FOR APPLICATIONS*: Position open until filled; completed > applications postmarked by November 15, 2010 will have priority. Possible > interviews at LSA in January 2011. > * ** > **SEND COVER LETTER, VITA, UNOFFICIAL TRANSCRIPT, STATEMENT OF TEACHING > PHILOSOPHY AND RESEARCH AGENDA, AND THREE LETTERS OF RECOMMENDATION TO:** > * > Linguistics Search Committee > Southern Illinois University Edwardsville > Department of English Language & Literature > Campus Box 1431 > Edwardsville, IL 62026-1431 > NOTE: Electronic applications will not be accepted for this position. > > > SIUE is a state university-benefits under state sponsored plans may not be > available to holders of F1 or J1 visas. Applicants may be subject to a > background check prior to an offer of employment. SIUE is an affirmative > action and equal opportunity employer. The SIUE ANNUAL SECURITY REPORT is > available on-line at: http://admin.siu.edu/studentrightto/. The report > contains safety and security information and crime statistics for the past > three (3) calendar years. This report is published in compliance with > Federal law, entitled *the ?Jeanne Clery Disclosure of Campus Security > Policy and Campus Crime Statistics Act.? * You may also access this report > through the SIUE Home Page: http://www.siue.edu under *Ready References, > Quick Links or Publications/Reports*. For those without computer access, a > paper copy of the report may be obtained from the Office of the Vice > Chancellor for Administration, Rendleman Hall, Room 2228. > > > -- > *Kristine A. Hildebrandt* > *Assistant Professor, Department of English Language & Literature > Southern Illinois University Edwardsville* > *Box 1431 > Edwardsville, IL 62026 U.S.A. > 618-650-3380 (office)* > *khildeb at siue.edu > http://www.siue.edu/~khildeb* > > > ------------------------------ > > Message: 27 > Date: Sat, 11 Sep 2010 11:45:27 -0400 > From: Ted Gibson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Cc: Richard Hudson , Daniel Everett > , Evelina Fedorenko > Message-ID: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9 at MIT.EDU> > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dear Brian, Dick, Dan et al: > > Thanks for the discussion. Here are a few responses: > > 1. Brian: > > "But I understand Dick's worry about how far Gibson and Fedorenko are > trying to push this. Neither their email nor their paper sets clear > limits on what we should be testing and we certainly don't want to > waste time checking out go-goed-went. So, Gibson and Fedorenko owe > us those clarifications." > > The answer that we give to this question in Gibson & Fedorenko (in > press) is as follows (the final paragraph in the paper): > > "Finally, a question that is often put to us is whether it is > necessary to evaluate every empirical claim quantitatively. A major > problem with the fields of syntax and semantics is that many papers > include no quantitative evidence in support of their research > hypotheses. Because conducting experiments is now so easy to do with > the advent of Amazon.com?s Mechanical Turk, we recommend gathering > quantitative evidence for all empirical claims. However, it would > clearly be a vast improvement to the field for all research papers to > include at least some quantitative evidence evaluating their research > hypotheses." > > Another possible answer to this question is: the more important some > observation is, the better your evidence should be. If the > observation is a key reason for some important theoretical claim, then > there should be solid quantitative data supporting that observation. > > In practice, once a linguist starts gathering quantitative data, s/he > will realize (a) how easy it is to do; and (b) how beneficial the > methods are, with the consequence that these researchers will probably > do most or all of their work quantitatively in the future. > > > 2. Dick (by the way, thank you for the kind responses, and your > positive tone): > > "Your [the psycholinguists'] goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality." > > in contrast to "my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers." > > Many cognitive psychologists / cognitive scientists (all the ones I > know at MIT for example) are interested in both cognitive > generalizations across people and ways in which people differ > cognitively. In fact, some methods (e.g., the individual differences > approach where co-variation of various behaviors / characteristics is > examined across individuals) have been specifically developed to study > differences among individuals. Both kinds of data are important for > understanding human cognition, including language. This applies to > language research directly: generalizations across people are > important, but so are individual differences. In either case, > quantitative data are necessary to evaluate research questions and > test hypotheses. > > On a related note, it is a mistake to characterize researchers with a > background in "psychology" or cognitive science as being interested in > "processing", and researchers with a background in "linguistics" as > being interested in "knowledge" or "representation / structure". Both > psychologists and linguists should be interested in *both* > representation and processing (and learning, for that matter). We > wrote a little about this confusion in Gibson & Fedorenko (in press), > which we include at the end of the message. > > This leads to something that Dan said: > > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > Both linguistics and psychology are big fields. We assume Dan is > referring to cognitive psychology / cognitive science here. (Of > course, there are sub-fields of psychology - e.g., personality > psychology or abnormal psychology - which are somewhat distinct from > linguistics, but those sub-fields are also distinct from cognitive > psychology.) It is true that historically linguistics is not treated > as a subfield of cognitive psychology / cognitive science. However, > key research questions in linguistics (i.e., the form of the knowledge > structures and algorithms underlying human language) are indeed a > subset of those investigated by cognitive psychologists / cognitive > scientists. We think that the biggest factor separating linguistics > from psychology is the methods used to explore the research questions, > rather than the research questions themselves. Consequently, we > would like to continue to see tighter connections among the fields of > psychology / cognitive science, linguistics, as well as other fields > like anthropology and computer science. > > Thanks to all for the interesting discussions. > > Ted & Ev > > We have encountered a claim that the reason for different kinds of > methods being used across the different fields of language study > (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the > research questions are different across these fields, and some methods > may be better suited to ask some questions than others. Although the > latter is likely true, the premise ? that the research questions are > different across the fields ? is false. The typical claim is that > researchers in the field of linguistics are investigating linguistic > representations, and researchers in the fields of psycho-/neuro- > linguistics are investigating the computations that take place as > language is understood or produced. However, many researchers in the > fields of psycho-/neuro-linguistics are also interested in the nature > of the linguistic representations (at all levels; e.g., phonological > representations, lexical representations, syntactic representations, > etc.) [1]. By the same token, many researchers in the field of > linguistics are interested in the computations that take place in the > course of online comprehension or production. However, inferences ? > drawn from any dependent measure ? about either the linguistic > representations or computations are always indirect. And these > inferences are no more indirect in reading times or event-related > potentials, etc., than in acceptability judgments: across all > dependent measures we take some observable (e.g., a participant?s > rating on an acceptability judgment task or the time it took a > participant to read a sentence) and we try to infer something about > the underlying cognitive representations / processes. More generally, > methods in cognitive science are often used to jointly learn about > representations and computations, because inferences about > representations can inform questions about the computations, and vice > versa. For example, certain data structures can make a computation > more or less difficult to perform, and certain representations may > require assumptions about the algorithms being used. > > In our opinion then, the distinction between the fields of linguistics > and psycho-/neuro-linguistics is purely along the lines of the kind of > data that are used as evidence for or against theoretical hypotheses: > typically non-quantitative data in linguistics vs. typically > quantitative data in psycho-/neuro-linguistics. Given the superficial > nature of this distinction, we think that there should be one field of > language study where a wide range of dependent measures is used to > investigate linguistic representations and computations. > > > [1] In fact, some methods in cognitive science and cognitive > neuroscience were specifically developed to get at representational > questions (e.g., lexical / syntactic priming methods, neural > adaptation or multi-voxel pattern analyses in functional MRI). > > > On Sep 10, 2010, at 9:05 PM, Daniel Everett wrote: > > > I think that Brian and Dick make excellent points. There are very > > good grammars written that could be improved by psycholinguistic > > experimentation and more quantitative approaches. But large sections > > of those grammars aren't going to change one bit (go-went) with > > quantitative tests and such tests would be completely > > counterproductive given the shortness of life and the vastness of > > the field linguist's tasks. > > > > Part of the problem is that linguistics is not simply a > > subdiscipline of psychology. Linguistics has its own objectives and > > those only occasionally overlap with psychology. The same for methods. > > > > On another note, I don't buy the 'in my head' 'out of my head' > > distinction either (that Matt seems to be urging upon us). We study > > different things and have different reasons for being satisfied with > > the results we achieve. > > > > I believe that we linguists are often complacent and fail to apply > > better methods. But of course that applies to all disciplines. > > > > In the meantime, checking corpora, collecting data as a result of > > careful interviews with native speakers, and the other aspects of > > the field linguist's task are vital parts of the linguist's task and > > much of this won't be improved by quantitative methods as we > > currently understand them. Maybe sometime. > > > > Dan > > > > P.S. In my original reference to Ted and Ev's paper, I said that > > they showed the danger of using intuitions. What I meant to say of > > using intuitions as standardly used by linguists. They convinced me > > that there is a lot to learn from quantitative methods. > > > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > >> Dear Ted and Ev, > >> Yes, I understand your view, but I think it's a psycholinguist's > >> view. Your goal is to find general processes and principles that > >> apply uniformly across individuals, so you have to use methods to > >> check for generality. And (as you know) I admire the way you pursue > >> that goal. But my goal, as a linguist, is different. I want to > >> explore the structure of a language so that I can understand how > >> all the bits fit together. Like you, I'm aiming to model cognition, > >> but my focus is on items and structures, and I start from the > >> assumption that these can and do vary across speakers. > >> > >> However, having said all that I do agree with you that linguists > >> should all get used to collecting and using quantitative data; and > >> with the help of Brian MacWhinney's typology we'd know what methods > >> to use when. And I do agree with your points about bid/bidded: in > >> cases like that, quantitative data would be at least a very good > >> starting point for a proper investigation. > >> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 19:30, Ted Gibson wrote: > >>> Dear Dick: > >>> > >>> Perhaps we are talking at cross purposes. I don't understand what > >>> is confusing about what Ev Fedorenko and I are claiming. All we > >>> are saying is that if you have some testable claim involving a > >>> general hypothesis about a language, then you need to get > >>> quantitative data from unbiased sources to evaluate that claim. If > >>> you are interested in English past tense morphology, then > >>> depending on the claims that you might want to investigate, there > >>> are lots of ways to get relevant quantitative evidence. Corpus > >>> data will probably be useful. For very low frequency words, you > >>> can run experiments to test behavior with respect to such words. > >>> > >>> Your example of the past tense of "bid" is a fine such example. > >>> You can run an experiment like the one you suggested to find out > >>> what people think the past tense is. If you then found that 20/50 > >>> people responded "bidded" and 30/50 respond "bid", that is a lot > >>> of useful information. As you suggest in your discussion, this > >>> result wouldn't answer the question of how past tense is stored in > >>> each individual. This result would be ambiguous among several > >>> possible explanations. One possibility is that the probability > >>> distribution that is being discovered reflects different dialects, > >>> such that 2/5 of the population has one past tense, and 3/5 has > >>> another. Another possibility is that each person has a similar > >>> probability distribution in their heads, such that 2/5 of the time > >>> I respond one way, and 3/5 of the time I respond another. Further > >>> experiments would be necessary to answer between these and other > >>> possible theories (e.g., with repeated trials from the same > >>> person, carefully planned so that the participants don't notice > >>> that they are being asked multiple times). Without the > >>> quantitative evidence in the first place, there is no way to > >>> answer these kinds of questions. > >>> > >>> Regarding the past tense of "go", this would be useful as a > >>> baseline in an experiment involving the less frequent ones. So, > >>> yes, it would useful to gather quantitative evidence in such a > >>> case also, as baselines with respect to the more interesting cases > >>> for theories. > >>> > >>> The bottom line: if you have a generalization about a language > >>> that you wish to evaluate (such that you hypothesize that it is > >>> true across the speakers of the language), then you need > >>> quantitative evidence from multiple individuals, using an unbiased > >>> data collection method, to evaluate such a claim. The point about > >>> Mechanical Turk is that it is really *easy* to do this now, at > >>> least for languages like English. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson & Ev Fedorenko > >>> > >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >>> > >>>> Dear Ted, > >>>> Thanks for the very interesting comment, but are you REALLY > >>>> saying that I shouldn't claim, for example, that the past tense > >>>> of GO is "went" without first cross-checking with 50 native > >>>> speakers? > >>>> > >>>> Isn't there a danger of missing the point that we all, as native > >>>> speakers, spend our whole lives scanning other people's > >>>> linguistic behaviour (language 'out there', E-language) and > >>>> trying to explain it to ourselves in terms of a language system > >>>> (language 'in here', I-language)? So every judgement we make is > >>>> based on thousands or millions of observed exemplars, and > >>>> reflects a unique experience of E-language filtered through a > >>>> unique I-language. > >>>> > >>>> Given that view of language development, I don't see how > >>>> quantitative data will help. Let's take a real uncertainty, such > >>>> as the past tense of BID. If I want to say I did it, do I say "I > >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>>> and 30 "bid". So what? Does that mean that the correct answer is > >>>> "bidded"? Surely not. How is it better than my judgement? I agree > >>>> you could record my speech and find how often I use each > >>>> alternative; but the reason I don't know is precisely because > >>>> it's a rare word, so in a sense quantitative data are irrelevant > >>>> even there. What would solve the problem of subjectivity, of > >>>> course, would be a machine for probing the bit of my mind (or > >>>> even brain) that holds BID and its details; but I suspect that > >>>> even that wouldn't move us much further forward than my original > >>>> "don't know". (Incidentally I write as a fan of quantitative > >>>> sociolinguistics, so I do accept that quantitative data are > >>>> relevant to linguistic analysis in some areas, where the I- > >>>> language phenomenon is frequent enough to produce usable data.) > >>>> > >>>> It seems to me that this discussion raises the really fundamental > >>>> question of what kind of thing we think language is: social or > >>>> individual. The problem isn't unique to linguistics of course; > >>>> it's the same throughout the social sciences. But what's special > >>>> about linguistics is that we deal in very fine details of culture > >>>> (e.g. details of how a particular word is used or pronounced) so > >>>> the differences between individuals really matter. I don't see > >>>> that we're ever going to have anything better than judgements to > >>>> go on, so what we need is a way to ensure that judgements are > >>>> accurate reports of individual I-language. A rotten situation for > >>>> a science, but I don't see how it can get better. > >>>> > >>>> Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> > >>>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>>> Dear Dan, Dick: > >>>>> > >>>>> I would like to clarify some points that Dan Everett makes, in > >>>>> response to Dick Hudson. > >>>>> > >>>>> Ev Fedorenko and I have written a couple of papers recently > >>>>> (Gibson & > >>>>> Fedorenko, 2010, in press, see references and links below) on > >>>>> what we > >>>>> think are weak methodological standards in syntax and semantics > >>>>> research over the past many years. The issue that we address is > >>>>> the > >>>>> prevalent method in syntax and semantics research, which involves > >>>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>>> pair, typically by just the author of the paper, sometimes with > >>>>> feedback from colleagues. As we address in our papers, this > >>>>> methodology does not allow proper testing of scientific hypotheses > >>>>> because of (a) the small number of experimental participants > >>>>> (typically one); (b) the small number of experimental stimuli > >>>>> (typically one); (c) cognitive biases on the part of the > >>>>> researcher > >>>>> and participants; and (d) the effect of the preceding context > >>>>> (e.g., > >>>>> other constructions the researcher may have been recently > >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>>> points, but with not as strong a conclusion as ours). > >>>>> > >>>>> Three issues need to be separated here: (1) the use of intuitive > >>>>> judgments as a dependent measure in a language experiment; (2) > >>>>> potential cognitive biases on the part of experimental subjects > >>>>> and > >>>>> experimenters in language experiments; and (3) the need for > >>>>> obtaining > >>>>> quantitative evidence, whatever the dependent measure might be. > >>>>> The > >>>>> paper that Ev and I wrote addresses the last two issues, but > >>>>> does not > >>>>> go into depth on the first issue (the use of intuitions as a > >>>>> dependent > >>>>> measure in language experiments). Regarding this issue, we don't > >>>>> think > >>>>> that there is anything wrong with gathering intuitive judgments > >>>>> as a > >>>>> dependent measure, as long as the task is clear to the > >>>>> experimental > >>>>> participants. > >>>>> > >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to > >>>>> some > >>>>> arguments that have been given in support of continuing to use the > >>>>> traditional non-quantitative method in syntax / semantics > >>>>> research. > >>>>> One recent defense of the traditional method comes from Phillips > >>>>> (2008), who argues that no harm has come from the non-quantitative > >>>>> approach in syntax research thus far. Phillips argues that there > >>>>> are > >>>>> no cases in the literature where an incorrect intuitive judgment > >>>>> has > >>>>> become the basis for a widely accepted generalization or an > >>>>> important > >>>>> theoretical claim. He therefore concludes that there is no > >>>>> reason to > >>>>> adopt more rigorous data collection standards. We challenge > >>>>> Philips? > >>>>> conclusion by presenting three cases from the literature where a > >>>>> faulty intuition has led to incorrect generalizations and mistaken > >>>>> theorizing, plausibly due to cognitive biases on the part of the > >>>>> researchers. > >>>>> > >>>>> A second argument that is sometimes presented for the continued > >>>>> use of > >>>>> the traditional non-quantitative method is that it would be too > >>>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff > >>>>> (2010) > >>>>> make this argument explicitly in their response to Gibson & > >>>>> Fedorenko > >>>>> (2010): ?It would cripple linguistic investigation if it were > >>>>> required > >>>>> that all judgments of ambiguity and grammaticality be subject to > >>>>> statistically rigorous experiments on naive subjects, especially > >>>>> when > >>>>> investigating languages whose speakers are hard to > >>>>> access? (Culicover > >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>>> earlier in the discussion here.) Whereas we agree that in > >>>>> circumstances where gathering data is difficult, some evidence is > >>>>> better than no evidence, we do not agree that research would be > >>>>> slowed > >>>>> with respect to languages where experimental participants are > >>>>> easy to > >>>>> access, such as English. In contrast, we think that the opposite > >>>>> is > >>>>> true: the field?s progress is probably slowed by not doing > >>>>> quantitative research. > >>>>> Suppose that a typical syntax / semantics paper that lacks > >>>>> quantitative evidence includes judgments for 50 or more > >>>>> sentences / > >>>>> meaning pairs, corresponding to 50 or more empirical claims. > >>>>> Even if > >>>>> most of the judgments from such a paper are correct or are on the > >>>>> right track, the problem is in knowing which judgments are > >>>>> correct. > >>>>> For example, suppose that 90% of the judgments from an arbitrary > >>>>> paper > >>>>> are correct (which is probably a high estimate). (Colin Phillips > >>>>> and > >>>>> some of his former students / postdocs have commented to us > >>>>> that, in > >>>>> their experience, quantitative acceptability judgment studies > >>>>> almost > >>>>> always validate the claim(s) in the literature. This is not our > >>>>> experience, however. Most experiments that we have run which > >>>>> attempt > >>>>> to test some syntactic / semantic hypothesis in the literature > >>>>> end up > >>>>> providing us with a pattern of data that had not been known > >>>>> before the > >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This > >>>>> means > >>>>> that in a paper with 50 empirical claims 45/50 are correct. But > >>>>> which > >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s > >>>>> over > >>>>> two million different theories. By quantitatively evaluating the > >>>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>>> progress, it is better to have theoretical claims supported by > >>>>> solid > >>>>> quantitative evidence, so that even if the interpretation of the > >>>>> data > >>>>> changes over time as new evidence becomes available ? as is > >>>>> often the > >>>>> case in any field of science ? the empirical pattern can be used > >>>>> as a > >>>>> basis for further theorizing. > >>>>> > >>>>> Furthermore, it is no longer expensive to run behavioral > >>>>> experiments, > >>>>> at least in English and other widely spoken languages. There now > >>>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? > >>>>> which > >>>>> can be used for collecting behavioral data over the internet > >>>>> quickly > >>>>> and inexpensively. The cost of using an interface like this is > >>>>> minimal, and the time that it takes for the results to be > >>>>> returned is > >>>>> short. For example, currently on Mechanical Turk, a survey of > >>>>> approximately 50 items will be answered by 50 or more participants > >>>>> within a couple of hours, at a cost of approximately $1 per > >>>>> participant. Thus a survey can be completed within a day, at a > >>>>> cost of > >>>>> less than $50. (The hard work of designing the experiment, and > >>>>> constructing controlled materials remains of course.) > >>>>> > >>>>> Sorry to be so verbose. But I think that these methodological > >>>>> points > >>>>> are very important. > >>>>> > >>>>> Best wishes, > >>>>> > >>>>> Ted Gibson > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>>> methods in syntax and semantics research. Language and Cognitive > >>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>>> & Fedorenko InPress LCP.pdf > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >>>>> Fedorenko > >>>>> 2010 TICS.pdf > >>>>> > >>>>> > >>>>> > >>>>> > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about methodology. I believe > >>>>>> that > >>>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>>> them - to a degree. But while it might not have been feasible for > >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>>> Grammar to conduct experiments on every point of the grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, corpora, and standard psycholinguistic > >>>>>> experimentation > >>>>>> (indeed, Standard Social Science Methodology) is vital for > >>>>>> taking the > >>>>>> field forward and for providing the best support for different > >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very > >>>>>> useful new > >>>>>> paper on this, showing serious shortcomings with intuitions as > >>>>>> the > >>>>>> sole source of evidence, in their paper: "The need for > >>>>>> quantitative > >>>>>> methods in syntax and semantics research". > >>>>>> > >>>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>>> convincingly on this. > >>>>>> > >>>>>> It is one reason that a team from Stanford, MIT (Brain and > >>>>>> Cognitive > >>>>>> Science), and researchers from Brazil are beginning a third > >>>>>> round of > >>>>>> experimental work among the Pirahas, since my own work on the > >>>>>> syntax > >>>>>> was, like almost every other field researcher's, based on native > >>>>>> speaker intuitions and corpora. > >>>>>> > >>>>>> The discussion of methodologies reminds me of the initial > >>>>>> reactions > >>>>>> to Greenberg's work on classifying the languages of the > >>>>>> Americas. His > >>>>>> methods were strongly (and justifiably) criticized. However, I > >>>>>> always > >>>>>> thought that his methods were a great way of generating > >>>>>> hypotheses, > >>>>>> so long as they were ultimately put to the test of standard > >>>>>> historical linguistics methods. And the same seems true for use > >>>>>> of > >>>>>> native-speaker intuitions. > >>>>>> > >>>>>> -- Dan > >>>>> > >>>>> > >>>>> > >>>>>>> We linguists can add a further layer of explanation to the > >>>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>>> every detailed analysis we make, our whole discipline will > >>>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>>> judgements are what put us linguists ahead of the rest in > >>>>>>> handling > >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the > >>>>>>> English > >>>>>>> Language (or the OED) without using native speaker judgements. > >>>>>>> > >>>>>>> Best wishes, Dick Hudson > >>>>> > >>>>> > >>>>> > >>> > >>> > >>> > >> > > > > > > ------------------------------ > > Message: 28 > Date: Sat, 11 Sep 2010 12:53:23 -0400 > From: Daniel Everett > Subject: Re: [FUNKNET] analysis: unhappiness > To: Ted Gibson > Cc: Richard Hudson , Daniel Everett > , Evelina Fedorenko , > funknet > > Message-ID: <0021EEE8-560C-4E0A-8A3B-9595384807D6 at daneverett.org> > Content-Type: text/plain; charset=us-ascii > > Ted, > > Let me clarify this: > > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > > > Both linguistics and psychology are big fields. We assume Dan is > referring to cognitive psychology / cognitive science here. (Of course, > there are sub-fields of psychology - e.g., personality psychology or > abnormal psychology - which are somewhat distinct from linguistics, but > those sub-fields are also distinct from cognitive psychology.) It is true > that historically linguistics is not treated as a subfield of cognitive > psychology / cognitive science. However, key research questions in > linguistics (i.e., the form of the knowledge structures and algorithms > underlying human language) are indeed a subset of those investigated by > cognitive psychologists / cognitive scientists. We think that the biggest > factor separating linguistics from psychology is the methods used to explore > the research questions, rather than the research questions themselves. > Consequently, we would like to continue to see tighter connections among the > fields of psychology / cognitive science, lingu > istics, as well as other fields like anthropology and computer science. > > > Correct, I meant cognitive psychology, not, say, psychoanalysis. There are > definitely overlapping concerns. But my main concern about language is less > about representations and more about the cultural and sociological values > that lead to sentences and expressions in the corpus, rather than the mind. > I used to think that my main interest was representations in the mind. But I > find the psychology less interesting than the anthropology these days. > > But this is not an excuse to avoid quantitative methods. I believe that you > and Ev, and others, have made a convincing case for quantitative methods. > Quantitative methods in field research on syntax and semantics is vital. > > -- dan > > > > > End of FUNKNET Digest, Vol 84, Issue 10 > *************************************** > From alifarghaly at yahoo.com Sun Sep 12 06:20:16 2010 From: alifarghaly at yahoo.com (Ali Farghaly) Date: Sat, 11 Sep 2010 23:20:16 -0700 Subject: Ali Farghaly wants to stay in touch on LinkedIn Message-ID: LinkedIn ------------ I'd like to add you to my professional network on LinkedIn. - Ali Farghaly Ali Farghaly Founder at Language Applications and Services San Francisco Bay Area Confirm that you know Ali Farghaly https://www.linkedin.com/e/-tffddx-gdzimymn-3l/isd/1659499320/pcnZ4P7w/ -- (c) 2010, LinkedIn Corporation From abergs at uos.de Sun Sep 12 07:40:21 2010 From: abergs at uos.de (Alexander Bergs) Date: Sun, 12 Sep 2010 09:40:21 +0200 Subject: ICLCE 4 - 2nd call for papers Message-ID: 4th International Conference on the Linguistics of Contemporary English ICLCE-4 Date: 19-Jul-2011 - 23-Jul-2011 Location: Osnabrueck, Germany Contact Person: Alexander Bergs Meeting Email: iclce4 at uos.de Web Site: http://www.blogs.uos.de/iclce4 2nd Call for Papers The attention devoted to the linguistics of the English language has resulted in a broad body of work in diverse research traditions. The aim of the ICLCE conference is to encourage the cross-fertilisation of ideas between different frameworks and research traditions, all of which may address any aspect of the linguistics of English. Previous ICLCE conferences were held in Edinburgh (2005), Toulouse (2007) and London (2009) along the same lines. We aim for the Osnabrueck conference to build on the success of those events. Plenary Speakers Scott F. Kiesling (University of Pittsburgh) Daniel Schreier (University of Zurich) Peter Stockwell (University of Nottingham) Graeme Trousdale (University of Edinburgh) Jessica de Villiers (University of British Columbia) Rachel Walker (University of Southern California) Gert Webelhuth (Goethe Universität Frankfurt am Main) We invite papers on any topic concerning the linguistics of contemporary English. Workshop proposals are also welcome. We are using the EasyABS system delivered by LinguistList. If there are any problems or if you have any questions, please feel free to contact us at iclce4uos.de. Abstracts should be no longer than 350 words, preferably format A4 with 2.5 cm margins, single-spaced, Times New Roman 12 pt, and with normal character spacing. All examples and references in the abstract should be included, but it is enough, when referring to previous work, to cite 'Author (Date)' in the body of the abstract - you do not need to include the full reference. Please only use common phonetic fonts such as SIL. +++++++++++++ Univ.-Prof. Dr. Alexander Bergs, M.A. Chair of English Language and Linguistics Institut für Anglistik und Amerikanistik (IfAA) Fachbereich 7 -Universitaet Osnabrueck Neuer Graben 40 D-49069 Osnabrueck Germany Tel: +49 541 969 4255 Tel: +49 541 969 6042 (secy) Fax: +49 541 969 4738 http:/www.ifaa.uni-osnabrueck.de/bergs +++++++++++++ Univ.-Prof. Dr. Alexander Bergs, M.A. Chair of English Language and Linguistics Institut für Anglistik und Amerikanistik (IfAA) Fachbereich 7 -Universitaet Osnabrueck Neuer Graben 40 D-49069 Osnabrueck Germany Tel: +49 541 969 4255 Tel: +49 541 969 6042 (secy) Fax: +49 541 969 4738 http:/www.ifaa.uni-osnabrueck.de/bergs From cbutler at ntlworld.com Sun Sep 12 10:03:31 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Sun, 12 Sep 2010 11:03:31 +0100 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya, I think two different things are getting a bit mixed up here. I don't for a moment dispute that expressions are often susceptible to multiple interpretations, that these interpretations are guided by all kinds of contextual information, or that different people, or even the same person at different times, may end up selecting differently from the various interpretations. Your example 'Open happiness' in another contribution to this thread illustrates these points very well. My point, though, is that each of these different interpretations, as well as the selection of one (or more) as more likely in a particular context, is achieved through mechanisms in the interpreter's brain which evolved in the course of the phylogenetic development of language in the human species, and developed ontogenetically in that particular interpreter's brain. It is surely likely that those mechanisms are highly similar in different human beings, even though there may be differences in the detailed wiring in different brains. What I'm saying is that in order to answer the question 'How do we communicate using language?' or, if you prefer, 'How does the language user work?', we need to investigate what those mechanisms are, and this is what psycholinguists can help us with. In particular, as linguists, we are interested (well, some of us are, though clearly not all) in whether the constructs we posit in our theories of language have psychological validity in the sense that they correspond to ways in which aspects of language are represented in the brain. [As an aside, I do realise that there are linguists who strenuously resist what they see as a misguided emphasis on mental representation in the work of cognitive scientists, but it seems clear that language must be represented in some way in the brain in order that we can engage in the sociosemiotic acts of meaning making which are the primary focus for many of these critics.] Taking your 'Open happiness' examples again, I think we need answers to questions such as: What kind of representation does the human language processing system have for 'open' and for 'happiness'? Are the phonological, syntactic, semantic and pragmatic (for those who distinguish the semantic from the pragmatic) properties of these items (and we might want to add 'for this particular interpreter', though there must be similarities across interpreters for communication to be possible) represented in the same or different ways, in the same or different locations (or sets of distributed locations)? Indeed, are we right in thinking that these familiar levels of linguistic description must be differentiated, as such, in the human language processing system? Does the representation for 'open' distinguish between what we call verbal and adjectival uses of this item, and if so how? Or are syntactic analyses computed on the fly, using semantic and contextual clues, rather than the neural equivalent of pigeon holes corresponding to verbs and adjectives? Is 'happiness' represented as 'happy + ness', or in its entirety, or both? All these questions, and many many more, are relevant to the construction of a model of language which reflects how language users communicate (as, of course, are a whole set of other questions about the sociocultural aspects of communication). I am not a psycholinguist, but my all too superficial reading in the area suggests that psycholinguists have gone some of the way towards answering some of the questions we might want to ask, but that there is still a long way to go. Chris -------------------------------------------------- From: "A. Katz" Sent: Saturday, September 11, 2010 3:18 PM To: "Chris Butler" Cc: "FUNKNET" Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > The problem is that once we achieve the psycholinguistic goal and see what > is happening in each person's head, and we see that conflicting analyses > are the norm, rather than the exception, among normal language speakers, > then we will realize that the way language works to transmit information > is despite individual differences and not because of uniform processing > strategies. > > Even when all are agreed as to the meaning of an utterance, they do not > process it the same way. Which means that processing is seocndary to > information transmission. > > --Aya > > > > > On Sat, 11 Sep 2010, Chris Butler wrote: > >> Dick's comment that "this discussion raises the really fundamental >> question of what kind of thing we think language is: social or >> individual" is, it seems to me, an important one, particularly for those >> of us who are committed functionalists. My own view is that a truly >> functional model of language would be one which aims to account for how >> human beings communicate using language, or in other words tries to >> answer the question which was posed by Simon Dik a long time ago now, but >> which was not tackled head-on in his own work: "How does the natural >> language user work?' In trying to answer this question we need to accept >> that language is BOTH social AND individual, and we need to explore both >> aspects to get as complete a picture as possible of how we communicate >> using language. We need to know BOTH how people create and respond to >> meanings and express those meanings in forms during social interaction >> AND the mechanisms which operate in the brains of individuals in order to >> make such interaction possible. Both are important parts of the answer to >> the question 'How do we communicate using language?', though this >> particular thread of the Funknet discussion has concentrated more on the >> second aspect, and so will I. >> >> This doesn't mean that all the work linguists have done on "exploring the >> structure of a language so that I can understand how all the bits fit >> together" and "exploring the connections between items", as Dick puts it, >> is useless - far from it. After all, the hypotheses that psycholinguists >> test are based on ideas about what languages are like. But it does mean, >> in my view, that ultimately we need to get evidence that the constructs >> and analyses we propose are ones that are at least consistent with what >> we know of the processes which go on when we use language. So I am with >> Matthew when he says that for him, "the only sense in which an analysis >> can be "the correct analysis" is in terms of what is inside of people's >> heads". Of course, this doesn't imply that linguists should just give up >> their jobs until such time as we know everything there is to know about >> language processing. But it does mean that we need to collaborate with >> psycholinguists, psychologists and neurologists, as has also been pointed >> out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan >> Nuyts. [We also need to collaborate much more with sociolinguists and >> sociologists, so that we can get a better handle on the sociocultural >> aspects of how we communicate.] And it also means that psycholinguists, >> for their part, need whenever possible to follow up tightly controlled >> lab experiments with studies under more naturalistic conditions, to avoid >> the criticism that what happens in artifical lab situations may not >> happen in natural communicative conditions. >> >> I also agree with Dick when he says that "the differences between >> individuals really matter", and with Lise when she points out that "we >> must also be careful not to idealize "what's in people's heads" as if it >> were a single coherent construct that we are trying to discover". >> However, there are surely processing mechanisms which are common to all >> language users by virtue of the evolution of the language faculty and >> which constitute the "general processes" which Dick says psycholinguists >> are interested in. >> >> On the issue of quantitative methodology, I'm sympathetic in general to >> Ted and Ev's views, though it does seem sensible to prioritise cases in >> terms of a hierarchy such as Brian suggests. One thing this means is that >> we should be giving our university students of linguistics (and some of >> our linguistics lecturers!) courses in quantitative aspects of >> linguistics that introduce them to the use of at least some of the basic >> statistical methods in language study, and I'm sure this is indeed going >> on in some enlightened places. To those who suspect this can't be done >> with maths-shy students who don't initially see the need for it, I offer >> my own experience, over quite a long period, of teaching such courses to >> people with little or no prior experience in quantitative techniques. For >> some years in the 1990s, I taught such courses to all linguistics >> students in an institution where we had many mature students who had come >> into university level studies with non-standard qualifications, and were >> not well equipped for courses of this kind by their previous experience. >> I'm glad to say that teaching the subject from their own perspective as >> language students rather than that of the statistician, and explaining >> the reasons for doing things in particular ways rather than just >> presenting formulae, paid off in the end, so that most students were able >> to appreciate the relevance of these courses and to turn in very >> creditable projects showing an understanding of research design and >> competence in the use of a range of basic statistical techniques. And I >> still find that bright graduate students respond well to similar courses >> which incorporate some of the rather more advanced techniques needed for >> many real research projects in various areas of linguistics. But I may >> well be out of date with what is now already happening in our fine >> institutions of higher education! >> >> Chris Butler >> >> From amnfn at well.com Sun Sep 12 13:13:05 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 06:13:05 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <0EEE18DF122B4EDFBB744F0A83BB52AE@OwnerPC> Message-ID: Chris, Thanks for addressing this question. I understand that many, many linguists are quite properly and approriately interested in this ultimate question: "How does the language user work?" (I am also intetested in this question some of the time.) My point is that "how does language work?" is also a valid question, and a central one to the field of linguistics. These two questions are not at all the same. Let me be very explicit: My aim is to separate out the "human" from the "language". There are many good reasons to do so. For anyone working in computerized language processing or in non-human language studies, this is a significant point. It does not matter if a computerized language processing system even remotely simulates what humans do with language in their brains. It does matter whether it comes up with comparable or indistinguishable results. It does not matter whether a parrot, a dolphin or a chimpanzee is doing the same things inside the same modules in his brain as a human does. It does matter if the results are functionally equivalent. We need to make that distinction, between humans and their language, or we will be caught inside a circular definition with no way to break out or to prove anything, not about others and not about ourselves! --Aya http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies On Sun, 12 Sep 2010, Chris Butler wrote: > Aya, I think two different things are getting a bit mixed up here. > > I don't for a moment dispute that expressions are often susceptible to > multiple interpretations, that these interpretations are guided by all kinds > of contextual information, or that different people, or even the same person > at different times, may end up selecting differently from the various > interpretations. Your example 'Open happiness' in another contribution to > this thread illustrates these points very well. > > My point, though, is that each of these different interpretations, as well as > the selection of one (or more) as more likely in a particular context, is > achieved through mechanisms in the interpreter's brain which evolved in the > course of the phylogenetic development of language in the human species, and > developed ontogenetically in that particular interpreter's brain. It is > surely likely that those mechanisms are highly similar in different human > beings, even though there may be differences in the detailed wiring in > different brains. What I'm saying is that in order to answer the question > 'How do we communicate using language?' or, if you prefer, 'How does the > language user work?', we need to investigate what those mechanisms are, and > this is what psycholinguists can help us with. > > In particular, as linguists, we are interested (well, some of us are, though > clearly not all) in whether the constructs we posit in our theories of > language have psychological validity in the sense that they correspond to > ways in which aspects of language are represented in the brain. [As an aside, > I do realise that there are linguists who strenuously resist what they see as > a misguided emphasis on mental representation in the work of cognitive > scientists, but it seems clear that language must be represented in some way > in the brain in order that we can engage in the sociosemiotic acts of meaning > making which are the primary focus for many of these critics.] Taking your > 'Open happiness' examples again, I think we need answers to questions such > as: What kind of representation does the human language processing system > have for 'open' and for 'happiness'? Are the phonological, syntactic, > semantic and pragmatic (for those who distinguish the semantic from the > pragmatic) properties of these items (and we might want to add 'for this > particular interpreter', though there must be similarities across > interpreters for communication to be possible) represented in the same or > different ways, in the same or different locations (or sets of distributed > locations)? Indeed, are we right in thinking that these familiar levels of > linguistic description must be differentiated, as such, in the human language > processing system? Does the representation for 'open' distinguish between > what we call verbal and adjectival uses of this item, and if so how? Or are > syntactic analyses computed on the fly, using semantic and contextual clues, > rather than the neural equivalent of pigeon holes corresponding to verbs and > adjectives? Is 'happiness' represented as 'happy + ness', or in its entirety, > or both? All these questions, and many many more, are relevant to the > construction of a model of language which reflects how language users > communicate (as, of course, are a whole set of other questions about the > sociocultural aspects of communication). I am not a psycholinguist, but my > all too superficial reading in the area suggests that psycholinguists have > gone some of the way towards answering some of the questions we might want to > ask, but that there is still a long way to go. > > Chris > -------------------------------------------------- > From: "A. Katz" > Sent: Saturday, September 11, 2010 3:18 PM > To: "Chris Butler" > Cc: "FUNKNET" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > >> The problem is that once we achieve the psycholinguistic goal and see what >> is happening in each person's head, and we see that conflicting analyses >> are the norm, rather than the exception, among normal language speakers, >> then we will realize that the way language works to transmit information >> is despite individual differences and not because of uniform processing >> strategies. >> >> Even when all are agreed as to the meaning of an utterance, they do not >> process it the same way. Which means that processing is seocndary to >> information transmission. >> >> --Aya >> >> >> >> >> On Sat, 11 Sep 2010, Chris Butler wrote: >> >>> Dick's comment that "this discussion raises the really fundamental >>> question of what kind of thing we think language is: social or individual" >>> is, it seems to me, an important one, particularly for those of us who are >>> committed functionalists. My own view is that a truly functional model of >>> language would be one which aims to account for how human beings >>> communicate using language, or in other words tries to answer the question >>> which was posed by Simon Dik a long time ago now, but which was not >>> tackled head-on in his own work: "How does the natural language user >>> work?' In trying to answer this question we need to accept that language >>> is BOTH social AND individual, and we need to explore both aspects to get >>> as complete a picture as possible of how we communicate using language. We >>> need to know BOTH how people create and respond to meanings and express >>> those meanings in forms during social interaction AND the mechanisms which >>> operate in the brains of individuals in order to make such interaction >>> possible. Both are important parts of the answer to the question 'How do >>> we communicate using language?', though this particular thread of the >>> Funknet discussion has concentrated more on the second aspect, and so will >>> I. >>> >>> This doesn't mean that all the work linguists have done on "exploring the >>> structure of a language so that I can understand how all the bits fit >>> together" and "exploring the connections between items", as Dick puts it, >>> is useless - far from it. After all, the hypotheses that psycholinguists >>> test are based on ideas about what languages are like. But it does mean, >>> in my view, that ultimately we need to get evidence that the constructs >>> and analyses we propose are ones that are at least consistent with what we >>> know of the processes which go on when we use language. So I am with >>> Matthew when he says that for him, "the only sense in which an analysis >>> can be "the correct analysis" is in terms of what is inside of people's >>> heads". Of course, this doesn't imply that linguists should just give up >>> their jobs until such time as we know everything there is to know about >>> language processing. But it does mean that we need to collaborate with >>> psycholinguists, psychologists and neurologists, as has also been pointed >>> out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan >>> Nuyts. [We also need to collaborate much more with sociolinguists and >>> sociologists, so that we can get a better handle on the sociocultural >>> aspects of how we communicate.] And it also means that psycholinguists, >>> for their part, need whenever possible to follow up tightly controlled lab >>> experiments with studies under more naturalistic conditions, to avoid the >>> criticism that what happens in artifical lab situations may not happen in >>> natural communicative conditions. >>> >>> I also agree with Dick when he says that "the differences between >>> individuals really matter", and with Lise when she points out that "we >>> must also be careful not to idealize "what's in people's heads" as if it >>> were a single coherent construct that we are trying to discover". However, >>> there are surely processing mechanisms which are common to all language >>> users by virtue of the evolution of the language faculty and which >>> constitute the "general processes" which Dick says psycholinguists are >>> interested in. >>> >>> On the issue of quantitative methodology, I'm sympathetic in general to >>> Ted and Ev's views, though it does seem sensible to prioritise cases in >>> terms of a hierarchy such as Brian suggests. One thing this means is that >>> we should be giving our university students of linguistics (and some of >>> our linguistics lecturers!) courses in quantitative aspects of linguistics >>> that introduce them to the use of at least some of the basic statistical >>> methods in language study, and I'm sure this is indeed going on in some >>> enlightened places. To those who suspect this can't be done with maths-shy >>> students who don't initially see the need for it, I offer my own >>> experience, over quite a long period, of teaching such courses to people >>> with little or no prior experience in quantitative techniques. For some >>> years in the 1990s, I taught such courses to all linguistics students in >>> an institution where we had many mature students who had come into >>> university level studies with non-standard qualifications, and were not >>> well equipped for courses of this kind by their previous experience. I'm >>> glad to say that teaching the subject from their own perspective as >>> language students rather than that of the statistician, and explaining the >>> reasons for doing things in particular ways rather than just presenting >>> formulae, paid off in the end, so that most students were able to >>> appreciate the relevance of these courses and to turn in very creditable >>> projects showing an understanding of research design and competence in the >>> use of a range of basic statistical techniques. And I still find that >>> bright graduate students respond well to similar courses which incorporate >>> some of the rather more advanced techniques needed for many real research >>> projects in various areas of linguistics. But I may well be out of date >>> with what is now already happening in our fine institutions of higher >>> education! >>> >>> Chris Butler >>> >>> > > From tgivon at uoregon.edu Sun Sep 12 15:46:12 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Sun, 12 Sep 2010 09:46:12 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: I wonder whether asking "how does language work?" is really a meaningful question without asking "how does the language user work?" One of the worst legacies good ol' Noam stuck us with is his (really, Saussure's) distinction of competence ("language", "knowledge") vs. performance ("language user", "processing"). It purported to limit linguists to the armchair methods that study competence, and relegated to psychology the quantified, distributional/variationist methods that study behavior, processing and on-line communication. The first breach in this artificial methodological wall occurred, leastwise for some of us, when we discovered the intermediate method of quantified studies of text, interaction, and conversation. As an ex-biologist, I am forever puzzled by the methodological purism we sill seem to embrace in linguistics, in the face of the manifest complexity and connectivity of language (mind, brain, culture, sociality, evolution, etc.). In biology, another extremely complex science with multiple connections (chemistry, geology, paleontology, behavior, sociality, economics, evolution, etc.), ANY method is welcome so long as it does the job of furthering our understanding. And by understanding we mean ever-wider circles of connectivity. Best, TG ================ A. Katz wrote: > Chris, > > Thanks for addressing this question. I understand that many, many > linguists are quite properly and approriately interested in this > ultimate question: "How does the language user work?" (I am also > intetested in this question some of the time.) > > My point is that "how does language work?" is also a valid question, > and a central one to the field of linguistics. These two questions are > not at all the same. > > Let me be very explicit: My aim is to separate out the "human" from > the "language". There are many good reasons to do so. For anyone > working in computerized language processing or in non-human language > studies, this is a significant point. > > It does not matter if a computerized language processing system even > remotely simulates what humans do with language in their brains. It > does matter whether it comes up with comparable or indistinguishable > results. > > It does not matter whether a parrot, a dolphin or a chimpanzee is > doing the same things inside the same modules in his brain as a human > does. It does matter if the results are functionally equivalent. > > We need to make that distinction, between humans and their language, > or we will be caught inside a circular definition with no way to break > out or to prove anything, not about others and not about ourselves! > > --Aya > > http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies > > > On Sun, 12 Sep 2010, Chris Butler wrote: > >> Aya, I think two different things are getting a bit mixed up here. >> >> I don't for a moment dispute that expressions are often susceptible >> to multiple interpretations, that these interpretations are guided by >> all kinds of contextual information, or that different people, or >> even the same person at different times, may end up selecting >> differently from the various interpretations. Your example 'Open >> happiness' in another contribution to this thread illustrates these >> points very well. >> >> My point, though, is that each of these different interpretations, as >> well as the selection of one (or more) as more likely in a particular >> context, is achieved through mechanisms in the interpreter's brain >> which evolved in the course of the phylogenetic development of >> language in the human species, and developed ontogenetically in that >> particular interpreter's brain. It is surely likely that those >> mechanisms are highly similar in different human beings, even though >> there may be differences in the detailed wiring in different brains. >> What I'm saying is that in order to answer the question 'How do we >> communicate using language?' or, if you prefer, 'How does the >> language user work?', we need to investigate what those mechanisms >> are, and this is what psycholinguists can help us with. >> >> In particular, as linguists, we are interested (well, some of us are, >> though clearly not all) in whether the constructs we posit in our >> theories of language have psychological validity in the sense that >> they correspond to ways in which aspects of language are represented >> in the brain. [As an aside, I do realise that there are linguists who >> strenuously resist what they see as a misguided emphasis on mental >> representation in the work of cognitive scientists, but it seems >> clear that language must be represented in some way in the brain in >> order that we can engage in the sociosemiotic acts of meaning making >> which are the primary focus for many of these critics.] Taking your >> 'Open happiness' examples again, I think we need answers to questions >> such as: What kind of representation does the human language >> processing system have for 'open' and for 'happiness'? Are the >> phonological, syntactic, semantic and pragmatic (for those who >> distinguish the semantic from the pragmatic) properties of these >> items (and we might want to add 'for this particular interpreter', >> though there must be similarities across interpreters for >> communication to be possible) represented in the same or different >> ways, in the same or different locations (or sets of distributed >> locations)? Indeed, are we right in thinking that these familiar >> levels of linguistic description must be differentiated, as such, in >> the human language processing system? Does the representation for >> 'open' distinguish between what we call verbal and adjectival uses of >> this item, and if so how? Or are syntactic analyses computed on the >> fly, using semantic and contextual clues, rather than the neural >> equivalent of pigeon holes corresponding to verbs and adjectives? Is >> 'happiness' represented as 'happy + ness', or in its entirety, or >> both? All these questions, and many many more, are relevant to the >> construction of a model of language which reflects how language users >> communicate (as, of course, are a whole set of other questions about >> the sociocultural aspects of communication). I am not a >> psycholinguist, but my all too superficial reading in the area >> suggests that psycholinguists have gone some of the way towards >> answering some of the questions we might want to ask, but that there >> is still a long way to go. >> >> Chris >> -------------------------------------------------- >> From: "A. Katz" >> Sent: Saturday, September 11, 2010 3:18 PM >> To: "Chris Butler" >> Cc: "FUNKNET" >> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >> >>> The problem is that once we achieve the psycholinguistic goal and >>> see what >>> is happening in each person's head, and we see that conflicting >>> analyses >>> are the norm, rather than the exception, among normal language >>> speakers, >>> then we will realize that the way language works to transmit >>> information >>> is despite individual differences and not because of uniform processing >>> strategies. >>> >>> Even when all are agreed as to the meaning of an utterance, they do not >>> process it the same way. Which means that processing is seocndary to >>> information transmission. >>> >>> --Aya >>> >>> >>> >>> >>> On Sat, 11 Sep 2010, Chris Butler wrote: >>> >>>> Dick's comment that "this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual" is, it seems to me, an important one, particularly for >>>> those of us who are committed functionalists. My own view is that a >>>> truly functional model of language would be one which aims to >>>> account for how human beings communicate using language, or in >>>> other words tries to answer the question which was posed by Simon >>>> Dik a long time ago now, but which was not tackled head-on in his >>>> own work: "How does the natural language user work?' In trying to >>>> answer this question we need to accept that language is BOTH social >>>> AND individual, and we need to explore both aspects to get as >>>> complete a picture as possible of how we communicate using >>>> language. We need to know BOTH how people create and respond to >>>> meanings and express those meanings in forms during social >>>> interaction AND the mechanisms which operate in the brains of >>>> individuals in order to make such interaction possible. Both are >>>> important parts of the answer to the question 'How do we >>>> communicate using language?', though this particular thread of the >>>> Funknet discussion has concentrated more on the second aspect, and >>>> so will I. >>>> >>>> This doesn't mean that all the work linguists have done on >>>> "exploring the structure of a language so that I can understand how >>>> all the bits fit together" and "exploring the connections between >>>> items", as Dick puts it, is useless - far from it. After all, the >>>> hypotheses that psycholinguists test are based on ideas about what >>>> languages are like. But it does mean, in my view, that ultimately >>>> we need to get evidence that the constructs and analyses we propose >>>> are ones that are at least consistent with what we know of the >>>> processes which go on when we use language. So I am with Matthew >>>> when he says that for him, "the only sense in which an analysis can >>>> be "the correct analysis" is in terms of what is inside of people's >>>> heads". Of course, this doesn't imply that linguists should just >>>> give up their jobs until such time as we know everything there is >>>> to know about language processing. But it does mean that we need to >>>> collaborate with psycholinguists, psychologists and neurologists, >>>> as has also been pointed out by linguists such as Ray Jackendoff, >>>> Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>> collaborate much more with sociolinguists and sociologists, so that >>>> we can get a better handle on the sociocultural aspects of how we >>>> communicate.] And it also means that psycholinguists, for their >>>> part, need whenever possible to follow up tightly controlled lab >>>> experiments with studies under more naturalistic conditions, to >>>> avoid the criticism that what happens in artifical lab situations >>>> may not happen in natural communicative conditions. >>>> >>>> I also agree with Dick when he says that "the differences between >>>> individuals really matter", and with Lise when she points out that >>>> "we must also be careful not to idealize "what's in people's heads" >>>> as if it were a single coherent construct that we are trying to >>>> discover". However, there are surely processing mechanisms which >>>> are common to all language users by virtue of the evolution of the >>>> language faculty and which constitute the "general processes" which >>>> Dick says psycholinguists are interested in. >>>> >>>> On the issue of quantitative methodology, I'm sympathetic in >>>> general to Ted and Ev's views, though it does seem sensible to >>>> prioritise cases in terms of a hierarchy such as Brian suggests. >>>> One thing this means is that we should be giving our university >>>> students of linguistics (and some of our linguistics lecturers!) >>>> courses in quantitative aspects of linguistics that introduce them >>>> to the use of at least some of the basic statistical methods in >>>> language study, and I'm sure this is indeed going on in some >>>> enlightened places. To those who suspect this can't be done with >>>> maths-shy students who don't initially see the need for it, I offer >>>> my own experience, over quite a long period, of teaching such >>>> courses to people with little or no prior experience in >>>> quantitative techniques. For some years in the 1990s, I taught such >>>> courses to all linguistics students in an institution where we had >>>> many mature students who had come into university level studies >>>> with non-standard qualifications, and were not well equipped for >>>> courses of this kind by their previous experience. I'm glad to say >>>> that teaching the subject from their own perspective as language >>>> students rather than that of the statistician, and explaining the >>>> reasons for doing things in particular ways rather than just >>>> presenting formulae, paid off in the end, so that most students >>>> were able to appreciate the relevance of these courses and to turn >>>> in very creditable projects showing an understanding of research >>>> design and competence in the use of a range of basic statistical >>>> techniques. And I still find that bright graduate students respond >>>> well to similar courses which incorporate some of the rather more >>>> advanced techniques needed for many real research projects in >>>> various areas of linguistics. But I may well be out of date with >>>> what is now already happening in our fine institutions of higher >>>> education! >>>> >>>> Chris Butler >>>> >>>> >> >> From amnfn at well.com Sun Sep 12 16:19:46 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 09:19:46 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <4C8CF5C4.8070005@uoregon.edu> Message-ID: Tom, I understand the uncomfortable association with Chomsky that speaking of language apart from people can have. Competence versus performance, the way Chomsky used those terms, never made sense. But that's precisely because he associated "competence" with the brain and suggested that it was hard wired there -- when there was never any evidence of that. However, if we don't distinguish language from humans, and language processing from language data, then how are we going to judge artificial language-using devices as to their efficacy at producing and interpreting language? How are we going to determine whether and to what extent a non-human has acquired language? We aren't born with it. We don't embody it. It's a tool that we use to communicate. Other tools can be studied separately from the people who use them or created them. Why not language? --Aya On Sun, 12 Sep 2010, Tom Givon wrote: > > > > I wonder whether asking "how does language work?" is really a meaningful > question without asking "how does the language user work?" One of the worst > legacies good ol' Noam stuck us with is his (really, Saussure's) distinction > of competence ("language", "knowledge") vs. performance ("language user", > "processing"). It purported to limit linguists to the armchair methods that > study competence, and relegated to psychology the quantified, > distributional/variationist methods that study behavior, processing and > on-line communication. The first breach in this artificial methodological > wall occurred, leastwise for some of us, when we discovered the intermediate > method of quantified studies of text, interaction, and conversation. As an > ex-biologist, I am forever puzzled by the methodological purism we sill seem > to embrace in linguistics, in the face of the manifest complexity and > connectivity of language (mind, brain, culture, sociality, evolution, etc.). > In biology, another extremely complex science with multiple connections > (chemistry, geology, paleontology, behavior, sociality, economics, evolution, > etc.), ANY method is welcome so long as it does the job of furthering our > understanding. And by understanding we mean ever-wider circles of > connectivity. > > Best, TG > ================ > > > A. Katz wrote: >> Chris, >> >> Thanks for addressing this question. I understand that many, many linguists >> are quite properly and approriately interested in this ultimate question: >> "How does the language user work?" (I am also intetested in this question >> some of the time.) >> >> My point is that "how does language work?" is also a valid question, and a >> central one to the field of linguistics. These two questions are not at all >> the same. >> >> Let me be very explicit: My aim is to separate out the "human" from the >> "language". There are many good reasons to do so. For anyone working in >> computerized language processing or in non-human language studies, this is >> a significant point. >> >> It does not matter if a computerized language processing system even >> remotely simulates what humans do with language in their brains. It does >> matter whether it comes up with comparable or indistinguishable results. >> >> It does not matter whether a parrot, a dolphin or a chimpanzee is doing the >> same things inside the same modules in his brain as a human does. It does >> matter if the results are functionally equivalent. >> >> We need to make that distinction, between humans and their language, or we >> will be caught inside a circular definition with no way to break out or to >> prove anything, not about others and not about ourselves! >> >> --Aya >> >> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >> >> >> On Sun, 12 Sep 2010, Chris Butler wrote: >> >>> Aya, I think two different things are getting a bit mixed up here. >>> >>> I don't for a moment dispute that expressions are often susceptible to >>> multiple interpretations, that these interpretations are guided by all >>> kinds of contextual information, or that different people, or even the >>> same person at different times, may end up selecting differently from the >>> various interpretations. Your example 'Open happiness' in another >>> contribution to this thread illustrates these points very well. >>> >>> My point, though, is that each of these different interpretations, as well >>> as the selection of one (or more) as more likely in a particular context, >>> is achieved through mechanisms in the interpreter's brain which evolved in >>> the course of the phylogenetic development of language in the human >>> species, and developed ontogenetically in that particular interpreter's >>> brain. It is surely likely that those mechanisms are highly similar in >>> different human beings, even though there may be differences in the >>> detailed wiring in different brains. What I'm saying is that in order to >>> answer the question 'How do we communicate using language?' or, if you >>> prefer, 'How does the language user work?', we need to investigate what >>> those mechanisms are, and this is what psycholinguists can help us with. >>> >>> In particular, as linguists, we are interested (well, some of us are, >>> though clearly not all) in whether the constructs we posit in our theories >>> of language have psychological validity in the sense that they correspond >>> to ways in which aspects of language are represented in the brain. [As an >>> aside, I do realise that there are linguists who strenuously resist what >>> they see as a misguided emphasis on mental representation in the work of >>> cognitive scientists, but it seems clear that language must be represented >>> in some way in the brain in order that we can engage in the sociosemiotic >>> acts of meaning making which are the primary focus for many of these >>> critics.] Taking your 'Open happiness' examples again, I think we need >>> answers to questions such as: What kind of representation does the human >>> language processing system have for 'open' and for 'happiness'? Are the >>> phonological, syntactic, semantic and pragmatic (for those who distinguish >>> the semantic from the pragmatic) properties of these items (and we might >>> want to add 'for this particular interpreter', though there must be >>> similarities across interpreters for communication to be possible) >>> represented in the same or different ways, in the same or different >>> locations (or sets of distributed locations)? Indeed, are we right in >>> thinking that these familiar levels of linguistic description must be >>> differentiated, as such, in the human language processing system? Does the >>> representation for 'open' distinguish between what we call verbal and >>> adjectival uses of this item, and if so how? Or are syntactic analyses >>> computed on the fly, using semantic and contextual clues, rather than the >>> neural equivalent of pigeon holes corresponding to verbs and adjectives? >>> Is 'happiness' represented as 'happy + ness', or in its entirety, or both? >>> All these questions, and many many more, are relevant to the construction >>> of a model of language which reflects how language users communicate (as, >>> of course, are a whole set of other questions about the sociocultural >>> aspects of communication). I am not a psycholinguist, but my all too >>> superficial reading in the area suggests that psycholinguists have gone >>> some of the way towards answering some of the questions we might want to >>> ask, but that there is still a long way to go. >>> >>> Chris >>> -------------------------------------------------- >>> From: "A. Katz" >>> Sent: Saturday, September 11, 2010 3:18 PM >>> To: "Chris Butler" >>> Cc: "FUNKNET" >>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>> >>>> The problem is that once we achieve the psycholinguistic goal and see >>>> what >>>> is happening in each person's head, and we see that conflicting analyses >>>> are the norm, rather than the exception, among normal language speakers, >>>> then we will realize that the way language works to transmit information >>>> is despite individual differences and not because of uniform processing >>>> strategies. >>>> >>>> Even when all are agreed as to the meaning of an utterance, they do not >>>> process it the same way. Which means that processing is seocndary to >>>> information transmission. >>>> >>>> --Aya >>>> >>>> >>>> >>>> >>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>> >>>>> Dick's comment that "this discussion raises the really fundamental >>>>> question of what kind of thing we think language is: social or >>>>> individual" is, it seems to me, an important one, particularly for those >>>>> of us who are committed functionalists. My own view is that a truly >>>>> functional model of language would be one which aims to account for how >>>>> human beings communicate using language, or in other words tries to >>>>> answer the question which was posed by Simon Dik a long time ago now, >>>>> but which was not tackled head-on in his own work: "How does the natural >>>>> language user work?' In trying to answer this question we need to accept >>>>> that language is BOTH social AND individual, and we need to explore both >>>>> aspects to get as complete a picture as possible of how we communicate >>>>> using language. We need to know BOTH how people create and respond to >>>>> meanings and express those meanings in forms during social interaction >>>>> AND the mechanisms which operate in the brains of individuals in order >>>>> to make such interaction possible. Both are important parts of the >>>>> answer to the question 'How do we communicate using language?', though >>>>> this particular thread of the Funknet discussion has concentrated more >>>>> on the second aspect, and so will I. >>>>> >>>>> This doesn't mean that all the work linguists have done on "exploring >>>>> the structure of a language so that I can understand how all the bits >>>>> fit together" and "exploring the connections between items", as Dick >>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>> psycholinguists test are based on ideas about what languages are like. >>>>> But it does mean, in my view, that ultimately we need to get evidence >>>>> that the constructs and analyses we propose are ones that are at least >>>>> consistent with what we know of the processes which go on when we use >>>>> language. So I am with Matthew when he says that for him, "the only >>>>> sense in which an analysis can be "the correct analysis" is in terms of >>>>> what is inside of people's heads". Of course, this doesn't imply that >>>>> linguists should just give up their jobs until such time as we know >>>>> everything there is to know about language processing. But it does mean >>>>> that we need to collaborate with psycholinguists, psychologists and >>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>> collaborate much more with sociolinguists and sociologists, so that we >>>>> can get a better handle on the sociocultural aspects of how we >>>>> communicate.] And it also means that psycholinguists, for their part, >>>>> need whenever possible to follow up tightly controlled lab experiments >>>>> with studies under more naturalistic conditions, to avoid the criticism >>>>> that what happens in artifical lab situations may not happen in natural >>>>> communicative conditions. >>>>> >>>>> I also agree with Dick when he says that "the differences between >>>>> individuals really matter", and with Lise when she points out that "we >>>>> must also be careful not to idealize "what's in people's heads" as if it >>>>> were a single coherent construct that we are trying to discover". >>>>> However, there are surely processing mechanisms which are common to all >>>>> language users by virtue of the evolution of the language faculty and >>>>> which constitute the "general processes" which Dick says psycholinguists >>>>> are interested in. >>>>> >>>>> On the issue of quantitative methodology, I'm sympathetic in general to >>>>> Ted and Ev's views, though it does seem sensible to prioritise cases in >>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>> that we should be giving our university students of linguistics (and >>>>> some of our linguistics lecturers!) courses in quantitative aspects of >>>>> linguistics that introduce them to the use of at least some of the basic >>>>> statistical methods in language study, and I'm sure this is indeed going >>>>> on in some enlightened places. To those who suspect this can't be done >>>>> with maths-shy students who don't initially see the need for it, I offer >>>>> my own experience, over quite a long period, of teaching such courses to >>>>> people with little or no prior experience in quantitative techniques. >>>>> For some years in the 1990s, I taught such courses to all linguistics >>>>> students in an institution where we had many mature students who had >>>>> come into university level studies with non-standard qualifications, and >>>>> were not well equipped for courses of this kind by their previous >>>>> experience. I'm glad to say that teaching the subject from their own >>>>> perspective as language students rather than that of the statistician, >>>>> and explaining the reasons for doing things in particular ways rather >>>>> than just presenting formulae, paid off in the end, so that most >>>>> students were able to appreciate the relevance of these courses and to >>>>> turn in very creditable projects showing an understanding of research >>>>> design and competence in the use of a range of basic statistical >>>>> techniques. And I still find that bright graduate students respond well >>>>> to similar courses which incorporate some of the rather more advanced >>>>> techniques needed for many real research projects in various areas of >>>>> linguistics. But I may well be out of date with what is now already >>>>> happening in our fine institutions of higher education! >>>>> >>>>> Chris Butler >>>>> >>>>> >>> >>> > > From mark at polymathix.com Mon Sep 13 00:16:10 2010 From: mark at polymathix.com (Mark P. Line) Date: Sun, 12 Sep 2010 19:16:10 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya -- You said: "Other tools can be studied separately from the people who use them or created them. Why not language?" Although language can certainly be considered a tool, I think it's unlike other tools in several very significant respects. 1. Although we're not born with language, we can't avoid (pathologies excluded) acquiring it unless we're not exposed to it. To that extent, language is a biological phenomenon. A prototypical tool is not a biological phenomenon, so I'm not sure how valid any conclusions might be that are drawn from a premise of language-as-tool. 2. A tool is as a tool does: Anything is a tool that is being used as a tool, including dead wombats, broken screwdrivers or decks of playing cards. (Completing the imagined scenarios is left as an exercise for the reader...) So saying that language is a tool is only saying that language is used as a tool. Quite a few conclusions can be and have been drawn from the fact that language is used as a tool, but I would have to be convinced in detail that almost everything worth knowing about language is dependent on the premise of language-as-tool. 3. If language is a "tool" for (say) communicating ideas, then eating is a "tool" for reducing hunger. In both cases, I worry about the tool metaphor being stretched so far from the prototype that we're left with an out-and-out category fallacy. 4. More prototypical tools can be studied separately from those who use or create them because those tools are easily observed separately from those who use or create them. I don't think the same thing can be said of language -- very little about language can be observed apart from its use, so very little about language can be observed apart from its user(s). 5. Any proposal to study something as complex as language separately from its embodiment is suspicious to me, smacking of reductionism -- something up with which linguistics has had to put a tad much. Anything that puts language back into its human context would be a step forward. -- Mark Mark P. Line A. Katz wrote: > Tom, > > I understand the uncomfortable association with Chomsky that speaking of > language apart from people can have. Competence versus performance, the > way Chomsky used those terms, never made sense. But that's precisely > because > he associated "competence" with the brain and suggested that it was hard > wired there -- when there was never any evidence of that. > > However, if we don't distinguish language from humans, and language > processing from language data, then how are we going to judge artificial > language-using devices as to their efficacy at producing and interpreting > language? How are we going to determine whether and to what extent a > non-human has acquired language? > > We aren't born with it. We don't embody it. It's a tool that we use to > communicate. Other tools can be studied separately from the people who use > them or created them. Why not language? > > --Aya > > > > > On Sun, 12 Sep 2010, Tom Givon wrote: > >> >> >> >> I wonder whether asking "how does language work?" is really a meaningful >> question without asking "how does the language user work?" One of the >> worst >> legacies good ol' Noam stuck us with is his (really, Saussure's) >> distinction >> of competence ("language", "knowledge") vs. performance ("language >> user", >> "processing"). It purported to limit linguists to the armchair methods >> that >> study competence, and relegated to psychology the quantified, >> distributional/variationist methods that study behavior, processing and >> on-line communication. The first breach in this artificial >> methodological >> wall occurred, leastwise for some of us, when we discovered the >> intermediate >> method of quantified studies of text, interaction, and conversation. As >> an >> ex-biologist, I am forever puzzled by the methodological purism we sill >> seem >> to embrace in linguistics, in the face of the manifest complexity and >> connectivity of language (mind, brain, culture, sociality, evolution, >> etc.). >> In biology, another extremely complex science with multiple connections >> (chemistry, geology, paleontology, behavior, sociality, economics, >> evolution, >> etc.), ANY method is welcome so long as it does the job of furthering >> our >> understanding. And by understanding we mean ever-wider circles of >> connectivity. >> >> Best, TG >> ================ >> >> >> A. Katz wrote: >>> Chris, >>> >>> Thanks for addressing this question. I understand that many, many >>> linguists >>> are quite properly and approriately interested in this ultimate >>> question: >>> "How does the language user work?" (I am also intetested in this >>> question >>> some of the time.) >>> >>> My point is that "how does language work?" is also a valid question, >>> and a >>> central one to the field of linguistics. These two questions are not at >>> all >>> the same. >>> >>> Let me be very explicit: My aim is to separate out the "human" from the >>> "language". There are many good reasons to do so. For anyone working in >>> computerized language processing or in non-human language studies, this >>> is >>> a significant point. >>> >>> It does not matter if a computerized language processing system even >>> remotely simulates what humans do with language in their brains. It >>> does >>> matter whether it comes up with comparable or indistinguishable >>> results. >>> >>> It does not matter whether a parrot, a dolphin or a chimpanzee is doing >>> the >>> same things inside the same modules in his brain as a human does. It >>> does >>> matter if the results are functionally equivalent. >>> >>> We need to make that distinction, between humans and their language, or >>> we >>> will be caught inside a circular definition with no way to break out or >>> to >>> prove anything, not about others and not about ourselves! >>> >>> --Aya >>> >>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>> >>> >>> On Sun, 12 Sep 2010, Chris Butler wrote: >>> >>>> Aya, I think two different things are getting a bit mixed up here. >>>> >>>> I don't for a moment dispute that expressions are often susceptible to >>>> multiple interpretations, that these interpretations are guided by all >>>> kinds of contextual information, or that different people, or even the >>>> same person at different times, may end up selecting differently from >>>> the >>>> various interpretations. Your example 'Open happiness' in another >>>> contribution to this thread illustrates these points very well. >>>> >>>> My point, though, is that each of these different interpretations, as >>>> well >>>> as the selection of one (or more) as more likely in a particular >>>> context, >>>> is achieved through mechanisms in the interpreter's brain which >>>> evolved in >>>> the course of the phylogenetic development of language in the human >>>> species, and developed ontogenetically in that particular >>>> interpreter's >>>> brain. It is surely likely that those mechanisms are highly similar in >>>> different human beings, even though there may be differences in the >>>> detailed wiring in different brains. What I'm saying is that in order >>>> to >>>> answer the question 'How do we communicate using language?' or, if you >>>> prefer, 'How does the language user work?', we need to investigate >>>> what >>>> those mechanisms are, and this is what psycholinguists can help us >>>> with. >>>> >>>> In particular, as linguists, we are interested (well, some of us are, >>>> though clearly not all) in whether the constructs we posit in our >>>> theories >>>> of language have psychological validity in the sense that they >>>> correspond >>>> to ways in which aspects of language are represented in the brain. [As >>>> an >>>> aside, I do realise that there are linguists who strenuously resist >>>> what >>>> they see as a misguided emphasis on mental representation in the work >>>> of >>>> cognitive scientists, but it seems clear that language must be >>>> represented >>>> in some way in the brain in order that we can engage in the >>>> sociosemiotic >>>> acts of meaning making which are the primary focus for many of these >>>> critics.] Taking your 'Open happiness' examples again, I think we need >>>> answers to questions such as: What kind of representation does the >>>> human >>>> language processing system have for 'open' and for 'happiness'? Are >>>> the >>>> phonological, syntactic, semantic and pragmatic (for those who >>>> distinguish >>>> the semantic from the pragmatic) properties of these items (and we >>>> might >>>> want to add 'for this particular interpreter', though there must be >>>> similarities across interpreters for communication to be possible) >>>> represented in the same or different ways, in the same or different >>>> locations (or sets of distributed locations)? Indeed, are we right in >>>> thinking that these familiar levels of linguistic description must be >>>> differentiated, as such, in the human language processing system? Does >>>> the >>>> representation for 'open' distinguish between what we call verbal and >>>> adjectival uses of this item, and if so how? Or are syntactic analyses >>>> computed on the fly, using semantic and contextual clues, rather than >>>> the >>>> neural equivalent of pigeon holes corresponding to verbs and >>>> adjectives? >>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>> both? >>>> All these questions, and many many more, are relevant to the >>>> construction >>>> of a model of language which reflects how language users communicate >>>> (as, >>>> of course, are a whole set of other questions about the sociocultural >>>> aspects of communication). I am not a psycholinguist, but my all too >>>> superficial reading in the area suggests that psycholinguists have >>>> gone >>>> some of the way towards answering some of the questions we might want >>>> to >>>> ask, but that there is still a long way to go. >>>> >>>> Chris >>>> -------------------------------------------------- >>>> From: "A. Katz" >>>> Sent: Saturday, September 11, 2010 3:18 PM >>>> To: "Chris Butler" >>>> Cc: "FUNKNET" >>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>> >>>>> The problem is that once we achieve the psycholinguistic goal and see >>>>> what >>>>> is happening in each person's head, and we see that conflicting >>>>> analyses >>>>> are the norm, rather than the exception, among normal language >>>>> speakers, >>>>> then we will realize that the way language works to transmit >>>>> information >>>>> is despite individual differences and not because of uniform >>>>> processing >>>>> strategies. >>>>> >>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>> not >>>>> process it the same way. Which means that processing is seocndary to >>>>> information transmission. >>>>> >>>>> --Aya >>>>> >>>>> >>>>> >>>>> >>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>> >>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>> question of what kind of thing we think language is: social or >>>>>> individual" is, it seems to me, an important one, particularly for >>>>>> those >>>>>> of us who are committed functionalists. My own view is that a truly >>>>>> functional model of language would be one which aims to account for >>>>>> how >>>>>> human beings communicate using language, or in other words tries to >>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>> now, >>>>>> but which was not tackled head-on in his own work: "How does the >>>>>> natural >>>>>> language user work?' In trying to answer this question we need to >>>>>> accept >>>>>> that language is BOTH social AND individual, and we need to explore >>>>>> both >>>>>> aspects to get as complete a picture as possible of how we >>>>>> communicate >>>>>> using language. We need to know BOTH how people create and respond >>>>>> to >>>>>> meanings and express those meanings in forms during social >>>>>> interaction >>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>> order >>>>>> to make such interaction possible. Both are important parts of the >>>>>> answer to the question 'How do we communicate using language?', >>>>>> though >>>>>> this particular thread of the Funknet discussion has concentrated >>>>>> more >>>>>> on the second aspect, and so will I. >>>>>> >>>>>> This doesn't mean that all the work linguists have done on >>>>>> "exploring >>>>>> the structure of a language so that I can understand how all the >>>>>> bits >>>>>> fit together" and "exploring the connections between items", as Dick >>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>> psycholinguists test are based on ideas about what languages are >>>>>> like. >>>>>> But it does mean, in my view, that ultimately we need to get >>>>>> evidence >>>>>> that the constructs and analyses we propose are ones that are at >>>>>> least >>>>>> consistent with what we know of the processes which go on when we >>>>>> use >>>>>> language. So I am with Matthew when he says that for him, "the only >>>>>> sense in which an analysis can be "the correct analysis" is in terms >>>>>> of >>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>> that >>>>>> linguists should just give up their jobs until such time as we know >>>>>> everything there is to know about language processing. But it does >>>>>> mean >>>>>> that we need to collaborate with psycholinguists, psychologists and >>>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>>> collaborate much more with sociolinguists and sociologists, so that >>>>>> we >>>>>> can get a better handle on the sociocultural aspects of how we >>>>>> communicate.] And it also means that psycholinguists, for their >>>>>> part, >>>>>> need whenever possible to follow up tightly controlled lab >>>>>> experiments >>>>>> with studies under more naturalistic conditions, to avoid the >>>>>> criticism >>>>>> that what happens in artifical lab situations may not happen in >>>>>> natural >>>>>> communicative conditions. >>>>>> >>>>>> I also agree with Dick when he says that "the differences between >>>>>> individuals really matter", and with Lise when she points out that >>>>>> "we >>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>> if it >>>>>> were a single coherent construct that we are trying to discover". >>>>>> However, there are surely processing mechanisms which are common to >>>>>> all >>>>>> language users by virtue of the evolution of the language faculty >>>>>> and >>>>>> which constitute the "general processes" which Dick says >>>>>> psycholinguists >>>>>> are interested in. >>>>>> >>>>>> On the issue of quantitative methodology, I'm sympathetic in general >>>>>> to >>>>>> Ted and Ev's views, though it does seem sensible to prioritise cases >>>>>> in >>>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>>> that we should be giving our university students of linguistics (and >>>>>> some of our linguistics lecturers!) courses in quantitative aspects >>>>>> of >>>>>> linguistics that introduce them to the use of at least some of the >>>>>> basic >>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>> going >>>>>> on in some enlightened places. To those who suspect this can't be >>>>>> done >>>>>> with maths-shy students who don't initially see the need for it, I >>>>>> offer >>>>>> my own experience, over quite a long period, of teaching such >>>>>> courses to >>>>>> people with little or no prior experience in quantitative >>>>>> techniques. >>>>>> For some years in the 1990s, I taught such courses to all >>>>>> linguistics >>>>>> students in an institution where we had many mature students who had >>>>>> come into university level studies with non-standard qualifications, >>>>>> and >>>>>> were not well equipped for courses of this kind by their previous >>>>>> experience. I'm glad to say that teaching the subject from their own >>>>>> perspective as language students rather than that of the >>>>>> statistician, >>>>>> and explaining the reasons for doing things in particular ways >>>>>> rather >>>>>> than just presenting formulae, paid off in the end, so that most >>>>>> students were able to appreciate the relevance of these courses and >>>>>> to >>>>>> turn in very creditable projects showing an understanding of >>>>>> research >>>>>> design and competence in the use of a range of basic statistical >>>>>> techniques. And I still find that bright graduate students respond >>>>>> well >>>>>> to similar courses which incorporate some of the rather more >>>>>> advanced >>>>>> techniques needed for many real research projects in various areas >>>>>> of >>>>>> linguistics. But I may well be out of date with what is now already >>>>>> happening in our fine institutions of higher education! >>>>>> >>>>>> Chris Butler >>>>>> >>>>>> >>>> >>>> >> >> > > -- Mark Mark P. Line Bartlesville, OK From amnfn at well.com Mon Sep 13 00:54:09 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 17:54:09 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <2a720add3be870cd8f39d93787201ee3.squirrel@sm.webmail.pair.com> Message-ID: Mark, Sorry, but "eating" is not a tool. It's a biological process. Eating cannot occur outside the biological context. A human being can avoid eating, but if so he starves. Feral children do not speak, but they eat, like everyone else, or they die. Eating does not have to be taught, there is no critical age of acquisition and it is not uniquely an artefact of human culture. If you are an animal, you eat. Eating cannot survive the death of the eater. Language can. Language can be transmitted over great distances by many different artificial methods. Someone's words may be heard long after he has died. They may be passed down verbatim from one generation to the next. Language survives its speakers. It can be re-established and re-used thousands of years after the last native speaker died. It can be generated by computers. It can be transmitted using light, sound, electrical pulses. It can be studied in the absence of speakers. The sensual modalities by which language can be transmitted are not themselbves language. Language is in the contrasts. It's a very abstract tool, not unlike morse code or music theory or mathematics. The fact that it's abstract is the biggest hurdle to understanding that language is a tool. But culture is abstract, too, and surely we can see that it's not a biologcally inherited property like eating or breathing or sleeping. You might as well say that a person cannot possibly avoid watching TV once he's exposed to it, as say the same about language. But people can survive just fine without television, and unless someone shows them how to make a TV set, most people will never figure out how to build one. The same goes for language. We're great at using it, not so great at generating it out of thin air with no ambient culture. --Aya On Sun, 12 Sep 2010, Mark P. Line wrote: > Aya -- > > You said: "Other tools can be studied separately from the people who use > them or created them. Why not language?" > > Although language can certainly be considered a tool, I think it's unlike > other tools in several very significant respects. > > 1. Although we're not born with language, we can't avoid (pathologies > excluded) acquiring it unless we're not exposed to it. To that extent, > language is a biological phenomenon. A prototypical tool is not a > biological phenomenon, so I'm not sure how valid any conclusions might be > that are drawn from a premise of language-as-tool. > > 2. A tool is as a tool does: Anything is a tool that is being used as a > tool, including dead wombats, broken screwdrivers or decks of playing > cards. (Completing the imagined scenarios is left as an exercise for the > reader...) So saying that language is a tool is only saying that language > is used as a tool. Quite a few conclusions can be and have been drawn from > the fact that language is used as a tool, but I would have to be convinced > in detail that almost everything worth knowing about language is dependent > on the premise of language-as-tool. > > 3. If language is a "tool" for (say) communicating ideas, then eating is a > "tool" for reducing hunger. In both cases, I worry about the tool metaphor > being stretched so far from the prototype that we're left with an > out-and-out category fallacy. > > 4. More prototypical tools can be studied separately from those who use or > create them because those tools are easily observed separately from those > who use or create them. I don't think the same thing can be said of > language -- very little about language can be observed apart from its use, > so very little about language can be observed apart from its user(s). > > 5. Any proposal to study something as complex as language separately from > its embodiment is suspicious to me, smacking of reductionism -- something > up with which linguistics has had to put a tad much. Anything that puts > language back into its human context would be a step forward. > > > -- Mark > > Mark P. Line > > > > A. Katz wrote: >> Tom, >> >> I understand the uncomfortable association with Chomsky that speaking of >> language apart from people can have. Competence versus performance, the >> way Chomsky used those terms, never made sense. But that's precisely >> because >> he associated "competence" with the brain and suggested that it was hard >> wired there -- when there was never any evidence of that. >> >> However, if we don't distinguish language from humans, and language >> processing from language data, then how are we going to judge artificial >> language-using devices as to their efficacy at producing and interpreting >> language? How are we going to determine whether and to what extent a >> non-human has acquired language? >> >> We aren't born with it. We don't embody it. It's a tool that we use to >> communicate. Other tools can be studied separately from the people who use >> them or created them. Why not language? >> >> --Aya >> >> >> >> >> On Sun, 12 Sep 2010, Tom Givon wrote: >> >>> >>> >>> >>> I wonder whether asking "how does language work?" is really a meaningful >>> question without asking "how does the language user work?" One of the >>> worst >>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>> distinction >>> of competence ("language", "knowledge") vs. performance ("language >>> user", >>> "processing"). It purported to limit linguists to the armchair methods >>> that >>> study competence, and relegated to psychology the quantified, >>> distributional/variationist methods that study behavior, processing and >>> on-line communication. The first breach in this artificial >>> methodological >>> wall occurred, leastwise for some of us, when we discovered the >>> intermediate >>> method of quantified studies of text, interaction, and conversation. As >>> an >>> ex-biologist, I am forever puzzled by the methodological purism we sill >>> seem >>> to embrace in linguistics, in the face of the manifest complexity and >>> connectivity of language (mind, brain, culture, sociality, evolution, >>> etc.). >>> In biology, another extremely complex science with multiple connections >>> (chemistry, geology, paleontology, behavior, sociality, economics, >>> evolution, >>> etc.), ANY method is welcome so long as it does the job of furthering >>> our >>> understanding. And by understanding we mean ever-wider circles of >>> connectivity. >>> >>> Best, TG >>> ================ >>> >>> >>> A. Katz wrote: >>>> Chris, >>>> >>>> Thanks for addressing this question. I understand that many, many >>>> linguists >>>> are quite properly and approriately interested in this ultimate >>>> question: >>>> "How does the language user work?" (I am also intetested in this >>>> question >>>> some of the time.) >>>> >>>> My point is that "how does language work?" is also a valid question, >>>> and a >>>> central one to the field of linguistics. These two questions are not at >>>> all >>>> the same. >>>> >>>> Let me be very explicit: My aim is to separate out the "human" from the >>>> "language". There are many good reasons to do so. For anyone working in >>>> computerized language processing or in non-human language studies, this >>>> is >>>> a significant point. >>>> >>>> It does not matter if a computerized language processing system even >>>> remotely simulates what humans do with language in their brains. It >>>> does >>>> matter whether it comes up with comparable or indistinguishable >>>> results. >>>> >>>> It does not matter whether a parrot, a dolphin or a chimpanzee is doing >>>> the >>>> same things inside the same modules in his brain as a human does. It >>>> does >>>> matter if the results are functionally equivalent. >>>> >>>> We need to make that distinction, between humans and their language, or >>>> we >>>> will be caught inside a circular definition with no way to break out or >>>> to >>>> prove anything, not about others and not about ourselves! >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>> >>>> >>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>> >>>>> Aya, I think two different things are getting a bit mixed up here. >>>>> >>>>> I don't for a moment dispute that expressions are often susceptible to >>>>> multiple interpretations, that these interpretations are guided by all >>>>> kinds of contextual information, or that different people, or even the >>>>> same person at different times, may end up selecting differently from >>>>> the >>>>> various interpretations. Your example 'Open happiness' in another >>>>> contribution to this thread illustrates these points very well. >>>>> >>>>> My point, though, is that each of these different interpretations, as >>>>> well >>>>> as the selection of one (or more) as more likely in a particular >>>>> context, >>>>> is achieved through mechanisms in the interpreter's brain which >>>>> evolved in >>>>> the course of the phylogenetic development of language in the human >>>>> species, and developed ontogenetically in that particular >>>>> interpreter's >>>>> brain. It is surely likely that those mechanisms are highly similar in >>>>> different human beings, even though there may be differences in the >>>>> detailed wiring in different brains. What I'm saying is that in order >>>>> to >>>>> answer the question 'How do we communicate using language?' or, if you >>>>> prefer, 'How does the language user work?', we need to investigate >>>>> what >>>>> those mechanisms are, and this is what psycholinguists can help us >>>>> with. >>>>> >>>>> In particular, as linguists, we are interested (well, some of us are, >>>>> though clearly not all) in whether the constructs we posit in our >>>>> theories >>>>> of language have psychological validity in the sense that they >>>>> correspond >>>>> to ways in which aspects of language are represented in the brain. [As >>>>> an >>>>> aside, I do realise that there are linguists who strenuously resist >>>>> what >>>>> they see as a misguided emphasis on mental representation in the work >>>>> of >>>>> cognitive scientists, but it seems clear that language must be >>>>> represented >>>>> in some way in the brain in order that we can engage in the >>>>> sociosemiotic >>>>> acts of meaning making which are the primary focus for many of these >>>>> critics.] Taking your 'Open happiness' examples again, I think we need >>>>> answers to questions such as: What kind of representation does the >>>>> human >>>>> language processing system have for 'open' and for 'happiness'? Are >>>>> the >>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>> distinguish >>>>> the semantic from the pragmatic) properties of these items (and we >>>>> might >>>>> want to add 'for this particular interpreter', though there must be >>>>> similarities across interpreters for communication to be possible) >>>>> represented in the same or different ways, in the same or different >>>>> locations (or sets of distributed locations)? Indeed, are we right in >>>>> thinking that these familiar levels of linguistic description must be >>>>> differentiated, as such, in the human language processing system? Does >>>>> the >>>>> representation for 'open' distinguish between what we call verbal and >>>>> adjectival uses of this item, and if so how? Or are syntactic analyses >>>>> computed on the fly, using semantic and contextual clues, rather than >>>>> the >>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>> adjectives? >>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>> both? >>>>> All these questions, and many many more, are relevant to the >>>>> construction >>>>> of a model of language which reflects how language users communicate >>>>> (as, >>>>> of course, are a whole set of other questions about the sociocultural >>>>> aspects of communication). I am not a psycholinguist, but my all too >>>>> superficial reading in the area suggests that psycholinguists have >>>>> gone >>>>> some of the way towards answering some of the questions we might want >>>>> to >>>>> ask, but that there is still a long way to go. >>>>> >>>>> Chris >>>>> -------------------------------------------------- >>>>> From: "A. Katz" >>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>> To: "Chris Butler" >>>>> Cc: "FUNKNET" >>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>> >>>>>> The problem is that once we achieve the psycholinguistic goal and see >>>>>> what >>>>>> is happening in each person's head, and we see that conflicting >>>>>> analyses >>>>>> are the norm, rather than the exception, among normal language >>>>>> speakers, >>>>>> then we will realize that the way language works to transmit >>>>>> information >>>>>> is despite individual differences and not because of uniform >>>>>> processing >>>>>> strategies. >>>>>> >>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>> not >>>>>> process it the same way. Which means that processing is seocndary to >>>>>> information transmission. >>>>>> >>>>>> --Aya >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>> >>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>> question of what kind of thing we think language is: social or >>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>> those >>>>>>> of us who are committed functionalists. My own view is that a truly >>>>>>> functional model of language would be one which aims to account for >>>>>>> how >>>>>>> human beings communicate using language, or in other words tries to >>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>> now, >>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>> natural >>>>>>> language user work?' In trying to answer this question we need to >>>>>>> accept >>>>>>> that language is BOTH social AND individual, and we need to explore >>>>>>> both >>>>>>> aspects to get as complete a picture as possible of how we >>>>>>> communicate >>>>>>> using language. We need to know BOTH how people create and respond >>>>>>> to >>>>>>> meanings and express those meanings in forms during social >>>>>>> interaction >>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>> order >>>>>>> to make such interaction possible. Both are important parts of the >>>>>>> answer to the question 'How do we communicate using language?', >>>>>>> though >>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>> more >>>>>>> on the second aspect, and so will I. >>>>>>> >>>>>>> This doesn't mean that all the work linguists have done on >>>>>>> "exploring >>>>>>> the structure of a language so that I can understand how all the >>>>>>> bits >>>>>>> fit together" and "exploring the connections between items", as Dick >>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>> like. >>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>> evidence >>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>> least >>>>>>> consistent with what we know of the processes which go on when we >>>>>>> use >>>>>>> language. So I am with Matthew when he says that for him, "the only >>>>>>> sense in which an analysis can be "the correct analysis" is in terms >>>>>>> of >>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>> that >>>>>>> linguists should just give up their jobs until such time as we know >>>>>>> everything there is to know about language processing. But it does >>>>>>> mean >>>>>>> that we need to collaborate with psycholinguists, psychologists and >>>>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>>>> collaborate much more with sociolinguists and sociologists, so that >>>>>>> we >>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>> part, >>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>> experiments >>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>> criticism >>>>>>> that what happens in artifical lab situations may not happen in >>>>>>> natural >>>>>>> communicative conditions. >>>>>>> >>>>>>> I also agree with Dick when he says that "the differences between >>>>>>> individuals really matter", and with Lise when she points out that >>>>>>> "we >>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>> if it >>>>>>> were a single coherent construct that we are trying to discover". >>>>>>> However, there are surely processing mechanisms which are common to >>>>>>> all >>>>>>> language users by virtue of the evolution of the language faculty >>>>>>> and >>>>>>> which constitute the "general processes" which Dick says >>>>>>> psycholinguists >>>>>>> are interested in. >>>>>>> >>>>>>> On the issue of quantitative methodology, I'm sympathetic in general >>>>>>> to >>>>>>> Ted and Ev's views, though it does seem sensible to prioritise cases >>>>>>> in >>>>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>>>> that we should be giving our university students of linguistics (and >>>>>>> some of our linguistics lecturers!) courses in quantitative aspects >>>>>>> of >>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>> basic >>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>> going >>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>> done >>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>> offer >>>>>>> my own experience, over quite a long period, of teaching such >>>>>>> courses to >>>>>>> people with little or no prior experience in quantitative >>>>>>> techniques. >>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>> linguistics >>>>>>> students in an institution where we had many mature students who had >>>>>>> come into university level studies with non-standard qualifications, >>>>>>> and >>>>>>> were not well equipped for courses of this kind by their previous >>>>>>> experience. I'm glad to say that teaching the subject from their own >>>>>>> perspective as language students rather than that of the >>>>>>> statistician, >>>>>>> and explaining the reasons for doing things in particular ways >>>>>>> rather >>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>> students were able to appreciate the relevance of these courses and >>>>>>> to >>>>>>> turn in very creditable projects showing an understanding of >>>>>>> research >>>>>>> design and competence in the use of a range of basic statistical >>>>>>> techniques. And I still find that bright graduate students respond >>>>>>> well >>>>>>> to similar courses which incorporate some of the rather more >>>>>>> advanced >>>>>>> techniques needed for many real research projects in various areas >>>>>>> of >>>>>>> linguistics. But I may well be out of date with what is now already >>>>>>> happening in our fine institutions of higher education! >>>>>>> >>>>>>> Chris Butler >>>>>>> >>>>>>> >>>>> >>>>> >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From wilcox at unm.edu Mon Sep 13 02:11:14 2010 From: wilcox at unm.edu (Sherman Wilcox) Date: Sun, 12 Sep 2010 20:11:14 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: On Sep 12, 2010, at 6:54 PM, "A. Katz" wrote: > Sorry, but "eating" is not a tool. It's a biological process. So is language. Sherman Wilcox University of New Mexico From cbutler at ntlworld.com Mon Sep 13 09:50:07 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Mon, 13 Sep 2010 10:50:07 +0100 Subject: FUNKNET] analysis: unhappiness Message-ID: I sympathise with Tom's view that the question 'How does language work?' may not be meaningful, or may at least not be an appropriate question to ask, independently of the question 'How do language users work?' The complex human activity that we call language exists only through the capabilities and activities of its users, pace Aya's comment that "we do not embody it". Furthermore, there is a considerable body of opinion nowadays, not only among linguists interested in sociosemiotic aspects of language but also in the work of many cognitive linguists and of psycholinguists such as Ray Gibbs, which agrees with Mark's suspicion of "any proposal to study something as complex as language separately from its embodiment". One problem with a strict separation between language and humans, such as Aya would like us to accept, is that if we do this we deny ourselves natural explanations of many phenomena in human language which are otherwise somewhat mysterious. As Tom astutely observed more than 30 years ago, many structural properties of language are most insightfully accounted for in terms of a set of explanatory parameters many of which are concerned with the properties of the human language-processing organism, including general cognitive and perceptual structure, the specific structure of the neurological, acoustic, articulatory and other mechanisms, the ontogenetic development of the young child, etc. Aya is also concerned about the need to distinguish language processing from language data. But those of us who want to ask the question 'How does the language user work?' do make this distinction. The data are the input to language understanding and the output from language production, so the two are distinct. I understand Aya's point that we need some way of evaluating what computers and non-humans can do. But I think it is dangerous to put this in terms of "artifical language-using devices" and "to what extent a non-human has acquired language", as if there were some monolithic entity out there called language, which humans have and computers and non-humans may have to some extent. A more fruitful way to approach the situation, in my view, would be to study how and why what can be produced (and also understood) by a non-human animal or a computer is similar to and different from what a human being can produce and understand, looking at the systems in the round, including analyses not only of structures but also of the range of communicative and social functions they perform, but without assuming that what we are looking at is a single entity 'language', present to variable extents. Sorry if I haven't put this last bit clearly enough - I found it quite hard to formulate without making too many unwarranted assumptions! Chris From kuzar at research.haifa.ac.il Mon Sep 13 10:01:54 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Mon, 13 Sep 2010 12:01:54 +0200 Subject: Methodological MA workshop Message-ID: Posted to Funknet and Cogling. -------- Dear fellow linguists, I have been assigned to teach a methodological MA workshop in linguistics (officially titled "Approaches to research in linguistics", however it is in the Department of English, so mainly English linguistics). Presumably, students have taken "Style & composition" during their BA, where they have learned the basics of academic writing. It is a two semester course meeting once a week for 2 academic hours, so there will be some 26-28 classes. The seminar is intended to help students at the MA level in several areas: 1. Choosing a research domain, formulating a research question, choosing the appropriate methodology. 2. The different parts of a research paper, how to develop an argument (probably the hardest of all tasks), how to write the introduction. What's the difference between presenting the findings/results and discussing/interpreting them? 3. Corpora: what English corpora are are available, search strategy and search tools, regular expressions, organization and manipulation of data (e.g. in Excel). 4. The LLBA as a resource (believe it or not: some students have made it to the MA level without knowing about the LLBA). 5. Organizing bibliographical data and note-taking files (Refworks, NotaBene). 6. Preparing students for linguistic presentations by reading and discussing an article by that linguist beforehand. If any of you have taught a similar course, I'd be very grateful to hear from you. Are there any other points that should be covered? Do you have any practical suggestions for points 1. and 2. or any literature that might be consulted? As for point 3: does anybody know of a user friendly introduction (or crash-course) to regular expressions at a rudimentary level for non-computationally oriented students? Looking forward to your advice. Ron Kuzar =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From dan at daneverett.org Mon Sep 13 12:45:43 2010 From: dan at daneverett.org (Daniel Everett) Date: Mon, 13 Sep 2010 08:45:43 -0400 Subject: Language as a Tool Message-ID: Dear Mark, These are all excellent points. Clearly there is some biology that must underwrite language, or plants could speak. The question is, how much. Less than eating. More than wearing a tie, perhaps. I think that all of your points, however, are compatible with the idea that language is a tool, so long as we don't think that, as you say, this explains everything. It does, however, explain more than many have recognized. My new book on this, Cognitive Fire: Language as a Cultural Tool, is due out from Pantheon (US) and Profile (UK) sometime in 2011. Hopefully, I will have answered your questions. -- Dan On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: > Aya -- > > You said: "Other tools can be studied separately from the people who use > them or created them. Why not language?" > > Although language can certainly be considered a tool, I think it's unlike > other tools in several very significant respects. > > 1. Although we're not born with language, we can't avoid (pathologies > excluded) acquiring it unless we're not exposed to it. To that extent, > language is a biological phenomenon. A prototypical tool is not a > biological phenomenon, so I'm not sure how valid any conclusions might be > that are drawn from a premise of language-as-tool. > > 2. A tool is as a tool does: Anything is a tool that is being used as a > tool, including dead wombats, broken screwdrivers or decks of playing > cards. (Completing the imagined scenarios is left as an exercise for the > reader...) So saying that language is a tool is only saying that language > is used as a tool. Quite a few conclusions can be and have been drawn from > the fact that language is used as a tool, but I would have to be convinced > in detail that almost everything worth knowing about language is dependent > on the premise of language-as-tool. > > 3. If language is a "tool" for (say) communicating ideas, then eating is a > "tool" for reducing hunger. In both cases, I worry about the tool metaphor > being stretched so far from the prototype that we're left with an > out-and-out category fallacy. > > 4. More prototypical tools can be studied separately from those who use or > create them because those tools are easily observed separately from those > who use or create them. I don't think the same thing can be said of > language -- very little about language can be observed apart from its use, > so very little about language can be observed apart from its user(s). > > 5. Any proposal to study something as complex as language separately from > its embodiment is suspicious to me, smacking of reductionism -- something > up with which linguistics has had to put a tad much. Anything that puts > language back into its human context would be a step forward. > > > -- Mark > > Mark P. Line > > From mark at polymathix.com Mon Sep 13 12:46:45 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 07:46:45 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya -- A. Katz wrote: > Mark, > > Sorry, but "eating" is not a tool. It's a biological process. > > Eating cannot occur outside the biological context. A human being can > avoid eating, but if so he starves. Feral children do not speak, but > they eat, like everyone else, or they die. Eating does not have to be > taught, there is no critical age of acquisition and it is not uniquely an > artefact of human culture. If you are an animal, you eat. Eating cannot > survive the death of the eater. > > Language can. I would say that not language, but the artifacts of language (texts, audio recordings) can survive the people who created them, because I try very hard not to reify the artifacts of language as "language". The (usually communicative) process I refer to as language cannot exist independently of its embodiment. That said, I don't care if the embodiment is human, computer, cetacean or non-human primate. -- Mark Mark P. Line > On Sun, 12 Sep 2010, Mark P. Line wrote: > >> Aya -- >> >> You said: "Other tools can be studied separately from the people who use >> them or created them. Why not language?" >> >> Although language can certainly be considered a tool, I think it's >> unlike >> other tools in several very significant respects. >> >> 1. Although we're not born with language, we can't avoid (pathologies >> excluded) acquiring it unless we're not exposed to it. To that extent, >> language is a biological phenomenon. A prototypical tool is not a >> biological phenomenon, so I'm not sure how valid any conclusions might >> be >> that are drawn from a premise of language-as-tool. >> >> 2. A tool is as a tool does: Anything is a tool that is being used as a >> tool, including dead wombats, broken screwdrivers or decks of playing >> cards. (Completing the imagined scenarios is left as an exercise for the >> reader...) So saying that language is a tool is only saying that >> language >> is used as a tool. Quite a few conclusions can be and have been drawn >> from >> the fact that language is used as a tool, but I would have to be >> convinced >> in detail that almost everything worth knowing about language is >> dependent >> on the premise of language-as-tool. >> >> 3. If language is a "tool" for (say) communicating ideas, then eating is >> a >> "tool" for reducing hunger. In both cases, I worry about the tool >> metaphor >> being stretched so far from the prototype that we're left with an >> out-and-out category fallacy. >> >> 4. More prototypical tools can be studied separately from those who use >> or >> create them because those tools are easily observed separately from >> those >> who use or create them. I don't think the same thing can be said of >> language -- very little about language can be observed apart from its >> use, >> so very little about language can be observed apart from its user(s). >> >> 5. Any proposal to study something as complex as language separately >> from >> its embodiment is suspicious to me, smacking of reductionism -- >> something >> up with which linguistics has had to put a tad much. Anything that puts >> language back into its human context would be a step forward. >> >> >> -- Mark >> >> Mark P. Line >> >> >> >> A. Katz wrote: >>> Tom, >>> >>> I understand the uncomfortable association with Chomsky that speaking >>> of >>> language apart from people can have. Competence versus performance, the >>> way Chomsky used those terms, never made sense. But that's precisely >>> because >>> he associated "competence" with the brain and suggested that it was >>> hard >>> wired there -- when there was never any evidence of that. >>> >>> However, if we don't distinguish language from humans, and language >>> processing from language data, then how are we going to judge >>> artificial >>> language-using devices as to their efficacy at producing and >>> interpreting >>> language? How are we going to determine whether and to what extent a >>> non-human has acquired language? >>> >>> We aren't born with it. We don't embody it. It's a tool that we use to >>> communicate. Other tools can be studied separately from the people who >>> use >>> them or created them. Why not language? >>> >>> --Aya >>> >>> >>> >>> >>> On Sun, 12 Sep 2010, Tom Givon wrote: >>> >>>> >>>> >>>> >>>> I wonder whether asking "how does language work?" is really a >>>> meaningful >>>> question without asking "how does the language user work?" One of >>>> the >>>> worst >>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>> distinction >>>> of competence ("language", "knowledge") vs. performance ("language >>>> user", >>>> "processing"). It purported to limit linguists to the armchair methods >>>> that >>>> study competence, and relegated to psychology the quantified, >>>> distributional/variationist methods that study behavior, processing >>>> and >>>> on-line communication. The first breach in this artificial >>>> methodological >>>> wall occurred, leastwise for some of us, when we discovered the >>>> intermediate >>>> method of quantified studies of text, interaction, and conversation. >>>> As >>>> an >>>> ex-biologist, I am forever puzzled by the methodological purism we >>>> sill >>>> seem >>>> to embrace in linguistics, in the face of the manifest complexity and >>>> connectivity of language (mind, brain, culture, sociality, evolution, >>>> etc.). >>>> In biology, another extremely complex science with multiple >>>> connections >>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>> evolution, >>>> etc.), ANY method is welcome so long as it does the job of furthering >>>> our >>>> understanding. And by understanding we mean ever-wider circles of >>>> connectivity. >>>> >>>> Best, TG >>>> ================ >>>> >>>> >>>> A. Katz wrote: >>>>> Chris, >>>>> >>>>> Thanks for addressing this question. I understand that many, many >>>>> linguists >>>>> are quite properly and approriately interested in this ultimate >>>>> question: >>>>> "How does the language user work?" (I am also intetested in this >>>>> question >>>>> some of the time.) >>>>> >>>>> My point is that "how does language work?" is also a valid question, >>>>> and a >>>>> central one to the field of linguistics. These two questions are not >>>>> at >>>>> all >>>>> the same. >>>>> >>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>> the >>>>> "language". There are many good reasons to do so. For anyone working >>>>> in >>>>> computerized language processing or in non-human language studies, >>>>> this >>>>> is >>>>> a significant point. >>>>> >>>>> It does not matter if a computerized language processing system even >>>>> remotely simulates what humans do with language in their brains. It >>>>> does >>>>> matter whether it comes up with comparable or indistinguishable >>>>> results. >>>>> >>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>> doing >>>>> the >>>>> same things inside the same modules in his brain as a human does. It >>>>> does >>>>> matter if the results are functionally equivalent. >>>>> >>>>> We need to make that distinction, between humans and their language, >>>>> or >>>>> we >>>>> will be caught inside a circular definition with no way to break out >>>>> or >>>>> to >>>>> prove anything, not about others and not about ourselves! >>>>> >>>>> --Aya >>>>> >>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>> >>>>> >>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>> >>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>> >>>>>> I don't for a moment dispute that expressions are often susceptible >>>>>> to >>>>>> multiple interpretations, that these interpretations are guided by >>>>>> all >>>>>> kinds of contextual information, or that different people, or even >>>>>> the >>>>>> same person at different times, may end up selecting differently >>>>>> from >>>>>> the >>>>>> various interpretations. Your example 'Open happiness' in another >>>>>> contribution to this thread illustrates these points very well. >>>>>> >>>>>> My point, though, is that each of these different interpretations, >>>>>> as >>>>>> well >>>>>> as the selection of one (or more) as more likely in a particular >>>>>> context, >>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>> evolved in >>>>>> the course of the phylogenetic development of language in the human >>>>>> species, and developed ontogenetically in that particular >>>>>> interpreter's >>>>>> brain. It is surely likely that those mechanisms are highly similar >>>>>> in >>>>>> different human beings, even though there may be differences in the >>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>> order >>>>>> to >>>>>> answer the question 'How do we communicate using language?' or, if >>>>>> you >>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>> what >>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>> with. >>>>>> >>>>>> In particular, as linguists, we are interested (well, some of us >>>>>> are, >>>>>> though clearly not all) in whether the constructs we posit in our >>>>>> theories >>>>>> of language have psychological validity in the sense that they >>>>>> correspond >>>>>> to ways in which aspects of language are represented in the brain. >>>>>> [As >>>>>> an >>>>>> aside, I do realise that there are linguists who strenuously resist >>>>>> what >>>>>> they see as a misguided emphasis on mental representation in the >>>>>> work >>>>>> of >>>>>> cognitive scientists, but it seems clear that language must be >>>>>> represented >>>>>> in some way in the brain in order that we can engage in the >>>>>> sociosemiotic >>>>>> acts of meaning making which are the primary focus for many of these >>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>> need >>>>>> answers to questions such as: What kind of representation does the >>>>>> human >>>>>> language processing system have for 'open' and for 'happiness'? Are >>>>>> the >>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>> distinguish >>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>> might >>>>>> want to add 'for this particular interpreter', though there must be >>>>>> similarities across interpreters for communication to be possible) >>>>>> represented in the same or different ways, in the same or different >>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>> in >>>>>> thinking that these familiar levels of linguistic description must >>>>>> be >>>>>> differentiated, as such, in the human language processing system? >>>>>> Does >>>>>> the >>>>>> representation for 'open' distinguish between what we call verbal >>>>>> and >>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>> analyses >>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>> than >>>>>> the >>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>> adjectives? >>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>>> both? >>>>>> All these questions, and many many more, are relevant to the >>>>>> construction >>>>>> of a model of language which reflects how language users communicate >>>>>> (as, >>>>>> of course, are a whole set of other questions about the >>>>>> sociocultural >>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>> too >>>>>> superficial reading in the area suggests that psycholinguists have >>>>>> gone >>>>>> some of the way towards answering some of the questions we might >>>>>> want >>>>>> to >>>>>> ask, but that there is still a long way to go. >>>>>> >>>>>> Chris >>>>>> -------------------------------------------------- >>>>>> From: "A. Katz" >>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>> To: "Chris Butler" >>>>>> Cc: "FUNKNET" >>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>> >>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>> see >>>>>>> what >>>>>>> is happening in each person's head, and we see that conflicting >>>>>>> analyses >>>>>>> are the norm, rather than the exception, among normal language >>>>>>> speakers, >>>>>>> then we will realize that the way language works to transmit >>>>>>> information >>>>>>> is despite individual differences and not because of uniform >>>>>>> processing >>>>>>> strategies. >>>>>>> >>>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>>> not >>>>>>> process it the same way. Which means that processing is seocndary >>>>>>> to >>>>>>> information transmission. >>>>>>> >>>>>>> --Aya >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>> >>>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>>> question of what kind of thing we think language is: social or >>>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>>> those >>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>> truly >>>>>>>> functional model of language would be one which aims to account >>>>>>>> for >>>>>>>> how >>>>>>>> human beings communicate using language, or in other words tries >>>>>>>> to >>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>> now, >>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>> natural >>>>>>>> language user work?' In trying to answer this question we need to >>>>>>>> accept >>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>> explore >>>>>>>> both >>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>> communicate >>>>>>>> using language. We need to know BOTH how people create and respond >>>>>>>> to >>>>>>>> meanings and express those meanings in forms during social >>>>>>>> interaction >>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>> order >>>>>>>> to make such interaction possible. Both are important parts of the >>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>> though >>>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>>> more >>>>>>>> on the second aspect, and so will I. >>>>>>>> >>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>> "exploring >>>>>>>> the structure of a language so that I can understand how all the >>>>>>>> bits >>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>> Dick >>>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>> like. >>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>> evidence >>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>> least >>>>>>>> consistent with what we know of the processes which go on when we >>>>>>>> use >>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>> only >>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>> terms >>>>>>>> of >>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>> that >>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>> know >>>>>>>> everything there is to know about language processing. But it does >>>>>>>> mean >>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>> and >>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>> Ray >>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need >>>>>>>> to >>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>> that >>>>>>>> we >>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>> part, >>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>> experiments >>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>> criticism >>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>> natural >>>>>>>> communicative conditions. >>>>>>>> >>>>>>>> I also agree with Dick when he says that "the differences between >>>>>>>> individuals really matter", and with Lise when she points out that >>>>>>>> "we >>>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>>> if it >>>>>>>> were a single coherent construct that we are trying to discover". >>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>> to >>>>>>>> all >>>>>>>> language users by virtue of the evolution of the language faculty >>>>>>>> and >>>>>>>> which constitute the "general processes" which Dick says >>>>>>>> psycholinguists >>>>>>>> are interested in. >>>>>>>> >>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>> general >>>>>>>> to >>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>> cases >>>>>>>> in >>>>>>>> terms of a hierarchy such as Brian suggests. One thing this means >>>>>>>> is >>>>>>>> that we should be giving our university students of linguistics >>>>>>>> (and >>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>> aspects >>>>>>>> of >>>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>>> basic >>>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>>> going >>>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>>> done >>>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>>> offer >>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>> courses to >>>>>>>> people with little or no prior experience in quantitative >>>>>>>> techniques. >>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>> linguistics >>>>>>>> students in an institution where we had many mature students who >>>>>>>> had >>>>>>>> come into university level studies with non-standard >>>>>>>> qualifications, >>>>>>>> and >>>>>>>> were not well equipped for courses of this kind by their previous >>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>> own >>>>>>>> perspective as language students rather than that of the >>>>>>>> statistician, >>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>> rather >>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>> and >>>>>>>> to >>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>> research >>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>> techniques. And I still find that bright graduate students respond >>>>>>>> well >>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>> advanced >>>>>>>> techniques needed for many real research projects in various areas >>>>>>>> of >>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>> already >>>>>>>> happening in our fine institutions of higher education! >>>>>>>> >>>>>>>> Chris Butler >>>>>>>> >>>>>>>> >>>>>> >>>>>> >>>> >>>> >>> >>> >> >> >> -- Mark >> >> Mark P. Line >> Bartlesville, OK >> >> > > -- Mark Mark P. Line Bartlesville, OK From mark at polymathix.com Mon Sep 13 12:59:47 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 07:59:47 -0500 Subject: Language as a Tool In-Reply-To: Message-ID: Thanks, Dan. I first got into linguistics through anthropology and never really lost that frame of reference, so I'm very much looking forward to your new book. Please rattle Pantheon's cage to get them to offer a Kindle edition of your book. :) -- Mark Mark P. Line Daniel Everett wrote: > > Dear Mark, > > These are all excellent points. Clearly there is some biology that must > underwrite language, or plants could speak. The question is, how much. > Less than eating. More than wearing a tie, perhaps. > > I think that all of your points, however, are compatible with the idea > that language is a tool, so long as we don't think that, as you say, this > explains everything. It does, however, explain more than many have > recognized. > > My new book on this, Cognitive Fire: Language as a Cultural Tool, is due > out from Pantheon (US) and Profile (UK) sometime in 2011. > > Hopefully, I will have answered your questions. > > -- Dan > > > On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: > >> Aya -- >> >> You said: "Other tools can be studied separately from the people who use >> them or created them. Why not language?" >> >> Although language can certainly be considered a tool, I think it's >> unlike >> other tools in several very significant respects. >> >> 1. Although we're not born with language, we can't avoid (pathologies >> excluded) acquiring it unless we're not exposed to it. To that extent, >> language is a biological phenomenon. A prototypical tool is not a >> biological phenomenon, so I'm not sure how valid any conclusions might >> be >> that are drawn from a premise of language-as-tool. >> >> 2. A tool is as a tool does: Anything is a tool that is being used as a >> tool, including dead wombats, broken screwdrivers or decks of playing >> cards. (Completing the imagined scenarios is left as an exercise for the >> reader...) So saying that language is a tool is only saying that >> language >> is used as a tool. Quite a few conclusions can be and have been drawn >> from >> the fact that language is used as a tool, but I would have to be >> convinced >> in detail that almost everything worth knowing about language is >> dependent >> on the premise of language-as-tool. >> >> 3. If language is a "tool" for (say) communicating ideas, then eating is >> a >> "tool" for reducing hunger. In both cases, I worry about the tool >> metaphor >> being stretched so far from the prototype that we're left with an >> out-and-out category fallacy. >> >> 4. More prototypical tools can be studied separately from those who use >> or >> create them because those tools are easily observed separately from >> those >> who use or create them. I don't think the same thing can be said of >> language -- very little about language can be observed apart from its >> use, >> so very little about language can be observed apart from its user(s). >> >> 5. Any proposal to study something as complex as language separately >> from >> its embodiment is suspicious to me, smacking of reductionism -- >> something >> up with which linguistics has had to put a tad much. Anything that puts >> language back into its human context would be a step forward. >> >> >> -- Mark >> >> Mark P. Line >> >> > > -- Mark Mark P. Line Bartlesville, OK From dlevere at ilstu.edu Mon Sep 13 13:07:38 2010 From: dlevere at ilstu.edu (Daniel Everett) Date: Mon, 13 Sep 2010 09:07:38 -0400 Subject: Language as a Tool In-Reply-To: Message-ID: Mark, Well, ahem, there *is* a kindle edition of Don't sleep there are snakes. There should also be one of Cognitive Fire. Dan http://www.amazon.com/Dont-Sleep-There-Snakes-ebook/dp/B0037Z8SMC/ref=tmm_kin_title_0?ie=UTF8&m=AG56TWVU5XWC2 On Sep 13, 2010, at 8:59 AM, Mark P. Line wrote: > Thanks, Dan. I first got into linguistics through anthropology and never > really lost that frame of reference, so I'm very much looking forward to > your new book. > > Please rattle Pantheon's cage to get them to offer a Kindle edition of > your book. :) > > -- Mark > > Mark P. Line > > > > Daniel Everett wrote: >> >> Dear Mark, >> >> These are all excellent points. Clearly there is some biology that must >> underwrite language, or plants could speak. The question is, how much. >> Less than eating. More than wearing a tie, perhaps. >> >> I think that all of your points, however, are compatible with the idea >> that language is a tool, so long as we don't think that, as you say, this >> explains everything. It does, however, explain more than many have >> recognized. >> >> My new book on this, Cognitive Fire: Language as a Cultural Tool, is due >> out from Pantheon (US) and Profile (UK) sometime in 2011. >> >> Hopefully, I will have answered your questions. >> >> -- Dan >> >> >> On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > From amnfn at well.com Mon Sep 13 13:17:28 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 06:17:28 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <149a0b7cd376d6c0f392a7d827ef7985.squirrel@sm.webmail.pair.com> Message-ID: Mark, Okay, I'm glad that you agree that a language user might be a machine or a non-human. That brings us much closer together. Nevertheless, there is also the fact that language can be preserved, but not used, for thousands of years, and then people start using it again. If it were an biological process, or even something like a live virus passed down from parent to child, then reviving it once it was dead would be impossible. But it's not. It's a tool. A tool that a group of people can stop using and making, but keep the blue prints for, and then make it again when they choose to. I wonder how many linguists realize that Hebrew was revived from a dead language. How many think it is just a slavic language with Hebrew lexemes? How many realize that the grammar, the triliteral roots, the derivational system, are all working, and that any changes in pronunciation are of little importance when the basic contrasts are preserved? Mark, if you can see that a language can be used by a non-human, then why not also acknowledge that it's a system for transferring information, and that the code can be preserved while in disuse? --Aya On Mon, 13 Sep 2010, Mark P. Line wrote: > Aya -- > > A. Katz wrote: >> Mark, >> >> Sorry, but "eating" is not a tool. It's a biological process. >> >> Eating cannot occur outside the biological context. A human being can >> avoid eating, but if so he starves. Feral children do not speak, but >> they eat, like everyone else, or they die. Eating does not have to be >> taught, there is no critical age of acquisition and it is not uniquely an >> artefact of human culture. If you are an animal, you eat. Eating cannot >> survive the death of the eater. >> >> Language can. > > I would say that not language, but the artifacts of language (texts, audio > recordings) can survive the people who created them, because I try very > hard not to reify the artifacts of language as "language". > > The (usually communicative) process I refer to as language cannot exist > independently of its embodiment. That said, I don't care if the embodiment > is human, computer, cetacean or non-human primate. > > > -- Mark > > Mark P. Line > > >> On Sun, 12 Sep 2010, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >>> >>> A. Katz wrote: >>>> Tom, >>>> >>>> I understand the uncomfortable association with Chomsky that speaking >>>> of >>>> language apart from people can have. Competence versus performance, the >>>> way Chomsky used those terms, never made sense. But that's precisely >>>> because >>>> he associated "competence" with the brain and suggested that it was >>>> hard >>>> wired there -- when there was never any evidence of that. >>>> >>>> However, if we don't distinguish language from humans, and language >>>> processing from language data, then how are we going to judge >>>> artificial >>>> language-using devices as to their efficacy at producing and >>>> interpreting >>>> language? How are we going to determine whether and to what extent a >>>> non-human has acquired language? >>>> >>>> We aren't born with it. We don't embody it. It's a tool that we use to >>>> communicate. Other tools can be studied separately from the people who >>>> use >>>> them or created them. Why not language? >>>> >>>> --Aya >>>> >>>> >>>> >>>> >>>> On Sun, 12 Sep 2010, Tom Givon wrote: >>>> >>>>> >>>>> >>>>> >>>>> I wonder whether asking "how does language work?" is really a >>>>> meaningful >>>>> question without asking "how does the language user work?" One of >>>>> the >>>>> worst >>>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>>> distinction >>>>> of competence ("language", "knowledge") vs. performance ("language >>>>> user", >>>>> "processing"). It purported to limit linguists to the armchair methods >>>>> that >>>>> study competence, and relegated to psychology the quantified, >>>>> distributional/variationist methods that study behavior, processing >>>>> and >>>>> on-line communication. The first breach in this artificial >>>>> methodological >>>>> wall occurred, leastwise for some of us, when we discovered the >>>>> intermediate >>>>> method of quantified studies of text, interaction, and conversation. >>>>> As >>>>> an >>>>> ex-biologist, I am forever puzzled by the methodological purism we >>>>> sill >>>>> seem >>>>> to embrace in linguistics, in the face of the manifest complexity and >>>>> connectivity of language (mind, brain, culture, sociality, evolution, >>>>> etc.). >>>>> In biology, another extremely complex science with multiple >>>>> connections >>>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>>> evolution, >>>>> etc.), ANY method is welcome so long as it does the job of furthering >>>>> our >>>>> understanding. And by understanding we mean ever-wider circles of >>>>> connectivity. >>>>> >>>>> Best, TG >>>>> ================ >>>>> >>>>> >>>>> A. Katz wrote: >>>>>> Chris, >>>>>> >>>>>> Thanks for addressing this question. I understand that many, many >>>>>> linguists >>>>>> are quite properly and approriately interested in this ultimate >>>>>> question: >>>>>> "How does the language user work?" (I am also intetested in this >>>>>> question >>>>>> some of the time.) >>>>>> >>>>>> My point is that "how does language work?" is also a valid question, >>>>>> and a >>>>>> central one to the field of linguistics. These two questions are not >>>>>> at >>>>>> all >>>>>> the same. >>>>>> >>>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>>> the >>>>>> "language". There are many good reasons to do so. For anyone working >>>>>> in >>>>>> computerized language processing or in non-human language studies, >>>>>> this >>>>>> is >>>>>> a significant point. >>>>>> >>>>>> It does not matter if a computerized language processing system even >>>>>> remotely simulates what humans do with language in their brains. It >>>>>> does >>>>>> matter whether it comes up with comparable or indistinguishable >>>>>> results. >>>>>> >>>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>>> doing >>>>>> the >>>>>> same things inside the same modules in his brain as a human does. It >>>>>> does >>>>>> matter if the results are functionally equivalent. >>>>>> >>>>>> We need to make that distinction, between humans and their language, >>>>>> or >>>>>> we >>>>>> will be caught inside a circular definition with no way to break out >>>>>> or >>>>>> to >>>>>> prove anything, not about others and not about ourselves! >>>>>> >>>>>> --Aya >>>>>> >>>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>>> >>>>>> >>>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>>> >>>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>>> >>>>>>> I don't for a moment dispute that expressions are often susceptible >>>>>>> to >>>>>>> multiple interpretations, that these interpretations are guided by >>>>>>> all >>>>>>> kinds of contextual information, or that different people, or even >>>>>>> the >>>>>>> same person at different times, may end up selecting differently >>>>>>> from >>>>>>> the >>>>>>> various interpretations. Your example 'Open happiness' in another >>>>>>> contribution to this thread illustrates these points very well. >>>>>>> >>>>>>> My point, though, is that each of these different interpretations, >>>>>>> as >>>>>>> well >>>>>>> as the selection of one (or more) as more likely in a particular >>>>>>> context, >>>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>>> evolved in >>>>>>> the course of the phylogenetic development of language in the human >>>>>>> species, and developed ontogenetically in that particular >>>>>>> interpreter's >>>>>>> brain. It is surely likely that those mechanisms are highly similar >>>>>>> in >>>>>>> different human beings, even though there may be differences in the >>>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>>> order >>>>>>> to >>>>>>> answer the question 'How do we communicate using language?' or, if >>>>>>> you >>>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>>> what >>>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>>> with. >>>>>>> >>>>>>> In particular, as linguists, we are interested (well, some of us >>>>>>> are, >>>>>>> though clearly not all) in whether the constructs we posit in our >>>>>>> theories >>>>>>> of language have psychological validity in the sense that they >>>>>>> correspond >>>>>>> to ways in which aspects of language are represented in the brain. >>>>>>> [As >>>>>>> an >>>>>>> aside, I do realise that there are linguists who strenuously resist >>>>>>> what >>>>>>> they see as a misguided emphasis on mental representation in the >>>>>>> work >>>>>>> of >>>>>>> cognitive scientists, but it seems clear that language must be >>>>>>> represented >>>>>>> in some way in the brain in order that we can engage in the >>>>>>> sociosemiotic >>>>>>> acts of meaning making which are the primary focus for many of these >>>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>>> need >>>>>>> answers to questions such as: What kind of representation does the >>>>>>> human >>>>>>> language processing system have for 'open' and for 'happiness'? Are >>>>>>> the >>>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>>> distinguish >>>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>>> might >>>>>>> want to add 'for this particular interpreter', though there must be >>>>>>> similarities across interpreters for communication to be possible) >>>>>>> represented in the same or different ways, in the same or different >>>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>>> in >>>>>>> thinking that these familiar levels of linguistic description must >>>>>>> be >>>>>>> differentiated, as such, in the human language processing system? >>>>>>> Does >>>>>>> the >>>>>>> representation for 'open' distinguish between what we call verbal >>>>>>> and >>>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>>> analyses >>>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>>> than >>>>>>> the >>>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>>> adjectives? >>>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>>>> both? >>>>>>> All these questions, and many many more, are relevant to the >>>>>>> construction >>>>>>> of a model of language which reflects how language users communicate >>>>>>> (as, >>>>>>> of course, are a whole set of other questions about the >>>>>>> sociocultural >>>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>>> too >>>>>>> superficial reading in the area suggests that psycholinguists have >>>>>>> gone >>>>>>> some of the way towards answering some of the questions we might >>>>>>> want >>>>>>> to >>>>>>> ask, but that there is still a long way to go. >>>>>>> >>>>>>> Chris >>>>>>> -------------------------------------------------- >>>>>>> From: "A. Katz" >>>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>>> To: "Chris Butler" >>>>>>> Cc: "FUNKNET" >>>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>>> >>>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>>> see >>>>>>>> what >>>>>>>> is happening in each person's head, and we see that conflicting >>>>>>>> analyses >>>>>>>> are the norm, rather than the exception, among normal language >>>>>>>> speakers, >>>>>>>> then we will realize that the way language works to transmit >>>>>>>> information >>>>>>>> is despite individual differences and not because of uniform >>>>>>>> processing >>>>>>>> strategies. >>>>>>>> >>>>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>>>> not >>>>>>>> process it the same way. Which means that processing is seocndary >>>>>>>> to >>>>>>>> information transmission. >>>>>>>> >>>>>>>> --Aya >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>>> >>>>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>>>> question of what kind of thing we think language is: social or >>>>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>>>> those >>>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>>> truly >>>>>>>>> functional model of language would be one which aims to account >>>>>>>>> for >>>>>>>>> how >>>>>>>>> human beings communicate using language, or in other words tries >>>>>>>>> to >>>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>>> now, >>>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>>> natural >>>>>>>>> language user work?' In trying to answer this question we need to >>>>>>>>> accept >>>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>>> explore >>>>>>>>> both >>>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>>> communicate >>>>>>>>> using language. We need to know BOTH how people create and respond >>>>>>>>> to >>>>>>>>> meanings and express those meanings in forms during social >>>>>>>>> interaction >>>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>>> order >>>>>>>>> to make such interaction possible. Both are important parts of the >>>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>>> though >>>>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>>>> more >>>>>>>>> on the second aspect, and so will I. >>>>>>>>> >>>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>>> "exploring >>>>>>>>> the structure of a language so that I can understand how all the >>>>>>>>> bits >>>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>>> Dick >>>>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>>> like. >>>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>>> evidence >>>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>>> least >>>>>>>>> consistent with what we know of the processes which go on when we >>>>>>>>> use >>>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>>> only >>>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>>> terms >>>>>>>>> of >>>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>>> that >>>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>>> know >>>>>>>>> everything there is to know about language processing. But it does >>>>>>>>> mean >>>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>>> and >>>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>>> Ray >>>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need >>>>>>>>> to >>>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>>> that >>>>>>>>> we >>>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>>> part, >>>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>>> experiments >>>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>>> criticism >>>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>>> natural >>>>>>>>> communicative conditions. >>>>>>>>> >>>>>>>>> I also agree with Dick when he says that "the differences between >>>>>>>>> individuals really matter", and with Lise when she points out that >>>>>>>>> "we >>>>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>>>> if it >>>>>>>>> were a single coherent construct that we are trying to discover". >>>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>>> to >>>>>>>>> all >>>>>>>>> language users by virtue of the evolution of the language faculty >>>>>>>>> and >>>>>>>>> which constitute the "general processes" which Dick says >>>>>>>>> psycholinguists >>>>>>>>> are interested in. >>>>>>>>> >>>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>>> general >>>>>>>>> to >>>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>>> cases >>>>>>>>> in >>>>>>>>> terms of a hierarchy such as Brian suggests. One thing this means >>>>>>>>> is >>>>>>>>> that we should be giving our university students of linguistics >>>>>>>>> (and >>>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>>> aspects >>>>>>>>> of >>>>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>>>> basic >>>>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>>>> going >>>>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>>>> done >>>>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>>>> offer >>>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>>> courses to >>>>>>>>> people with little or no prior experience in quantitative >>>>>>>>> techniques. >>>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>>> linguistics >>>>>>>>> students in an institution where we had many mature students who >>>>>>>>> had >>>>>>>>> come into university level studies with non-standard >>>>>>>>> qualifications, >>>>>>>>> and >>>>>>>>> were not well equipped for courses of this kind by their previous >>>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>>> own >>>>>>>>> perspective as language students rather than that of the >>>>>>>>> statistician, >>>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>>> rather >>>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>>> and >>>>>>>>> to >>>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>>> research >>>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>>> techniques. And I still find that bright graduate students respond >>>>>>>>> well >>>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>>> advanced >>>>>>>>> techniques needed for many real research projects in various areas >>>>>>>>> of >>>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>>> already >>>>>>>>> happening in our fine institutions of higher education! >>>>>>>>> >>>>>>>>> Chris Butler >>>>>>>>> >>>>>>>>> >>>>>>> >>>>>>> >>>>> >>>>> >>>> >>>> >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> Bartlesville, OK >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From amnfn at well.com Mon Sep 13 13:19:36 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 06:19:36 -0700 Subject: Language as a Tool In-Reply-To: Message-ID: I also very much look forward to reading Daniel Everett's new book! --Aya On Mon, 13 Sep 2010, Mark P. Line wrote: > Thanks, Dan. I first got into linguistics through anthropology and never > really lost that frame of reference, so I'm very much looking forward to > your new book. > > Please rattle Pantheon's cage to get them to offer a Kindle edition of > your book. :) > > -- Mark > > Mark P. Line > > > > Daniel Everett wrote: >> >> Dear Mark, >> >> These are all excellent points. Clearly there is some biology that must >> underwrite language, or plants could speak. The question is, how much. >> Less than eating. More than wearing a tie, perhaps. >> >> I think that all of your points, however, are compatible with the idea >> that language is a tool, so long as we don't think that, as you say, this >> explains everything. It does, however, explain more than many have >> recognized. >> >> My new book on this, Cognitive Fire: Language as a Cultural Tool, is due >> out from Pantheon (US) and Profile (UK) sometime in 2011. >> >> Hopefully, I will have answered your questions. >> >> -- Dan >> >> >> On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From geoffnathan at wayne.edu Mon Sep 13 13:36:40 2010 From: geoffnathan at wayne.edu (Geoff Nathan) Date: Mon, 13 Sep 2010 09:36:40 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: ----- Original Message ----- > From: "A. Katz" > You might as well say that a person cannot possibly avoid watching TV > once > he's exposed to it, as say the same about language. But people can > survive > just fine without television, and unless someone shows them how to > make a > TV set, most people will never figure out how to build one. The same > goes > for language. We're great at using it, not so great at generating it > out > of thin air with no ambient culture. > > --Aya Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage Tübingen 1994 or, for those, who, like me are Germanically-challenged, On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. Geoff Geoffrey S. Nathan Faculty Liaison, C&IT and Professor, Linguistics Program +1 (313) 577-1259 (C&IT) +1 (313) 577-8621 (English/Linguistics) From phonosemantics at earthlink.net Mon Sep 13 13:39:59 2010 From: phonosemantics at earthlink.net (jess tauber) Date: Mon, 13 Sep 2010 09:39:59 -0400 Subject: Language as a Tool Message-ID: Actually I'd say the true prototypical tools are body parts doing particular jobs- nails, teeth, hairs, wings, fins, legs etc. I doubt it escaped our early ancestors that their created or found tools had functions similar or identical to parts they had themselves or saw in other creatures. Even other lineages must have some inkling (as when a jay uses a thorn to prize a grub out of rotten wood, or a sea otter bashes shellfish with stones fished up from the sea floor). Not everything is pure instinct. Sometimes these parts come out- we lose teeth, nails, hair, birds feathers, and so on, and you can also yank them out of corpses, skeletons, etc., and the occasional unwilling live victim. We may find them loose. This sets the stage for alienability, and generalization perhaps? I can get a sharp tool from a sabertooth, or from the living rock if I knock politely. In languages with 'bipartite' structure, effector bodypart and instrument terms are often dealt with identically, and stand in contrast with pathway/position terms, which may have a mirror in the way the brain deals with tools and gait/posture related motion. Jess Tauber phonosemantics at earthlink.net From dan at daneverett.org Mon Sep 13 14:37:11 2010 From: dan at daneverett.org (Daniel Everett) Date: Mon, 13 Sep 2010 10:37:11 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <1143045877.645687.1284385000582.JavaMail.root@starship.merit.edu> Message-ID: Dear Geoff, These are all valid points, but none of them support nativism. One place to start is with Fiona Cowie's relatively recent book, What's Within (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) There are many cultural artifacts and values and types of knowledge that are never taught, but are all acquired just fine, from knowing to sit still in a canoe to how to use a bow and arrow. They are learned out of necessity and by observation, with no explicit instruction. In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev Fedorenko, in Cognition, support was offered for the idea that there are cognitive tools, number and counting being the examples we discussed. The reasoning chain for innatism often goes like this: 1. There is evidence that someone knows something that they were not taught. 2. I cannot think of how they learned it. 3. Therefore, they didn't learn it. 4. Therefore, it is innate. Both 3 and 4 are non-sequiturs. I don't want to get into a big discussion of this here. And I talk about this quite a bit in the book. So I will try to resist the temptation to respond to further postings. So I find your concerns both quite understandable, but not insurmountable or even particularly difficult problems for understanding language as a cognitive tool. -- Dan On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > ----- Original Message ----- >> From: "A. Katz" > >> You might as well say that a person cannot possibly avoid watching TV >> once >> he's exposed to it, as say the same about language. But people can >> survive >> just fine without television, and unless someone shows them how to >> make a >> TV set, most people will never figure out how to build one. The same >> goes >> for language. We're great at using it, not so great at generating it >> out >> of thin air with no ambient culture. >> >> --Aya > > Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. > This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. > All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. > I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. > I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book > > Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage Tübingen 1994 > > or, for those, who, like me are Germanically-challenged, > > On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. > > Geoff > > > Geoffrey S. Nathan > Faculty Liaison, C&IT > and Professor, Linguistics Program > +1 (313) 577-1259 (C&IT) > +1 (313) 577-8621 (English/Linguistics) > From vanvalin at buffalo.edu Mon Sep 13 14:46:06 2010 From: vanvalin at buffalo.edu (Robert Van Valin) Date: Mon, 13 Sep 2010 08:46:06 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <911A69BB-4302-4257-8452-B91B7CF61EB6@daneverett.org> Message-ID: Dan Slobin refers to the argument that Dan outlined as 'the argument from the poverty of the imagination'. Robert Van Valin On Sep 13, 2010, at 8:37 AM, Daniel Everett wrote: > Dear Geoff, > > These are all valid points, but none of them support nativism. > > One place to start is with Fiona Cowie's relatively recent book, What's Within (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) > > There are many cultural artifacts and values and types of knowledge that are never taught, but are all acquired just fine, from knowing to sit still in a canoe to how to use a bow and arrow. They are learned out of necessity and by observation, with no explicit instruction. > > In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev Fedorenko, in Cognition, support was offered for the idea that there are cognitive tools, number and counting being the examples we discussed. > > The reasoning chain for innatism often goes like this: > > 1. There is evidence that someone knows something that they were not taught. > > 2. I cannot think of how they learned it. > > 3. Therefore, they didn't learn it. > > 4. Therefore, it is innate. > > Both 3 and 4 are non-sequiturs. I don't want to get into a big discussion of this here. And I talk about this quite a bit in the book. So I will try to resist the temptation to respond to further postings. > > So I find your concerns both quite understandable, but not insurmountable or even particularly difficult problems for understanding language as a cognitive tool. > > -- Dan > > > > > > > On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > >> ----- Original Message ----- >>> From: "A. Katz" >> >>> You might as well say that a person cannot possibly avoid watching TV >>> once >>> he's exposed to it, as say the same about language. But people can >>> survive >>> just fine without television, and unless someone shows them how to >>> make a >>> TV set, most people will never figure out how to build one. The same >>> goes >>> for language. We're great at using it, not so great at generating it >>> out >>> of thin air with no ambient culture. >>> >>> --Aya >> >> Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. >> This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. >> All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. >> I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. >> I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book >> >> Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage Tübingen 1994 >> >> or, for those, who, like me are Germanically-challenged, >> >> On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. >> >> Geoff >> >> >> Geoffrey S. Nathan >> Faculty Liaison, C&IT >> and Professor, Linguistics Program >> +1 (313) 577-1259 (C&IT) >> +1 (313) 577-8621 (English/Linguistics) >> > > From geoffnathan at wayne.edu Mon Sep 13 15:39:03 2010 From: geoffnathan at wayne.edu (Geoff Nathan) Date: Mon, 13 Sep 2010 11:39:03 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <911A69BB-4302-4257-8452-B91B7CF61EB6@daneverett.org> Message-ID: I did not consider my argument to be one supporting Chomsky's version of nativism, in which linguistic knowledge is independent of (and prior to) experience. Rather I was arguing that the acquisition of language was quite different from the acquisition of a carpenter's skills in that language is not a consciously designed object, unlike what I would think of as a prototypical tool, such as a hammer or a pair of glasses. The structure of language is constrained, not by our genes, but by the 'equipment' that is used to produce, perceive and understand it. I am arguing here for the 'natural' view in 'Natural' Phonology. Stampe and Donegan, in their seminal paper 'The Study of Natural Phonology' argued for a different kind of innateness, in which universals arise from the interaction of the nature of the materials used with the general cognitive processes of the user. This is most clear in phonology, of course, where phonological structures are (partly) determined by the anatomy and physiology of the vocal tract, the physics of sound production and the psychology of perception. We make the sounds that our mouths make it easy to produce clearly (fortitions), and in the way that it's easy to produce them (lenitions), and we hear what other people say by assuming they have the same vocal apparatus and we hear what we would have intended should we have said that. But folks like Karen van Hoek have shown that syntactic island constraints have a similar basis in processing limitations, and others have also found extrinsic causes for other universal constraints on morphology and syntax. So language is certainly learned, but not in the same way that arithmetic is learned--it's more like how riding a bicycle is learned, through interaction with the physical stuff, and additionally, guided by our categorization and perceptual systems. Donegan, Patricia J., and David Stampe, 1979. “The Study of Natural Phonology,” in Current Approaches to Phonological Theory., ed. by Dan Dinnsen. Bloomington: Indiana University Press. Geoffrey S. Nathan Faculty Liaison, C&IT and Professor, Linguistics Program +1 (313) 577-1259 (C&IT) +1 (313) 577-8621 (English/Linguistics) ----- Original Message ----- > From: "Daniel Everett" > To: "Geoff Nathan" > Cc: "Funknet Funknet" > Sent: Monday, September 13, 2010 10:37:11 AM > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > Dear Geoff, > > These are all valid points, but none of them support nativism. > > One place to start is with Fiona Cowie's relatively recent book, > What's Within > (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) > > There are many cultural artifacts and values and types of knowledge > that are never taught, but are all acquired just fine, from knowing to > sit still in a canoe to how to use a bow and arrow. They are learned > out of necessity and by observation, with no explicit instruction. > > In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev > Fedorenko, in Cognition, support was offered for the idea that there > are cognitive tools, number and counting being the examples we > discussed. > > The reasoning chain for innatism often goes like this: > > 1. There is evidence that someone knows something that they were not > taught. > > 2. I cannot think of how they learned it. > > 3. Therefore, they didn't learn it. > > 4. Therefore, it is innate. > > Both 3 and 4 are non-sequiturs. I don't want to get into a big > discussion of this here. And I talk about this quite a bit in the > book. So I will try to resist the temptation to respond to further > postings. > > So I find your concerns both quite understandable, but not > insurmountable or even particularly difficult problems for > understanding language as a cognitive tool. > > -- Dan > > > > > > > On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > > > ----- Original Message ----- > >> From: "A. Katz" > > > >> You might as well say that a person cannot possibly avoid watching > >> TV > >> once > >> he's exposed to it, as say the same about language. But people can > >> survive > >> just fine without television, and unless someone shows them how to > >> make a > >> TV set, most people will never figure out how to build one. The > >> same > >> goes > >> for language. We're great at using it, not so great at generating > >> it > >> out > >> of thin air with no ambient culture. > >> > >> --Aya > > > > Those who are familiar with my work know that I'm anything but a > > Chomskyan, but I'm sorry, there's an enormous difference between the > > acquisition of language and the acquisition of the knowledge > > necessary to build a television set (or the brodcasting and > > recording technology behind it). As generativists have pointed out > > since the early sixties, nobody is explicitly taught language, yet > > we all acquire it. Conversely many are intensively taught elementary > > physics, engineering etc. and DON'T aquire it. This is a difference > > in kind, not in degree. > > This is not to say that culture is taught either (of course nobody > > learns in school the correct distance to stand apart from an > > interlocutor, or how many milliseconds of silence in a conversation > > constitutes a 'pregnant pause'), but these are different kinds of > > knowledge from academic knowledge explicitly taught in some cultures > > and not in others. > > All cultures have correct social distance rules, syntactic > > structures and other tacitly acquired knowledge, but not all > > cultures learn physics, or which mushrooms are edible and which > > fatal. > > I'm looking forward to reading Dan's book too, but I find 'tool' an > > inappropriate metaphor for a cultural artifact that is never > > explicitly 'taught', is learned without effort in all cultures > > regardless of level of technology and is never improved by explicit > > experimentation or accidental innovation--there will never be the > > linguistic equivalent of a 'better mousetrap'. > > I prefer Rudi Keller's view that language is an object of the 'third > > kind'--an artifact that is neither wholly natural nor man-made, but > > that develops as a spontaneous order, without being designed, and > > with 'improvements' developing in different directions from the > > intentions of the developers. See his book > > > > Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage > > Tübingen 1994 > > > > or, for those, who, like me are Germanically-challenged, > > > > On Language Change: The Invisible Hand in Language, Routledge 1995, > > translated by Brigitte Nerlich. > > > > Geoff > > > > > > Geoffrey S. Nathan > > Faculty Liaison, C&IT > > and Professor, Linguistics Program > > +1 (313) 577-1259 (C&IT) > > +1 (313) 577-8621 (English/Linguistics) > > From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 17:49:39 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 19:49:39 +0200 Subject: unhappiness debate-psycholinguistic research methods Message-ID: It seems that we are getting caught in an escalation of intellectual firepower, when a look at the research literature could focus more specifically on the issue Shannon raised. First of all, in reference to Tom's suggestions, it is not merely semantic priming that would be utilized, but a variety of priming techniques. Most important would be that the priming be masked, so that the subject would be unaware of the prime or of the nature of the real experimental question so as not to bias the results. The use of psycholinguistics as a research method in this case is not only not incompatible with a review of the literature of theoretical proposals from English diachronic phonologists and morphologists, it actually could be a necessity. To adequately structure a psycholinguistic experiment, information regarding the history and the age of the affixes could be required as background information. For example, if the bracketing paradox has as its origin that -ness and un- are on the same morpho-phonological level at some diachronic stage (the English of older speakers-who would not use the un-prefix on nouns), but younger speakers having been exposed to modern advertising language that uses it (Remember the "un-Cola"?) tend to think "un + happiness" the psycholinguist should separate subjects by age and use it is a factor within the statistical analysis. Not doing so might give inconclusive results. The differences between older and younger speakers would demonstrate a mini-stage in the historical process of affix integration. How does psycholinguistic research with agglutinating languages better inform us? We need to look at experiments with Hungarian, Turkish, Hungarian and Finnish-a special issue of Language and Cognitive Processes at the link below is a good summary of the state of the art as of 2008, and an earlier book by Baayen & Schreuder (2003) Morphological Structure in Language Processing is also informative. http://www.cognitivepsychologyarena.com/books/Advances-in-Morphological- Processing-isbn9781841698670 Ken Forster advised me, when I was consulting him some years ago about possibly applying this line of experimentation to my African language research, to study the literature on Finnish. At the time it seemed somewhat inconclusive, but I intend to keep checking back. There are people working in Hungarian, another vowel harmony language, so it looks like there will eventually be some convergence of research methods in the future. In fact, Vannerst and Boland (1999) already successfully tested the presence of morpho-phonological levels in English, explicitly invoking the lexical phonology model of English. It can be accessed at this link: http://www.ling.ohio-state.edu/~skang/371/DownloadablePapers/VannestBola nd99.pdf Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From amnfn at well.com Mon Sep 13 17:50:08 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 10:50:08 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <416964165.651968.1284392343336.JavaMail.root@starship.merit.edu> Message-ID: Geoff, Let's say we accept your bicycle analogy. Learning language is like learning to ride a bicycle. Children in bicycle riding cultures pick it up fairly quickly, barring any physical disability. Children in cultures where bicycles are not present do not pick it up at all. Very few people could design a bicycle, even in cultures where bicycles abound. They usually buy their bicycles ready made. If they make adgjustments to the frame, they are fairly minor. --Aya On Mon, 13 Sep 2010, Geoff Nathan wrote: > I did not consider my argument to be one supporting Chomsky's version of nativism, in which linguistic knowledge is independent of (and prior to) experience. Rather I was arguing that the acquisition of language was quite different from the acquisition of a carpenter's skills in that language is not a consciously designed object, unlike what I would think of as a prototypical tool, such as a hammer or a pair of glasses. The structure of language is constrained, not by our genes, but by the 'equipment' that is used to produce, perceive and understand it. I am arguing here for the 'natural' view in 'Natural' Phonology. > > Stampe and Donegan, in their seminal paper 'The Study of Natural Phonology' argued for a different kind of innateness, in which universals arise from the interaction of the nature of the materials used with the general cognitive processes of the user. This is most clear in phonology, of course, where phonological structures are (partly) determined by the anatomy and physiology of the vocal tract, the physics of sound production and the psychology of perception. We make the sounds that our mouths make it easy to produce clearly (fortitions), and in the way that it's easy to produce them (lenitions), and we hear what other people say by assuming they have the same vocal apparatus and we hear what we would have intended should we have said that. But folks like Karen van Hoek have shown that syntactic island constraints have a similar basis in processing limitations, and others have also found extrinsic causes for other universal constraints on morphology and syntax. > > So language is certainly learned, but not in the same way that arithmetic is learned--it's more like how riding a bicycle is learned, through interaction with the physical stuff, and additionally, guided by our categorization and perceptual systems. > > Donegan, Patricia J., and David Stampe, 1979. “The Study of Natural Phonology,” in Current Approaches to Phonological Theory., ed. by Dan Dinnsen. Bloomington: Indiana University Press. > > > Geoffrey S. Nathan > Faculty Liaison, C&IT > and Professor, Linguistics Program > +1 (313) 577-1259 (C&IT) > +1 (313) 577-8621 (English/Linguistics) > > ----- Original Message ----- >> From: "Daniel Everett" >> To: "Geoff Nathan" >> Cc: "Funknet Funknet" >> Sent: Monday, September 13, 2010 10:37:11 AM >> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >> Dear Geoff, >> >> These are all valid points, but none of them support nativism. >> >> One place to start is with Fiona Cowie's relatively recent book, >> What's Within >> (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) >> >> There are many cultural artifacts and values and types of knowledge >> that are never taught, but are all acquired just fine, from knowing to >> sit still in a canoe to how to use a bow and arrow. They are learned >> out of necessity and by observation, with no explicit instruction. >> >> In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev >> Fedorenko, in Cognition, support was offered for the idea that there >> are cognitive tools, number and counting being the examples we >> discussed. >> >> The reasoning chain for innatism often goes like this: >> >> 1. There is evidence that someone knows something that they were not >> taught. >> >> 2. I cannot think of how they learned it. >> >> 3. Therefore, they didn't learn it. >> >> 4. Therefore, it is innate. >> >> Both 3 and 4 are non-sequiturs. I don't want to get into a big >> discussion of this here. And I talk about this quite a bit in the >> book. So I will try to resist the temptation to respond to further >> postings. >> >> So I find your concerns both quite understandable, but not >> insurmountable or even particularly difficult problems for >> understanding language as a cognitive tool. >> >> -- Dan >> >> >> >> >> >> >> On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: >> >>> ----- Original Message ----- >>>> From: "A. Katz" >>> >>>> You might as well say that a person cannot possibly avoid watching >>>> TV >>>> once >>>> he's exposed to it, as say the same about language. But people can >>>> survive >>>> just fine without television, and unless someone shows them how to >>>> make a >>>> TV set, most people will never figure out how to build one. The >>>> same >>>> goes >>>> for language. We're great at using it, not so great at generating >>>> it >>>> out >>>> of thin air with no ambient culture. >>>> >>>> --Aya >>> >>> Those who are familiar with my work know that I'm anything but a >>> Chomskyan, but I'm sorry, there's an enormous difference between the >>> acquisition of language and the acquisition of the knowledge >>> necessary to build a television set (or the brodcasting and >>> recording technology behind it). As generativists have pointed out >>> since the early sixties, nobody is explicitly taught language, yet >>> we all acquire it. Conversely many are intensively taught elementary >>> physics, engineering etc. and DON'T aquire it. This is a difference >>> in kind, not in degree. >>> This is not to say that culture is taught either (of course nobody >>> learns in school the correct distance to stand apart from an >>> interlocutor, or how many milliseconds of silence in a conversation >>> constitutes a 'pregnant pause'), but these are different kinds of >>> knowledge from academic knowledge explicitly taught in some cultures >>> and not in others. >>> All cultures have correct social distance rules, syntactic >>> structures and other tacitly acquired knowledge, but not all >>> cultures learn physics, or which mushrooms are edible and which >>> fatal. >>> I'm looking forward to reading Dan's book too, but I find 'tool' an >>> inappropriate metaphor for a cultural artifact that is never >>> explicitly 'taught', is learned without effort in all cultures >>> regardless of level of technology and is never improved by explicit >>> experimentation or accidental innovation--there will never be the >>> linguistic equivalent of a 'better mousetrap'. >>> I prefer Rudi Keller's view that language is an object of the 'third >>> kind'--an artifact that is neither wholly natural nor man-made, but >>> that develops as a spontaneous order, without being designed, and >>> with 'improvements' developing in different directions from the >>> intentions of the developers. See his book >>> >>> Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage >>> Tübingen 1994 >>> >>> or, for those, who, like me are Germanically-challenged, >>> >>> On Language Change: The Invisible Hand in Language, Routledge 1995, >>> translated by Brigitte Nerlich. >>> >>> Geoff >>> >>> >>> Geoffrey S. Nathan >>> Faculty Liaison, C&IT >>> and Professor, Linguistics Program >>> +1 (313) 577-1259 (C&IT) >>> +1 (313) 577-8621 (English/Linguistics) >>> > > From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 18:02:25 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 20:02:25 +0200 Subject: Query of interest to Funknetters Message-ID: Given the debate, I should think that this query on the Linguist List should be of interest. ===========================Directory============================== 1) Date: 09-Sep-2010 From: Edouard Machery < machery at pitt.edu > Subject: Request for Participation for Innateness Study -------------------------Message 1 ---------------------------------- Date: Mon, 13 Sep 2010 11:41:07 From: Edouard Machery [machery at pitt.edu] Subject: Request for Participation for Innateness Study E-mail this message to a friend: http://linguistlist.org/issues/emailmessage/verification.cfm?iss=21-3616 .html&submissionid=2647414&topicid=8&msgnumber=1 Dear colleagues, Paul Griffiths (Sydney), Joshua Knobe (Yale), Stefan Linquist (Guelph), Edouard Machery (Pittsburgh), Richard Samuels (OSU), and Karola Stotz (Sydney) are collecting data about scientists' use of the notion of innateness in several disciplines. We are particularly interested in how linguists use this notion. We would like to invite you to take part to a short web study (at most 10 minutes). You will be asked to read a few vignettes and to judge whether some traits are innate. You will also be asked to fill out a brief demographic questionnaire. This survey is entirely anonymous, and the study has be approved by Guelph IRB, OSU IRB, Pittsburgh IRB, Sydney IRB, and Yale IRB. To participate to this study, follow the following link: http://yale.qualtrics.com/SE/?SID=SV_5yTK6uiKUc9HPy4 We would also appreciate if you could forward this e-mail to your colleagues and graduate students! Many thanks for your help!! Edouard Machery Linguistic Field(s): Cognitive Science Pragmatics Semantics Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From erin.shay at colorado.edu Mon Sep 13 18:46:56 2010 From: erin.shay at colorado.edu (Erin Shay) Date: Mon, 13 Sep 2010 12:46:56 -0600 Subject: Unsubscribe Message-ID: Would you mind taking me off the subscription list, please? Thank you. Erin Erin Shay, Ph.D. Research Ass't Professor Dept. of Linguistics, Box 295 University of Colorado Boulder, CO 80309-0295 USA Phone: 303-882-3786 From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 20:14:13 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 22:14:13 +0200 Subject: unhappiness Message-ID: In response to Craig's observation: This is important. Now we would need to look at the history of the formation of -ness nouns to provide some clues as to the interactions of the two processes for a proper design of the study. If the observations of our list participants is correct, a study controlling age and other social factors might provide some clues regarding some type of historical change in progress, or a recycling of older meanings, further underscoring the importance of cross-fertilization between the fields of historical linguistics and psychology. Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From mark at polymathix.com Mon Sep 13 22:04:31 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 17:04:31 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: A. Katz wrote: > Mark, > > Okay, I'm glad that you agree that a language user might be a machine or a > non-human. That brings us much closer together. Nevertheless, there is > also the fact that language can be preserved, but not used, for > thousands of years, and then people start using it again. Again, I would say that artifacts of language can be preserved separately from language users, but not language itself. To my way of thinking, language is the processes by which such artifacts are created and used (when artifacts are involved at all). Those processes are synonymous with language use and they cannot be divorced from their embodiment, the language user. > If it were an biological process, or even something like a live virus > passed down from parent to child, then reviving it once it was dead would > be > impossible. But it's not. It's a tool. A tool that a group of people can > stop using and making, but keep the blue prints for, and then make it > again when they choose to. > > I wonder how many linguists realize that Hebrew was revived from a dead > language. I would assume most of us do. But I would also assume that most of us realize that Modern Hebrew was engineered to capture the available data in and about the dead language, not the language itself. There's no way to revive a dead language because all of its embodiments are, like, dead. To my mind, Modern Hebrew was created by tracing over a palimpsest. > How many think it is just a slavic language with Hebrew lexemes? I can't imagine any linguist who would think that. I might be wrong, I guess. > How many realize that the grammar, the triliteral roots, the derivational > system, are all working, and that any changes in pronunciation are of > little importance when the basic contrasts are preserved? This is all Ling 101 stuff, where I'm from, since Hebrew is the most culturally salient Semitic language in these parts, and Semitic is an important part of typological space with its triliteral and quadriliteral roots. Some people's mileage might vary, I guess. > Mark, if you can see that a language can be used by a non-human, then why > not also acknowledge that it's a system for transferring information, and > that the code can be preserved while in disuse? First, I accept that language is a tool in a far-from-prototypical sense (it's not a wrench, but it's not eating, either). As I said, a tool is as a tool does, and since language can be used as a tool, it's a tool, trivially, when it's being used as a tool. My point has been that not everything worth knowing about language profits from the insight that language can be a tool, and that that insight is not enough to justify the study of language separately from any of its observed or potential embodiments (and that, in fact, I know of nothing that might justify a reductionist approach of that order). Second, though, and more importantly, I most emphatically do not believe that ***THE*** code of ***A*** language is preserved by any finite volume of preserved text, for at least two related reasons: 1. There's no such thing as THE code of A language, so the whole endeavor suffers from presupposition failure. Every embodiment (like you, and me) has its own code. Communication ensues more or less effectively when a speaker and hearer are using codes that overlap to a sufficient degree, all the way up and down the chain from phonology to ontological commitment. 2. Even if we stipulated some "common code" that was shared (perfectly, or within some variationist framework) by some community of speakers, no amount of preserved text and/or grammatical description would suffice to preserve that common code. I would also note that, the farther you go from the artifacts of language to the code we postulate as existing in the minds of speaker/hearers, the less likely you're going to understand anything about it without taking the exigencies of the embodiment into account. My favorite example of this is the ease with which a language can be constructed (constructed as in Esperanto and Klingon) in which center embedding is rampant, the ease with which software parsers and generators can be built for such languages, and the abject difficulties any human being will have speaking or understanding any such language. -- Mark > On Mon, 13 Sep 2010, Mark P. Line wrote: > >> Aya -- >> >> A. Katz wrote: >>> Mark, >>> >>> Sorry, but "eating" is not a tool. It's a biological process. >>> >>> Eating cannot occur outside the biological context. A human being can >>> avoid eating, but if so he starves. Feral children do not speak, but >>> they eat, like everyone else, or they die. Eating does not have to be >>> taught, there is no critical age of acquisition and it is not uniquely >>> an >>> artefact of human culture. If you are an animal, you eat. Eating cannot >>> survive the death of the eater. >>> >>> Language can. >> >> I would say that not language, but the artifacts of language (texts, >> audio >> recordings) can survive the people who created them, because I try very >> hard not to reify the artifacts of language as "language". >> >> The (usually communicative) process I refer to as language cannot exist >> independently of its embodiment. That said, I don't care if the >> embodiment >> is human, computer, cetacean or non-human primate. >> >> >> -- Mark >> >> Mark P. Line >> >> >>> On Sun, 12 Sep 2010, Mark P. Line wrote: >>> >>>> Aya -- >>>> >>>> You said: "Other tools can be studied separately from the people who >>>> use >>>> them or created them. Why not language?" >>>> >>>> Although language can certainly be considered a tool, I think it's >>>> unlike >>>> other tools in several very significant respects. >>>> >>>> 1. Although we're not born with language, we can't avoid (pathologies >>>> excluded) acquiring it unless we're not exposed to it. To that extent, >>>> language is a biological phenomenon. A prototypical tool is not a >>>> biological phenomenon, so I'm not sure how valid any conclusions might >>>> be >>>> that are drawn from a premise of language-as-tool. >>>> >>>> 2. A tool is as a tool does: Anything is a tool that is being used as >>>> a >>>> tool, including dead wombats, broken screwdrivers or decks of playing >>>> cards. (Completing the imagined scenarios is left as an exercise for >>>> the >>>> reader...) So saying that language is a tool is only saying that >>>> language >>>> is used as a tool. Quite a few conclusions can be and have been drawn >>>> from >>>> the fact that language is used as a tool, but I would have to be >>>> convinced >>>> in detail that almost everything worth knowing about language is >>>> dependent >>>> on the premise of language-as-tool. >>>> >>>> 3. If language is a "tool" for (say) communicating ideas, then eating >>>> is >>>> a >>>> "tool" for reducing hunger. In both cases, I worry about the tool >>>> metaphor >>>> being stretched so far from the prototype that we're left with an >>>> out-and-out category fallacy. >>>> >>>> 4. More prototypical tools can be studied separately from those who >>>> use >>>> or >>>> create them because those tools are easily observed separately from >>>> those >>>> who use or create them. I don't think the same thing can be said of >>>> language -- very little about language can be observed apart from its >>>> use, >>>> so very little about language can be observed apart from its user(s). >>>> >>>> 5. Any proposal to study something as complex as language separately >>>> from >>>> its embodiment is suspicious to me, smacking of reductionism -- >>>> something >>>> up with which linguistics has had to put a tad much. Anything that >>>> puts >>>> language back into its human context would be a step forward. >>>> >>>> >>>> -- Mark >>>> >>>> Mark P. Line >>>> >>>> >>>> >>>> A. Katz wrote: >>>>> Tom, >>>>> >>>>> I understand the uncomfortable association with Chomsky that speaking >>>>> of >>>>> language apart from people can have. Competence versus performance, >>>>> the >>>>> way Chomsky used those terms, never made sense. But that's precisely >>>>> because >>>>> he associated "competence" with the brain and suggested that it was >>>>> hard >>>>> wired there -- when there was never any evidence of that. >>>>> >>>>> However, if we don't distinguish language from humans, and language >>>>> processing from language data, then how are we going to judge >>>>> artificial >>>>> language-using devices as to their efficacy at producing and >>>>> interpreting >>>>> language? How are we going to determine whether and to what extent a >>>>> non-human has acquired language? >>>>> >>>>> We aren't born with it. We don't embody it. It's a tool that we use >>>>> to >>>>> communicate. Other tools can be studied separately from the people >>>>> who >>>>> use >>>>> them or created them. Why not language? >>>>> >>>>> --Aya >>>>> >>>>> >>>>> >>>>> >>>>> On Sun, 12 Sep 2010, Tom Givon wrote: >>>>> >>>>>> >>>>>> >>>>>> >>>>>> I wonder whether asking "how does language work?" is really a >>>>>> meaningful >>>>>> question without asking "how does the language user work?" One of >>>>>> the >>>>>> worst >>>>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>>>> distinction >>>>>> of competence ("language", "knowledge") vs. performance ("language >>>>>> user", >>>>>> "processing"). It purported to limit linguists to the armchair >>>>>> methods >>>>>> that >>>>>> study competence, and relegated to psychology the quantified, >>>>>> distributional/variationist methods that study behavior, processing >>>>>> and >>>>>> on-line communication. The first breach in this artificial >>>>>> methodological >>>>>> wall occurred, leastwise for some of us, when we discovered the >>>>>> intermediate >>>>>> method of quantified studies of text, interaction, and conversation. >>>>>> As >>>>>> an >>>>>> ex-biologist, I am forever puzzled by the methodological purism we >>>>>> sill >>>>>> seem >>>>>> to embrace in linguistics, in the face of the manifest complexity >>>>>> and >>>>>> connectivity of language (mind, brain, culture, sociality, >>>>>> evolution, >>>>>> etc.). >>>>>> In biology, another extremely complex science with multiple >>>>>> connections >>>>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>>>> evolution, >>>>>> etc.), ANY method is welcome so long as it does the job of >>>>>> furthering >>>>>> our >>>>>> understanding. And by understanding we mean ever-wider circles of >>>>>> connectivity. >>>>>> >>>>>> Best, TG >>>>>> ================ >>>>>> >>>>>> >>>>>> A. Katz wrote: >>>>>>> Chris, >>>>>>> >>>>>>> Thanks for addressing this question. I understand that many, many >>>>>>> linguists >>>>>>> are quite properly and approriately interested in this ultimate >>>>>>> question: >>>>>>> "How does the language user work?" (I am also intetested in this >>>>>>> question >>>>>>> some of the time.) >>>>>>> >>>>>>> My point is that "how does language work?" is also a valid >>>>>>> question, >>>>>>> and a >>>>>>> central one to the field of linguistics. These two questions are >>>>>>> not >>>>>>> at >>>>>>> all >>>>>>> the same. >>>>>>> >>>>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>>>> the >>>>>>> "language". There are many good reasons to do so. For anyone >>>>>>> working >>>>>>> in >>>>>>> computerized language processing or in non-human language studies, >>>>>>> this >>>>>>> is >>>>>>> a significant point. >>>>>>> >>>>>>> It does not matter if a computerized language processing system >>>>>>> even >>>>>>> remotely simulates what humans do with language in their brains. It >>>>>>> does >>>>>>> matter whether it comes up with comparable or indistinguishable >>>>>>> results. >>>>>>> >>>>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>>>> doing >>>>>>> the >>>>>>> same things inside the same modules in his brain as a human does. >>>>>>> It >>>>>>> does >>>>>>> matter if the results are functionally equivalent. >>>>>>> >>>>>>> We need to make that distinction, between humans and their >>>>>>> language, >>>>>>> or >>>>>>> we >>>>>>> will be caught inside a circular definition with no way to break >>>>>>> out >>>>>>> or >>>>>>> to >>>>>>> prove anything, not about others and not about ourselves! >>>>>>> >>>>>>> --Aya >>>>>>> >>>>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>>>> >>>>>>> >>>>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>>>> >>>>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>>>> >>>>>>>> I don't for a moment dispute that expressions are often >>>>>>>> susceptible >>>>>>>> to >>>>>>>> multiple interpretations, that these interpretations are guided by >>>>>>>> all >>>>>>>> kinds of contextual information, or that different people, or even >>>>>>>> the >>>>>>>> same person at different times, may end up selecting differently >>>>>>>> from >>>>>>>> the >>>>>>>> various interpretations. Your example 'Open happiness' in another >>>>>>>> contribution to this thread illustrates these points very well. >>>>>>>> >>>>>>>> My point, though, is that each of these different interpretations, >>>>>>>> as >>>>>>>> well >>>>>>>> as the selection of one (or more) as more likely in a particular >>>>>>>> context, >>>>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>>>> evolved in >>>>>>>> the course of the phylogenetic development of language in the >>>>>>>> human >>>>>>>> species, and developed ontogenetically in that particular >>>>>>>> interpreter's >>>>>>>> brain. It is surely likely that those mechanisms are highly >>>>>>>> similar >>>>>>>> in >>>>>>>> different human beings, even though there may be differences in >>>>>>>> the >>>>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>>>> order >>>>>>>> to >>>>>>>> answer the question 'How do we communicate using language?' or, if >>>>>>>> you >>>>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>>>> what >>>>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>>>> with. >>>>>>>> >>>>>>>> In particular, as linguists, we are interested (well, some of us >>>>>>>> are, >>>>>>>> though clearly not all) in whether the constructs we posit in our >>>>>>>> theories >>>>>>>> of language have psychological validity in the sense that they >>>>>>>> correspond >>>>>>>> to ways in which aspects of language are represented in the brain. >>>>>>>> [As >>>>>>>> an >>>>>>>> aside, I do realise that there are linguists who strenuously >>>>>>>> resist >>>>>>>> what >>>>>>>> they see as a misguided emphasis on mental representation in the >>>>>>>> work >>>>>>>> of >>>>>>>> cognitive scientists, but it seems clear that language must be >>>>>>>> represented >>>>>>>> in some way in the brain in order that we can engage in the >>>>>>>> sociosemiotic >>>>>>>> acts of meaning making which are the primary focus for many of >>>>>>>> these >>>>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>>>> need >>>>>>>> answers to questions such as: What kind of representation does the >>>>>>>> human >>>>>>>> language processing system have for 'open' and for 'happiness'? >>>>>>>> Are >>>>>>>> the >>>>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>>>> distinguish >>>>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>>>> might >>>>>>>> want to add 'for this particular interpreter', though there must >>>>>>>> be >>>>>>>> similarities across interpreters for communication to be possible) >>>>>>>> represented in the same or different ways, in the same or >>>>>>>> different >>>>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>>>> in >>>>>>>> thinking that these familiar levels of linguistic description must >>>>>>>> be >>>>>>>> differentiated, as such, in the human language processing system? >>>>>>>> Does >>>>>>>> the >>>>>>>> representation for 'open' distinguish between what we call verbal >>>>>>>> and >>>>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>>>> analyses >>>>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>>>> than >>>>>>>> the >>>>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>>>> adjectives? >>>>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, >>>>>>>> or >>>>>>>> both? >>>>>>>> All these questions, and many many more, are relevant to the >>>>>>>> construction >>>>>>>> of a model of language which reflects how language users >>>>>>>> communicate >>>>>>>> (as, >>>>>>>> of course, are a whole set of other questions about the >>>>>>>> sociocultural >>>>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>>>> too >>>>>>>> superficial reading in the area suggests that psycholinguists have >>>>>>>> gone >>>>>>>> some of the way towards answering some of the questions we might >>>>>>>> want >>>>>>>> to >>>>>>>> ask, but that there is still a long way to go. >>>>>>>> >>>>>>>> Chris >>>>>>>> -------------------------------------------------- >>>>>>>> From: "A. Katz" >>>>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>>>> To: "Chris Butler" >>>>>>>> Cc: "FUNKNET" >>>>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>>>> >>>>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>>>> see >>>>>>>>> what >>>>>>>>> is happening in each person's head, and we see that conflicting >>>>>>>>> analyses >>>>>>>>> are the norm, rather than the exception, among normal language >>>>>>>>> speakers, >>>>>>>>> then we will realize that the way language works to transmit >>>>>>>>> information >>>>>>>>> is despite individual differences and not because of uniform >>>>>>>>> processing >>>>>>>>> strategies. >>>>>>>>> >>>>>>>>> Even when all are agreed as to the meaning of an utterance, they >>>>>>>>> do >>>>>>>>> not >>>>>>>>> process it the same way. Which means that processing is seocndary >>>>>>>>> to >>>>>>>>> information transmission. >>>>>>>>> >>>>>>>>> --Aya >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>>>> >>>>>>>>>> Dick's comment that "this discussion raises the really >>>>>>>>>> fundamental >>>>>>>>>> question of what kind of thing we think language is: social or >>>>>>>>>> individual" is, it seems to me, an important one, particularly >>>>>>>>>> for >>>>>>>>>> those >>>>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>>>> truly >>>>>>>>>> functional model of language would be one which aims to account >>>>>>>>>> for >>>>>>>>>> how >>>>>>>>>> human beings communicate using language, or in other words tries >>>>>>>>>> to >>>>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>>>> now, >>>>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>>>> natural >>>>>>>>>> language user work?' In trying to answer this question we need >>>>>>>>>> to >>>>>>>>>> accept >>>>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>>>> explore >>>>>>>>>> both >>>>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>>>> communicate >>>>>>>>>> using language. We need to know BOTH how people create and >>>>>>>>>> respond >>>>>>>>>> to >>>>>>>>>> meanings and express those meanings in forms during social >>>>>>>>>> interaction >>>>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>>>> order >>>>>>>>>> to make such interaction possible. Both are important parts of >>>>>>>>>> the >>>>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>>>> though >>>>>>>>>> this particular thread of the Funknet discussion has >>>>>>>>>> concentrated >>>>>>>>>> more >>>>>>>>>> on the second aspect, and so will I. >>>>>>>>>> >>>>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>>>> "exploring >>>>>>>>>> the structure of a language so that I can understand how all the >>>>>>>>>> bits >>>>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>>>> Dick >>>>>>>>>> puts it, is useless - far from it. After all, the hypotheses >>>>>>>>>> that >>>>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>>>> like. >>>>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>>>> evidence >>>>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>>>> least >>>>>>>>>> consistent with what we know of the processes which go on when >>>>>>>>>> we >>>>>>>>>> use >>>>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>>>> only >>>>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>>>> terms >>>>>>>>>> of >>>>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>>>> that >>>>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>>>> know >>>>>>>>>> everything there is to know about language processing. But it >>>>>>>>>> does >>>>>>>>>> mean >>>>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>>>> and >>>>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>>>> Ray >>>>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also >>>>>>>>>> need >>>>>>>>>> to >>>>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>>>> that >>>>>>>>>> we >>>>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>>>> part, >>>>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>>>> experiments >>>>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>>>> criticism >>>>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>>>> natural >>>>>>>>>> communicative conditions. >>>>>>>>>> >>>>>>>>>> I also agree with Dick when he says that "the differences >>>>>>>>>> between >>>>>>>>>> individuals really matter", and with Lise when she points out >>>>>>>>>> that >>>>>>>>>> "we >>>>>>>>>> must also be careful not to idealize "what's in people's heads" >>>>>>>>>> as >>>>>>>>>> if it >>>>>>>>>> were a single coherent construct that we are trying to >>>>>>>>>> discover". >>>>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>>>> to >>>>>>>>>> all >>>>>>>>>> language users by virtue of the evolution of the language >>>>>>>>>> faculty >>>>>>>>>> and >>>>>>>>>> which constitute the "general processes" which Dick says >>>>>>>>>> psycholinguists >>>>>>>>>> are interested in. >>>>>>>>>> >>>>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>>>> general >>>>>>>>>> to >>>>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>>>> cases >>>>>>>>>> in >>>>>>>>>> terms of a hierarchy such as Brian suggests. One thing this >>>>>>>>>> means >>>>>>>>>> is >>>>>>>>>> that we should be giving our university students of linguistics >>>>>>>>>> (and >>>>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>>>> aspects >>>>>>>>>> of >>>>>>>>>> linguistics that introduce them to the use of at least some of >>>>>>>>>> the >>>>>>>>>> basic >>>>>>>>>> statistical methods in language study, and I'm sure this is >>>>>>>>>> indeed >>>>>>>>>> going >>>>>>>>>> on in some enlightened places. To those who suspect this can't >>>>>>>>>> be >>>>>>>>>> done >>>>>>>>>> with maths-shy students who don't initially see the need for it, >>>>>>>>>> I >>>>>>>>>> offer >>>>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>>>> courses to >>>>>>>>>> people with little or no prior experience in quantitative >>>>>>>>>> techniques. >>>>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>>>> linguistics >>>>>>>>>> students in an institution where we had many mature students who >>>>>>>>>> had >>>>>>>>>> come into university level studies with non-standard >>>>>>>>>> qualifications, >>>>>>>>>> and >>>>>>>>>> were not well equipped for courses of this kind by their >>>>>>>>>> previous >>>>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>>>> own >>>>>>>>>> perspective as language students rather than that of the >>>>>>>>>> statistician, >>>>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>>>> rather >>>>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>>>> and >>>>>>>>>> to >>>>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>>>> research >>>>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>>>> techniques. And I still find that bright graduate students >>>>>>>>>> respond >>>>>>>>>> well >>>>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>>>> advanced >>>>>>>>>> techniques needed for many real research projects in various >>>>>>>>>> areas >>>>>>>>>> of >>>>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>>>> already >>>>>>>>>> happening in our fine institutions of higher education! >>>>>>>>>> >>>>>>>>>> Chris Butler >>>>>>>>>> >>>>>>>>>> >>>>>>>> >>>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>> >>>> >>>> -- Mark >>>> >>>> Mark P. Line >>>> Bartlesville, OK >>>> >>>> >>> >>> >> >> >> -- Mark >> >> Mark P. Line >> Bartlesville, OK >> >> > > -- Mark Mark P. Line Bartlesville, OK From amnfn at well.com Mon Sep 13 23:33:06 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 16:33:06 -0700 Subject: Language as a Tool In-Reply-To: <24374061.1284385200100.JavaMail.root@wamui-june.atl.sa.earthlink.net> Message-ID: Jess, Your observation about body parts serving as prototypical tools is apt. I would add that teeth serving as cutting tools came first, and that their use in making dental consonants was discovered later. But we have other tools now to help us represent dental consonants, including but not limited to the writing system and speech synthesizers. Parrots can make sounds that pass for dental consonants, though they have no teeth. Chimpanzees have teeth, but cannot make those sounds. The point? There is more than one object that can serve as a tool to produce language contrasts. Some are embodied, and some are not. Anatomical similarity isn't everything. --Aya On Mon, 13 Sep 2010, jess tauber wrote: > Actually I'd say the true prototypical tools are body parts doing particular jobs- nails, teeth, hairs, wings, fins, legs etc. I doubt it escaped our early ancestors that their created or found tools had functions similar or identical to parts they had themselves or saw in other creatures. Even other lineages must have some inkling (as when a jay uses a thorn to prize a grub out of rotten wood, or a sea otter bashes shellfish with stones fished up from the sea floor). Not everything is pure instinct. > > Sometimes these parts come out- we lose teeth, nails, hair, birds feathers, and so on, and you can also yank them out of corpses, skeletons, etc., and the occasional unwilling live victim. We may find them loose. This sets the stage for alienability, and generalization perhaps? I can get a sharp tool from a sabertooth, or from the living rock if I knock politely. > > In languages with 'bipartite' structure, effector bodypart and instrument terms are often dealt with identically, and stand in contrast with pathway/position terms, which may have a mirror in the way the brain deals with tools and gait/posture related motion. > > Jess Tauber > phonosemantics at earthlink.net > > From LGarrison at gc.cuny.edu Mon Sep 20 15:11:21 2010 From: LGarrison at gc.cuny.edu (Garrison, Leigh) Date: Mon, 20 Sep 2010 11:11:21 -0400 Subject: Columbia School Linguistics Conference Announcement Message-ID: 10th International Columbia School Conference on the Interaction of Linguistic Form and Meaning with Human Behavior Conference theme: Grammatical analysis and the discovery of meaning October 9-11, 2010 Rutgers University New Brunswick, New Jersey Invited speakers: Flora Klein-Andreu (Stony Brook University) Linguistics for non-linguists Andrea Tyler (Georgetown University) Connecting Spatial Particles and Aspect Markers: Applying the Principled Polysemy Model to Russian za List of presenters: Tanya Karoli Christensen (Copenhagen University) Sign combinations in context: Imperatives and modal particles in Danish Ellen Contini-Morava (University of Virginia) The meaning(s?) of non-animate deictic markers in Swahili Joseph Davis (The City College – CUNY) Diver’s Latin voice and case Bob de Jonge (University of Groningen) Phonology as Human Behaviour revisited: The case of Romance languages Thomas Eccardt (Independent scholar) Pitch and aperture: Two articulatory scalars in comparison Richard Epstein (Rutgers University) Some discourse uses of the distal demonstrative determiner in Beowulf Elena Even-Simkin, Yishai Tobin (Ben-Gurion University of the Negev) Internal vowel alternation as a phonological-semantic sign system in English according to the sign-oriented theory of the Columbia School Alan Huffman (The Graduate Center – CUNY) The phonological motivation for Verner's Law and Grimm's Law Robert Kirsner (University of California, Los Angeles) Minimal units, their context, and the insufficiency of conceptual metaphor: Revisiting the Dutch dismissive idiom ho maar ‘fuhgeddaboudit, of course not!’ Robert Leonard (Hofstra University) Linguistic meaning, pragmatics and context: Semantic analysis of evidence in a double homicide trial seeking to weigh intent Lin Lin (University of California, Los Angeles) Rethinking of the Chinese demonstratives in the Columbia School framework Carol Moder (Oklahoma State University) Dirty hands, dirty work: Usage-based noun modification Ricardo Otheguy (The Graduate Center – CUNY) A report on current research on Spanish in New York Wallis Reid (Rutgers University) English verb number: Syntactic or semantic? Hidemi Riggs (Soka University of America) The structure of Japanese conditionals in Modern Japanese: A grammatical account from a functional linguistics perspective Nancy Stern (The City College – CUNY) Ourself, themself, and more: The communicative function of Number in -self pronouns Lavi Wolf, Yishai Tobin (Ben-Gurion University of the Negev) Phonological proclivities across languages according to the theory of Phonology as Human Behavior The Columbia School is a group of linguists developing the theoretical framework origi¬nally established by the late William Diver. Language is seen as a symbolic tool whose structure is shaped both by its communicative function and by the characteristics of its human users. Grammatical analyses account for the distribution of linguistic forms as an interaction between linguistic meaning and pragmatic and functional factors such as inference, ease of processing, and iconicity. Phonological analyses explain the syntag¬matic and paradigmatic distribution of phonological units within signals, also drawing on both communicative function and human physiological and psychological characteristics. * * * * * * * * The support of The Columbia School Linguistic Society is gratefully acknowledged www.csling.org * * * * * * * * Selected Columbia School bibliography: Contini-Morava, Ellen, Robert S. Kirsner, and Betsy Rodriguez-Bachiller (eds.). 2005. Cognitive and Communicative Approaches to Linguistic Analysis. Amsterdam/Philadelphia: John Benjamins. Contini-Morava, Ellen, and Barbara Sussman Goldberg (eds.). 1995. Meaning as Explanation: Advances in Linguistic Sign Theory. Berlin: Mouton de Gruyter. Davis, Joseph, Radmila Gorup, and Nancy Stern (eds.). 2006. Advances in Functional Linguistics: Columbia School beyond its origins. Amsterdam/Philadelphia: John Benjamins. Huffman, Alan. 1997. The Categories of Grammar: French lui and le. Amsterdam/Philadelphia: John Benjamins. Huffman, Alan. 2001. “The Linguistics of William Diver and the Columbia School.” WORD 52:1, 29-68. Reid, Wallis. 1991. Verb and Noun Number in English: A Functional Explanation. London: Longman. Reid, Wallis, Ricardo Otheguy, and Nancy Stern (eds.). 2002. Signal, Meaning, and Message: Perspectives on Sign-Based Linguistics. Amsterdam/Philadelphia: John Benjamins. Tobin, Yishai. 1997. Phonology as Human Behavior: Theoretical Implications and Clinical Applications. Durham: Duke U Press. * * * * * * * * For more information, please contact Joseph Davis at jdavis at ccny.cuny.edu From yutamb at mail.ru Mon Sep 20 19:54:12 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Tue, 21 Sep 2010 02:54:12 +0700 Subject: References to your own articles Message-ID: Dear Funknet colleagues, selfreferences are prohibited. But what can I do if nobody writes anything in my narrow field of research? If I omit references to my own articles, then there'd be no references at all. What can you advise me? Is it not noble to refer to your own opus? Looking forward to hearing from you soon to yutamb at mail.ru Bewell, Yuri Tambovtsev From yutamb at mail.ru Wed Sep 22 11:23:33 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Wed, 22 Sep 2010 18:23:33 +0700 Subject: Language variation Message-ID: Dear Funknet colleagues, I study language variation on phonetic and semantic level. I mean I research how the speech souns vary in different languages from the point of view of their occurrence. That is, how many particular phonemes occur in this or that world language. I studied the occurrence of phonemes in the speech chain of some 300 world languages. However, in every language speech sound chain the element vary. I study their variation with the help of the coefficient of variation. I wonder if I could compare my results to some other results of this sort. Who studies the variation of the occurrence speech sounds in different languages now? Or is it not in fashion any more? It was very fashionable in the 1960-1970. The other question is if it is all right to use the coefficient of variation to study the occurrence of speech sounds? Looking forward to your advice to yutamb at mail.ru Yours sincerely Yuri Tambovtsev From francisco.ruizdemendoza at unirioja.es Wed Sep 22 13:26:59 2010 From: francisco.ruizdemendoza at unirioja.es (=?iso-8859-1?Q?=22Francisco_Jos=E9_Ruiz_De_Mendoza_Ib=E1=F1ez=22?=) Date: Wed, 22 Sep 2010 15:26:59 +0200 Subject: 44th Annual Meeting of the Societas Linguistica Europaea Sept 8-11, 2011 University of La Rioja Message-ID: Dear list members, The first call for papers for the 44th Annual Meeting of the Societas Linguistica Europaea, to be held at the University of La Rioja, September 8-11, 2011 can be found at: http://sle2011.cilap.es/call-for-papers Best regards, Francisco Ruiz de Mendoza =========================== Francisco Ruiz de Mendoza www.cilap.es www.lexicom.es From djh514 at york.ac.uk Wed Sep 22 17:23:29 2010 From: djh514 at york.ac.uk (Damien Hall) Date: Wed, 22 Sep 2010 18:23:29 +0100 Subject: Language variation In-Reply-To: Message-ID: Dear Yuri > I research how the speech souns vary in different > languages from the point of view of their occurrence. That is, how many > particular phonemes occur in this or that world language. I studied the > occurrence of phonemes in the speech chain of some 300 world languages. > However, in every language speech sound chain the element vary. I study > their variation with the help of the coefficient of variation. I wonder > if I could compare my results to some other results of this sort. Who > studies the variation of the occurrence speech sounds in different > languages now? If you can get hold of it, you should have a look at work like that of Ian Maddieson http://en.wikipedia.org/wiki/Ian_Maddieson and Peter Ladefoged (particularly his 1996 book with Ian Maddieson and his 2001 sole-authored volume): http://en.wikipedia.org/wiki/Peter_Ladefoged#Bibliography > The other question is if it is all right to use the > coefficient of variation to study the occurrence of speech sounds? In order to know what to say about this, I'd need to know more about how you calculate the coefficient. I might not be able to say more even then, but that would be a start! Sorry not to be more helpful, but good luck! Best wishes Damien Hall -- Damien Hall University of York Department of Language and Linguistic Science Heslington YORK YO10 5DD UK Tel. (office) +44 (0)1904 432665 (mobile) +44 (0)771 853 5634 Fax +44 (0)1904 432673 http://www.york.ac.uk/res/aiseb http://www.york.ac.uk/depts/lang/people/pages/hall.htm DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm From yasshiraijp at gmail.com Wed Sep 22 18:00:43 2010 From: yasshiraijp at gmail.com (Yasuhiro Shirai) Date: Wed, 22 Sep 2010 14:00:43 -0400 Subject: Hispanic Linguistics Position (open-rank), University of Pittsburgh In-Reply-To: Message-ID: Hispanic Linguistics The University of Pittsburgh, Department of Linguistics, invites applications for a tenure stream professor (open rank) position in Hispanic Linguistics, pending budgetary approval. Applicants *must* have a solid training in a core area of linguistics. Competence in one or more of the following areas will be an advantage: second language acquisition, discourse analysis, bilingualism, language policy, and bilingual education. Candidates should send a CV (including a list of funded research if applicable), a statement of research and teaching interests, copies of 2 reprints or other written work, teaching evaluations (if available), and the names of three references. Hard copies of reference letters should be sent *directly* to the search committee. Send materials to: Search Committee, Hispanic Linguistics, Dept. of Linguistics, 2816 CL, University of Pittsburgh, Pittsburgh, PA 15260; (412) 624 5900; Fax (412) 624 6130. E-mail inquiries should be directed to Professor Yasuhiro Shirai, Chair, at yshirai at pitt.edu. The web page for the department is http://www.linguistics.pitt.edu/*.* To ensure full consideration, complete applications should be received by December 1, 2010. Preliminary interviews will be held at the LSA meeting in Pittsburgh in January 2011. The University of Pittsburgh is an Affirmative Action, Equal Opportunity Employer. Women and members of minority groups under-represented in academia are especially encouraged to apply. Yasuhiro Shirai Professor and Chair Department of Linguistics 2801 Cathedral of Learning University of Pittsburgh Pittsburgh, PA 15260 Tel: 412-624-5933 URL: http://www.linguistics.pitt.edu/people/faculty/shirai.htm From d.f.lesley-neuman.2 at hum.leidenuniv.nl Wed Sep 22 18:55:01 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Wed, 22 Sep 2010 20:55:01 +0200 Subject: Variation of speech sounds Message-ID: I know of work that has been done on distinctive features rather than on phonemes: Jeff Mielke (2008) The Emergence of Distinctive Features. Oxford University Press. In this Mielke describes distinctive features as constituting a complex system. If this is indeed so, then like other phenomena in the physical and social sciences, it the system would obey a Zipfian distribution in which the frequency of any one distinctive feature would be inversely proportional to its rank in the frequency table. I do not know if anyone has checked out this hypothesis. Other work was done by the late George Clements. Clements (2009) The Role of Features in Phonological Inventories in Raimy & Cairns (2009). Based on a survey of 451 inventories he proposes principles Feature Bounding and Feature Economy. Unfortunately, I have found from looking at the historical evolution of Kwa languages in Stewart (1971) and those noted in the Nilotic literature that the principle of feature economy does not pan out with regards to historical changes in inventories common to African languages with vowel harmony. San Duanmu at the University of Michigan has been examining Clements proposals in other languages; he and his student Huili Zhang have upheld the feature economy principle in an analysis of the historical evolution of Chinese. This paper was presented last fall at the Michigan Linguistic Society, the program of which can be accessed here: http://ling.lsa.umich.edu/mls2009/program/MLS_Program_Posted.pdf I know that Dr. Duanmu has a very strong interest in this area and would probably gladly participate in a fruitful exchange on this issue. Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BX Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From jkaplan at mail.sdsu.edu Thu Sep 23 22:25:26 2010 From: jkaplan at mail.sdsu.edu (Jeffrey P. Kaplan) Date: Thu, 23 Sep 2010 15:25:26 -0700 Subject: job posting Message-ID: Hi, How can one post a job ad on FUNKnet? Is it as simple as just sending you an e-mail containing the ad copy? Is there a charge? Thanks, Jeff Kaplan -- Jeffrey P. Kaplan Professor of linguistics Dept. of Linguistics & Asian/Middle Eastern Languages San Diego State University San Diego, CA 92182-7727 619-594-5879 fax 619-594-4877 http://www-rohan.sdsu.edu/~jeff315/ This e-mail is a natural product. The slight variations in spelling and grammar enhance its individual character and beauty and in no way are to be considered flaws or defects. From jkaplan at mail.sdsu.edu Fri Sep 24 01:53:00 2010 From: jkaplan at mail.sdsu.edu (Jeffrey P. Kaplan) Date: Thu, 23 Sep 2010 18:53:00 -0700 Subject: Sociolinguistics opening Message-ID: The Department of Linguistics and Asian/Middle Eastern Languages at San Diego State University seeks a tenure-track assistant professor with a specialization in sociolinguistics, to begin in fall 2011. Areas of particular interest include, but are not limited to, variation theory and language contact. Preference will be shown to candidates interested in carrying out field work in the linguistically and culturally diverse speech communities of San Diego. The successful candidate will show a promising research agenda, demonstrated excellence in teaching, and the potential to attract students from across the university. Candidates should send a letter of application, CV, documentation of teaching excellence, and sample publications; and should arrange for three letters of recommendation to be sent. All materials should be sent as e-mail attachments, if possible (MS WORD or .pdf files), to Dr. Jeffrey P. Kaplan, chair, Sociolinguistics Search Committee, at jkaplan at mail.sdsu.edu, or in hard copy form to Dept. of Linguistics & Asian/Middle Eastern Languages, San Diego State University, San Diego, CA 92182-7727. Screening will begin on Nov. 1, 2010 and will continue until the position is filled. San Diego State University is an equal opportunity employer and does not discriminate against persons on the basis of race, religion, national origin, sexual orientation, gender, gender identity and expression, marital status, age, disability, pregnancy, medical condition, or covered veteran status. -- Jeffrey P. Kaplan Professor of linguistics Dept. of Linguistics & Asian/Middle Eastern Languages San Diego State University San Diego, CA 92182-7727 619-594-5879 fax 619-594-4877 http://www-rohan.sdsu.edu/~jeff315/ This e-mail is a natural product. The slight variations in spelling and grammar enhance its individual character and beauty and in no way are to be considered flaws or defects. From rdbusser at gmail.com Fri Sep 24 05:40:45 2010 From: rdbusser at gmail.com (Rik) Date: Fri, 24 Sep 2010 15:40:45 +1000 Subject: Research Centre for Linguistic Typology Podcast Message-ID: (Apologies for cross-posting) The Research Centre for Linguistic Typology at La Trobe University is proud to announce its own podcast on iTunesU: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewiTunesUInstitution?id= 380435656 (Alternatively, go to iTunesU in the iTunes store and search for "Research Centre for Linguistic Typology") In the future, we aim at making all talks given at RCLT freely available to the public through this channel, as long as speakers give permission. At the moment, three are available for download: - Prof. Randy LaPolla, "Preserving Languages of the World" - Prof. Christian Lehmann, "Distance Iconicity" - Prof. Christian Lehmann, "The Function of Numeral Classifiers" Some of them might be of interest to the members of Funknet. Rik ------------------------------ Rik De Busser Research Centre for Linguistic Typology La Trobe University, Bundoora 3086 VIC, Australia T: +61 3 9479 6413 www.rdbusser.com From Florence.Chenu at univ-lyon2.fr Fri Sep 24 20:18:02 2010 From: Florence.Chenu at univ-lyon2.fr (Florence Chenu (FC)) Date: Fri, 24 Sep 2010 22:18:02 +0200 Subject: [conference] 4th International conference of the French Cognitive Linguistics Association (AFLiCo IV; 24-27 May 2011; Lyon, France) Message-ID: AFLiCo IV Fourth International Conference of the French Cognitive Linguistics Association, Lyon, France, 24th-27th May 2011 (French version follows) CONFERENCE THEME of AFLiCo IV The theme of the 2011 conference is: 'Cognitive Linguistics and Typology: Language diversity, variation and change '. This conference aims to bring together linguists engaged in cognitively-oriented research with those working in a functional-typological framework on cross-linguistic variation and on language description. The emphasis will be on (1) language diversity of both spoken and signed languages; (2) inter- and intra-linguistic variation; (3) language change. The conference will bring together linguists working with various methodological approaches and using various kinds of spontaneous and elicited data, including spoken and written corpora, fieldwork data, and experimental data. Proposals are invited for workshops/thematic sessions, for general session papers, and for posters, on topics related to the theme, and on topics in Cognitive Linguistics generally. Papers that report empirically-grounded research on less-studied languages and on typologically, genetically and areally diverse languages will be particularly welcome. Topics include, but are not limited to: - methods and data in cognitive linguistics and in language typology and description - convergence and divergence between cognitive linguistics and functional-typological linguistics - studies from a cognitive and/or typological perspective in phonetics, phonology, morphosyntax, semantics and pragmatics - language variation within and across languages, both spoken and signed - language change from a cognitive and/or typological perspective - language acquisition - studies and advances in construction grammar - language and gesture in cross-linguistic perspective LANGUAGES OF THE CONFERENCE The languages of the conference are French and English. ORAL PRESENTATIONS AND POSTERS Proposals are invited for 30-minute slots (20-minute presentation plus question time) in the general sessions and for posters (A1 size). WORKSHOPS, INCLUDING THEMATIC SESSIONS Proposals are invited for half-day or full-day workshops/thematic sessions. Each workshop proposal should contain the following information: - the names and contact details of two workshop organizers - the title of the proposed workshop - an overview of the topic and aims of the workshop (up to 2 pages) - an indication of the desired schedule (number of slots: 4, 6 or 10; half day or full day; number and nature of presentations, discussions, round tables, etc. that the workshop will comprise). Note that, within a workshop, each presentation, discussion or round table will occupy one 30-minute slot in parallel with one general session slot. - an abstract (consistent with the indications below under 'Submission procedure') for each proposed 30-minute presentation Workshop proposals will be refereed in the same way as general session and poster proposals. SUBMISSION PROCEDURE Proposals should be submitted online following the instructions to be found at the following address: http://www.ddl.ish-lyon.cnrs.fr/colloques/AFLICO_IV/ Author information (name, affiliation, email address) will be required on the submission website. An author may submit a maximum of two abstracts, of which at least one must be co-authored. In the case of co-authored abstracts, the first-named author will be the contact person. Abstracts will be anonymously reviewed and notification of acceptance will be sent out from 25th February 2011. The anonymous abstracts must be in 12 point Times or Times New Roman font, formatted for A4 or US Letter size paper with margins of 2.5 cm or 1 inch. The maximum length for the text of the abstract is one page; a second page may be used only for figures, glossed examples and bibliographical references. Deadline for general session papers: 22nd December 2010 Deadline for workshops/thematic sessions: 18th December 2010 ******************************************************* ******************************************************* AFLiCo IV Quatrième Colloque International de l’Association Française de Linguistique Cognitive Lyon, France, 24-27 Mai 2011 THÈME DU COLLOQUE AFLiCo IV ‘Linguistique cognitive et typologie : diversité des langues, variation et changement’. L’objectif de ce colloque est de réunir des linguistes travaillant dans le domaine de la linguistique cognitive et/ou dans le domaine de la linguistique fonctionnelle-typologique sur la variation inter-linguistique et la description des langues. L’accent du colloque sera mis sur (1) la diversité des systèmes linguistiques aussi bien oraux que signés, (2) la variation qui s’opère sur les plans inter- et intra- linguistiques et (3) les changements des systèmes linguistiques. Dans cette perspective, le colloque rassemblera des chercheurs qui travaillent sur des terrains linguistiques variés, qui abordent leur objet d’étude dans une perspective synchronique et/ou diachronique et qui utilisent différentes méthodes et différents types de données telles que des données spontanées ou élicitées, y compris orales ou écrites, des données de terrain ou encore des données expérimentales. Nous attendons des propositions de sessions thématiques, des propositions de présentations orales de sessions générales et de posters sur des problématiques en lien avec le thème du colloque et dans le domaine de la linguistique cognitive en général. Les propositions portant sur des langues moins bien décrites et des langues qui varient du point de vue typologique, génétique et aréal seront particulièrement appréciées. Les thématiques incluent, mais ne se limitent pas aux suivantes : - méthodes et données en linguistique cognitive, typologie et description des langues ; - convergence et divergence entre linguistique cognitive et linguistique fonctionnelle-typologique ; - études menées dans une perspective cognitive et/ou typologique dans les domaines de la phonétique, phonologie, morphosyntaxe, sémantique et pragmatique ; - variation inter- et intra-linguistique dans les langues parlées et les langues signées ; - changements linguistiques dans une perspective cognitive et/ou typologique ; - acquisition du langage ; - recherches et avancées dans le domaine de la grammaire des constructions ; - langue et geste dans une perspective inter-linguistique. LANGUES OFFICIELLES DU COLLOQUE Les deux langues du colloque sont le français et l’anglais. COMMUNICATIONS ET POSTERS Nous invitons des propositions de communication aux sessions générales de 30 minutes (20 minutes de présentation et 10 minutes de questions) et des propositions de posters (format A1). ATELIERS ET SESSIONS THÉMATIQUES Nous accueillons des propositions d’une demi-journée ou d’une journée entière pour des ateliers et/ou sessions thématiques. Ces ateliers/sessions thématiques doivent être proposés par deux organisateurs. Chaque proposition doit inclure les informations suivantes : - les noms et les coordonnées des deux organisateurs - le titre de la session - une présentation du thème et des objectifs de la session (2 pages maximum) - une précision concernant le temps souhaité (nombre de créneaux horaires : 4, 6 ou 10 ; une journée ou une journée entière ; nombre et nature des présentations, discussions, tables rondes, etc.). - un résumé d’une page pour chaque présentation (une deuxième page peut être utilisée pour des figures, exemples glosés et références bibliographiques) Les propositions d’ateliers et/ou de sessions thématiques seront soumises à la même procédure d’évaluation que les propositions pour les sessions générales et les posters. La notification d’acceptation sera envoyée aux deux organisateurs à partir du 25 février 2011. SOUMISSION DES PROPOSITIONS Les propositions seront soumises en ligne suivant les instructions indiquées à l’adresse suivante : http://www.ddl.ish-lyon.cnrs.fr/colloques/AFLICO_IV/ Un auteur ne peut soumettre que deux propositions de communication dont une au moins devrait être en co-auteur. Les informations concernant l’auteur (nom, affiliation, adresse email) seront requises lors de la soumission en ligne mais les propositions seront évaluées de façon anonyme. Dans le cas des propositions en co-auteur le premier auteur sera la personne référente/contact. Les propositions seront examinées de façon anonyme par 2 membres experts du comité scientifique. La notification d’acceptation sera envoyée aux auteurs à partir du 25 février 2011. Les propositions ne devront pas dépasser une page. Une deuxième page peut être utilisée pour des figures, exemples glosés et références bibliographiques. Format des propositions : papier A4, marges 2,5 cm, police Times ou Times New Roman. Date limite pour les sessions générales : 22 décembre 2010 Date limite pour les sessions thématiques : 18 décembre 2010 From mg246 at cornell.edu Sat Sep 25 22:17:01 2010 From: mg246 at cornell.edu (monica gonzalez-marquez) Date: Sun, 26 Sep 2010 00:17:01 +0200 Subject: CFP: EMCL 5.1 - Freiburg, Germany Message-ID: Empirical Methods in Cognitive Linguistics 5.1 -- Freiburg **** March 6 -- 11, 2011 https://sites.google.com/site/emcl5freiburg/ --------------------Application deadline: December 15, 2010---------------------- We invite applications for the, 5th Empirical Methods in Cognitive Linguistics workshop, to be held in, Freiburg, Germany, March 6 -- 11, 2011 The goal of EMCL is to facilitate dialogue among language researchers with different methodological backgrounds, i.e. theorists, experimentalists, corpus linguists, etc. We do this by creating an environment where specialists learn from each other by developing a research project together where their various skills are combined. Intended audience: Language researchers with an embodiment, situated cognition and/or cognitive linguistics background. No prior experimental or corpus training is required though an understanding of the theoretical issues is necessary. Participants can be at different early stages in their careers, i.e. graduate students, post-grads, post-docs, junior faculty, etc. Format: During the course of a week, participants will join one of 5 hands-on mini-labs. Each mini-lab will be responsible for completing a joint research project. A select group of students (max. 8 per group for a total of 40***) will be invited to participate. Each group will work with two researchers who will guide the group in selecting an idea for the group to investigate, structuring and organizing a research project, and carrying it out. The session will end with the presentation of findings and a general discussion. Topics to be covered include, - Deciding on a research topic - Transforming the research topic into a research question - Developing experimental hypotheses and designing an experiment - Data collection - Statistical analysis and interpretation - Presentation of findings to an audience Workshop Faculty Group 1: Rolf Zwaan University of Rotterdam Interests: The relationship between cognition and systems of perception, action, and emotion, Language comprehension, Embodied cognition http://www.brain-cognition.eu/index.html?personal.php?id=Zwaan Alan Cienki, Vrije Universiteit Interests: cognitive linguistics, pragmatics, spoken language, gesture, political discourse, contrastive linguistics http://www.let.vu.nl/en/about-the-faculty/academic-staff/staff-listed-alphabetically/staff-a-d/dr-a-cienki/index.asp Group 2: Kenny Coventry, Northumbria University Interests: language and perception, spatial language, embodiment, decision making http://kenny.coventry.googlepages.com/home Katharina Rohlfing, Bielefeld University Interests: emergentist semantics, early literacy, human-machine interaction, rhetoric and communication https://www.techfak.uni-bielefeld.de/~rohlfing/website/data/index.html http://www.cit-ec.de/es Group 3: Lars Konieczny, University of Freiburg Interests: Theoretical, Empirical, and Computational Psycholinguistics, Eye-movements research, Reading, Spoken language comprehension in the Visual-World-paradigm, Spatial reasoning and wayfinding, Cognitive modeling (ACT-R, Connectionist Modeling), Embodied Cognition http://portal.uni-freiburg.de/cognition/Members/konieczny Michele Feist, University of Louisiana at Lafayette Interests: lexical semantics, spatial language, psycholinguistics, acquisition of semantics, language and cognition http://www.ucs.louisiana.edu/~mif8232/ Group 4: Seana Coulson, University of California, San Diego Interests: Conceptual Blending, Joke Comprehension, Metaphor, Analogical Reasoning, Concept Combination, Sentence Processing http://www.cogsci.ucsd.edu/~coulson/ Panos Athanasopoulos, Bangor University Interests: Bilingualism and Cognition, Language and Thought, Emotion, Language Acquisition, conceptual development http://www.bangor.ac.uk/linguistics/about/panos.php.en Group 5: Pia Knoeferle, Bielefeld University Interests: Influences of visual context on real-time language comprehension, Picture-sentence verification, Models and processing accounts of situated language comprehension http://wwwhomes.uni-bielefeld.de/pknoeferle/Staff_PK/PK.html Anatol Stefanowitsch, University of Hamburg Interests: Encoding of motion events, Second language research, Construction Grammar, Quantitative Corpus Linguistics, Metaphor, Negative evidence http://www1.uni-hamburg.de/stefanowitsch/ Accommodation: Accommodation at walking distance to the university will be arranged for all student participants. Cost will be EUR20 per night. (We "may" receive funds to cover student accommodation, in which case all applicants will be notified.) Participation Fee: EUR125 **, payable by bank transfer or upon arrival by prior arrangement. (This fee helps cover the costs of organization and faculty travel.) Application: To apply, please send the following by December 15, 2010. All materials must be submitted electronically to emcl5.freiburg (at) googlemail.com PLEASE WRITE 'APPLICATION' IN THE SUBJECT LINE. 1. A maximum of two (2) pages, (1000 words), describing, - your background, - your reasons for wanting to participate, - the research group you would like to work in and why. Please include in this section a brief description of your research interests. All topics listed above must be addressed. Incomplete applications will not be reviewed. 2. A copy of your curriculum vitae. The application deadline is December 15, 2010 Accepted applicants will be notified on or before January 15, 2011 This workshop is supported by: the FRIAS at Freiburg University http://www.frias.uni-freiburg.de/ the Research training group (GRK DFG 1624/1) Frequency effects in language http://frequenz.uni-freiburg.de/abstract&language=de and the DFG (pending). www.dfg.de ** 2 (two) tuition scholarships will be awarded by lottery to students traveling from Eastern Europe and 3rd world countries. Please state in your application whether you would like to be included in the lottery. *** Please note: Attendance is strictly limited to invited participants. No exceptions will be made so as to preserve pedagogical integrity. **** EMCL 5.2 will be held in Chicago, USA, June 2011 with a different set of faculty. That notice will follow in January, 2011. --- EMCL 5 Organizing Committee: Monica Gonzalez-Marquez, Chair, Cornell University Martin Hilpert, University of Freiburg Pia Knoeferle, Bielefeld University Lars Konieczny, University of Freiburg From hallowel at ohio.edu Mon Sep 27 18:34:19 2010 From: hallowel at ohio.edu (Hallowell, Brooke) Date: Mon, 27 Sep 2010 14:34:19 -0400 Subject: indexing word difficulty Message-ID: Dear colleagues: We are working on deriving a list of "easy" and "difficult" words for an auditory comprehension study in American English. As we're unaware of databases of words indexed by "difficulty," we are, as many have done, approximating an index of difficulty by using multiple indices, including word length, familiarity, imageability, and frequency. Do know of a more index of word difficulty that more explicitly taps how hard it is to understand a word? Thanks very much for any suggestions. Brooke Hallowell Brooke Hallowell, Ph.D., CCC-SLP, F-ASHA President, Council of Academic Programs in Communication Sciences and Disorders Professor and Director, Neurolinguistics Laboratory Communication Sciences and Disorders School of Rehabilitation and Communication Sciences College of Health Sciences and Professions W237 Grover Center Ohio University Athens, OH 45701 USA From crm5 at rice.edu Mon Sep 27 22:55:32 2010 From: crm5 at rice.edu (crm5 at rice.edu) Date: Mon, 27 Sep 2010 17:55:32 -0500 Subject: Call for papers: Rice Working Papers in Linguistics Vol 3 Message-ID: Dear Funknet members, The Rice Working Papers in Linguistics is currently soliciting submissions for its third volume (you can see published volumes at http://owling.blogs.rice.edu/rwpl-vol-1/ and http://owling.blogs.rice.edu/rwpl-vol-2/). The deadline is November 15th. Please see the guidelines below and consider submitting your work to rwpl at rice.edu. Carlos Molina-Vital Rice Working Papers in Linguistics EIC *** Rice Working Papers in Linguistics, Volume 3 Deadline: **November 15th, 2010** The Rice Linguistics Society (RLS) solicits submissions from all subfields of linguistics (with the exception of ESL/TESOL and related areas of applied linguistics) for publication in the Rice Working Papers in Linguistics. Students and post-docs are strongly encouraged to submit. We especially welcome submissions in line with our department's focus on functional, usage-based approaches to language study using empirical data, including but not limited to the following: -cognitive/functional linguistics -typology and language universals -field studies in less commonly researched languages -sociolinguistics, including sociophonetics -phonetics and speech processing -laboratory phonology -forensic linguistics -corpus linguistics -discourse -neurolinguistics -psycholinguistics and language processing -language change and grammaticalization Submitted papers must meet the following minimum style requirements: -recommended length 15-25 pages (normally 5000-8000 words); significantly longer or shorter papers will be considered on a case-by-base basis (contact the editorial board) -For comprehensive details on format (such as font, margins, examples, references, etc.) please refer to the RWPL template available on the Style sheet link at http://owling.blogs.rice.edu/ -submit an abstract (maximum 500 words), including 3-5 keywords, as a separate Word file -submit two copies (in addition to your abstract): (1) one copy in Word (2003 or 2007) (2) in addition to the Word submission, you must send a PDF version to ensure fonts are preserved RLS accepts only electronic submissions for the working papers. These must be sent to rwpl at rice.edu, and the body of the e-mail should include: -title of paper -name of author(s) -affiliation -address -phone number -contact e-mail address The deadline for receipt of submissions is **November 15th, 2010**. Questions regarding the submissions process or style requirements may be addressed to the editorial board at rwpl at rice.edu. ******* From yutamb at mail.ru Wed Sep 29 09:01:31 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Wed, 29 Sep 2010 16:01:31 +0700 Subject: Variation of different functions of gerund and participle Message-ID: Dear Funknet List members, do you use the coefficient of variation and Chi-square to study functioning of phonemes or prepositions in language? In fact, the application of coefficient of variation and Chi-square to investigate the variation of linguistic elements in language may stop endless debates about language variation problems. I use them all right. They proved quite useful. With their help I also studied variation of phomenes and groups of phonemes (labials, velars, sonorants, fricatives, etc.) across languages. Usually they were used to study the variation of phonemes in texts. I did that as well. The degree of variation of different functions of gerund and participle allowed us to distinquish between authors. It contributed much to the theory of authorship. I wonder if you read my publications? Who is researching in the same area? I call this area phonostatistical typology. Looking forward to hearing from you soon to yutamb at mail.ru Yours sincerely Yuri Tambovtsev, Novosibirsk, Russia From yutamb at mail.ru Wed Sep 29 20:10:52 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Thu, 30 Sep 2010 03:10:52 +0700 Subject: Different authors use different functions differently Message-ID: Dear Funknet colleagues, if corpora of English texts is not one unity but a conglomerate, how can we use statistic criteria? I started the discussion what tools to use for the analysis because many of linguists do not use all these complex statistical packets correctly. The other thing is who knows what and how the data are being analysed in them. We have a sort of a "black box" which has the entrance in which you put your data and the outcome where you receive your results. You must be quite sure that the data are analysed correctly. The more simple criteria you use, the better. This is why I stopped using all the stat. packets and began using very simple criteria like the coefficient of variation, the Chi-square and the t-test. At least, I am sure about the outcome. More often than not, it is advisable to try you data manually than to use the statpack. Looking forward to hearing from you if you agree with me that using statpacks may give you strange results. The fact that so many scholar spoke on the list about how to use the statpacks showed that it is a burning question for linguists and other researchers who are not specialists in mathematics. Yours sincerely Yuri Tambovtsev, yutamb at mail.ru From tgivon at uoregon.edu Thu Sep 30 06:03:25 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Thu, 30 Sep 2010 00:03:25 -0600 Subject: A poet slips Message-ID: A GREAT POET CAN STILL GET IT WRONG I came to know of Salman Masalha ("Arabs, speak Hebrew!", Haaretz/English, International Herald Tribune 9-27-10) accidentally by stumbling a on his truly great quote: "All fixed identities are imposed from the outside. Whoever has a clear identity knows it can assume multiple forms". In the context of Palestine/Israel, what a breath of fresh, rare clarity. Still, like the rest of us mortals, a great poet can on occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got only one third of the story right. At first, he was led astray be the academic researchers he cited, who claimed that the lagging reading skills of Israeli-Arab students is correlated to lagging R-hemisphere activity, then explained this neurological lag by suggesting that the Arab script requires more contextual analysis. But it is the R-hemisphere of the human cortex that is more context oriented, less automated. If Arab-reading students required more contextual labor, it should have been registered as a higher R-hemisphere activity, not lower. Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among many others, have established beyond reasonable doubt that written words in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere module on the boundary of the occipital and temporal lobes (just past Brodman's Area 19), along the ventral visual object-recognition 'stream' that flows from the back to the front of the L-cortex. And the L-cortex is in general responsible for the more automated--less context-dependent--processing of language (as well as visual, motor and other skills). The visual word-recognition module is, in turn, recruited from the pre-existing visual object-recognition ventral-stream module. A considerable amount of life-time practice and repetition is required to affect this late-cultural adaptation. The human brain is not (yet) genetically configured at birth for visual-word recognition, only for visual-object recognition. A similar cultural adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) in the L-pareita lobe, an area originally configured for analysis of object-location in space. Mr. Masalha then, on his own, points out to a more plausible right answer: Arab students, in Israel as well as all over the Arab world, are not taught literacy in their native language (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 years or more. That is, in a foreign language. The discrepancy would be just as great if Israeli kids were taught their Hebrew literacy first in the language of Genesis; or if French students were taught literacy first in the language of La Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or English-speaking kids in the language of Beowolf. As far as my frail guessing powers go, remedying the situation would be much easier by combining two well-known verities of second language acquisition: (a) Teach them both early, together--'co-ordinated bilingualism'. And (b), teach literacy first in the student's spoken native language; only then gradually 'stretch' it to more literary genres. This method, bhy the way, was suggested in the late 1930's by no other than L. Bloomfield, in a book outlining a 'phonics-first ' literacy program for native English speakers. Rather than depriving Israeli-Arab students of literacy in their own--equally glorious--native language, just teach them smart. For his last culprit, the presumed--tho hardly unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized prejudices of the educated classes. While readily endorsing his aesthetic sentiments about modern media, I would still like to point out that the 'vulgar' genre is much closer to the Arab students' spoken native language, and if anything should facilitate the easier initial acquisition of native-language literacy. Respectuosamente, ma'-salaam, T. Givón From john at research.haifa.ac.il Thu Sep 30 08:06:57 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 10:06:57 +0200 Subject: A poet slips In-Reply-To: <4CA4282D.6080303@uoregon.edu> Message-ID: Tom, I've been trying for several years to get Israeli Arabs and Jews to seriously consider the possibility of educating Israeli Arabs in a written version of their spoken language, as you suggest, at least through 3rd grade. I can send you some things I've written on this topic, with a lot of cross-linguistic data. But thus far, it isn't working. As with most aspects of 'the situation' here, politically active Arabs think that the solution is to pursue the same self-destructive strategy which they've been following for almost a century but with even more vigor and steadfastness (e.g. Masalha points the blame at the Arab media because they do not 'provide the linguistic richness of formal Arabic') while Jews are basically content to let the Arabs stew in their own juices so that they can reap the benefits. There may, however, be hope for the future in the form of the radically increased usage of written forms of colloquial Arabic dialects in electronic media such as Facebook, blogs, emails, etc., by Arabic speakers below the age of 30, which will soon come to seriously threaten the status of classical Arabic in the same way that the invention of the printing press overturned the linguistic hierarchy in Western Europe by overthrowing Latin. I'm working on this too. We'll see what happens. I realize that Masalha as well as the researchers he referred to don't necessarily know what they're talking about regarding the brain. But the general point is still potentially significant--that the connected script which Arabic uses as well as the multiple forms which many Arabic letters have may be a significant obstacle to literacy, however this may be related to processes in the brain. Best wishes, John Quoting Tom Givon : > > > A GREAT POET CAN STILL GET IT WRONG > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > Haaretz/English, International Herald Tribune 9-27-10) accidentally > by stumbling a on his truly great quote: > "All fixed identities are imposed from the outside. > Whoever has a clear identity knows it can assume > multiple forms". > In the context of Palestine/Israel, what a breath of fresh, rare > clarity. Still, like the rest of us mortals, a great poet can on > occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > intended--surely got only one third of the story right. At first, he was > led astray be the academic researchers he cited, who claimed that the > lagging reading skills of Israeli-Arab students is correlated to > lagging R-hemisphere activity, then explained this neurological lag > by suggesting that the Arab script requires more contextual analysis. > But it is the R-hemisphere of the human cortex that is more context > oriented, less automated. If Arab-reading students required more > contextual labor, it should have been registered as a higher > R-hemisphere activity, not lower. > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > among many others, have established beyond reasonable doubt that > written words in all languages (English, Mandarin, Hindi, Amharic, > Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > module on the boundary of the occipital and temporal lobes (just past > Brodman's Area 19), along the ventral visual object-recognition > 'stream' that flows from the back to the front of the L-cortex. And > the L-cortex is in general responsible for the more automated--less > context-dependent--processing of language (as well as visual, motor and > other skills). The visual word-recognition module is, in turn, > recruited from the pre-existing visual object-recognition > ventral-stream module. A considerable amount of life-time practice and > repetition is required to affect this late-cultural adaptation. The > human brain is not (yet) genetically configured at birth for > visual-word recognition, only for visual-object recognition. A similar > cultural adaptation, this one for math, has been shown for (Dahaene & > Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > in the L-pareita lobe, an area originally configured for analysis of > object-location in space. > Mr. Masalha then, on his own, points out to a more > plausible right answer: Arab students, in Israel as well as all over the > Arab world, are not taught literacy in their native language > (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > instrument harking back 1,400 years or more. That is, in a foreign > language. The discrepancy would be just as great if Israeli kids were > taught their Hebrew literacy first in the language of Genesis; or if > French students were taught literacy first in the language of La > Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or > English-speaking kids in the language of Beowolf. As far as my frail > guessing powers go, remedying the situation would be much easier by > combining two well-known verities of second language acquisition: (a) > Teach them both early, together--'co-ordinated bilingualism'. And (b), > teach literacy first in the student's spoken native language; only then > gradually 'stretch' it to more literary genres. This method, bhy the > way, was suggested in the late 1930's by no other than L. Bloomfield, > in a book outlining a 'phonics-first ' literacy program for native > English speakers. Rather than depriving Israeli-Arab students of > literacy in their own--equally glorious--native language, just teach > them smart. > For his last culprit, the presumed--tho hardly > unique--vulgarity of Arab media, Mr. Masalha lapses into well > recognized prejudices of the educated classes. While readily endorsing > his aesthetic sentiments about modern media, I would still like to point > out that the 'vulgar' genre is much closer to the Arab students' > spoken native language, and if anything should facilitate the easier > initial acquisition of native-language literacy. Respectuosamente, > ma'-salaam, > > > T. Givón > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From tgivon at uoregon.edu Thu Sep 30 08:25:21 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Thu, 30 Sep 2010 02:25:21 -0600 Subject: A poet slips In-Reply-To: <1285834017.4ca445214a9c7@webmail.haifa.ac.il> Message-ID: When I learned to read Arabic on the kibbutz (Maabarot) as a child, our textbook was of written COLLOQUIAL Falastini Arabic. We never got to Classical (after 1949 things changed...). That book may still exist, you might track it down. It was easy, a cinch really. I concluded it was really just Hebrew with a few trivial transformations in Phonology & Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat on a few sessions of a faculty study group at UO who were trying to learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, it was like pullin' teeth. But Leonard Bloomfield said it already, and well, in 1939 (or was it 1943?). Cheers, TG ============== john at research.haifa.ac.il wrote: > Tom, > I've been trying for several years to get Israeli Arabs and Jews to seriously > consider the possibility of educating Israeli Arabs in a written version of > their spoken language, as you suggest, at least through 3rd grade. I can send > you some things I've written on this topic, with a lot of cross-linguistic > data. But thus far, it isn't working. As with most aspects of 'the situation' > here, politically active Arabs think that the solution is to pursue the same > self-destructive strategy which they've been following for almost a century but > with even more vigor and steadfastness (e.g. Masalha points the blame at the > Arab media because they do not 'provide the linguistic richness of formal > Arabic') while Jews are basically content to let the Arabs stew in their own > juices so that they can reap the benefits. There may, however, be hope for the > future in the form of the radically increased usage of written forms of > colloquial Arabic dialects in electronic media such as Facebook, blogs, emails, > etc., by Arabic speakers below the age of 30, which will soon come to seriously > threaten the status of classical Arabic in the same way that the invention of > the printing press overturned the linguistic hierarchy in Western Europe by > overthrowing Latin. I'm working on this too. We'll see what happens. > > I realize that Masalha as well as the researchers he referred to don't > necessarily know what they're talking about regarding the brain. But the > general point is still potentially significant--that the connected script which > Arabic uses as well as the multiple forms which many Arabic letters have may be > a significant obstacle to literacy, however this may be related to processes in > the brain. > Best wishes, > John > > > > Quoting Tom Givon : > > >> A GREAT POET CAN STILL GET IT WRONG >> >> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >> Haaretz/English, International Herald Tribune 9-27-10) accidentally >> by stumbling a on his truly great quote: >> "All fixed identities are imposed from the outside. >> Whoever has a clear identity knows it can assume >> multiple forms". >> In the context of Palestine/Israel, what a breath of fresh, rare >> clarity. Still, like the rest of us mortals, a great poet can on >> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >> intended--surely got only one third of the story right. At first, he was >> led astray be the academic researchers he cited, who claimed that the >> lagging reading skills of Israeli-Arab students is correlated to >> lagging R-hemisphere activity, then explained this neurological lag >> by suggesting that the Arab script requires more contextual analysis. >> But it is the R-hemisphere of the human cortex that is more context >> oriented, less automated. If Arab-reading students required more >> contextual labor, it should have been registered as a higher >> R-hemisphere activity, not lower. >> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >> among many others, have established beyond reasonable doubt that >> written words in all languages (English, Mandarin, Hindi, Amharic, >> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >> module on the boundary of the occipital and temporal lobes (just past >> Brodman's Area 19), along the ventral visual object-recognition >> 'stream' that flows from the back to the front of the L-cortex. And >> the L-cortex is in general responsible for the more automated--less >> context-dependent--processing of language (as well as visual, motor and >> other skills). The visual word-recognition module is, in turn, >> recruited from the pre-existing visual object-recognition >> ventral-stream module. A considerable amount of life-time practice and >> repetition is required to affect this late-cultural adaptation. The >> human brain is not (yet) genetically configured at birth for >> visual-word recognition, only for visual-object recognition. A similar >> cultural adaptation, this one for math, has been shown for (Dahaene & >> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >> in the L-pareita lobe, an area originally configured for analysis of >> object-location in space. >> Mr. Masalha then, on his own, points out to a more >> plausible right answer: Arab students, in Israel as well as all over the >> Arab world, are not taught literacy in their native language >> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >> instrument harking back 1,400 years or more. That is, in a foreign >> language. The discrepancy would be just as great if Israeli kids were >> taught their Hebrew literacy first in the language of Genesis; or if >> French students were taught literacy first in the language of La >> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or >> English-speaking kids in the language of Beowolf. As far as my frail >> guessing powers go, remedying the situation would be much easier by >> combining two well-known verities of second language acquisition: (a) >> Teach them both early, together--'co-ordinated bilingualism'. And (b), >> teach literacy first in the student's spoken native language; only then >> gradually 'stretch' it to more literary genres. This method, bhy the >> way, was suggested in the late 1930's by no other than L. Bloomfield, >> in a book outlining a 'phonics-first ' literacy program for native >> English speakers. Rather than depriving Israeli-Arab students of >> literacy in their own--equally glorious--native language, just teach >> them smart. >> For his last culprit, the presumed--tho hardly >> unique--vulgarity of Arab media, Mr. Masalha lapses into well >> recognized prejudices of the educated classes. While readily endorsing >> his aesthetic sentiments about modern media, I would still like to point >> out that the 'vulgar' genre is much closer to the Arab students' >> spoken native language, and if anything should facilitate the easier >> initial acquisition of native-language literacy. Respectuosamente, >> ma'-salaam, >> >> >> T. Givón >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > From john at research.haifa.ac.il Thu Sep 30 08:40:23 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 10:40:23 +0200 Subject: A poet slips In-Reply-To: <4CA44971.2050105@uoregon.edu> Message-ID: Learning the classical language is like pulling teeth for Arabs kids also. They just can't publicly say it because that would make them bad Arabs. It's just ridiculous. I have an even better plan for the Jewish kids--teach them to use the written version of the spoken language which Arab kids are using for Facebook. That's how kids make friends these days anyway. Best wishes, John Quoting Tom Givon : > > > When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > textbook was of written COLLOQUIAL Falastini Arabic. We never got to > Classical (after 1949 things changed...). That book may still exist, > you might track it down. It was easy, a cinch really. I concluded it was > really just Hebrew with a few trivial transformations in Phonology & > Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > on a few sessions of a faculty study group at UO who were trying to > learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > it was like pullin' teeth. But Leonard Bloomfield said it already, and > well, in 1939 (or was it 1943?). > Cheers, TG > > ============== > > > john at research.haifa.ac.il wrote: > > Tom, > > I've been trying for several years to get Israeli Arabs and Jews to > seriously > > consider the possibility of educating Israeli Arabs in a written version of > > their spoken language, as you suggest, at least through 3rd grade. I can > send > > you some things I've written on this topic, with a lot of cross-linguistic > > data. But thus far, it isn't working. As with most aspects of 'the > situation' > > here, politically active Arabs think that the solution is to pursue the > same > > self-destructive strategy which they've been following for almost a century > but > > with even more vigor and steadfastness (e.g. Masalha points the blame at > the > > Arab media because they do not 'provide the linguistic richness of formal > > Arabic') while Jews are basically content to let the Arabs stew in their > own > > juices so that they can reap the benefits. There may, however, be hope for > the > > future in the form of the radically increased usage of written forms of > > colloquial Arabic dialects in electronic media such as Facebook, blogs, > emails, > > etc., by Arabic speakers below the age of 30, which will soon come to > seriously > > threaten the status of classical Arabic in the same way that the invention > of > > the printing press overturned the linguistic hierarchy in Western Europe by > > overthrowing Latin. I'm working on this too. We'll see what happens. > > > > I realize that Masalha as well as the researchers he referred to don't > > necessarily know what they're talking about regarding the brain. But the > > general point is still potentially significant--that the connected script > which > > Arabic uses as well as the multiple forms which many Arabic letters have > may be > > a significant obstacle to literacy, however this may be related to > processes in > > the brain. > > Best wishes, > > John > > > > > > > > Quoting Tom Givon : > > > > > >> A GREAT POET CAN STILL GET IT WRONG > >> > >> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >> by stumbling a on his truly great quote: > >> "All fixed identities are imposed from the outside. > >> Whoever has a clear identity knows it can assume > >> multiple forms". > >> In the context of Palestine/Israel, what a breath of fresh, rare > >> clarity. Still, like the rest of us mortals, a great poet can on > >> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >> intended--surely got only one third of the story right. At first, he was > >> led astray be the academic researchers he cited, who claimed that the > >> lagging reading skills of Israeli-Arab students is correlated to > >> lagging R-hemisphere activity, then explained this neurological lag > >> by suggesting that the Arab script requires more contextual analysis. > >> But it is the R-hemisphere of the human cortex that is more context > >> oriented, less automated. If Arab-reading students required more > >> contextual labor, it should have been registered as a higher > >> R-hemisphere activity, not lower. > >> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >> among many others, have established beyond reasonable doubt that > >> written words in all languages (English, Mandarin, Hindi, Amharic, > >> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >> module on the boundary of the occipital and temporal lobes (just past > >> Brodman's Area 19), along the ventral visual object-recognition > >> 'stream' that flows from the back to the front of the L-cortex. And > >> the L-cortex is in general responsible for the more automated--less > >> context-dependent--processing of language (as well as visual, motor and > >> other skills). The visual word-recognition module is, in turn, > >> recruited from the pre-existing visual object-recognition > >> ventral-stream module. A considerable amount of life-time practice and > >> repetition is required to affect this late-cultural adaptation. The > >> human brain is not (yet) genetically configured at birth for > >> visual-word recognition, only for visual-object recognition. A similar > >> cultural adaptation, this one for math, has been shown for (Dahaene & > >> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >> in the L-pareita lobe, an area originally configured for analysis of > >> object-location in space. > >> Mr. Masalha then, on his own, points out to a more > >> plausible right answer: Arab students, in Israel as well as all over the > >> Arab world, are not taught literacy in their native language > >> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >> instrument harking back 1,400 years or more. That is, in a foreign > >> language. The discrepancy would be just as great if Israeli kids were > >> taught their Hebrew literacy first in the language of Genesis; or if > >> French students were taught literacy first in the language of La > >> Chanחon de Roland, Guilhome de Machaut, or Chrך tien de Troyes. Or > >> English-speaking kids in the language of Beowolf. As far as my frail > >> guessing powers go, remedying the situation would be much easier by > >> combining two well-known verities of second language acquisition: (a) > >> Teach them both early, together--'co-ordinated bilingualism'. And (b), > >> teach literacy first in the student's spoken native language; only then > >> gradually 'stretch' it to more literary genres. This method, bhy the > >> way, was suggested in the late 1930's by no other than L. Bloomfield, > >> in a book outlining a 'phonics-first ' literacy program for native > >> English speakers. Rather than depriving Israeli-Arab students of > >> literacy in their own--equally glorious--native language, just teach > >> them smart. > >> For his last culprit, the presumed--tho hardly > >> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >> recognized prejudices of the educated classes. While readily endorsing > >> his aesthetic sentiments about modern media, I would still like to point > >> out that the 'vulgar' genre is much closer to the Arab students' > >> spoken native language, and if anything should facilitate the easier > >> initial acquisition of native-language literacy. Respectuosamente, > >> ma'-salaam, > >> > >> > >> T. Givףn > >> > >> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From cbutler at ntlworld.com Thu Sep 30 09:51:34 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Thu, 30 Sep 2010 10:51:34 +0100 Subject: Different authors use different functions differently In-Reply-To: <174B792CEB644021811F83E8342A0A97@ngufa28a6c2639> Message-ID: Dear Yuri, Statistical packages are indeed sometimes quite difficult for linguists and other researchers who are not statisticians. But they are extremely useful - sometimes essential - and it is worth getting to know how to use at least one of them, for anyone whose work is primarily quantitative. One of the most important aspects of using statistics in linguistic research is knowing the basis for the various measures and tests that are available, and what criteria you can use in order to decide which measure or test is appropriate for your data. This is something you need to know, whether you are using a statistical package such as SPSS or doing the calculations manually. However, although it is not too hard to do the calculations manually for simple tests such as chi square or the t-test, it is much harder for more advanced tests, and almost impossible for some techniques such as various kinds of multivariate analysis on large data sets, which require the use of computing power. Manual analysis is error-prone, and this is avoided by using a computer package. On the other hand, the statistical packages often give you a lot of output, some of it very important, other parts of it less so, so that you have to know how to interpret the output. There are some good books around to help researchers to get to grips with the standard packages. For SPSS I would recommend the 3rd edition of Andy Field's 'Discovering Statistics Using SPSS' (Sage Publications, 2009), which is comprehensive and detailed, but well explained and even entertaining. There is also the statistical programming language R, which is being used by an increasing number of adventurous quantitative linguists. For this, there are at least two recent books: Stefan Gries' 'Statistics for Linguists with R: A Practical Introduction' (Routledge [Taylor and Francis Group], 2009), and Harald Baayen's 'Analyzing Linguistic Data. A Practical Introduction to Statistics Using R' (Cambridge University Press 2008), which is rather difficult in parts, but useful nevertheless. R can be downloaded freely from http://www.r-project.org/, and free packages with some of the functionality of the standard ones can be found (e.g. PSPP from http://www.gnu.org/software/pspp/, which I have not tried myself), and may be useful for researchers who do not have access, through their institutions, to the (expensive) standard packages. There are also books which are introductions to statistics for linguists, but are not linked to specific packages. If you are interested in details of these, please get in touch with me directly. Best wishes, Chris Butler -------------------------------------------------- From: "Yuri Tambovtsev" Sent: Wednesday, September 29, 2010 9:10 PM To: Subject: [FUNKNET] Different authors use different functions differently > Dear Funknet colleagues, if corpora of English texts is not one unity but > a conglomerate, how can we use statistic criteria? I started the > discussion what tools to use for the analysis because many of linguists do > not use all these complex statistical packets correctly. The other thing > is who knows what and how the data are being analysed in them. We have a > sort of a "black box" which has the entrance in which you put your data > and the outcome where you receive your results. You must be quite sure > that the data are analysed correctly. The more simple criteria you use, > the better. This is why I stopped using all the stat. packets and began > using very simple criteria like the coefficient of variation, the > Chi-square and the t-test. At least, I am sure about the outcome. More > often than not, it is advisable to try you data manually than to use the > statpack. Looking forward to hearing from you if you agree with me that > using statpacks may give you strange results. The fact that so many > scholar spoke on the list about how to use the statpacks showed that it is > a burning question for linguists and other researchers who are not > specialists in mathematics. Yours sincerely Yuri Tambovtsev, > yutamb at mail.ru From haspelmath at eva.mpg.de Thu Sep 30 11:09:29 2010 From: haspelmath at eva.mpg.de (Martin Haspelmath) Date: Thu, 30 Sep 2010 13:09:29 +0200 Subject: A poet slips In-Reply-To: <1285836023.4ca44cf7aeb68@webmail.haifa.ac.il> Message-ID: Could it be that linguists, who tend to be more interested in structural forms, underestimate the social value of a particular form of a language? I'm a native speaker of a language (Common German of Germany) that didn't have native speakers 150 years ago. It had a lot of social prestige, so the southern dialects have been becoming more and more similar to it, and in northern Germany, Low German (of which my father was still a semi-speaker) has been abandoned in favour of the school language (which happened to be the sacred language of protestantism). Nowadays there are millions of native speakers oft this artificial language, whose front rounded ö and ü umlaut vowels were kept alive by the spelling (the vernaculars lost them centuries ago). From a linguist's point of view, this is a deplorable development, but the speakers sometimes view things differently. Martin john at research.haifa.ac.il wrote: > Learning the classical language is like pulling teeth for Arabs kids also. They > just can't publicly say it because that would make them bad Arabs. It's just > ridiculous. I have an even better plan for the Jewish kids--teach them to use > the written version of the spoken language which Arab kids are using for > Facebook. That's how kids make friends these days anyway. > Best wishes, > John > > > > > > Quoting Tom Givon : > > >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >> Classical (after 1949 things changed...). That book may still exist, >> you might track it down. It was easy, a cinch really. I concluded it was >> really just Hebrew with a few trivial transformations in Phonology & >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >> on a few sessions of a faculty study group at UO who were trying to >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >> it was like pullin' teeth. But Leonard Bloomfield said it already, and >> well, in 1939 (or was it 1943?). >> Cheers, TG >> >> ============== >> >> >> john at research.haifa.ac.il wrote: >> >>> Tom, >>> I've been trying for several years to get Israeli Arabs and Jews to >>> >> seriously >> >>> consider the possibility of educating Israeli Arabs in a written version of >>> their spoken language, as you suggest, at least through 3rd grade. I can >>> >> send >> >>> you some things I've written on this topic, with a lot of cross-linguistic >>> data. But thus far, it isn't working. As with most aspects of 'the >>> >> situation' >> >>> here, politically active Arabs think that the solution is to pursue the >>> >> same >> >>> self-destructive strategy which they've been following for almost a century >>> >> but >> >>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>> >> the >> >>> Arab media because they do not 'provide the linguistic richness of formal >>> Arabic') while Jews are basically content to let the Arabs stew in their >>> >> own >> >>> juices so that they can reap the benefits. There may, however, be hope for >>> >> the >> >>> future in the form of the radically increased usage of written forms of >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>> >> emails, >> >>> etc., by Arabic speakers below the age of 30, which will soon come to >>> >> seriously >> >>> threaten the status of classical Arabic in the same way that the invention >>> >> of >> >>> the printing press overturned the linguistic hierarchy in Western Europe by >>> overthrowing Latin. I'm working on this too. We'll see what happens. >>> >>> I realize that Masalha as well as the researchers he referred to don't >>> necessarily know what they're talking about regarding the brain. But the >>> general point is still potentially significant--that the connected script >>> >> which >> >>> Arabic uses as well as the multiple forms which many Arabic letters have >>> >> may be >> >>> a significant obstacle to literacy, however this may be related to >>> >> processes in >> >>> the brain. >>> Best wishes, >>> John >>> >>> >>> >>> Quoting Tom Givon : >>> >>> >>> >>>> A GREAT POET CAN STILL GET IT WRONG >>>> >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>> by stumbling a on his truly great quote: >>>> "All fixed identities are imposed from the outside. >>>> Whoever has a clear identity knows it can assume >>>> multiple forms". >>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>> clarity. Still, like the rest of us mortals, a great poet can on >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>> intended--surely got only one third of the story right. At first, he was >>>> led astray be the academic researchers he cited, who claimed that the >>>> lagging reading skills of Israeli-Arab students is correlated to >>>> lagging R-hemisphere activity, then explained this neurological lag >>>> by suggesting that the Arab script requires more contextual analysis. >>>> But it is the R-hemisphere of the human cortex that is more context >>>> oriented, less automated. If Arab-reading students required more >>>> contextual labor, it should have been registered as a higher >>>> R-hemisphere activity, not lower. >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>> among many others, have established beyond reasonable doubt that >>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>> module on the boundary of the occipital and temporal lobes (just past >>>> Brodman's Area 19), along the ventral visual object-recognition >>>> 'stream' that flows from the back to the front of the L-cortex. And >>>> the L-cortex is in general responsible for the more automated--less >>>> context-dependent--processing of language (as well as visual, motor and >>>> other skills). The visual word-recognition module is, in turn, >>>> recruited from the pre-existing visual object-recognition >>>> ventral-stream module. A considerable amount of life-time practice and >>>> repetition is required to affect this late-cultural adaptation. The >>>> human brain is not (yet) genetically configured at birth for >>>> visual-word recognition, only for visual-object recognition. A similar >>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>> in the L-pareita lobe, an area originally configured for analysis of >>>> object-location in space. >>>> Mr. Masalha then, on his own, points out to a more >>>> plausible right answer: Arab students, in Israel as well as all over the >>>> Arab world, are not taught literacy in their native language >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>> instrument harking back 1,400 years or more. That is, in a foreign >>>> language. The discrepancy would be just as great if Israeli kids were >>>> taught their Hebrew literacy first in the language of Genesis; or if >>>> French students were taught literacy first in the language of La >>>> Chanחon de Roland, Guilhome de Machaut, or Chrך tien de Troyes. Or >>>> English-speaking kids in the language of Beowolf. As far as my frail >>>> guessing powers go, remedying the situation would be much easier by >>>> combining two well-known verities of second language acquisition: (a) >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>> teach literacy first in the student's spoken native language; only then >>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>> in a book outlining a 'phonics-first ' literacy program for native >>>> English speakers. Rather than depriving Israeli-Arab students of >>>> literacy in their own--equally glorious--native language, just teach >>>> them smart. >>>> For his last culprit, the presumed--tho hardly >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>> recognized prejudices of the educated classes. While readily endorsing >>>> his aesthetic sentiments about modern media, I would still like to point >>>> out that the 'vulgar' genre is much closer to the Arab students' >>>> spoken native language, and if anything should facilitate the easier >>>> initial acquisition of native-language literacy. Respectuosamente, >>>> ma'-salaam, >>>> >>>> >>>> T. Givףn >>>> >>>> >>>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > > -- Martin Haspelmath (haspelmath at eva.mpg.de) Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 From cbutler at ntlworld.com Thu Sep 30 12:06:15 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Thu, 30 Sep 2010 13:06:15 +0100 Subject: Books on statistics for linguistics Message-ID: Shanley Allen has asked me to post to the list some references to general books on statistics for linguistics. Here is an annotated selection, which I hope will be useful. Other list members will no doubt be able to suggest other books which they have found helpful. Chris Butler ****************************** Butler, C. S. (1985) Statistics in Linguistics. Oxford: Blackwell. [Covers basic techniques only; explains the maths on the way; many examples. Last chapter hopelessly out of date 25 years on! Now out of print, but available via the internet at http://www.uwe.ac.uk/hlss/llas/statistics-in-linguistics/bkindex.shtml] Hatch, E. & A. Lazaraton (1991) The Research Manual: Design and Statistics for Applied Linguistics. Boston: Heinle & Heinle. [Fairly high level treatment, with many useful examples.] Johnson, K. (2008) Quantitative Methods in Linguistics. Oxford, Malden MA (USA) and Carlton, Victoria (Australia): Blackwell Publishing. [A fairly advanced treatment of a wide range of statistical techniques for the study of language, with statistical routines in the programming environment 'R'. Deals with the basics rather quickly, but also covers regression methods in detail. Oakes, M. P. (1998) Statistics for Corpus Linguistics. Edinburgh Textbooks in Linguistics. Edinburgh: Edinburgh University Press. [One chapter covers basic techniques rather briefly. Good discussion of multivariate methods. Also deals with other statistically-related procedures relevant to corpus linguistics, such as word tagging.] Rasinger, S. M. (2008) Quantitative Research in Linguistics: An Introduction. London: Continuum. [An introduction to quantitative data and its analysis, including basic statistical techniques.] Woods, A., P. Fletcher & A. Hughes (1986) Statistics in Language Studies. Cambridge: Cambridge University Press. [A more comprehensive treatment than Butler 1985, and at a considerably higher level, so correspondingly more difficult to read. Covers multivariate techniques, regression, etc in some detail. Very good examples.] From amnfn at well.com Thu Sep 30 12:42:11 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:42:11 -0700 Subject: A poet slips In-Reply-To: <4CA4282D.6080303@uoregon.edu> Message-ID: Tom, I feel the need to point out a few things: 1) There are no Arab students in the Israeli primary schools. There are those whose native language is Arabic, but they are not ethnically Arabs. Israeli citizens who do not identify as Israeli used to think of themselves as fellahin and now often identify as Palestinian -- precisely because they were not allowed to go to an integrated school where the only language spoken is Hebrew. They are ethnically Judeans who never went away. 2) The language of Genesis is the language that was revived by Zionists in the 19th century. I learned to read Hebrew in the language of Genesis. (And yes, I was born in Israel.) My father learned to read Hebrew in the language of Genesis. He was born in Poland. It was his first language, and the first language he was literate in. His parents were Zionists who learned to read Hebrew in the language of the OT. If it were not for people who learned to read Hebrew in the language of Genesis, there would be no native Hebrew speakers in the world today. Right cortex or left is a personal matter, depending on how you are wired. Left handers and right handers do it differently. But native speaker or oursider has everything to do with how you are treated and which school you go to. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, Tom Givon wrote: > > > A GREAT POET CAN STILL GET IT WRONG > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > Haaretz/English, International Herald Tribune 9-27-10) accidentally by > stumbling a on his truly great quote: > "All fixed identities are imposed from the outside. > Whoever has a clear identity knows it can assume multiple > forms". > In the context of Palestine/Israel, what a breath of fresh, rare clarity. > Still, like the rest of us mortals, a great poet can on occasion get it > wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got > only one third of the story right. At first, he was led astray be the > academic researchers he cited, who claimed that the lagging reading skills of > Israeli-Arab students is correlated to lagging R-hemisphere activity, > then explained this neurological lag by suggesting that the Arab script > requires more contextual analysis. But it is the R-hemisphere of the human > cortex that is more context oriented, less automated. If Arab-reading > students required more contextual labor, it should have been registered as a > higher R-hemisphere activity, not lower. > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among > many others, have established beyond reasonable doubt that written words > in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) > are decoded automatically in an L-hemisphere module on the boundary of the > occipital and temporal lobes (just past Brodman's Area 19), along the ventral > visual object-recognition 'stream' that flows from the back to the front of > the L-cortex. And the L-cortex is in general responsible for the more > automated--less context-dependent--processing of language (as well as > visual, motor and other skills). The visual word-recognition module is, in > turn, recruited from the pre-existing visual object-recognition > ventral-stream module. A considerable amount of life-time practice and > repetition is required to affect this late-cultural adaptation. The human > brain is not (yet) genetically configured at birth for visual-word > recognition, only for visual-object recognition. A similar cultural > adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see > recent article in The New Yorker by Oliver Sachs) > in the L-pareita lobe, an area originally configured for analysis of > object-location in space. > Mr. Masalha then, on his own, points out to a more plausible > right answer: Arab students, in Israel as well as all over the Arab world, > are not taught literacy in their native language (Falastini, Maghrebi, > Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 > years or more. That is, in a foreign language. The discrepancy would be just > as great if Israeli kids were taught their Hebrew literacy first in the > language of Genesis; or if French students were taught literacy first in > the language of La Chançon de Roland, Guilhome de Machaut, or Chrê tien de > Troyes. Or English-speaking kids in the language of Beowolf. As far as my > frail guessing powers go, remedying the situation would be much easier by > combining two well-known verities of second language acquisition: (a) Teach > them both early, together--'co-ordinated bilingualism'. And (b), teach > literacy first in the student's spoken native language; only then gradually > 'stretch' it to more literary genres. This method, bhy the way, was > suggested in the late 1930's by no other than L. Bloomfield, in a book > outlining a 'phonics-first ' literacy program for native English speakers. > Rather than depriving Israeli-Arab students of literacy in their own--equally > glorious--native language, just teach them smart. > For his last culprit, the presumed--tho hardly > unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized > prejudices of the educated classes. While readily endorsing his aesthetic > sentiments about modern media, I would still like to point out that the > 'vulgar' genre is much closer to the Arab students' spoken native language, > and if anything should facilitate the easier initial acquisition of > native-language literacy. Respectuosamente, ma'-salaam, > > T. > Givón > > From amnfn at well.com Thu Sep 30 12:51:34 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:51:34 -0700 Subject: A poet slips In-Reply-To: <1285836023.4ca44cf7aeb68@webmail.haifa.ac.il> Message-ID: John, These people you speak of are not Arabs. Some of them are Moslem and they read the Quran in the original. Some of them are not Moslem. All of them speak a local dialect of Arabic. Ask them sometimes if they think they are Arabs. Trying to turn every dialect into a separate language with a separate writing system is a way to try to disunite people. But a common language, however differently it is pronounced, unites disparate people. Australians and Cockneys and Indians and Americans speak sometimes mutually unintelligible versions of English. Using the same writing system and the same classic texts unites them. Instead of telling people they should magnify every difference, why not offer to share your language with them? Hebrew could be a uniting factor if spoken in all Israeli schools. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Learning the classical language is like pulling teeth for Arabs kids also. They > just can't publicly say it because that would make them bad Arabs. It's just > ridiculous. I have an even better plan for the Jewish kids--teach them to use > the written version of the spoken language which Arab kids are using for > Facebook. That's how kids make friends these days anyway. > Best wishes, > John > > > > > > Quoting Tom Givon : > >> >> >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >> Classical (after 1949 things changed...). That book may still exist, >> you might track it down. It was easy, a cinch really. I concluded it was >> really just Hebrew with a few trivial transformations in Phonology & >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >> on a few sessions of a faculty study group at UO who were trying to >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >> it was like pullin' teeth. But Leonard Bloomfield said it already, and >> well, in 1939 (or was it 1943?). >> Cheers, TG >> >> ============== >> >> >> john at research.haifa.ac.il wrote: >>> Tom, >>> I've been trying for several years to get Israeli Arabs and Jews to >> seriously >>> consider the possibility of educating Israeli Arabs in a written version of >>> their spoken language, as you suggest, at least through 3rd grade. I can >> send >>> you some things I've written on this topic, with a lot of cross-linguistic >>> data. But thus far, it isn't working. As with most aspects of 'the >> situation' >>> here, politically active Arabs think that the solution is to pursue the >> same >>> self-destructive strategy which they've been following for almost a century >> but >>> with even more vigor and steadfastness (e.g. Masalha points the blame at >> the >>> Arab media because they do not 'provide the linguistic richness of formal >>> Arabic') while Jews are basically content to let the Arabs stew in their >> own >>> juices so that they can reap the benefits. There may, however, be hope for >> the >>> future in the form of the radically increased usage of written forms of >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >> emails, >>> etc., by Arabic speakers below the age of 30, which will soon come to >> seriously >>> threaten the status of classical Arabic in the same way that the invention >> of >>> the printing press overturned the linguistic hierarchy in Western Europe by >>> overthrowing Latin. I'm working on this too. We'll see what happens. >>> >>> I realize that Masalha as well as the researchers he referred to don't >>> necessarily know what they're talking about regarding the brain. But the >>> general point is still potentially significant--that the connected script >> which >>> Arabic uses as well as the multiple forms which many Arabic letters have >> may be >>> a significant obstacle to literacy, however this may be related to >> processes in >>> the brain. >>> Best wishes, >>> John >>> >>> >>> >>> Quoting Tom Givon : >>> >>> >>>> A GREAT POET CAN STILL GET IT WRONG >>>> >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>> by stumbling a on his truly great quote: >>>> "All fixed identities are imposed from the outside. >>>> Whoever has a clear identity knows it can assume >>>> multiple forms". >>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>> clarity. Still, like the rest of us mortals, a great poet can on >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>> intended--surely got only one third of the story right. At first, he was >>>> led astray be the academic researchers he cited, who claimed that the >>>> lagging reading skills of Israeli-Arab students is correlated to >>>> lagging R-hemisphere activity, then explained this neurological lag >>>> by suggesting that the Arab script requires more contextual analysis. >>>> But it is the R-hemisphere of the human cortex that is more context >>>> oriented, less automated. If Arab-reading students required more >>>> contextual labor, it should have been registered as a higher >>>> R-hemisphere activity, not lower. >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>> among many others, have established beyond reasonable doubt that >>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>> module on the boundary of the occipital and temporal lobes (just past >>>> Brodman's Area 19), along the ventral visual object-recognition >>>> 'stream' that flows from the back to the front of the L-cortex. And >>>> the L-cortex is in general responsible for the more automated--less >>>> context-dependent--processing of language (as well as visual, motor and >>>> other skills). The visual word-recognition module is, in turn, >>>> recruited from the pre-existing visual object-recognition >>>> ventral-stream module. A considerable amount of life-time practice and >>>> repetition is required to affect this late-cultural adaptation. The >>>> human brain is not (yet) genetically configured at birth for >>>> visual-word recognition, only for visual-object recognition. A similar >>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>> in the L-pareita lobe, an area originally configured for analysis of >>>> object-location in space. >>>> Mr. Masalha then, on his own, points out to a more >>>> plausible right answer: Arab students, in Israel as well as all over the >>>> Arab world, are not taught literacy in their native language >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>> instrument harking back 1,400 years or more. That is, in a foreign >>>> language. The discrepancy would be just as great if Israeli kids were >>>> taught their Hebrew literacy first in the language of Genesis; or if >>>> French students were taught literacy first in the language of La >>>> Chanחon de Roland, Guilhome de Machaut, or Chrך tien de Troyes. Or >>>> English-speaking kids in the language of Beowolf. As far as my frail >>>> guessing powers go, remedying the situation would be much easier by >>>> combining two well-known verities of second language acquisition: (a) >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>> teach literacy first in the student's spoken native language; only then >>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>> in a book outlining a 'phonics-first ' literacy program for native >>>> English speakers. Rather than depriving Israeli-Arab students of >>>> literacy in their own--equally glorious--native language, just teach >>>> them smart. >>>> For his last culprit, the presumed--tho hardly >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>> recognized prejudices of the educated classes. While readily endorsing >>>> his aesthetic sentiments about modern media, I would still like to point >>>> out that the 'vulgar' genre is much closer to the Arab students' >>>> spoken native language, and if anything should facilitate the easier >>>> initial acquisition of native-language literacy. Respectuosamente, >>>> ma'-salaam, >>>> >>>> >>>> T. Givףn >>>> >>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 12:55:16 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:55:16 -0700 Subject: A poet slips In-Reply-To: <4CA46FE9.2000008@eva.mpg.de> Message-ID: Martin, Language is a tool. People want to communicate and fit in. Ultimately, that's more important than the ornamental differences between one tool and another. As functionalists, we should be cognizant of that. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, Martin Haspelmath wrote: > Could it be that linguists, who tend to be more interested in structural > forms, underestimate the social value of a particular form of a language? > > I'm a native speaker of a language (Common German of Germany) that didn't > have native speakers 150 years ago. It had a lot of social prestige, so the > southern dialects have been becoming more and more similar to it, and in > northern Germany, Low German (of which my father was still a semi-speaker) > has been abandoned in favour of the school language (which happened to be the > sacred language of protestantism). Nowadays there are millions of native > speakers oft this artificial language, whose front rounded ö and ü umlaut > vowels were kept alive by the spelling (the vernaculars lost them centuries > ago). > > From a linguist's point of view, this is a deplorable development, but the > speakers sometimes view things differently. > > Martin > > john at research.haifa.ac.il wrote: >> Learning the classical language is like pulling teeth for Arabs kids also. >> They >> just can't publicly say it because that would make them bad Arabs. It's >> just >> ridiculous. I have an even better plan for the Jewish kids--teach them to >> use >> the written version of the spoken language which Arab kids are using for >> Facebook. That's how kids make friends these days anyway. >> Best wishes, >> John >> >> >> >> >> >> Quoting Tom Givon : >> >> >>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>> Classical (after 1949 things changed...). That book may still exist, >>> you might track it down. It was easy, a cinch really. I concluded it was >>> really just Hebrew with a few trivial transformations in Phonology & >>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>> on a few sessions of a faculty study group at UO who were trying to >>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>> well, in 1939 (or was it 1943?). >>> Cheers, TG >>> >>> ============== >>> >>> >>> john at research.haifa.ac.il wrote: >>> >>>> Tom, >>>> I've been trying for several years to get Israeli Arabs and Jews to >>>> >>> seriously >>> >>>> consider the possibility of educating Israeli Arabs in a written version >>>> of >>>> their spoken language, as you suggest, at least through 3rd grade. I can >>>> >>> send >>> >>>> you some things I've written on this topic, with a lot of >>>> cross-linguistic >>>> data. But thus far, it isn't working. As with most aspects of 'the >>>> >>> situation' >>> >>>> here, politically active Arabs think that the solution is to pursue the >>>> >>> same >>> >>>> self-destructive strategy which they've been following for almost a >>>> century >>>> >>> but >>> >>>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>>> >>> the >>> >>>> Arab media because they do not 'provide the linguistic richness of formal >>>> Arabic') while Jews are basically content to let the Arabs stew in their >>>> >>> own >>> >>>> juices so that they can reap the benefits. There may, however, be hope >>>> for >>>> >>> the >>> >>>> future in the form of the radically increased usage of written forms of >>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>> >>> emails, >>> >>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>> >>> seriously >>> >>>> threaten the status of classical Arabic in the same way that the >>>> invention >>>> >>> of >>> >>>> the printing press overturned the linguistic hierarchy in Western Europe >>>> by >>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>> >>>> I realize that Masalha as well as the researchers he referred to don't >>>> necessarily know what they're talking about regarding the brain. But the >>>> general point is still potentially significant--that the connected script >>>> >>> which >>> >>>> Arabic uses as well as the multiple forms which many Arabic letters have >>>> >>> may be >>> >>>> a significant obstacle to literacy, however this may be related to >>>> >>> processes in >>> >>>> the brain. >>>> Best wishes, >>>> John >>>> >>>> >>>> >>>> Quoting Tom Givon : >>>> >>>> >>>> >>>>> A GREAT POET CAN STILL GET IT >>>>> WRONG >>>>> >>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>> by stumbling a on his truly great quote: >>>>> "All fixed identities are imposed from the outside. >>>>> Whoever has a clear identity knows it can assume >>>>> multiple forms". >>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>> intended--surely got only one third of the story right. At first, he was >>>>> led astray be the academic researchers he cited, who claimed that the >>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>> by suggesting that the Arab script requires more contextual analysis. >>>>> But it is the R-hemisphere of the human cortex that is more context >>>>> oriented, less automated. If Arab-reading students required more >>>>> contextual labor, it should have been registered as a higher >>>>> R-hemisphere activity, not lower. >>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>> among many others, have established beyond reasonable doubt that >>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>> module on the boundary of the occipital and temporal lobes (just past >>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>> the L-cortex is in general responsible for the more automated--less >>>>> context-dependent--processing of language (as well as visual, motor and >>>>> other skills). The visual word-recognition module is, in turn, >>>>> recruited from the pre-existing visual object-recognition >>>>> ventral-stream module. A considerable amount of life-time practice and >>>>> repetition is required to affect this late-cultural adaptation. The >>>>> human brain is not (yet) genetically configured at birth for >>>>> visual-word recognition, only for visual-object recognition. A similar >>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>> object-location in space. >>>>> Mr. Masalha then, on his own, points out to a more >>>>> plausible right answer: Arab students, in Israel as well as all over the >>>>> Arab world, are not taught literacy in their native language >>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>> language. The discrepancy would be just as great if Israeli kids were >>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>> French students were taught literacy first in the language of La >>>>> Chanחon de Roland, Guilhome de Machaut, or Chrך tien de Troyes. Or >>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>> guessing powers go, remedying the situation would be much easier by >>>>> combining two well-known verities of second language acquisition: (a) >>>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>>> teach literacy first in the student's spoken native language; only then >>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>> literacy in their own--equally glorious--native language, just teach >>>>> them smart. >>>>> For his last culprit, the presumed--tho hardly >>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>> recognized prejudices of the educated classes. While readily endorsing >>>>> his aesthetic sentiments about modern media, I would still like to point >>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>> spoken native language, and if anything should facilitate the easier >>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>> ma'-salaam, >>>>> >>>>> >>>>> T. Givףn >>>>> >>>>> >>>>> >>>> >>>> >>>> ------------------------------------------------------------------------ >>>> This message was sent using IMP, the Webmail Program of Haifa University >>>> >>>> >>> >> >> >> >> >> ------------------------------------------------------------------------ >> This message was sent using IMP, the Webmail Program of Haifa University >> >> >> > > > -- > Martin Haspelmath (haspelmath at eva.mpg.de) > Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 > D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 > > > > > > From mariel at post.tau.ac.il Thu Sep 30 14:36:24 2010 From: mariel at post.tau.ac.il (Mira Ariel) Date: Thu, 30 Sep 2010 06:36:24 -0800 Subject: Learning to read Hebrew Message-ID: Hi, I'm surprised to hear that Aya Katz learnt to read Hebrew using Biblical texts. Growing up in Israel even before Aya, we leant to read and write from Modern Hebrew texts, albeit, somewhat formal, written Hebrew. We never started to read the Bible till we knew how to read (2nd grade), and at the beginning, the Biblical text was a simplified one ("Bible for children"). So, no doubt one CAN learn Hebrew from Biblical texts, but we don't impose it on all kids in Israel. Best, Mira _____________ Mira Ariel http://www.tau.ac.il/~mariel From amnfn at well.com Thu Sep 30 14:27:30 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 07:27:30 -0700 Subject: Learning to read Hebrew In-Reply-To: <001001cb60ac$dd8ae5a0$98a0b0e0$@tau.ac.il> Message-ID: Mira, My story is in fact a little unusual. While I grew up in Israel, I went to first and second grade in the United States. I had picked up a little reading when I was five, but as you can imagine, I was rusty by the time we returned to Israel and I was expected to go into third grade. My grandfather, who was then the Rector of the University of Tel-Aviv, brought me a Bible and made me start reading Genesis out loud to him. He made me do this until he was satisfied that my reading was good enough for third grade. In third grade, we read the old testament in its unexpurgated and uncut form. I believe we started on Judges, because I had missed out on some of the earlier books. I would like to emphasize that none of my classmates needed a translation from Biblical to "modern" Hebrew. I know that in academic circles much is made of the difference, but believe me Biblical texts are intelligible to modern day readers. The difference in more like that between Shakespeare and today's English. It is not like the difference between Chaucer and modern English. There's a lot of politics that goes on in academic circles to make mountains out of molehills where minute differences between different versions of the same language are concerned. However, if one insists on modern day school children unable to understand the old language, not making them read it in school is the best first step. We know what we use. --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, Mira Ariel wrote: > Hi, > > > > I'm surprised to hear that Aya Katz learnt to read Hebrew using Biblical > texts. Growing up in Israel even before Aya, we leant to read and write from > Modern Hebrew texts, albeit, somewhat formal, written Hebrew. We never > started to read the Bible till we knew how to read (2nd grade), and at the > beginning, the Biblical text was a simplified one ("Bible for children"). > So, no doubt one CAN learn Hebrew from Biblical texts, but we don't impose > it on all kids in Israel. > > > > Best, > > > > Mira > > > > _____________ > > Mira Ariel > > http://www.tau.ac.il/~mariel > > > > From kuzar at research.haifa.ac.il Thu Sep 30 14:25:21 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Thu, 30 Sep 2010 16:25:21 +0200 Subject: A poet slips In-Reply-To: Message-ID: I see that all (ex)Israelis on this list are throwing in their 2Cs. Here are mine. One of the slippery things about identity is that it is dynamic, multiple, and defies one-word definitions. Israeli Arabs/Palestinians are a sector of Israel's citizenry. I even hesitate to say that they are a community, but even if they are, they are as diversified among themselves, as any other citizenry nowadays. Just because they happen to be Arabs (people of the Orient) doesn't mean you can easily draw generalizing cultural conclusions any more than you could do this about SAE nations. I am saying all this, because I hear an Orientalistic tone in both John's and Aya's postings, who just simply know and are so quick to define Arab/Palestinian/Arabic speaking citizens of Israel as either this or that. Some Israeli Arabs put more emphasis on their general ethnic common origin, others more on their local Palestinian side of their identity, others yet are first and foremost Muslims, or Christians, some slip in and out of this or that identity every five years, and some couldn't care less and would in fact become Hebrew speaking Israeli's, if the Israeli system had been more forthcoming. Masalha himself is from Druze origin, a minority that has been persecuted by Muslim majority for centuries. This resulted in massive support of Zionism by the Druze and they go to the army and fight the Arabs without hesitation. Except that this is also a massive generalization. Israel's Jewish ethnocentric laws and practices don't even fully embrace the Druze population with great love, a fact which brought some of them to identify with and to view themselves as Palestinian and to struggle for the common Palestinian cause. Add to this the fact that any two communities living together produce hybridity, and you will find all sorts of creative (and wonderful) hybrid people who combine Jewish and Arab identities in their personality. Masalha himself is from Druze origin, anti Zionist, but he is also a poet, writing in both Hebrew and Arabic, and also a translator both ways. This is only a shallow introduction to the complexities of identity for those of you who are not so well informed in this discussion. The question of language is a very complex one, precisely because the educational system has to devise a model that would be equally beneficial for a widely diversified community. Assuming that there is an Arab speaking population in Israel (which we don't have to define culturally for the sake of this discussion) and assuming that many of them support the idea that Arabic should be maintained and should be the first schooling language for these people, we may indeed observe that Arab children are faced with two languages, their native "dialect" and the language of broad cultural exchange in the Arab world today. If one day the native tongues are going to become official languages of Egypt, Syria, Jordan, Morocco, Yemen, etc. I am sure that the Palestinians will follow suit, and the Israeli Arabs will be a part of this process. It is not reasonable to expect Israeli Arabs to be the pioneers here. Now, Modern Standard Arabic is not Classical (Quran) Arabic, it is a modern language that has been revived parallel to the revival of Hebrew, except that the revival of Hebrew included the spoken language and the revival of Arabic was only in the written language. This is however a full-fledged communication system, in which anything can be written, from engineering to modern philosophy. Another complicating factor is the fact that MSA and local dialects are not a dichotomy but rather a scale. The is the more formal form, which is a simplified version of Classical Arabic, but then there is also a simplified version of this language that is used without the case endings, and sometimes in the SVO order of the spoken language, and so on. While it is true that the Internet and Facebook have brought about new ways of communicating in some koine forms of the spoken dialects, it is not true that the whole population shares this mode of communication (only the more privileged ones) and you certainly cannot write books on philosophy and engineering in these language forms. So let's be a little more careful about giving sweeping simple answers to the complex situation of learner of Arabic as a first language. In fact, the Arabs are not the first ones to have to face such a complex linguistic situation. I would suggest that this issue be approached with more cultural sensitivity and attempts to to operate within the given complex situation. Believe me, with a good schooling system that is well funded and has a good educational leadership (which is not the case in Israel at large, and is much worse in the Arab sector), the left and right hemispheres of the brain will adjust just fine, and will learn what they need to do to deal with all the complexities, despite the "difficult" cursive alphabet, despite the right-to-left writing direction, and despite all other "objective" difficulties. Ron Kuzar ---------------- On Thu, 30 Sep 2010 05:42:11 -0700 (PDT) "A. Katz" wrote: > Tom, > > I feel the need to point out a few things: > > 1) There are no Arab students in the Israeli primary schools. There are > those whose native language is Arabic, but they are not ethnically Arabs. > Israeli citizens who do not identify as Israeli used to think of themselves as > fellahin and now often identify as Palestinian -- precisely because they > were not allowed to go to an integrated school where the only language > spoken is Hebrew. They are ethnically Judeans who never went away. > > 2) The language of Genesis is the language that was revived by Zionists in > the 19th century. I learned to read Hebrew in the language of Genesis. > (And yes, I was born in Israel.) My father learned to read Hebrew in the > language of Genesis. He was born in Poland. It was his first language, > and the first language he was literate in. His parents were Zionists > who learned to read Hebrew in the language of the OT. If it were not for > people who learned to read Hebrew in the language of Genesis, there would be no > native Hebrew speakers in the world today. > > Right cortex or left is a personal matter, depending on how you are wired. > Left handers and right handers do it differently. But native speaker or > oursider has everything to do with how you are treated and which school > you go to. > > > --Aya > > http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > On Thu, 30 Sep 2010, Tom Givon wrote: > > > > > > > A GREAT POET CAN STILL GET IT WRONG > > > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > > Haaretz/English, International Herald Tribune 9-27-10) accidentally by > > stumbling a on his truly great quote: > > "All fixed identities are imposed from the outside. > > Whoever has a clear identity knows it can assume multiple > > forms". > > In the context of Palestine/Israel, what a breath of fresh, rare clarity. > > Still, like the rest of us mortals, a great poet can on occasion get it > > wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got > > only one third of the story right. At first, he was led astray be the > > academic researchers he cited, who claimed that the lagging reading skills of > > Israeli-Arab students is correlated to lagging R-hemisphere activity, > > then explained this neurological lag by suggesting that the Arab script > > requires more contextual analysis. But it is the R-hemisphere of the human > > cortex that is more context oriented, less automated. If Arab-reading > > students required more contextual labor, it should have been registered as a > > higher R-hemisphere activity, not lower. > > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among > > many others, have established beyond reasonable doubt that written words > > in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) > > are decoded automatically in an L-hemisphere module on the boundary of the > > occipital and temporal lobes (just past Brodman's Area 19), along the ventral > > visual object-recognition 'stream' that flows from the back to the front of > > the L-cortex. And the L-cortex is in general responsible for the more > > automated--less context-dependent--processing of language (as well as > > visual, motor and other skills). The visual word-recognition module is, in > > turn, recruited from the pre-existing visual object-recognition > > ventral-stream module. A considerable amount of life-time practice and > > repetition is required to affect this late-cultural adaptation. The human > > brain is not (yet) genetically configured at birth for visual-word > > recognition, only for visual-object recognition. A similar cultural > > adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see > > recent article in The New Yorker by Oliver Sachs) > > in the L-pareita lobe, an area originally configured for analysis of > > object-location in space. > > Mr. Masalha then, on his own, points out to a more plausible > > right answer: Arab students, in Israel as well as all over the Arab world, > > are not taught literacy in their native language (Falastini, Maghrebi, > > Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 > > years or more. That is, in a foreign language. The discrepancy would be just > > as great if Israeli kids were taught their Hebrew literacy first in the > > language of Genesis; or if French students were taught literacy first in > > the language of La Chançon de Roland, Guilhome de Machaut, or Chrê tien de > > Troyes. Or English-speaking kids in the language of Beowolf. As far as my > > frail guessing powers go, remedying the situation would be much easier by > > combining two well-known verities of second language acquisition: (a) Teach > > them both early, together--'co-ordinated bilingualism'. And (b), teach > > literacy first in the student's spoken native language; only then gradually > > 'stretch' it to more literary genres. This method, bhy the way, was > > suggested in the late 1930's by no other than L. Bloomfield, in a book > > outlining a 'phonics-first ' literacy program for native English speakers. > > Rather than depriving Israeli-Arab students of literacy in their own--equally > > glorious--native language, just teach them smart. > > For his last culprit, the presumed--tho hardly > > unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized > > prejudices of the educated classes. While readily endorsing his aesthetic > > sentiments about modern media, I would still like to point out that the > > 'vulgar' genre is much closer to the Arab students' spoken native language, > > and if anything should facilitate the easier initial acquisition of > > native-language literacy. Respectuosamente, ma'-salaam, > > > > T. > > Givón > > > > =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From john at research.haifa.ac.il Thu Sep 30 14:43:30 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 16:43:30 +0200 Subject: A poet slips In-Reply-To: Message-ID: Aya, I ask them constantly if they think they are Arabs (60% of my students are Arabic speakers and this is a central topic of many classes that I teach) and almost without exception they do. The only exceptions are some Druze (mostly males) and most Maronites. I am aware that 100 years ago Arabic speakers living in this area did not consider themselves to be Arabs (except for the Bedouins), but the situation has completely changed. It is totally unrealistic to put Israeli Arabs as a group into Hebrew-speaking schools. No one wants it, not the Jews and not the Arabs. There are a tiny number of Arabs who for one reason or another send their children to Hebrew-speaking schools (for example the author Sayed Kashua) but this is insignificant. I don't know what the situation was when you were a child, and I don't know about the situation among religious Jews, but secular Jewish children do not learn to read by reading Genesis today. They learn to read with secular texts in 1st grade and go on to the Bible in 2nd grade, but far from learning to read by reading the Bible, teachers have to explain what's written in the Bible to the students. NO ONE can understand it without an explanation from the teacher or their parents. Ron--I agree with everything you say except that I believe that the radical difference between spoken and written Arabic definitely is a serious problem for literacy. I would not be so quick to dismiss the effect of the Arabic writing system--we had a conference on this topic in Haifa in May and the general consensus of people who sounded like they knew what they were talking about was that it is a significant problem-- but they hadn't done cross-linguistic research and this isn't something I know enough about to have strong opinions one way or the other. Also, the usage of written colloquial Arabic is basically universal among all Israeli Arabs under the age of 25 who have a cellular phone. This is definitely not just a privileged section of the population, it's most people of the relevant age group. John Quoting "A. Katz" : > John, > > These people you speak of are not Arabs. Some of them are Moslem and they > read the Quran in the original. Some of them are not Moslem. All of them > speak a local dialect of Arabic. Ask them sometimes if they think they are > Arabs. > > Trying to turn every dialect into a separate language with a separate > writing system is a way to try to disunite people. But a common language, > however differently it is pronounced, unites disparate people. Australians > and Cockneys and Indians and Americans speak sometimes mutually > unintelligible versions of English. Using the same writing system and > the same classic texts unites them. > > Instead of telling people they should magnify every difference, why not > offer to share your language with them? Hebrew could be a uniting factor > if spoken in all Israeli schools. > > --Aya > > http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Learning the classical language is like pulling teeth for Arabs kids also. > They > > just can't publicly say it because that would make them bad Arabs. It's > just > > ridiculous. I have an even better plan for the Jewish kids--teach them to > use > > the written version of the spoken language which Arab kids are using for > > Facebook. That's how kids make friends these days anyway. > > Best wishes, > > John > > > > > > > > > > > > Quoting Tom Givon : > > > >> > >> > >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >> Classical (after 1949 things changed...). That book may still exist, > >> you might track it down. It was easy, a cinch really. I concluded it was > >> really just Hebrew with a few trivial transformations in Phonology & > >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > >> on a few sessions of a faculty study group at UO who were trying to > >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > >> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >> well, in 1939 (or was it 1943?). > >> Cheers, TG > >> > >> ============== > >> > >> > >> john at research.haifa.ac.il wrote: > >>> Tom, > >>> I've been trying for several years to get Israeli Arabs and Jews to > >> seriously > >>> consider the possibility of educating Israeli Arabs in a written version > of > >>> their spoken language, as you suggest, at least through 3rd grade. I can > >> send > >>> you some things I've written on this topic, with a lot of > cross-linguistic > >>> data. But thus far, it isn't working. As with most aspects of 'the > >> situation' > >>> here, politically active Arabs think that the solution is to pursue the > >> same > >>> self-destructive strategy which they've been following for almost a > century > >> but > >>> with even more vigor and steadfastness (e.g. Masalha points the blame at > >> the > >>> Arab media because they do not 'provide the linguistic richness of formal > >>> Arabic') while Jews are basically content to let the Arabs stew in their > >> own > >>> juices so that they can reap the benefits. There may, however, be hope > for > >> the > >>> future in the form of the radically increased usage of written forms of > >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, > >> emails, > >>> etc., by Arabic speakers below the age of 30, which will soon come to > >> seriously > >>> threaten the status of classical Arabic in the same way that the > invention > >> of > >>> the printing press overturned the linguistic hierarchy in Western Europe > by > >>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>> > >>> I realize that Masalha as well as the researchers he referred to don't > >>> necessarily know what they're talking about regarding the brain. But the > >>> general point is still potentially significant--that the connected script > >> which > >>> Arabic uses as well as the multiple forms which many Arabic letters have > >> may be > >>> a significant obstacle to literacy, however this may be related to > >> processes in > >>> the brain. > >>> Best wishes, > >>> John > >>> > >>> > >>> > >>> Quoting Tom Givon : > >>> > >>> > >>>> A GREAT POET CAN STILL GET IT > WRONG > >>>> > >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >>>> by stumbling a on his truly great quote: > >>>> "All fixed identities are imposed from the outside. > >>>> Whoever has a clear identity knows it can assume > >>>> multiple forms". > >>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>> intended--surely got only one third of the story right. At first, he was > >>>> led astray be the academic researchers he cited, who claimed that the > >>>> lagging reading skills of Israeli-Arab students is correlated to > >>>> lagging R-hemisphere activity, then explained this neurological lag > >>>> by suggesting that the Arab script requires more contextual analysis. > >>>> But it is the R-hemisphere of the human cortex that is more context > >>>> oriented, less automated. If Arab-reading students required more > >>>> contextual labor, it should have been registered as a higher > >>>> R-hemisphere activity, not lower. > >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>> among many others, have established beyond reasonable doubt that > >>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>> module on the boundary of the occipital and temporal lobes (just past > >>>> Brodman's Area 19), along the ventral visual object-recognition > >>>> 'stream' that flows from the back to the front of the L-cortex. And > >>>> the L-cortex is in general responsible for the more automated--less > >>>> context-dependent--processing of language (as well as visual, motor and > >>>> other skills). The visual word-recognition module is, in turn, > >>>> recruited from the pre-existing visual object-recognition > >>>> ventral-stream module. A considerable amount of life-time practice and > >>>> repetition is required to affect this late-cultural adaptation. The > >>>> human brain is not (yet) genetically configured at birth for > >>>> visual-word recognition, only for visual-object recognition. A similar > >>>> cultural adaptation, this one for math, has been shown for (Dahaene & > >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>> in the L-pareita lobe, an area originally configured for analysis of > >>>> object-location in space. > >>>> Mr. Masalha then, on his own, points out to a more > >>>> plausible right answer: Arab students, in Israel as well as all over the > >>>> Arab world, are not taught literacy in their native language > >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>> language. The discrepancy would be just as great if Israeli kids were > >>>> taught their Hebrew literacy first in the language of Genesis; or if > >>>> French students were taught literacy first in the language of La > >>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or > >>>> English-speaking kids in the language of Beowolf. As far as my frail > >>>> guessing powers go, remedying the situation would be much easier by > >>>> combining two well-known verities of second language acquisition: (a) > >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), > >>>> teach literacy first in the student's spoken native language; only then > >>>> gradually 'stretch' it to more literary genres. This method, bhy the > >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, > >>>> in a book outlining a 'phonics-first ' literacy program for native > >>>> English speakers. Rather than depriving Israeli-Arab students of > >>>> literacy in their own--equally glorious--native language, just teach > >>>> them smart. > >>>> For his last culprit, the presumed--tho hardly > >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>> recognized prejudices of the educated classes. While readily endorsing > >>>> his aesthetic sentiments about modern media, I would still like to point > >>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>> spoken native language, and if anything should facilitate the easier > >>>> initial acquisition of native-language literacy. Respectuosamente, > >>>> ma'-salaam, > >>>> > >>>> > >>>> T. Givón > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >> > >> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From amnfn at well.com Thu Sep 30 14:51:29 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 07:51:29 -0700 Subject: A poet slips In-Reply-To: <20100930162521.B22D.BA0BAB47@research.haifa.ac.il> Message-ID: Ron, Most of what you said made sense to me. I am not familiar with the revival of Arabic and would be happy if you posted a link for all us Funknetters on this topic. It makes sense to see different versions of the langauge as gradient. One thing you said, though, struck me as weird. "Despite the right-to-left direction...." You're not seriously suggesting that RTL or LTR makes a difference in ease of reading, are you? --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz From grvsmth at panix.com Thu Sep 30 15:01:07 2010 From: grvsmth at panix.com (Angus B. Grieve-Smith) Date: Thu, 30 Sep 2010 11:01:07 -0400 Subject: A poet slips In-Reply-To: <1285857810.4ca4a212286fb@webmail.haifa.ac.il> Message-ID: On Thu, September 30, 2010 10:43 am, john at research.haifa.ac.il wrote: > Also, the usage of written colloquial > Arabic is basically universal among all Israeli Arabs under the age of 25 > who have a cellular phone. > This is definitely not just a privileged section of the population, it's > most people of the relevant age group. That's very interesting. What alphabet do they use? -- -Angus B. Grieve-Smith grvsmth at panix.com From john at research.haifa.ac.il Thu Sep 30 15:06:52 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 17:06:52 +0200 Subject: A poet slips In-Reply-To: Message-ID: For learning to read for the first time, obviously it doesn't make a difference. But once you've learned to read in one direction, it's tough to start reading in the opposite direction, particularly if you've started late in life. I know from personal experience. John Quoting "A. Katz" : > Ron, > > Most of what you said made sense to me. I am not familiar with the revival > of Arabic and would be happy if you posted a link for all us Funknetters > on this topic. It makes sense to see different versions of the langauge as > gradient. > > One thing you said, though, struck me as weird. "Despite the right-to-left > direction...." You're not seriously suggesting that RTL or LTR makes a > difference in ease of reading, are you? > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From kuzar at research.haifa.ac.il Thu Sep 30 15:14:04 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Thu, 30 Sep 2010 17:14:04 +0200 Subject: A poet slips In-Reply-To: Message-ID: The most informative source, albeit a bit old, is: Blau, Joshua. 1981. The renaissance of modern Hebrew and modern standard Arabic : parallels and differences in the revival of two semitic languages http://books.google.com/books?id=EwbvrNRcaNIC&printsec=frontcover&dq=blau+revival+arabic&source=bl&ots=P_1nIzaFDb&sig=zCfhko8v9UY8rbZC3-OKp_P3kq4&hl=en&ei=pqikTM3NOYiPswbJoYWiCA&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBIQ6AEwAA#v=onepage&q&f=false Directionality has been suggested in scholarly circles as a problem. I am not an expert on this issue, and I do NOT believe it could be true. Ron On Thu, 30 Sep 2010 07:51:29 -0700 (PDT) "A. Katz" wrote: > Ron, > > Most of what you said made sense to me. I am not familiar with the revival > of Arabic and would be happy if you posted a link for all us Funknetters > on this topic. It makes sense to see different versions of the langauge as > gradient. > > One thing you said, though, struck me as weird. "Despite the right-to-left > direction...." You're not seriously suggesting that RTL or LTR makes a > difference in ease of reading, are you? > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From amnfn at well.com Thu Sep 30 15:18:51 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:18:51 -0700 Subject: A poet slips In-Reply-To: <1285857810.4ca4a212286fb@webmail.haifa.ac.il> Message-ID: John, Well, at least I'm glad you realize that one hundred years ago, and possibly even more recently, they did not think of themselves as Arabs. I still wonder whether among themselves they say "Arabs". Isn't there another word? If any of them tried to emigrate to Saudi Arabia, I doubt very much they'd be called "Arabs." For that matter, what do Jordanians call non-Jewish Israelis? The biggest obstacle to allowing for integration in the schools is not the Palestinian population -- it's those among the Israelis who adopt a "Jews first" attitude. --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Aya, > I ask them constantly if they think they are Arabs (60% of my students > are Arabic speakers and this is a central topic of many classes that I teach) > and almost without exception they do. The only exceptions are some Druze > (mostly males) and most Maronites. I am aware that 100 years ago Arabic > speakers living in this area did not consider themselves to be Arabs (except > for the Bedouins), but the situation has completely changed. > > It is totally unrealistic to put Israeli Arabs as a group into Hebrew-speaking > schools. No one wants it, not the Jews and not the Arabs. There are a tiny > number of Arabs who for one reason or another send their children to > Hebrew-speaking schools (for example the author Sayed Kashua) but this > is insignificant. > > I don't know what the situation was when you were a child, and I don't know > about the situation among religious Jews, but secular Jewish children do not > learn to read by reading Genesis today. They learn to read with secular texts > in 1st grade and go on to the Bible in 2nd grade, but far from learning to > read by reading the Bible, teachers have to explain what's written in the > Bible to the students. NO ONE can understand it without an explanation from the > teacher or their parents. > > Ron--I agree with everything you say except that I believe that the radical > difference between spoken and written Arabic definitely is a serious problem > for literacy. I would not be so quick to dismiss the effect of the Arabic > writing system--we had a conference on this topic in Haifa in May and the > general consensus of people who sounded like they knew what they were talking > about was that it is a significant problem-- but they hadn't done > cross-linguistic research and this isn't something I know enough about to have > strong opinions one way or the other. Also, the usage of written colloquial > Arabic is basically universal among all Israeli Arabs under the age of 25 who > have a cellular phone. > This is definitely not just a privileged section of the population, it's most > people of the relevant age group. > > John > > > > > > Quoting "A. Katz" : > >> John, >> >> These people you speak of are not Arabs. Some of them are Moslem and they >> read the Quran in the original. Some of them are not Moslem. All of them >> speak a local dialect of Arabic. Ask them sometimes if they think they are >> Arabs. >> >> Trying to turn every dialect into a separate language with a separate >> writing system is a way to try to disunite people. But a common language, >> however differently it is pronounced, unites disparate people. Australians >> and Cockneys and Indians and Americans speak sometimes mutually >> unintelligible versions of English. Using the same writing system and >> the same classic texts unites them. >> >> Instead of telling people they should magnify every difference, why not >> offer to share your language with them? Hebrew could be a uniting factor >> if spoken in all Israeli schools. >> >> --Aya >> >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Learning the classical language is like pulling teeth for Arabs kids also. >> They >>> just can't publicly say it because that would make them bad Arabs. It's >> just >>> ridiculous. I have an even better plan for the Jewish kids--teach them to >> use >>> the written version of the spoken language which Arab kids are using for >>> Facebook. That's how kids make friends these days anyway. >>> Best wishes, >>> John >>> >>> >>> >>> >>> >>> Quoting Tom Givon : >>> >>>> >>>> >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>> Classical (after 1949 things changed...). That book may still exist, >>>> you might track it down. It was easy, a cinch really. I concluded it was >>>> really just Hebrew with a few trivial transformations in Phonology & >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>>> on a few sessions of a faculty study group at UO who were trying to >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>> well, in 1939 (or was it 1943?). >>>> Cheers, TG >>>> >>>> ============== >>>> >>>> >>>> john at research.haifa.ac.il wrote: >>>>> Tom, >>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>> seriously >>>>> consider the possibility of educating Israeli Arabs in a written version >> of >>>>> their spoken language, as you suggest, at least through 3rd grade. I can >>>> send >>>>> you some things I've written on this topic, with a lot of >> cross-linguistic >>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>> situation' >>>>> here, politically active Arabs think that the solution is to pursue the >>>> same >>>>> self-destructive strategy which they've been following for almost a >> century >>>> but >>>>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>>> the >>>>> Arab media because they do not 'provide the linguistic richness of formal >>>>> Arabic') while Jews are basically content to let the Arabs stew in their >>>> own >>>>> juices so that they can reap the benefits. There may, however, be hope >> for >>>> the >>>>> future in the form of the radically increased usage of written forms of >>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>> emails, >>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>> seriously >>>>> threaten the status of classical Arabic in the same way that the >> invention >>>> of >>>>> the printing press overturned the linguistic hierarchy in Western Europe >> by >>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>> >>>>> I realize that Masalha as well as the researchers he referred to don't >>>>> necessarily know what they're talking about regarding the brain. But the >>>>> general point is still potentially significant--that the connected script >>>> which >>>>> Arabic uses as well as the multiple forms which many Arabic letters have >>>> may be >>>>> a significant obstacle to literacy, however this may be related to >>>> processes in >>>>> the brain. >>>>> Best wishes, >>>>> John >>>>> >>>>> >>>>> >>>>> Quoting Tom Givon : >>>>> >>>>> >>>>>> A GREAT POET CAN STILL GET IT >> WRONG >>>>>> >>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>>> by stumbling a on his truly great quote: >>>>>> "All fixed identities are imposed from the outside. >>>>>> Whoever has a clear identity knows it can assume >>>>>> multiple forms". >>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>> intended--surely got only one third of the story right. At first, he was >>>>>> led astray be the academic researchers he cited, who claimed that the >>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>>> by suggesting that the Arab script requires more contextual analysis. >>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>> oriented, less automated. If Arab-reading students required more >>>>>> contextual labor, it should have been registered as a higher >>>>>> R-hemisphere activity, not lower. >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>> among many others, have established beyond reasonable doubt that >>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>> module on the boundary of the occipital and temporal lobes (just past >>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>>> the L-cortex is in general responsible for the more automated--less >>>>>> context-dependent--processing of language (as well as visual, motor and >>>>>> other skills). The visual word-recognition module is, in turn, >>>>>> recruited from the pre-existing visual object-recognition >>>>>> ventral-stream module. A considerable amount of life-time practice and >>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>> human brain is not (yet) genetically configured at birth for >>>>>> visual-word recognition, only for visual-object recognition. A similar >>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>> object-location in space. >>>>>> Mr. Masalha then, on his own, points out to a more >>>>>> plausible right answer: Arab students, in Israel as well as all over the >>>>>> Arab world, are not taught literacy in their native language >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>> language. The discrepancy would be just as great if Israeli kids were >>>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>>> French students were taught literacy first in the language of La >>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or >>>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>>> guessing powers go, remedying the situation would be much easier by >>>>>> combining two well-known verities of second language acquisition: (a) >>>>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>>>> teach literacy first in the student's spoken native language; only then >>>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>> literacy in their own--equally glorious--native language, just teach >>>>>> them smart. >>>>>> For his last culprit, the presumed--tho hardly >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>> recognized prejudices of the educated classes. While readily endorsing >>>>>> his aesthetic sentiments about modern media, I would still like to point >>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>> spoken native language, and if anything should facilitate the easier >>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>> ma'-salaam, >>>>>> >>>>>> >>>>>> T. Givón >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>> >>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 15:23:09 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:23:09 -0700 Subject: A poet slips In-Reply-To: <1285859212.4ca4a78cc9912@webmail.haifa.ac.il> Message-ID: Well, everything is hard if you start too late. But if you start almost simultaneously in both directions in different languages, then, in my experience, it's not a problem. --Aya On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > For learning to read for the first time, obviously it doesn't make a difference. > But once you've learned to read in one direction, it's tough to start reading > in the opposite direction, particularly if you've started late in life. I > know from personal experience. > John > > > > > > Quoting "A. Katz" : > >> Ron, >> >> Most of what you said made sense to me. I am not familiar with the revival >> of Arabic and would be happy if you posted a link for all us Funknetters >> on this topic. It makes sense to see different versions of the langauge as >> gradient. >> >> One thing you said, though, struck me as weird. "Despite the right-to-left >> direction...." You're not seriously suggesting that RTL or LTR makes a >> difference in ease of reading, are you? >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 15:24:14 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:24:14 -0700 Subject: A poet slips In-Reply-To: <20100930171403.B234.BA0BAB47@research.haifa.ac.il> Message-ID: Ron, Thanks for the link! --Aya On Thu, 30 Sep 2010, Ron Kuzar wrote: > The most informative source, albeit a bit old, is: > Blau, Joshua. 1981. The renaissance of modern Hebrew and modern standard > Arabic : parallels and differences in the revival of two semitic languages > http://books.google.com/books?id=EwbvrNRcaNIC&printsec=frontcover&dq=blau+revival+arabic&source=bl&ots=P_1nIzaFDb&sig=zCfhko8v9UY8rbZC3-OKp_P3kq4&hl=en&ei=pqikTM3NOYiPswbJoYWiCA&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBIQ6AEwAA#v=onepage&q&f=false > Directionality has been suggested in scholarly circles as a problem. I > am not an expert on this issue, and I do NOT believe it could be true. > Ron > > > > On Thu, 30 Sep 2010 07:51:29 -0700 (PDT) > "A. Katz" wrote: > >> Ron, >> >> Most of what you said made sense to me. I am not familiar with the revival >> of Arabic and would be happy if you posted a link for all us Funknetters >> on this topic. It makes sense to see different versions of the langauge as >> gradient. >> >> One thing you said, though, struck me as weird. "Despite the right-to-left >> direction...." You're not seriously suggesting that RTL or LTR makes a >> difference in ease of reading, are you? >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> > > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > From john at research.haifa.ac.il Thu Sep 30 15:26:50 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 17:26:50 +0200 Subject: A poet slips In-Reply-To: Message-ID: Yes, actually I thought more and there is some truth to what you said. It seems that at a psychological level Arabs in other countries don't consider them to be 'real Arabs'. A student of mine who went to the Egyptian national museum in Cairo saw that there were radically different entry fees posted for 'Arabs' and 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in fact he was a Druze who had served in the Israeli Army, but they had no way of knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker asked to see his passport and when he saw that it was from Israel informed him that he was not an Arab and would have to pay the higher 'non-Arab' fee. I have told this story to many of my Israeli Arab students and not a single one has even expressed surprise--they completely expect such treatment. When I suggest that this means that they are not regarded as being real Arabs by Arabs from other countries, they admit that this seems to be true--but they still regard themselves as being Arabs. I then sum up the situation by saying that the people called 'Israeli Arabs' are not considered to be Israelis by Israelis and they are not considered to be Arabs by Arabs. It's complicated... With regard to the obstacles to integration, it really is true that neither side wants integration if this threatens their distinctive identities. In no sense can it be said that the Jewish side rejects assimilation more than the Arab side. I remember one class in which a Christian Arab woman reported temporarily sending her child to a Jewish kindergarten because it was the most convenient one--but when the child came home one day and insisted on lighting candles on Friday evening, the parents realized that this was too much and they needed another solution. Everyone in the class, Jews and Arabs, recognized this--except for an American visiting scholar who said 'What's the problem, it's cute!'. Everyone looked at him like he'd landed from outer space. The difference is more that the Arab side wants EQUALITY, but (contrary to what Americans generally think) that isn't the same thing as integration. John Quoting "A. Katz" : > John, > > Well, at least I'm glad you realize that one hundred years ago, and > possibly even more recently, they did not think of themselves as Arabs. > I still wonder whether among themselves they say "Arabs". Isn't there > another word? If any of them tried to emigrate to Saudi Arabia, I doubt > very much they'd be called "Arabs." For that matter, what do Jordanians > call non-Jewish Israelis? > > The biggest obstacle to allowing for integration in the schools is not the > Palestinian population -- it's those among the Israelis who adopt a "Jews > first" attitude. > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Aya, > > I ask them constantly if they think they are Arabs (60% of my students > > are Arabic speakers and this is a central topic of many classes that I > teach) > > and almost without exception they do. The only exceptions are some Druze > > (mostly males) and most Maronites. I am aware that 100 years ago Arabic > > speakers living in this area did not consider themselves to be Arabs > (except > > for the Bedouins), but the situation has completely changed. > > > > It is totally unrealistic to put Israeli Arabs as a group into > Hebrew-speaking > > schools. No one wants it, not the Jews and not the Arabs. There are a tiny > > number of Arabs who for one reason or another send their children to > > Hebrew-speaking schools (for example the author Sayed Kashua) but this > > is insignificant. > > > > I don't know what the situation was when you were a child, and I don't know > > about the situation among religious Jews, but secular Jewish children do > not > > learn to read by reading Genesis today. They learn to read with secular > texts > > in 1st grade and go on to the Bible in 2nd grade, but far from learning to > > read by reading the Bible, teachers have to explain what's written in the > > Bible to the students. NO ONE can understand it without an explanation from > the > > teacher or their parents. > > > > Ron--I agree with everything you say except that I believe that the radical > > difference between spoken and written Arabic definitely is a serious > problem > > for literacy. I would not be so quick to dismiss the effect of the Arabic > > writing system--we had a conference on this topic in Haifa in May and the > > general consensus of people who sounded like they knew what they were > talking > > about was that it is a significant problem-- but they hadn't done > > cross-linguistic research and this isn't something I know enough about to > have > > strong opinions one way or the other. Also, the usage of written colloquial > > Arabic is basically universal among all Israeli Arabs under the age of 25 > who > > have a cellular phone. > > This is definitely not just a privileged section of the population, it's > most > > people of the relevant age group. > > > > John > > > > > > > > > > > > Quoting "A. Katz" : > > > >> John, > >> > >> These people you speak of are not Arabs. Some of them are Moslem and they > >> read the Quran in the original. Some of them are not Moslem. All of them > >> speak a local dialect of Arabic. Ask them sometimes if they think they are > >> Arabs. > >> > >> Trying to turn every dialect into a separate language with a separate > >> writing system is a way to try to disunite people. But a common language, > >> however differently it is pronounced, unites disparate people. Australians > >> and Cockneys and Indians and Americans speak sometimes mutually > >> unintelligible versions of English. Using the same writing system and > >> the same classic texts unites them. > >> > >> Instead of telling people they should magnify every difference, why not > >> offer to share your language with them? Hebrew could be a uniting factor > >> if spoken in all Israeli schools. > >> > >> --Aya > >> > >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > >> > >> > >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >> > >>> Learning the classical language is like pulling teeth for Arabs kids > also. > >> They > >>> just can't publicly say it because that would make them bad Arabs. It's > >> just > >>> ridiculous. I have an even better plan for the Jewish kids--teach them to > >> use > >>> the written version of the spoken language which Arab kids are using for > >>> Facebook. That's how kids make friends these days anyway. > >>> Best wishes, > >>> John > >>> > >>> > >>> > >>> > >>> > >>> Quoting Tom Givon : > >>> > >>>> > >>>> > >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >>>> Classical (after 1949 things changed...). That book may still exist, > >>>> you might track it down. It was easy, a cinch really. I concluded it was > >>>> really just Hebrew with a few trivial transformations in Phonology & > >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > >>>> on a few sessions of a faculty study group at UO who were trying to > >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >>>> well, in 1939 (or was it 1943?). > >>>> Cheers, TG > >>>> > >>>> ============== > >>>> > >>>> > >>>> john at research.haifa.ac.il wrote: > >>>>> Tom, > >>>>> I've been trying for several years to get Israeli Arabs and Jews to > >>>> seriously > >>>>> consider the possibility of educating Israeli Arabs in a written > version > >> of > >>>>> their spoken language, as you suggest, at least through 3rd grade. I > can > >>>> send > >>>>> you some things I've written on this topic, with a lot of > >> cross-linguistic > >>>>> data. But thus far, it isn't working. As with most aspects of 'the > >>>> situation' > >>>>> here, politically active Arabs think that the solution is to pursue the > >>>> same > >>>>> self-destructive strategy which they've been following for almost a > >> century > >>>> but > >>>>> with even more vigor and steadfastness (e.g. Masalha points the blame > at > >>>> the > >>>>> Arab media because they do not 'provide the linguistic richness of > formal > >>>>> Arabic') while Jews are basically content to let the Arabs stew in > their > >>>> own > >>>>> juices so that they can reap the benefits. There may, however, be hope > >> for > >>>> the > >>>>> future in the form of the radically increased usage of written forms of > >>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, > >>>> emails, > >>>>> etc., by Arabic speakers below the age of 30, which will soon come to > >>>> seriously > >>>>> threaten the status of classical Arabic in the same way that the > >> invention > >>>> of > >>>>> the printing press overturned the linguistic hierarchy in Western > Europe > >> by > >>>>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>>>> > >>>>> I realize that Masalha as well as the researchers he referred to don't > >>>>> necessarily know what they're talking about regarding the brain. But > the > >>>>> general point is still potentially significant--that the connected > script > >>>> which > >>>>> Arabic uses as well as the multiple forms which many Arabic letters > have > >>>> may be > >>>>> a significant obstacle to literacy, however this may be related to > >>>> processes in > >>>>> the brain. > >>>>> Best wishes, > >>>>> John > >>>>> > >>>>> > >>>>> > >>>>> Quoting Tom Givon : > >>>>> > >>>>> > >>>>>> A GREAT POET CAN STILL GET IT > >> WRONG > >>>>>> > >>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >>>>>> by stumbling a on his truly great quote: > >>>>>> "All fixed identities are imposed from the outside. > >>>>>> Whoever has a clear identity knows it can assume > >>>>>> multiple forms". > >>>>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>>>> intended--surely got only one third of the story right. At first, he > was > >>>>>> led astray be the academic researchers he cited, who claimed that the > >>>>>> lagging reading skills of Israeli-Arab students is correlated to > >>>>>> lagging R-hemisphere activity, then explained this neurological lag > >>>>>> by suggesting that the Arab script requires more contextual analysis. > >>>>>> But it is the R-hemisphere of the human cortex that is more context > >>>>>> oriented, less automated. If Arab-reading students required more > >>>>>> contextual labor, it should have been registered as a higher > >>>>>> R-hemisphere activity, not lower. > >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>>>> among many others, have established beyond reasonable doubt that > >>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>>>> module on the boundary of the occipital and temporal lobes (just past > >>>>>> Brodman's Area 19), along the ventral visual object-recognition > >>>>>> 'stream' that flows from the back to the front of the L-cortex. And > >>>>>> the L-cortex is in general responsible for the more automated--less > >>>>>> context-dependent--processing of language (as well as visual, motor > and > >>>>>> other skills). The visual word-recognition module is, in turn, > >>>>>> recruited from the pre-existing visual object-recognition > >>>>>> ventral-stream module. A considerable amount of life-time practice > and > >>>>>> repetition is required to affect this late-cultural adaptation. The > >>>>>> human brain is not (yet) genetically configured at birth for > >>>>>> visual-word recognition, only for visual-object recognition. A similar > >>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & > >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>>>> in the L-pareita lobe, an area originally configured for analysis of > >>>>>> object-location in space. > >>>>>> Mr. Masalha then, on his own, points out to a more > >>>>>> plausible right answer: Arab students, in Israel as well as all over > the > >>>>>> Arab world, are not taught literacy in their native language > >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>>>> language. The discrepancy would be just as great if Israeli kids > were > >>>>>> taught their Hebrew literacy first in the language of Genesis; or if > >>>>>> French students were taught literacy first in the language of La > >>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or > >>>>>> English-speaking kids in the language of Beowolf. As far as my frail > >>>>>> guessing powers go, remedying the situation would be much easier by > >>>>>> combining two well-known verities of second language acquisition: (a) > >>>>>> Teach them both early, together--'co-ordinated bilingualism'. And > (b), > >>>>>> teach literacy first in the student's spoken native language; only > then > >>>>>> gradually 'stretch' it to more literary genres. This method, bhy the > >>>>>> way, was suggested in the late 1930's by no other than L. > Bloomfield, > >>>>>> in a book outlining a 'phonics-first ' literacy program for native > >>>>>> English speakers. Rather than depriving Israeli-Arab students of > >>>>>> literacy in their own--equally glorious--native language, just teach > >>>>>> them smart. > >>>>>> For his last culprit, the presumed--tho hardly > >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>>>> recognized prejudices of the educated classes. While readily > endorsing > >>>>>> his aesthetic sentiments about modern media, I would still like to > point > >>>>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>>>> spoken native language, and if anything should facilitate the easier > >>>>>> initial acquisition of native-language literacy. Respectuosamente, > >>>>>> ma'-salaam, > >>>>>> > >>>>>> > >>>>>> T. Givón > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > ------------------------------------------------------------------------ > >>>>> This message was sent using IMP, the Webmail Program of Haifa > University > >>>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >>> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From rcameron at uic.edu Thu Sep 30 15:45:48 2010 From: rcameron at uic.edu (Cameron, Richard) Date: Thu, 30 Sep 2010 10:45:48 -0500 Subject: A poet slips In-Reply-To: <1285860410.4ca4ac3acc01a@webmail.haifa.ac.il> Message-ID: John's point about "that the people called 'Israeli Arabs' are not considered to be Israelis by Israelis and they are not considered to be Arabs by Arabs." may (I say, may) have some curious parallels among Puerto Ricans and Mexicans in the States in their relationships with those countries as they will claim that back home they are called Americanos but in the States they are not. The old "Ni de aqui, ni de alla." Not from here, not from there. - Richard Cameron On Thu, September 30, 2010 10:26 am, john at research.haifa.ac.il wrote: > Yes, actually I thought more and there is some truth to what you said. It > seems > that at a psychological level Arabs in other countries don't consider them > to > be 'real Arabs'. A student of mine who went to the Egyptian national > museum in > Cairo saw that there were radically different entry fees posted for > 'Arabs' and > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in > fact > he was a Druze who had served in the Israeli Army, but they had no way of > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker > asked > to see his passport and when he saw that > it was from Israel informed him that he was not an Arab and would have to > pay > the higher 'non-Arab' fee. I have told this story to many of my Israeli > Arab > students and not a single one has even expressed surprise--they completely > expect such treatment. When I suggest that this means that they are not > regarded as being real Arabs by Arabs from other countries, they admit > that > this seems to be true--but they still regard themselves as being Arabs. I > then > sum up the situation by saying that the people called 'Israeli Arabs' are > not > considered to be Israelis by Israelis and they are not considered to be > Arabs > by Arabs. > It's complicated... > > With regard to the obstacles to integration, it really is true that > neither side > wants integration if this threatens their distinctive identities. In no > sense > can it be said that the Jewish side rejects assimilation more than the > Arab > side. I remember one class in which a Christian Arab woman reported > temporarily > sending her child to a Jewish kindergarten because it was the most > convenient > one--but when the child came home one day and insisted on lighting candles > on > Friday evening, the parents realized that this was too much and they > needed > another solution. Everyone in the class, Jews and Arabs, recognized > this--except for an American visiting scholar who said 'What's the > problem, > it's cute!'. Everyone looked at him like he'd landed from outer space. The > difference is more that the Arab side wants EQUALITY, but (contrary to > what > Americans generally think) that isn't the same thing as integration. > > John > > > > Quoting "A. Katz" : > >> John, >> >> Well, at least I'm glad you realize that one hundred years ago, and >> possibly even more recently, they did not think of themselves as Arabs. >> I still wonder whether among themselves they say "Arabs". Isn't there >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >> very much they'd be called "Arabs." For that matter, what do Jordanians >> call non-Jewish Israelis? >> >> The biggest obstacle to allowing for integration in the schools is not >> the >> Palestinian population -- it's those among the Israelis who adopt a >> "Jews >> first" attitude. >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >> > Aya, >> > I ask them constantly if they think they are Arabs (60% of my students >> > are Arabic speakers and this is a central topic of many classes that I >> teach) >> > and almost without exception they do. The only exceptions are some >> Druze >> > (mostly males) and most Maronites. I am aware that 100 years ago >> Arabic >> > speakers living in this area did not consider themselves to be Arabs >> (except >> > for the Bedouins), but the situation has completely changed. >> > >> > It is totally unrealistic to put Israeli Arabs as a group into >> Hebrew-speaking >> > schools. No one wants it, not the Jews and not the Arabs. There are a >> tiny >> > number of Arabs who for one reason or another send their children to >> > Hebrew-speaking schools (for example the author Sayed Kashua) but this >> > is insignificant. >> > >> > I don't know what the situation was when you were a child, and I don't >> know >> > about the situation among religious Jews, but secular Jewish children >> do >> not >> > learn to read by reading Genesis today. They learn to read with >> secular >> texts >> > in 1st grade and go on to the Bible in 2nd grade, but far from >> learning to >> > read by reading the Bible, teachers have to explain what's written in >> the >> > Bible to the students. NO ONE can understand it without an explanation >> from >> the >> > teacher or their parents. >> > >> > Ron--I agree with everything you say except that I believe that the >> radical >> > difference between spoken and written Arabic definitely is a serious >> problem >> > for literacy. I would not be so quick to dismiss the effect of the >> Arabic >> > writing system--we had a conference on this topic in Haifa in May and >> the >> > general consensus of people who sounded like they knew what they were >> talking >> > about was that it is a significant problem-- but they hadn't done >> > cross-linguistic research and this isn't something I know enough about >> to >> have >> > strong opinions one way or the other. Also, the usage of written >> colloquial >> > Arabic is basically universal among all Israeli Arabs under the age of >> 25 >> who >> > have a cellular phone. >> > This is definitely not just a privileged section of the population, >> it's >> most >> > people of the relevant age group. >> > >> > John >> > >> > >> > >> > >> > >> > Quoting "A. Katz" : >> > >> >> John, >> >> >> >> These people you speak of are not Arabs. Some of them are Moslem and >> they >> >> read the Quran in the original. Some of them are not Moslem. All of >> them >> >> speak a local dialect of Arabic. Ask them sometimes if they think >> they are >> >> Arabs. >> >> >> >> Trying to turn every dialect into a separate language with a separate >> >> writing system is a way to try to disunite people. But a common >> language, >> >> however differently it is pronounced, unites disparate people. >> Australians >> >> and Cockneys and Indians and Americans speak sometimes mutually >> >> unintelligible versions of English. Using the same writing system and >> >> the same classic texts unites them. >> >> >> >> Instead of telling people they should magnify every difference, why >> not >> >> offer to share your language with them? Hebrew could be a uniting >> factor >> >> if spoken in all Israeli schools. >> >> >> >> --Aya >> >> >> >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >> >> >> >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >> >> >>> Learning the classical language is like pulling teeth for Arabs kids >> also. >> >> They >> >>> just can't publicly say it because that would make them bad Arabs. >> It's >> >> just >> >>> ridiculous. I have an even better plan for the Jewish kids--teach >> them to >> >> use >> >>> the written version of the spoken language which Arab kids are using >> for >> >>> Facebook. That's how kids make friends these days anyway. >> >>> Best wishes, >> >>> John >> >>> >> >>> >> >>> >> >>> >> >>> >> >>> Quoting Tom Givon : >> >>> >> >>>> >> >>>> >> >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, >> our >> >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got >> to >> >>>> Classical (after 1949 things changed...). That book may still >> exist, >> >>>> you might track it down. It was easy, a cinch really. I concluded >> it was >> >>>> really just Hebrew with a few trivial transformations in Phonology >> & >> >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I >> sat >> >>>> on a few sessions of a faculty study group at UO who were trying to >> >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. >> Boy, >> >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, >> and >> >>>> well, in 1939 (or was it 1943?). >> >>>> Cheers, TG >> >>>> >> >>>> ============== >> >>>> >> >>>> >> >>>> john at research.haifa.ac.il wrote: >> >>>>> Tom, >> >>>>> I've been trying for several years to get Israeli Arabs and Jews >> to >> >>>> seriously >> >>>>> consider the possibility of educating Israeli Arabs in a written >> version >> >> of >> >>>>> their spoken language, as you suggest, at least through 3rd grade. >> I >> can >> >>>> send >> >>>>> you some things I've written on this topic, with a lot of >> >> cross-linguistic >> >>>>> data. But thus far, it isn't working. As with most aspects of 'the >> >>>> situation' >> >>>>> here, politically active Arabs think that the solution is to >> pursue the >> >>>> same >> >>>>> self-destructive strategy which they've been following for almost >> a >> >> century >> >>>> but >> >>>>> with even more vigor and steadfastness (e.g. Masalha points the >> blame >> at >> >>>> the >> >>>>> Arab media because they do not 'provide the linguistic richness of >> formal >> >>>>> Arabic') while Jews are basically content to let the Arabs stew in >> their >> >>>> own >> >>>>> juices so that they can reap the benefits. There may, however, be >> hope >> >> for >> >>>> the >> >>>>> future in the form of the radically increased usage of written >> forms of >> >>>>> colloquial Arabic dialects in electronic media such as Facebook, >> blogs, >> >>>> emails, >> >>>>> etc., by Arabic speakers below the age of 30, which will soon come >> to >> >>>> seriously >> >>>>> threaten the status of classical Arabic in the same way that the >> >> invention >> >>>> of >> >>>>> the printing press overturned the linguistic hierarchy in Western >> Europe >> >> by >> >>>>> overthrowing Latin. I'm working on this too. We'll see what >> happens. >> >>>>> >> >>>>> I realize that Masalha as well as the researchers he referred to >> don't >> >>>>> necessarily know what they're talking about regarding the brain. >> But >> the >> >>>>> general point is still potentially significant--that the connected >> script >> >>>> which >> >>>>> Arabic uses as well as the multiple forms which many Arabic >> letters >> have >> >>>> may be >> >>>>> a significant obstacle to literacy, however this may be related to >> >>>> processes in >> >>>>> the brain. >> >>>>> Best wishes, >> >>>>> John >> >>>>> >> >>>>> >> >>>>> >> >>>>> Quoting Tom Givon : >> >>>>> >> >>>>> >> >>>>>> A GREAT POET CAN STILL GET >> IT >> >> WRONG >> >>>>>> >> >>>>>> I came to know of Salman Masalha ("Arabs, speak >> Hebrew!", >> >>>>>> Haaretz/English, International Herald Tribune 9-27-10) >> accidentally >> >>>>>> by stumbling a on his truly great quote: >> >>>>>> "All fixed identities are imposed from the >> outside. >> >>>>>> Whoever has a clear identity knows it can >> assume >> >>>>>> multiple forms". >> >>>>>> In the context of Palestine/Israel, what a breath of fresh, >> rare >> >>>>>> clarity. Still, like the rest of us mortals, a great poet can >> on >> >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; >> pun >> >>>>>> intended--surely got only one third of the story right. At first, >> he >> was >> >>>>>> led astray be the academic researchers he cited, who claimed that >> the >> >>>>>> lagging reading skills of Israeli-Arab students is correlated >> to >> >>>>>> lagging R-hemisphere activity, then explained this >> neurological lag >> >>>>>> by suggesting that the Arab script requires more contextual >> analysis. >> >>>>>> But it is the R-hemisphere of the human cortex that is more >> context >> >>>>>> oriented, less automated. If Arab-reading students required more >> >>>>>> contextual labor, it should have been registered as a higher >> >>>>>> R-hemisphere activity, not lower. >> >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. >> Dahane, >> >>>>>> among many others, have established beyond reasonable doubt >> that >> >>>>>> written words in all languages (English, Mandarin, Hindi, >> Amharic, >> >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an >> L-hemisphere >> >>>>>> module on the boundary of the occipital and temporal lobes (just >> past >> >>>>>> Brodman's Area 19), along the ventral visual object-recognition >> >>>>>> 'stream' that flows from the back to the front of the L-cortex. >> And >> >>>>>> the L-cortex is in general responsible for the more >> automated--less >> >>>>>> context-dependent--processing of language (as well as visual, >> motor >> and >> >>>>>> other skills). The visual word-recognition module is, in turn, >> >>>>>> recruited from the pre-existing visual object-recognition >> >>>>>> ventral-stream module. A considerable amount of life-time >> practice >> and >> >>>>>> repetition is required to affect this late-cultural adaptation. >> The >> >>>>>> human brain is not (yet) genetically configured at birth for >> >>>>>> visual-word recognition, only for visual-object recognition. A >> similar >> >>>>>> cultural adaptation, this one for math, has been shown for >> (Dahaene & >> >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver >> Sachs) >> >>>>>> in the L-pareita lobe, an area originally configured for analysis >> of >> >>>>>> object-location in space. >> >>>>>> Mr. Masalha then, on his own, points out to a more >> >>>>>> plausible right answer: Arab students, in Israel as well as all >> over >> the >> >>>>>> Arab world, are not taught literacy in their native language >> >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen >> literary >> >>>>>> instrument harking back 1,400 years or more. That is, in a >> foreign >> >>>>>> language. The discrepancy would be just as great if Israeli >> kids >> were >> >>>>>> taught their Hebrew literacy first in the language of Genesis; >> or if >> >>>>>> French students were taught literacy first in the language of >> La >> >>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. >> Or >> >>>>>> English-speaking kids in the language of Beowolf. As far as my >> frail >> >>>>>> guessing powers go, remedying the situation would be much easier >> by >> >>>>>> combining two well-known verities of second language acquisition: >> (a) >> >>>>>> Teach them both early, together--'co-ordinated bilingualism'. >> And >> (b), >> >>>>>> teach literacy first in the student's spoken native language; >> only >> then >> >>>>>> gradually 'stretch' it to more literary genres. This method, bhy >> the >> >>>>>> way, was suggested in the late 1930's by no other than L. >> Bloomfield, >> >>>>>> in a book outlining a 'phonics-first ' literacy program for >> native >> >>>>>> English speakers. Rather than depriving Israeli-Arab students of >> >>>>>> literacy in their own--equally glorious--native language, just >> teach >> >>>>>> them smart. >> >>>>>> For his last culprit, the presumed--tho hardly >> >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >> >>>>>> recognized prejudices of the educated classes. While readily >> endorsing >> >>>>>> his aesthetic sentiments about modern media, I would still like >> to >> point >> >>>>>> out that the 'vulgar' genre is much closer to the Arab students' >> >>>>>> spoken native language, and if anything should facilitate the >> easier >> >>>>>> initial acquisition of native-language literacy. >> Respectuosamente, >> >>>>>> ma'-salaam, >> >>>>>> >> >>>>>> >> >>>>>> T. Givón >> >>>>>> >> >>>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> ------------------------------------------------------------------------ >> >>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >> >>>>> >> >>>> >> >>>> >> >>> >> >>> >> >>> >> >>> >> >>> ------------------------------------------------------------------------ >> >>> This message was sent using IMP, the Webmail Program of Haifa >> University >> >>> >> >>> >> > >> > >> > >> > >> > ------------------------------------------------------------------------ >> > This message was sent using IMP, the Webmail Program of Haifa >> University >> > >> > > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 16:10:45 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 09:10:45 -0700 Subject: A poet slips In-Reply-To: <1285860410.4ca4ac3acc01a@webmail.haifa.ac.il> Message-ID: John, Thanks for sharing the story about the Egyptian national museum. It's ironic that this would happen there, of all places, because ethnically most Egyptians are not Arab, either. Yes, it is complicated. Everybody wants to belong somewhere, and the reason this particular group of people are denied this opportunity is through a weird fluke of history. To Arabs, they are Judeans or Palestinians. To Jews, they are Arabs. The story about the Kindergarten is very sad. Why does religion have to be part of the curriculum in a public school? Separation of Church and State would go a long way toward integration and ultimately peace. --Aya But, I agree, it is complicated. On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Yes, actually I thought more and there is some truth to what you said. It seems > that at a psychological level Arabs in other countries don't consider them to > be 'real Arabs'. A student of mine who went to the Egyptian national museum in > Cairo saw that there were radically different entry fees posted for 'Arabs' and > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in fact > he was a Druze who had served in the Israeli Army, but they had no way of > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker asked > to see his passport and when he saw that > it was from Israel informed him that he was not an Arab and would have to pay > the higher 'non-Arab' fee. I have told this story to many of my Israeli Arab > students and not a single one has even expressed surprise--they completely > expect such treatment. When I suggest that this means that they are not > regarded as being real Arabs by Arabs from other countries, they admit that > this seems to be true--but they still regard themselves as being Arabs. I then > sum up the situation by saying that the people called 'Israeli Arabs' are not > considered to be Israelis by Israelis and they are not considered to be Arabs > by Arabs. > It's complicated... > > With regard to the obstacles to integration, it really is true that neither side > wants integration if this threatens their distinctive identities. In no sense > can it be said that the Jewish side rejects assimilation more than the Arab > side. I remember one class in which a Christian Arab woman reported temporarily > sending her child to a Jewish kindergarten because it was the most convenient > one--but when the child came home one day and insisted on lighting candles on > Friday evening, the parents realized that this was too much and they needed > another solution. Everyone in the class, Jews and Arabs, recognized > this--except for an American visiting scholar who said 'What's the problem, > it's cute!'. Everyone looked at him like he'd landed from outer space. The > difference is more that the Arab side wants EQUALITY, but (contrary to what > Americans generally think) that isn't the same thing as integration. > > John > > > > Quoting "A. Katz" : > >> John, >> >> Well, at least I'm glad you realize that one hundred years ago, and >> possibly even more recently, they did not think of themselves as Arabs. >> I still wonder whether among themselves they say "Arabs". Isn't there >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >> very much they'd be called "Arabs." For that matter, what do Jordanians >> call non-Jewish Israelis? >> >> The biggest obstacle to allowing for integration in the schools is not the >> Palestinian population -- it's those among the Israelis who adopt a "Jews >> first" attitude. >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Aya, >>> I ask them constantly if they think they are Arabs (60% of my students >>> are Arabic speakers and this is a central topic of many classes that I >> teach) >>> and almost without exception they do. The only exceptions are some Druze >>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic >>> speakers living in this area did not consider themselves to be Arabs >> (except >>> for the Bedouins), but the situation has completely changed. >>> >>> It is totally unrealistic to put Israeli Arabs as a group into >> Hebrew-speaking >>> schools. No one wants it, not the Jews and not the Arabs. There are a tiny >>> number of Arabs who for one reason or another send their children to >>> Hebrew-speaking schools (for example the author Sayed Kashua) but this >>> is insignificant. >>> >>> I don't know what the situation was when you were a child, and I don't know >>> about the situation among religious Jews, but secular Jewish children do >> not >>> learn to read by reading Genesis today. They learn to read with secular >> texts >>> in 1st grade and go on to the Bible in 2nd grade, but far from learning to >>> read by reading the Bible, teachers have to explain what's written in the >>> Bible to the students. NO ONE can understand it without an explanation from >> the >>> teacher or their parents. >>> >>> Ron--I agree with everything you say except that I believe that the radical >>> difference between spoken and written Arabic definitely is a serious >> problem >>> for literacy. I would not be so quick to dismiss the effect of the Arabic >>> writing system--we had a conference on this topic in Haifa in May and the >>> general consensus of people who sounded like they knew what they were >> talking >>> about was that it is a significant problem-- but they hadn't done >>> cross-linguistic research and this isn't something I know enough about to >> have >>> strong opinions one way or the other. Also, the usage of written colloquial >>> Arabic is basically universal among all Israeli Arabs under the age of 25 >> who >>> have a cellular phone. >>> This is definitely not just a privileged section of the population, it's >> most >>> people of the relevant age group. >>> >>> John >>> >>> >>> >>> >>> >>> Quoting "A. Katz" : >>> >>>> John, >>>> >>>> These people you speak of are not Arabs. Some of them are Moslem and they >>>> read the Quran in the original. Some of them are not Moslem. All of them >>>> speak a local dialect of Arabic. Ask them sometimes if they think they are >>>> Arabs. >>>> >>>> Trying to turn every dialect into a separate language with a separate >>>> writing system is a way to try to disunite people. But a common language, >>>> however differently it is pronounced, unites disparate people. Australians >>>> and Cockneys and Indians and Americans speak sometimes mutually >>>> unintelligible versions of English. Using the same writing system and >>>> the same classic texts unites them. >>>> >>>> Instead of telling people they should magnify every difference, why not >>>> offer to share your language with them? Hebrew could be a uniting factor >>>> if spoken in all Israeli schools. >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >>>> >>>> >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>> >>>>> Learning the classical language is like pulling teeth for Arabs kids >> also. >>>> They >>>>> just can't publicly say it because that would make them bad Arabs. It's >>>> just >>>>> ridiculous. I have an even better plan for the Jewish kids--teach them to >>>> use >>>>> the written version of the spoken language which Arab kids are using for >>>>> Facebook. That's how kids make friends these days anyway. >>>>> Best wishes, >>>>> John >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Quoting Tom Givon : >>>>> >>>>>> >>>>>> >>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>>>> Classical (after 1949 things changed...). That book may still exist, >>>>>> you might track it down. It was easy, a cinch really. I concluded it was >>>>>> really just Hebrew with a few trivial transformations in Phonology & >>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>>>>> on a few sessions of a faculty study group at UO who were trying to >>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>>>> well, in 1939 (or was it 1943?). >>>>>> Cheers, TG >>>>>> >>>>>> ============== >>>>>> >>>>>> >>>>>> john at research.haifa.ac.il wrote: >>>>>>> Tom, >>>>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>>>> seriously >>>>>>> consider the possibility of educating Israeli Arabs in a written >> version >>>> of >>>>>>> their spoken language, as you suggest, at least through 3rd grade. I >> can >>>>>> send >>>>>>> you some things I've written on this topic, with a lot of >>>> cross-linguistic >>>>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>>>> situation' >>>>>>> here, politically active Arabs think that the solution is to pursue the >>>>>> same >>>>>>> self-destructive strategy which they've been following for almost a >>>> century >>>>>> but >>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame >> at >>>>>> the >>>>>>> Arab media because they do not 'provide the linguistic richness of >> formal >>>>>>> Arabic') while Jews are basically content to let the Arabs stew in >> their >>>>>> own >>>>>>> juices so that they can reap the benefits. There may, however, be hope >>>> for >>>>>> the >>>>>>> future in the form of the radically increased usage of written forms of >>>>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>>>> emails, >>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>>>> seriously >>>>>>> threaten the status of classical Arabic in the same way that the >>>> invention >>>>>> of >>>>>>> the printing press overturned the linguistic hierarchy in Western >> Europe >>>> by >>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>>>> >>>>>>> I realize that Masalha as well as the researchers he referred to don't >>>>>>> necessarily know what they're talking about regarding the brain. But >> the >>>>>>> general point is still potentially significant--that the connected >> script >>>>>> which >>>>>>> Arabic uses as well as the multiple forms which many Arabic letters >> have >>>>>> may be >>>>>>> a significant obstacle to literacy, however this may be related to >>>>>> processes in >>>>>>> the brain. >>>>>>> Best wishes, >>>>>>> John >>>>>>> >>>>>>> >>>>>>> >>>>>>> Quoting Tom Givon : >>>>>>> >>>>>>> >>>>>>>> A GREAT POET CAN STILL GET IT >>>> WRONG >>>>>>>> >>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>>>>> by stumbling a on his truly great quote: >>>>>>>> "All fixed identities are imposed from the outside. >>>>>>>> Whoever has a clear identity knows it can assume >>>>>>>> multiple forms". >>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>>>> intended--surely got only one third of the story right. At first, he >> was >>>>>>>> led astray be the academic researchers he cited, who claimed that the >>>>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>>>>> by suggesting that the Arab script requires more contextual analysis. >>>>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>>>> oriented, less automated. If Arab-reading students required more >>>>>>>> contextual labor, it should have been registered as a higher >>>>>>>> R-hemisphere activity, not lower. >>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>>>> among many others, have established beyond reasonable doubt that >>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>>>> module on the boundary of the occipital and temporal lobes (just past >>>>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>>>>> the L-cortex is in general responsible for the more automated--less >>>>>>>> context-dependent--processing of language (as well as visual, motor >> and >>>>>>>> other skills). The visual word-recognition module is, in turn, >>>>>>>> recruited from the pre-existing visual object-recognition >>>>>>>> ventral-stream module. A considerable amount of life-time practice >> and >>>>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>>>> human brain is not (yet) genetically configured at birth for >>>>>>>> visual-word recognition, only for visual-object recognition. A similar >>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>>>> object-location in space. >>>>>>>> Mr. Masalha then, on his own, points out to a more >>>>>>>> plausible right answer: Arab students, in Israel as well as all over >> the >>>>>>>> Arab world, are not taught literacy in their native language >>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>>>> language. The discrepancy would be just as great if Israeli kids >> were >>>>>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>>>>> French students were taught literacy first in the language of La >>>>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or >>>>>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>>>>> guessing powers go, remedying the situation would be much easier by >>>>>>>> combining two well-known verities of second language acquisition: (a) >>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And >> (b), >>>>>>>> teach literacy first in the student's spoken native language; only >> then >>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>>>>> way, was suggested in the late 1930's by no other than L. >> Bloomfield, >>>>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>>>> literacy in their own--equally glorious--native language, just teach >>>>>>>> them smart. >>>>>>>> For his last culprit, the presumed--tho hardly >>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>>>> recognized prejudices of the educated classes. While readily >> endorsing >>>>>>>> his aesthetic sentiments about modern media, I would still like to >> point >>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>>>> spoken native language, and if anything should facilitate the easier >>>>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>>>> ma'-salaam, >>>>>>>> >>>>>>>> >>>>>>>> T. Givón >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >> ------------------------------------------------------------------------ >>>>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From john at research.haifa.ac.il Thu Sep 30 16:26:14 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 18:26:14 +0200 Subject: A poet slips In-Reply-To: Message-ID: Stop thinking like an American and thinking that other places should be like the wonderful United States. No one here thinks the Kindergarten situation is sad except for American tourists. On the other hand, the situation of Israeli Arabs is sad, and there is pretty widespread recognition of this here. Richard--yes, this situation isn't so unusual. Russians Israelis say that they aren't thought of as Israelis here or as Russians in Russia. John Quoting "A. Katz" : > John, > > Thanks for sharing the story about the Egyptian national museum. It's > ironic that this would happen there, of all places, because ethnically > most Egyptians are not Arab, either. > > Yes, it is complicated. Everybody wants to belong somewhere, and the > reason this particular group of people are denied this opportunity is > through a weird fluke of history. To Arabs, they are Judeans or > Palestinians. To Jews, they are Arabs. > > The story about the Kindergarten is very sad. Why does religion have to be > part of the curriculum in a public school? Separation of Church and State > would go a long way toward integration and ultimately peace. > > --Aya > > > But, I agree, it is complicated. > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Yes, actually I thought more and there is some truth to what you said. It > seems > > that at a psychological level Arabs in other countries don't consider them > to > > be 'real Arabs'. A student of mine who went to the Egyptian national museum > in > > Cairo saw that there were radically different entry fees posted for 'Arabs' > and > > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in > fact > > he was a Druze who had served in the Israeli Army, but they had no way of > > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker > asked > > to see his passport and when he saw that > > it was from Israel informed him that he was not an Arab and would have to > pay > > the higher 'non-Arab' fee. I have told this story to many of my Israeli > Arab > > students and not a single one has even expressed surprise--they completely > > expect such treatment. When I suggest that this means that they are not > > regarded as being real Arabs by Arabs from other countries, they admit that > > this seems to be true--but they still regard themselves as being Arabs. I > then > > sum up the situation by saying that the people called 'Israeli Arabs' are > not > > considered to be Israelis by Israelis and they are not considered to be > Arabs > > by Arabs. > > It's complicated... > > > > With regard to the obstacles to integration, it really is true that neither > side > > wants integration if this threatens their distinctive identities. In no > sense > > can it be said that the Jewish side rejects assimilation more than the Arab > > side. I remember one class in which a Christian Arab woman reported > temporarily > > sending her child to a Jewish kindergarten because it was the most > convenient > > one--but when the child came home one day and insisted on lighting candles > on > > Friday evening, the parents realized that this was too much and they needed > > another solution. Everyone in the class, Jews and Arabs, recognized > > this--except for an American visiting scholar who said 'What's the problem, > > it's cute!'. Everyone looked at him like he'd landed from outer space. The > > difference is more that the Arab side wants EQUALITY, but (contrary to what > > Americans generally think) that isn't the same thing as integration. > > > > John > > > > > > > > Quoting "A. Katz" : > > > >> John, > >> > >> Well, at least I'm glad you realize that one hundred years ago, and > >> possibly even more recently, they did not think of themselves as Arabs. > >> I still wonder whether among themselves they say "Arabs". Isn't there > >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt > >> very much they'd be called "Arabs." For that matter, what do Jordanians > >> call non-Jewish Israelis? > >> > >> The biggest obstacle to allowing for integration in the schools is not the > >> Palestinian population -- it's those among the Israelis who adopt a "Jews > >> first" attitude. > >> > >> --Aya > >> > >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > >> > >> > >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >> > >>> Aya, > >>> I ask them constantly if they think they are Arabs (60% of my students > >>> are Arabic speakers and this is a central topic of many classes that I > >> teach) > >>> and almost without exception they do. The only exceptions are some Druze > >>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic > >>> speakers living in this area did not consider themselves to be Arabs > >> (except > >>> for the Bedouins), but the situation has completely changed. > >>> > >>> It is totally unrealistic to put Israeli Arabs as a group into > >> Hebrew-speaking > >>> schools. No one wants it, not the Jews and not the Arabs. There are a > tiny > >>> number of Arabs who for one reason or another send their children to > >>> Hebrew-speaking schools (for example the author Sayed Kashua) but this > >>> is insignificant. > >>> > >>> I don't know what the situation was when you were a child, and I don't > know > >>> about the situation among religious Jews, but secular Jewish children do > >> not > >>> learn to read by reading Genesis today. They learn to read with secular > >> texts > >>> in 1st grade and go on to the Bible in 2nd grade, but far from learning > to > >>> read by reading the Bible, teachers have to explain what's written in the > >>> Bible to the students. NO ONE can understand it without an explanation > from > >> the > >>> teacher or their parents. > >>> > >>> Ron--I agree with everything you say except that I believe that the > radical > >>> difference between spoken and written Arabic definitely is a serious > >> problem > >>> for literacy. I would not be so quick to dismiss the effect of the Arabic > >>> writing system--we had a conference on this topic in Haifa in May and the > >>> general consensus of people who sounded like they knew what they were > >> talking > >>> about was that it is a significant problem-- but they hadn't done > >>> cross-linguistic research and this isn't something I know enough about to > >> have > >>> strong opinions one way or the other. Also, the usage of written > colloquial > >>> Arabic is basically universal among all Israeli Arabs under the age of 25 > >> who > >>> have a cellular phone. > >>> This is definitely not just a privileged section of the population, it's > >> most > >>> people of the relevant age group. > >>> > >>> John > >>> > >>> > >>> > >>> > >>> > >>> Quoting "A. Katz" : > >>> > >>>> John, > >>>> > >>>> These people you speak of are not Arabs. Some of them are Moslem and > they > >>>> read the Quran in the original. Some of them are not Moslem. All of them > >>>> speak a local dialect of Arabic. Ask them sometimes if they think they > are > >>>> Arabs. > >>>> > >>>> Trying to turn every dialect into a separate language with a separate > >>>> writing system is a way to try to disunite people. But a common > language, > >>>> however differently it is pronounced, unites disparate people. > Australians > >>>> and Cockneys and Indians and Americans speak sometimes mutually > >>>> unintelligible versions of English. Using the same writing system and > >>>> the same classic texts unites them. > >>>> > >>>> Instead of telling people they should magnify every difference, why not > >>>> offer to share your language with them? Hebrew could be a uniting factor > >>>> if spoken in all Israeli schools. > >>>> > >>>> --Aya > >>>> > >>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > >>>> > >>>> > >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >>>> > >>>>> Learning the classical language is like pulling teeth for Arabs kids > >> also. > >>>> They > >>>>> just can't publicly say it because that would make them bad Arabs. It's > >>>> just > >>>>> ridiculous. I have an even better plan for the Jewish kids--teach them > to > >>>> use > >>>>> the written version of the spoken language which Arab kids are using > for > >>>>> Facebook. That's how kids make friends these days anyway. > >>>>> Best wishes, > >>>>> John > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> Quoting Tom Givon : > >>>>> > >>>>>> > >>>>>> > >>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, > our > >>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >>>>>> Classical (after 1949 things changed...). That book may still exist, > >>>>>> you might track it down. It was easy, a cinch really. I concluded it > was > >>>>>> really just Hebrew with a few trivial transformations in Phonology & > >>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I > sat > >>>>>> on a few sessions of a faculty study group at UO who were trying to > >>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. > Boy, > >>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >>>>>> well, in 1939 (or was it 1943?). > >>>>>> Cheers, TG > >>>>>> > >>>>>> ============== > >>>>>> > >>>>>> > >>>>>> john at research.haifa.ac.il wrote: > >>>>>>> Tom, > >>>>>>> I've been trying for several years to get Israeli Arabs and Jews to > >>>>>> seriously > >>>>>>> consider the possibility of educating Israeli Arabs in a written > >> version > >>>> of > >>>>>>> their spoken language, as you suggest, at least through 3rd grade. I > >> can > >>>>>> send > >>>>>>> you some things I've written on this topic, with a lot of > >>>> cross-linguistic > >>>>>>> data. But thus far, it isn't working. As with most aspects of 'the > >>>>>> situation' > >>>>>>> here, politically active Arabs think that the solution is to pursue > the > >>>>>> same > >>>>>>> self-destructive strategy which they've been following for almost a > >>>> century > >>>>>> but > >>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame > >> at > >>>>>> the > >>>>>>> Arab media because they do not 'provide the linguistic richness of > >> formal > >>>>>>> Arabic') while Jews are basically content to let the Arabs stew in > >> their > >>>>>> own > >>>>>>> juices so that they can reap the benefits. There may, however, be > hope > >>>> for > >>>>>> the > >>>>>>> future in the form of the radically increased usage of written forms > of > >>>>>>> colloquial Arabic dialects in electronic media such as Facebook, > blogs, > >>>>>> emails, > >>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to > >>>>>> seriously > >>>>>>> threaten the status of classical Arabic in the same way that the > >>>> invention > >>>>>> of > >>>>>>> the printing press overturned the linguistic hierarchy in Western > >> Europe > >>>> by > >>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>>>>>> > >>>>>>> I realize that Masalha as well as the researchers he referred to > don't > >>>>>>> necessarily know what they're talking about regarding the brain. But > >> the > >>>>>>> general point is still potentially significant--that the connected > >> script > >>>>>> which > >>>>>>> Arabic uses as well as the multiple forms which many Arabic letters > >> have > >>>>>> may be > >>>>>>> a significant obstacle to literacy, however this may be related to > >>>>>> processes in > >>>>>>> the brain. > >>>>>>> Best wishes, > >>>>>>> John > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> Quoting Tom Givon : > >>>>>>> > >>>>>>> > >>>>>>>> A GREAT POET CAN STILL GET IT > >>>> WRONG > >>>>>>>> > >>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) > accidentally > >>>>>>>> by stumbling a on his truly great quote: > >>>>>>>> "All fixed identities are imposed from the outside. > >>>>>>>> Whoever has a clear identity knows it can assume > >>>>>>>> multiple forms". > >>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>>>>>> intended--surely got only one third of the story right. At first, he > >> was > >>>>>>>> led astray be the academic researchers he cited, who claimed that > the > >>>>>>>> lagging reading skills of Israeli-Arab students is correlated to > >>>>>>>> lagging R-hemisphere activity, then explained this neurological > lag > >>>>>>>> by suggesting that the Arab script requires more contextual > analysis. > >>>>>>>> But it is the R-hemisphere of the human cortex that is more context > >>>>>>>> oriented, less automated. If Arab-reading students required more > >>>>>>>> contextual labor, it should have been registered as a higher > >>>>>>>> R-hemisphere activity, not lower. > >>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>>>>>> among many others, have established beyond reasonable doubt > that > >>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>>>>>> module on the boundary of the occipital and temporal lobes (just > past > >>>>>>>> Brodman's Area 19), along the ventral visual object-recognition > >>>>>>>> 'stream' that flows from the back to the front of the L-cortex. > And > >>>>>>>> the L-cortex is in general responsible for the more automated--less > >>>>>>>> context-dependent--processing of language (as well as visual, motor > >> and > >>>>>>>> other skills). The visual word-recognition module is, in turn, > >>>>>>>> recruited from the pre-existing visual object-recognition > >>>>>>>> ventral-stream module. A considerable amount of life-time practice > >> and > >>>>>>>> repetition is required to affect this late-cultural adaptation. The > >>>>>>>> human brain is not (yet) genetically configured at birth for > >>>>>>>> visual-word recognition, only for visual-object recognition. A > similar > >>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene > & > >>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>>>>>> in the L-pareita lobe, an area originally configured for analysis of > >>>>>>>> object-location in space. > >>>>>>>> Mr. Masalha then, on his own, points out to a more > >>>>>>>> plausible right answer: Arab students, in Israel as well as all over > >> the > >>>>>>>> Arab world, are not taught literacy in their native language > >>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>>>>>> language. The discrepancy would be just as great if Israeli kids > >> were > >>>>>>>> taught their Hebrew literacy first in the language of Genesis; or > if > >>>>>>>> French students were taught literacy first in the language of La > >>>>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or > >>>>>>>> English-speaking kids in the language of Beowolf. As far as my > frail > >>>>>>>> guessing powers go, remedying the situation would be much easier by > >>>>>>>> combining two well-known verities of second language acquisition: > (a) > >>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And > >> (b), > >>>>>>>> teach literacy first in the student's spoken native language; only > >> then > >>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy > the > >>>>>>>> way, was suggested in the late 1930's by no other than L. > >> Bloomfield, > >>>>>>>> in a book outlining a 'phonics-first ' literacy program for native > >>>>>>>> English speakers. Rather than depriving Israeli-Arab students of > >>>>>>>> literacy in their own--equally glorious--native language, just teach > >>>>>>>> them smart. > >>>>>>>> For his last culprit, the presumed--tho hardly > >>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>>>>>> recognized prejudices of the educated classes. While readily > >> endorsing > >>>>>>>> his aesthetic sentiments about modern media, I would still like to > >> point > >>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>>>>>> spoken native language, and if anything should facilitate the > easier > >>>>>>>> initial acquisition of native-language literacy. Respectuosamente, > >>>>>>>> ma'-salaam, > >>>>>>>> > >>>>>>>> > >>>>>>>> T. Givón > >>>>>>>> > >>>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >> ------------------------------------------------------------------------ > >>>>>>> This message was sent using IMP, the Webmail Program of Haifa > >> University > >>>>>>> > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > ------------------------------------------------------------------------ > >>>>> This message was sent using IMP, the Webmail Program of Haifa > University > >>>>> > >>>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >>> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From amnfn at well.com Thu Sep 30 16:50:12 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 09:50:12 -0700 Subject: A poet slips In-Reply-To: <1285863974.4ca4ba269d9ed@webmail.haifa.ac.il> Message-ID: John, You may think of me as an American, and correctly so, but the reason I am an American is that my father left Israel because separation of Church and State was not the law. Many Israelis whose ancestors came from the European diaspora are also not members of the Jewish religion and do not want that forced on their children. I know of one family that left because the State was trying to force them to circumcize their son. Their children are no longer Israeli. But they could have been! Wanting freedom of religion is not an exclusively American desire. It's universal. Your Kindergarten example showed that children do assimilate easily in the public schools, but that religion is a touchy subject for everybody. Even if you don't care about secular Israelis, please keep in mind that not all Palestinians who speak Arabic are Moslem. There is already a split between religious Islamic fundamentalists and others among the Palestinians, and many lives have been lost because of religious issues. The solution to the problem has to be secular. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Stop thinking like an American and thinking that other places should be like the > wonderful United States. No one here thinks the Kindergarten situation is sad > except for American tourists. > > On the other hand, the situation of Israeli Arabs is sad, and there is pretty > widespread recognition of this here. > > Richard--yes, this situation isn't so unusual. Russians Israelis say that they > aren't thought of as Israelis here or as Russians in Russia. > John > > > Quoting "A. Katz" : > >> John, >> >> Thanks for sharing the story about the Egyptian national museum. It's >> ironic that this would happen there, of all places, because ethnically >> most Egyptians are not Arab, either. >> >> Yes, it is complicated. Everybody wants to belong somewhere, and the >> reason this particular group of people are denied this opportunity is >> through a weird fluke of history. To Arabs, they are Judeans or >> Palestinians. To Jews, they are Arabs. >> >> The story about the Kindergarten is very sad. Why does religion have to be >> part of the curriculum in a public school? Separation of Church and State >> would go a long way toward integration and ultimately peace. >> >> --Aya >> >> >> But, I agree, it is complicated. >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Yes, actually I thought more and there is some truth to what you said. It >> seems >>> that at a psychological level Arabs in other countries don't consider them >> to >>> be 'real Arabs'. A student of mine who went to the Egyptian national museum >> in >>> Cairo saw that there were radically different entry fees posted for 'Arabs' >> and >>> 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in >> fact >>> he was a Druze who had served in the Israeli Army, but they had no way of >>> knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker >> asked >>> to see his passport and when he saw that >>> it was from Israel informed him that he was not an Arab and would have to >> pay >>> the higher 'non-Arab' fee. I have told this story to many of my Israeli >> Arab >>> students and not a single one has even expressed surprise--they completely >>> expect such treatment. When I suggest that this means that they are not >>> regarded as being real Arabs by Arabs from other countries, they admit that >>> this seems to be true--but they still regard themselves as being Arabs. I >> then >>> sum up the situation by saying that the people called 'Israeli Arabs' are >> not >>> considered to be Israelis by Israelis and they are not considered to be >> Arabs >>> by Arabs. >>> It's complicated... >>> >>> With regard to the obstacles to integration, it really is true that neither >> side >>> wants integration if this threatens their distinctive identities. In no >> sense >>> can it be said that the Jewish side rejects assimilation more than the Arab >>> side. I remember one class in which a Christian Arab woman reported >> temporarily >>> sending her child to a Jewish kindergarten because it was the most >> convenient >>> one--but when the child came home one day and insisted on lighting candles >> on >>> Friday evening, the parents realized that this was too much and they needed >>> another solution. Everyone in the class, Jews and Arabs, recognized >>> this--except for an American visiting scholar who said 'What's the problem, >>> it's cute!'. Everyone looked at him like he'd landed from outer space. The >>> difference is more that the Arab side wants EQUALITY, but (contrary to what >>> Americans generally think) that isn't the same thing as integration. >>> >>> John >>> >>> >>> >>> Quoting "A. Katz" : >>> >>>> John, >>>> >>>> Well, at least I'm glad you realize that one hundred years ago, and >>>> possibly even more recently, they did not think of themselves as Arabs. >>>> I still wonder whether among themselves they say "Arabs". Isn't there >>>> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >>>> very much they'd be called "Arabs." For that matter, what do Jordanians >>>> call non-Jewish Israelis? >>>> >>>> The biggest obstacle to allowing for integration in the schools is not the >>>> Palestinian population -- it's those among the Israelis who adopt a "Jews >>>> first" attitude. >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >>>> >>>> >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>> >>>>> Aya, >>>>> I ask them constantly if they think they are Arabs (60% of my students >>>>> are Arabic speakers and this is a central topic of many classes that I >>>> teach) >>>>> and almost without exception they do. The only exceptions are some Druze >>>>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic >>>>> speakers living in this area did not consider themselves to be Arabs >>>> (except >>>>> for the Bedouins), but the situation has completely changed. >>>>> >>>>> It is totally unrealistic to put Israeli Arabs as a group into >>>> Hebrew-speaking >>>>> schools. No one wants it, not the Jews and not the Arabs. There are a >> tiny >>>>> number of Arabs who for one reason or another send their children to >>>>> Hebrew-speaking schools (for example the author Sayed Kashua) but this >>>>> is insignificant. >>>>> >>>>> I don't know what the situation was when you were a child, and I don't >> know >>>>> about the situation among religious Jews, but secular Jewish children do >>>> not >>>>> learn to read by reading Genesis today. They learn to read with secular >>>> texts >>>>> in 1st grade and go on to the Bible in 2nd grade, but far from learning >> to >>>>> read by reading the Bible, teachers have to explain what's written in the >>>>> Bible to the students. NO ONE can understand it without an explanation >> from >>>> the >>>>> teacher or their parents. >>>>> >>>>> Ron--I agree with everything you say except that I believe that the >> radical >>>>> difference between spoken and written Arabic definitely is a serious >>>> problem >>>>> for literacy. I would not be so quick to dismiss the effect of the Arabic >>>>> writing system--we had a conference on this topic in Haifa in May and the >>>>> general consensus of people who sounded like they knew what they were >>>> talking >>>>> about was that it is a significant problem-- but they hadn't done >>>>> cross-linguistic research and this isn't something I know enough about to >>>> have >>>>> strong opinions one way or the other. Also, the usage of written >> colloquial >>>>> Arabic is basically universal among all Israeli Arabs under the age of 25 >>>> who >>>>> have a cellular phone. >>>>> This is definitely not just a privileged section of the population, it's >>>> most >>>>> people of the relevant age group. >>>>> >>>>> John >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Quoting "A. Katz" : >>>>> >>>>>> John, >>>>>> >>>>>> These people you speak of are not Arabs. Some of them are Moslem and >> they >>>>>> read the Quran in the original. Some of them are not Moslem. All of them >>>>>> speak a local dialect of Arabic. Ask them sometimes if they think they >> are >>>>>> Arabs. >>>>>> >>>>>> Trying to turn every dialect into a separate language with a separate >>>>>> writing system is a way to try to disunite people. But a common >> language, >>>>>> however differently it is pronounced, unites disparate people. >> Australians >>>>>> and Cockneys and Indians and Americans speak sometimes mutually >>>>>> unintelligible versions of English. Using the same writing system and >>>>>> the same classic texts unites them. >>>>>> >>>>>> Instead of telling people they should magnify every difference, why not >>>>>> offer to share your language with them? Hebrew could be a uniting factor >>>>>> if spoken in all Israeli schools. >>>>>> >>>>>> --Aya >>>>>> >>>>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >>>>>> >>>>>> >>>>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>>>> >>>>>>> Learning the classical language is like pulling teeth for Arabs kids >>>> also. >>>>>> They >>>>>>> just can't publicly say it because that would make them bad Arabs. It's >>>>>> just >>>>>>> ridiculous. I have an even better plan for the Jewish kids--teach them >> to >>>>>> use >>>>>>> the written version of the spoken language which Arab kids are using >> for >>>>>>> Facebook. That's how kids make friends these days anyway. >>>>>>> Best wishes, >>>>>>> John >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> Quoting Tom Givon : >>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, >> our >>>>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>>>>>> Classical (after 1949 things changed...). That book may still exist, >>>>>>>> you might track it down. It was easy, a cinch really. I concluded it >> was >>>>>>>> really just Hebrew with a few trivial transformations in Phonology & >>>>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I >> sat >>>>>>>> on a few sessions of a faculty study group at UO who were trying to >>>>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. >> Boy, >>>>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>>>>>> well, in 1939 (or was it 1943?). >>>>>>>> Cheers, TG >>>>>>>> >>>>>>>> ============== >>>>>>>> >>>>>>>> >>>>>>>> john at research.haifa.ac.il wrote: >>>>>>>>> Tom, >>>>>>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>>>>>> seriously >>>>>>>>> consider the possibility of educating Israeli Arabs in a written >>>> version >>>>>> of >>>>>>>>> their spoken language, as you suggest, at least through 3rd grade. I >>>> can >>>>>>>> send >>>>>>>>> you some things I've written on this topic, with a lot of >>>>>> cross-linguistic >>>>>>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>>>>>> situation' >>>>>>>>> here, politically active Arabs think that the solution is to pursue >> the >>>>>>>> same >>>>>>>>> self-destructive strategy which they've been following for almost a >>>>>> century >>>>>>>> but >>>>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame >>>> at >>>>>>>> the >>>>>>>>> Arab media because they do not 'provide the linguistic richness of >>>> formal >>>>>>>>> Arabic') while Jews are basically content to let the Arabs stew in >>>> their >>>>>>>> own >>>>>>>>> juices so that they can reap the benefits. There may, however, be >> hope >>>>>> for >>>>>>>> the >>>>>>>>> future in the form of the radically increased usage of written forms >> of >>>>>>>>> colloquial Arabic dialects in electronic media such as Facebook, >> blogs, >>>>>>>> emails, >>>>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>>>>>> seriously >>>>>>>>> threaten the status of classical Arabic in the same way that the >>>>>> invention >>>>>>>> of >>>>>>>>> the printing press overturned the linguistic hierarchy in Western >>>> Europe >>>>>> by >>>>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>>>>>> >>>>>>>>> I realize that Masalha as well as the researchers he referred to >> don't >>>>>>>>> necessarily know what they're talking about regarding the brain. But >>>> the >>>>>>>>> general point is still potentially significant--that the connected >>>> script >>>>>>>> which >>>>>>>>> Arabic uses as well as the multiple forms which many Arabic letters >>>> have >>>>>>>> may be >>>>>>>>> a significant obstacle to literacy, however this may be related to >>>>>>>> processes in >>>>>>>>> the brain. >>>>>>>>> Best wishes, >>>>>>>>> John >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> Quoting Tom Givon : >>>>>>>>> >>>>>>>>> >>>>>>>>>> A GREAT POET CAN STILL GET IT >>>>>> WRONG >>>>>>>>>> >>>>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) >> accidentally >>>>>>>>>> by stumbling a on his truly great quote: >>>>>>>>>> "All fixed identities are imposed from the outside. >>>>>>>>>> Whoever has a clear identity knows it can assume >>>>>>>>>> multiple forms". >>>>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>>>>>> intended--surely got only one third of the story right. At first, he >>>> was >>>>>>>>>> led astray be the academic researchers he cited, who claimed that >> the >>>>>>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>>>>>> lagging R-hemisphere activity, then explained this neurological >> lag >>>>>>>>>> by suggesting that the Arab script requires more contextual >> analysis. >>>>>>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>>>>>> oriented, less automated. If Arab-reading students required more >>>>>>>>>> contextual labor, it should have been registered as a higher >>>>>>>>>> R-hemisphere activity, not lower. >>>>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>>>>>> among many others, have established beyond reasonable doubt >> that >>>>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>>>>>> module on the boundary of the occipital and temporal lobes (just >> past >>>>>>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>>>>>> 'stream' that flows from the back to the front of the L-cortex. >> And >>>>>>>>>> the L-cortex is in general responsible for the more automated--less >>>>>>>>>> context-dependent--processing of language (as well as visual, motor >>>> and >>>>>>>>>> other skills). The visual word-recognition module is, in turn, >>>>>>>>>> recruited from the pre-existing visual object-recognition >>>>>>>>>> ventral-stream module. A considerable amount of life-time practice >>>> and >>>>>>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>>>>>> human brain is not (yet) genetically configured at birth for >>>>>>>>>> visual-word recognition, only for visual-object recognition. A >> similar >>>>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene >> & >>>>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>>>>>> object-location in space. >>>>>>>>>> Mr. Masalha then, on his own, points out to a more >>>>>>>>>> plausible right answer: Arab students, in Israel as well as all over >>>> the >>>>>>>>>> Arab world, are not taught literacy in their native language >>>>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>>>>>> language. The discrepancy would be just as great if Israeli kids >>>> were >>>>>>>>>> taught their Hebrew literacy first in the language of Genesis; or >> if >>>>>>>>>> French students were taught literacy first in the language of La >>>>>>>>>> Chançon de Roland, Guilhome de Machaut, or Chrê tien de Troyes. Or >>>>>>>>>> English-speaking kids in the language of Beowolf. As far as my >> frail >>>>>>>>>> guessing powers go, remedying the situation would be much easier by >>>>>>>>>> combining two well-known verities of second language acquisition: >> (a) >>>>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And >>>> (b), >>>>>>>>>> teach literacy first in the student's spoken native language; only >>>> then >>>>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy >> the >>>>>>>>>> way, was suggested in the late 1930's by no other than L. >>>> Bloomfield, >>>>>>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>>>>>> literacy in their own--equally glorious--native language, just teach >>>>>>>>>> them smart. >>>>>>>>>> For his last culprit, the presumed--tho hardly >>>>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>>>>>> recognized prejudices of the educated classes. While readily >>>> endorsing >>>>>>>>>> his aesthetic sentiments about modern media, I would still like to >>>> point >>>>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>>>>>> spoken native language, and if anything should facilitate the >> easier >>>>>>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>>>>>> ma'-salaam, >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> T. Givón >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>> ------------------------------------------------------------------------ >>>>>>>>> This message was sent using IMP, the Webmail Program of Haifa >>>> University >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >> ------------------------------------------------------------------------ >>>>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >>>>>>> >>>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From Jordan.Zlatev at ling.lu.se Thu Sep 30 22:45:56 2010 From: Jordan.Zlatev at ling.lu.se (Jordan Zlatev) Date: Fri, 1 Oct 2010 00:45:56 +0200 Subject: CfP Evolition at ICLC 11, 2011 Xian Message-ID: Call for Papers for a Theme Session at the ICLC 11, July 11-17, 2011, Xi'an, China Language Evolution: Biological, Cultural and Bio-Cultural Organizers: Arie Verhagen (Leiden University, email Arie.Verhagen at hum.leidenuniv.nl) Jordan Zlatev (Lund University, email Jordan.Zlatev at ling.lu.se) Oct. 10, 2010: Deadline for submitting presentation proposals to the session organizers (Title, and mini-abstract of at most three lines) Nov. 15, 2010: Deadline for submitting full abstracts to ICLC 11 (http://www.iclc11.org) Evolution of language is emerging as a prominent interdisciplinary field of research, bringing together linguists, biologists, psychologists, anthropologists, literary scholars, semioticians and cognitive scientists. A key challenge is to unify theories of language and cognition with the theory of biological evolution: language has evidently evolved in the human lineage, but it has some very special features that make it “one of the most significant and interesting evolutionary events […] during the entire history of life on Earth” (Fitch 2010: 1): it is a “cheap”, honest, flexible, and powerful system of communication (and thought) that no other animal species appears to have and that must have required some special circumstances to have evolved. A central issue of debate is the relation between biological and cultural factors involved in the evolution of Homo Sapiens on the hand, and of modern languages on the other. It is undeniable that mechanisms and principles beyond those involved in biological evolution play a role in cultural evolution, but there is considerable room for variation and even disagreement on what these mechanisms are, how big a role each of them plays, how they interact with each other and with biological factors, and –very fundamentally– to what extent an overall account, incorporating such mechanisms, can still be considered Darwinian. At one end, the view is found that both the biological evolution of the brain (and body) necessary for language, and the historical changes that have given rise to the 6000+ languages spoken today are explainable in terms of standard evolutionary theory, based on the dynamics of variation, selection and replication combined (e.g. Croft 2000). Others hold that even biological evolution, and much more so cultural evolution of language and other institutional practices involves principles such as “autopoiesis” and “qualitative emergence” that intrinsically lead to complexification (e.g. Kull 2009). Whether such processes are fundamentally different from (if somewhat analogous to) Darwinian processes is an additional issue. Yet another one is the extent to which genetic evolution per se, cultural evolution per se and/or the interaction of biological and cultural processes (among which gene-culture co-evolution and niche-construction) account for modern human cognition, language, and culture in general. Different positions and discussion of the nature and consequences of mechanisms and principles being proposed can be found in publications by researchers from different disciplinary backgrounds, e.g. Deacon (1997), Levinson (2000), Laland & Brown (2002), Odling-Smee et al. (2003), Heine and Kuteva (2007), Brier (2008), Sinha (2010). We invite contributions addressing linguistic issues, based on theoretical argumentation and empirical evidence, that contribute to this debate. Some examples (non-exhaustive!) of possible topics include: ‑ Construal: how does the capacity to construe the same object of conceptualization in different ways fit into an evolutionary perspective (biological and/or cultural)? ‑ Grammaticalization and semantic change as cultural evolution. ‑ Grammatical and lexical systems as adaptations, to cultural and/or biological environments. ‑ Co-evolution of cognition (e.g. intersubjectivity, shared intentionality), language, and culture; cultural/linguistic niche-construction. ‑ Co-evolution of communicative and collaborative practices and linguistic systems. - The roles of narrative, writing and symbolic artifacts in the cultural evolution of language. References Brier, Søren (2008). Biosemiotics: Why Information is not Enough. Toronto: University of Toronto Press. Croft, William (2000), Explaining Language Change. An Evolutionary Approach. London, etc.: Longman. Deacon, Terrence W. (1997), The Symbolic Species: The Coevolution of Language and the Brain. New York: W.W. Norton. Fitch, Tecumseh W. (2010), The Evolution of Language. Cambridge: Cambridge University Press. Heine, Bernd and Tania Kuteva (2007), The Genesis of Grammar. A Reconstruction. Oxford: Oxford University Press. Kull, Kalevi, (2009) Vegetative, animal, and propositional semiosis: The semiotic threshold zones. Cognitive Semiotics #4. Laland, Kevin H. & Gillian R. Brown (2002), Sense and Nonsense. Evolutionary Perspectives on Human Behaviour. Oxford: Oxford University Press. Levinson, Stephen C. (2000), Language as nature and language as art. In: R. Hide, J. Mittelstrass, W. Singer (eds.), Changing concepts of nature at the turn of the millennium. Vatican City: Pontifical Academy of Sciences, 257‑287. Odling-Smee, F.J., Laland, K.N. and Feldman, M.W.: 2003.,Niche Construction. The Neglected Process in Evolution. Princeton: Princeton University Press. Sinha, C. (2010). Language as a biocultural niche and social institution. In V. Evans, & S. Pourcel (eds.), New Directions in Cognitive Linguistics. Amsterdam: John Benjamins. From jan.nuyts at ua.ac.be Thu Sep 2 19:49:21 2010 From: jan.nuyts at ua.ac.be (jan nuyts) Date: Thu, 2 Sep 2010 21:49:21 +0200 Subject: Int'l Conference on Grammaticalization and (Inter)Subjectification - Update Message-ID: All information (program, abstracts, registration info, etc.) regarding the International Conference on Grammaticalization and (Inter)Subjectification, to be held in Brussels on November 11-13, 2010, is now available at the conference website at http://webh01.ua.ac.be/gramis/conference/conference.html From tono at ualberta.ca Fri Sep 3 19:02:10 2010 From: tono at ualberta.ca (Tsuyoshi Ono) Date: Fri, 3 Sep 2010 13:02:10 -0600 Subject: Position in Chinese Lingusitics and Pedagogy Message-ID: Assistant Professor - Assistant Professor in Chinese Linguistics and Pedagogy Department of East Asian Studies, University of Alberta Application deadline: 11 October 2010 The Department of East Asian Studies in the Faculty of Arts at the University of Alberta invites applications for a tenure track position at the rank of Assistant Professor in Chinese Linguistics and Pedagogy, which may include any or several of the following areas; Chinese Linguistics, Chinese Applied Linguistics, Chinese Second Language Acquisition, Chinese Language Pedagogy. Applicants should have, or be close to completing, a Ph.D. in an appropriate discipline, and demonstrate outstanding potential for a research career. They should also demonstrate native or near-native fluency in spoken and written Mandarin and English. Responsibilities will include teaching in both undergraduate and graduate student programs, and maintaining an active research program. Experience in teaching Chinese at the college/university level in North America, interest in instructional technology and experience in coordinating a Chinese language program are important assets. The teaching load is four courses per year. Salary is competitive, and will be commensurate with qualifications and experience. Applicants should send a letter of application together with a CV and writing sample, and, if available, a teaching dossier and evaluations of teaching performance to: Professor Yoshi Ono, Chair . Chinese Linguistics Search Committee Department of East Asian Studies 331 Pembina Hall, University of Alberta Edmonton, AB, T6G 2H8, Canada. Applicants must also arrange for three letters of reference to be sent to the Chair. Consideration of applications will begin October 11, 2010 and continue until the position is filled. The effective date of employment will be July 1, 2011. The Department of East Asian Studies (http://www.eastasianstudies.ualberta.ca/) has recently undergone a significant rebuilding process with the addition of several important positions, and efforts are continuing to further expand the Department.s offerings and expertise. The filling of this position is an important part of this process. Inquiries may be directed to Professor Ono at tono at ualberta.ca, or Professor Mikael Adolphson, Acting Chair at madolphs at ualberta.ca. Established in 1908 as a board-governed, public institution, the University of Alberta has earned the reputation of being one of the best universities in Canada based on our strengths in teaching, research, and services. The University of Alberta serves over 36,000 students in more than 200 undergraduate programs and 170 graduate programs (www.ualberta.ca/). The Faculty of Arts is the oldest and most diverse faculty on campus, and one of the largest research and teaching centres in western Canada (www.arts.ualberta.ca). The University.s main campus is located in Edmonton, the vibrant, cosmopolitan capital of the province of Alberta. The Edmonton metropolitan area is the sixth largest in the country with a population of approximately one million (www.edmonton.ca) with excellent health care facilities and recreation opportunities. Edmonton is located only a few hours drive from Banff and Jasper National Parks, which offer skiing in winter and excellent hiking and sightseeing in summer. All qualified candidates are encouraged to apply; however, Canadians and permanent residents will be given priority. The University of Alberta hires on the basis of merit. We are committed to the principle of equity in employment. We welcome diversity and encourage applications from all qualified women and men, including persons with disabilities, members of visible minorities, and Aboriginal persons. From edith at uwm.edu Fri Sep 3 20:26:38 2010 From: edith at uwm.edu (Edith A Moravcsik) Date: Fri, 3 Sep 2010 15:26:38 -0500 Subject: Call for papers re endangered languages Message-ID: ? CALL FOR PAPERS ? LANGUAGE DEATH, ENDANGERMENT, DOCUMENTATION AND REVITALIZATION ? 26th UWM Linguistics Symposium University of Wisconsin-Milwaukee Milwaukee, WI, USA October 20-22, 2011 ? ORGANIZING COMMITTEE Fred Eckman, Elena Mihas, Edith Moravcsik, Sally Noonan, Hamid Ouali, Bernard Perley, Gabriel Rei-Doval , Bozena Tieszen, Kathleen Wheatley ? DESCRIPTION In a globalized world where hundreds of languages are expected to become extinct in the 21st century, it is highly relevant to analyze the viability and continuity of threatened languages. The purpose of this symposium is to discuss this impending loss to humankind from a multidisciplinary perspective. ? We invite contributions for the assessment of this process from Linguistics, Sociology, Psychology, Anthropology, Education, and related fields. Equally welcome is the participation of practitioners in language revitalization efforts. We wish to combine theoretical and practical perspectives for the analysis of the linguistic and social processes involved in language death, endangerment, documentation and revitalization. Possible topics include the following: - The genetic and areal distribution of endangered ? languages - Structural characteristics of endangered ? ? languages - Cultural characteristics of endangered language ? communities - Causes of language endangerment - Documentation of endangered languages - Language revitalization programs and practices - Academic ethics and advocacy in language ? endangerment ? SUBMISSION OF ABSTRACTS (a) ?? Length: ?? ?? ? The abstract may be up to one page of text ? ?? ?? ? plus up to half a page containing possible ??? ?? examples, charts, and references. ? (b) ?? Format: ??? ?? The abstract should include the title of the ? ??? ?? paper and the text of the abstract but not ? ????? the author?s name or affiliation. The e-mail ? ???? message to which it is attached should list ????? the title, the author?s name, and the ????? author?s affiliation. Abstracts will be ? ????? evaluated anonymously. ? Please send the message to reidoval at uwm.edu ? SUBMISSION DEADLINE : FEBRUARY 1 st , 2011 Authors will be notified on their acceptance status by April 30 th , 2011. ? CONFERENCE WEBSITE: https://www4.uwm.edu/letsci/conferences/linguistics2011 or search for UWM Linguistics Symposium From jdavis at ccny.cuny.edu Fri Sep 3 21:30:16 2010 From: jdavis at ccny.cuny.edu (Joseph Davis) Date: Fri, 3 Sep 2010 17:30:16 -0400 Subject: Chi-square Message-ID: A colleague sent to me the June 28 posting by Yuri Tambovtsev below, to which I offer a belated reply that may be useful. The Use of Chi-square by Yuri Tambovtsev Adam Kilgariff wrote that it is not possible to use Chi-square in corpus linguistics. I do not think it is true. One can use Chi-square in linguistics in all cases under the condition that one keeps to the principle of commensurability. That is here, if two samples are equal. I have counted the occurrence of labial consonants in the equal samples of 10000 speech sounds of different Estonian and Russian authors. For instance, in the text of the Estonian writer Aarne Biin ?Moetleja? and Enn Vetemaa ?Neitsist Suendinud? labial consonants occur 896 and 962 times. Could we say that statistically it is the same? So, we put forward the null hypothesis under the 5% level of significance and one degree of freedom. The theoretical threshold value for Chi-square is 3.841. The actual Chi-square value should be less than 3.841 to state that the occurrence of labial consonant in these two samples is the same. We calculated the Chi-square between 896 and 962. It is 2.344. Thus, it is less than 3.841. So, the two text samples enter the same general sample or in other words it is statistically the same. I wonder if my reasoning is correct. [End of quotation from June 28 posting by Yuri Tambovtsev] The main requirement for the use of the chi-square test of significance is that the observations (data points, tokens) in the sample of some population be statistically independent. That is, there should be no statistical relation between one observation and another in the data set. It should not be possible, given the occurrence of one observation, to predict the next observation or any other observation. In my experience, such independence among observations typically is not a property of connected discourse. Rather, the occurrence of one observation typically raises the probability of the same type of observation occurring next or later in the discourse, no doubt because connected discourse is typically coherent, not random. For instance, if a text in English concerns largely the topic of ?peace,? then there will likely be many instances in the text of the labial [p], due to the frequency of the word ?peace? and related words (?peaceful,? ?pacify,? ?peacenik,? etc.). By contrast, if another text is about ?health,? then it will have a disproportionately high frequency of [h], relative to [p]. Consequently, given any occurrence of a labial in the first text, there will be a somewhat elevated probability for occurrence of a labial next or soon; versus the possibility of predicting another [h] in the second text. This is statistical dependence, not independence. As a result, chi-square is not appropriate as a test of significance; it will likely give an inaccurate measure of the degree to which the sample of labials is representative of the larger population of discourse from which the sample was drawn. (In this case, I suppose we can only imagine a hypothetical population of ?English? discourse from which our text was in some idealistic sense ?drawn??-another reason the use of a statistical test of significance may be inappropriate: a text is not in any real sense a sample from a population.) I have a chapter from several years ago that addresses this problem in relation to somewhat different analytical concerns. The reference is: Joseph Davis, 2002, ?Rethinking the place of statistics in Columbia School analysis,? in Wallis Reid, Ricardo Otheguy, and Nancy Stern (eds.), Signal, meaning, and message: Perspectives on sign-based linguistics (pp. 65-90). Amsterdam/Philadelphia: John Benjamins. Joseph Davis, Ph.D. Associate Professor School of Education, NAC 6207 The City College New York, NY 10031 From tgivon at uoregon.edu Fri Sep 3 21:59:19 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Fri, 3 Sep 2010 15:59:19 -0600 Subject: Alexandre Kimenyi, RIP In-Reply-To: Message-ID: ALEXANDRE KIMENYI It is with sad heart that I pass on, belatedly, the news of the sad departure of Prof. Alexandre Kimenyi. Out in the boonies, news travel slow, so this is 2 months after the sad facts but it is still a shock. Alexandre was a gentle soul who lighted my early years in linguistics, and his untimely passing brings back faded old memories. I have last seen him ca. 10 years ago when he came to an Africanist conference at Oregon & hung out with one of my last students there, Boniface Kawasha. Otherwise, I had lost touch with Alexandre almost the minute he finished his dissertation. So this takes me back to ancient days at UCLA, when we were all just beginning to learn what the heck we were doing. For me, Alexandre was part of that very slow learning process. Tho I was technically his dissertation adviser, Ed Keenan was the real inspiration for that marvelous work. In fact, it took me a few years before I realized why and how that work was so important. In those days, I was still experimenting (another detour...) with descriptions that had zero formal components. Only after moving to Oregon (1981) and supervising Noel Rude's work on Nez Perce did I realize how important Alexandre's work had been and how central GR's were to the whole mechanism of grammar. Larry Hyman may recall my less-than-clement comments on a beautiful paper he and Annie Hawkinson published in SAL. True, I did accept it without modifications, but under loud protest. I thought it was 'too formal'. I guess I have always been a slow learner. It is always a sad occasion to see a person much younger than oneself depart early, prematurely (well, is there really a timely departure?). But from the obit Larry Hyman furnished, it appears Alexandre has left behind a large family and many friends, students and colleagues who will surely cherish the memory of this gentle, caring, beautiful man. The Ruanda Genocide was, of course a big part of the story. But Alexandre had great difficulties with visas, permits and just plain survival long before that officially-designated horror. I watched him scurry around helplessly in those pre-genocide days, trying to save as many of his numerous relations. His (and our) utter helplessness about Ruanda, and indeed Africa and what still goes on there, is one of the reason I have never gone back to that beautiful horror-plagued continent. Hard to keep watching. But Alexandre didn't have a choice, so he struggled to do as much as could be done from afar. For those of you who work with native peoples of this continent, it is not that difficult to just close your eyes and almost see how just as much of this horror was happening right here a century and a half ago. I live among some of the survivors, indeed on their land. They are thriving by all official measures, but the scars are still there. Will we ever learn? May Alexandre's gentle soul rest in peace. TG ============================================ From yutamb at mail.ru Sat Sep 4 19:04:28 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Sun, 5 Sep 2010 02:04:28 +0700 Subject: Phono-typological distances received by Tambovtsev Message-ID: Dear Funknet colleagues, We have computed the phono-typological distances between Old Russian and the other Slavonic and Baltic languages. The ordered series of the TMB coefficient is as follows: Old Russian - Lithuanian (2.84); Old Russian - Latvian (3,65); Old Russian - Russian (4,71); Old Russian - Ukrainian (5,20); Belorussian - 6,42; Slovenian - 8,60; Czech - 10,29; Bulgarian - 11,08; Macedonian - 13,92; Slovak - 14,20; Serbo-Croatian - 15,31; Sorbian (Luzhits) - 20,16; Polish - 30,54. Therefore, the distribution of the elements of the speech sound chains makes us conclude that Baltic (Lithuanian and Latvian) sound more similarly to the Old Russian language, but not to Russian, Belorussian or Ukrainian as one may expect. Can you share your opinion and write me to yutamb at mail.ru Yuri Tambovtsev, Novosibirsk Ped.University, Russia. From sepkit at utu.fi Mon Sep 6 04:57:16 2010 From: sepkit at utu.fi (=?iso-8859-1?B?IlNlcHBvIEtpdHRpbOQi?=) Date: Mon, 6 Sep 2010 07:57:16 +0300 Subject: FYI: Workshop on role complexes (Zurich, Switzerland) Message-ID: (apologies for multiple postings) Role complexes: (new) approaches to defining semantic roles Since one of the important functions of many communicative acts is to make clear ?who is doing what to whom? when portraying a particular state of affairs, it is hardly surprising that semantic roles, thematic roles or thematic relations constitute the topic of countless studies in linguistics and are also always discussed, either explicitly or implicitly, in reference grammars. Numerous studies have dealt with agents and patients within and across languages since the 1970s, and there are several comparatively recent studies that address other roles as well (e.g., Stolz et al 2006 for comitatives, N?ss 2007 for transitivity in general, Kittil? 2008 for recipients and goals, Z??iga & Kittil? 2010 for benefactives among numerous others). Among the many interesting characteristics of accounts of semantic roles, it is noteworthy that semantic relationships between predicates and their arguments are treated in different ways. On the one hand, the explicit formal distinction made in natural languages between agents and patients is typically reflected in their analytical status: the volitionally acting instigator of an event (agent) and the inactive, thoroughly affected target of the event (patient) are invariably regarded as two separate roles. On the other hand, different kinds of beneficiaries (e.g., the first person in John tossed me a salad and John mowed the lawn instead of me) are usually considered instances of one and the same role despite their different meanings. Similarly, different subtypes of agents have tended to be treated as different roles while different kinds of experiencers have not. Against this background, the goal of this workshop is to explore approaches to the notion of semantic role in terms of ROLE COMPLEXES, i.e., of clusters of several related sub-roles that might be distinguished by some constructions in certain languages but are otherwise subsumed under a general umbrella notion. For example, different instances of goals differ according to the exact nature of motion (e.g., he threw the ball to the box / behind the box / on the box). The basic definition of the goal role remains unchanged: we are dealing with an endpoint of motion in all cases. Nevertheless, the potential differences between the roles are thus determined by features not typically considered in studies of semantic roles; features usually used for distinguishing between semantic roles, such as instigation, volitionality and affectedness (cf. e.g.. N?ss 2007), can explain neither the semantic nor the formal differences between these three subtypes of goals, or the different codi ng of goals and beneficiaries. We welcome all abstracts dealing with role complexes within and across languages. Possible topics for papers include (but, as always, are not restricted to) the following: - When should we speak of distinct roles, and when are two slightly different (potentially differently coded) roles rather manifestations of one basic role? Are, e.g., inanimate goals and animate goals manifestations of a single role or should they rather be treated separately? - What consequences does role synonymy have for our understanding of semantic roles? What are the features that any adequate theory of semantic roles should consider, what is the ?correct? number of semantic roles, etc.? - How should we treat partial formal mismatches between roles? - How do we deal with semantically/pragmatically determined differences in the coding of roles (e.g., marking conditioned by definiteness, referentiality, specificity, topicality, focality)? - Corpus-based studies of role synonymy: What determines the use of different (yet semantically similar) manifestations of a role in actual language use? - How do we best treat the diachronic development of multifunctional coding devices (syncretisms, polysemies, homonymies, etc.)? - Formal manifestation of semantic role synonymy: case marking, verbal marking, lexical differences, etc. - Role synonymy of core and peripheral roles: Are there any differences, is synonymy more common for one of these? Organizers of the workshop Fernando Z??iga (Zurich) and Seppo Kittil? (Helsinki) Venue University of Zurich, Zurich, Switzerland Dates April 4-5, 2011 Abstract submission Please send your (maximally) 500-word abstract (excluding data and references) to both fernando.zuniga at spw.uzh.ch and kittila at mappi.helsinki.fi no later than November 14, 2010. The letters of acceptance will be sent by December 12, 2010. Abstracts must be anonymous, but the body of the e-mail should include the following information: Name of the author(s) Title of the paper Affiliation(s) E-mail In case you have any questions about the workshop, please don?t hesitate to contact us. We are looking forward to welcoming you all to Zurich. Fernando and Seppo From n.m.stukker at hum.leidenuniv.nl Tue Sep 7 11:47:36 2010 From: n.m.stukker at hum.leidenuniv.nl (Stukker, N.M.) Date: Tue, 7 Sep 2010 13:47:36 +0200 Subject: Call for papers: Stylistics across disciplines Message-ID: Stylistics across disciplines University of Leiden, The Netherlands June 16-17, 2011 Confirmed keynote speakers: Prof. dr. Barbara Dancygier, University of British Columbia, Vancouver (Canada) Prof. dr. Arie Verhagen, Leiden University (The Netherlands) Stylistics is a field of study that is growing and developing fast. Its central concern is the way cognitive and communicative effects are achieved by means of linguistic choices. It therefore encompasses literary studies and linguistics as well as discourse studies. In spite of the shared, overarching definition of what it is, the field of study of Stylistics is highly fragmented. It mainly takes place within the boundaries of the various, more traditional, domains of study, e.g. literary analysis, rhetoric, (critical) discourse analysis, applied linguistics, etc. As a result, a comprehensive understanding of the wide variety of interests and foci of attention in stylistic studies, as well as exchange of knowledge between these research domains, is developing relatively slowly. In recent years, successful attempts have been made to take an integrative, cross-disciplinary perspective on Stylistics, focusing on the shared research object: language use. An example is the expanding body of studies associated with the International Poetics and Linguistics Association (PALA). Especially fruitful has proven to be the developing area of 'cognitive poetics', a field closely allied with the theoretical framework of cognitive linguistics, which includes attention for contextual factors and the inherently 'subjective' basis of language in linguistic analysis. This Stylistics across disciplines conference links up with these developments and intends to offer a platform for exchange of ideas and to stimulate fruitful collaboration among linguists, literary scholars and discourse scholars studying 'style'. We invite participants from all relevant fields to participate in the Stylistics across disciplines conference to discuss the opportunities and problems regarding the development of stylistics as a coherent and methodologically sound research discipline. We welcome papers on (but not limited to) the subject of: * Possibilities and limitations of an interdisciplinary perspective: what can literary scholars learn from the way style is studied in linguistics or rhetoric, and vice versa? * Opportunities and problems of a 'linguistic stylistics' * Methodological issues: qualitative (interpretive analysis) or quantitative methods (digital humanities, corpus stylistics) and different research methods (corpus analysis, experimental effect studies) in relation to various research contexts * Development of theoretical notions and analytical tools especially suited for stylistic analysis * Context-sensitivity of stylistic patterns and analysis: how does stylistic choice interact with contextual factors such as institution, genre characteristics, etc.? * Language specificity and culture specificity of stylistic phenomena and analysis Please submit your abstract (in Word or PDF format, containing the title of your paper, author's name(s) and affiliation(s), max. 500 words) to stylistics at hum.leidenuniv.nl. The deadline for abstract submission is December 15, 2010. Notification of acceptance will be by February 1, 2011. Organizing committee: Suzanne Fagel Maarten van Leeuwen Ninke Stukker stylistics at hum.leidenuniv.nl Scientific committee: Jaap Goedegebuure (literary studies) Ton van Haaften (language and communication) Jaap de Jong (rhetoric) Arie Verhagen (linguistics) The Stylistics across disciplines conference is organized by researchers from the NWO research project Stylistics of Dutch (Leiden University 2007-2012); website: www.stylistics.leidenuniv.nl From Freek.VanDeVelde at arts.kuleuven.be Tue Sep 7 13:17:26 2010 From: Freek.VanDeVelde at arts.kuleuven.be (Freek Van de Velde) Date: Tue, 7 Sep 2010 15:17:26 +0200 Subject: workshop proposal 'exaptation' at ICHL 20 Message-ID: Preliminary call for papers for a workshop at ICHL 20 (Osaka, 25-30 July 2011) Subject: Exaptation Convenors: Muriel Norde (University of Groningen) & Freek Van de Velde (University of Leuven) Contact: freek.vandevelde at arts.kuleuven.be Deadline: 13 september 2010 (no specific title or abstract needed at this point. Just let us know whether you are interested in participating). Although some morphological changes seem to obey general tendencies, as formulated for instance by Kury?owicz or Ma?czak (see Hock 1986, ch.10) or Van Loon (2005), most of these tendencies can just as easily be violated. Diachronic morphology is largely idiosyncratic (Joseph 1998). Morphological paradigms appear to be ripped up at random in order to establish "local generalizations" (Joseph 1992). One particular way in which unpredictable changes come about is through 'bricolage' with junk morphology, which goes under the name of exaptation (Lass 1990, 1997: 316ff.). Exaptation is a concept that was first used in evolutionary biology, to refer to co-optation of a certain trait for a new function. A typical example is the use of feathers, originally serving a thermo-regulatory function, for flight. In linguistics, exaptation is defined as follows: "Say a language has a grammatical distinction of some sort, coded by means of morphology. Then say this distinction is jettisoned, PRIOR TO the loss of the morphological material that codes it. This morphology is now, functionally speaking, junk; and there are three things that can in principle be done with it: (i) it can be dumped entirely; (ii) it can be kept as marginal garbage or nonfunctional/nonexpressive residue (suppletion, 'irregularity'); (iii) it can be kept, but instead of being relegated as in (ii), it can be used for something else, perhaps just as systematic. (...) Option (iii) is linguistic exaptation." (Lass 1990: 81-82) Lass originally understood exaptation in a rather narrow sense. First, the term exaptation was reserved for changes affecting functionless (or 'junk') morphology. Second, in order to qualify as exaptation, the new function of a morpheme needed to be entirely novel. In Lass's own words: "Exaptation then is the opportunistic co-optation of a feature whose origin is unrelated or only marginally related to its later use. In other words (loosely) a 'conceptual novelty' or 'invention'." Both criteria have been criticized. With regard to the first criterion, Vincent (1995: 435), Giacalone Ramat (1998), Smith (2006) and Willis (ms.) pointed out difficulties with regard to the notion of junk. And indeed, Lass later stretched his notion of exaptation, admitting that linguistic exaptation - just like biological exaptation - could also affect non-junk morphology (see Lass 1997: 318), to the effect that the old and the new function may co-exist. Doubt has also been raised with regard to the second criterion, the novelty of the new function, which is central to the notion of exaptation according to Lass (1990: 82) (see also Norde 2001: 244, 2009: 117 and Traugott 2004). Some scholars have argued against the purported novelty of the function after exaptation (Vincent 1995: 436; Giacalone Ramat 1998, Hopper & Traugott 2003: 135-136). If this criterion is jettisoned, we arrive at a fairly broad definition of exaptation, like for instance in Booij (2010: 211), who defines it as "[t]he re-use of morphological markers". Such a broad conception of exaptation is in line with the notion in evolutionary biology, where neither of the two criteria is decisive for the application of the term to shifts in function, but the question then arises whether this does not make the concept vacuous (see De Cuypere 2005). Despite these criticisms, exaptation has been used as a convenient label for morphological changes that at first sight seem to proceed unpredictably, e.g. by running counter to grammaticalization clines (see Norde 2009: 115-118). It has been applied to various cases of morphological change, discussed in Lass (1990), Norde (2002), Fudeman (2004), Van de Velde (2005, 2006), Narrog (2007), Booij (2010, ms.), Willis (ms.) among others. In this workshop, we aim to come to terms with exaptation. Apart from specific case studies drawing on original data, we welcome papers that address the following issues: (1) Do we need exaptation in diachronic morphology, or does it reduce to more traditional mechanisms such as reanalysis and analogy, as e.g. De Cuypere (2005) argues? (2) Does exaptation only apply to morphology (Heine 2003: 173), or is it relevant to syntactic change as well, as Brinton & Stein (1995) have argued? (3) Does exaptation presuppose irregularity and unpredictability? If so, does this entail that exaptation is language-specific (as argued by Heine 2003: 173), and that cross-linguistics generalizations are not possible? See, however, Narrog (2007) for evidence to the contrary. (4) Does exaptation happen primarily in cases of 'system disruption', such as typological word order change or deflection (see Norde 2002: 49, 60, 61)? (5) How should we define the concept of 'novelty', and is it a useful criterion for a change to be qualified as exaptation? Currently, there seem to be different views in the literature on what is exactly understood by a 'new' function. Does this mean (a) an entirely new category in the grammar, (b) a function unrelated to the morpheme's old function, or (c) a different though perhaps not totally unrelated function from the old function? (6) Is exaptation infrequent (Heine 2003:174, Traugott 2004) and non-recurrent (as argued by Heine 2003: 172)? Or can one morpheme undergo several successive stages of exaptation (as argued by Giacalone Ramat 1998: 110-111 with regard to the -sk- suffix and by Van de Velde 2006 with regard to the Germanic adjective inflection)? (7) What is the relation between exaptation and grammaticalization? Do they refer to fundamentally different kinds of changes (Vincent 1995), is exaptation a final stage of grammaticalization (Greenberg 1991, Traugott 2004), or are exaptation and grammaticalization just two different labels for the same type of change? After all, both processes involve reanalysis (Narrog 2007), both processes can come about through pragmatic strengthening (see Croft 2000: 126-130). Furthermore, if the old and new function of the exaptatum co-exist (see above) and if the new function is related to the old one, then exaptation involves 'layering' and 'persistence', respectively (see Van de Velde 2006: 61-62), which are also key features of grammaticalization (see Hopper 1991). (8) What is the relation between exaptation and degrammaticalization? Does exaptation always entail some sort of 'degrammaticalization' (as argued by Heine 2003 and arguably Narrog 2007: 9, 18), or does exaptation often, but not always, go together with degrammaticalization (Norde 2009: 118)? (9) Is exaptation the same thing as what Greenberg (1991) understands by 'regrammaticalization' and as what Croft (2000) understands by 'hypoanalysis', or are there significant differences between these concepts? And what is the overlap with related concept such as 'functional renewal' (Brinton & Stein 1995)? References Booij, G. 2010 (to appear). Construction morphology. Oxford: Oxford University Press. Booij, G. manuscript. Recycling morphology: Case endings as markers of Dutch constructions. . Brinton, L. & D. Stein. 1995. Functional renewal. In: H. Andersen (ed.), Historical Linguistics 1993. Amsterdam: Benjamins. 33-47. Croft, W. 2000. Explaining language change. An evolutionary approach. Harlow : Longman. De Cuypere, L. 2005. Exploring exaptation in language change. Folia Linguistica Historica 26: 13-26. Fudeman, K. 2004. Adjectival agreement vs. adverbal inflection in Balanta. Lingua 114: 105-23. Giacalone Ramat, A. 1998. Testing the boundaries of grammaticalization. In: A. Giacalone Ramat & P.J. Hopper (eds.), The limits of grammaticalization. Amsterdam: Benjamins. 227-270. Greenberg, J.H. 1991. The last stages of grammatical elements: Contractive and expansive desemanticization. In: E.C. Traugott & B. Heine (eds.), Approaches to grammaticalization. Amsterdam: Benjamins. 301-314. Heine, B. 2003. On degrammaticalization. In: B.J. Blake & K. Burridge (eds.), Historical linguistics 2001. Amsterdam: Benjamins. 163-179. Hock, H.H. 1986. Principles of historical linguistics. Berlin: de Gruyter. Hopper, P.J. 1991. On some principles of grammaticalization. In: E.C. Traugott & B. Heine (eds.), Approaches to grammaticalization. Amsterdam: Benjamins. 17-35. Hopper, P.J. & E.C. Traugott. 2003. Grammaticalization. 2nd edn. Cambridge: Cambridge University Press. Joseph, B.D. 1992. Diachronic explanation: Putting the speaker back into the picture. In: G.W. Davis & G.K. Iverson (eds.), Explanations in historical linguistics. Amsterdam: Benjamins. 123-144. Joseph, B.D. 1998. Diachronic morphology. In: A. Spencer & A.M. Zwicky (eds.), Handbook of morphology. Oxford: Blackell. 351-373. Lass, R. 1990. How to do things with junk: Exaptation in language evolution. Journal of Linguistics 26: 79-102. Lass, R. 1997. Historical linguistics and language change. Cambridge: Cambridge University Press. Narrog, H. 2007. Exaptation, grammaticalization, and reanalysis. California Linguistic Notes 32 (1). . Norde, M. 2001. Deflexion as a counterdirectional factor in grammatical change. Language Sciences 23: 231-264. Norde, M. 2002. The final stages of grammaticalization: Affixhood and beyond. In: I. Wischer & G. Diewald (eds.), New reflections on grammaticalization. Amsterdam: Benjamins. 45-81. Norde, M. 2009. Degrammaticalization. Oxford: Oxford University Press. Smith, J.C. 2006. How to do things without junk: the refunctionalization of a pronominal subsystem between Latin and Romance. In: J.-P.Y. Montreuil (ed.), New perspectives on Romance linguistics. Volume II: Phonetics, phonology and dialectology. Amsterdam: Benjamins. 183-205. Traugott, E.C. 2004. Exaptation and grammaticalization. In: M. Akimoto (ed.), Linguistic studies based on corpora. Tokyo: Hituzi Syobo. 133-156. Van de Velde, F. 2005. Exaptatie en subjectificatie in de Nederlandse adverbiale morfologie [Exaptation and subjectification in Dutch adverbial morphology]. Handelingen der Koninklijke Zuid-Nederlandse Maatschappij voor Taal- en Letterkunde en Geschiedenis 58: 105-124. Van de Velde, F. 2006. Herhaalde exaptatie. Een diachrone analyse van de Germaanse adjectiefflexie [Iterative exaptation. A diachronic analysis of the Germanic adjectival inflection]. In: M. H?ning, A. Verhagen, U. Vogl & T. van der Wouden (eds.), Nederlands tussen Duits en Engels. Leiden: Stichting Neerlandistiek Leiden. 47-69. Van Loon, J. 2005. Principles of historical morphology. Heidelberg: Universit?tsverlag Winter. Vincent, N. 1995. Exaptation and grammaticalization. In: H. Andersen (ed.), Historical linguistics 1993. Amsterdam: Benjamins. 433-445. Willis, D. Manuscript. Degrammaticalization and obsolescent morphology: Evidence from Slavonic. < http://people.pwf.cam.ac.uk/dwew2/willis_degramm_berlin.pdf>. Freek Van de Velde Postdoctoral research fellow Research Foundation Flanders (FWO), University of Leuven Fac. of Arts, Dept. of Linguistics Blijde Inkomststraat 21, P.O. Box 3308 BE-3000 Leuven Tel. 0032 16 32 47 81 Fax 0032 16 32 47 67 http://wwwling.arts.kuleuven.be/nedling/fvandevelde/ From bischoff.st at gmail.com Wed Sep 8 12:46:55 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Wed, 8 Sep 2010 08:46:55 -0400 Subject: analysis: unhappiness Message-ID: Hi all, I had an interesting exchange with a few generative syntacticians/morphologists (former classmates of mine) regarding an analysis of "unhappiness". Two things that they said surprised me a bit, they are the following: (1) un- (negation, 'not') only attaches to adjectives (now this clearly isn't the case, a simple cursory view of the etymology in the OED provides a number of examples of un- with nouns and verbs...though to significantly lesser degrees...in addition works on English morphology contain examples as well) (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis such as [un-[happy-ness]] is impossible (due to (1) above according to my former colleagues). My questions are the following: (1) Is there a good/well grounded reason to believe un- "only" attaches to adjectives? (2) What would be the consensus on an analysis of "unhappiness" that most linguists would agree on? Thanks, Shannon From amnfn at well.com Wed Sep 8 13:05:11 2010 From: amnfn at well.com (A. Katz) Date: Wed, 8 Sep 2010 06:05:11 -0700 Subject: analysis: unhappiness In-Reply-To: Message-ID: Many of the functionalist linguists I know take the opposite view that derviational morphology is not really active in English, and that people just memorize the lexemes "unhappy", "unhappiness", and "happy"/"happiness" separately. The idea is that while speakers recognize that these words are related, they do not rederive them every time, and they learned them from others, rather than putting them together by themselves. I personally think derivational morphology is a little more active than that, but there you have your two theoretical extremes, the one you related about your former colleagues and the one I related about my former colleagues. Best, --Aya On Wed, 8 Sep 2010, s.t. bischoff wrote: > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon > > From dan at daneverett.org Wed Sep 8 13:07:26 2010 From: dan at daneverett.org (Daniel Everett) Date: Wed, 8 Sep 2010 09:07:26 -0400 Subject: analysis: unhappiness Message-ID: Shannon, There is a long literature on this under the heading of 'bracketing paradoxes'. One of the best articles came out in the early 90s in Language, by Andy Spencer. Though just about everyone and their dog was writing on it back then. The generative analysis is what leads to the 'paradoxes', which are either a discovery or an error depending on your perspective. -- Dan On Sep 8, 2010, at 8:46 AM, s.t. bischoff wrote: > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon From amber at cs.toronto.edu Wed Sep 8 13:38:09 2010 From: amber at cs.toronto.edu (L. Amber Wilcox-O'Hearn) Date: Wed, 8 Sep 2010 09:38:09 -0400 Subject: analysis: unhappiness Message-ID: On Wed, Sep 8, 2010 at 6:46 AM, s.t. bischoff wrote: > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) Maybe what they are saying is not that there is no un- attaching to verbs, there obviously are many examples. But in those cases, un- is not negation, it is a reverse action - 'untie' does not mean 'not tie'. \ L. Amber Wilcox-O'Hearn * http://www.cs.toronto.edu/~amber/ / -\ Graduate student * Computational Linguistics Research Group /- --\ Department of Computer Science * University of Toronto /-- From jrubba at calpoly.edu Wed Sep 8 15:26:29 2010 From: jrubba at calpoly.edu (Johanna Rubba) Date: Wed, 8 Sep 2010 08:26:29 -0700 Subject: analysis: unhappiness In-Reply-To: <20100908133810.00EE92B01BD@apps0.cs.toronto.edu> Message-ID: One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. Dr. Johanna Rubba, Ph. D. Professor, Linguistics Linguistics Minor Advisor English Dept. Cal Poly State University San Luis Obispo San Luis Obispo, CA 93407 Ofc. tel. : 805-756-2184 Dept. tel.: 805-756-2596 Dept. fax: 805-756-6374 E-mail: jrubba at calpoly.edu URL: http://cla.calpoly.edu/~jrubba From dubartell at edinboro.edu Wed Sep 8 16:41:52 2010 From: dubartell at edinboro.edu (DUBARTELL) Date: Wed, 8 Sep 2010 12:41:52 -0400 Subject: cc: Re: analysis: unhappiness Message-ID: Hello All, I think this issue is not perhaps one of analysis or intuition, but one of K-12 teaching/learning. Many teachers provide exercises over the school years that reinforce prefixation with less lesson time given to the suffixes. My students can easily explain the meanings of pre-, anti-, and ex-, but are often unable to explain the meaning of -ness, -tion-, and -ence, for example, as readily. I know from teaching student teachers that many of their grammar lesson plans focus primarily on prefixes. (In addition, some students do not realize that suffixes appear in dictionaries). I must also wonder if the stress difficulties are related to the possible infrequency of such lessons in the K-12 curricula. Very generally, many college students are better at those grammatical exercises with which they may have at least some familiarity from their K-12 experience. Regards, Deborah DuBartell On Wednesday, September 08, 2010 11:26 AM, Johanna Rubba wrote: > >Date: Wed, 8 Sep 2010 08:26:29 -0700 >From: Johanna Rubba >To: funknet at mailman.rice.edu >cc: >Subject: Re: [FUNKNET] analysis: unhappiness > >One thing that consistently occurs in my intro linguistics classes is >that at least half of my students do not analyze complex words the >way a linguist would -- many would analyze "unhappiness" as "un" + >"happiness." They make such analyses over and over. It makes one >wonder, of course, about how much native-speaker intuition is in >agreement with some linguistic analyses. I can *feel* that the >analysis is [[un-happy]-ness], but, apparently, large numbers of >native speakers cannot. > >Another thing I often find is that many students cannot locate either >primary or (especially) secondary stress in words. This is very >bizarre, considering that they produce the stresses correctly and >hear them correctly in others' speech. So many are unsuccessful at >this that I have stopped requiring them to find stress in words on >tests. I give them tricks like singing the word and monitoring for >the highest-pitched syllable, but the tricks don't work. That many >students can't be tone-deaf. > >Dr. Johanna Rubba, Ph. D. >Professor, Linguistics >Linguistics Minor Advisor >English Dept. >Cal Poly State University San Luis Obispo >San Luis Obispo, CA 93407 >Ofc. Tel. : 805-756-2184 >Dept. Tel.: 805-756-2596 >Dept. fax: 805-756-6374 >E-mail: jrubba at calpoly.edu >URL: http://cla.calpoly.edu/~jrubba > > > Deborah DuBartell, Ph.D. Professor of Linguistics Department of English 295 Meadville St. Centennial Hall 238 Edinboro University of Pennsylvania Edinboro, PA 16444 tele: 814.732.2269 web: users.edinboro.edu/DuBartell From djh514 at york.ac.uk Wed Sep 8 17:18:49 2010 From: djh514 at york.ac.uk (Damien Hall) Date: Wed, 8 Sep 2010 18:18:49 +0100 Subject: Audit of School-Leavers' Grammatical Knowledge In-Reply-To: Message-ID: Inspired by the 'unhappiness' thread and the subsequent reflections on what grammar / grammatical terms students do and do not know, I'm reposting here a link to something that I saw on the LINGUIST List this morning, in case others did not see it. Dick Hudson would appreciate it if colleagues not in the UK would administer an apparently simple questionnaire to their beginning-level Linguistics classes, to gauge the amount of grammatical terminology they know when they enter University (ie after their high-school education). Further details are on LINGUIST, here: http://linguistlist.org/pubs/sums/query-details.cfm?submissionid=2646814 Damien -- Damien Hall University of York Department of Language and Linguistic Science Heslington YORK YO10 5DD UK Tel. (office) +44 (0)1904 432665 (mobile) +44 (0)771 853 5634 Fax +44 (0)1904 432673 http://www.york.ac.uk/res/aiseb http://www.york.ac.uk/depts/lang/people/pages/hall.htm DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm From lesleyne at msu.edu Wed Sep 8 19:02:50 2010 From: lesleyne at msu.edu (Diane Frances Lesley-Neuman) Date: Wed, 8 Sep 2010 15:02:50 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: ?You are getting into the Lexical Phonology of English debate over strata.? To begin with consult: April Mcmahon (1999) Lexical Strata and the History of English Hans Giegerich? Lexical Strata in English The old literature was debated throughout the 1980's: Paul Kiparsky, K.P. Mohanan (with Morris Halle), Jerzy Rubach and Geert Booij come to mind.? Toni Borowsky as well, although her analysis has come under question.? i also think that Ricardo Bermudez-Otero, an English lexical?phonology and stratal?OT?specialist at Manchester?somewhere has a critique of the various models or can guide you further.? In general this is the object of study of the English phonology faculty at Edinburgh where Mcmahon and Giegerich are?and I think an another specialist?person to weigh in on the issue would be Patrick Honeybone,?a specialist in the diachronic phonology of English also Edinburgh-based. ? ______________________________ Diane Lesley-Neuman Linguistics Program Wells A-614 Michigan State University East Lansing, MI 48824 Quoting "s.t. bischoff" : > Hi all, > > I had an interesting exchange with a few generative > syntacticians/morphologists (former classmates of mine) regarding an > analysis of "unhappiness". Two things that they said surprised me a bit, > they are the? following: > > (1) un- (negation, 'not') only attaches to adjectives (now this clearly > isn't the case, a simple cursory view of the etymology in the OED provides a > number of examples of un- with nouns and verbs...though to significantly > lesser degrees...in addition works on English morphology contain examples as > well) > > (2) the analysis of unhappiness can only be [[un-happy]-ness]...an analysis > such as [un-[happy-ness]] is impossible (due to (1) above according to my > former colleagues). > > My questions are the following: > > (1) Is there a good/well grounded reason to? believe un- "only" attaches to > adjectives? > > (2) What would be the consensus on an analysis of "unhappiness" that most > linguists would agree on? > > Thanks, > Shannon > From Lise.Menn at Colorado.EDU Thu Sep 9 00:16:15 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Wed, 8 Sep 2010 18:16:15 -0600 Subject: analysis: unhappiness In-Reply-To: Message-ID: So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > One thing that consistently occurs in my intro linguistics classes > is that at least half of my students do not analyze complex words > the way a linguist would -- many would analyze "unhappiness" as "un" > + "happiness." They make such analyses over and over. It makes one > wonder, of course, about how much native-speaker intuition is in > agreement with some linguistic analyses. I can *feel* that the > analysis is [[un-happy]-ness], but, apparently, large numbers of > native speakers cannot. > > Another thing I often find is that many students cannot locate > either primary or (especially) secondary stress in words. This is > very bizarre, considering that they produce the stresses correctly > and hear them correctly in others' speech. So many are unsuccessful > at this that I have stopped requiring them to find stress in words > on tests. I give them tricks like singing the word and monitoring > for the highest-pitched syllable, but the tricks don't work. That > many students can't be tone-deaf. > > Dr. Johanna Rubba, Ph. D. > Professor, Linguistics > Linguistics Minor Advisor > English Dept. > Cal Poly State University San Luis Obispo > San Luis Obispo, CA 93407 > Ofc. tel. : 805-756-2184 > Dept. tel.: 805-756-2596 > Dept. fax: 805-756-6374 > E-mail: jrubba at calpoly.edu > URL: http://cla.calpoly.edu/~jrubba > > > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 http://spot.colorado.edu/~menn/index.html Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Fellow, Linguistic Society of America Campus Mail Address: UCB 594, Institute for Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From tgivon at uoregon.edu Thu Sep 9 01:03:48 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Wed, 8 Sep 2010 19:03:48 -0600 Subject: analysis: unhappiness In-Reply-To: <9DE8281B-3822-43CF-8001-C70B10E98441@colorado.edu> Message-ID: Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG ============== Lise Menn wrote: > So we see an important phenomenon: Tacit knowledge really IS tacit, > and 'intuitions' are very poor guides to what our minds are doing when > we are using the patterns of our language as speakers/hearers. > Introspection cannot replace observation of actual usage and > psycholinguistic experiments; it can only act as a suggestion of where > to dig. After all, we can't figure out vision or digestion by > thinking about how they feel, although we certainly have to account > for subjective feelings of contrast and indigestion. The same is true > for language, mutatis mutandis. > > On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > >> One thing that consistently occurs in my intro linguistics classes is >> that at least half of my students do not analyze complex words the >> way a linguist would -- many would analyze "unhappiness" as "un" + >> "happiness." They make such analyses over and over. It makes one >> wonder, of course, about how much native-speaker intuition is in >> agreement with some linguistic analyses. I can *feel* that the >> analysis is [[un-happy]-ness], but, apparently, large numbers of >> native speakers cannot. >> >> Another thing I often find is that many students cannot locate either >> primary or (especially) secondary stress in words. This is very >> bizarre, considering that they produce the stresses correctly and >> hear them correctly in others' speech. So many are unsuccessful at >> this that I have stopped requiring them to find stress in words on >> tests. I give them tricks like singing the word and monitoring for >> the highest-pitched syllable, but the tricks don't work. That many >> students can't be tone-deaf. >> >> Dr. Johanna Rubba, Ph. D. >> Professor, Linguistics >> Linguistics Minor Advisor >> English Dept. >> Cal Poly State University San Luis Obispo >> San Luis Obispo, CA 93407 >> Ofc. tel. : 805-756-2184 >> Dept. tel.: 805-756-2596 >> Dept. fax: 805-756-6374 >> E-mail: jrubba at calpoly.edu >> URL: http://cla.calpoly.edu/~jrubba >> >> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > http://spot.colorado.edu/~menn/index.html > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > Fellow, Linguistic Society of America > > Campus Mail Address: > UCB 594, Institute for Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From jrubba at calpoly.edu Thu Sep 9 01:20:17 2010 From: jrubba at calpoly.edu (Johanna Rubba) Date: Wed, 8 Sep 2010 18:20:17 -0700 Subject: analysis: unhappiness In-Reply-To: <4C883274.3020503@uoregon.edu> Message-ID: The reason I make much of my students' disagreements with linguists' analyses is that, at least as I recall from my graduate education, native-speaker intuitions were routinely used in defense of analyses of sentence structure. I agree completely with Lise that introspection alone is not sufficient for defending an analysis. I never meant to suggest that my students store complex words disconnected from their component parts. Such a notion would be bizarre for a practitioner of Cognitive Grammar. Plus, the behavior I described doesn't contradict network connections; it challenges the usefulness of NS intuitions, which would, in turn, challenge analyses defended based on same. It's odd that, in some cases, students' tacit knowledge can be brought to consciousness, and in other cases not. It seems clear that context (including priming) affects a language-user's analysis. Certainly, many expressions that sound odd out of context sound perfectly fine in a suitable context. Jo On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG ============== Lise Menn wrote: > So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. > > On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: > >> One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. >> >> Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. >> >> Dr. Johanna Rubba, Ph. D. >> Professor, Linguistics >> Linguistics Minor Advisor >> English Dept. >> Cal Poly State University San Luis Obispo >> San Luis Obispo, CA 93407 >> Ofc. tel. : 805-756-2184 >> Dept. tel.: 805-756-2596 >> Dept. fax: 805-756-6374 >> E-mail: jrubba at calpoly.edu >> URL: http://cla.calpoly.edu/~jrubba >> >> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > http://spot.colorado.edu/~menn/index.html > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > Fellow, Linguistic Society of America > > Campus Mail Address: > UCB 594, Institute for Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > Dr. Johanna Rubba, Professor, Linguistics Linguistics Minor Advisor English Department California Polytechnic State University, San Luis Obispo E-mail: jrubba at calpoly.edu Tel.: 805.756.2184 Dept. Ofc. Tel.: 805.756.2596 Dept. Fax: 805.756.6374 URL: http://www.cla.calpoly.edu/~jrubba From Lise.Menn at Colorado.EDU Thu Sep 9 02:31:31 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Wed, 8 Sep 2010 20:31:31 -0600 Subject: analysis: unhappiness In-Reply-To: <4C883274.3020503@uoregon.edu> Message-ID: if you-all will forgive me for a little advance self-promotion, my new book, Psycholinguistics: Introduction and Applications, will be available from Plural Publishing by the end of the year. It's intended to be completely accessible for anyone (linguist, psychologist, speech-language pathologist, language teacher...) who needs to be able to think about brain and language, and what Tom says about experimental techniques and word storage is one of the things I explore, starting from analyses of slips of the tongue. Also a chapter each on the brain, reading, language development, aphasia, and second language learning. You can see the full table of contents on the web page for the book: http://www.pluralpublishing.com/publication_psl.htm Best regards to all, Lise On Sep 8, 2010, at 7:03 PM, Tom Givon wrote: > > Right on, Lise. And further, there is a well-known experimental > technique called "semantic priming" that is admirably well suited > for investigating whether when a language used hears "unhappiness", > "happy" and "happiness" are activated ('come to mind'). This > technique will probably not answer the question of the differential > bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to > answer questions of directionality (does "unhappy" prime "happy" > stronger than vice versa?). But it does tends to suggest that we > don't store complex words in total disconnect from their parts, at > least not as frequent adult users. And that phonological similarity > (shared parts of words) has semantic consequences. Cheers, TG > > ============== > > > Lise Menn wrote: >> So we see an important phenomenon: Tacit knowledge really IS tacit, >> and 'intuitions' are very poor guides to what our minds are doing >> when we are using the patterns of our language as speakers/ >> hearers. Introspection cannot replace observation of actual usage >> and psycholinguistic experiments; it can only act as a suggestion >> of where to dig. After all, we can't figure out vision or >> digestion by thinking about how they feel, although we certainly >> have to account for subjective feelings of contrast and >> indigestion. The same is true for language, mutatis mutandis. >> >> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >> >>> One thing that consistently occurs in my intro linguistics classes >>> is that at least half of my students do not analyze complex words >>> the way a linguist would -- many would analyze "unhappiness" as >>> "un" + "happiness." They make such analyses over and over. It >>> makes one wonder, of course, about how much native-speaker >>> intuition is in agreement with some linguistic analyses. I can >>> *feel* that the analysis is [[un-happy]-ness], but, apparently, >>> large numbers of native speakers cannot. >>> >>> Another thing I often find is that many students cannot locate >>> either primary or (especially) secondary stress in words. This is >>> very bizarre, considering that they produce the stresses correctly >>> and hear them correctly in others' speech. So many are >>> unsuccessful at this that I have stopped requiring them to find >>> stress in words on tests. I give them tricks like singing the word >>> and monitoring for the highest-pitched syllable, but the tricks >>> don't work. That many students can't be tone-deaf. >>> >>> Dr. Johanna Rubba, Ph. D. >>> Professor, Linguistics >>> Linguistics Minor Advisor >>> English Dept. >>> Cal Poly State University San Luis Obispo >>> San Luis Obispo, CA 93407 >>> Ofc. tel. : 805-756-2184 >>> Dept. tel.: 805-756-2596 >>> Dept. fax: 805-756-6374 >>> E-mail: jrubba at calpoly.edu >>> URL: http://cla.calpoly.edu/~jrubba >>> >>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> http://spot.colorado.edu/~menn/index.html >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> Fellow, Linguistic Society of America >> >> Campus Mail Address: >> UCB 594, Institute for Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From dick at ling.ucl.ac.uk Thu Sep 9 07:51:19 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Thu, 9 Sep 2010 08:51:19 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: This discussion about the role of native-speaker intuition treats all native speakers and all levels of language equally. And yet we all agree with Joanna when she says "I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. " Why? Certainly not because we've got psycholinguistic evidence on this particular word. Isn't it something to do with maturity and training for the speakers, and meaningfulness for the levels? Think of that classic 1979 experiment by the Gleitmans that found massive differences in sensitivity to language structure with both age and education, so that /eat house bird/ is interpreted as 'a house-bird who is very eat' by students with a PhD (but not in linguistics) but as 'everybody is eating up their pet birds' by clerical staff. In this case, as in the other examples they tried, the more educated were right (by our standards) and the less educated were wrong. They also found big differences in reliability from level to level, with semantic judgements easiest and most reliable and phonological judgements least reliable, and syntax in between. That's presumably because ordinary speakers spend most of their time grappling with meaning ('Look after the sense and the sounds will look after themselves', as someone said to Alice in Wonderland). We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. Best wishes, Dick Hudson Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 09/09/2010 02:20, Johanna Rubba wrote: > The reason I make much of my students' disagreements with linguists' analyses is that, at least as I recall from my graduate education, native-speaker intuitions were routinely used in defense of analyses of sentence structure. I agree completely with Lise that introspection alone is not sufficient for defending an analysis. > > I never meant to suggest that my students store complex words disconnected from their component parts. Such a notion would be bizarre for a practitioner of Cognitive Grammar. Plus, the behavior I described doesn't contradict network connections; it challenges the usefulness of NS intuitions, which would, in turn, challenge analyses defended based on same. It's odd that, in some cases, students' tacit knowledge can be brought to consciousness, and in other cases not. > > It seems clear that context (including priming) affects a language-user's analysis. Certainly, many expressions that sound odd out of context sound perfectly fine in a suitable context. > > Jo > > On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: > > > Right on, Lise. And further, there is a well-known experimental technique called "semantic priming" that is admirably well suited for investigating whether when a language used hears "unhappiness", "happy" and "happiness" are activated ('come to mind'). This technique will probably not answer the question of the differential bracketing (un[happiness] vs. [unhappy]ness). And it is too rough to answer questions of directionality (does "unhappy" prime "happy" stronger than vice versa?). But it does tends to suggest that we don't store complex words in total disconnect from their parts, at least not as frequent adult users. And that phonological similarity (shared parts of words) has semantic consequences. Cheers, TG > > ============== > > > Lise Menn wrote: >> So we see an important phenomenon: Tacit knowledge really IS tacit, and 'intuitions' are very poor guides to what our minds are doing when we are using the patterns of our language as speakers/hearers. Introspection cannot replace observation of actual usage and psycholinguistic experiments; it can only act as a suggestion of where to dig. After all, we can't figure out vision or digestion by thinking about how they feel, although we certainly have to account for subjective feelings of contrast and indigestion. The same is true for language, mutatis mutandis. >> >> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >> >>> One thing that consistently occurs in my intro linguistics classes is that at least half of my students do not analyze complex words the way a linguist would -- many would analyze "unhappiness" as "un" + "happiness." They make such analyses over and over. It makes one wonder, of course, about how much native-speaker intuition is in agreement with some linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], but, apparently, large numbers of native speakers cannot. >>> >>> Another thing I often find is that many students cannot locate either primary or (especially) secondary stress in words. This is very bizarre, considering that they produce the stresses correctly and hear them correctly in others' speech. So many are unsuccessful at this that I have stopped requiring them to find stress in words on tests. I give them tricks like singing the word and monitoring for the highest-pitched syllable, but the tricks don't work. That many students can't be tone-deaf. >>> >>> Dr. Johanna Rubba, Ph. D. >>> Professor, Linguistics >>> Linguistics Minor Advisor >>> English Dept. >>> Cal Poly State University San Luis Obispo >>> San Luis Obispo, CA 93407 >>> Ofc. tel. : 805-756-2184 >>> Dept. tel.: 805-756-2596 >>> Dept. fax: 805-756-6374 >>> E-mail: jrubba at calpoly.edu >>> URL: http://cla.calpoly.edu/~jrubba >>> >>> >>> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> http://spot.colorado.edu/~menn/index.html >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> Fellow, Linguistic Society of America >> >> Campus Mail Address: >> UCB 594, Institute for Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> > > Dr. Johanna Rubba, Professor, Linguistics > Linguistics Minor Advisor > English Department > California Polytechnic State University, San Luis Obispo > E-mail: jrubba at calpoly.edu > Tel.: 805.756.2184 > Dept. Ofc. Tel.: 805.756.2596 > Dept. Fax: 805.756.6374 > URL: http://www.cla.calpoly.edu/~jrubba > > > From dan at daneverett.org Thu Sep 9 10:39:42 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 06:39:42 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8891F7.2070608@ling.ucl.ac.uk> Message-ID: Dick, You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. -- Dan > > We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. > > Best wishes, Dick Hudson > From dick at ling.ucl.ac.uk Thu Sep 9 12:16:50 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Thu, 9 Sep 2010 13:16:50 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 09/09/2010 11:39, Daniel Everett wrote: > Dick, > > You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". > > Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. > > It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. > > The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. > > -- Dan > >> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >> >> Best wishes, Dick Hudson >> > > From dan at daneverett.org Thu Sep 9 12:21:30 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 08:21:30 -0400 Subject: analysis: unhappiness In-Reply-To: <4C88D032.1060406@ling.ucl.ac.uk> Message-ID: Dick, I just am not so sure I share the faith in linguistic judgements across the board. Now *that* is something to test experimentally! Dan On 9 Sep 2010, at 08:16, Richard Hudson wrote: > Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. > > But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 09/09/2010 11:39, Daniel Everett wrote: >> Dick, >> >> You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". >> >> Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. >> >> It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. >> >> The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. >> >> -- Dan >> >>> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >>> >>> Best wishes, Dick Hudson >>> >> >> > From dan at daneverett.org Thu Sep 9 12:23:00 2010 From: dan at daneverett.org (Daniel Everett) Date: Thu, 9 Sep 2010 08:23:00 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: I mean, 'judgements by linguists'. -- Dan On 9 Sep 2010, at 08:21, Daniel Everett wrote: > Dick, > > I just am not so sure I share the faith in linguistic judgements across the board. > > Now *that* is something to test experimentally! > > Dan > > > > On 9 Sep 2010, at 08:16, Richard Hudson wrote: > >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that conscious judgements are only one kind of evidence that we need to take into account. I admire Carson Schutze's work (which I reviewed in fact), and of course I've been aware of complaints about judgements by people like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't equally reliable. If you want to know how /unhappiness/ is structured, ask a linguist, not a five-year old. And one of the by-products of education may be increased sensitivity to syntax - which is one of the many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that intuitions are a fine way to generate hypotheses and even to test them - to a degree. But while it might not have been feasible for Huddleston, Pullum, and the other contributors to the Cambridge Grammar to conduct experiments on every point of the grammar, experiments could have only made the grammar better. The use of intuitions, corpora, and standard psycholinguistic experimentation (indeed, Standard Social Science Methodology) is vital for taking the field forward and for providing the best support for different analyses. Ted Gibson and Ev Fedorenko have written a very useful new paper on this, showing serious shortcomings with intuitions as the sole source of evidence, in their paper: "The need for quantitative methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive Science), and researchers from Brazil are beginning a third round of experimental work among the Pirahas, since my own work on the syntax was, like almost every other field researcher's, based on native speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions to Greenberg's work on classifying the languages of the Americas. His methods were strongly (and justifiably) criticized. However, I always thought that his methods were a great way of generating hypotheses, so long as they were ultimately put to the test of standard historical linguistics methods. And the same seems true for use of native-speaker intuitions. >>> >>> -- Dan >>> >>>> We linguists can add a further layer of explanation to the judgements, but some judgements do seem to be more reliable than others. And if we have to wait for psycholinguistic evidence for every detailed analysis we make, our whole discipline will immediately grind to a halt. Like it or not, native speaker judgements are what put us linguists ahead of the rest in handling fine detail. Imagine writing the Cambridge Grammar of the English Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> > > From amnfn at well.com Thu Sep 9 12:26:02 2010 From: amnfn at well.com (A. Katz) Date: Thu, 9 Sep 2010 05:26:02 -0700 Subject: analysis: unhappiness In-Reply-To: <4C8891F7.2070608@ling.ucl.ac.uk> Message-ID: Richard, What you just pointed out -- that speaker sensitivity to language structure varies from individual to individual and can be affected by experience and training -- goes toward an even more fundamental point: that language structure exists separate and apart from how individual speakers process it. --Aya On Thu, 9 Sep 2010, Richard Hudson wrote: > This discussion about the role of native-speaker intuition treats all native > speakers and all levels of language equally. And yet we all agree with Joanna > when she says "I can *feel* that the analysis is [[un-happy]-ness], but, > apparently, large numbers of native speakers cannot. " Why? Certainly not > because we've got psycholinguistic evidence on this particular word. Isn't it > something to do with maturity and training for the speakers, and > meaningfulness for the levels? > > Think of that classic 1979 experiment by the Gleitmans that found massive > differences in sensitivity to language structure with both age and education, > so that /eat house bird/ is interpreted as 'a house-bird who is very eat' by > students with a PhD (but not in linguistics) but as 'everybody is eating up > their pet birds' by clerical staff. In this case, as in the other examples > they tried, the more educated were right (by our standards) and the less > educated were wrong. They also found big differences in reliability from > level to level, with semantic judgements easiest and most reliable and > phonological judgements least reliable, and syntax in between. That's > presumably because ordinary speakers spend most of their time grappling with > meaning ('Look after the sense and the sounds will look after themselves', as > someone said to Alice in Wonderland). > > We linguists can add a further layer of explanation to the judgements, but > some judgements do seem to be more reliable than others. And if we have to > wait for psycholinguistic evidence for every detailed analysis we make, our > whole discipline will immediately grind to a halt. Like it or not, native > speaker judgements are what put us linguists ahead of the rest in handling > fine detail. Imagine writing the Cambridge Grammar of the English Language > (or the OED) without using native speaker judgements. > > Best wishes, Dick Hudson > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 09/09/2010 02:20, Johanna Rubba wrote: >> The reason I make much of my students' disagreements with linguists' >> analyses is that, at least as I recall from my graduate education, >> native-speaker intuitions were routinely used in defense of analyses of >> sentence structure. I agree completely with Lise that introspection alone >> is not sufficient for defending an analysis. >> >> I never meant to suggest that my students store complex words disconnected >> from their component parts. Such a notion would be bizarre for a >> practitioner of Cognitive Grammar. Plus, the behavior I described doesn't >> contradict network connections; it challenges the usefulness of NS >> intuitions, which would, in turn, challenge analyses defended based on >> same. It's odd that, in some cases, students' tacit knowledge can be >> brought to consciousness, and in other cases not. >> >> It seems clear that context (including priming) affects a language-user's >> analysis. Certainly, many expressions that sound odd out of context sound >> perfectly fine in a suitable context. >> >> Jo >> >> On Sep 8, 2010, at 6:03 PM, Tom Givon wrote: >> >> >> Right on, Lise. And further, there is a well-known experimental technique >> called "semantic priming" that is admirably well suited for investigating >> whether when a language used hears "unhappiness", "happy" and "happiness" >> are activated ('come to mind'). This technique will probably not answer the >> question of the differential bracketing (un[happiness] vs. [unhappy]ness). >> And it is too rough to answer questions of directionality (does "unhappy" >> prime "happy" stronger than vice versa?). But it does tends to suggest that >> we don't store complex words in total disconnect from their parts, at least >> not as frequent adult users. And that phonological similarity (shared parts >> of words) has semantic consequences. Cheers, TG >> >> ============== >> >> >> Lise Menn wrote: >>> So we see an important phenomenon: Tacit knowledge really IS tacit, and >>> 'intuitions' are very poor guides to what our minds are doing when we are >>> using the patterns of our language as speakers/hearers. Introspection >>> cannot replace observation of actual usage and psycholinguistic >>> experiments; it can only act as a suggestion of where to dig. After all, >>> we can't figure out vision or digestion by thinking about how they feel, >>> although we certainly have to account for subjective feelings of contrast >>> and indigestion. The same is true for language, mutatis mutandis. >>> >>> On Sep 8, 2010, at 9:26 AM, Johanna Rubba wrote: >>> >>>> One thing that consistently occurs in my intro linguistics classes is >>>> that at least half of my students do not analyze complex words the way a >>>> linguist would -- many would analyze "unhappiness" as "un" + "happiness." >>>> They make such analyses over and over. It makes one wonder, of course, >>>> about how much native-speaker intuition is in agreement with some >>>> linguistic analyses. I can *feel* that the analysis is [[un-happy]-ness], >>>> but, apparently, large numbers of native speakers cannot. >>>> >>>> Another thing I often find is that many students cannot locate either >>>> primary or (especially) secondary stress in words. This is very bizarre, >>>> considering that they produce the stresses correctly and hear them >>>> correctly in others' speech. So many are unsuccessful at this that I have >>>> stopped requiring them to find stress in words on tests. I give them >>>> tricks like singing the word and monitoring for the highest-pitched >>>> syllable, but the tricks don't work. That many students can't be >>>> tone-deaf. >>>> >>>> Dr. Johanna Rubba, Ph. D. >>>> Professor, Linguistics >>>> Linguistics Minor Advisor >>>> English Dept. >>>> Cal Poly State University San Luis Obispo >>>> San Luis Obispo, CA 93407 >>>> Ofc. tel. : 805-756-2184 >>>> Dept. tel.: 805-756-2596 >>>> Dept. fax: 805-756-6374 >>>> E-mail: jrubba at calpoly.edu >>>> URL: http://cla.calpoly.edu/~jrubba >>>> >>>> >>>> >>> Lise Menn Home Office: 303-444-4274 >>> 1625 Mariposa Ave Fax: 303-413-0017 >>> Boulder CO 80302 >>> http://spot.colorado.edu/~menn/index.html >>> Professor Emerita of Linguistics >>> Fellow, Institute of Cognitive Science >>> University of Colorado >>> >>> Secretary, AAAS Section Z [Linguistics] >>> Fellow, Linguistic Society of America >>> >>> Campus Mail Address: >>> UCB 594, Institute for Cognitive Science >>> >>> Campus Physical Address: >>> CINC 234 >>> 1777 Exposition Ave, Boulder >>> >>> >>> >> >> Dr. Johanna Rubba, Professor, Linguistics >> Linguistics Minor Advisor >> English Department >> California Polytechnic State University, San Luis Obispo >> E-mail: jrubba at calpoly.edu >> Tel.: 805.756.2184 >> Dept. Ofc. Tel.: 805.756.2596 >> Dept. Fax: 805.756.6374 >> URL: http://www.cla.calpoly.edu/~jrubba >> >> >> > > From bischoff.st at gmail.com Thu Sep 9 12:53:48 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Thu, 9 Sep 2010 08:53:48 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Thanks to all for the replies. One thing that strikes me as a result of these responses is how my conception of linguistic concepts, objects, and processes, if you will, is so surprisingly skewed and limited. I say surprisingly because while I have been trained as a generativist I have aspired to look at language from various perspectives and find that many of those are more appealing various reasons. Yet, at some basic level, generative linguistics seems to be my point of reference. I recall as a graduate student one instructor noting that "they could no longer think of language in any other way [than generative]"...that when they saw a sentence they "saw trees." Curious... On Wed, Sep 8, 2010 at 4:35 PM, George Lakoff wrote: > Think about "undo" and "untangle." > > In addition, morphology may not reflect semantics. One might have > un+happiness in surface morphology and [un+happy] ness for semantics, which > is perfectly natural. > > George > > > On Wed, Sep 8, 2010 at 5:46 AM, s.t. bischoff wrote: > >> Hi all, >> >> I had an interesting exchange with a few generative >> syntacticians/morphologists (former classmates of mine) regarding an >> analysis of "unhappiness". Two things that they said surprised me a bit, >> they are the following: >> >> (1) un- (negation, 'not') only attaches to adjectives (now this clearly >> isn't the case, a simple cursory view of the etymology in the OED provides >> a >> number of examples of un- with nouns and verbs...though to significantly >> lesser degrees...in addition works on English morphology contain examples >> as >> well) >> >> (2) the analysis of unhappiness can only be [[un-happy]-ness]...an >> analysis >> such as [un-[happy-ness]] is impossible (due to (1) above according to my >> former colleagues). >> >> My questions are the following: >> >> (1) Is there a good/well grounded reason to believe un- "only" attaches >> to >> adjectives? >> >> (2) What would be the consensus on an analysis of "unhappiness" that most >> linguists would agree on? >> >> Thanks, >> Shannon >> > > From dryer at buffalo.edu Thu Sep 9 14:00:47 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Thu, 9 Sep 2010 10:00:47 -0400 Subject: analysis: unhappiness Message-ID: Two comments. First (elaborating perhaps on Dick Hudson's comment), I think there is an important distinction between low-level linguistic intuitions (like whether a word or sentence is well-formed or what it means) and higher-level intuitions (like what the structure of a word or sentence is). One can take the position that we need to account for the former (while recognizing that they are not always reliable) but not the latter. Second, the tension here is not only between evidence from speaker intuitions versus evidence from psycholinguistic experiments. There is also a tension between deciding on the correct analysis on the basis of a priori simplicity arguments versus deciding on the correct analysis on the basis of psycholinguistic evidence (see Derwing 1973). The bracketing paradox that Dan referred to that arises with the word (semantics argues for [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] (the comparative suffix can only be attached to adjectives containing one or two syllables) is only a paradox if one assumes that speakers adopt the simplest analysis. For example, if speakers adopt a more complex rule for either of these (e.g. perhaps the rule for attaching -er can apply exceptionally to trisyllabic words beginning with un-), then the bracketing paradox disappears. Matthew On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > Thanks Dan. I'm sure you're right, and I'd be the first to agree that > conscious judgements are only one kind of evidence that we need to take > into account. I admire Carson Schutze's work (which I reviewed in fact), > and of course I've been aware of complaints about judgements by people > like Labov for decades. > > But you're missing my main point, which is that all judgements aren't > equally reliable. If you want to know how /unhappiness/ is structured, > ask a linguist, not a five-year old. And one of the by-products of > education may be increased sensitivity to syntax - which is one of the > many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > On 09/09/2010 11:39, Daniel Everett wrote: > > Dick, > > > > You raise an important issue here about > methodology. I believe that intuitions are a fine way to generate > hypotheses and even to test them - to a degree. But while it might not have > been feasible for Huddleston, Pullum, and the other contributors to the > Cambridge Grammar to conduct experiments on every point of the grammar, > experiments could have only made the grammar better. The use of intuitions, > corpora, and standard psycholinguistic experimentation (indeed, Standard > Social Science Methodology) is vital for taking the field forward and for > providing the best support for different analyses. Ted Gibson and Ev > Fedorenko have written a very useful new paper on this, showing serious > shortcomings with intuitions as the sole source of evidence, in their > paper: "The need for quantitative methods in syntax and semantics > research".> > > Carson Schutze and Wayne Cowart, among others, > have also written convincingly on this.> > > It is one reason that a team from Stanford, MIT > (Brain and Cognitive Science), and researchers from Brazil are beginning a > third round of experimental work among the Pirahas, since my own work on > the syntax was, like almost every other field researcher's, based on native > speaker intuitions and corpora.> > > The discussion of methodologies reminds me of > the initial reactions to Greenberg's work on classifying the languages of > the Americas. His methods were strongly (and justifiably) criticized. > However, I always thought that his methods were a great way of generating > hypotheses, so long as they were ultimately put to the test of standard > historical linguistics methods. And the same seems true for use of > native-speaker intuitions.> > > -- Dan > > > >> We linguists can add a further layer of > explanation to the judgements, but some judgements do seem to be more > reliable than others. And if we have to wait for psycholinguistic evidence > for every detailed analysis we make, our whole discipline will immediately > grind to a halt. Like it or not, native speaker judgements are what put us > linguists ahead of the rest in handling fine detail. Imagine writing the > Cambridge Grammar of the English Language (or the OED) without using native > speaker judgements.>> > >> Best wishes, Dick Hudson > >> > > > > > > > > > From amnfn at well.com Thu Sep 9 14:27:08 2010 From: amnfn at well.com (A. Katz) Date: Thu, 9 Sep 2010 07:27:08 -0700 Subject: analysis: unhappiness In-Reply-To: <7866.1284040847@buffalo.edu> Message-ID: The use of "unhappier" is idiosyncratic and not universal among English speakers. Rather than assuming that a person who says "unhappier" has a very complicated rule about trisyllabic words beginning in "un", would it not make more sense to see this as an indication that for this particular speaker, the derivation of "unhappy" is opague? That is, the word is treated as an indivisible whole. On the other hand, a speaker who says "more unhappy" is probably aware of the derivation of unhappy and takes it into account when applying the rule for comparatives. Best, --Aya On Thu, 9 Sep 2010, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there is an > important distinction between low-level linguistic intuitions (like whether a > word or sentence is well-formed or what it means) and higher-level intuitions > (like what the structure of a word or sentence is). One can take the position > that we need to account for the former (while recognizing that they are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker intuitions > versus evidence from psycholinguistic experiments. There is also a tension > between deciding on the correct analysis on the basis of a priori simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing paradox that Dan > referred to that arises with the word (semantics argues for > [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] (the > comparative suffix can only be attached to adjectives containing one or two > syllables) is only a paradox if one assumes that speakers adopt the simplest > analysis. For example, if speakers adopt a more complex rule for either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to take >> into account. I admire Carson Schutze's work (which I reviewed in fact), >> and of course I've been aware of complaints about judgements by people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable. If you want to know how /unhappiness/ is structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might not have >> been feasible for Huddleston, Pullum, and the other contributors to the >> Cambridge Grammar to conduct experiments on every point of the grammar, >> experiments could have only made the grammar better. The use of intuitions, >> corpora, and standard psycholinguistic experimentation (indeed, Standard >> Social Science Methodology) is vital for taking the field forward and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are beginning a >> third round of experimental work among the Pirahas, since my own work on >> the syntax was, like almost every other field researcher's, based on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of generating >> hypotheses, so long as they were ultimately put to the test of standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic evidence >> for every detailed analysis we make, our whole discipline will immediately >> grind to a halt. Like it or not, native speaker judgements are what put us >> linguists ahead of the rest in handling fine detail. Imagine writing the >> Cambridge Grammar of the English Language (or the OED) without using native >> speaker judgements.>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> >> >> >> >> > > From dryer at buffalo.edu Thu Sep 9 14:42:16 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Thu, 9 Sep 2010 10:42:16 -0400 Subject: analysis: unhappiness Message-ID: Aya proposes a possible alternative hypothesis, but my point is that whether this alternative is correct or not is not a question of whether it "makes more sense", but an empirical question that can really only be decided on the basis of psycholinguistic evidence. Matthew On Thu 09/09/10 10:27 AM , "A. Katz" amnfn at well.com sent: > The use of "unhappier" is idiosyncratic and not universal among > English speakers. Rather than assuming that a person who says "unhappier" > has a very complicated rule about trisyllabic words beginning in "un", > would it not make more sense to see this as an indication that for this particular > speaker, the derivation of "unhappy" is opague? That is, the word > is treated as an indivisible whole. > > On the other hand, a speaker who says "more unhappy" is probably > aware of the derivation of unhappy and takes it into account when applying the rule > for comparatives. > > Best, > > --Aya > > > On Thu, 9 Sep 2010, Matthew S. Dryer wrote: > > > > > Two comments. > > > > First (elaborating perhaps on Dick Hudson's > comment), I think there is an> important distinction between low-level > linguistic intuitions (like whether a> word or sentence is well-formed or what it > means) and higher-level intuitions> (like what the structure of a word or sentence > is). One can take the position> that we need to account for the former (while > recognizing that they are not> always reliable) but not the latter. > > > > Second, the tension here is not only between > evidence from speaker intuitions> versus evidence from psycholinguistic > experiments. There is also a tension> between deciding on the correct analysis on the > basis of a priori simplicity> arguments versus deciding on the correct > analysis on the basis of> psycholinguistic evidence (see Derwing 1973). > The bracketing paradox that Dan> referred to that arises with the word > (semantics argues for> [[un+happi] + er], morphology and phonology > argues for [un + [happi+er]] (the> comparative suffix can only be attached to > adjectives containing one or two> syllables) is only a paradox if one assumes that > speakers adopt the simplest> analysis. For example, if speakers adopt a more > complex rule for either of these> (e.g. perhaps the rule for attaching -er can > apply exceptionally to trisyllabic> words beginning with un-), then the bracketing > paradox disappears.> > > Matthew > > > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling > .ucl.ac.uk sent:>> Thanks Dan. I'm sure you're right, and I'd > be the first to agree that>> conscious judgements are only one kind of > evidence that we need to take>> into account. I admire Carson Schutze's work > (which I reviewed in fact),>> and of course I've been aware of complaints > about judgements by people>> like Labov for decades. > >> > >> But you're missing my main point, which is > that all judgements aren't>> equally reliable. If you want to know how > /unhappiness/ is structured,>> ask a linguist, not a five-year old. And one > of the by-products of>> education may be increased sensitivity to > syntax - which is one of the>> many reasons why linguists need to pay more > attention to education.>> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>> On 09/09/2010 11:39, Daniel Everett > wrote:>>> Dick, > >>> > >>> You raise an important issue here > about>> methodology. I believe that intuitions are a > fine way to generate>> hypotheses and even to test them - to a > degree. But while it might not have>> been feasible for Huddleston, Pullum, and > the other contributors to the>> Cambridge Grammar to conduct experiments on > every point of the grammar,>> experiments could have only made the grammar > better. The use of intuitions,>> corpora, and standard psycholinguistic > experimentation (indeed, Standard>> Social Science Methodology) is vital for > taking the field forward and for>> providing the best support for different > analyses. Ted Gibson and Ev>> Fedorenko have written a very useful new > paper on this, showing serious>> shortcomings with intuitions as the sole > source of evidence, in their>> paper: "The need for quantitative > methods in syntax and semantics>> research".> > >>> Carson Schutze and Wayne Cowart, among > others,>> have also written convincingly on > this.>>>> It is one reason that a team from > Stanford, MIT>> (Brain and Cognitive Science), and > researchers from Brazil are beginning a>> third round of experimental work among the > Pirahas, since my own work on>> the syntax was, like almost every other > field researcher's, based on native>> speaker intuitions and > corpora.>>>> The discussion of methodologies reminds > me of>> the initial reactions to Greenberg's work on > classifying the languages of>> the Americas. His methods were strongly (and > justifiably) criticized.>> However, I always thought that his methods > were a great way of generating>> hypotheses, so long as they were ultimately > put to the test of standard>> historical linguistics methods. And the same > seems true for use of>> native-speaker intuitions.> > >>> -- Dan > >>> > >>>> We linguists can add a further layer > of>> explanation to the judgements, but some > judgements do seem to be more>> reliable than others. And if we have to wait > for psycholinguistic evidence>> for every detailed analysis we make, our > whole discipline will immediately>> grind to a halt. Like it or not, native > speaker judgements are what put us>> linguists ahead of the rest in handling fine > detail. Imagine writing the>> Cambridge Grammar of the English Language > (or the OED) without using native>> speaker judgements.>> > >>>> Best wishes, Dick Hudson > >>>> > >>> > >>> > >> > >> > >> > >> > >> > > > > > > > > > From khildeb at siue.edu Thu Sep 9 14:58:22 2010 From: khildeb at siue.edu (Kristine Hildebrandt) Date: Thu, 9 Sep 2010 09:58:22 -0500 Subject: Call For Papers: Special Issue of Himalayan Linguistics Message-ID: Himalayan Linguistics 10.1 (June 2011) Guest Editors: Yogendra Yadava, Karen Grunow-Harsta, Kristine Hildebrandt, Stephen Watters Himalayan Linguistics, a free peer-reviewed web journal and archive devoted to the study of the languages of the Himalayas, is now accepting submissions to a special issue in memory of our late colleagues, HL Associate Editors Michael (Mickey) Noonan and David Watters. Articles on all languages of the Himalayan region are welcome, as are those that significantly draw on work by Noonan or Watters on Himalayan languages. Deadline for submissions: 15 October 2010 Address inquiries to Guest Editor Yogendra Yadava (ypyadava at gmail.com) or HL Editor Carol Genetti (cgenetti at linguistics.ucsb.edu) http://www.linguistics.ucsb.edu/HimalayanLinguistics/ -- Kristine A. Hildebrandt Department of English Language & Literature Southern Illinois University Edwardsville Edwardsville, IL 62026 U.S.A. 618-650-3380 (office) khildeb at siue.edu http://www.siue.edu/~khildeb From Jean-Christophe.Verstraete at arts.kuleuven.be Thu Sep 9 18:54:23 2010 From: Jean-Christophe.Verstraete at arts.kuleuven.be (Jean-Christophe Verstraete) Date: Thu, 9 Sep 2010 20:54:23 +0200 Subject: First call: Greenberg Award 2011 Message-ID: The Joseph Greenberg Award 2011 The Association for Linguistic Typology's Joseph Greenberg Award recognizes and honours the best piece of typological research embodied in a doctoral dissertation or equivalent in 2009-2010. Theses are eligible if they were accepted by a university between 1 January 2009 and 31 December 2010. The award will consist of payment of travel, per diem expenses and registration fee to attend the ALT IX Conference, to be held in Hong Kong, July 21-24, 2011, and to present a synopsis or element of the prize-winning work as a plenary lecture at that meeting. The Joseph Greenberg Award was named to remember Joseph Greenberg's (1915-2001) fundamental contributions to typology and the interest he showed in encouraging young researchers. Between 1998 and 2006, it was known as the "ALT Junior Award". To be eligible, those submitting their dissertation must be members of the Association for Linguistic Typology (ALT). They are asked to submit their manuscript by email in pdf format, with all non-standard fonts in Unicode, to the chair of the jury, to arrive no later than January 31st 2011. If this proves technically difficult, the candidate is asked to discuss the problem with the chair. A jury, consisting of about ten ALT members, will be appointed by ALT's president, appropriate to the work submitted. The chair will be: Isabelle Bril Laboratoire LACITO 7, rue Guy M?quet (b?t. D) 94801 Villejuif Cedex France From sclancy at uchicago.edu Thu Sep 9 19:10:15 2010 From: sclancy at uchicago.edu (Steven Clancy) Date: Thu, 9 Sep 2010 14:10:15 -0500 Subject: SCLC-2010 Conference Program In-Reply-To: <4C827393.1060608@sheffield.ac.uk> Message-ID: THE TENTH ANNUAL CONFERENCE OF THE SLAVIC COGNITIVE LINGUISTICS ASSOCIATION (SCLC-2010) October 9-11, 2010 Brown University (Providence, Rhode Island, USA) Hosted by the Department of Slavic Languages, Brown University (Providence, Rhode Island) with support from the Office of International Affairs, the Colver Lectureship Fund from the Office of the Dean of the Faculty, the Center for Language Studies, the Department of Cognitive, Linguistics and Psychologial Sciences, and John Benjamins Publishing Company. THE TENTH ANNUAL CONFERENCE OF THE SLAVIC COGNITIVE LINGUISTICS ASSOCIATION (SCLC-2010) October 9-11, 2010 The Slavic Cognitive Linguistics Association (SCLA) announces the program for the 2010 annual conference. The conference will be held on the campus of Brown University (Providence, Rhode Island) on Saturday, October 9 through Monday, October 11, 2010. SCLC-2010 Keynote Speakers Cognitively Plausible Parsing Eugene Charniak Brown University, USA The role of rational discounting: what to say, what not to say Adele E. Goldberg Princeton University, USA Access, Activation, and Overlap: Focusing on the Differential Ronald W. Langacker University of California, San Diego, USA MAIN SESSIONS (Saturday, Sunday, and Monday) Each presentation for the main sessions will be given 20 minutes and will be followed by a 10-minute discussion period. SCHEDULE View the Conference Schedule in PDF Format. CONFERENCE INFORMATION Membership in the SCLA is free, please see "Joining SCLA" for more information. Attendance at SCLC-2010 is open and all interested individuals are encouraged to attend. Information on Conference Fees and Registration are provided below. Brown University is located in Providence, Rhode Island and is accessible from Boston Logan International Airport (BOS, 55 miles away) or T.F. Green Airport (PVD) in Providence. Transportation from Airports: If you arrive at Boston Logan International Airport (BOS), you can get to Providence by bus (Peter Pan Bus, currently $36 round trip) from Boston Logan to Kennedy Plaza (in downtown Providence, tell the driver you are dropping off at Kennedy Plaza when you board), then by taxi to the Radisson Hotel in Providence (1.6 miles to the hotel) or by subway to South Station and by bus or train to Providence, then by taxi to the Radisson Hotel in Providence. Taxi from Logan to Providence would cost more than $100 and is not recommended. Ground transportation from Boston: further information. If you arrive directly at Providence T.F. Green International Airport (PVD) To and from PVD to downtown area: further information. Taxi from Providence Airport to Providence will cost around $30 one way. Ground transportation from Providence is much simpler. But international travelers may find more flights flying into Boston. Conference Hotel: Radisson Hotel Providence Harbor 220 India Street, Providence, RI 02903, Reservations: 1-800-395-7046 US/Canada Toll-free Telephone: (401) 272-5577 Fax: (401) 272-0251 Email: RHI_PROV at radisson.com Participants should mention the Slavic Cognitive Linguistics Association Conference to reserve a room at the special conference rate ($79USD plus 13% tax). Special rates are valid from Thursday 10/7 to Monday 10/11 (fewer rooms are available on Thursday and Monday). Additional people per room is $10 per person. Reservations at the conference rate must be made by 5:00pm local time, September 8, 2010. Please make your reservations early to ensure a room at the conference hotel. Conference Fees: Registration Fee: Regular participants $60USD Graduate student participants $40USD Conference dinner: $50USD Please make your checks payable to "Brown University" and send them to: Masako Fidler SCLC Organizer 20 Manning Walk, Box E Department of Slavic Languages Brown University Providence, RI02912 If you wish to preregister by wiring money for conference fees, please contact Gisela Belton for details. If you will need to register on-site, please contactSteven Clancy to confirm your participation. We hope you will be able to join us for SCLC-2010. Please forward this call for papers to your colleagues and graduate students who may be interested in presenting or attending. Sincerely, Steven Clancy President, SCLA Tore Nesset Vice-President, SCLA Masako Fidler Conference Organizer and Host Brown University on behalf of the SCLA officers and the 2010 SCLA organizing committee From jaejung.song at stonebow.otago.ac.nz Thu Sep 9 22:46:38 2010 From: jaejung.song at stonebow.otago.ac.nz (Jae Jung Song) Date: Fri, 10 Sep 2010 10:46:38 +1200 Subject: Oxford Handbook of Linguistic Typology Message-ID: Dear Funknetters, May I draw your attention to the imminent publication of the Oxford Handbook of Linguistic Typology? The table of contents appears below. If you would like (to ask your university/institution library) to pre-order a copy, you can go to: http://www.oup.com/us/catalog/general/subject/Linguistics/SyntaxMorphology/?view=usa&ci=9780199281251 With best wishes, Jae Jung Song ----------------------------------------------------------------- The Oxford Handbook of Linguistic Typology Edited by Jae Jung Song November 2010, 776 pages, 978-0-19-928125-1, hardback, ? 85.00 Oxford University Press Part I. Setting the Stage - Jae Jung Song Part II. Foundations: History, Theory and Method 1. The (Early) History of Linguistic Typology - Paolo Ramat 2. The Pioneers of Linguistic Typology: From Gabelentz to Greenberg - Giorgio Graffi 3. Linguistic Typology and the Study of Language - Michael Daniel 4. Explaining Language Universals - Edith Moravcsik 5. The Problem of Cross-Linguistic Identification - Leon Stassen 6. Language Sampling - Dik Bakker Part III. Theoretical Dimensions of Linguistic Typology 7. Markedness: Iconicity, Economy and Frequency - Joan Bybee 8. Competing Motivations - John Haiman 9. Categories and Prototypes - Johan van der Auwera and Volker Gast 10. Implicational Hierarchies - Greville Corbett 11. Processing Efficiency and Complexity in Typological Patterns - John Hawkins 12. Language Universals and Linguistic Knowledge - Sonia Cristofaro Part IV. Empirical Dimensions of Linguistic Typology 13. Word Order Typology - Jae Jung Song 14. Word Classes - Walter Bisang 15. Case-Marking Typology - Beatrice Primus 16. Person Marking - Anna Siewierska 17. Transitivity Typology - Seppo Kittil? 18. Voice Typology - Leonid Kulikov 19. Grammatical Relation Typology - Balthasar Bickel 20.Typology of Tense, Aspect and Modality Systems - Ferdinand de Haan 21. Syntactic Typology - Lindsay Whaley 22. Morphological Typology - Dunstan Brown 23. Semantic Typology - Nicholas Evans 24. Typology of Phonological Systems - Ian Maddieson Part V. Linguistic Typology in a Wider Context 25. Linguistic Typology and Historical Linguistics - Kenneth Shields 26. Linguistic Typology and Language Contact - Maria Koptjevskaja-Tamm 27. Linguistic Typology and First Language Acquisition - Melissa Bowerman 28. Linguistic Typology and Second Language Acquisition - Fred Eckman 29. Linguistic Typology and Language Documentation - Patience Epps 30. Linguistic Typology and Formal Grammar - Maria Polinsky References Author Index Language Index Subject Index From Lise.Menn at Colorado.EDU Thu Sep 9 23:26:13 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Thu, 9 Sep 2010 17:26:13 -0600 Subject: analysis: unhappiness In-Reply-To: <7866.1284040847@buffalo.edu> Message-ID: I wish we had better terminology for keeping track of whether, at a given time, we are talking about the patterns that are 'out there' in the language and might possibly be apprehended (subconsciously) by a speaker, and when we are talking about the patterns that a particular speaker actually does apprehend, as indicated by experiments, from simple 'wug tests' up to brain wave and eye-gaze studies. And for distinguishing among the degrees of pattern apprehension that a person may have, from vague preferences detectable in reaction times or other behavior all the way up through clear metalinguistic insights. Dick Hudson's note reminding us of the Gleitman and Gleitman study is right on target. Since we don't in fact have such an agreed-on terminology, we have to be quite careful in making clear what we are referring to when we talk about 'the correct analysis' of a form like 'unhappiness'. We know, but tend to forget - and tend to forget to tell our students! - that it's an empirical question as to whether the formal simplicity and coherence of description of forms 'out there' (e.g. lovely abstract morphophonemics) is any kind of approximation to the way knowledge of the same forms is organized in a particular person's head. If we remember that a very large proportion of what we know about our language is 'out there' when we are infants and has to be internalized through experience with the language (even if you believe in innate 'core language'), the variation in internal knowledge from one person to another is more understandable. We especially need to consider (and try to test) the possibility that since the brain can make multiple cross-connections, multiple patterns are involved simultaneously in morphological and syntactic analyses. I suggest that that's the case with 'unhappiness' - and the linguistic analyses that I know about are not good at handling that kind of idea. Lise On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there > is an > important distinction between low-level linguistic intuitions (like > whether a > word or sentence is well-formed or what it means) and higher-level > intuitions > (like what the structure of a word or sentence is). One can take > the position > that we need to account for the former (while recognizing that they > are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker > intuitions > versus evidence from psycholinguistic experiments. There is also a > tension > between deciding on the correct analysis on the basis of a priori > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing > paradox that Dan > referred to that arises with the word (semantics argues > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one > or two > syllables) is only a paradox if one assumes that speakers adopt the > simplest > analysis. For example, if speakers adopt a more complex rule for > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to >> take >> into account. I admire Carson Schutze's work (which I reviewed in >> fact), >> and of course I've been aware of complaints about judgements by >> people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable. If you want to know how /unhappiness/ is >> structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of >> the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might >> not have >> been feasible for Huddleston, Pullum, and the other contributors to >> the >> Cambridge Grammar to conduct experiments on every point of the >> grammar, >> experiments could have only made the grammar better. The use of >> intuitions, >> corpora, and standard psycholinguistic experimentation (indeed, >> Standard >> Social Science Methodology) is vital for taking the field forward >> and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing >> serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are >> beginning a >> third round of experimental work among the Pirahas, since my own >> work on >> the syntax was, like almost every other field researcher's, based >> on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the >> languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of >> generating >> hypotheses, so long as they were ultimately put to the test of >> standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic >> evidence >> for every detailed analysis we make, our whole discipline will >> immediately >> grind to a halt. Like it or not, native speaker judgements are what >> put us >> linguists ahead of the rest in handling fine detail. Imagine >> writing the >> Cambridge Grammar of the English Language (or the OED) without >> using native >> speaker judgements.>> >>>> Best wishes, Dick Hudson >>>> >>> >>> >> >> >> >> >> > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From edith at uwm.edu Thu Sep 9 23:55:17 2010 From: edith at uwm.edu (Edith A Moravcsik) Date: Thu, 9 Sep 2010 18:55:17 -0500 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: It seems to me that it is important to construct grammars that are based simply on the form-meaning correspondences of?the language?'out there', as Lise said, apart from psycholinguistic data, and that are constrained by the usual requirements of scientific descriptions, such as generality and simplicity. Such grammars may not be psycholinguistically real at all nor do they try to be; but they provide a baseline in reference to which we can then assess people's actual ways of learning, processing, and analyzing structures. In the absence of such basic descriptions, it?would be?hard to know what psycholinguistic data we should be "surprised at" - i.e., what it is that needs to be explained about acquisition, processing, and people's intuitive analyses of structures. Edith Moravcsik From: "Lise Menn" To: dryer at buffalo.edu, "Funknet" Cc: "Richard Hudson" Sent: Thursday, September 9, 2010 6:26:13 PM Subject: Re: [FUNKNET] analysis: unhappiness I wish we had better terminology for keeping track of whether, at a ? given time, we are talking about the patterns that are 'out there' in ? the language and might possibly be apprehended (subconsciously) by a ? speaker, and when we are talking about the patterns that a particular ? speaker actually does apprehend, as indicated by experiments, from ? simple 'wug tests' up to brain wave and eye-gaze studies. ?And for ? distinguishing among the degrees of pattern apprehension that a person ? may have, from vague preferences detectable in reaction times or other ? behavior all the way up through clear metalinguistic insights. ?Dick ? Hudson's note reminding us of the Gleitman and Gleitman study is right ? on target. ????????Since we don't in fact have such an agreed-on terminology, we have to ? be quite careful in making clear what we are referring to when we talk ? about 'the correct analysis' of a form like 'unhappiness'. ?We know, ? but tend to forget - and tend to forget to tell our students! - that ? it's an empirical question as to whether the formal simplicity and ? coherence of description of forms 'out there' (e.g. lovely abstract ? morphophonemics) is any kind of approximation to the way knowledge of ? the same forms is organized in a particular person's head. ?If we ? remember that a very large proportion of what we know about our ? language is 'out there' when we are infants and has to be internalized ? through experience with the language (even if you believe in innate ? 'core language'), the variation in internal knowledge from one person ? to another is more understandable. ???????? ????????We especially need to consider (and try to test) the possibility that ? since the brain can make multiple cross-connections, multiple patterns are ? involved simultaneously in morphological and syntactic analyses. I suggest that ? that's the case with 'unhappiness' - and the linguistic analyses that ? I know about are not good at handling that kind of idea. ????????Lise On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there ? > is an > important distinction between low-level linguistic intuitions (like ? > whether a > word or sentence is well-formed or what it means) and higher-level ? > intuitions > (like what the structure of a word or sentence is). ?One can take ? > the position > that we need to account for the former (while recognizing that they ? > are not > always reliable) but not the latter. > > Second, the tension here is not only between evidence from speaker ? > intuitions > versus evidence from psycholinguistic experiments. ?There is also a ? > tension > between deciding on the correct analysis on the basis of a priori ? > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). ?The bracketing ? > paradox that Dan > referred to that arises with the word (semantics argues ? > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one ? > or two > syllables) is only a paradox if one assumes that speakers adopt the ? > simplest > analysis. ?For example, if speakers adopt a more complex rule for ? > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to ? > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > > Matthew > > On Thu 09/09/10 ?8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >> conscious judgements are only one kind of evidence that we need to ? >> take >> into account. I admire Carson Schutze's work (which I reviewed in ? >> fact), >> and of course I've been aware of complaints about judgements by ? >> people >> like Labov for decades. >> >> But you're missing my main point, which is that all judgements aren't >> equally reliable. ?If you want to know how /unhappiness/ is ? >> structured, >> ask a linguist, not a five-year old. And one of the by-products of >> education may be increased sensitivity to syntax - which is one of ? >> the >> many reasons why linguists need to pay more attention to education. >> >> Best wishes, ?Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> On 09/09/2010 11:39, Daniel Everett wrote: >>> Dick, >>> >>> You raise an important issue here about >> methodology. I believe that intuitions are a fine way to generate >> hypotheses and even to test them - to a degree. But while it might ? >> not have >> been feasible for Huddleston, Pullum, and the other contributors to ? >> the >> Cambridge Grammar to conduct experiments on every point of the ? >> grammar, >> experiments could have only made the grammar better. The use of ? >> intuitions, >> corpora, and standard psycholinguistic experimentation (indeed, ? >> Standard >> Social Science Methodology) ?is vital for taking the field forward ? >> and for >> providing the best support for different analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new paper on this, showing ? >> serious >> shortcomings with intuitions as the sole source of evidence, in their >> paper: "The need for quantitative methods in syntax and semantics >> research".> >>> Carson Schutze and Wayne Cowart, among others, >> have also written convincingly on this.> >>> It is one reason that a team from Stanford, MIT >> (Brain and Cognitive Science), and researchers from Brazil are ? >> beginning a >> third round of experimental work among the Pirahas, since my own ? >> work on >> the syntax was, like almost every other field researcher's, based ? >> on native >> speaker intuitions and corpora.> >>> The discussion of methodologies reminds me of >> the initial reactions to Greenberg's work on classifying the ? >> languages of >> the Americas. His methods were strongly (and justifiably) criticized. >> However, I always thought that his methods were a great way of ? >> generating >> hypotheses, so long as they were ultimately put to the test of ? >> standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions.> >>> -- Dan >>> >>>> We linguists can add a further layer of >> explanation to the judgements, but some judgements do seem to be more >> reliable than others. And if we have to wait for psycholinguistic ? >> evidence >> for every detailed analysis we make, our whole discipline will ? >> immediately >> grind to a halt. Like it or not, native speaker judgements are what ? >> put us >> linguists ahead of the rest in handling fine detail. Imagine ? >> writing the >> Cambridge Grammar of the English Language (or the OED) without ? >> using native >> speaker judgements.>> >>>> Best wishes, ?Dick Hudson >>>> >>> >>> >> >> >> >> >> > Lise Menn ? ? ? ? ? ? ? ? ? ? ?Home Office: 303-444-4274 1625 Mariposa Ave ? ? ? Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of ?Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder -- Professor Emerita of Linguistics Department of Linguistics University of Wisconsin-Milwaukee Milwaukee, WI 53201-0413 USA From dcyr at yorku.ca Fri Sep 10 02:38:59 2010 From: dcyr at yorku.ca (Danielle E. Cyr) Date: Thu, 9 Sep 2010 22:38:59 -0400 Subject: analysis: unhappiness In-Reply-To: <960039811.375031.1284076517806.JavaMail.root@mail03.pantherlink.uwm.edu> Message-ID: We might also want to take into account 1) the historical time span extension of different people's grammatical knowledge and 2) the fact that grammaticalization occurs progressively over generations so that not all people of a same generation have the same grammatical analysis in mind when they think about morphological analysis. By 1) I mean that the more one knows about the history/evolution of a language's morphology, spread over the longer a period of time, the finest morphological analysis one will be able to make. For instance, when I present my students with the French adverb MAINTENANT 'now' at first they can't parse it into separate morphemes. After some minutes and a little bit of coaxing they come too see MAIN+TEMANT 'in hand' + 'holding', and finally MAIN+TEN+ANT 'in hand'+'hold'+ '-ing'. From that moment on, these students will integrate that precise morphological parsing. It will become part of their "internal historical grammar" and they will not really be able to go back to the "feeling" that MAINTENANT is only one morpheme. Now when I present them with the adverb AUJOURD'HUI 'today' and we go back in time to AD+ILLU(M)+DIURN(UM)+DE+HO(C)+DIE(M), the evolution of which spans over more than 2000 years, it totally transforms their inner historical grammar. They become different speakers of French from who they were before. By 2) I mean what is made totally explicit in Hopper and Traugott Grammaticalization (1993) and Marchello-Nizia (2006) Grammaticalisation et changement linguistique, Bruxelles, De Boeck). Speakers of a same language community do not all have the exact same grammar in mind simply because language change occurs constantly and progressively among different social groups, classes and generations. So it will always be impossible to get everyone to produce morphological parsing in the same exact way. Language, and grammar with it, are realities in constant flux. Conceiving that there is a unique stable grammar "out there" is "une vue de l'esprit". It helps us to think about the flux with a sensation of being on solid ground. Perhaps just like believing there is a God out there helps some of us to cope with impermanence. Cordialement, Danielle Cyr From haspelmath at eva.mpg.de Fri Sep 10 11:18:34 2010 From: haspelmath at eva.mpg.de (Martin Haspelmath) Date: Fri, 10 Sep 2010 13:18:34 +0200 Subject: analysis: phenomenological vs. cognitive In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. I suggested two terms for these kinds of patterns in a 2004 paper (reference below): ? phenomenological description (for what is "out there", or "grammars that are based simply on the form-meaning correspondences", in Edith Moravcsik's terms) ? cognitive description (for what a speaker has in their head) In the paper my main claim is that we don't really need cognitive description in order to explain the patterns of languages in functional terms (just as Darwin didn't need full descriptions of genomes to come up with functional explanations of the phenomenological properties of species). In generative linguistics, of course, one needs a unique "analysis" for each structure, because explanation and description/analysis are the same enterprise (just two different aspects of it), whereas in functional linguistics, description and explanation are separate. Thus, the fact that the best cognitive description is somewhat elusive doesn't matter to this approach. Martin Reference: Haspelmath, Martin. 2004. Does linguistic explanation presuppose linguistic description? Studies in Language 28. 554-579. (doi:10.1075/sl.28.3.06has) -- Martin Haspelmath (haspelmath at eva.mpg.de) Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 From A.Foolen at let.ru.nl Fri Sep 10 11:21:05 2010 From: A.Foolen at let.ru.nl (Ad Foolen) Date: Fri, 10 Sep 2010 13:21:05 +0200 Subject: analysis: unhappiness In-Reply-To: <1284086339.4c899a434ddd8@mymail.yorku.ca> Message-ID: When the historical dimension is taken into consideration, as Danielle Cyr proposes, it might be worthwhile to re-read Saussure's Cours, in particular his methodological reflections in the Appendices A, B, and C, following part 3 (in my 1968 Payot edition page 251-260). Saussure distinguishes between 'analyse subjective' and 'analyse objective'. The first is made by the native speaker, the second one by the linguist ('le grammairien'). The subjective analysis is the one that plays a role in language change via 'formations analogiques'. Analyses that come to the surface in psycholinguistic experiments as discussed in the present exchange did not play a role in Saussure's methodology. Two examples from Appendix A (Cours, p. 251): "L'analyse objective voit quatre sous-unit?s dans amabas (am-a-ba-s); les Latins coupaient ama-ba-s; il est m?me probable qu'ils regardaient -bas comme un tout flexionell oppos? au radical. Dans les mots fran?ais 'entier' (lat. in-teger 'intact'), 'enfant' (lat. infans 'qui ne parle pas') 'enceinte' (lat. in-cincta 'sans ceinture'), l'historien d?gagera un pr?fixe commun 'en-', identique au 'in-'privatif du latin; l'analyse subjective des sujets parlants l'ignore totalement." In Saussure's view, both perspectives have their value, but "en dernier resort celle [l'analyse] des sujets importe seule, car elle est fond? directement sur les faits de langue." (Cours, p. 252). Ad Foolen -------------------------------------------------- From: "Danielle E. Cyr" Sent: Friday, September 10, 2010 4:38 AM To: "Edith A Moravcsik" Cc: "Lise Menn" ; "Richard Hudson" ; ; "Funknet" Subject: Re: [FUNKNET] analysis: unhappiness > We might also want to take into account 1) the historical time span > extension of > different people's grammatical knowledge and 2) the fact that > grammaticalization > occurs progressively over generations so that not all people of a same > generation have the same grammatical analysis in mind when they think > about > morphological analysis. > > By 1) I mean that the more one knows about the history/evolution of a > language's > morphology, spread over the longer a period of time, the finest > morphological > analysis one will be able to make. For instance, when I present my > students > with the French adverb MAINTENANT 'now' at first they can't parse it into > separate morphemes. After some minutes and a little bit of coaxing they > come > too see > MAIN+TEMANT 'in hand' + 'holding', and finally MAIN+TEN+ANT 'in > hand'+'hold'+ > '-ing'. From that moment on, these students will integrate that precise > morphological parsing. It will become part of their "internal historical > grammar" and they will not really be able to go back to the "feeling" that > MAINTENANT is only one morpheme. > > Now when I present them with the adverb AUJOURD'HUI 'today' and we go back > in > time to AD+ILLU(M)+DIURN(UM)+DE+HO(C)+DIE(M), the evolution of which spans > over > more than 2000 years, it totally transforms their inner historical > grammar. They > become different speakers of French from who they were before. > > By 2) I mean what is made totally explicit in Hopper and Traugott > Grammaticalization (1993) and Marchello-Nizia (2006) Grammaticalisation et > changement linguistique, Bruxelles, De Boeck). Speakers of a same language > community do not all have the exact same grammar in mind simply because > language change occurs constantly and progressively among different social > groups, classes and generations. So it will always be impossible to get > everyone to produce morphological parsing in the same exact way. > > Language, and grammar with it, are realities in constant flux. Conceiving > that > there is a unique stable grammar "out there" is "une vue de l'esprit". It > helps > us to think about the flux with a sensation of being on solid ground. > Perhaps > just like believing there is a God out there helps some of us to cope with > impermanence. > > Cordialement, > Danielle Cyr > From egibson at MIT.EDU Fri Sep 10 13:03:20 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Fri, 10 Sep 2010 09:03:20 -0400 Subject: analysis: unhappiness Message-ID: Dear Dan, Dick: I would like to clarify some points that Dan Everett makes, in response to Dick Hudson. Ev Fedorenko and I have written a couple of papers recently (Gibson & Fedorenko, 2010, in press, see references and links below) on what we think are weak methodological standards in syntax and semantics research over the past many years. The issue that we address is the prevalent method in syntax and semantics research, which involves obtaining a judgment of the acceptability of a sentence / meaning pair, typically by just the author of the paper, sometimes with feedback from colleagues. As we address in our papers, this methodology does not allow proper testing of scientific hypotheses because of (a) the small number of experimental participants (typically one); (b) the small number of experimental stimuli (typically one); (c) cognitive biases on the part of the researcher and participants; and (d) the effect of the preceding context (e.g., other constructions the researcher may have been recently considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and several others cited in Gibson & Fedorenko, in press; for similar points, but with not as strong a conclusion as ours). Three issues need to be separated here: (1) the use of intuitive judgments as a dependent measure in a language experiment; (2) potential cognitive biases on the part of experimental subjects and experimenters in language experiments; and (3) the need for obtaining quantitative evidence, whatever the dependent measure might be. The paper that Ev and I wrote addresses the last two issues, but does not go into depth on the first issue (the use of intuitions as a dependent measure in language experiments). Regarding this issue, we don't think that there is anything wrong with gathering intuitive judgments as a dependent measure, as long as the task is clear to the experimental participants. In the longer paper (Gibson & Fedorenko, in press) we respond to some arguments that have been given in support of continuing to use the traditional non-quantitative method in syntax / semantics research. One recent defense of the traditional method comes from Phillips (2008), who argues that no harm has come from the non-quantitative approach in syntax research thus far. Phillips argues that there are no cases in the literature where an incorrect intuitive judgment has become the basis for a widely accepted generalization or an important theoretical claim. He therefore concludes that there is no reason to adopt more rigorous data collection standards. We challenge Philips? conclusion by presenting three cases from the literature where a faulty intuition has led to incorrect generalizations and mistaken theorizing, plausibly due to cognitive biases on the part of the researchers. A second argument that is sometimes presented for the continued use of the traditional non-quantitative method is that it would be too inefficient to evaluate every syntactic / semantic hypothesis or phenomenon quantitatively. For example, Culicover & Jackendoff (2010) make this argument explicitly in their response to Gibson & Fedorenko (2010): ?It would cripple linguistic investigation if it were required that all judgments of ambiguity and grammaticality be subject to statistically rigorous experiments on naive subjects, especially when investigating languages whose speakers are hard to access? (Culicover & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point earlier in the discussion here.) Whereas we agree that in circumstances where gathering data is difficult, some evidence is better than no evidence, we do not agree that research would be slowed with respect to languages where experimental participants are easy to access, such as English. In contrast, we think that the opposite is true: the field?s progress is probably slowed by not doing quantitative research. Suppose that a typical syntax / semantics paper that lacks quantitative evidence includes judgments for 50 or more sentences / meaning pairs, corresponding to 50 or more empirical claims. Even if most of the judgments from such a paper are correct or are on the right track, the problem is in knowing which judgments are correct. For example, suppose that 90% of the judgments from an arbitrary paper are correct (which is probably a high estimate). (Colin Phillips and some of his former students / postdocs have commented to us that, in their experience, quantitative acceptability judgment studies almost always validate the claim(s) in the literature. This is not our experience, however. Most experiments that we have run which attempt to test some syntactic / semantic hypothesis in the literature end up providing us with a pattern of data that had not been known before the experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in press; Patel et al., 2009; Scontras & Gibson, submitted).) This means that in a paper with 50 empirical claims 45/50 are correct. But which 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over two million different theories. By quantitatively evaluating the empirical claims, we reduce the uncertainty a great deal. To make progress, it is better to have theoretical claims supported by solid quantitative evidence, so that even if the interpretation of the data changes over time as new evidence becomes available ? as is often the case in any field of science ? the empirical pattern can be used as a basis for further theorizing. Furthermore, it is no longer expensive to run behavioral experiments, at least in English and other widely spoken languages. There now exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which can be used for collecting behavioral data over the internet quickly and inexpensively. The cost of using an interface like this is minimal, and the time that it takes for the results to be returned is short. For example, currently on Mechanical Turk, a survey of approximately 50 items will be answered by 50 or more participants within a couple of hours, at a cost of approximately $1 per participant. Thus a survey can be completed within a day, at a cost of less than $50. (The hard work of designing the experiment, and constructing controlled materials remains of course.) Sorry to be so verbose. But I think that these methodological points are very important. Best wishes, Ted Gibson Gibson, E. & Fedorenko, E. (In press). The need for quantitative methods in syntax and semantics research. Language and Cognitive Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko InPress LCP.pdf Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in linguistics research. Trends in Cognitive Science, 14, 233-234. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko 2010 TICS.pdf > Dick, > > You raise an important issue here about methodology. I believe that > intuitions are a fine way to generate hypotheses and even to test > them - to a degree. But while it might not have been feasible for > Huddleston, Pullum, and the other contributors to the Cambridge > Grammar to conduct experiments on every point of the grammar, > experiments could have only made the grammar better. The use of > intuitions, corpora, and standard psycholinguistic experimentation > (indeed, Standard Social Science Methodology) is vital for taking > the field forward and for providing the best support for different > analyses. Ted Gibson and Ev Fedorenko have written a very useful new > paper on this, showing serious shortcomings with intuitions as the > sole source of evidence, in their paper: "The need for quantitative > methods in syntax and semantics research". > > Carson Schutze and Wayne Cowart, among others, have also written > convincingly on this. > > It is one reason that a team from Stanford, MIT (Brain and Cognitive > Science), and researchers from Brazil are beginning a third round of > experimental work among the Pirahas, since my own work on the syntax > was, like almost every other field researcher's, based on native > speaker intuitions and corpora. > > The discussion of methodologies reminds me of the initial reactions > to Greenberg's work on classifying the languages of the Americas. > His methods were strongly (and justifiably) criticized. However, I > always thought that his methods were a great way of generating > hypotheses, so long as they were ultimately put to the test of > standard historical linguistics methods. And the same seems true for > use of native-speaker intuitions. > > -- Dan >> We linguists can add a further layer of explanation to the >> judgements, but some judgements do seem to be more reliable than >> others. And if we have to wait for psycholinguistic evidence for >> every detailed analysis we make, our whole discipline will >> immediately grind to a halt. Like it or not, native speaker >> judgements are what put us linguists ahead of the rest in handling >> fine detail. Imagine writing the Cambridge Grammar of the English >> Language (or the OED) without using native speaker judgements. >> >> Best wishes, Dick Hudson From amnfn at well.com Fri Sep 10 13:12:03 2010 From: amnfn at well.com (A. Katz) Date: Fri, 10 Sep 2010 06:12:03 -0700 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: Lise, I like very much the terminology that you used here: "what is out there in a language" versus the way it is "apprehended by a speaker." The problem is that too many of our colleagues don't agree that there is such a distinction-- or that language could ever possibly be out there! They think language exists only in the mind of speakers, and they look for uniformity of processing where none exists. That language is an abstraction and that each speaker has to crack the code individually and alone is not universally accepted among linguists. Best, --Aya On Thu, 9 Sep 2010, Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a given > time, we are talking about the patterns that are 'out there' in the language > and might possibly be apprehended (subconsciously) by a speaker, and when we > are talking about the patterns that a particular speaker actually does > apprehend, as indicated by experiments, from simple 'wug tests' up to brain > wave and eye-gaze studies. And for distinguishing among the degrees of > pattern apprehension that a person may have, from vague preferences > detectable in reaction times or other behavior all the way up through clear > metalinguistic insights. Dick Hudson's note reminding us of the Gleitman and > Gleitman study is right on target. > > Since we don't in fact have such an agreed-on terminology, we have to > be quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'. We know, but tend to > forget - and tend to forget to tell our students! - that it's an empirical > question as to whether the formal simplicity and coherence of description of > forms 'out there' (e.g. lovely abstract morphophonemics) is any kind of > approximation to the way knowledge of the same forms is organized in a > particular person's head. If we remember that a very large proportion of > what we know about our language is 'out there' when we are infants and has to > be internalized through experience with the language (even if you believe in > innate 'core language'), the variation in internal knowledge from one person > to another is more understandable. > We especially need to consider (and try to test) the > possibility that since > the brain can make multiple cross-connections, multiple patterns are involved > simultaneously in morphological and syntactic analyses. I suggest that that's > the case with 'unhappiness' - and the linguistic analyses that I know about > are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> >> Two comments. >> >> First (elaborating perhaps on Dick Hudson's comment), I think there is an >> important distinction between low-level linguistic intuitions (like whether >> a >> word or sentence is well-formed or what it means) and higher-level >> intuitions >> (like what the structure of a word or sentence is). One can take the >> position >> that we need to account for the former (while recognizing that they are not >> always reliable) but not the latter. >> >> Second, the tension here is not only between evidence from speaker >> intuitions >> versus evidence from psycholinguistic experiments. There is also a tension >> between deciding on the correct analysis on the basis of a priori >> simplicity >> arguments versus deciding on the correct analysis on the basis of >> psycholinguistic evidence (see Derwing 1973). The bracketing paradox that >> Dan >> referred to that arises with the word (semantics argues for >> [[un+happi] + er], morphology and phonology argues for [un + [happi+er]] >> (the >> comparative suffix can only be attached to adjectives containing one or two >> syllables) is only a paradox if one assumes that speakers adopt the >> simplest >> analysis. For example, if speakers adopt a more complex rule for either of >> these >> (e.g. perhaps the rule for attaching -er can apply exceptionally to >> trisyllabic >> words beginning with un-), then the bracketing paradox disappears. >> >> Matthew >> >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>> conscious judgements are only one kind of evidence that we need to take >>> into account. I admire Carson Schutze's work (which I reviewed in fact), >>> and of course I've been aware of complaints about judgements by people >>> like Labov for decades. >>> >>> But you're missing my main point, which is that all judgements aren't >>> equally reliable. If you want to know how /unhappiness/ is structured, >>> ask a linguist, not a five-year old. And one of the by-products of >>> education may be increased sensitivity to syntax - which is one of the >>> many reasons why linguists need to pay more attention to education. >>> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> On 09/09/2010 11:39, Daniel Everett wrote: >>>> Dick, >>>> >>>> You raise an important issue here about >>> methodology. I believe that intuitions are a fine way to generate >>> hypotheses and even to test them - to a degree. But while it might not >>> have >>> been feasible for Huddleston, Pullum, and the other contributors to the >>> Cambridge Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, >>> corpora, and standard psycholinguistic experimentation (indeed, Standard >>> Social Science Methodology) is vital for taking the field forward and for >>> providing the best support for different analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new paper on this, showing serious >>> shortcomings with intuitions as the sole source of evidence, in their >>> paper: "The need for quantitative methods in syntax and semantics >>> research".> >>>> Carson Schutze and Wayne Cowart, among others, >>> have also written convincingly on this.> >>>> It is one reason that a team from Stanford, MIT >>> (Brain and Cognitive Science), and researchers from Brazil are beginning a >>> third round of experimental work among the Pirahas, since my own work on >>> the syntax was, like almost every other field researcher's, based on >>> native >>> speaker intuitions and corpora.> >>>> The discussion of methodologies reminds me of >>> the initial reactions to Greenberg's work on classifying the languages of >>> the Americas. His methods were strongly (and justifiably) criticized. >>> However, I always thought that his methods were a great way of generating >>> hypotheses, so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions.> >>>> -- Dan >>>> >>>>> We linguists can add a further layer of >>> explanation to the judgements, but some judgements do seem to be more >>> reliable than others. And if we have to wait for psycholinguistic evidence >>> for every detailed analysis we make, our whole discipline will immediately >>> grind to a halt. Like it or not, native speaker judgements are what put us >>> linguists ahead of the rest in handling fine detail. Imagine writing the >>> Cambridge Grammar of the English Language (or the OED) without using >>> native >>> speaker judgements.>> >>>>> Best wishes, Dick Hudson >>>>> >>>> >>>> >>> >>> >>> >>> >>> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From tpayne at uoregon.edu Fri Sep 10 14:16:56 2010 From: tpayne at uoregon.edu (Thomas E. Payne) Date: Fri, 10 Sep 2010 07:16:56 -0700 Subject: "Relative clauses" with no relativized element Message-ID: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne From eitkonen at utu.fi Fri Sep 10 14:01:38 2010 From: eitkonen at utu.fi (Esa Itkonen) Date: Fri, 10 Sep 2010 17:01:38 +0300 Subject: analysis: phenomenological vs. cognitive In-Reply-To: <4C8A140A.8000804@eva.mpg.de> Message-ID: Dear Friends: These questions were given what I consider the definitive answer in the context of the 'psychological reality' debate in the mid and late 70's. But there is no harm in reinventing (or renaming) the wheel. Esa Homepage: http://users.utu.fi/eitkonen ----- Original Message ----- From: Martin Haspelmath Date: Friday, September 10, 2010 2:19 pm Subject: Re: [FUNKNET] analysis: phenomenological vs. cognitive To: Funknet > Lise Menn wrote: > > I wish we had better terminology for keeping track of whether, at a > > > given time, we are talking about the patterns that are 'out there' > in > > the language and might possibly be apprehended (subconsciously) by > a > > speaker, and when we are talking about the patterns that a > particular > > speaker actually does apprehend, as indicated by experiments, from > > > simple 'wug tests' up to brain wave and eye-gaze studies. > I suggested two terms for these kinds of patterns in a 2004 paper > (reference below): > > ? phenomenological description (for what is "out there", or "grammars > > that are based simply on the form-meaning correspondences", in Edith > > Moravcsik's terms) > > ? cognitive description (for what a speaker has in their head) > > In the paper my main claim is that we don't really need cognitive > description in order to explain the patterns of languages in > functional > terms (just as Darwin didn't need full descriptions of genomes to > come > up with functional explanations of the phenomenological properties of > > species). > > In generative linguistics, of course, one needs a unique "analysis" > for > each structure, because explanation and description/analysis are the > > same enterprise (just two different aspects of it), whereas in > functional linguistics, description and explanation are separate. > Thus, > the fact that the best cognitive description is somewhat elusive > doesn't > matter to this approach. > > Martin > > > Reference: > Haspelmath, Martin. 2004. Does linguistic explanation presuppose > linguistic description? Studies in Language 28. 554-579. > (doi:10.1075/sl.28.3.06has) > > -- > Martin Haspelmath (haspelmath at eva.mpg.de) > Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz > 6 > D-04103 Leipzig > Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 > > > > > > From jscheibm at odu.edu Fri Sep 10 15:15:22 2010 From: jscheibm at odu.edu (Scheibman, Joanne) Date: Fri, 10 Sep 2010 11:15:22 -0400 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Hi Tom, Quirk et al. analyze these as appositive clauses (in 17.26 and elsewhere in that chapter). The head of the NP is a general abstract noun, e.g. "fact", "idea", "claim", "belief", (or "condition"), and "that" has no role in the subordinate clause, as you mentioned. Joanne on 9/10/10 10:16 AM, Thomas E. Payne wrote: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne -- BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach CanIt if this mail (ID 358169235) is spam: Spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=s Not spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=n Forget vote: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=f ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS From rchen at csusb.edu Fri Sep 10 15:42:43 2010 From: rchen at csusb.edu (Rong Chen) Date: Fri, 10 Sep 2010 23:42:43 +0800 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: To add to Joanne's comments: There are basically three ways to distinguish an appositive clause (AC) from a relative clause (RC). 1) An AC can only be led by *that* while an RC can be led by other pronouns. 2) The AC and the noun it modifies display an equative relationship--one can say X (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't (except, perhaps, when the relative clause is sentential). 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun is always part of the clause in an RC. Rong Chen -----Original Message----- From: funknet-bounces at mailman.rice.edu [mailto:funknet-bounces at mailman.rice.edu] On Behalf Of Scheibman, Joanne Sent: Friday, September 10, 2010 11:15 PM To: Thomas E. Payne; FUNKNET Subject: Re: [FUNKNET] "Relative clauses" with no relativized element Hi Tom, Quirk et al. analyze these as appositive clauses (in 17.26 and elsewhere in that chapter). The head of the NP is a general abstract noun, e.g. "fact", "idea", "claim", "belief", (or "condition"), and "that" has no role in the subordinate clause, as you mentioned. Joanne on 9/10/10 10:16 AM, Thomas E. Payne wrote: Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne -- BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach CanIt if this mail (ID 358169235) is spam: Spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=s Not spam: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=n Forget vote: https://www.spamtrap.odu.edu/b.php?i=358169235&m=d1f8830c0a73&t=20100910&c=f ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS From carol.moder at okstate.edu Fri Sep 10 15:52:37 2010 From: carol.moder at okstate.edu (Moder, Carol) Date: Fri, 10 Sep 2010 10:52:37 -0500 Subject: Job advertisement Message-ID: Dear Funknetters: Please distribute the following position description to interested colleagues and doctoral students completing their degrees. Our linguistics program has a cognitive-functional orientation and we seek a new colleague with this perspective. Carol Moder Professor & Head Department of English 205 Morrill Hall Stillwater, OK 74078 (405)744-6140 Assistant Professor -Linguistics with specialization in phonetics or phonology Department of English Oklahoma State University Tenure-track position. Ph.D. in Linguistics, or related area with a specialization in phonetics or phonology. Ability to teach a broad range of undergraduate and graduate linguistics courses in support of specializations in linguistics and Teaching English as a Second Language. 3-2 teaching load beginning August 2011. Strong research agenda and demonstrated excellence in teaching required. Salary competitive and commensurate with experience. OSU offers the BA, MA, and the PhD in English with an M.A. option in Teaching English as a Second Language and M.A. and Ph.D. specializations in Linguistics.. For further information on the department see our webpage at http://english.okstate.edu. To ensure full consideration, applications must be received by November 3, 2010 However, we will continue to accept and consider applications until the position has been filled. Send letter of application, cv, writing sample, and dossier, including three letters of recommendation and transcript to Carol Moder, Head, English Department, Oklahoma State University, 205 Morrill Hall, Stillwater, OK 74078-4069. Filling of this position is contingent upon availability of funding. Oklahoma State University is an AA/EEO/E-Verify employer committed to diversity. OSU-Stillwater is a tobacco-free campus. From Arie.Verhagen at hum.LeidenUniv.nl Fri Sep 10 16:21:10 2010 From: Arie.Verhagen at hum.LeidenUniv.nl (Arie Verhagen) Date: Fri, 10 Sep 2010 18:21:10 +0200 Subject: "Relative clauses" with no relativized ele Message-ID: And as another addition: the clauses that can only be introduced by *that* (with no role to play in the subordinate clause) may be seen as (subtypes of) complement clauses, expressing a proposition with the noun functioning as Complement Taking Predicate (CTP), expressing a propostional attitude, epistemic/evaluative stance, etc. (following analyses by Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. constructions like "The claim is that X" (traditionally analysed as subject clauses), "I claim that X", "I put forward the claim that X", in which the relationship between the verb or noun and the that-clause is comparable to the one in "The claim that X". --Arie Verhagen ---------------- Message from Rong Chen 10 Sep 2010, 23:42 Subject: Re: [FUNKNET] "Relative clauses" with no relativi > To add to Joanne's comments: > > There are basically three ways to distinguish an appositive clause > (AC) from a relative clause (RC). > > 1) An AC can only be led by *that* while an RC can be led by other > pronouns. > > 2) The AC and the noun it modifies display an equative relationship--one can say X > (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't > (except, perhaps, when the relative clause is sentential). > 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun > is always part of the clause in an RC. > > Rong Chen > From shinja_hwang at gial.edu Fri Sep 10 17:03:47 2010 From: shinja_hwang at gial.edu (Hwang, Shinja) Date: Fri, 10 Sep 2010 17:03:47 +0000 Subject: "Relative clauses" with no relativized ele In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: I also prefer to refer to them as complement clauses. They differ from the relative clauses since the embedded clause is complete in itself, not lacking anything. These clauses are in an appositive relation with the head noun like 'fact, claim, evidence' and in English they can be linked by 'that', but not by 'which', unlike the relative clause. Compare the examples below: The evidence [that(*which) he stole the book] is clearly shown to us. The evidence [that/which they used 0 in the trial] was very powerful. In some languages there is no syntactic difference (such as 'that' vs. 'which' in English), and the criteria may need to resort to some other clues such as whether there is a missing element in the embedded clause which is the head noun. Shin Ja Hwang -----Original Message----- From: funknet-bounces at mailman.rice.edu [mailto:funknet-bounces at mailman.rice.edu] On Behalf Of Arie Verhagen Sent: Friday, September 10, 2010 11:21 AM To: 'FUNKNET' Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele And as another addition: the clauses that can only be introduced by *that* (with no role to play in the subordinate clause) may be seen as (subtypes of) complement clauses, expressing a proposition with the noun functioning as Complement Taking Predicate (CTP), expressing a propostional attitude, epistemic/evaluative stance, etc. (following analyses by Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. constructions like "The claim is that X" (traditionally analysed as subject clauses), "I claim that X", "I put forward the claim that X", in which the relationship between the verb or noun and the that-clause is comparable to the one in "The claim that X". --Arie Verhagen ---------------- Message from Rong Chen 10 Sep 2010, 23:42 Subject: Re: [FUNKNET] "Relative clauses" with no relativi > To add to Joanne's comments: > > There are basically three ways to distinguish an appositive clause > (AC) from a relative clause (RC). > > 1) An AC can only be led by *that* while an RC can be led by other > pronouns. > > 2) The AC and the noun it modifies display an equative relationship--one can say X > (denoted by the noun) is Y (presented by the appositive)--while an RC often doesn't > (except, perhaps, when the relative clause is sentential). > 3)--which Tom noted--*that* is not part of the clause in an AC; but a relative pronoun > is always part of the clause in an RC. > > Rong Chen > From kemmer at rice.edu Fri Sep 10 17:16:09 2010 From: kemmer at rice.edu (Suzanne Kemmer) Date: Fri, 10 Sep 2010 12:16:09 -0500 Subject: "Relative clauses" with no relativized element In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: In Generative Syntax these clauses were viewed as complement clauses with an NP head, distinct from relative clauses but having some parallels with them. I think it was Joan Bresnan that brought out the parallels and distinctions, maybe in her doctoral dissertation . As I recall (but my remembrance may be faulty), Bresnan named the THAT element a COMP for complementizer. The term 'appositive' isn't very good because in traditional grammar that is reserved for an 'UNrestrictive' relation of a noun and its complement--an incidental description of a head N's referent rather than a specification of which referent ("the tree, a live oak, survived another 100 years or so"). In Cognitive Grammar nouns like claim, statement, idea, realization, belief etc. are in almost all cases nominalizations of 'viewing predicates' (verbs like claim, believe, etc.) that introduce on-stage predications 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the viewing predicates are space builders in Fauconnier's mental spaces terminology) For the nominalizations of these predicates, the semantics of the nouns intrinsically has an "e-site" or elaboration site that allows for spelling out the content of the viewed predicate in the form of a complement clause. The e-site inherent to the semantics of the nouns is parallel to the e-site inherent to the semantics of the corresponding verbs. There are a few cases I can think of of nouns that have 'viewing predicate' e-sites but don't have corresponding verbs . For example the noun _view_ "The view that global climate change is anthropogenic is widely held by scientists" ( ' X views that (proposition)' is not possible, only 'X views Y as ...' , with a restriction to equative or descriptive propositions). Also _idea_----the verb has to be changed to something like 'believe' to make a corresponding full predicate. I view (!) these nouns as semantically parallel in interesting ways to picture nouns. The conceptualizer (viewer) in both cases can designate the noun in a possessive phrase, but after that the syntax diverges. --Suzanne On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > And as another addition: the clauses that can only be introduced by > *that* (with no role to > play in the subordinate clause) may be seen as (subtypes of) > complement clauses, > expressing a proposition with the noun functioning as Complement > Taking Predicate (CTP), > expressing a propostional attitude, epistemic/evaluative stance, > etc. (following analyses by > Thompson, Diessel, Langacker, myself, and others), i.e. not > relatives. Cf. constructions like > "The claim is that X" (traditionally analysed as subject clauses), > "I claim that X", "I put forward > the claim that X", in which the relationship between the verb or > noun and the that-clause is > comparable to the one in "The claim that X". > > --Arie Verhagen > > ---------------- > Message from Rong Chen > 10 Sep 2010, 23:42 > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > >> To add to Joanne's comments: >> >> There are basically three ways to distinguish an appositive clause >> (AC) from a relative clause (RC). >> >> 1) An AC can only be led by *that* while an RC can be led by other >> pronouns. >> >> 2) The AC and the noun it modifies display an equative >> relationship--one can say X >> (denoted by the noun) is Y (presented by the appositive)--while an >> RC often doesn't >> (except, perhaps, when the relative clause is sentential). > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but >> a relative pronoun >> is always part of the clause in an RC. >> >> Rong Chen >> > > From kuzar at research.haifa.ac.il Fri Sep 10 17:42:24 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Fri, 10 Sep 2010 20:42:24 +0300 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: A thorough discussion of the head nouns and their relation with their complement clauses may be found in Hans-Joerg Schmid's book on Shell Nouns (this is his term for the head nouns). Ron Kuzar --------- On Fri, Sep 10, 2010 at 8:16 PM, Suzanne Kemmer wrote: > In Generative Syntax these clauses were viewed as complement clauses with > an NP head, distinct > from relative clauses but having some parallels with them. I think it was > Joan Bresnan that > brought out the parallels and distinctions, maybe in her doctoral > dissertation . As I recall (but > my remembrance may be faulty), Bresnan named the > THAT element a COMP for complementizer. > > The term 'appositive' isn't very good because in traditional grammar > that is reserved for an 'UNrestrictive' relation of a noun and its > complement--an incidental description of > a head N's referent rather than a specification of which referent ("the > tree, a live oak, survived another 100 years or so"). > > In Cognitive Grammar nouns like claim, statement, idea, realization, > belief etc. are in almost all cases nominalizations of 'viewing predicates' > (verbs like claim, believe, etc.) that introduce on-stage predications > 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the > viewing > predicates are space builders in Fauconnier's mental spaces terminology) > > For the nominalizations of these predicates, the semantics of the nouns > intrinsically has an "e-site" or elaboration site > that allows for spelling out the content of the viewed predicate in the > form of a complement clause. The e-site > inherent to the semantics of the nouns is parallel to the e-site inherent > to the semantics of the corresponding verbs. > > There are a few cases I can think of of nouns that have 'viewing > predicate' e-sites but don't have corresponding verbs . > For example > the noun _view_ "The view that global climate change is anthropogenic is > widely held by scientists" > ( ' X views that (proposition)' is not possible, only 'X views Y as ...' , > with a restriction to equative or descriptive propositions). > Also _idea_----the verb has to be changed to something like 'believe' to > make a corresponding full predicate. > > I view (!) these nouns as semantically parallel in interesting ways to > picture nouns. The conceptualizer (viewer) in > both cases can designate the noun in a possessive phrase, but after that > the syntax diverges. > > --Suzanne > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > And as another addition: the clauses that can only be introduced by *that* >> (with no role to >> play in the subordinate clause) may be seen as (subtypes of) complement >> clauses, >> expressing a proposition with the noun functioning as Complement Taking >> Predicate (CTP), >> expressing a propostional attitude, epistemic/evaluative stance, etc. >> (following analyses by >> Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. >> constructions like >> "The claim is that X" (traditionally analysed as subject clauses), "I >> claim that X", "I put forward >> the claim that X", in which the relationship between the verb or noun and >> the that-clause is >> comparable to the one in "The claim that X". >> >> --Arie Verhagen >> >> ---------------- >> Message from Rong Chen >> 10 Sep 2010, 23:42 >> >> Subject: Re: [FUNKNET] "Relative clauses" with no relativi >> >> To add to Joanne's comments: >>> >>> There are basically three ways to distinguish an appositive clause >>> (AC) from a relative clause (RC). >>> >>> 1) An AC can only be led by *that* while an RC can be led by other >>> pronouns. >>> >>> 2) The AC and the noun it modifies display an equative relationship--one >>> can say X >>> (denoted by the noun) is Y (presented by the appositive)--while an RC >>> often doesn't >>> (except, perhaps, when the relative clause is sentential). >>> >> >> 3)--which Tom noted--*that* is not part of the clause in an AC; but a >>> relative pronoun >>> is always part of the clause in an RC. >>> >>> Rong Chen >>> >>> >> >> > -- =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From tgivon at uoregon.edu Fri Sep 10 17:43:34 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Fri, 10 Sep 2010 11:43:34 -0600 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this construction described and analyzed as the product of nominalization of clauses with verbs that take verbal complements ('know', 'think', 'say' etc. The preceding section (6.6.3. "Complex NP's arising through nominalization", p. 287) deals more generally with nominalizations. The term "noun complements" was used in syntax classes at UCLA in the mid 1960s, so certainly Joan Bresnan did not invent it. Best, TG =========== Thomas E. Payne wrote: > Can anyone help me name the following structure in English, and maybe point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > > From giuliana.fiorentino at unimol.it Fri Sep 10 17:53:15 2010 From: giuliana.fiorentino at unimol.it (Giuliana Fiorentino) Date: Fri, 10 Sep 2010 19:53:15 +0200 Subject: "Relative clauses" with no relativized element Message-ID: Hi Tom, clauses like: The importance of being Earnest the fact of being late the fact that you are late the idea that world is round etcetera are not relative clauses but can be considered among syntactic strategies in order to nominalise events after a generic noun (working as a classifier for nominalised events). Giuliana ----- Original Message ----- From: Thomas E. Payne To: FUNKNET Sent: Friday, September 10, 2010 4:16 PM Subject: [FUNKNET] "Relative clauses" with no relativized element Can anyone help me name the following structure in English, and maybe point me to some references? I do not find reference to this in the Cambridge Grammar of the English Language or any other of my English grammar books. But then, maybe I just don't know where to look. Here are two examples from a play: His protestations of devotion in the trial scene are, in our opinion, genuine, as is his confession [that his affair with the Countess is platonic]. The bracketed clause seems to modify "confession", though there is no position for a confession in the clause itself. . . . forced hither with an impious black design [to have my innocence and youth become the sacrifice of brutal violence]. Here the bracketed non-finite clause seems to modify "design." These are not all that rare. I'm reminded of examples like: "The claim [that my client is a murderer] is totally false." Are these relative clauses? If so what kind? Thanks for any help. Tom Payne From eitan.eg at gmail.com Fri Sep 10 17:54:06 2010 From: eitan.eg at gmail.com (E.G.) Date: Fri, 10 Sep 2010 20:54:06 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: <4C8A5AF6.5209.210287F@Arie.Verhagen.hum.LeidenUniv.nl> Message-ID: Hi all, I'd agree with Arie Verhagen. But there's a way that cross-linguistic comparison can help what seems to be a purely theoretical question based on a single language. The problem here is that English uses the same element to mark regular relatives and these "appositional" relatives. But if at least one language encodes them by different means, then there's at least a good case for seeing them as distinct functions. It's basically the same principle that's used to decide whether to put a meaning on a semantic map. So here are two languages that I know that encode them differently. In Modern Hebrew, these clauses can be encoded as a dedicated complement clause (ki), which differs from the relative clause marker (Se-), e.g. ha-hoda'a Se-kibalnu the-announcment rel-we_got "The announcement that we got." ha-hoda'a ki hitbatel ha-mifgaS the-message CMP was_cancelled the-meeting "The announcement that the meeting was cancelled." In Coptic, these clauses are marked by ce-, which marks complement clauses, *inter alia*, but not relative clauses: ph-mewi ce- (complement clause) 'the-thought that (we are angry)' ph-mewi ete- (relative clause) 'the thought that (we used to think)' This seems to be a pretty clear indication that these are complement clauses rather than relatives. Even if one doesn't like the notion of nouns taking complement clauses (and why not? nominalizations in some languages can take accusative modifiers as well as genitives), it still probably isn't incidental that the nominalizations are of verbs that take complement clauses when finite. As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth looking at. Best, Eitan On 10 September 2010 19:21, Arie Verhagen wrote: > And as another addition: the clauses that can only be introduced by *that* > (with no role to > play in the subordinate clause) may be seen as (subtypes of) complement > clauses, > expressing a proposition with the noun functioning as Complement Taking > Predicate (CTP), > expressing a propostional attitude, epistemic/evaluative stance, etc. > (following analyses by > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. Cf. > constructions like > "The claim is that X" (traditionally analysed as subject clauses), "I claim > that X", "I put forward > the claim that X", in which the relationship between the verb or noun and > the that-clause is > comparable to the one in "The claim that X". > > --Arie Verhagen > > ---------------- > Message from Rong Chen > 10 Sep 2010, 23:42 > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > To add to Joanne's comments: > > > > There are basically three ways to distinguish an appositive clause > > (AC) from a relative clause (RC). > > > > 1) An AC can only be led by *that* while an RC can be led by other > > pronouns. > > > > 2) The AC and the noun it modifies display an equative relationship--one > can say X > > (denoted by the noun) is Y (presented by the appositive)--while an RC > often doesn't > > (except, perhaps, when the relative clause is sentential). > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > relative pronoun > > is always part of the clause in an RC. > > > > Rong Chen > > > > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From kemmer at rice.edu Fri Sep 10 17:55:30 2010 From: kemmer at rice.edu (Suzanne Kemmer) Date: Fri, 10 Sep 2010 12:55:30 -0500 Subject: "Relative clauses" with no relativized element In-Reply-To: <4C8A6E46.3040301@uoregon.edu> Message-ID: Talmy, yes Joan B did not invent noun complements nor the term for them; but my recollection, in passing, was that she named 'that' in such structures as "complementizer" (and I also recall that she referred to 'that' relativizers with the same term, while recognizing other differences between the two structures.) I may be wrong on that, but it's a different recollection, claim, or whatever, than the one you refer to. Not all the head nouns are nominalizations, but most are. S. On Sep 10, 2010, at 12:43 PM, Tom Givon wrote: > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this construction described and analyzed as the product of nominalization of clauses with verbs that take verbal complements ('know', 'think', 'say' etc. The preceding section (6.6.3. "Complex NP's arising through nominalization", p. 287) deals more generally with nominalizations. The term "noun complements" was used in syntax classes at UCLA in the mid 1960s, so certainly Joan Bresnan did not invent it. Best, TG > > =========== > > Thomas E. Payne wrote: >> Can anyone help me name the following structure in English, and maybe point >> me to some references? I do not find reference to this in the Cambridge >> Grammar of the English Language or any other of my English grammar books. >> But then, maybe I just don't know where to look. >> >> Here are two examples from a play: >> >> His protestations of devotion in the trial scene are, in our opinion, >> genuine, as is his confession [that his affair with the Countess is >> platonic]. >> >> The bracketed clause seems to modify "confession", though there is no >> position for a confession in the clause itself. >> >> . . . forced hither with an impious black design [to have my innocence and >> youth become the sacrifice of brutal violence]. >> >> Here the bracketed non-finite clause seems to modify "design." >> >> These are not all that rare. I'm reminded of examples like: >> >> "The claim [that my client is a murderer] is totally false." >> >> Are these relative clauses? If so what kind? Thanks for any help. >> >> Tom Payne >> >> > > From eitan.eg at gmail.com Fri Sep 10 17:56:23 2010 From: eitan.eg at gmail.com (E.G.) Date: Fri, 10 Sep 2010 20:56:23 +0300 Subject: "Relative clauses" with no relativized element In-Reply-To: <036386D3933D4CB99389FFB2576ADBE8@giuliana> Message-ID: Jespersen and his nexus-substantives should be mentioned (Philosophy of Grammar, 1924). Also in his MEG and Analytic Syntax one could find interesting discussions. Eitan On 10 September 2010 20:53, Giuliana Fiorentino < giuliana.fiorentino at unimol.it> wrote: > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From dick at ling.ucl.ac.uk Fri Sep 10 17:59:12 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Fri, 10 Sep 2010 18:59:12 +0100 Subject: analysis: unhappiness In-Reply-To: <676152F6-8654-4BCD-BC30-B39E807C07B7@mit.edu> Message-ID: Dear Ted, Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 14:03, Ted Gibson wrote: > Dear Dan, Dick: > > I would like to clarify some points that Dan Everett makes, in > response to Dick Hudson. > > Ev Fedorenko and I have written a couple of papers recently (Gibson & > Fedorenko, 2010, in press, see references and links below) on what we > think are weak methodological standards in syntax and semantics > research over the past many years. The issue that we address is the > prevalent method in syntax and semantics research, which involves > obtaining a judgment of the acceptability of a sentence / meaning > pair, typically by just the author of the paper, sometimes with > feedback from colleagues. As we address in our papers, this > methodology does not allow proper testing of scientific hypotheses > because of (a) the small number of experimental participants > (typically one); (b) the small number of experimental stimuli > (typically one); (c) cognitive biases on the part of the researcher > and participants; and (d) the effect of the preceding context (e.g., > other constructions the researcher may have been recently > considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > several others cited in Gibson & Fedorenko, in press; for similar > points, but with not as strong a conclusion as ours). > > Three issues need to be separated here: (1) the use of intuitive > judgments as a dependent measure in a language experiment; (2) > potential cognitive biases on the part of experimental subjects and > experimenters in language experiments; and (3) the need for obtaining > quantitative evidence, whatever the dependent measure might be. The > paper that Ev and I wrote addresses the last two issues, but does not > go into depth on the first issue (the use of intuitions as a dependent > measure in language experiments). Regarding this issue, we don't think > that there is anything wrong with gathering intuitive judgments as a > dependent measure, as long as the task is clear to the experimental > participants. > > In the longer paper (Gibson & Fedorenko, in press) we respond to some > arguments that have been given in support of continuing to use the > traditional non-quantitative method in syntax / semantics research. > One recent defense of the traditional method comes from Phillips > (2008), who argues that no harm has come from the non-quantitative > approach in syntax research thus far. Phillips argues that there are > no cases in the literature where an incorrect intuitive judgment has > become the basis for a widely accepted generalization or an important > theoretical claim. He therefore concludes that there is no reason to > adopt more rigorous data collection standards. We challenge Philips? > conclusion by presenting three cases from the literature where a > faulty intuition has led to incorrect generalizations and mistaken > theorizing, plausibly due to cognitive biases on the part of the > researchers. > > A second argument that is sometimes presented for the continued use of > the traditional non-quantitative method is that it would be too > inefficient to evaluate every syntactic / semantic hypothesis or > phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > make this argument explicitly in their response to Gibson & Fedorenko > (2010): ?It would cripple linguistic investigation if it were required > that all judgments of ambiguity and grammaticality be subject to > statistically rigorous experiments on naive subjects, especially when > investigating languages whose speakers are hard to access? (Culicover > & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > earlier in the discussion here.) Whereas we agree that in > circumstances where gathering data is difficult, some evidence is > better than no evidence, we do not agree that research would be slowed > with respect to languages where experimental participants are easy to > access, such as English. In contrast, we think that the opposite is > true: the field?s progress is probably slowed by not doing > quantitative research. > Suppose that a typical syntax / semantics paper that lacks > quantitative evidence includes judgments for 50 or more sentences / > meaning pairs, corresponding to 50 or more empirical claims. Even if > most of the judgments from such a paper are correct or are on the > right track, the problem is in knowing which judgments are correct. > For example, suppose that 90% of the judgments from an arbitrary paper > are correct (which is probably a high estimate). (Colin Phillips and > some of his former students / postdocs have commented to us that, in > their experience, quantitative acceptability judgment studies almost > always validate the claim(s) in the literature. This is not our > experience, however. Most experiments that we have run which attempt > to test some syntactic / semantic hypothesis in the literature end up > providing us with a pattern of data that had not been known before the > experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > that in a paper with 50 empirical claims 45/50 are correct. But which > 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > two million different theories. By quantitatively evaluating the > empirical claims, we reduce the uncertainty a great deal. To make > progress, it is better to have theoretical claims supported by solid > quantitative evidence, so that even if the interpretation of the data > changes over time as new evidence becomes available ? as is often the > case in any field of science ? the empirical pattern can be used as a > basis for further theorizing. > > Furthermore, it is no longer expensive to run behavioral experiments, > at least in English and other widely spoken languages. There now > exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > can be used for collecting behavioral data over the internet quickly > and inexpensively. The cost of using an interface like this is > minimal, and the time that it takes for the results to be returned is > short. For example, currently on Mechanical Turk, a survey of > approximately 50 items will be answered by 50 or more participants > within a couple of hours, at a cost of approximately $1 per > participant. Thus a survey can be completed within a day, at a cost of > less than $50. (The hard work of designing the experiment, and > constructing controlled materials remains of course.) > > Sorry to be so verbose. But I think that these methodological points > are very important. > > Best wishes, > > Ted Gibson > > Gibson, E. & Fedorenko, E. (In press). The need for quantitative > methods in syntax and semantics research. Language and Cognitive > Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > & Fedorenko InPress LCP.pdf > > Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > linguistics research. Trends in Cognitive Science, 14, 233-234. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > 2010 TICS.pdf > > > > >> Dick, >> >> You raise an important issue here about methodology. I believe that >> intuitions are a fine way to generate hypotheses and even to test >> them - to a degree. But while it might not have been feasible for >> Huddleston, Pullum, and the other contributors to the Cambridge >> Grammar to conduct experiments on every point of the grammar, >> experiments could have only made the grammar better. The use of >> intuitions, corpora, and standard psycholinguistic experimentation >> (indeed, Standard Social Science Methodology) is vital for taking the >> field forward and for providing the best support for different >> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >> paper on this, showing serious shortcomings with intuitions as the >> sole source of evidence, in their paper: "The need for quantitative >> methods in syntax and semantics research". >> >> Carson Schutze and Wayne Cowart, among others, have also written >> convincingly on this. >> >> It is one reason that a team from Stanford, MIT (Brain and Cognitive >> Science), and researchers from Brazil are beginning a third round of >> experimental work among the Pirahas, since my own work on the syntax >> was, like almost every other field researcher's, based on native >> speaker intuitions and corpora. >> >> The discussion of methodologies reminds me of the initial reactions >> to Greenberg's work on classifying the languages of the Americas. His >> methods were strongly (and justifiably) criticized. However, I always >> thought that his methods were a great way of generating hypotheses, >> so long as they were ultimately put to the test of standard >> historical linguistics methods. And the same seems true for use of >> native-speaker intuitions. >> >> -- Dan > > > >>> We linguists can add a further layer of explanation to the >>> judgements, but some judgements do seem to be more reliable than >>> others. And if we have to wait for psycholinguistic evidence for >>> every detailed analysis we make, our whole discipline will >>> immediately grind to a halt. Like it or not, native speaker >>> judgements are what put us linguists ahead of the rest in handling >>> fine detail. Imagine writing the Cambridge Grammar of the English >>> Language (or the OED) without using native speaker judgements. >>> >>> Best wishes, Dick Hudson > > > From egibson at MIT.EDU Fri Sep 10 18:30:16 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Fri, 10 Sep 2010 14:30:16 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8A71F0.4050507@ling.ucl.ac.uk> Message-ID: Dear Dick: Perhaps we are talking at cross purposes. I don't understand what is confusing about what Ev Fedorenko and I are claiming. All we are saying is that if you have some testable claim involving a general hypothesis about a language, then you need to get quantitative data from unbiased sources to evaluate that claim. If you are interested in English past tense morphology, then depending on the claims that you might want to investigate, there are lots of ways to get relevant quantitative evidence. Corpus data will probably be useful. For very low frequency words, you can run experiments to test behavior with respect to such words. Your example of the past tense of "bid" is a fine such example. You can run an experiment like the one you suggested to find out what people think the past tense is. If you then found that 20/50 people responded "bidded" and 30/50 respond "bid", that is a lot of useful information. As you suggest in your discussion, this result wouldn't answer the question of how past tense is stored in each individual. This result would be ambiguous among several possible explanations. One possibility is that the probability distribution that is being discovered reflects different dialects, such that 2/5 of the population has one past tense, and 3/5 has another. Another possibility is that each person has a similar probability distribution in their heads, such that 2/5 of the time I respond one way, and 3/5 of the time I respond another. Further experiments would be necessary to answer between these and other possible theories (e.g., with repeated trials from the same person, carefully planned so that the participants don't notice that they are being asked multiple times). Without the quantitative evidence in the first place, there is no way to answer these kinds of questions. Regarding the past tense of "go", this would be useful as a baseline in an experiment involving the less frequent ones. So, yes, it would useful to gather quantitative evidence in such a case also, as baselines with respect to the more interesting cases for theories. The bottom line: if you have a generalization about a language that you wish to evaluate (such that you hypothesize that it is true across the speakers of the language), then you need quantitative evidence from multiple individuals, using an unbiased data collection method, to evaluate such a claim. The point about Mechanical Turk is that it is really *easy* to do this now, at least for languages like English. Best wishes, Ted Gibson & Ev Fedorenko On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying > that I shouldn't claim, for example, that the past tense of GO is > "went" without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic > behaviour (language 'out there', E-language) and trying to explain > it to ourselves in terms of a language system (language 'in here', I- > language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language. > > Given that view of language development, I don't see how > quantitative data will help. Let's take a real uncertainty, such as > the past tense of BID. If I want to say I did it, do I say "I > bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > people to oblige on Mechanical Turk, and 20 of them give "bidded" > and 30 "bid". So what? Does that mean that the correct answer is > "bidded"? Surely not. How is it better than my judgement? I agree > you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a > rare word, so in a sense quantitative data are irrelevant even > there. What would solve the problem of subjectivity, of course, > would be a machine for probing the bit of my mind (or even brain) > that holds BID and its details; but I suspect that even that > wouldn't move us much further forward than my original "don't know". > (Incidentally I write as a fan of quantitative sociolinguistics, so > I do accept that quantitative data are relevant to linguistic > analysis in some areas, where the I-language phenomenon is frequent > enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or > individual. The problem isn't unique to linguistics of course; it's > the same throughout the social sciences. But what's special about > linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the > differences between individuals really matter. I don't see that > we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate > reports of individual I-language. A rotten situation for a science, > but I don't see how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: >> Dear Dan, Dick: >> >> I would like to clarify some points that Dan Everett makes, in >> response to Dick Hudson. >> >> Ev Fedorenko and I have written a couple of papers recently (Gibson & >> Fedorenko, 2010, in press, see references and links below) on what we >> think are weak methodological standards in syntax and semantics >> research over the past many years. The issue that we address is the >> prevalent method in syntax and semantics research, which involves >> obtaining a judgment of the acceptability of a sentence / meaning >> pair, typically by just the author of the paper, sometimes with >> feedback from colleagues. As we address in our papers, this >> methodology does not allow proper testing of scientific hypotheses >> because of (a) the small number of experimental participants >> (typically one); (b) the small number of experimental stimuli >> (typically one); (c) cognitive biases on the part of the researcher >> and participants; and (d) the effect of the preceding context (e.g., >> other constructions the researcher may have been recently >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >> several others cited in Gibson & Fedorenko, in press; for similar >> points, but with not as strong a conclusion as ours). >> >> Three issues need to be separated here: (1) the use of intuitive >> judgments as a dependent measure in a language experiment; (2) >> potential cognitive biases on the part of experimental subjects and >> experimenters in language experiments; and (3) the need for obtaining >> quantitative evidence, whatever the dependent measure might be. The >> paper that Ev and I wrote addresses the last two issues, but does not >> go into depth on the first issue (the use of intuitions as a >> dependent >> measure in language experiments). Regarding this issue, we don't >> think >> that there is anything wrong with gathering intuitive judgments as a >> dependent measure, as long as the task is clear to the experimental >> participants. >> >> In the longer paper (Gibson & Fedorenko, in press) we respond to some >> arguments that have been given in support of continuing to use the >> traditional non-quantitative method in syntax / semantics research. >> One recent defense of the traditional method comes from Phillips >> (2008), who argues that no harm has come from the non-quantitative >> approach in syntax research thus far. Phillips argues that there are >> no cases in the literature where an incorrect intuitive judgment has >> become the basis for a widely accepted generalization or an important >> theoretical claim. He therefore concludes that there is no reason to >> adopt more rigorous data collection standards. We challenge Philips? >> conclusion by presenting three cases from the literature where a >> faulty intuition has led to incorrect generalizations and mistaken >> theorizing, plausibly due to cognitive biases on the part of the >> researchers. >> >> A second argument that is sometimes presented for the continued use >> of >> the traditional non-quantitative method is that it would be too >> inefficient to evaluate every syntactic / semantic hypothesis or >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >> make this argument explicitly in their response to Gibson & Fedorenko >> (2010): ?It would cripple linguistic investigation if it were >> required >> that all judgments of ambiguity and grammaticality be subject to >> statistically rigorous experiments on naive subjects, especially when >> investigating languages whose speakers are hard to access? (Culicover >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >> earlier in the discussion here.) Whereas we agree that in >> circumstances where gathering data is difficult, some evidence is >> better than no evidence, we do not agree that research would be >> slowed >> with respect to languages where experimental participants are easy to >> access, such as English. In contrast, we think that the opposite is >> true: the field?s progress is probably slowed by not doing >> quantitative research. >> Suppose that a typical syntax / semantics paper that lacks >> quantitative evidence includes judgments for 50 or more sentences / >> meaning pairs, corresponding to 50 or more empirical claims. Even if >> most of the judgments from such a paper are correct or are on the >> right track, the problem is in knowing which judgments are correct. >> For example, suppose that 90% of the judgments from an arbitrary >> paper >> are correct (which is probably a high estimate). (Colin Phillips and >> some of his former students / postdocs have commented to us that, in >> their experience, quantitative acceptability judgment studies almost >> always validate the claim(s) in the literature. This is not our >> experience, however. Most experiments that we have run which attempt >> to test some syntactic / semantic hypothesis in the literature end up >> providing us with a pattern of data that had not been known before >> the >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >> that in a paper with 50 empirical claims 45/50 are correct. But which >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over >> two million different theories. By quantitatively evaluating the >> empirical claims, we reduce the uncertainty a great deal. To make >> progress, it is better to have theoretical claims supported by solid >> quantitative evidence, so that even if the interpretation of the data >> changes over time as new evidence becomes available ? as is often the >> case in any field of science ? the empirical pattern can be used as a >> basis for further theorizing. >> >> Furthermore, it is no longer expensive to run behavioral experiments, >> at least in English and other widely spoken languages. There now >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which >> can be used for collecting behavioral data over the internet quickly >> and inexpensively. The cost of using an interface like this is >> minimal, and the time that it takes for the results to be returned is >> short. For example, currently on Mechanical Turk, a survey of >> approximately 50 items will be answered by 50 or more participants >> within a couple of hours, at a cost of approximately $1 per >> participant. Thus a survey can be completed within a day, at a cost >> of >> less than $50. (The hard work of designing the experiment, and >> constructing controlled materials remains of course.) >> >> Sorry to be so verbose. But I think that these methodological points >> are very important. >> >> Best wishes, >> >> Ted Gibson >> >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >> methods in syntax and semantics research. Language and Cognitive >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >> & Fedorenko InPress LCP.pdf >> >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >> linguistics research. Trends in Cognitive Science, 14, 233-234. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >> Fedorenko >> 2010 TICS.pdf >> >> >> >> >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that >>> intuitions are a fine way to generate hypotheses and even to test >>> them - to a degree. But while it might not have been feasible for >>> Huddleston, Pullum, and the other contributors to the Cambridge >>> Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, corpora, and standard psycholinguistic experimentation >>> (indeed, Standard Social Science Methodology) is vital for taking >>> the >>> field forward and for providing the best support for different >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>> paper on this, showing serious shortcomings with intuitions as the >>> sole source of evidence, in their paper: "The need for quantitative >>> methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written >>> convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>> Science), and researchers from Brazil are beginning a third round of >>> experimental work among the Pirahas, since my own work on the syntax >>> was, like almost every other field researcher's, based on native >>> speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions >>> to Greenberg's work on classifying the languages of the Americas. >>> His >>> methods were strongly (and justifiably) criticized. However, I >>> always >>> thought that his methods were a great way of generating hypotheses, >>> so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions. >>> >>> -- Dan >> >> >> >>>> We linguists can add a further layer of explanation to the >>>> judgements, but some judgements do seem to be more reliable than >>>> others. And if we have to wait for psycholinguistic evidence for >>>> every detailed analysis we make, our whole discipline will >>>> immediately grind to a halt. Like it or not, native speaker >>>> judgements are what put us linguists ahead of the rest in handling >>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>> Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >> >> >> From dryer at buffalo.edu Fri Sep 10 18:51:45 2010 From: dryer at buffalo.edu (dryer at buffalo.edu) Date: Fri, 10 Sep 2010 14:51:45 -0400 Subject: analysis: unhappiness In-Reply-To: <63A5D436-4384-4F43-AC8E-0DF7D03AE8B8@colorado.edu> Message-ID: The following sentence of Lise's "Since we don't in fact have such an agreed-on terminology, we have to be quite careful in making clear what we are referring to when we talk about 'the correct analysis' of a form like 'unhappiness'" suggests that there are two senses in which an analysis can be "the correct analysis", one in terms of what is in people's heads, the other in terms of what is "out there". There are a vast number of patterns "out there". The only distinction, on my view, amongst this vast number of patterns, is between those that reflect something inside people's heads and those that don't. But if that is the case, then there is no coherent sense in which one can talk of "the correct analysis" of what is "out there", except in terms of what is in people's heads, and thus no second sense of "the correct analysis". The patterns that don't correspond to things in people's heads fall into (at least) two categories. There are those that are akin to constellations of stars and, as with constellations, there is no reality to these patterns, except in the minds of linguists. And there are those patterns which are the fossil remains of what was in the heads of speakers of an earlier stage of the language but which no longer are. These latter patterns are real, and they are relevant to exlaining why the language is now the way it is, but they are not relevant, I think many would agree, as to what is the "correct analysis" of the language today. For this reason, I claim that the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads. Again, I recommend the work of Bruce Derwing for lengthy discussion of these issues. Matthew --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn wrote: > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. And for > distinguishing among the degrees of pattern apprehension that a person > may have, from vague preferences detectable in reaction times or other > behavior all the way up through clear metalinguistic insights. Dick > Hudson's note reminding us of the Gleitman and Gleitman study is right > on target. > > Since we don't in fact have such an agreed-on terminology, we have to > be quite careful in making clear what we are referring to when we talk > about 'the correct analysis' of a form like 'unhappiness'. We know, but > tend to forget - and tend to forget to tell our students! - that it's an > empirical question as to whether the formal simplicity and coherence of > description of forms 'out there' (e.g. lovely abstract morphophonemics) > is any kind of approximation to the way knowledge of the same forms is > organized in a particular person's head. If we remember that a very > large proportion of what we know about our language is 'out there' when > we are infants and has to be internalized through experience with the > language (even if you believe in innate 'core language'), the variation > in internal knowledge from one person to another is more understandable. > > We especially need to consider (and try to test) the possibility that > since > the brain can make multiple cross-connections, multiple patterns are > involved > simultaneously in morphological and syntactic analyses. I suggest that > that's the case with 'unhappiness' - and the linguistic analyses that I > know about are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> >> Two comments. >> >> First (elaborating perhaps on Dick Hudson's comment), I think there >> is an >> important distinction between low-level linguistic intuitions (like >> whether a >> word or sentence is well-formed or what it means) and higher-level >> intuitions >> (like what the structure of a word or sentence is). One can take >> the position >> that we need to account for the former (while recognizing that they >> are not >> always reliable) but not the latter. >> >> Second, the tension here is not only between evidence from speaker >> intuitions >> versus evidence from psycholinguistic experiments. There is also a >> tension >> between deciding on the correct analysis on the basis of a priori >> simplicity >> arguments versus deciding on the correct analysis on the basis of >> psycholinguistic evidence (see Derwing 1973). The bracketing >> paradox that Dan >> referred to that arises with the word (semantics argues >> for >> [[un+happi] + er], morphology and phonology argues for [un + [happi >> +er]] (the >> comparative suffix can only be attached to adjectives containing one >> or two >> syllables) is only a paradox if one assumes that speakers adopt the >> simplest >> analysis. For example, if speakers adopt a more complex rule for >> either of these >> (e.g. perhaps the rule for attaching -er can apply exceptionally to >> trisyllabic >> words beginning with un-), then the bracketing paradox disappears. >> >> Matthew >> >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>> conscious judgements are only one kind of evidence that we need to >>> take >>> into account. I admire Carson Schutze's work (which I reviewed in >>> fact), >>> and of course I've been aware of complaints about judgements by >>> people >>> like Labov for decades. >>> >>> But you're missing my main point, which is that all judgements aren't >>> equally reliable. If you want to know how /unhappiness/ is >>> structured, >>> ask a linguist, not a five-year old. And one of the by-products of >>> education may be increased sensitivity to syntax - which is one of >>> the >>> many reasons why linguists need to pay more attention to education. >>> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> On 09/09/2010 11:39, Daniel Everett wrote: >>>> Dick, >>>> >>>> You raise an important issue here about >>> methodology. I believe that intuitions are a fine way to generate >>> hypotheses and even to test them - to a degree. But while it might >>> not have >>> been feasible for Huddleston, Pullum, and the other contributors to >>> the >>> Cambridge Grammar to conduct experiments on every point of the >>> grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, >>> corpora, and standard psycholinguistic experimentation (indeed, >>> Standard >>> Social Science Methodology) is vital for taking the field forward >>> and for >>> providing the best support for different analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new paper on this, showing >>> serious >>> shortcomings with intuitions as the sole source of evidence, in their >>> paper: "The need for quantitative methods in syntax and semantics >>> research".> >>>> Carson Schutze and Wayne Cowart, among others, >>> have also written convincingly on this.> >>>> It is one reason that a team from Stanford, MIT >>> (Brain and Cognitive Science), and researchers from Brazil are >>> beginning a >>> third round of experimental work among the Pirahas, since my own >>> work on >>> the syntax was, like almost every other field researcher's, based >>> on native >>> speaker intuitions and corpora.> >>>> The discussion of methodologies reminds me of >>> the initial reactions to Greenberg's work on classifying the >>> languages of >>> the Americas. His methods were strongly (and justifiably) criticized. >>> However, I always thought that his methods were a great way of >>> generating >>> hypotheses, so long as they were ultimately put to the test of >>> standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions.> >>>> -- Dan >>>> >>>>> We linguists can add a further layer of >>> explanation to the judgements, but some judgements do seem to be more >>> reliable than others. And if we have to wait for psycholinguistic >>> evidence >>> for every detailed analysis we make, our whole discipline will >>> immediately >>> grind to a halt. Like it or not, native speaker judgements are what >>> put us >>> linguists ahead of the rest in handling fine detail. Imagine >>> writing the >>> Cambridge Grammar of the English Language (or the OED) without >>> using native >>> speaker judgements.>> >>>>> Best wishes, Dick Hudson >>>>> >>>> >>>> >>> >>> >>> >>> >>> >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > From amnfn at well.com Fri Sep 10 19:09:07 2010 From: amnfn at well.com (A. Katz) Date: Fri, 10 Sep 2010 12:09:07 -0700 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284130304@cast-dryerm2.caset.buffalo.edu> Message-ID: Matthew, Thanks for stating that, because I was almost beginning to imagine that there was no essential disagreement, and that all of us agree that there is more -- and less -- to language than what is found in people's heads. Your position is the one I am familiar with from the functionalist point of view, and I was beginning to feel that it was underrepresented on Funknet. Those of us who disagree with your stated position -- but are very familiar with it -- are interested not just in psycholinguistics and how people process language -- but also in the communicative function of language as a system whereby information is transferred. Just as you and I may not be aware of the way our emails are encoded and then decoded by the computers that help us send emails back and forth, speakers may be compeltely unaware of what language does in order to transmit information. After speakers have finished sending forth their linguistic output, it matters not at all how they arrived at this output. Language processing is separate from language in the same way that data processing is separate from data. Best, --Aya On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have to be > quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the correct > analysis", one in terms of what is in people's heads, the other in terms of > what is "out there". > > There are a vast number of patterns "out there". The only distinction, on my > view, amongst this vast number of patterns, is between those that reflect > something inside people's heads and those that don't. But if that is the > case, then there is no coherent sense in which one can talk of "the correct > analysis" of what is "out there", except in terms of what is in people's > heads, and thus no second sense of "the correct analysis". The patterns that > don't correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of stars and, as > with constellations, there is no reality to these patterns, except in the > minds of linguists. And there are those patterns which are the fossil > remains of what was in the heads of speakers of an earlier stage of the > language but which no longer are. These latter patterns are real, and they > are relevant to exlaining why the language is now the way it is, but they are > not relevant, I think many would agree, as to what is the "correct analysis" > of the language today. > > For this reason, I claim that the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion of these > issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > >> I wish we had better terminology for keeping track of whether, at a >> given time, we are talking about the patterns that are 'out there' in >> the language and might possibly be apprehended (subconsciously) by a >> speaker, and when we are talking about the patterns that a particular >> speaker actually does apprehend, as indicated by experiments, from >> simple 'wug tests' up to brain wave and eye-gaze studies. And for >> distinguishing among the degrees of pattern apprehension that a person >> may have, from vague preferences detectable in reaction times or other >> behavior all the way up through clear metalinguistic insights. Dick >> Hudson's note reminding us of the Gleitman and Gleitman study is right >> on target. >> >> Since we don't in fact have such an agreed-on terminology, we have to >> be quite careful in making clear what we are referring to when we talk >> about 'the correct analysis' of a form like 'unhappiness'. We know, but >> tend to forget - and tend to forget to tell our students! - that it's an >> empirical question as to whether the formal simplicity and coherence of >> description of forms 'out there' (e.g. lovely abstract morphophonemics) >> is any kind of approximation to the way knowledge of the same forms is >> organized in a particular person's head. If we remember that a very >> large proportion of what we know about our language is 'out there' when >> we are infants and has to be internalized through experience with the >> language (even if you believe in innate 'core language'), the variation >> in internal knowledge from one person to another is more understandable. >> We especially need to consider (and try to test) the >> possibility that >> since >> the brain can make multiple cross-connections, multiple patterns are >> involved >> simultaneously in morphological and syntactic analyses. I suggest that >> that's the case with 'unhappiness' - and the linguistic analyses that I >> know about are not good at handling that kind of idea. >> >> Lise >> >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >> >>> >>> Two comments. >>> >>> First (elaborating perhaps on Dick Hudson's comment), I think there >>> is an >>> important distinction between low-level linguistic intuitions (like >>> whether a >>> word or sentence is well-formed or what it means) and higher-level >>> intuitions >>> (like what the structure of a word or sentence is). One can take >>> the position >>> that we need to account for the former (while recognizing that they >>> are not >>> always reliable) but not the latter. >>> >>> Second, the tension here is not only between evidence from speaker >>> intuitions >>> versus evidence from psycholinguistic experiments. There is also a >>> tension >>> between deciding on the correct analysis on the basis of a priori >>> simplicity >>> arguments versus deciding on the correct analysis on the basis of >>> psycholinguistic evidence (see Derwing 1973). The bracketing >>> paradox that Dan >>> referred to that arises with the word (semantics argues >>> for >>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>> +er]] (the >>> comparative suffix can only be attached to adjectives containing one >>> or two >>> syllables) is only a paradox if one assumes that speakers adopt the >>> simplest >>> analysis. For example, if speakers adopt a more complex rule for >>> either of these >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>> trisyllabic >>> words beginning with un-), then the bracketing paradox disappears. >>> >>> Matthew >>> >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>> conscious judgements are only one kind of evidence that we need to >>>> take >>>> into account. I admire Carson Schutze's work (which I reviewed in >>>> fact), >>>> and of course I've been aware of complaints about judgements by >>>> people >>>> like Labov for decades. >>>> >>>> But you're missing my main point, which is that all judgements aren't >>>> equally reliable. If you want to know how /unhappiness/ is >>>> structured, >>>> ask a linguist, not a five-year old. And one of the by-products of >>>> education may be increased sensitivity to syntax - which is one of >>>> the >>>> many reasons why linguists need to pay more attention to education. >>>> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>> Dick, >>>>> >>>>> You raise an important issue here about >>>> methodology. I believe that intuitions are a fine way to generate >>>> hypotheses and even to test them - to a degree. But while it might >>>> not have >>>> been feasible for Huddleston, Pullum, and the other contributors to >>>> the >>>> Cambridge Grammar to conduct experiments on every point of the >>>> grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, >>>> corpora, and standard psycholinguistic experimentation (indeed, >>>> Standard >>>> Social Science Methodology) is vital for taking the field forward >>>> and for >>>> providing the best support for different analyses. Ted Gibson and Ev >>>> Fedorenko have written a very useful new paper on this, showing >>>> serious >>>> shortcomings with intuitions as the sole source of evidence, in their >>>> paper: "The need for quantitative methods in syntax and semantics >>>> research".> >>>>> Carson Schutze and Wayne Cowart, among others, >>>> have also written convincingly on this.> >>>>> It is one reason that a team from Stanford, MIT >>>> (Brain and Cognitive Science), and researchers from Brazil are >>>> beginning a >>>> third round of experimental work among the Pirahas, since my own >>>> work on >>>> the syntax was, like almost every other field researcher's, based >>>> on native >>>> speaker intuitions and corpora.> >>>>> The discussion of methodologies reminds me of >>>> the initial reactions to Greenberg's work on classifying the >>>> languages of >>>> the Americas. His methods were strongly (and justifiably) criticized. >>>> However, I always thought that his methods were a great way of >>>> generating >>>> hypotheses, so long as they were ultimately put to the test of >>>> standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions.> >>>>> -- Dan >>>>> >>>>>> We linguists can add a further layer of >>>> explanation to the judgements, but some judgements do seem to be more >>>> reliable than others. And if we have to wait for psycholinguistic >>>> evidence >>>> for every detailed analysis we make, our whole discipline will >>>> immediately >>>> grind to a halt. Like it or not, native speaker judgements are what >>>> put us >>>> linguists ahead of the rest in handling fine detail. Imagine >>>> writing the >>>> Cambridge Grammar of the English Language (or the OED) without >>>> using native >>>> speaker judgements.>> >>>>>> Best wishes, Dick Hudson >>>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> >>>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> >> Campus Mail Address: >> UCB 594, Institute of Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> >> > > > > > From kuzar at research.haifa.ac.il Fri Sep 10 20:26:18 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Fri, 10 Sep 2010 23:26:18 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: Message-ID: The Modern Hebrew data supplied by Eitan are incomplete. Hebrew distinguishes between locution (say, hear, think, etc.) and situation (action, event, state, etc.). What Eitan describes is only true with regard to nouns (and clauses) expressing locution. 'Announcement' is indeed such a noun. Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', etc., whose denotatum is a situation, cannot be followed by ki, but only by Se-, e.g.: margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) annoys me the-situation that-all went 'I am upset about the situation that all have gone' On the other hand, the relative Se- may be replaced by the more elegant and classical aSer, while the Se- of situation clauses may not. Sorry about the invented example. I am overseas now. All this has been described (with corpus data) in: Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit [Hebrew Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. The article is somewhat outdated and contains some inaccuracies I would formulate differently today, but the basic distinction is valid in my opinion. Best, Ron Kuzar --------------- On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > Hi all, > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > comparison can help what seems to be a purely theoretical question based on > a single language. The problem here is that English uses the same element > to > mark regular relatives and these "appositional" relatives. But if at least > one language encodes them by different means, then there's at least a good > case for seeing them as distinct functions. It's basically the same > principle that's used to decide whether to put a meaning on a semantic map. > So here are two languages that I know that encode them differently. > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > clause (ki), which differs from the relative clause marker (Se-), e.g. > > ha-hoda'a Se-kibalnu > the-announcment rel-we_got > "The announcement that we got." > > ha-hoda'a ki hitbatel ha-mifgaS > the-message CMP was_cancelled the-meeting > "The announcement that the meeting was cancelled." > > In Coptic, these clauses are marked by ce-, which marks complement clauses, > *inter alia*, but not relative clauses: > > ph-mewi ce- (complement clause) > 'the-thought that (we are angry)' > > ph-mewi ete- (relative clause) > 'the thought that (we used to think)' > > This seems to be a pretty clear indication that these are complement > clauses > rather than relatives. Even if one doesn't like the notion of nouns taking > complement clauses (and why not? nominalizations in some languages can take > accusative modifiers as well as genitives), it still probably isn't > incidental that the nominalizations are of verbs that take complement > clauses when finite. > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > looking at. > > Best, > Eitan > > > On 10 September 2010 19:21, Arie Verhagen > wrote: > > > And as another addition: the clauses that can only be introduced by > *that* > > (with no role to > > play in the subordinate clause) may be seen as (subtypes of) complement > > clauses, > > expressing a proposition with the noun functioning as Complement Taking > > Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > > constructions like > > "The claim is that X" (traditionally analysed as subject clauses), "I > claim > > that X", "I put forward > > the claim that X", in which the relationship between the verb or noun and > > the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > To add to Joanne's comments: > > > > > > There are basically three ways to distinguish an appositive clause > > > (AC) from a relative clause (RC). > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > pronouns. > > > > > > 2) The AC and the noun it modifies display an equative > relationship--one > > can say X > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > often doesn't > > > (except, perhaps, when the relative clause is sentential). > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > relative pronoun > > > is always part of the clause in an RC. > > > > > > Rong Chen > > > > > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > -- =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From macw at cmu.edu Fri Sep 10 22:23:54 2010 From: macw at cmu.edu (Brian MacWhinney) Date: Fri, 10 Sep 2010 18:23:54 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8A71F0.4050507@ling.ucl.ac.uk> Message-ID: Dick and Ted, I agree with parts of what each of you are saying. Which means that I also disagree with other parts. In practice,, Gibson and Fedorenko, in press, (which I downloaded and scanned) deals with no more than two or three constructions. They mention the fact that people don't have problems with sentences such as "Susan muttered him the news" despite claims that verbs such as "mutter" cannot take the double object construction. They also note that the claims from Jackendoff and Culicover about the differences between the two sentences below are not supported by results from the Mechanical Turk: 1. Peter was trying to remember who carried what. 2. Peter was trying to remember who carried what when. These are interesting facts. If these sentences are supposed to be different and people judge them to be similarly grammatical, then theories based on the supposed differences should be reexamined. There are big chunks of syntactic theory resting on shaky judgments about complex sentences of this type. Getting some of this straight would be a big win, I would say, particularly if linguists would pay attention to the results. But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications. But, Dick, you then move on to questioning data on bid-bidded. Here we have a case of true variation in the population. I would love to know its distribution. As a "fan of quantitative sociolinguistics" shouldn't you too? My take on this is that constructions are not created equal. The three types mentioned here are probably just a start on an inventory of evidentiary types. We need to correctly pair up appropriate methods with each of the types. And we to make sure that people pay attention to the results, once they are in --Brian MacWhinney On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. > > Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: >> Dear Dan, Dick: >> >> I would like to clarify some points that Dan Everett makes, in >> response to Dick Hudson. >> >> Ev Fedorenko and I have written a couple of papers recently (Gibson & >> Fedorenko, 2010, in press, see references and links below) on what we >> think are weak methodological standards in syntax and semantics >> research over the past many years. The issue that we address is the >> prevalent method in syntax and semantics research, which involves >> obtaining a judgment of the acceptability of a sentence / meaning >> pair, typically by just the author of the paper, sometimes with >> feedback from colleagues. As we address in our papers, this >> methodology does not allow proper testing of scientific hypotheses >> because of (a) the small number of experimental participants >> (typically one); (b) the small number of experimental stimuli >> (typically one); (c) cognitive biases on the part of the researcher >> and participants; and (d) the effect of the preceding context (e.g., >> other constructions the researcher may have been recently >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >> several others cited in Gibson & Fedorenko, in press; for similar >> points, but with not as strong a conclusion as ours). >> >> Three issues need to be separated here: (1) the use of intuitive >> judgments as a dependent measure in a language experiment; (2) >> potential cognitive biases on the part of experimental subjects and >> experimenters in language experiments; and (3) the need for obtaining >> quantitative evidence, whatever the dependent measure might be. The >> paper that Ev and I wrote addresses the last two issues, but does not >> go into depth on the first issue (the use of intuitions as a dependent >> measure in language experiments). Regarding this issue, we don't think >> that there is anything wrong with gathering intuitive judgments as a >> dependent measure, as long as the task is clear to the experimental >> participants. >> >> In the longer paper (Gibson & Fedorenko, in press) we respond to some >> arguments that have been given in support of continuing to use the >> traditional non-quantitative method in syntax / semantics research. >> One recent defense of the traditional method comes from Phillips >> (2008), who argues that no harm has come from the non-quantitative >> approach in syntax research thus far. Phillips argues that there are >> no cases in the literature where an incorrect intuitive judgment has >> become the basis for a widely accepted generalization or an important >> theoretical claim. He therefore concludes that there is no reason to >> adopt more rigorous data collection standards. We challenge Philips? >> conclusion by presenting three cases from the literature where a >> faulty intuition has led to incorrect generalizations and mistaken >> theorizing, plausibly due to cognitive biases on the part of the >> researchers. >> >> A second argument that is sometimes presented for the continued use of >> the traditional non-quantitative method is that it would be too >> inefficient to evaluate every syntactic / semantic hypothesis or >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >> make this argument explicitly in their response to Gibson & Fedorenko >> (2010): ?It would cripple linguistic investigation if it were required >> that all judgments of ambiguity and grammaticality be subject to >> statistically rigorous experiments on naive subjects, especially when >> investigating languages whose speakers are hard to access? (Culicover >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >> earlier in the discussion here.) Whereas we agree that in >> circumstances where gathering data is difficult, some evidence is >> better than no evidence, we do not agree that research would be slowed >> with respect to languages where experimental participants are easy to >> access, such as English. In contrast, we think that the opposite is >> true: the field?s progress is probably slowed by not doing >> quantitative research. >> Suppose that a typical syntax / semantics paper that lacks >> quantitative evidence includes judgments for 50 or more sentences / >> meaning pairs, corresponding to 50 or more empirical claims. Even if >> most of the judgments from such a paper are correct or are on the >> right track, the problem is in knowing which judgments are correct. >> For example, suppose that 90% of the judgments from an arbitrary paper >> are correct (which is probably a high estimate). (Colin Phillips and >> some of his former students / postdocs have commented to us that, in >> their experience, quantitative acceptability judgment studies almost >> always validate the claim(s) in the literature. This is not our >> experience, however. Most experiments that we have run which attempt >> to test some syntactic / semantic hypothesis in the literature end up >> providing us with a pattern of data that had not been known before the >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >> that in a paper with 50 empirical claims 45/50 are correct. But which >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over >> two million different theories. By quantitatively evaluating the >> empirical claims, we reduce the uncertainty a great deal. To make >> progress, it is better to have theoretical claims supported by solid >> quantitative evidence, so that even if the interpretation of the data >> changes over time as new evidence becomes available ? as is often the >> case in any field of science ? the empirical pattern can be used as a >> basis for further theorizing. >> >> Furthermore, it is no longer expensive to run behavioral experiments, >> at least in English and other widely spoken languages. There now >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which >> can be used for collecting behavioral data over the internet quickly >> and inexpensively. The cost of using an interface like this is >> minimal, and the time that it takes for the results to be returned is >> short. For example, currently on Mechanical Turk, a survey of >> approximately 50 items will be answered by 50 or more participants >> within a couple of hours, at a cost of approximately $1 per >> participant. Thus a survey can be completed within a day, at a cost of >> less than $50. (The hard work of designing the experiment, and >> constructing controlled materials remains of course.) >> >> Sorry to be so verbose. But I think that these methodological points >> are very important. >> >> Best wishes, >> >> Ted Gibson >> >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >> methods in syntax and semantics research. Language and Cognitive >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >> & Fedorenko InPress LCP.pdf >> >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >> linguistics research. Trends in Cognitive Science, 14, 233-234. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >> 2010 TICS.pdf >> >> >> >> >>> Dick, >>> >>> You raise an important issue here about methodology. I believe that >>> intuitions are a fine way to generate hypotheses and even to test >>> them - to a degree. But while it might not have been feasible for >>> Huddleston, Pullum, and the other contributors to the Cambridge >>> Grammar to conduct experiments on every point of the grammar, >>> experiments could have only made the grammar better. The use of >>> intuitions, corpora, and standard psycholinguistic experimentation >>> (indeed, Standard Social Science Methodology) is vital for taking the >>> field forward and for providing the best support for different >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>> paper on this, showing serious shortcomings with intuitions as the >>> sole source of evidence, in their paper: "The need for quantitative >>> methods in syntax and semantics research". >>> >>> Carson Schutze and Wayne Cowart, among others, have also written >>> convincingly on this. >>> >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>> Science), and researchers from Brazil are beginning a third round of >>> experimental work among the Pirahas, since my own work on the syntax >>> was, like almost every other field researcher's, based on native >>> speaker intuitions and corpora. >>> >>> The discussion of methodologies reminds me of the initial reactions >>> to Greenberg's work on classifying the languages of the Americas. His >>> methods were strongly (and justifiably) criticized. However, I always >>> thought that his methods were a great way of generating hypotheses, >>> so long as they were ultimately put to the test of standard >>> historical linguistics methods. And the same seems true for use of >>> native-speaker intuitions. >>> >>> -- Dan >> >> >> >>>> We linguists can add a further layer of explanation to the >>>> judgements, but some judgements do seem to be more reliable than >>>> others. And if we have to wait for psycholinguistic evidence for >>>> every detailed analysis we make, our whole discipline will >>>> immediately grind to a halt. Like it or not, native speaker >>>> judgements are what put us linguists ahead of the rest in handling >>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>> Language (or the OED) without using native speaker judgements. >>>> >>>> Best wishes, Dick Hudson >> >> >> > From dryer at buffalo.edu Fri Sep 10 22:33:49 2010 From: dryer at buffalo.edu (dryer at buffalo.edu) Date: Fri, 10 Sep 2010 18:33:49 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Aya, I actually agree with everything you say here. Personally, I am MORE interested in the communicative function of language than I am in psycholinguistics and how people process language. But none of that is relevant, I believe, to the very specific question of what it means for an analysis to be correct. While one might conclude from what I said that one ought to do psycholinguistics, that is not my intention. Rather, my conclusion is that since I myself prefer not to do psycholinguistics, I cannot really claim that the analyses I come up with are "the correct" ones. And if it is really important to someone that they identify "correct" analyses, then they ought to be doing psycholinguistics, since there is no coherent notion of correct analysis outside of what is inside of people's heads. Matthew --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" wrote: > Matthew, > > Thanks for stating that, because I was almost beginning to imagine that > there was no essential disagreement, and that all of us agree that there > is more -- and less -- to language than what is found in people's heads. > > Your position is the one I am familiar with from the functionalist point > of view, and I was beginning to feel that it was underrepresented on > Funknet. > > Those of us who disagree with your stated position -- but are very > familiar with it -- are interested not just in psycholinguistics and how > people process language -- but also in the communicative function of > language as a system whereby information is transferred. Just as you and > I may not be aware of the way our emails are encoded and then decoded by > the computers that help us send emails back and forth, speakers may be > compeltely unaware of what language does in order to transmit information. > > After speakers have finished sending forth their linguistic output, it > matters not at all how they arrived at this output. Language processing > is separate from language in the same way that data processing is > separate from data. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > >> >> The following sentence of Lise's >> >> "Since we don't in fact have such an agreed-on terminology, we have to >> be quite careful in making clear what we are referring to when we talk >> about 'the correct analysis' of a form like 'unhappiness'" >> >> suggests that there are two senses in which an analysis can be "the >> correct analysis", one in terms of what is in people's heads, the other >> in terms of what is "out there". >> >> There are a vast number of patterns "out there". The only distinction, >> on my view, amongst this vast number of patterns, is between those that >> reflect something inside people's heads and those that don't. But if >> that is the case, then there is no coherent sense in which one can talk >> of "the correct analysis" of what is "out there", except in terms of >> what is in people's heads, and thus no second sense of "the correct >> analysis". The patterns that don't correspond to things in people's >> heads fall into (at least) two categories. There are those that are >> akin to constellations of stars and, as with constellations, there is >> no reality to these patterns, except in the minds of linguists. And >> there are those patterns which are the fossil remains of what was in >> the heads of speakers of an earlier stage of the language but which no >> longer are. These latter patterns are real, and they are relevant to >> exlaining why the language is now the way it is, but they are not >> relevant, I think many would agree, as to what is the "correct analysis" >> of the language today. >> >> For this reason, I claim that the only sense in which an analysis can be >> "the correct analysis" is in terms of what is inside of people's heads. >> >> Again, I recommend the work of Bruce Derwing for lengthy discussion of >> these issues. >> >> Matthew >> >> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn >> wrote: >> >>> I wish we had better terminology for keeping track of whether, at a >>> given time, we are talking about the patterns that are 'out there' in >>> the language and might possibly be apprehended (subconsciously) by a >>> speaker, and when we are talking about the patterns that a particular >>> speaker actually does apprehend, as indicated by experiments, from >>> simple 'wug tests' up to brain wave and eye-gaze studies. And for >>> distinguishing among the degrees of pattern apprehension that a person >>> may have, from vague preferences detectable in reaction times or other >>> behavior all the way up through clear metalinguistic insights. Dick >>> Hudson's note reminding us of the Gleitman and Gleitman study is right >>> on target. >>> >>> Since we don't in fact have such an agreed-on terminology, we have to >>> be quite careful in making clear what we are referring to when we talk >>> about 'the correct analysis' of a form like 'unhappiness'. We know, >>> but tend to forget - and tend to forget to tell our students! - that >>> it's an empirical question as to whether the formal simplicity and >>> coherence of description of forms 'out there' (e.g. lovely abstract >>> morphophonemics) is any kind of approximation to the way knowledge of >>> the same forms is organized in a particular person's head. If we >>> remember that a very large proportion of what we know about our >>> language is 'out there' when we are infants and has to be internalized >>> through experience with the language (even if you believe in innate >>> 'core language'), the variation in internal knowledge from one person >>> to another is more understandable. We especially need to consider (and >>> try to test) the >>> possibility that >>> since >>> the brain can make multiple cross-connections, multiple patterns are >>> involved >>> simultaneously in morphological and syntactic analyses. I suggest that >>> that's the case with 'unhappiness' - and the linguistic analyses that I >>> know about are not good at handling that kind of idea. >>> >>> Lise >>> >>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >>> >>>> >>>> Two comments. >>>> >>>> First (elaborating perhaps on Dick Hudson's comment), I think there >>>> is an >>>> important distinction between low-level linguistic intuitions (like >>>> whether a >>>> word or sentence is well-formed or what it means) and higher-level >>>> intuitions >>>> (like what the structure of a word or sentence is). One can take >>>> the position >>>> that we need to account for the former (while recognizing that they >>>> are not >>>> always reliable) but not the latter. >>>> >>>> Second, the tension here is not only between evidence from speaker >>>> intuitions >>>> versus evidence from psycholinguistic experiments. There is also a >>>> tension >>>> between deciding on the correct analysis on the basis of a priori >>>> simplicity >>>> arguments versus deciding on the correct analysis on the basis of >>>> psycholinguistic evidence (see Derwing 1973). The bracketing >>>> paradox that Dan >>>> referred to that arises with the word (semantics argues >>>> for >>>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>>> +er]] (the >>>> comparative suffix can only be attached to adjectives containing one >>>> or two >>>> syllables) is only a paradox if one assumes that speakers adopt the >>>> simplest >>>> analysis. For example, if speakers adopt a more complex rule for >>>> either of these >>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>>> trisyllabic >>>> words beginning with un-), then the bracketing paradox disappears. >>>> >>>> Matthew >>>> >>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>>> conscious judgements are only one kind of evidence that we need to >>>>> take >>>>> into account. I admire Carson Schutze's work (which I reviewed in >>>>> fact), >>>>> and of course I've been aware of complaints about judgements by >>>>> people >>>>> like Labov for decades. >>>>> >>>>> But you're missing my main point, which is that all judgements aren't >>>>> equally reliable. If you want to know how /unhappiness/ is >>>>> structured, >>>>> ask a linguist, not a five-year old. And one of the by-products of >>>>> education may be increased sensitivity to syntax - which is one of >>>>> the >>>>> many reasons why linguists need to pay more attention to education. >>>>> >>>>> Best wishes, Dick >>>>> >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about >>>>> methodology. I believe that intuitions are a fine way to generate >>>>> hypotheses and even to test them - to a degree. But while it might >>>>> not have >>>>> been feasible for Huddleston, Pullum, and the other contributors to >>>>> the >>>>> Cambridge Grammar to conduct experiments on every point of the >>>>> grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, >>>>> corpora, and standard psycholinguistic experimentation (indeed, >>>>> Standard >>>>> Social Science Methodology) is vital for taking the field forward >>>>> and for >>>>> providing the best support for different analyses. Ted Gibson and Ev >>>>> Fedorenko have written a very useful new paper on this, showing >>>>> serious >>>>> shortcomings with intuitions as the sole source of evidence, in their >>>>> paper: "The need for quantitative methods in syntax and semantics >>>>> research".> >>>>>> Carson Schutze and Wayne Cowart, among others, >>>>> have also written convincingly on this.> >>>>>> It is one reason that a team from Stanford, MIT >>>>> (Brain and Cognitive Science), and researchers from Brazil are >>>>> beginning a >>>>> third round of experimental work among the Pirahas, since my own >>>>> work on >>>>> the syntax was, like almost every other field researcher's, based >>>>> on native >>>>> speaker intuitions and corpora.> >>>>>> The discussion of methodologies reminds me of >>>>> the initial reactions to Greenberg's work on classifying the >>>>> languages of >>>>> the Americas. His methods were strongly (and justifiably) criticized. >>>>> However, I always thought that his methods were a great way of >>>>> generating >>>>> hypotheses, so long as they were ultimately put to the test of >>>>> standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions.> >>>>>> -- Dan >>>>>> >>>>>>> We linguists can add a further layer of >>>>> explanation to the judgements, but some judgements do seem to be more >>>>> reliable than others. And if we have to wait for psycholinguistic >>>>> evidence >>>>> for every detailed analysis we make, our whole discipline will >>>>> immediately >>>>> grind to a halt. Like it or not, native speaker judgements are what >>>>> put us >>>>> linguists ahead of the rest in handling fine detail. Imagine >>>>> writing the >>>>> Cambridge Grammar of the English Language (or the OED) without >>>>> using native >>>>> speaker judgements.>> >>>>>>> Best wishes, Dick Hudson >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> >>> >>> Lise Menn Home Office: 303-444-4274 >>> 1625 Mariposa Ave Fax: 303-413-0017 >>> Boulder CO 80302 >>> >>> Professor Emerita of Linguistics >>> Fellow, Institute of Cognitive Science >>> University of Colorado >>> >>> Secretary, AAAS Section Z [Linguistics] >>> >>> Campus Mail Address: >>> UCB 594, Institute of Cognitive Science >>> >>> Campus Physical Address: >>> CINC 234 >>> 1777 Exposition Ave, Boulder >>> >>> >>> >>> >> >> >> >> >> > > From dick at ling.ucl.ac.uk Fri Sep 10 23:20:15 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 00:20:15 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Brian, What a helpful message! I think you're right: we need a typology of cases, each needing a different range of methods, ranging from the linguist's own judgements for really easy cases to more complicated quantitative methods for more complicated ones. The trouble with our discipline is that for any community of N speakers, and a language consisting of M 'items' (however you may choose to define 'community' and 'item'), we have N*M datapoints that, in principle, all need to be validated somehow. We might reduce the number by focusing on one speaker, but then you can't use data from other speakers as evidence for that speaker's language; or we might try to construct a 'typical' speaker, but we don't know how to do that; or we might reduce the size of the community by trying to find a 'dialect' (but dialects don't really exist); or we might ignore most of the linguistic items and focus on, say, the modal verbs - but then we miss their links to all the other items. It's different for psycholinguists because they're only interested in general processes, for which linguistic items are just evidence, not the thing under investigation; but for us linguists, the fine detail is everything because we're the people who explore the connections between items. So I look forward to the day when your typology of cases will guide us through a range of different methods to the appropriate ones for any given item. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 23:23, Brian MacWhinney wrote: > Dick and Ted, > I agree with parts of what each of you are saying. Which means that I also disagree with other parts. In practice,, Gibson and Fedorenko, in press, (which I downloaded and scanned) deals with no more than two or three constructions. They mention the fact that people don't have problems with sentences such as "Susan muttered him the news" despite claims that verbs such as "mutter" cannot take the double object construction. They also note that the claims from Jackendoff and Culicover about the differences between the two sentences below are not supported by results from the Mechanical Turk: > 1. Peter was trying to remember who carried what. > 2. Peter was trying to remember who carried what when. > These are interesting facts. If these sentences are supposed to be different and people judge them to be similarly grammatical, then theories based on the supposed differences should be reexamined. There are big chunks of syntactic theory resting on shaky judgments about complex sentences of this type. Getting some of this straight would be a big win, I would say, particularly if linguists would pay attention to the results. > But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications. > But, Dick, you then move on to questioning data on bid-bidded. Here we have a case of true variation in the population. I would love to know its distribution. As a "fan of quantitative sociolinguistics" shouldn't you too? > My take on this is that constructions are not created equal. The three types mentioned here are probably just a start on an inventory of evidentiary types. We need to correctly pair up appropriate methods with each of the types. And we to make sure that people pay attention to the results, once they are in > > --Brian MacWhinney > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> Dear Ted, >> Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? >> >> Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. >> >> Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) >> >> It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. >> >> Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 14:03, Ted Gibson wrote: >>> Dear Dan, Dick: >>> >>> I would like to clarify some points that Dan Everett makes, in >>> response to Dick Hudson. >>> >>> Ev Fedorenko and I have written a couple of papers recently (Gibson& >>> Fedorenko, 2010, in press, see references and links below) on what we >>> think are weak methodological standards in syntax and semantics >>> research over the past many years. The issue that we address is the >>> prevalent method in syntax and semantics research, which involves >>> obtaining a judgment of the acceptability of a sentence / meaning >>> pair, typically by just the author of the paper, sometimes with >>> feedback from colleagues. As we address in our papers, this >>> methodology does not allow proper testing of scientific hypotheses >>> because of (a) the small number of experimental participants >>> (typically one); (b) the small number of experimental stimuli >>> (typically one); (c) cognitive biases on the part of the researcher >>> and participants; and (d) the effect of the preceding context (e.g., >>> other constructions the researcher may have been recently >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>> several others cited in Gibson& Fedorenko, in press; for similar >>> points, but with not as strong a conclusion as ours). >>> >>> Three issues need to be separated here: (1) the use of intuitive >>> judgments as a dependent measure in a language experiment; (2) >>> potential cognitive biases on the part of experimental subjects and >>> experimenters in language experiments; and (3) the need for obtaining >>> quantitative evidence, whatever the dependent measure might be. The >>> paper that Ev and I wrote addresses the last two issues, but does not >>> go into depth on the first issue (the use of intuitions as a dependent >>> measure in language experiments). Regarding this issue, we don't think >>> that there is anything wrong with gathering intuitive judgments as a >>> dependent measure, as long as the task is clear to the experimental >>> participants. >>> >>> In the longer paper (Gibson& Fedorenko, in press) we respond to some >>> arguments that have been given in support of continuing to use the >>> traditional non-quantitative method in syntax / semantics research. >>> One recent defense of the traditional method comes from Phillips >>> (2008), who argues that no harm has come from the non-quantitative >>> approach in syntax research thus far. Phillips argues that there are >>> no cases in the literature where an incorrect intuitive judgment has >>> become the basis for a widely accepted generalization or an important >>> theoretical claim. He therefore concludes that there is no reason to >>> adopt more rigorous data collection standards. We challenge Philips? >>> conclusion by presenting three cases from the literature where a >>> faulty intuition has led to incorrect generalizations and mistaken >>> theorizing, plausibly due to cognitive biases on the part of the >>> researchers. >>> >>> A second argument that is sometimes presented for the continued use of >>> the traditional non-quantitative method is that it would be too >>> inefficient to evaluate every syntactic / semantic hypothesis or >>> phenomenon quantitatively. For example, Culicover& Jackendoff (2010) >>> make this argument explicitly in their response to Gibson& Fedorenko >>> (2010): ?It would cripple linguistic investigation if it were required >>> that all judgments of ambiguity and grammaticality be subject to >>> statistically rigorous experiments on naive subjects, especially when >>> investigating languages whose speakers are hard to access? (Culicover >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>> earlier in the discussion here.) Whereas we agree that in >>> circumstances where gathering data is difficult, some evidence is >>> better than no evidence, we do not agree that research would be slowed >>> with respect to languages where experimental participants are easy to >>> access, such as English. In contrast, we think that the opposite is >>> true: the field?s progress is probably slowed by not doing >>> quantitative research. >>> Suppose that a typical syntax / semantics paper that lacks >>> quantitative evidence includes judgments for 50 or more sentences / >>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>> most of the judgments from such a paper are correct or are on the >>> right track, the problem is in knowing which judgments are correct. >>> For example, suppose that 90% of the judgments from an arbitrary paper >>> are correct (which is probably a high estimate). (Colin Phillips and >>> some of his former students / postdocs have commented to us that, in >>> their experience, quantitative acceptability judgment studies almost >>> always validate the claim(s) in the literature. This is not our >>> experience, however. Most experiments that we have run which attempt >>> to test some syntactic / semantic hypothesis in the literature end up >>> providing us with a pattern of data that had not been known before the >>> experiment (e.g., Breen et al., in press; Fedorenko& Gibson, in >>> press; Patel et al., 2009; Scontras& Gibson, submitted).) This means >>> that in a paper with 50 empirical claims 45/50 are correct. But which >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over >>> two million different theories. By quantitatively evaluating the >>> empirical claims, we reduce the uncertainty a great deal. To make >>> progress, it is better to have theoretical claims supported by solid >>> quantitative evidence, so that even if the interpretation of the data >>> changes over time as new evidence becomes available ? as is often the >>> case in any field of science ? the empirical pattern can be used as a >>> basis for further theorizing. >>> >>> Furthermore, it is no longer expensive to run behavioral experiments, >>> at least in English and other widely spoken languages. There now >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which >>> can be used for collecting behavioral data over the internet quickly >>> and inexpensively. The cost of using an interface like this is >>> minimal, and the time that it takes for the results to be returned is >>> short. For example, currently on Mechanical Turk, a survey of >>> approximately 50 items will be answered by 50 or more participants >>> within a couple of hours, at a cost of approximately $1 per >>> participant. Thus a survey can be completed within a day, at a cost of >>> less than $50. (The hard work of designing the experiment, and >>> constructing controlled materials remains of course.) >>> >>> Sorry to be so verbose. But I think that these methodological points >>> are very important. >>> >>> Best wishes, >>> >>> Ted Gibson >>> >>> Gibson, E.& Fedorenko, E. (In press). The need for quantitative >>> methods in syntax and semantics research. Language and Cognitive >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>> & Fedorenko InPress LCP.pdf >>> >>> Gibson, E.& Fedorenko, E. (2010). Weak quantitative standards in >>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson& Fedorenko >>> 2010 TICS.pdf >>> >>> >>> >>> >>>> Dick, >>>> >>>> You raise an important issue here about methodology. I believe that >>>> intuitions are a fine way to generate hypotheses and even to test >>>> them - to a degree. But while it might not have been feasible for >>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>> Grammar to conduct experiments on every point of the grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, corpora, and standard psycholinguistic experimentation >>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>> field forward and for providing the best support for different >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>> paper on this, showing serious shortcomings with intuitions as the >>>> sole source of evidence, in their paper: "The need for quantitative >>>> methods in syntax and semantics research". >>>> >>>> Carson Schutze and Wayne Cowart, among others, have also written >>>> convincingly on this. >>>> >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>> Science), and researchers from Brazil are beginning a third round of >>>> experimental work among the Pirahas, since my own work on the syntax >>>> was, like almost every other field researcher's, based on native >>>> speaker intuitions and corpora. >>>> >>>> The discussion of methodologies reminds me of the initial reactions >>>> to Greenberg's work on classifying the languages of the Americas. His >>>> methods were strongly (and justifiably) criticized. However, I always >>>> thought that his methods were a great way of generating hypotheses, >>>> so long as they were ultimately put to the test of standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions. >>>> >>>> -- Dan >>> >>> >>>>> We linguists can add a further layer of explanation to the >>>>> judgements, but some judgements do seem to be more reliable than >>>>> others. And if we have to wait for psycholinguistic evidence for >>>>> every detailed analysis we make, our whole discipline will >>>>> immediately grind to a halt. Like it or not, native speaker >>>>> judgements are what put us linguists ahead of the rest in handling >>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>> Language (or the OED) without using native speaker judgements. >>>>> >>>>> Best wishes, Dick Hudson >>> >>> > > From dick at ling.ucl.ac.uk Fri Sep 10 23:40:06 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 00:40:06 +0100 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Ted and Ev, Yes, I understand your view, but I think it's a psycholinguist's view. Your goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality. And (as you know) I admire the way you pursue that goal. But my goal, as a linguist, is different. I want to explore the structure of a language so that I can understand how all the bits fit together. Like you, I'm aiming to model cognition, but my focus is on items and structures, and I start from the assumption that these can and do vary across speakers. However, having said all that I do agree with you that linguists should all get used to collecting and using quantitative data; and with the help of Brian MacWhinney's typology we'd know what methods to use when. And I do agree with your points about bid/bidded: in cases like that, quantitative data would be at least a very good starting point for a proper investigation. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 10/09/2010 19:30, Ted Gibson wrote: > Dear Dick: > > Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are > saying is that if you have some testable claim involving a general > hypothesis about a language, then you need to get quantitative data > from unbiased sources to evaluate that claim. If you are interested in > English past tense morphology, then depending on the claims that you > might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very > low frequency words, you can run experiments to test behavior with > respect to such words. > > Your example of the past tense of "bid" is a fine such example. You > can run an experiment like the one you suggested to find out what > people think the past tense is. If you then found that 20/50 people > responded "bidded" and 30/50 respond "bid", that is a lot of useful > information. As you suggest in your discussion, this result wouldn't > answer the question of how past tense is stored in each individual. > This result would be ambiguous among several possible explanations. > One possibility is that the probability distribution that is being > discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution > in their heads, such that 2/5 of the time I respond one way, and 3/5 > of the time I respond another. Further experiments would be necessary > to answer between these and other possible theories (e.g., with > repeated trials from the same person, carefully planned so that the > participants don't notice that they are being asked multiple times). > Without the quantitative evidence in the first place, there is no way > to answer these kinds of questions. > > Regarding the past tense of "go", this would be useful as a baseline > in an experiment involving the less frequent ones. So, yes, it would > useful to gather quantitative evidence in such a case also, as > baselines with respect to the more interesting cases for theories. > > The bottom line: if you have a generalization about a language that > you wish to evaluate (such that you hypothesize that it is true across > the speakers of the language), then you need quantitative evidence > from multiple individuals, using an unbiased data collection method, > to evaluate such a claim. The point about Mechanical Turk is that it > is really *easy* to do this now, at least for languages like English. > > Best wishes, > > Ted Gibson & Ev Fedorenko > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> Dear Ted, >> Thanks for the very interesting comment, but are you REALLY saying >> that I shouldn't claim, for example, that the past tense of GO is >> "went" without first cross-checking with 50 native speakers? >> >> Isn't there a danger of missing the point that we all, as native >> speakers, spend our whole lives scanning other people's linguistic >> behaviour (language 'out there', E-language) and trying to explain it >> to ourselves in terms of a language system (language 'in here', >> I-language)? So every judgement we make is based on thousands or >> millions of observed exemplars, and reflects a unique experience of >> E-language filtered through a unique I-language. >> >> Given that view of language development, I don't see how quantitative >> data will help. Let's take a real uncertainty, such as the past tense >> of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My >> judgement: I don't know. Ok, you get 50 people to oblige on >> Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? >> Does that mean that the correct answer is "bidded"? Surely not. How >> is it better than my judgement? I agree you could record my speech >> and find how often I use each alternative; but the reason I don't >> know is precisely because it's a rare word, so in a sense >> quantitative data are irrelevant even there. What would solve the >> problem of subjectivity, of course, would be a machine for probing >> the bit of my mind (or even brain) that holds BID and its details; >> but I suspect that even that wouldn't move us much further forward >> than my original "don't know". (Incidentally I write as a fan of >> quantitative sociolinguistics, so I do accept that quantitative data >> are relevant to linguistic analysis in some areas, where the >> I-language phenomenon is frequent enough to produce usable data.) >> >> It seems to me that this discussion raises the really fundamental >> question of what kind of thing we think language is: social or >> individual. The problem isn't unique to linguistics of course; it's >> the same throughout the social sciences. But what's special about >> linguistics is that we deal in very fine details of culture (e.g. >> details of how a particular word is used or pronounced) so the >> differences between individuals really matter. I don't see that we're >> ever going to have anything better than judgements to go on, so what >> we need is a way to ensure that judgements are accurate reports of >> individual I-language. A rotten situation for a science, but I don't >> see how it can get better. >> >> Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 14:03, Ted Gibson wrote: >>> Dear Dan, Dick: >>> >>> I would like to clarify some points that Dan Everett makes, in >>> response to Dick Hudson. >>> >>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>> Fedorenko, 2010, in press, see references and links below) on what we >>> think are weak methodological standards in syntax and semantics >>> research over the past many years. The issue that we address is the >>> prevalent method in syntax and semantics research, which involves >>> obtaining a judgment of the acceptability of a sentence / meaning >>> pair, typically by just the author of the paper, sometimes with >>> feedback from colleagues. As we address in our papers, this >>> methodology does not allow proper testing of scientific hypotheses >>> because of (a) the small number of experimental participants >>> (typically one); (b) the small number of experimental stimuli >>> (typically one); (c) cognitive biases on the part of the researcher >>> and participants; and (d) the effect of the preceding context (e.g., >>> other constructions the researcher may have been recently >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>> several others cited in Gibson & Fedorenko, in press; for similar >>> points, but with not as strong a conclusion as ours). >>> >>> Three issues need to be separated here: (1) the use of intuitive >>> judgments as a dependent measure in a language experiment; (2) >>> potential cognitive biases on the part of experimental subjects and >>> experimenters in language experiments; and (3) the need for obtaining >>> quantitative evidence, whatever the dependent measure might be. The >>> paper that Ev and I wrote addresses the last two issues, but does not >>> go into depth on the first issue (the use of intuitions as a dependent >>> measure in language experiments). Regarding this issue, we don't think >>> that there is anything wrong with gathering intuitive judgments as a >>> dependent measure, as long as the task is clear to the experimental >>> participants. >>> >>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>> arguments that have been given in support of continuing to use the >>> traditional non-quantitative method in syntax / semantics research. >>> One recent defense of the traditional method comes from Phillips >>> (2008), who argues that no harm has come from the non-quantitative >>> approach in syntax research thus far. Phillips argues that there are >>> no cases in the literature where an incorrect intuitive judgment has >>> become the basis for a widely accepted generalization or an important >>> theoretical claim. He therefore concludes that there is no reason to >>> adopt more rigorous data collection standards. We challenge Philips? >>> conclusion by presenting three cases from the literature where a >>> faulty intuition has led to incorrect generalizations and mistaken >>> theorizing, plausibly due to cognitive biases on the part of the >>> researchers. >>> >>> A second argument that is sometimes presented for the continued use of >>> the traditional non-quantitative method is that it would be too >>> inefficient to evaluate every syntactic / semantic hypothesis or >>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>> make this argument explicitly in their response to Gibson & Fedorenko >>> (2010): ?It would cripple linguistic investigation if it were required >>> that all judgments of ambiguity and grammaticality be subject to >>> statistically rigorous experiments on naive subjects, especially when >>> investigating languages whose speakers are hard to access? (Culicover >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>> earlier in the discussion here.) Whereas we agree that in >>> circumstances where gathering data is difficult, some evidence is >>> better than no evidence, we do not agree that research would be slowed >>> with respect to languages where experimental participants are easy to >>> access, such as English. In contrast, we think that the opposite is >>> true: the field?s progress is probably slowed by not doing >>> quantitative research. >>> Suppose that a typical syntax / semantics paper that lacks >>> quantitative evidence includes judgments for 50 or more sentences / >>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>> most of the judgments from such a paper are correct or are on the >>> right track, the problem is in knowing which judgments are correct. >>> For example, suppose that 90% of the judgments from an arbitrary paper >>> are correct (which is probably a high estimate). (Colin Phillips and >>> some of his former students / postdocs have commented to us that, in >>> their experience, quantitative acceptability judgment studies almost >>> always validate the claim(s) in the literature. This is not our >>> experience, however. Most experiments that we have run which attempt >>> to test some syntactic / semantic hypothesis in the literature end up >>> providing us with a pattern of data that had not been known before the >>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>> that in a paper with 50 empirical claims 45/50 are correct. But which >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over >>> two million different theories. By quantitatively evaluating the >>> empirical claims, we reduce the uncertainty a great deal. To make >>> progress, it is better to have theoretical claims supported by solid >>> quantitative evidence, so that even if the interpretation of the data >>> changes over time as new evidence becomes available ? as is often the >>> case in any field of science ? the empirical pattern can be used as a >>> basis for further theorizing. >>> >>> Furthermore, it is no longer expensive to run behavioral experiments, >>> at least in English and other widely spoken languages. There now >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which >>> can be used for collecting behavioral data over the internet quickly >>> and inexpensively. The cost of using an interface like this is >>> minimal, and the time that it takes for the results to be returned is >>> short. For example, currently on Mechanical Turk, a survey of >>> approximately 50 items will be answered by 50 or more participants >>> within a couple of hours, at a cost of approximately $1 per >>> participant. Thus a survey can be completed within a day, at a cost of >>> less than $50. (The hard work of designing the experiment, and >>> constructing controlled materials remains of course.) >>> >>> Sorry to be so verbose. But I think that these methodological points >>> are very important. >>> >>> Best wishes, >>> >>> Ted Gibson >>> >>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>> methods in syntax and semantics research. Language and Cognitive >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>> & Fedorenko InPress LCP.pdf >>> >>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>> 2010 TICS.pdf >>> >>> >>> >>> >>>> Dick, >>>> >>>> You raise an important issue here about methodology. I believe that >>>> intuitions are a fine way to generate hypotheses and even to test >>>> them - to a degree. But while it might not have been feasible for >>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>> Grammar to conduct experiments on every point of the grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, corpora, and standard psycholinguistic experimentation >>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>> field forward and for providing the best support for different >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>> paper on this, showing serious shortcomings with intuitions as the >>>> sole source of evidence, in their paper: "The need for quantitative >>>> methods in syntax and semantics research". >>>> >>>> Carson Schutze and Wayne Cowart, among others, have also written >>>> convincingly on this. >>>> >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>> Science), and researchers from Brazil are beginning a third round of >>>> experimental work among the Pirahas, since my own work on the syntax >>>> was, like almost every other field researcher's, based on native >>>> speaker intuitions and corpora. >>>> >>>> The discussion of methodologies reminds me of the initial reactions >>>> to Greenberg's work on classifying the languages of the Americas. His >>>> methods were strongly (and justifiably) criticized. However, I always >>>> thought that his methods were a great way of generating hypotheses, >>>> so long as they were ultimately put to the test of standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions. >>>> >>>> -- Dan >>> >>> >>> >>>>> We linguists can add a further layer of explanation to the >>>>> judgements, but some judgements do seem to be more reliable than >>>>> others. And if we have to wait for psycholinguistic evidence for >>>>> every detailed analysis we make, our whole discipline will >>>>> immediately grind to a halt. Like it or not, native speaker >>>>> judgements are what put us linguists ahead of the rest in handling >>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>> Language (or the OED) without using native speaker judgements. >>>>> >>>>> Best wishes, Dick Hudson >>> >>> >>> > > > From Lise.Menn at Colorado.EDU Sat Sep 11 00:40:47 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Fri, 10 Sep 2010 18:40:47 -0600 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284130304@cast-dryerm2.caset.buffalo.edu> Message-ID: Matt, I have to disagree with you on the validity of describing what's 'out there' (what Dick Hudson says is his interest, in his contribution of 5:40:06 PM MDT today). We DO have to account for it in order to understand how 'the language in speakers' heads' gets into those heads in the first place. In more detail: Each of us is immersed from (before) birth in a sampling of utterances (and if we are literate, eventually also written forms of the language). In order to understand how we really create our internal representations of our language, we have to know (or be able to estimate) something about the data our brains get as input. There are at least better and worse descriptions of the patterns in those data, and certainly there are wrong ones, though in many cases - for example in the 'unhappiness' case - there are probably conflicting right ones, rather than any single correct one. (OT offers some help in thinking about this.) To take a concrete example, in order to account for the still- unstable changes in English pronominal case marking in compound NP objects of prepositions from a system based on syntactic case (He gave the cookies to Mary and me) to a system apparently based partly on whether the pronoun is next to the governing preposition (He gave the cookies to Mary and I/ to me and Mary), you first have to do an analysis of usage and figure out what the pattern is. And usage is not in our heads (although it's the result of what's in our heads), it's 'out there'. Even fossils and obscure patterns contribute to the redundancy of the language, making it more learnable and and helping to create the resonances used by great poets and orators. (I admit to having oversimplified in speaking as if there were always one 'correct' analysis of the patterns 'out there' that might be (subconsciously) discoverable by speakers. That's not true.) And because not all speakers are equally sensitive to language patterns - again, the Gleitman and Gleitman book is a terrific example - it's also an oversimplification to talk about 'what is in speaker's heads' as if the same thing is in everyone's head. (K.P. Mohanan has also published on this.) At the lexical level, Danielle Cyr's examples (September 9, 2010 8:38:59 PM MDT) further remind us that what's inside each person's head changes over time. So we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover. It's not - it's more like a complex mosaic that does not fit together perfectly. Lise On Sep 10, 2010, at 12:51 PM, dryer at buffalo.edu wrote: > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have > to be quite careful in making clear what we are referring to when > we talk about 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the > correct analysis", one in terms of what is in people's heads, the > other in terms of what is "out there". > > There are a vast number of patterns "out there". The only > distinction, on my view, amongst this vast number of patterns, is > between those that reflect something inside people's heads and those > that don't. But if that is the case, then there is no coherent > sense in which one can talk of "the correct analysis" of what is > "out there", except in terms of what is in people's heads, and thus > no second sense of "the correct analysis". The patterns that don't > correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of > stars and, as with constellations, there is no reality to these > patterns, except in the minds of linguists. And there are those > patterns which are the fossil remains of what was in the heads of > speakers of an earlier stage of the language but which no longer > are. These latter patterns are real, and they are relevant to > exlaining why the language is now the way it is, but they are not > relevant, I think many would agree, as to what is the "correct > analysis" of the language today. > > For this reason, I claim that the only sense in which an analysis > can be "the correct analysis" is in terms of what is inside of > people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion > of these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > >> I wish we had better terminology for keeping track of whether, at a >> given time, we are talking about the patterns that are 'out there' in >> the language and might possibly be apprehended (subconsciously) by a >> speaker, and when we are talking about the patterns that a particular >> speaker actually does apprehend, as indicated by experiments, from >> simple 'wug tests' up to brain wave and eye-gaze studies. And for >> distinguishing among the degrees of pattern apprehension that a >> person >> may have, from vague preferences detectable in reaction times or >> other >> behavior all the way up through clear metalinguistic insights. Dick >> Hudson's note reminding us of the Gleitman and Gleitman study is >> right >> on target. >> >> Since we don't in fact have such an agreed-on terminology, we have >> to >> be quite careful in making clear what we are referring to when we >> talk >> about 'the correct analysis' of a form like 'unhappiness'. We >> know, but >> tend to forget - and tend to forget to tell our students! - that >> it's an >> empirical question as to whether the formal simplicity and >> coherence of >> description of forms 'out there' (e.g. lovely abstract >> morphophonemics) >> is any kind of approximation to the way knowledge of the same >> forms is >> organized in a particular person's head. If we remember that a very >> large proportion of what we know about our language is 'out there' >> when >> we are infants and has to be internalized through experience with >> the >> language (even if you believe in innate 'core language'), the >> variation >> in internal knowledge from one person to another is more >> understandable. >> >> We especially need to consider (and try to test) the possibility >> that >> since >> the brain can make multiple cross-connections, multiple patterns are >> involved >> simultaneously in morphological and syntactic analyses. I suggest >> that >> that's the case with 'unhappiness' - and the linguistic analyses >> that I >> know about are not good at handling that kind of idea. >> >> Lise >> >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >> >>> >>> Two comments. >>> >>> First (elaborating perhaps on Dick Hudson's comment), I think there >>> is an >>> important distinction between low-level linguistic intuitions (like >>> whether a >>> word or sentence is well-formed or what it means) and higher-level >>> intuitions >>> (like what the structure of a word or sentence is). One can take >>> the position >>> that we need to account for the former (while recognizing that they >>> are not >>> always reliable) but not the latter. >>> >>> Second, the tension here is not only between evidence from speaker >>> intuitions >>> versus evidence from psycholinguistic experiments. There is also a >>> tension >>> between deciding on the correct analysis on the basis of a priori >>> simplicity >>> arguments versus deciding on the correct analysis on the basis of >>> psycholinguistic evidence (see Derwing 1973). The bracketing >>> paradox that Dan >>> referred to that arises with the word (semantics argues >>> for >>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>> +er]] (the >>> comparative suffix can only be attached to adjectives containing one >>> or two >>> syllables) is only a paradox if one assumes that speakers adopt the >>> simplest >>> analysis. For example, if speakers adopt a more complex rule for >>> either of these >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>> trisyllabic >>> words beginning with un-), then the bracketing paradox disappears. >>> >>> Matthew >>> >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree >>>> that >>>> conscious judgements are only one kind of evidence that we need to >>>> take >>>> into account. I admire Carson Schutze's work (which I reviewed in >>>> fact), >>>> and of course I've been aware of complaints about judgements by >>>> people >>>> like Labov for decades. >>>> >>>> But you're missing my main point, which is that all judgements >>>> aren't >>>> equally reliable. If you want to know how /unhappiness/ is >>>> structured, >>>> ask a linguist, not a five-year old. And one of the by-products of >>>> education may be increased sensitivity to syntax - which is one of >>>> the >>>> many reasons why linguists need to pay more attention to education. >>>> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>> Dick, >>>>> >>>>> You raise an important issue here about >>>> methodology. I believe that intuitions are a fine way to generate >>>> hypotheses and even to test them - to a degree. But while it might >>>> not have >>>> been feasible for Huddleston, Pullum, and the other contributors to >>>> the >>>> Cambridge Grammar to conduct experiments on every point of the >>>> grammar, >>>> experiments could have only made the grammar better. The use of >>>> intuitions, >>>> corpora, and standard psycholinguistic experimentation (indeed, >>>> Standard >>>> Social Science Methodology) is vital for taking the field forward >>>> and for >>>> providing the best support for different analyses. Ted Gibson and >>>> Ev >>>> Fedorenko have written a very useful new paper on this, showing >>>> serious >>>> shortcomings with intuitions as the sole source of evidence, in >>>> their >>>> paper: "The need for quantitative methods in syntax and semantics >>>> research".> >>>>> Carson Schutze and Wayne Cowart, among others, >>>> have also written convincingly on this.> >>>>> It is one reason that a team from Stanford, MIT >>>> (Brain and Cognitive Science), and researchers from Brazil are >>>> beginning a >>>> third round of experimental work among the Pirahas, since my own >>>> work on >>>> the syntax was, like almost every other field researcher's, based >>>> on native >>>> speaker intuitions and corpora.> >>>>> The discussion of methodologies reminds me of >>>> the initial reactions to Greenberg's work on classifying the >>>> languages of >>>> the Americas. His methods were strongly (and justifiably) >>>> criticized. >>>> However, I always thought that his methods were a great way of >>>> generating >>>> hypotheses, so long as they were ultimately put to the test of >>>> standard >>>> historical linguistics methods. And the same seems true for use of >>>> native-speaker intuitions.> >>>>> -- Dan >>>>> >>>>>> We linguists can add a further layer of >>>> explanation to the judgements, but some judgements do seem to be >>>> more >>>> reliable than others. And if we have to wait for psycholinguistic >>>> evidence >>>> for every detailed analysis we make, our whole discipline will >>>> immediately >>>> grind to a halt. Like it or not, native speaker judgements are what >>>> put us >>>> linguists ahead of the rest in handling fine detail. Imagine >>>> writing the >>>> Cambridge Grammar of the English Language (or the OED) without >>>> using native >>>> speaker judgements.>> >>>>>> Best wishes, Dick Hudson >>>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> >>>> >>> >> >> Lise Menn Home Office: 303-444-4274 >> 1625 Mariposa Ave Fax: 303-413-0017 >> Boulder CO 80302 >> >> Professor Emerita of Linguistics >> Fellow, Institute of Cognitive Science >> University of Colorado >> >> Secretary, AAAS Section Z [Linguistics] >> >> Campus Mail Address: >> UCB 594, Institute of Cognitive Science >> >> Campus Physical Address: >> CINC 234 >> 1777 Exposition Ave, Boulder >> >> >> >> > > > > Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From Lise.Menn at Colorado.EDU Sat Sep 11 01:03:41 2010 From: Lise.Menn at Colorado.EDU (Lise Menn) Date: Fri, 10 Sep 2010 19:03:41 -0600 Subject: analysis: unhappiness In-Reply-To: <4C8AC1D6.3010203@ling.ucl.ac.uk> Message-ID: Dick, I think that what you say is true, but it is only (I hope) a temporary description of the state of the art of psycholinguistics. At least some of us would like our methods to become sensitive enough to individual differences so that we can look at how 'the general processes and principles' interact with the level of an individual person's knowledge of particular constructions, to find out how much each person knows of the patterns 'out there' in the language. Some experimental methods are almost at that point already; they can distinguish degrees of mastery of particular constructions of a language among groups of second-language learners. Have a look at Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. 2002. Overhearing a language during childhood. Psychological Science 13.3, 238-243. Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on to childhood language memory. Cognition, 86(3), B53-B64. Lise On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's > view. Your goal is to find general processes and principles that > apply uniformly across individuals, so you have to use methods to > check for generality. And (as you know) I admire the way you pursue > that goal. But my goal, as a linguist, is different. I want to > explore the structure of a language so that I can understand how all > the bits fit together. Like you, I'm aiming to model cognition, but > my focus is on items and structures, and I start from the assumption > that these can and do vary across speakers. > > However, having said all that I do agree with you that linguists > should all get used to collecting and using quantitative data; and > with the help of Brian MacWhinney's typology we'd know what methods > to use when. And I do agree with your points about bid/bidded: in > cases like that, quantitative data would be at least a very good > starting point for a proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: >> Dear Dick: >> >> Perhaps we are talking at cross purposes. I don't understand what >> is confusing about what Ev Fedorenko and I are claiming. All we are >> saying is that if you have some testable claim involving a general >> hypothesis about a language, then you need to get quantitative data >> from unbiased sources to evaluate that claim. If you are interested >> in English past tense morphology, then depending on the claims that >> you might want to investigate, there are lots of ways to get >> relevant quantitative evidence. Corpus data will probably be >> useful. For very low frequency words, you can run experiments to >> test behavior with respect to such words. >> >> Your example of the past tense of "bid" is a fine such example. You >> can run an experiment like the one you suggested to find out what >> people think the past tense is. If you then found that 20/50 people >> responded "bidded" and 30/50 respond "bid", that is a lot of useful >> information. As you suggest in your discussion, this result >> wouldn't answer the question of how past tense is stored in each >> individual. This result would be ambiguous among several possible >> explanations. One possibility is that the probability distribution >> that is being discovered reflects different dialects, such that 2/5 >> of the population has one past tense, and 3/5 has another. Another >> possibility is that each person has a similar probability >> distribution in their heads, such that 2/5 of the time I respond >> one way, and 3/5 of the time I respond another. Further experiments >> would be necessary to answer between these and other possible >> theories (e.g., with repeated trials from the same person, >> carefully planned so that the participants don't notice that they >> are being asked multiple times). Without the quantitative evidence >> in the first place, there is no way to answer these kinds of >> questions. >> >> Regarding the past tense of "go", this would be useful as a >> baseline in an experiment involving the less frequent ones. So, >> yes, it would useful to gather quantitative evidence in such a case >> also, as baselines with respect to the more interesting cases for >> theories. >> >> The bottom line: if you have a generalization about a language that >> you wish to evaluate (such that you hypothesize that it is true >> across the speakers of the language), then you need quantitative >> evidence from multiple individuals, using an unbiased data >> collection method, to evaluate such a claim. The point about >> Mechanical Turk is that it is really *easy* to do this now, at >> least for languages like English. >> >> Best wishes, >> >> Ted Gibson & Ev Fedorenko >> >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >> >>> Dear Ted, >>> Thanks for the very interesting comment, but are you REALLY saying >>> that I shouldn't claim, for example, that the past tense of GO is >>> "went" without first cross-checking with 50 native speakers? >>> >>> Isn't there a danger of missing the point that we all, as native >>> speakers, spend our whole lives scanning other people's linguistic >>> behaviour (language 'out there', E-language) and trying to explain >>> it to ourselves in terms of a language system (language 'in here', >>> I-language)? So every judgement we make is based on thousands or >>> millions of observed exemplars, and reflects a unique experience >>> of E-language filtered through a unique I-language. >>> >>> Given that view of language development, I don't see how >>> quantitative data will help. Let's take a real uncertainty, such >>> as the past tense of BID. If I want to say I did it, do I say "I >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>> and 30 "bid". So what? Does that mean that the correct answer is >>> "bidded"? Surely not. How is it better than my judgement? I agree >>> you could record my speech and find how often I use each >>> alternative; but the reason I don't know is precisely because it's >>> a rare word, so in a sense quantitative data are irrelevant even >>> there. What would solve the problem of subjectivity, of course, >>> would be a machine for probing the bit of my mind (or even brain) >>> that holds BID and its details; but I suspect that even that >>> wouldn't move us much further forward than my original "don't >>> know". (Incidentally I write as a fan of quantitative >>> sociolinguistics, so I do accept that quantitative data are >>> relevant to linguistic analysis in some areas, where the I- >>> language phenomenon is frequent enough to produce usable data.) >>> >>> It seems to me that this discussion raises the really fundamental >>> question of what kind of thing we think language is: social or >>> individual. The problem isn't unique to linguistics of course; >>> it's the same throughout the social sciences. But what's special >>> about linguistics is that we deal in very fine details of culture >>> (e.g. details of how a particular word is used or pronounced) so >>> the differences between individuals really matter. I don't see >>> that we're ever going to have anything better than judgements to >>> go on, so what we need is a way to ensure that judgements are >>> accurate reports of individual I-language. A rotten situation for >>> a science, but I don't see how it can get better. >>> >>> Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> >>> On 10/09/2010 14:03, Ted Gibson wrote: >>>> Dear Dan, Dick: >>>> >>>> I would like to clarify some points that Dan Everett makes, in >>>> response to Dick Hudson. >>>> >>>> Ev Fedorenko and I have written a couple of papers recently >>>> (Gibson & >>>> Fedorenko, 2010, in press, see references and links below) on >>>> what we >>>> think are weak methodological standards in syntax and semantics >>>> research over the past many years. The issue that we address is the >>>> prevalent method in syntax and semantics research, which involves >>>> obtaining a judgment of the acceptability of a sentence / meaning >>>> pair, typically by just the author of the paper, sometimes with >>>> feedback from colleagues. As we address in our papers, this >>>> methodology does not allow proper testing of scientific hypotheses >>>> because of (a) the small number of experimental participants >>>> (typically one); (b) the small number of experimental stimuli >>>> (typically one); (c) cognitive biases on the part of the researcher >>>> and participants; and (d) the effect of the preceding context >>>> (e.g., >>>> other constructions the researcher may have been recently >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>> several others cited in Gibson & Fedorenko, in press; for similar >>>> points, but with not as strong a conclusion as ours). >>>> >>>> Three issues need to be separated here: (1) the use of intuitive >>>> judgments as a dependent measure in a language experiment; (2) >>>> potential cognitive biases on the part of experimental subjects and >>>> experimenters in language experiments; and (3) the need for >>>> obtaining >>>> quantitative evidence, whatever the dependent measure might be. The >>>> paper that Ev and I wrote addresses the last two issues, but does >>>> not >>>> go into depth on the first issue (the use of intuitions as a >>>> dependent >>>> measure in language experiments). Regarding this issue, we don't >>>> think >>>> that there is anything wrong with gathering intuitive judgments >>>> as a >>>> dependent measure, as long as the task is clear to the experimental >>>> participants. >>>> >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to >>>> some >>>> arguments that have been given in support of continuing to use the >>>> traditional non-quantitative method in syntax / semantics research. >>>> One recent defense of the traditional method comes from Phillips >>>> (2008), who argues that no harm has come from the non-quantitative >>>> approach in syntax research thus far. Phillips argues that there >>>> are >>>> no cases in the literature where an incorrect intuitive judgment >>>> has >>>> become the basis for a widely accepted generalization or an >>>> important >>>> theoretical claim. He therefore concludes that there is no reason >>>> to >>>> adopt more rigorous data collection standards. We challenge >>>> Philips? >>>> conclusion by presenting three cases from the literature where a >>>> faulty intuition has led to incorrect generalizations and mistaken >>>> theorizing, plausibly due to cognitive biases on the part of the >>>> researchers. >>>> >>>> A second argument that is sometimes presented for the continued >>>> use of >>>> the traditional non-quantitative method is that it would be too >>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>> phenomenon quantitatively. For example, Culicover & Jackendoff >>>> (2010) >>>> make this argument explicitly in their response to Gibson & >>>> Fedorenko >>>> (2010): ?It would cripple linguistic investigation if it were >>>> required >>>> that all judgments of ambiguity and grammaticality be subject to >>>> statistically rigorous experiments on naive subjects, especially >>>> when >>>> investigating languages whose speakers are hard to >>>> access? (Culicover >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>> earlier in the discussion here.) Whereas we agree that in >>>> circumstances where gathering data is difficult, some evidence is >>>> better than no evidence, we do not agree that research would be >>>> slowed >>>> with respect to languages where experimental participants are >>>> easy to >>>> access, such as English. In contrast, we think that the opposite is >>>> true: the field?s progress is probably slowed by not doing >>>> quantitative research. >>>> Suppose that a typical syntax / semantics paper that lacks >>>> quantitative evidence includes judgments for 50 or more sentences / >>>> meaning pairs, corresponding to 50 or more empirical claims. Even >>>> if >>>> most of the judgments from such a paper are correct or are on the >>>> right track, the problem is in knowing which judgments are correct. >>>> For example, suppose that 90% of the judgments from an arbitrary >>>> paper >>>> are correct (which is probably a high estimate). (Colin Phillips >>>> and >>>> some of his former students / postdocs have commented to us that, >>>> in >>>> their experience, quantitative acceptability judgment studies >>>> almost >>>> always validate the claim(s) in the literature. This is not our >>>> experience, however. Most experiments that we have run which >>>> attempt >>>> to test some syntactic / semantic hypothesis in the literature >>>> end up >>>> providing us with a pattern of data that had not been known >>>> before the >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This >>>> means >>>> that in a paper with 50 empirical claims 45/50 are correct. But >>>> which >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s >>>> over >>>> two million different theories. By quantitatively evaluating the >>>> empirical claims, we reduce the uncertainty a great deal. To make >>>> progress, it is better to have theoretical claims supported by >>>> solid >>>> quantitative evidence, so that even if the interpretation of the >>>> data >>>> changes over time as new evidence becomes available ? as is often >>>> the >>>> case in any field of science ? the empirical pattern can be used >>>> as a >>>> basis for further theorizing. >>>> >>>> Furthermore, it is no longer expensive to run behavioral >>>> experiments, >>>> at least in English and other widely spoken languages. There now >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? >>>> which >>>> can be used for collecting behavioral data over the internet >>>> quickly >>>> and inexpensively. The cost of using an interface like this is >>>> minimal, and the time that it takes for the results to be >>>> returned is >>>> short. For example, currently on Mechanical Turk, a survey of >>>> approximately 50 items will be answered by 50 or more participants >>>> within a couple of hours, at a cost of approximately $1 per >>>> participant. Thus a survey can be completed within a day, at a >>>> cost of >>>> less than $50. (The hard work of designing the experiment, and >>>> constructing controlled materials remains of course.) >>>> >>>> Sorry to be so verbose. But I think that these methodological >>>> points >>>> are very important. >>>> >>>> Best wishes, >>>> >>>> Ted Gibson >>>> >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>> methods in syntax and semantics research. Language and Cognitive >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>> & Fedorenko InPress LCP.pdf >>>> >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >>>> Fedorenko >>>> 2010 TICS.pdf >>>> >>>> >>>> >>>> >>>>> Dick, >>>>> >>>>> You raise an important issue here about methodology. I believe >>>>> that >>>>> intuitions are a fine way to generate hypotheses and even to test >>>>> them - to a degree. But while it might not have been feasible for >>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>> Grammar to conduct experiments on every point of the grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>> (indeed, Standard Social Science Methodology) is vital for >>>>> taking the >>>>> field forward and for providing the best support for different >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful >>>>> new >>>>> paper on this, showing serious shortcomings with intuitions as the >>>>> sole source of evidence, in their paper: "The need for >>>>> quantitative >>>>> methods in syntax and semantics research". >>>>> >>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>> convincingly on this. >>>>> >>>>> It is one reason that a team from Stanford, MIT (Brain and >>>>> Cognitive >>>>> Science), and researchers from Brazil are beginning a third >>>>> round of >>>>> experimental work among the Pirahas, since my own work on the >>>>> syntax >>>>> was, like almost every other field researcher's, based on native >>>>> speaker intuitions and corpora. >>>>> >>>>> The discussion of methodologies reminds me of the initial >>>>> reactions >>>>> to Greenberg's work on classifying the languages of the >>>>> Americas. His >>>>> methods were strongly (and justifiably) criticized. However, I >>>>> always >>>>> thought that his methods were a great way of generating >>>>> hypotheses, >>>>> so long as they were ultimately put to the test of standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions. >>>>> >>>>> -- Dan >>>> >>>> >>>> >>>>>> We linguists can add a further layer of explanation to the >>>>>> judgements, but some judgements do seem to be more reliable than >>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>> every detailed analysis we make, our whole discipline will >>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>> judgements are what put us linguists ahead of the rest in >>>>>> handling >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>> Language (or the OED) without using native speaker judgements. >>>>>> >>>>>> Best wishes, Dick Hudson >>>> >>>> >>>> >> >> >> Lise Menn Home Office: 303-444-4274 1625 Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 Professor Emerita of Linguistics Fellow, Institute of Cognitive Science University of Colorado Secretary, AAAS Section Z [Linguistics] Campus Mail Address: UCB 594, Institute of Cognitive Science Campus Physical Address: CINC 234 1777 Exposition Ave, Boulder From dan at daneverett.org Sat Sep 11 01:05:18 2010 From: dan at daneverett.org (Daniel Everett) Date: Fri, 10 Sep 2010 21:05:18 -0400 Subject: analysis: unhappiness In-Reply-To: <4C8AC1D6.3010203@ling.ucl.ac.uk> Message-ID: I think that Brian and Dick make excellent points. There are very good grammars written that could be improved by psycholinguistic experimentation and more quantitative approaches. But large sections of those grammars aren't going to change one bit (go-went) with quantitative tests and such tests would be completely counterproductive given the shortness of life and the vastness of the field linguist's tasks. Part of the problem is that linguistics is not simply a subdiscipline of psychology. Linguistics has its own objectives and those only occasionally overlap with psychology. The same for methods. On another note, I don't buy the 'in my head' 'out of my head' distinction either (that Matt seems to be urging upon us). We study different things and have different reasons for being satisfied with the results we achieve. I believe that we linguists are often complacent and fail to apply better methods. But of course that applies to all disciplines. In the meantime, checking corpora, collecting data as a result of careful interviews with native speakers, and the other aspects of the field linguist's task are vital parts of the linguist's task and much of this won't be improved by quantitative methods as we currently understand them. Maybe sometime. Dan P.S. In my original reference to Ted and Ev's paper, I said that they showed the danger of using intuitions. What I meant to say of using intuitions as standardly used by linguists. They convinced me that there is a lot to learn from quantitative methods. On 10 Sep 2010, at 19:40, Richard Hudson wrote: > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's view. Your goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality. And (as you know) I admire the way you pursue that goal. But my goal, as a linguist, is different. I want to explore the structure of a language so that I can understand how all the bits fit together. Like you, I'm aiming to model cognition, but my focus is on items and structures, and I start from the assumption that these can and do vary across speakers. > > However, having said all that I do agree with you that linguists should all get used to collecting and using quantitative data; and with the help of Brian MacWhinney's typology we'd know what methods to use when. And I do agree with your points about bid/bidded: in cases like that, quantitative data would be at least a very good starting point for a proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: >> Dear Dick: >> >> Perhaps we are talking at cross purposes. I don't understand what is confusing about what Ev Fedorenko and I are claiming. All we are saying is that if you have some testable claim involving a general hypothesis about a language, then you need to get quantitative data from unbiased sources to evaluate that claim. If you are interested in English past tense morphology, then depending on the claims that you might want to investigate, there are lots of ways to get relevant quantitative evidence. Corpus data will probably be useful. For very low frequency words, you can run experiments to test behavior with respect to such words. >> >> Your example of the past tense of "bid" is a fine such example. You can run an experiment like the one you suggested to find out what people think the past tense is. If you then found that 20/50 people responded "bidded" and 30/50 respond "bid", that is a lot of useful information. As you suggest in your discussion, this result wouldn't answer the question of how past tense is stored in each individual. This result would be ambiguous among several possible explanations. One possibility is that the probability distribution that is being discovered reflects different dialects, such that 2/5 of the population has one past tense, and 3/5 has another. Another possibility is that each person has a similar probability distribution in their heads, such that 2/5 of the time I respond one way, and 3/5 of the time I respond another. Further experiments would be necessary to answer between these and other possible theories (e.g., with repeated trials from the same person, carefully planned so that the participants don't notice that they are being asked multiple times). Without the quantitative evidence in the first place, there is no way to answer these kinds of questions. >> >> Regarding the past tense of "go", this would be useful as a baseline in an experiment involving the less frequent ones. So, yes, it would useful to gather quantitative evidence in such a case also, as baselines with respect to the more interesting cases for theories. >> >> The bottom line: if you have a generalization about a language that you wish to evaluate (such that you hypothesize that it is true across the speakers of the language), then you need quantitative evidence from multiple individuals, using an unbiased data collection method, to evaluate such a claim. The point about Mechanical Turk is that it is really *easy* to do this now, at least for languages like English. >> >> Best wishes, >> >> Ted Gibson & Ev Fedorenko >> >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >> >>> Dear Ted, >>> Thanks for the very interesting comment, but are you REALLY saying that I shouldn't claim, for example, that the past tense of GO is "went" without first cross-checking with 50 native speakers? >>> >>> Isn't there a danger of missing the point that we all, as native speakers, spend our whole lives scanning other people's linguistic behaviour (language 'out there', E-language) and trying to explain it to ourselves in terms of a language system (language 'in here', I-language)? So every judgement we make is based on thousands or millions of observed exemplars, and reflects a unique experience of E-language filtered through a unique I-language. >>> >>> Given that view of language development, I don't see how quantitative data will help. Let's take a real uncertainty, such as the past tense of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the correct answer is "bidded"? Surely not. How is it better than my judgement? I agree you could record my speech and find how often I use each alternative; but the reason I don't know is precisely because it's a rare word, so in a sense quantitative data are irrelevant even there. What would solve the problem of subjectivity, of course, would be a machine for probing the bit of my mind (or even brain) that holds BID and its details; but I suspect that even that wouldn't move us much further forward than my original "don't know". (Incidentally I write as a fan of quantitative sociolinguistics, so I do accept that quantitative data are relevant to linguistic analysis in some areas, where the I-language phenomenon is frequent enough to produce usable data.) >>> >>> It seems to me that this discussion raises the really fundamental question of what kind of thing we think language is: social or individual. The problem isn't unique to linguistics of course; it's the same throughout the social sciences. But what's special about linguistics is that we deal in very fine details of culture (e.g. details of how a particular word is used or pronounced) so the differences between individuals really matter. I don't see that we're ever going to have anything better than judgements to go on, so what we need is a way to ensure that judgements are accurate reports of individual I-language. A rotten situation for a science, but I don't see how it can get better. >>> >>> Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>> >>> On 10/09/2010 14:03, Ted Gibson wrote: >>>> Dear Dan, Dick: >>>> >>>> I would like to clarify some points that Dan Everett makes, in >>>> response to Dick Hudson. >>>> >>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>>> Fedorenko, 2010, in press, see references and links below) on what we >>>> think are weak methodological standards in syntax and semantics >>>> research over the past many years. The issue that we address is the >>>> prevalent method in syntax and semantics research, which involves >>>> obtaining a judgment of the acceptability of a sentence / meaning >>>> pair, typically by just the author of the paper, sometimes with >>>> feedback from colleagues. As we address in our papers, this >>>> methodology does not allow proper testing of scientific hypotheses >>>> because of (a) the small number of experimental participants >>>> (typically one); (b) the small number of experimental stimuli >>>> (typically one); (c) cognitive biases on the part of the researcher >>>> and participants; and (d) the effect of the preceding context (e.g., >>>> other constructions the researcher may have been recently >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>> several others cited in Gibson & Fedorenko, in press; for similar >>>> points, but with not as strong a conclusion as ours). >>>> >>>> Three issues need to be separated here: (1) the use of intuitive >>>> judgments as a dependent measure in a language experiment; (2) >>>> potential cognitive biases on the part of experimental subjects and >>>> experimenters in language experiments; and (3) the need for obtaining >>>> quantitative evidence, whatever the dependent measure might be. The >>>> paper that Ev and I wrote addresses the last two issues, but does not >>>> go into depth on the first issue (the use of intuitions as a dependent >>>> measure in language experiments). Regarding this issue, we don't think >>>> that there is anything wrong with gathering intuitive judgments as a >>>> dependent measure, as long as the task is clear to the experimental >>>> participants. >>>> >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>>> arguments that have been given in support of continuing to use the >>>> traditional non-quantitative method in syntax / semantics research. >>>> One recent defense of the traditional method comes from Phillips >>>> (2008), who argues that no harm has come from the non-quantitative >>>> approach in syntax research thus far. Phillips argues that there are >>>> no cases in the literature where an incorrect intuitive judgment has >>>> become the basis for a widely accepted generalization or an important >>>> theoretical claim. He therefore concludes that there is no reason to >>>> adopt more rigorous data collection standards. We challenge Philips? >>>> conclusion by presenting three cases from the literature where a >>>> faulty intuition has led to incorrect generalizations and mistaken >>>> theorizing, plausibly due to cognitive biases on the part of the >>>> researchers. >>>> >>>> A second argument that is sometimes presented for the continued use of >>>> the traditional non-quantitative method is that it would be too >>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>>> make this argument explicitly in their response to Gibson & Fedorenko >>>> (2010): ?It would cripple linguistic investigation if it were required >>>> that all judgments of ambiguity and grammaticality be subject to >>>> statistically rigorous experiments on naive subjects, especially when >>>> investigating languages whose speakers are hard to access? (Culicover >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>> earlier in the discussion here.) Whereas we agree that in >>>> circumstances where gathering data is difficult, some evidence is >>>> better than no evidence, we do not agree that research would be slowed >>>> with respect to languages where experimental participants are easy to >>>> access, such as English. In contrast, we think that the opposite is >>>> true: the field?s progress is probably slowed by not doing >>>> quantitative research. >>>> Suppose that a typical syntax / semantics paper that lacks >>>> quantitative evidence includes judgments for 50 or more sentences / >>>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>>> most of the judgments from such a paper are correct or are on the >>>> right track, the problem is in knowing which judgments are correct. >>>> For example, suppose that 90% of the judgments from an arbitrary paper >>>> are correct (which is probably a high estimate). (Colin Phillips and >>>> some of his former students / postdocs have commented to us that, in >>>> their experience, quantitative acceptability judgment studies almost >>>> always validate the claim(s) in the literature. This is not our >>>> experience, however. Most experiments that we have run which attempt >>>> to test some syntactic / semantic hypothesis in the literature end up >>>> providing us with a pattern of data that had not been known before the >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>>> that in a paper with 50 empirical claims 45/50 are correct. But which >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over >>>> two million different theories. By quantitatively evaluating the >>>> empirical claims, we reduce the uncertainty a great deal. To make >>>> progress, it is better to have theoretical claims supported by solid >>>> quantitative evidence, so that even if the interpretation of the data >>>> changes over time as new evidence becomes available ? as is often the >>>> case in any field of science ? the empirical pattern can be used as a >>>> basis for further theorizing. >>>> >>>> Furthermore, it is no longer expensive to run behavioral experiments, >>>> at least in English and other widely spoken languages. There now >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which >>>> can be used for collecting behavioral data over the internet quickly >>>> and inexpensively. The cost of using an interface like this is >>>> minimal, and the time that it takes for the results to be returned is >>>> short. For example, currently on Mechanical Turk, a survey of >>>> approximately 50 items will be answered by 50 or more participants >>>> within a couple of hours, at a cost of approximately $1 per >>>> participant. Thus a survey can be completed within a day, at a cost of >>>> less than $50. (The hard work of designing the experiment, and >>>> constructing controlled materials remains of course.) >>>> >>>> Sorry to be so verbose. But I think that these methodological points >>>> are very important. >>>> >>>> Best wishes, >>>> >>>> Ted Gibson >>>> >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>> methods in syntax and semantics research. Language and Cognitive >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>> & Fedorenko InPress LCP.pdf >>>> >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>>> 2010 TICS.pdf >>>> >>>> >>>> >>>> >>>>> Dick, >>>>> >>>>> You raise an important issue here about methodology. I believe that >>>>> intuitions are a fine way to generate hypotheses and even to test >>>>> them - to a degree. But while it might not have been feasible for >>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>> Grammar to conduct experiments on every point of the grammar, >>>>> experiments could have only made the grammar better. The use of >>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>>> field forward and for providing the best support for different >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>>> paper on this, showing serious shortcomings with intuitions as the >>>>> sole source of evidence, in their paper: "The need for quantitative >>>>> methods in syntax and semantics research". >>>>> >>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>> convincingly on this. >>>>> >>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>>> Science), and researchers from Brazil are beginning a third round of >>>>> experimental work among the Pirahas, since my own work on the syntax >>>>> was, like almost every other field researcher's, based on native >>>>> speaker intuitions and corpora. >>>>> >>>>> The discussion of methodologies reminds me of the initial reactions >>>>> to Greenberg's work on classifying the languages of the Americas. His >>>>> methods were strongly (and justifiably) criticized. However, I always >>>>> thought that his methods were a great way of generating hypotheses, >>>>> so long as they were ultimately put to the test of standard >>>>> historical linguistics methods. And the same seems true for use of >>>>> native-speaker intuitions. >>>>> >>>>> -- Dan >>>> >>>> >>>> >>>>>> We linguists can add a further layer of explanation to the >>>>>> judgements, but some judgements do seem to be more reliable than >>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>> every detailed analysis we make, our whole discipline will >>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>> judgements are what put us linguists ahead of the rest in handling >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>> Language (or the OED) without using native speaker judgements. >>>>>> >>>>>> Best wishes, Dick Hudson >>>> >>>> >>>> >> >> >> > From phdebrab at yahoo.co.uk Sat Sep 11 10:03:35 2010 From: phdebrab at yahoo.co.uk (Philippe De Brabanter) Date: Sat, 11 Sep 2010 10:03:35 +0000 Subject: "Relative clauses" with no relativized element In-Reply-To: Message-ID: Dear all, this is just to say that the Cambridge Grammar of the English Language, like several other grammars I'm aware of, does mention the sorts of constructions Tom is wondering about (esp. pp. 964-67). They're treated as noun complements, whereas relative clauses usually function as modifiers of nouns. H&P give a useful list of nouns licensing these complements (a list which confirms Suzanne Kemmerer's point that these nouns do not always have a verbal counterpart taking a content clause as its complement ? H&P suggest that the most frequent of those licensing nouns is fact). They also point out that content clauses can also sometimes function as supplements (i.e. appositives), as in I'm inclined to favour your first suggestion, that we shelve the proposal until after the election. This confirms Suzanne's suggestion that we shouldn't say that the clausal noun complements are appositives. One last interesting point. On p. 967, H&P show that the licensor may sometimes be more than just a noun, with certain constructions like have + licensing NP or existential there + be facilitating (or being conditions for) the clausal noun complement: The present system has the disadvantage that it is inordinately complicated. vs. ? The disadvantage that it is inordinately complicated has been overlooked. Probably an example like This principle may ground some optimism that the account can be usefully pursued. (M. Sainsbury 2002: "Reference and anaphora", Mind & Language) also derives its acceptability from a construction rather than from just optimism. Best, Philippe De Brabanter Paris 4 - Sorbonne ________________________________ From: E.G. To: funknet at mailman.rice.edu Sent: Fri, 10 September, 2010 19:56:23 Subject: Re: [FUNKNET] "Relative clauses" with no relativized element Jespersen and his nexus-substantives should be mentioned (Philosophy of Grammar, 1924). Also in his MEG and Analytic Syntax one could find interesting discussions. Eitan On 10 September 2010 20:53, Giuliana Fiorentino < giuliana.fiorentino at unimol.it> wrote: > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From dick at ling.ucl.ac.uk Sat Sep 11 11:03:56 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 12:03:56 +0100 Subject: analysis: unhappiness In-Reply-To: <10638E94-DC8F-4181-8135-CE74F045079E@colorado.edu> Message-ID: Dear Lise, Many thanks for these references, which I've just looked at. As you say, they do offer hope that we can apply the methods of psycholinguists to the questions of linguists (e.g. how do individuals categorise Korean or Spanish consonants?), and at the same time build a clearer understanding of how these individuals' I-language is related to their E-language (i.e. the language they've heard). That is encouraging, not least because it breaks down what I see as a gulf between linguists and psycholinguists. But even better, it complements very nicely some work in sociolinguistics which I think you'd enjoy reading. It's by a young Scottish sociolinguist called Jennifer Smith, who did a very careful quantitative study (with two colleagues) of 24 3-year olds in a small fishing town in the north of Scotland. She recorded each of them with their mother, and then analysed two sociolinguistic variables that she'd also analysed in adult speech: a phonological variable (pronunciation of the /au/ vowel in "cow", and a morphosyntactic one, the use of -s on a verb with a plural subject, e.g. "My trousers is falling doon." In both cases usage is variable, so the analysis produces a percentage score for each speaker (e.g. 5% of words with the /au/ vowel have a monophthong). She then compared the children's scores with those of their mothers, and found an astonishingly close match for the phonological variable but no match at all for the morphosyntactic one. The reference is Jennifer Smith, Mercedes Durham & Liane Fortune, 2007. 'Mam, my trousers is fa'in doon!' Community, caregiver and child in the acquisition of variation in a Scottish dialect. (Language, Variation and Change 19. 63-99) Once again we have evidence that output can be closely linked to input, we have a nice quantitative method, and we see an example of the 'global' analysis of language that we should all be striving for: one which embraces both E-language (directly observable to the learner as well as to the linguist) and I-language (only indirectly accessible to both), and which tries to both describe and explain the relation between the two. Neither E-language nor I-language is the 'real' language - they're both part of it. And this global enterprise needs all the methods we can muster. Best wishes, Dick Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm On 11/09/2010 02:03, Lise Menn wrote: > Dick, I think that what you say is true, but it is only (I hope) a > temporary description of the state of the art of psycholinguistics. > At least some of us would like our methods to become sensitive enough > to individual differences so that we can look at how 'the general > processes and principles' interact with the level of an individual > person's knowledge of particular constructions, to find out how much > each person knows of the patterns 'out there' in the language. Some > experimental methods are almost at that point already; they can > distinguish degrees of mastery of particular constructions of a > language among groups of second-language learners. > Have a look at > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > 2002. Overhearing a language during childhood. /Psychological Science/ > 13.3, 238-243. > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on > to childhood language memory. /Cognition, 86/(3), B53-B64. > > Lise > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > >> Dear Ted and Ev, >> Yes, I understand your view, but I think it's a psycholinguist's >> view. Your goal is to find general processes and principles that >> apply uniformly across individuals, so you have to use methods to >> check for generality. And (as you know) I admire the way you pursue >> that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how all >> the bits fit together. Like you, I'm aiming to model cognition, but >> my focus is on items and structures, and I start from the assumption >> that these can and do vary across speakers. >> >> However, having said all that I do agree with you that linguists >> should all get used to collecting and using quantitative data; and >> with the help of Brian MacWhinney's typology we'd know what methods >> to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good >> starting point for a proper investigation. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> >> On 10/09/2010 19:30, Ted Gibson wrote: >>> Dear Dick: >>> >>> Perhaps we are talking at cross purposes. I don't understand what is >>> confusing about what Ev Fedorenko and I are claiming. All we are >>> saying is that if you have some testable claim involving a general >>> hypothesis about a language, then you need to get quantitative data >>> from unbiased sources to evaluate that claim. If you are interested >>> in English past tense morphology, then depending on the claims that >>> you might want to investigate, there are lots of ways to get >>> relevant quantitative evidence. Corpus data will probably be useful. >>> For very low frequency words, you can run experiments to test >>> behavior with respect to such words. >>> >>> Your example of the past tense of "bid" is a fine such example. You >>> can run an experiment like the one you suggested to find out what >>> people think the past tense is. If you then found that 20/50 people >>> responded "bidded" and 30/50 respond "bid", that is a lot of useful >>> information. As you suggest in your discussion, this result wouldn't >>> answer the question of how past tense is stored in each individual. >>> This result would be ambiguous among several possible explanations. >>> One possibility is that the probability distribution that is being >>> discovered reflects different dialects, such that 2/5 of the >>> population has one past tense, and 3/5 has another. Another >>> possibility is that each person has a similar probability >>> distribution in their heads, such that 2/5 of the time I respond one >>> way, and 3/5 of the time I respond another. Further experiments >>> would be necessary to answer between these and other possible >>> theories (e.g., with repeated trials from the same person, carefully >>> planned so that the participants don't notice that they are being >>> asked multiple times). Without the quantitative evidence in the >>> first place, there is no way to answer these kinds of questions. >>> >>> Regarding the past tense of "go", this would be useful as a baseline >>> in an experiment involving the less frequent ones. So, yes, it would >>> useful to gather quantitative evidence in such a case also, as >>> baselines with respect to the more interesting cases for theories. >>> >>> The bottom line: if you have a generalization about a language that >>> you wish to evaluate (such that you hypothesize that it is true >>> across the speakers of the language), then you need quantitative >>> evidence from multiple individuals, using an unbiased data >>> collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at least >>> for languages like English. >>> >>> Best wishes, >>> >>> Ted Gibson & Ev Fedorenko >>> >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >>> >>>> Dear Ted, >>>> Thanks for the very interesting comment, but are you REALLY saying >>>> that I shouldn't claim, for example, that the past tense of GO is >>>> "went" without first cross-checking with 50 native speakers? >>>> >>>> Isn't there a danger of missing the point that we all, as native >>>> speakers, spend our whole lives scanning other people's linguistic >>>> behaviour (language 'out there', E-language) and trying to explain >>>> it to ourselves in terms of a language system (language 'in here', >>>> I-language)? So every judgement we make is based on thousands or >>>> millions of observed exemplars, and reflects a unique experience of >>>> E-language filtered through a unique I-language. >>>> >>>> Given that view of language development, I don't see how >>>> quantitative data will help. Let's take a real uncertainty, such as >>>> the past tense of BID. If I want to say I did it, do I say "I >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>>> and 30 "bid". So what? Does that mean that the correct answer is >>>> "bidded"? Surely not. How is it better than my judgement? I agree >>>> you could record my speech and find how often I use each >>>> alternative; but the reason I don't know is precisely because it's >>>> a rare word, so in a sense quantitative data are irrelevant even >>>> there. What would solve the problem of subjectivity, of course, >>>> would be a machine for probing the bit of my mind (or even brain) >>>> that holds BID and its details; but I suspect that even that >>>> wouldn't move us much further forward than my original "don't >>>> know". (Incidentally I write as a fan of quantitative >>>> sociolinguistics, so I do accept that quantitative data are >>>> relevant to linguistic analysis in some areas, where the I-language >>>> phenomenon is frequent enough to produce usable data.) >>>> >>>> It seems to me that this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual. The problem isn't unique to linguistics of course; it's >>>> the same throughout the social sciences. But what's special about >>>> linguistics is that we deal in very fine details of culture (e.g. >>>> details of how a particular word is used or pronounced) so the >>>> differences between individuals really matter. I don't see that >>>> we're ever going to have anything better than judgements to go on, >>>> so what we need is a way to ensure that judgements are accurate >>>> reports of individual I-language. A rotten situation for a science, >>>> but I don't see how it can get better. >>>> >>>> Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> >>>> >>>> On 10/09/2010 14:03, Ted Gibson wrote: >>>>> Dear Dan, Dick: >>>>> >>>>> I would like to clarify some points that Dan Everett makes, in >>>>> response to Dick Hudson. >>>>> >>>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & >>>>> Fedorenko, 2010, in press, see references and links below) on what we >>>>> think are weak methodological standards in syntax and semantics >>>>> research over the past many years. The issue that we address is the >>>>> prevalent method in syntax and semantics research, which involves >>>>> obtaining a judgment of the acceptability of a sentence / meaning >>>>> pair, typically by just the author of the paper, sometimes with >>>>> feedback from colleagues. As we address in our papers, this >>>>> methodology does not allow proper testing of scientific hypotheses >>>>> because of (a) the small number of experimental participants >>>>> (typically one); (b) the small number of experimental stimuli >>>>> (typically one); (c) cognitive biases on the part of the researcher >>>>> and participants; and (d) the effect of the preceding context (e.g., >>>>> other constructions the researcher may have been recently >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>>> several others cited in Gibson & Fedorenko, in press; for similar >>>>> points, but with not as strong a conclusion as ours). >>>>> >>>>> Three issues need to be separated here: (1) the use of intuitive >>>>> judgments as a dependent measure in a language experiment; (2) >>>>> potential cognitive biases on the part of experimental subjects and >>>>> experimenters in language experiments; and (3) the need for obtaining >>>>> quantitative evidence, whatever the dependent measure might be. The >>>>> paper that Ev and I wrote addresses the last two issues, but does not >>>>> go into depth on the first issue (the use of intuitions as a dependent >>>>> measure in language experiments). Regarding this issue, we don't think >>>>> that there is anything wrong with gathering intuitive judgments as a >>>>> dependent measure, as long as the task is clear to the experimental >>>>> participants. >>>>> >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some >>>>> arguments that have been given in support of continuing to use the >>>>> traditional non-quantitative method in syntax / semantics research. >>>>> One recent defense of the traditional method comes from Phillips >>>>> (2008), who argues that no harm has come from the non-quantitative >>>>> approach in syntax research thus far. Phillips argues that there are >>>>> no cases in the literature where an incorrect intuitive judgment has >>>>> become the basis for a widely accepted generalization or an important >>>>> theoretical claim. He therefore concludes that there is no reason to >>>>> adopt more rigorous data collection standards. We challenge Philips' >>>>> conclusion by presenting three cases from the literature where a >>>>> faulty intuition has led to incorrect generalizations and mistaken >>>>> theorizing, plausibly due to cognitive biases on the part of the >>>>> researchers. >>>>> >>>>> A second argument that is sometimes presented for the continued use of >>>>> the traditional non-quantitative method is that it would be too >>>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) >>>>> make this argument explicitly in their response to Gibson & Fedorenko >>>>> (2010): "It would cripple linguistic investigation if it were required >>>>> that all judgments of ambiguity and grammaticality be subject to >>>>> statistically rigorous experiments on naive subjects, especially when >>>>> investigating languages whose speakers are hard to access" (Culicover >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>>> earlier in the discussion here.) Whereas we agree that in >>>>> circumstances where gathering data is difficult, some evidence is >>>>> better than no evidence, we do not agree that research would be slowed >>>>> with respect to languages where experimental participants are easy to >>>>> access, such as English. In contrast, we think that the opposite is >>>>> true: the field's progress is probably slowed by not doing >>>>> quantitative research. >>>>> Suppose that a typical syntax / semantics paper that lacks >>>>> quantitative evidence includes judgments for 50 or more sentences / >>>>> meaning pairs, corresponding to 50 or more empirical claims. Even if >>>>> most of the judgments from such a paper are correct or are on the >>>>> right track, the problem is in knowing which judgments are correct. >>>>> For example, suppose that 90% of the judgments from an arbitrary paper >>>>> are correct (which is probably a high estimate). (Colin Phillips and >>>>> some of his former students / postdocs have commented to us that, in >>>>> their experience, quantitative acceptability judgment studies almost >>>>> always validate the claim(s) in the literature. This is not our >>>>> experience, however. Most experiments that we have run which attempt >>>>> to test some syntactic / semantic hypothesis in the literature end up >>>>> providing us with a pattern of data that had not been known before the >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means >>>>> that in a paper with 50 empirical claims 45/50 are correct. But which >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That's over >>>>> two million different theories. By quantitatively evaluating the >>>>> empirical claims, we reduce the uncertainty a great deal. To make >>>>> progress, it is better to have theoretical claims supported by solid >>>>> quantitative evidence, so that even if the interpretation of the data >>>>> changes over time as new evidence becomes available -- as is often the >>>>> case in any field of science -- the empirical pattern can be used as a >>>>> basis for further theorizing. >>>>> >>>>> Furthermore, it is no longer expensive to run behavioral experiments, >>>>> at least in English and other widely spoken languages. There now >>>>> exists a marketplace interface -- Amazon.com's Mechanical Turk -- >>>>> which >>>>> can be used for collecting behavioral data over the internet quickly >>>>> and inexpensively. The cost of using an interface like this is >>>>> minimal, and the time that it takes for the results to be returned is >>>>> short. For example, currently on Mechanical Turk, a survey of >>>>> approximately 50 items will be answered by 50 or more participants >>>>> within a couple of hours, at a cost of approximately $1 per >>>>> participant. Thus a survey can be completed within a day, at a cost of >>>>> less than $50. (The hard work of designing the experiment, and >>>>> constructing controlled materials remains of course.) >>>>> >>>>> Sorry to be so verbose. But I think that these methodological points >>>>> are very important. >>>>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson >>>>> >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>>> methods in syntax and semantics research. Language and Cognitive >>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>>> & Fedorenko InPress LCP.pdf >>>>> >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko >>>>> 2010 TICS.pdf >>>>> >>>>> >>>>> >>>>> >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about methodology. I believe that >>>>>> intuitions are a fine way to generate hypotheses and even to test >>>>>> them - to a degree. But while it might not have been feasible for >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>>> Grammar to conduct experiments on every point of the grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, corpora, and standard psycholinguistic experimentation >>>>>> (indeed, Standard Social Science Methodology) is vital for taking the >>>>>> field forward and for providing the best support for different >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new >>>>>> paper on this, showing serious shortcomings with intuitions as the >>>>>> sole source of evidence, in their paper: "The need for quantitative >>>>>> methods in syntax and semantics research". >>>>>> >>>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>>> convincingly on this. >>>>>> >>>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive >>>>>> Science), and researchers from Brazil are beginning a third round of >>>>>> experimental work among the Pirahas, since my own work on the syntax >>>>>> was, like almost every other field researcher's, based on native >>>>>> speaker intuitions and corpora. >>>>>> >>>>>> The discussion of methodologies reminds me of the initial reactions >>>>>> to Greenberg's work on classifying the languages of the Americas. His >>>>>> methods were strongly (and justifiably) criticized. However, I always >>>>>> thought that his methods were a great way of generating hypotheses, >>>>>> so long as they were ultimately put to the test of standard >>>>>> historical linguistics methods. And the same seems true for use of >>>>>> native-speaker intuitions. >>>>>> >>>>>> -- Dan >>>>> >>>>> >>>>> >>>>>>> We linguists can add a further layer of explanation to the >>>>>>> judgements, but some judgements do seem to be more reliable than >>>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>>> every detailed analysis we make, our whole discipline will >>>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>>> judgements are what put us linguists ahead of the rest in handling >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the English >>>>>>> Language (or the OED) without using native speaker judgements. >>>>>>> >>>>>>> Best wishes, Dick Hudson >>>>> >>>>> >>>>> >>> >>> >>> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > From amnfn at well.com Sat Sep 11 13:00:08 2010 From: amnfn at well.com (A. Katz) Date: Sat, 11 Sep 2010 06:00:08 -0700 Subject: analysis: unhappiness In-Reply-To: <2147483647.1284143629@cast-dryerm2.caset.buffalo.edu> Message-ID: Matt, We appear to be fairly close in our approach, but I would have to add that sometimes there isn't a unique "correct" analysis, because the language allows equally for several different ones, and how any particular speaker analyzes a phrase out of context says more about how their individual brain is wired and less about the language. As an example, take the slogan the Coca-Cola company is currently using: "Open happiness". I first saw it on a cocktail napkin in flight. Reading the English slogan, my first analysis was that "open" was an adjective modifying the noun happiness, as opposed to say "closed happiness." That seemed weird, so I considered a few other possiblities. Maybe "open" is a verb in the imperative, and "happiness" is a proper noun in the vocative, as in "Open Sesame!" Then again, it could be that "Happiness" was just a proper noun in objective case: as in "open America (to tourism)." Then I read the French translation on the napkin: "Ouvrez du bonheur." "Oh! So this means "open some happiness"!" I said to myself. English lexemes are underspecified for category, which is why we need little words like "some" to disamnbiguate. That's how the language works. But... all those different analyses could have been correct, given the proper context, and experimenting even with a large population as to which one they thought of first would tell you less about the language and more about the people. The only analysis that seems a bit doubtful is the one suggested by the translation. Best, --Aya On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > Aya, > > I actually agree with everything you say here. Personally, I am MORE > interested in the communicative function of language than I am in > psycholinguistics and how people process language. > > But none of that is relevant, I believe, to the very specific question of > what it means for an analysis to be correct. While one might conclude from > what I said that one ought to do psycholinguistics, that is not my intention. > Rather, my conclusion is that since I myself prefer not to do > psycholinguistics, I cannot really claim that the analyses I come up with are > "the correct" ones. And if it is really important to someone that they > identify "correct" analyses, then they ought to be doing psycholinguistics, > since there is no coherent notion of correct analysis outside of what is > inside of people's heads. > > Matthew > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > wrote: > >> Matthew, >> >> Thanks for stating that, because I was almost beginning to imagine that >> there was no essential disagreement, and that all of us agree that there >> is more -- and less -- to language than what is found in people's heads. >> >> Your position is the one I am familiar with from the functionalist point >> of view, and I was beginning to feel that it was underrepresented on >> Funknet. >> >> Those of us who disagree with your stated position -- but are very >> familiar with it -- are interested not just in psycholinguistics and how >> people process language -- but also in the communicative function of >> language as a system whereby information is transferred. Just as you and >> I may not be aware of the way our emails are encoded and then decoded by >> the computers that help us send emails back and forth, speakers may be >> compeltely unaware of what language does in order to transmit information. >> >> After speakers have finished sending forth their linguistic output, it >> matters not at all how they arrived at this output. Language processing >> is separate from language in the same way that data processing is >> separate from data. >> >> Best, >> >> --Aya >> >> >> On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: >> >>> >>> The following sentence of Lise's >>> >>> "Since we don't in fact have such an agreed-on terminology, we have to >>> be quite careful in making clear what we are referring to when we talk >>> about 'the correct analysis' of a form like 'unhappiness'" >>> >>> suggests that there are two senses in which an analysis can be "the >>> correct analysis", one in terms of what is in people's heads, the other >>> in terms of what is "out there". >>> >>> There are a vast number of patterns "out there". The only distinction, >>> on my view, amongst this vast number of patterns, is between those that >>> reflect something inside people's heads and those that don't. But if >>> that is the case, then there is no coherent sense in which one can talk >>> of "the correct analysis" of what is "out there", except in terms of >>> what is in people's heads, and thus no second sense of "the correct >>> analysis". The patterns that don't correspond to things in people's >>> heads fall into (at least) two categories. There are those that are >>> akin to constellations of stars and, as with constellations, there is >>> no reality to these patterns, except in the minds of linguists. And >>> there are those patterns which are the fossil remains of what was in >>> the heads of speakers of an earlier stage of the language but which no >>> longer are. These latter patterns are real, and they are relevant to >>> exlaining why the language is now the way it is, but they are not >>> relevant, I think many would agree, as to what is the "correct analysis" >>> of the language today. >>> >>> For this reason, I claim that the only sense in which an analysis can be >>> "the correct analysis" is in terms of what is inside of people's heads. >>> >>> Again, I recommend the work of Bruce Derwing for lengthy discussion of >>> these issues. >>> >>> Matthew >>> >>> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn >>> wrote: >>> >>>> I wish we had better terminology for keeping track of whether, at a >>>> given time, we are talking about the patterns that are 'out there' in >>>> the language and might possibly be apprehended (subconsciously) by a >>>> speaker, and when we are talking about the patterns that a particular >>>> speaker actually does apprehend, as indicated by experiments, from >>>> simple 'wug tests' up to brain wave and eye-gaze studies. And for >>>> distinguishing among the degrees of pattern apprehension that a person >>>> may have, from vague preferences detectable in reaction times or other >>>> behavior all the way up through clear metalinguistic insights. Dick >>>> Hudson's note reminding us of the Gleitman and Gleitman study is right >>>> on target. >>>> >>>> Since we don't in fact have such an agreed-on terminology, we have to >>>> be quite careful in making clear what we are referring to when we talk >>>> about 'the correct analysis' of a form like 'unhappiness'. We know, >>>> but tend to forget - and tend to forget to tell our students! - that >>>> it's an empirical question as to whether the formal simplicity and >>>> coherence of description of forms 'out there' (e.g. lovely abstract >>>> morphophonemics) is any kind of approximation to the way knowledge of >>>> the same forms is organized in a particular person's head. If we >>>> remember that a very large proportion of what we know about our >>>> language is 'out there' when we are infants and has to be internalized >>>> through experience with the language (even if you believe in innate >>>> 'core language'), the variation in internal knowledge from one person >>>> to another is more understandable. We especially need to consider (and >>>> try to test) the >>>> possibility that >>>> since >>>> the brain can make multiple cross-connections, multiple patterns are >>>> involved >>>> simultaneously in morphological and syntactic analyses. I suggest that >>>> that's the case with 'unhappiness' - and the linguistic analyses that I >>>> know about are not good at handling that kind of idea. >>>> >>>> Lise >>>> >>>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: >>>> >>>>> >>>>> Two comments. >>>>> >>>>> First (elaborating perhaps on Dick Hudson's comment), I think there >>>>> is an >>>>> important distinction between low-level linguistic intuitions (like >>>>> whether a >>>>> word or sentence is well-formed or what it means) and higher-level >>>>> intuitions >>>>> (like what the structure of a word or sentence is). One can take >>>>> the position >>>>> that we need to account for the former (while recognizing that they >>>>> are not >>>>> always reliable) but not the latter. >>>>> >>>>> Second, the tension here is not only between evidence from speaker >>>>> intuitions >>>>> versus evidence from psycholinguistic experiments. There is also a >>>>> tension >>>>> between deciding on the correct analysis on the basis of a priori >>>>> simplicity >>>>> arguments versus deciding on the correct analysis on the basis of >>>>> psycholinguistic evidence (see Derwing 1973). The bracketing >>>>> paradox that Dan >>>>> referred to that arises with the word (semantics argues >>>>> for >>>>> [[un+happi] + er], morphology and phonology argues for [un + [happi >>>>> +er]] (the >>>>> comparative suffix can only be attached to adjectives containing one >>>>> or two >>>>> syllables) is only a paradox if one assumes that speakers adopt the >>>>> simplest >>>>> analysis. For example, if speakers adopt a more complex rule for >>>>> either of these >>>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to >>>>> trisyllabic >>>>> words beginning with un-), then the bracketing paradox disappears. >>>>> >>>>> Matthew >>>>> >>>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: >>>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that >>>>>> conscious judgements are only one kind of evidence that we need to >>>>>> take >>>>>> into account. I admire Carson Schutze's work (which I reviewed in >>>>>> fact), >>>>>> and of course I've been aware of complaints about judgements by >>>>>> people >>>>>> like Labov for decades. >>>>>> >>>>>> But you're missing my main point, which is that all judgements aren't >>>>>> equally reliable. If you want to know how /unhappiness/ is >>>>>> structured, >>>>>> ask a linguist, not a five-year old. And one of the by-products of >>>>>> education may be increased sensitivity to syntax - which is one of >>>>>> the >>>>>> many reasons why linguists need to pay more attention to education. >>>>>> >>>>>> Best wishes, Dick >>>>>> >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>>>> On 09/09/2010 11:39, Daniel Everett wrote: >>>>>>> Dick, >>>>>>> >>>>>>> You raise an important issue here about >>>>>> methodology. I believe that intuitions are a fine way to generate >>>>>> hypotheses and even to test them - to a degree. But while it might >>>>>> not have >>>>>> been feasible for Huddleston, Pullum, and the other contributors to >>>>>> the >>>>>> Cambridge Grammar to conduct experiments on every point of the >>>>>> grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, >>>>>> corpora, and standard psycholinguistic experimentation (indeed, >>>>>> Standard >>>>>> Social Science Methodology) is vital for taking the field forward >>>>>> and for >>>>>> providing the best support for different analyses. Ted Gibson and Ev >>>>>> Fedorenko have written a very useful new paper on this, showing >>>>>> serious >>>>>> shortcomings with intuitions as the sole source of evidence, in their >>>>>> paper: "The need for quantitative methods in syntax and semantics >>>>>> research".> >>>>>>> Carson Schutze and Wayne Cowart, among others, >>>>>> have also written convincingly on this.> >>>>>>> It is one reason that a team from Stanford, MIT >>>>>> (Brain and Cognitive Science), and researchers from Brazil are >>>>>> beginning a >>>>>> third round of experimental work among the Pirahas, since my own >>>>>> work on >>>>>> the syntax was, like almost every other field researcher's, based >>>>>> on native >>>>>> speaker intuitions and corpora.> >>>>>>> The discussion of methodologies reminds me of >>>>>> the initial reactions to Greenberg's work on classifying the >>>>>> languages of >>>>>> the Americas. His methods were strongly (and justifiably) criticized. >>>>>> However, I always thought that his methods were a great way of >>>>>> generating >>>>>> hypotheses, so long as they were ultimately put to the test of >>>>>> standard >>>>>> historical linguistics methods. And the same seems true for use of >>>>>> native-speaker intuitions.> >>>>>>> -- Dan >>>>>>> >>>>>>>> We linguists can add a further layer of >>>>>> explanation to the judgements, but some judgements do seem to be more >>>>>> reliable than others. And if we have to wait for psycholinguistic >>>>>> evidence >>>>>> for every detailed analysis we make, our whole discipline will >>>>>> immediately >>>>>> grind to a halt. Like it or not, native speaker judgements are what >>>>>> put us >>>>>> linguists ahead of the rest in handling fine detail. Imagine >>>>>> writing the >>>>>> Cambridge Grammar of the English Language (or the OED) without >>>>>> using native >>>>>> speaker judgements.>> >>>>>>>> Best wishes, Dick Hudson >>>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>> >>>> Lise Menn Home Office: 303-444-4274 >>>> 1625 Mariposa Ave Fax: 303-413-0017 >>>> Boulder CO 80302 >>>> >>>> Professor Emerita of Linguistics >>>> Fellow, Institute of Cognitive Science >>>> University of Colorado >>>> >>>> Secretary, AAAS Section Z [Linguistics] >>>> >>>> Campus Mail Address: >>>> UCB 594, Institute of Cognitive Science >>>> >>>> Campus Physical Address: >>>> CINC 234 >>>> 1777 Exposition Ave, Boulder >>>> >>>> >>>> >>>> >>> >>> >>> >>> >>> >> >> > > > > > From cbutler at ntlworld.com Sat Sep 11 11:17:29 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Sat, 11 Sep 2010 12:17:29 +0100 Subject: FUNKNET] analysis: unhappiness Message-ID: Dick's comment that "this discussion raises the really fundamental question of what kind of thing we think language is: social or individual" is, it seems to me, an important one, particularly for those of us who are committed functionalists. My own view is that a truly functional model of language would be one which aims to account for how human beings communicate using language, or in other words tries to answer the question which was posed by Simon Dik a long time ago now, but which was not tackled head-on in his own work: "How does the natural language user work?' In trying to answer this question we need to accept that language is BOTH social AND individual, and we need to explore both aspects to get as complete a picture as possible of how we communicate using language. We need to know BOTH how people create and respond to meanings and express those meanings in forms during social interaction AND the mechanisms which operate in the brains of individuals in order to make such interaction possible. Both are important parts of the answer to the question 'How do we communicate using language?', though this particular thread of the Funknet discussion has concentrated more on the second aspect, and so will I. This doesn't mean that all the work linguists have done on "exploring the structure of a language so that I can understand how all the bits fit together" and "exploring the connections between items", as Dick puts it, is useless - far from it. After all, the hypotheses that psycholinguists test are based on ideas about what languages are like. But it does mean, in my view, that ultimately we need to get evidence that the constructs and analyses we propose are ones that are at least consistent with what we know of the processes which go on when we use language. So I am with Matthew when he says that for him, "the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads". Of course, this doesn't imply that linguists should just give up their jobs until such time as we know everything there is to know about language processing. But it does mean that we need to collaborate with psycholinguists, psychologists and neurologists, as has also been pointed out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more with sociolinguists and sociologists, so that we can get a better handle on the sociocultural aspects of how we communicate.] And it also means that psycholinguists, for their part, need whenever possible to follow up tightly controlled lab experiments with studies under more naturalistic conditions, to avoid the criticism that what happens in artifical lab situations may not happen in natural communicative conditions. I also agree with Dick when he says that "the differences between individuals really matter", and with Lise when she points out that "we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover". However, there are surely processing mechanisms which are common to all language users by virtue of the evolution of the language faculty and which constitute the "general processes" which Dick says psycholinguists are interested in. On the issue of quantitative methodology, I'm sympathetic in general to Ted and Ev's views, though it does seem sensible to prioritise cases in terms of a hierarchy such as Brian suggests. One thing this means is that we should be giving our university students of linguistics (and some of our linguistics lecturers!) courses in quantitative aspects of linguistics that introduce them to the use of at least some of the basic statistical methods in language study, and I'm sure this is indeed going on in some enlightened places. To those who suspect this can't be done with maths-shy students who don't initially see the need for it, I offer my own experience, over quite a long period, of teaching such courses to people with little or no prior experience in quantitative techniques. For some years in the 1990s, I taught such courses to all linguistics students in an institution where we had many mature students who had come into university level studies with non-standard qualifications, and were not well equipped for courses of this kind by their previous experience. I'm glad to say that teaching the subject from their own perspective as language students rather than that of the statistician, and explaining the reasons for doing things in particular ways rather than just presenting formulae, paid off in the end, so that most students were able to appreciate the relevance of these courses and to turn in very creditable projects showing an understanding of research design and competence in the use of a range of basic statistical techniques. And I still find that bright graduate students respond well to similar courses which incorporate some of the rather more advanced techniques needed for many real research projects in various areas of linguistics. But I may well be out of date with what is now already happening in our fine institutions of higher education! Chris Butler From amnfn at well.com Sat Sep 11 14:18:29 2010 From: amnfn at well.com (A. Katz) Date: Sat, 11 Sep 2010 07:18:29 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <24693E13135D4D6AAD5A098FAA952480@OwnerPC> Message-ID: The problem is that once we achieve the psycholinguistic goal and see what is happening in each person's head, and we see that conflicting analyses are the norm, rather than the exception, among normal language speakers, then we will realize that the way language works to transmit information is despite individual differences and not because of uniform processing strategies. Even when all are agreed as to the meaning of an utterance, they do not process it the same way. Which means that processing is seocndary to information transmission. --Aya On Sat, 11 Sep 2010, Chris Butler wrote: > Dick's comment that "this discussion raises the really fundamental question of what kind of thing we think language is: social or individual" is, it seems to me, an important one, particularly for those of us who are committed functionalists. My own view is that a truly functional model of language would be one which aims to account for how human beings communicate using language, or in other words tries to answer the question which was posed by Simon Dik a long time ago now, but which was not tackled head-on in his own work: "How does the natural language user work?' In trying to answer this question we need to accept that language is BOTH social AND individual, and we need to explore both aspects to get as complete a picture as possible of how we communicate using language. We need to know BOTH how people create and respond to meanings and express those meanings in forms during social interaction AND the mechanisms which operate in the brains of individuals in order to make such interaction possible. Both are important parts of the answer to the question 'How do we communicate using language?', though this particular thread of the Funknet discussion has concentrated more on the second aspect, and so will I. > > This doesn't mean that all the work linguists have done on "exploring the structure of a language so that I can understand how all the bits fit together" and "exploring the connections between items", as Dick puts it, is useless - far from it. After all, the hypotheses that psycholinguists test are based on ideas about what languages are like. But it does mean, in my view, that ultimately we need to get evidence that the constructs and analyses we propose are ones that are at least consistent with what we know of the processes which go on when we use language. So I am with Matthew when he says that for him, "the only sense in which an analysis can be "the correct analysis" is in terms of what is inside of people's heads". Of course, this doesn't imply that linguists should just give up their jobs until such time as we know everything there is to know about language processing. But it does mean that we need to collaborate with psycholinguists, psychologists and neurologists, as has also been pointed out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more with sociolinguists and sociologists, so that we can get a better handle on the sociocultural aspects of how we communicate.] And it also means that psycholinguists, for their part, need whenever possible to follow up tightly controlled lab experiments with studies under more naturalistic conditions, to avoid the criticism that what happens in artifical lab situations may not happen in natural communicative conditions. > > I also agree with Dick when he says that "the differences between individuals really matter", and with Lise when she points out that "we must also be careful not to idealize "what's in people's heads" as if it were a single coherent construct that we are trying to discover". However, there are surely processing mechanisms which are common to all language users by virtue of the evolution of the language faculty and which constitute the "general processes" which Dick says psycholinguists are interested in. > > On the issue of quantitative methodology, I'm sympathetic in general to Ted and Ev's views, though it does seem sensible to prioritise cases in terms of a hierarchy such as Brian suggests. One thing this means is that we should be giving our university students of linguistics (and some of our linguistics lecturers!) courses in quantitative aspects of linguistics that introduce them to the use of at least some of the basic statistical methods in language study, and I'm sure this is indeed going on in some enlightened places. To those who suspect this can't be done with maths-shy students who don't initially see the need for it, I offer my own experience, over quite a long period, of teaching such courses to people with little or no prior experience in quantitative techniques. For some years in the 1990s, I taught such courses to all linguistics students in an institution where we had many mature students who had come into university level studies with non-standard qualifications, and were not well equipped for courses of this kind by their previous experience. I'm glad to say that teaching the subject from their own perspective as language students rather than that of the statistician, and explaining the reasons for doing things in particular ways rather than just presenting formulae, paid off in the end, so that most students were able to appreciate the relevance of these courses and to turn in very creditable projects showing an understanding of research design and competence in the use of a range of basic statistical techniques. And I still find that bright graduate students respond well to similar courses which incorporate some of the rather more advanced techniques needed for many real research projects in various areas of linguistics. But I may well be out of date with what is now already happening in our fine institutions of higher education! > > Chris Butler > > From eitan.eg at gmail.com Sat Sep 11 14:54:58 2010 From: eitan.eg at gmail.com (E.G.) Date: Sat, 11 Sep 2010 17:54:58 +0300 Subject: "Relative clauses" with no relativized ele In-Reply-To: Message-ID: Hi all, I'd like to thank Ron for pointing out the incompleteness of what I wrote, and for the reference to his article. I hope I didn't give the impression I was trying to give a complete description of noun complementation in Modern Hebrew in an email. What I *was* trying to say is that in some languages, unlike English, some nominalizations (of utterance and cognition verbs, as Ron points out, but also of some perception verbs too) can occur with a construction that is explicitly and unmistakably marked as a complement clause. Moreover, the nominalizations that take these explicit complement clauses are related to verbs that can take the same type of complement clause. In such languages, then, it's pretty clear that these instances involve complement clauses. It doesn't mean that other noun complementation strategies don't exist for other types of nouns. However, my main point was more general, albeit poorly expressed. It's that we can turn to cross-linguistic comparison in order to try to reach generalizations about how languages encode meaning. These generalizations are useful, because they can be used to ask "why" questions. For example, it's not really possible to ask *why* Hebrew has two distinct strategies for noun complementation, how *why* English has one. That's because it could always be otherwise, and language change can alter the picture (and has!). However, if we find that in languages that have two strategies, one is limited to nominalizations of PCU verbs, then we have the beginnings of a hierarchy that is amenable to functional explanation. Anyway, it seems that Thomas Payne's question has turned up a pretty general consensus that these constructions are complement clauses. Best wishes, Eitan On 10 September 2010 23:26, Ron Kuzar wrote: > The Modern Hebrew data supplied by Eitan are incomplete. > Hebrew distinguishes between locution (say, hear, think, etc.) and > situation (action, event, state, etc.). > What Eitan describes is only true with regard to nouns (and clauses) > expressing locution. 'Announcement' is indeed such a noun. > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > etc., whose denotatum is a situation, cannot be followed by ki, but only > by Se-, e.g.: > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > annoys me the-situation that-all went > 'I am upset about the situation that all have gone' > > On the other hand, the relative Se- may be replaced by the more > elegant and classical aSer, while the Se- of situation clauses may not. > Sorry about the invented example. I am overseas now. > All this has been described (with corpus data) in: > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit > [Hebrew > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > The article is somewhat outdated and contains some inaccuracies I would > formulate differently today, but the basic distinction is valid in my > opinion. > Best, > Ron Kuzar > --------------- > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > Hi all, > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > comparison can help what seems to be a purely theoretical question based > on > > a single language. The problem here is that English uses the same element > > to > > mark regular relatives and these "appositional" relatives. But if at > least > > one language encodes them by different means, then there's at least a > good > > case for seeing them as distinct functions. It's basically the same > > principle that's used to decide whether to put a meaning on a semantic > map. > > So here are two languages that I know that encode them differently. > > > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > ha-hoda'a Se-kibalnu > > the-announcment rel-we_got > > "The announcement that we got." > > > > ha-hoda'a ki hitbatel ha-mifgaS > > the-message CMP was_cancelled the-meeting > > "The announcement that the meeting was cancelled." > > > > In Coptic, these clauses are marked by ce-, which marks complement > clauses, > > *inter alia*, but not relative clauses: > > > > ph-mewi ce- (complement clause) > > 'the-thought that (we are angry)' > > > > ph-mewi ete- (relative clause) > > 'the thought that (we used to think)' > > > > This seems to be a pretty clear indication that these are complement > > clauses > > rather than relatives. Even if one doesn't like the notion of nouns > taking > > complement clauses (and why not? nominalizations in some languages can > take > > accusative modifiers as well as genitives), it still probably isn't > > incidental that the nominalizations are of verbs that take complement > > clauses when finite. > > > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > > looking at. > > > > Best, > > Eitan > > > > > > On 10 September 2010 19:21, Arie Verhagen > > wrote: > > > > > And as another addition: the clauses that can only be introduced by > > *that* > > > (with no role to > > > play in the subordinate clause) may be seen as (subtypes of) complement > > > clauses, > > > expressing a proposition with the noun functioning as Complement Taking > > > Predicate (CTP), > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > (following analyses by > > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > > Cf. > > > constructions like > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > claim > > > that X", "I put forward > > > the claim that X", in which the relationship between the verb or noun > and > > > the that-clause is > > > comparable to the one in "The claim that X". > > > > > > --Arie Verhagen > > > > > > ---------------- > > > Message from Rong Chen > > > 10 Sep 2010, 23:42 > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > To add to Joanne's comments: > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > (AC) from a relative clause (RC). > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > pronouns. > > > > > > > > 2) The AC and the noun it modifies display an equative > > relationship--one > > > can say X > > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > > often doesn't > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > > relative pronoun > > > > is always part of the clause in an RC. > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > -- > > Eitan Grossman > > Martin Buber Society of Fellows > > Hebrew University of Jerusalem > > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > -- Eitan Grossman Martin Buber Society of Fellows Hebrew University of Jerusalem From khildeb at siue.edu Sat Sep 11 15:19:05 2010 From: khildeb at siue.edu (Kristine Hildebrandt) Date: Sat, 11 Sep 2010 10:19:05 -0500 Subject: Job Advertisement Message-ID: Dear Funknetters: Please distribute this job advertisement to interested colleagues and doctoral students completing their degrees. *HIRING UNIT*: *DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE* ** *TITLE/RANK*: Assistant Professor of English-Linguistics *DESCRIPTION OF DUTIES*: The Department of English Language and Literature invites applications for a tenure-track position in general linguistics, with secondary specialization in applied linguistics. The candidate will teach courses in the MA TESL program, along with undergraduate courses in linguistics, composition (ESL and regular), and some general education courses. Academic year: 3/3 load. *TERMS OF APPOINTMENT*: Academic, tenure-track beginning August 16, 2011, 100% appointment. *SOURCE OF FUNDS*: State *SALARY RANGE*: commensurate with training and experience *QUALIFICATIONS REQUIRED: *A Ph.D. in Linguistics, Applied Linguistics, or related field required. If Ph.D. is not completed by the beginning of the contract period, appointment will be at the rank of Instructor until all degree requirements are fulfilled. A record of ESL and/or TESL experience is desirable. *CLOSING DATE FOR APPLICATIONS*: Position open until filled; completed applications postmarked by November 15, 2010 will have priority. Possible interviews at LSA in January 2011. * ** **SEND COVER LETTER, VITA, UNOFFICIAL TRANSCRIPT, STATEMENT OF TEACHING PHILOSOPHY AND RESEARCH AGENDA, AND THREE LETTERS OF RECOMMENDATION TO:** * Linguistics Search Committee Southern Illinois University Edwardsville Department of English Language & Literature Campus Box 1431 Edwardsville, IL 62026-1431 NOTE: Electronic applications will not be accepted for this position. SIUE is a state university-benefits under state sponsored plans may not be available to holders of F1 or J1 visas. Applicants may be subject to a background check prior to an offer of employment. SIUE is an affirmative action and equal opportunity employer. The SIUE ANNUAL SECURITY REPORT is available on-line at: http://admin.siu.edu/studentrightto/. The report contains safety and security information and crime statistics for the past three (3) calendar years. This report is published in compliance with Federal law, entitled *the ?Jeanne Clery Disclosure of Campus Security Policy and Campus Crime Statistics Act.? * You may also access this report through the SIUE Home Page: http://www.siue.edu under *Ready References, Quick Links or Publications/Reports*. For those without computer access, a paper copy of the report may be obtained from the Office of the Vice Chancellor for Administration, Rendleman Hall, Room 2228. -- *Kristine A. Hildebrandt* *Assistant Professor, Department of English Language & Literature Southern Illinois University Edwardsville* *Box 1431 Edwardsville, IL 62026 U.S.A. 618-650-3380 (office)* *khildeb at siue.edu http://www.siue.edu/~khildeb* From egibson at MIT.EDU Sat Sep 11 15:45:27 2010 From: egibson at MIT.EDU (Ted Gibson) Date: Sat, 11 Sep 2010 11:45:27 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: Dear Brian, Dick, Dan et al: Thanks for the discussion. Here are a few responses: 1. Brian: "But I understand Dick's worry about how far Gibson and Fedorenko are trying to push this. Neither their email nor their paper sets clear limits on what we should be testing and we certainly don't want to waste time checking out go-goed-went. So, Gibson and Fedorenko owe us those clarifications." The answer that we give to this question in Gibson & Fedorenko (in press) is as follows (the final paragraph in the paper): "Finally, a question that is often put to us is whether it is necessary to evaluate every empirical claim quantitatively. A major problem with the fields of syntax and semantics is that many papers include no quantitative evidence in support of their research hypotheses. Because conducting experiments is now so easy to do with the advent of Amazon.com?s Mechanical Turk, we recommend gathering quantitative evidence for all empirical claims. However, it would clearly be a vast improvement to the field for all research papers to include at least some quantitative evidence evaluating their research hypotheses." Another possible answer to this question is: the more important some observation is, the better your evidence should be. If the observation is a key reason for some important theoretical claim, then there should be solid quantitative data supporting that observation. In practice, once a linguist starts gathering quantitative data, s/he will realize (a) how easy it is to do; and (b) how beneficial the methods are, with the consequence that these researchers will probably do most or all of their work quantitatively in the future. 2. Dick (by the way, thank you for the kind responses, and your positive tone): "Your [the psycholinguists'] goal is to find general processes and principles that apply uniformly across individuals, so you have to use methods to check for generality." in contrast to "my focus is on items and structures, and I start from the assumption that these can and do vary across speakers." Many cognitive psychologists / cognitive scientists (all the ones I know at MIT for example) are interested in both cognitive generalizations across people and ways in which people differ cognitively. In fact, some methods (e.g., the individual differences approach where co-variation of various behaviors / characteristics is examined across individuals) have been specifically developed to study differences among individuals. Both kinds of data are important for understanding human cognition, including language. This applies to language research directly: generalizations across people are important, but so are individual differences. In either case, quantitative data are necessary to evaluate research questions and test hypotheses. On a related note, it is a mistake to characterize researchers with a background in "psychology" or cognitive science as being interested in "processing", and researchers with a background in "linguistics" as being interested in "knowledge" or "representation / structure". Both psychologists and linguists should be interested in *both* representation and processing (and learning, for that matter). We wrote a little about this confusion in Gibson & Fedorenko (in press), which we include at the end of the message. This leads to something that Dan said: 3. Dan says: "linguistics is not simply a subdiscipline of psychology" Both linguistics and psychology are big fields. We assume Dan is referring to cognitive psychology / cognitive science here. (Of course, there are sub-fields of psychology - e.g., personality psychology or abnormal psychology - which are somewhat distinct from linguistics, but those sub-fields are also distinct from cognitive psychology.) It is true that historically linguistics is not treated as a subfield of cognitive psychology / cognitive science. However, key research questions in linguistics (i.e., the form of the knowledge structures and algorithms underlying human language) are indeed a subset of those investigated by cognitive psychologists / cognitive scientists. We think that the biggest factor separating linguistics from psychology is the methods used to explore the research questions, rather than the research questions themselves. Consequently, we would like to continue to see tighter connections among the fields of psychology / cognitive science, linguistics, as well as other fields like anthropology and computer science. Thanks to all for the interesting discussions. Ted & Ev We have encountered a claim that the reason for different kinds of methods being used across the different fields of language study (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the research questions are different across these fields, and some methods may be better suited to ask some questions than others. Although the latter is likely true, the premise ? that the research questions are different across the fields ? is false. The typical claim is that researchers in the field of linguistics are investigating linguistic representations, and researchers in the fields of psycho-/neuro- linguistics are investigating the computations that take place as language is understood or produced. However, many researchers in the fields of psycho-/neuro-linguistics are also interested in the nature of the linguistic representations (at all levels; e.g., phonological representations, lexical representations, syntactic representations, etc.) [1]. By the same token, many researchers in the field of linguistics are interested in the computations that take place in the course of online comprehension or production. However, inferences ? drawn from any dependent measure ? about either the linguistic representations or computations are always indirect. And these inferences are no more indirect in reading times or event-related potentials, etc., than in acceptability judgments: across all dependent measures we take some observable (e.g., a participant?s rating on an acceptability judgment task or the time it took a participant to read a sentence) and we try to infer something about the underlying cognitive representations / processes. More generally, methods in cognitive science are often used to jointly learn about representations and computations, because inferences about representations can inform questions about the computations, and vice versa. For example, certain data structures can make a computation more or less difficult to perform, and certain representations may require assumptions about the algorithms being used. In our opinion then, the distinction between the fields of linguistics and psycho-/neuro-linguistics is purely along the lines of the kind of data that are used as evidence for or against theoretical hypotheses: typically non-quantitative data in linguistics vs. typically quantitative data in psycho-/neuro-linguistics. Given the superficial nature of this distinction, we think that there should be one field of language study where a wide range of dependent measures is used to investigate linguistic representations and computations. [1] In fact, some methods in cognitive science and cognitive neuroscience were specifically developed to get at representational questions (e.g., lexical / syntactic priming methods, neural adaptation or multi-voxel pattern analyses in functional MRI). On Sep 10, 2010, at 9:05 PM, Daniel Everett wrote: > I think that Brian and Dick make excellent points. There are very > good grammars written that could be improved by psycholinguistic > experimentation and more quantitative approaches. But large sections > of those grammars aren't going to change one bit (go-went) with > quantitative tests and such tests would be completely > counterproductive given the shortness of life and the vastness of > the field linguist's tasks. > > Part of the problem is that linguistics is not simply a > subdiscipline of psychology. Linguistics has its own objectives and > those only occasionally overlap with psychology. The same for methods. > > On another note, I don't buy the 'in my head' 'out of my head' > distinction either (that Matt seems to be urging upon us). We study > different things and have different reasons for being satisfied with > the results we achieve. > > I believe that we linguists are often complacent and fail to apply > better methods. But of course that applies to all disciplines. > > In the meantime, checking corpora, collecting data as a result of > careful interviews with native speakers, and the other aspects of > the field linguist's task are vital parts of the linguist's task and > much of this won't be improved by quantitative methods as we > currently understand them. Maybe sometime. > > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that > they showed the danger of using intuitions. What I meant to say of > using intuitions as standardly used by linguists. They convinced me > that there is a lot to learn from quantitative methods. > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > >> Dear Ted and Ev, >> Yes, I understand your view, but I think it's a psycholinguist's >> view. Your goal is to find general processes and principles that >> apply uniformly across individuals, so you have to use methods to >> check for generality. And (as you know) I admire the way you pursue >> that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how >> all the bits fit together. Like you, I'm aiming to model cognition, >> but my focus is on items and structures, and I start from the >> assumption that these can and do vary across speakers. >> >> However, having said all that I do agree with you that linguists >> should all get used to collecting and using quantitative data; and >> with the help of Brian MacWhinney's typology we'd know what methods >> to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good >> starting point for a proper investigation. >> >> Best wishes, Dick >> >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >> >> On 10/09/2010 19:30, Ted Gibson wrote: >>> Dear Dick: >>> >>> Perhaps we are talking at cross purposes. I don't understand what >>> is confusing about what Ev Fedorenko and I are claiming. All we >>> are saying is that if you have some testable claim involving a >>> general hypothesis about a language, then you need to get >>> quantitative data from unbiased sources to evaluate that claim. If >>> you are interested in English past tense morphology, then >>> depending on the claims that you might want to investigate, there >>> are lots of ways to get relevant quantitative evidence. Corpus >>> data will probably be useful. For very low frequency words, you >>> can run experiments to test behavior with respect to such words. >>> >>> Your example of the past tense of "bid" is a fine such example. >>> You can run an experiment like the one you suggested to find out >>> what people think the past tense is. If you then found that 20/50 >>> people responded "bidded" and 30/50 respond "bid", that is a lot >>> of useful information. As you suggest in your discussion, this >>> result wouldn't answer the question of how past tense is stored in >>> each individual. This result would be ambiguous among several >>> possible explanations. One possibility is that the probability >>> distribution that is being discovered reflects different dialects, >>> such that 2/5 of the population has one past tense, and 3/5 has >>> another. Another possibility is that each person has a similar >>> probability distribution in their heads, such that 2/5 of the time >>> I respond one way, and 3/5 of the time I respond another. Further >>> experiments would be necessary to answer between these and other >>> possible theories (e.g., with repeated trials from the same >>> person, carefully planned so that the participants don't notice >>> that they are being asked multiple times). Without the >>> quantitative evidence in the first place, there is no way to >>> answer these kinds of questions. >>> >>> Regarding the past tense of "go", this would be useful as a >>> baseline in an experiment involving the less frequent ones. So, >>> yes, it would useful to gather quantitative evidence in such a >>> case also, as baselines with respect to the more interesting cases >>> for theories. >>> >>> The bottom line: if you have a generalization about a language >>> that you wish to evaluate (such that you hypothesize that it is >>> true across the speakers of the language), then you need >>> quantitative evidence from multiple individuals, using an unbiased >>> data collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at >>> least for languages like English. >>> >>> Best wishes, >>> >>> Ted Gibson & Ev Fedorenko >>> >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: >>> >>>> Dear Ted, >>>> Thanks for the very interesting comment, but are you REALLY >>>> saying that I shouldn't claim, for example, that the past tense >>>> of GO is "went" without first cross-checking with 50 native >>>> speakers? >>>> >>>> Isn't there a danger of missing the point that we all, as native >>>> speakers, spend our whole lives scanning other people's >>>> linguistic behaviour (language 'out there', E-language) and >>>> trying to explain it to ourselves in terms of a language system >>>> (language 'in here', I-language)? So every judgement we make is >>>> based on thousands or millions of observed exemplars, and >>>> reflects a unique experience of E-language filtered through a >>>> unique I-language. >>>> >>>> Given that view of language development, I don't see how >>>> quantitative data will help. Let's take a real uncertainty, such >>>> as the past tense of BID. If I want to say I did it, do I say "I >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" >>>> and 30 "bid". So what? Does that mean that the correct answer is >>>> "bidded"? Surely not. How is it better than my judgement? I agree >>>> you could record my speech and find how often I use each >>>> alternative; but the reason I don't know is precisely because >>>> it's a rare word, so in a sense quantitative data are irrelevant >>>> even there. What would solve the problem of subjectivity, of >>>> course, would be a machine for probing the bit of my mind (or >>>> even brain) that holds BID and its details; but I suspect that >>>> even that wouldn't move us much further forward than my original >>>> "don't know". (Incidentally I write as a fan of quantitative >>>> sociolinguistics, so I do accept that quantitative data are >>>> relevant to linguistic analysis in some areas, where the I- >>>> language phenomenon is frequent enough to produce usable data.) >>>> >>>> It seems to me that this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual. The problem isn't unique to linguistics of course; >>>> it's the same throughout the social sciences. But what's special >>>> about linguistics is that we deal in very fine details of culture >>>> (e.g. details of how a particular word is used or pronounced) so >>>> the differences between individuals really matter. I don't see >>>> that we're ever going to have anything better than judgements to >>>> go on, so what we need is a way to ensure that judgements are >>>> accurate reports of individual I-language. A rotten situation for >>>> a science, but I don't see how it can get better. >>>> >>>> Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm >>>> >>>> On 10/09/2010 14:03, Ted Gibson wrote: >>>>> Dear Dan, Dick: >>>>> >>>>> I would like to clarify some points that Dan Everett makes, in >>>>> response to Dick Hudson. >>>>> >>>>> Ev Fedorenko and I have written a couple of papers recently >>>>> (Gibson & >>>>> Fedorenko, 2010, in press, see references and links below) on >>>>> what we >>>>> think are weak methodological standards in syntax and semantics >>>>> research over the past many years. The issue that we address is >>>>> the >>>>> prevalent method in syntax and semantics research, which involves >>>>> obtaining a judgment of the acceptability of a sentence / meaning >>>>> pair, typically by just the author of the paper, sometimes with >>>>> feedback from colleagues. As we address in our papers, this >>>>> methodology does not allow proper testing of scientific hypotheses >>>>> because of (a) the small number of experimental participants >>>>> (typically one); (b) the small number of experimental stimuli >>>>> (typically one); (c) cognitive biases on the part of the >>>>> researcher >>>>> and participants; and (d) the effect of the preceding context >>>>> (e.g., >>>>> other constructions the researcher may have been recently >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and >>>>> several others cited in Gibson & Fedorenko, in press; for similar >>>>> points, but with not as strong a conclusion as ours). >>>>> >>>>> Three issues need to be separated here: (1) the use of intuitive >>>>> judgments as a dependent measure in a language experiment; (2) >>>>> potential cognitive biases on the part of experimental subjects >>>>> and >>>>> experimenters in language experiments; and (3) the need for >>>>> obtaining >>>>> quantitative evidence, whatever the dependent measure might be. >>>>> The >>>>> paper that Ev and I wrote addresses the last two issues, but >>>>> does not >>>>> go into depth on the first issue (the use of intuitions as a >>>>> dependent >>>>> measure in language experiments). Regarding this issue, we don't >>>>> think >>>>> that there is anything wrong with gathering intuitive judgments >>>>> as a >>>>> dependent measure, as long as the task is clear to the >>>>> experimental >>>>> participants. >>>>> >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to >>>>> some >>>>> arguments that have been given in support of continuing to use the >>>>> traditional non-quantitative method in syntax / semantics >>>>> research. >>>>> One recent defense of the traditional method comes from Phillips >>>>> (2008), who argues that no harm has come from the non-quantitative >>>>> approach in syntax research thus far. Phillips argues that there >>>>> are >>>>> no cases in the literature where an incorrect intuitive judgment >>>>> has >>>>> become the basis for a widely accepted generalization or an >>>>> important >>>>> theoretical claim. He therefore concludes that there is no >>>>> reason to >>>>> adopt more rigorous data collection standards. We challenge >>>>> Philips? >>>>> conclusion by presenting three cases from the literature where a >>>>> faulty intuition has led to incorrect generalizations and mistaken >>>>> theorizing, plausibly due to cognitive biases on the part of the >>>>> researchers. >>>>> >>>>> A second argument that is sometimes presented for the continued >>>>> use of >>>>> the traditional non-quantitative method is that it would be too >>>>> inefficient to evaluate every syntactic / semantic hypothesis or >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff >>>>> (2010) >>>>> make this argument explicitly in their response to Gibson & >>>>> Fedorenko >>>>> (2010): ?It would cripple linguistic investigation if it were >>>>> required >>>>> that all judgments of ambiguity and grammaticality be subject to >>>>> statistically rigorous experiments on naive subjects, especially >>>>> when >>>>> investigating languages whose speakers are hard to >>>>> access? (Culicover >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point >>>>> earlier in the discussion here.) Whereas we agree that in >>>>> circumstances where gathering data is difficult, some evidence is >>>>> better than no evidence, we do not agree that research would be >>>>> slowed >>>>> with respect to languages where experimental participants are >>>>> easy to >>>>> access, such as English. In contrast, we think that the opposite >>>>> is >>>>> true: the field?s progress is probably slowed by not doing >>>>> quantitative research. >>>>> Suppose that a typical syntax / semantics paper that lacks >>>>> quantitative evidence includes judgments for 50 or more >>>>> sentences / >>>>> meaning pairs, corresponding to 50 or more empirical claims. >>>>> Even if >>>>> most of the judgments from such a paper are correct or are on the >>>>> right track, the problem is in knowing which judgments are >>>>> correct. >>>>> For example, suppose that 90% of the judgments from an arbitrary >>>>> paper >>>>> are correct (which is probably a high estimate). (Colin Phillips >>>>> and >>>>> some of his former students / postdocs have commented to us >>>>> that, in >>>>> their experience, quantitative acceptability judgment studies >>>>> almost >>>>> always validate the claim(s) in the literature. This is not our >>>>> experience, however. Most experiments that we have run which >>>>> attempt >>>>> to test some syntactic / semantic hypothesis in the literature >>>>> end up >>>>> providing us with a pattern of data that had not been known >>>>> before the >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This >>>>> means >>>>> that in a paper with 50 empirical claims 45/50 are correct. But >>>>> which >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s >>>>> over >>>>> two million different theories. By quantitatively evaluating the >>>>> empirical claims, we reduce the uncertainty a great deal. To make >>>>> progress, it is better to have theoretical claims supported by >>>>> solid >>>>> quantitative evidence, so that even if the interpretation of the >>>>> data >>>>> changes over time as new evidence becomes available ? as is >>>>> often the >>>>> case in any field of science ? the empirical pattern can be used >>>>> as a >>>>> basis for further theorizing. >>>>> >>>>> Furthermore, it is no longer expensive to run behavioral >>>>> experiments, >>>>> at least in English and other widely spoken languages. There now >>>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? >>>>> which >>>>> can be used for collecting behavioral data over the internet >>>>> quickly >>>>> and inexpensively. The cost of using an interface like this is >>>>> minimal, and the time that it takes for the results to be >>>>> returned is >>>>> short. For example, currently on Mechanical Turk, a survey of >>>>> approximately 50 items will be answered by 50 or more participants >>>>> within a couple of hours, at a cost of approximately $1 per >>>>> participant. Thus a survey can be completed within a day, at a >>>>> cost of >>>>> less than $50. (The hard work of designing the experiment, and >>>>> constructing controlled materials remains of course.) >>>>> >>>>> Sorry to be so verbose. But I think that these methodological >>>>> points >>>>> are very important. >>>>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson >>>>> >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative >>>>> methods in syntax and semantics research. Language and Cognitive >>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson >>>>> & Fedorenko InPress LCP.pdf >>>>> >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & >>>>> Fedorenko >>>>> 2010 TICS.pdf >>>>> >>>>> >>>>> >>>>> >>>>>> Dick, >>>>>> >>>>>> You raise an important issue here about methodology. I believe >>>>>> that >>>>>> intuitions are a fine way to generate hypotheses and even to test >>>>>> them - to a degree. But while it might not have been feasible for >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge >>>>>> Grammar to conduct experiments on every point of the grammar, >>>>>> experiments could have only made the grammar better. The use of >>>>>> intuitions, corpora, and standard psycholinguistic >>>>>> experimentation >>>>>> (indeed, Standard Social Science Methodology) is vital for >>>>>> taking the >>>>>> field forward and for providing the best support for different >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very >>>>>> useful new >>>>>> paper on this, showing serious shortcomings with intuitions as >>>>>> the >>>>>> sole source of evidence, in their paper: "The need for >>>>>> quantitative >>>>>> methods in syntax and semantics research". >>>>>> >>>>>> Carson Schutze and Wayne Cowart, among others, have also written >>>>>> convincingly on this. >>>>>> >>>>>> It is one reason that a team from Stanford, MIT (Brain and >>>>>> Cognitive >>>>>> Science), and researchers from Brazil are beginning a third >>>>>> round of >>>>>> experimental work among the Pirahas, since my own work on the >>>>>> syntax >>>>>> was, like almost every other field researcher's, based on native >>>>>> speaker intuitions and corpora. >>>>>> >>>>>> The discussion of methodologies reminds me of the initial >>>>>> reactions >>>>>> to Greenberg's work on classifying the languages of the >>>>>> Americas. His >>>>>> methods were strongly (and justifiably) criticized. However, I >>>>>> always >>>>>> thought that his methods were a great way of generating >>>>>> hypotheses, >>>>>> so long as they were ultimately put to the test of standard >>>>>> historical linguistics methods. And the same seems true for use >>>>>> of >>>>>> native-speaker intuitions. >>>>>> >>>>>> -- Dan >>>>> >>>>> >>>>> >>>>>>> We linguists can add a further layer of explanation to the >>>>>>> judgements, but some judgements do seem to be more reliable than >>>>>>> others. And if we have to wait for psycholinguistic evidence for >>>>>>> every detailed analysis we make, our whole discipline will >>>>>>> immediately grind to a halt. Like it or not, native speaker >>>>>>> judgements are what put us linguists ahead of the rest in >>>>>>> handling >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the >>>>>>> English >>>>>>> Language (or the OED) without using native speaker judgements. >>>>>>> >>>>>>> Best wishes, Dick Hudson >>>>> >>>>> >>>>> >>> >>> >>> >> > From dan at daneverett.org Sat Sep 11 16:53:23 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 12:53:23 -0400 Subject: analysis: unhappiness In-Reply-To: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9@MIT.EDU> Message-ID: Ted, Let me clarify this: > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > Both linguistics and psychology are big fields. We assume Dan is referring to cognitive psychology / cognitive science here. (Of course, there are sub-fields of psychology - e.g., personality psychology or abnormal psychology - which are somewhat distinct from linguistics, but those sub-fields are also distinct from cognitive psychology.) It is true that historically linguistics is not treated as a subfield of cognitive psychology / cognitive science. However, key research questions in linguistics (i.e., the form of the knowledge structures and algorithms underlying human language) are indeed a subset of those investigated by cognitive psychologists / cognitive scientists. We think that the biggest factor separating linguistics from psychology is the methods used to explore the research questions, rather than the research questions themselves. Consequently, we would like to continue to see tighter connections among the fields of psychology / cognitive science, linguistics, as well as other fields like anthropology and computer science. Correct, I meant cognitive psychology, not, say, psychoanalysis. There are definitely overlapping concerns. But my main concern about language is less about representations and more about the cultural and sociological values that lead to sentences and expressions in the corpus, rather than the mind. I used to think that my main interest was representations in the mind. But I find the psychology less interesting than the anthropology these days. But this is not an excuse to avoid quantitative methods. I believe that you and Ev, and others, have made a convincing case for quantitative methods. Quantitative methods in field research on syntax and semantics is vital. -- dan From dryer at buffalo.edu Sat Sep 11 20:25:54 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Sat, 11 Sep 2010 16:25:54 -0400 Subject: analysis: unhappiness Message-ID: Lise, First, let me concede the obvious: what is inside people's heads varies from individual to individual and, for a given individual, from one time to another. And that whether a given pattern is internalized by a given speaker at a given time is not a black and white issue, but often a matter of degree. But while that makes the story more complicated, it seems largely orthogonal to the fundamental issue we are discussing. It's not clear to me at this point whether the issue we are disagreeing about is substantive or terminological, whether we are meaning different things when we talk about patterns in the data. For example, the frequency with which different sorts of sentences occur in the data a child is exposed to has profound effects on what the child internalizes as rules and hence is an important part of explanations for why languages are the way they are. In so far as such relative frequencies in the data are "patterns in the data", we need to know what those patterns are in order to explain why languages are the way they are. However, that is not what I was talking about when I referred to patterns in the data. I was intending the sorts of patterns that distinguish two analyses of the same set of data, and relative frequency in the data does not play a role in distinguishing what two analyses of the same set of data say. For example, when we compare a more abstract analysis of some phonological phenomenon that posits some rule based on a perceived pattern in the data with a more concrete analysis of the same phenomenon that denies the generalization covered by the rule in the more abstract analysis, and ask which is "the correct" analysis, my claim is that the only sense of that question is whether the generalization corresponds to something inside speakers' heads (admitting, again, that this can vary from speaker to speaker, from time to time, and is a matter of degree). There is no second sense in which the question can be interpreted as a question about which analysis is correct as far as what is "out there" is concerned. As far as I can see, everything you say in your response to me is consistent with this claim of mine. And when you say " I have to disagree with you on the validity of describing what's 'out there'", your wording implies that I have said that describing what's "other there" is not a valid activity. But, to the contrary, since I am not a psycholinguist, that is ALL that I do. I am currently (co-)writing a grammatical description of a language, and that is describing what's "out there". However, when confronted with two analyses which are consistent with the data, which differ in that one describes the data in terms of a rule that the other does not, I do not believe that one of the two analyses must be the "correct" one in terms of what is "out there". At most, one of them may be more correct in terms of what is inside speakers' heads (and perhaps one is correct for some speakers and the other for other speakers), but since we are not doing psycholinguistic research on those speakers, I can't tell. But we have enough work to do describing the language that I needn't worry about which analysis is correct. Some old notions from Chomsky are useful here. We have enough to do trying to achieve observational adequacy in describing the language we are working on that I really don't worry about descriptive adequacy. When we have competing analyses that we need to decide on, they normally differ in observational adequacy, so we look for data that will tell us which is correct in terms of observational adequacy. Now it's true that our description is full of generalizations based on patterns in the data; after all, a listing of our raw data would achieve observational adequacy. But in doing so, we are not claiming that the generalizations we are making are necessarily real. I suspect some of them aren't. (And by "real", I mean psychologically real, since my whole point in this discussion is that the only sense in which generalizations can or cannot be real is psychological.) So why don't we just provide a listing of our raw data? Because we want to communicate to others what the language is like and a description that describes it in terms of patterns in the data better serves that purpose, whether or not the patterns are real. Matthew On Fri 09/10/10 8:40 PM , Lise Menn Lise.Menn at Colorado.EDU sent: > Matt, I have to disagree with you on the validity of describing what's > 'out there' (what Dick Hudson says is his interest, in his > contribution of 5:40:06 PM MDT today). We DO have to account for it > in order to understand how 'the language in speakers' heads' gets > into those heads in the first place. In more detail: Each of us is > immersed from (before) birth in a sampling of utterances (and if we > are literate, eventually also written forms of the language). In > order to understand how we really create our internal representations > of our language, we have to know (or be able to estimate) something > about the data our brains get as input. There are at least better and > worse descriptions of the patterns in those data, and certainly there > are wrong ones, though in many cases - for example in the > 'unhappiness' case - there are probably conflicting right ones, > rather than any single correct one. (OT offers some help in thinking > about this.) > To take a concrete example, in order to account for the > still-unstable changes in English pronominal case marking in compound > NP objects of prepositions from a system based on syntactic case (He > gave the cookies to Mary and me) to a system apparently based partly > on whether the pronoun is next to the governing preposition (He gave > the cookies to Mary and I/ to me and Mary), you first have to do an > analysis of usage and figure out what the pattern is. And usage is > not in our heads (although it's the result of what's in our heads), > it's 'out there' . > Even fossils and obscure patterns contribute to the redundancy of > the language, making it more learnable and and helping to create the > resonances used by great poets and orators. (I admit to having > oversimplified in speaking as if there were always one 'correct' > analysis of the patterns 'out there' that might be (subconsciously) > discoverable by speakers. That's not true.) And because not all > speakers are equally sensitive to language patterns - again, the > Gleitman and Gleitman book is a terrific example - it's also an > oversimplification to talk about 'what is in speaker's heads' as if > the same thing is in everyone's head. (K.P. Mohanan has also > published on this.) At the lexical level, Danielle Cyr's examples ( > September 9, 2010 8:38:59 PM MDT) further remind us that what's > inside each person's head changes over time. So we must also be > careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover. It's not - > it's more like a complex mosaic that does not fit together perfectly. > Lise > On Sep 10, 2010, at 12:51 PM, wrote: > > The following sentence of Lise's > "Since we don't in fact have such an agreed-on terminology, we have > to be quite careful in making clear what we are referring to when we > talk about 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the > correct analysis", one in terms of what is in people's heads, the > other in terms of what is "out there". > > There are a vast number of patterns "out there". The only > distinction, on my view, amongst this vast number of patterns, is > between those that reflect something inside people's heads and those > that don't. But if that is the case, then there is no coherent sense > in which one can talk of "the correct analysis" of what is "out > there", except in terms of what is in people's heads, and thus no > second sense of "the correct analysis". The patterns that don't > correspond to things in people's heads fall into (at least) two > categories. There are those that are akin to constellations of stars > and, as with constellations, there is no reality to these patterns, > except in the minds of linguists. And there are those patterns which > are the fossil remains of what was in the heads of speakers of an > earlier stage of the language but which no longer are. These latter > patterns are real, and they are relevant to exlaining why the > language is now the way it is, but they are not relevant, I think > many would agree, as to what is the "correct analysis" of the > language today. > For this reason, I claim that the only sense in which an analysis can > be "the correct analysis" is in terms of what is inside of people's > heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion > of these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn wrote: > > I wish we had better terminology for keeping track of whether, at a > given time, we are talking about the patterns that are 'out there' > in > the language and might possibly be apprehended (subconsciously) by a > speaker, and when we are talking about the patterns that a particular > speaker actually does apprehend, as indicated by experiments, from > simple 'wug tests' up to brain wave and eye-gaze studies. And for > distinguishing among the degrees of pattern apprehension that a > person > may have, from vague preferences detectable in reaction times or > other > behavior all the way up through clear metalinguistic insights. Dick > Hudson's note reminding us of the Gleitman and Gleitman study is > right > on target. > Since we don't in fact have such an agreed-on terminology, we have > to > be quite careful in making clear what we are referring to when we > talk > about 'the correct analysis' of a form like 'unhappiness'. We know, > but > tend to forget - and tend to forget to tell our students! - that > it's an > empirical question as to whether the formal simplicity and > coherence of > description of forms 'out there' (e.g. lovely abstract > morphophonemics) > is any kind of approximation to the way knowledge of the same forms > is > organized in a particular person's head. If we remember that a very > large proportion of what we know about our language is 'out there' > when > we are infants and has to be internalized through experience with > the > language (even if you believe in innate 'core language'), the > variation > in internal knowledge from one person to another is more > understandable. > We especially need to consider (and try to test) the possibility > that > since > the brain can make multiple cross-connections, multiple patterns are > involved > simultaneously in morphological and syntactic analyses. I suggest > that > that's the case with 'unhappiness' - and the linguistic analyses > that I > know about are not good at handling that kind of idea. > > Lise > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > Two comments. > > First (elaborating perhaps on Dick Hudson's comment), I think there > is an > important distinction between low-level linguistic intuitions (like > whether a > word or sentence is well-formed or what it means) and higher-level > intuitions > (like what the structure of a word or sentence is). One can take > the position > that we need to account for the former (while recognizing that they > are not > always reliable) but not the latter. > Second, the tension here is not only between evidence from speaker > intuitions > versus evidence from psycholinguistic experiments. There is also a > tension > between deciding on the correct analysis on the basis of a priori > simplicity > arguments versus deciding on the correct analysis on the basis of > psycholinguistic evidence (see Derwing 1973). The bracketing > paradox that Dan > referred to that arises with the word (semantics argues > for > [[un+happi] + er], morphology and phonology argues for [un + [happi > +er]] (the > comparative suffix can only be attached to adjectives containing one > or two > syllables) is only a paradox if one assumes that speakers adopt the > simplest > analysis. For example, if speakers adopt a more complex rule for > either of these > (e.g. perhaps the rule for attaching -er can apply exceptionally to > trisyllabic > words beginning with un-), then the bracketing paradox disappears. > Matthew > > On Thu 09/09/10 8:16 AM , Richard Hudson sent: > Thanks Dan. I'm sure you're right, and I'd be the first to agree > that > conscious judgements are only one kind of evidence that we need to > take > into account. I admire Carson Schutze's work (which I reviewed in > fact), > and of course I've been aware of complaints about judgements by > people > like Labov for decades. > > But you're missing my main point, which is that all judgements aren't > equally reliable. If you want to know how /unhappiness/ is > structured, > ask a linguist, not a five-year old. And one of the by-products of > education may be increased sensitivity to syntax - which is one of > the > many reasons why linguists need to pay more attention to education. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm [4] > On 09/09/2010 11:39, Daniel Everett wrote: > Dick, > > You raise an important issue here about > methodology. I believe that intuitions are a fine way to generate > hypotheses and even to test them - to a degree. But while it might > not have > been feasible for Huddleston, Pullum, and the other contributors to > the > Cambridge Grammar to conduct experiments on every point of the > grammar, > experiments could have only made the grammar better. The use of > intuitions, > corpora, and standard psycholinguistic experimentation (indeed, > Standard > Social Science Methodology) is vital for taking the field forward > and for > providing the best support for different analyses. Ted Gibson and Ev > Fedorenko have written a very useful new paper on this, showing > serious > shortcomings with intuitions as the sole source of evidence, in > their > paper: "The need for quantitative methods in syntax and semantics > research".> > Carson Schutze and Wayne Cowart, among others, > have also written convincingly on this.> > It is one reason that a team from Stanford, MIT > (Brain and Cognitive Science), and researchers from Brazil are > beginning a > third round of experimental work among the Pirahas, since my own > work on > the syntax was, like almost every other field researcher's, based > on native > speaker intuitions and corpora.> > The discussion of methodologies reminds me of > the initial reactions to Greenberg's work on classifying the > languages of > the Americas. His methods were strongly (and justifiably) criticized. > However, I always thought that his methods were a great way of > generating > hypotheses, so long as they were ultimately put to the test of > standard > historical linguistics methods. And the same seems true for use of > native-speaker intuitions.> > -- Dan > > We linguists can add a further layer of > explanation to the judgements, but some judgements do seem to be more > reliable than others. And if we have to wait for psycholinguistic > evidence > for every detailed analysis we make, our whole discipline will > immediately > grind to a halt. Like it or not, native speaker judgements are what > put us > linguists ahead of the rest in handling fine detail. Imagine > writing the > Cambridge Grammar of the English Language (or the OED) without > using native > speaker judgements.>> > Best wishes, Dick Hudson > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > Lise Menn Home Office: 303-444-42741625 > Mariposa Ave Fax: 303-413-0017 Boulder CO 80302 > Professor Emerita of Linguistics Fellow, Institute of Cognitive > Science University of Colorado > Secretary, AAAS Section Z [Linguistics] > Campus Mail Address:UCB 594, Institute of Cognitive Science > Campus Physical Address:CINC 234 1777 Exposition Ave, Boulder > > > Links: > ------ > [4] http://www.phon.ucl.ac.uk/home/dick/home.htm > > From dick at ling.ucl.ac.uk Sat Sep 11 21:14:16 2010 From: dick at ling.ucl.ac.uk (Richard Hudson) Date: Sat, 11 Sep 2010 22:14:16 +0100 Subject: analysis: unhappiness In-Reply-To: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9@MIT.EDU> Message-ID: Dear Ted, Thanks for this. Sorry I misrepresented your goals; I'm afraid I was stereotyping you. But then, I was also stereotyping myself. It looks as though we both share the same global goal that I sketched in my earlier message: a comprehensive model for language which covers both structure and processing, and both behaviour and cognition; and which allows individual variation in all four. Hooray! All we need is the wisdom to 'think global and act local', as they keep telling us. I'm afraid it's all too easy to get side-tracked into something narrow and easier. Best wishes, Dick > 2. Dick (by the way, thank you for the kind responses, and your > positive tone): > > "Your [the psycholinguists'] goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality." > > in contrast to "my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers." > > Many cognitive psychologists / cognitive scientists (all the ones I > know at MIT for example) are interested in both cognitive > generalizations across people and ways in which people differ > cognitively. In fact, some methods (e.g., the individual differences > approach where co-variation of various behaviors / characteristics is > examined across individuals) have been specifically developed to study > differences among individuals. Both kinds of data are important for > understanding human cognition, including language. This applies to > language research directly: generalizations across people are > important, but so are individual differences. In either case, > quantitative data are necessary to evaluate research questions and > test hypotheses. > > On a related note, it is a mistake to characterize researchers with a > background in "psychology" or cognitive science as being interested in > "processing", and researchers with a background in "linguistics" as > being interested in "knowledge" or "representation / structure". Both > psychologists and linguists should be interested in *both* > representation and processing (and learning, for that matter). We > wrote a little about this confusion in Gibson & Fedorenko (in press), > which we include at the end of the message. > From dryer at buffalo.edu Sun Sep 12 01:26:25 2010 From: dryer at buffalo.edu (Matthew S. Dryer) Date: Sat, 11 Sep 2010 21:26:25 -0400 Subject: analysis: unhappiness Message-ID: Dan etc, There have unfortunately been two sub-threads with the same subject heading "Re: [FUNKNET] analysis: unhappiness". Even I have found the Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have pursued the other one only because it wasn't clear that others hadn't misunderstood what I was trying to say. Unfortunately, the following comment from Dan seems to illustrate further that I haven't made myself clear: "On another note, I don't buy the 'in my head' 'out of my head' distinction either (that Matt seems to be urging upon us)." BUT, it was Lise who urged that distinction on us. The whole point of my emails has been to deny such a distinction, to argue that the only reality is the "in the head" one. In fact, the gradual convergence of thinking in the Gibson-etc subthread seems to reflect the idea that despite apparent differences, there is a common underlying goal. Matthew On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: > I think that Brian and Dick make excellent points. There are very good > grammars written that could be improved by psycholinguistic experimentation > and more quantitative approaches. But large sections of those grammars > aren't going to change one bit (go-went) with quantitative tests and such > tests would be completely counterproductive given the shortness of life and > the vastness of the field linguist's tasks. > Part of the problem is that linguistics is not simply a subdiscipline of > psychology. Linguistics has its own objectives and those only occasionally > overlap with psychology. The same for methods. > On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us). We study different things > and have different reasons for being satisfied with the results we > achieve. > I believe that we linguists are often complacent and fail to apply better > methods. But of course that applies to all disciplines. > In the meantime, checking corpora, collecting data as a result of careful > interviews with native speakers, and the other aspects of the field > linguist's task are vital parts of the linguist's task and much of this > won't be improved by quantitative methods as we currently understand them. > Maybe sometime. > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that they > showed the danger of using intuitions. What I meant to say of using > intuitions as standardly used by linguists. They convinced me that there is > a lot to learn from quantitative methods. > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a > psycholinguist's view. Your goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality. And (as you know) I admire the way you > pursue that goal. But my goal, as a linguist, is different. I want to > explore the structure of a language so that I can understand how all the > bits fit together. Like you, I'm aiming to model cognition, but my focus is > on items and structures, and I start from the assumption that these can and > do vary across speakers.> > > However, having said all that I do agree with > you that linguists should all get used to collecting and using quantitative > data; and with the help of Brian MacWhinney's typology we'd know what > methods to use when. And I do agree with your points about bid/bidded: in > cases like that, quantitative data would be at least a very good starting > point for a proper investigation.> > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> > > On 10/09/2010 19:30, Ted Gibson > wrote:>> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I > don't understand what is confusing about what Ev Fedorenko and I are > claiming. All we are saying is that if you have some testable claim > involving a general hypothesis about a language, then you need to get > quantitative data from unbiased sources to evaluate that claim. If you are > interested in English past tense morphology, then depending on the claims > that you might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very low > frequency words, you can run experiments to test behavior with respect to > such words.>> > >> Your example of the past tense of > "bid" is a fine such example. You can run an experiment like the > one you suggested to find out what people think the past tense is. If you > then found that 20/50 people responded "bidded" and 30/50 respond > "bid", that is a lot of useful information. As you suggest in > your discussion, this result wouldn't answer the question of how past tense > is stored in each individual. This result would be ambiguous among several > possible explanations. One possibility is that the probability distribution > that is being discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another possibility is > that each person has a similar probability distribution in their heads, > such that 2/5 of the time I respond one way, and 3/5 of the time I respond > another. Further experiments would be necessary to answer between these and > other possible theories (e.g., with repeated trials from the same person, > carefully planned so that the participants don't notice that they are being > asked multiple times). Without the quantitative evidence in the first > place, there is no way to answer these kinds of questions.>> > >> Regarding the past tense of "go", > this would be useful as a baseline in an experiment involving the less > frequent ones. So, yes, it would useful to gather quantitative evidence in > such a case also, as baselines with respect to the more interesting cases > for theories.>> > >> The bottom line: if you have a > generalization about a language that you wish to evaluate (such that you > hypothesize that it is true across the speakers of the language), then you > need quantitative evidence from multiple individuals, using an unbiased > data collection method, to evaluate such a claim. The point about > Mechanical Turk is that it is really *easy* to do this now, at least for > languages like English.>> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson > wrote:>> > >>> Dear Ted, > >>> Thanks for the very interesting comment, > but are you REALLY saying that I shouldn't claim, for example, that the > past tense of GO is "went" without first cross-checking with 50 > native speakers?>>> > >>> Isn't there a danger of missing the > point that we all, as native speakers, spend our whole lives scanning other > people's linguistic behaviour (language 'out there', E-language) and trying > to explain it to ourselves in terms of a language system (language 'in > here', I-language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language.>>> > >>> Given that view of language development, > I don't see how quantitative data will help. Let's take a real uncertainty, > such as the past tense of BID. If I want to say I did it, do I say "I > bidded" or "I bid"? My judgement: I don't know. Ok, you get > 50 people to oblige on Mechanical Turk, and 20 of them give > "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my > judgement? I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for > probing the bit of my mind (or even brain) that holds BID and its details; > but I suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of > quantitative sociolinguistics, so I do accept that quantitative data are > relevant to linguistic analysis in some areas, where the I-language > phenomenon is frequent enough to produce usable data.)>>> > >>> It seems to me that this discussion > raises the really fundamental question of what kind of thing we think > language is: social or individual. The problem isn't unique to linguistics > of course; it's the same throughout the social sciences. But what's special > about linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the differences > between individuals really matter. I don't see that we're ever going to > have anything better than judgements to go on, so what we need is a way to > ensure that judgements are accurate reports of individual I-language. A > rotten situation for a science, but I don't see how it can get > better.>>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> > >>> On 10/09/2010 14:03, Ted Gibson > wrote:>>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points > that Dan Everett makes, in>>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a > couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see > references and links below) on what we>>>> think are weak methodological > standards in syntax and semantics>>>> research over the past many years. > The issue that we address is the>>>> prevalent method in syntax and > semantics research, which involves>>>> obtaining a judgment of the > acceptability of a sentence / meaning>>>> pair, typically by just the author > of the paper, sometimes with>>>> feedback from colleagues. As we > address in our papers, this>>>> methodology does not allow proper > testing of scientific hypotheses>>>> because of (a) the small number of > experimental participants>>>> (typically one); (b) the small > number of experimental stimuli>>>> (typically one); (c) cognitive > biases on the part of the researcher>>>> and participants; and (d) the effect > of the preceding context (e.g.,>>>> other constructions the researcher > may have been recently>>>> considering). (As Dan said, see > Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & > Fedorenko, in press; for similar>>>> points, but with not as strong a > conclusion as ours).>>>> > >>>> Three issues need to be separated > here: (1) the use of intuitive>>>> judgments as a dependent measure in > a language experiment; (2)>>>> potential cognitive biases on the > part of experimental subjects and>>>> experimenters in language > experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the > dependent measure might be. The>>>> paper that Ev and I wrote addresses > the last two issues, but does not>>>> go into depth on the first issue > (the use of intuitions as a dependent>>>> measure in language experiments). > Regarding this issue, we don't think>>>> that there is anything wrong with > gathering intuitive judgments as a>>>> dependent measure, as long as the > task is clear to the experimental>>>> participants. > >>>> > >>>> In the longer paper (Gibson & > Fedorenko, in press) we respond to some>>>> arguments that have been given in > support of continuing to use the>>>> traditional non-quantitative method > in syntax / semantics research.>>>> One recent defense of the > traditional method comes from Phillips>>>> (2008), who argues that no harm has > come from the non-quantitative>>>> approach in syntax research thus > far. Phillips argues that there are>>>> no cases in the literature where an > incorrect intuitive judgment has>>>> become the basis for a widely > accepted generalization or an important>>>> theoretical claim. He therefore > concludes that there is no reason to>>>> adopt more rigorous data collection > standards. We challenge Philips???>>>> conclusion by presenting three cases > from the literature where a>>>> faulty intuition has led to > incorrect generalizations and mistaken>>>> theorizing, plausibly due to > cognitive biases on the part of the>>>> researchers. > >>>> > >>>> A second argument that is sometimes > presented for the continued use of>>>> the traditional non-quantitative > method is that it would be too>>>> inefficient to evaluate every > syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For > example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in > their response to Gibson & Fedorenko>>>> (2010): ???It would cripple > linguistic investigation if it were required>>>> that all judgments of ambiguity and > grammaticality be subject to>>>> statistically rigorous experiments > on naive subjects, especially when>>>> investigating languages whose > speakers are hard to access??? (Culicover>>>> & Jackendoff, 2010, p. 234). > (Dick Hudson makes a similar point>>>> earlier in the discussion here.) > Whereas we agree that in>>>> circumstances where gathering data > is difficult, some evidence is>>>> better than no evidence, we do not > agree that research would be slowed>>>> with respect to languages where > experimental participants are easy to>>>> access, such as English. In > contrast, we think that the opposite is>>>> true: the field???s progress > is probably slowed by not doing>>>> quantitative research. > >>>> Suppose that a typical syntax / > semantics paper that lacks>>>> quantitative evidence includes > judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 > or more empirical claims. Even if>>>> most of the judgments from such a > paper are correct or are on the>>>> right track, the problem is in > knowing which judgments are correct.>>>> For example, suppose that 90% of the > judgments from an arbitrary paper>>>> are correct (which is probably a > high estimate). (Colin Phillips and>>>> some of his former students / > postdocs have commented to us that, in>>>> their experience, quantitative > acceptability judgment studies almost>>>> always validate the claim(s) in the > literature. This is not our>>>> experience, however. Most > experiments that we have run which attempt>>>> to test some syntactic / semantic > hypothesis in the literature end up>>>> providing us with a pattern of data > that had not been known before the>>>> experiment (e.g., Breen et al., in > press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras > & Gibson, submitted).) This means>>>> that in a paper with 50 empirical > claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to > choose 45 items from 50. That???s over>>>> two million different theories. By > quantitatively evaluating the>>>> empirical claims, we reduce the > uncertainty a great deal. To make>>>> progress, it is better to have > theoretical claims supported by solid>>>> quantitative evidence, so that even > if the interpretation of the data>>>> changes over time as new evidence > becomes available ??? as is often the>>>> case in any field of science > ??? the empirical pattern can be used as a>>>> basis for further > theorizing.>>>> > >>>> Furthermore, it is no longer > expensive to run behavioral experiments,>>>> at least in English and other widely > spoken languages. There now>>>> exists a marketplace interface > ??? Amazon.com???s Mechanical Turk ??? which>>>> can be used for collecting > behavioral data over the internet quickly>>>> and inexpensively. The cost of using > an interface like this is>>>> minimal, and the time that it takes > for the results to be returned is>>>> short. For example, currently on > Mechanical Turk, a survey of>>>> approximately 50 items will be > answered by 50 or more participants>>>> within a couple of hours, at a cost > of approximately $1 per>>>> participant. Thus a survey can be > completed within a day, at a cost of>>>> less than $50. (The hard work of > designing the experiment, and>>>> constructing controlled materials > remains of course.)>>>> > >>>> Sorry to be so verbose. But I think > that these methodological points>>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In > press). The need for quantitative>>>> methods in syntax and semantics > research. Language and Cognitive>>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress > LCP.pdf>>>> > >>>> Gibson, E. & Fedorenko, E. > (2010). Weak quantitative standards in>>>> linguistics research. Trends in > Cognitive Science, 14, 233-234.>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue > here about methodology. I believe that>>>>> intuitions are a fine way to > generate hypotheses and even to test>>>>> them - to a degree. But while it > might not have been feasible for>>>>> Huddleston, Pullum, and the > other contributors to the Cambridge>>>>> Grammar to conduct experiments > on every point of the grammar,>>>>> experiments could have only made > the grammar better. The use of>>>>> intuitions, corpora, and > standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science > Methodology) is vital for taking the>>>>> field forward and for providing > the best support for different>>>>> analyses. Ted Gibson and Ev > Fedorenko have written a very useful new>>>>> paper on this, showing serious > shortcomings with intuitions as the>>>>> sole source of evidence, in > their paper: "The need for quantitative>>>>> methods in syntax and semantics > research".>>>>> > >>>>> Carson Schutze and Wayne Cowart, > among others, have also written>>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team > from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from > Brazil are beginning a third round of>>>>> experimental work among the > Pirahas, since my own work on the syntax>>>>> was, like almost every other > field researcher's, based on native>>>>> speaker intuitions and > corpora.>>>>> > >>>>> The discussion of methodologies > reminds me of the initial reactions>>>>> to Greenberg's work on > classifying the languages of the Americas. His>>>>> methods were strongly (and > justifiably) criticized. However, I always>>>>> thought that his methods were a > great way of generating hypotheses,>>>>> so long as they were ultimately > put to the test of standard>>>>> historical linguistics methods. > And the same seems true for use of>>>>> native-speaker > intuitions.>>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a > further layer of explanation to the>>>>>> judgements, but some > judgements do seem to be more reliable than>>>>>> others. And if we have to > wait for psycholinguistic evidence for>>>>>> every detailed analysis we > make, our whole discipline will>>>>>> immediately grind to a halt. > Like it or not, native speaker>>>>>> judgements are what put us > linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing > the Cambridge Grammar of the English>>>>>> Language (or the OED) > without using native speaker judgements.>>>>>> > >>>>>> Best wishes, Dick > Hudson>>>> > >>>> > >>>> > >> > >> > >> > > > > > > > From dan at daneverett.org Sun Sep 12 01:32:59 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 21:32:59 -0400 Subject: analysis: unhappiness In-Reply-To: <12429.1284254785@buffalo.edu> Message-ID: Sorry, Matt, for being careless in my attribution of that distinction. I do believe that there is such a distinction, however, and that it is very important. If you are after careful and deep understanding of the nature of representations, then the 'in the head' focus is important. If you are more interested (always, occasionally, etc) in how discourses, sentences, words, phrases and so on are underwritten by/structured by or merely interact with cultural values, then there is a sense in which the focus is on language outside the head, in the community. Even cognition works this way, I think. For example, I know somethings about trees. But even greater knowledge, knowledge I can access, about trees is found in my culture. In studying knowledge, I might want to know what people know individually, or I might want to know about the knowledge of cultures, some of which will not be known/mastered by any one member of the culture, and how the cultural knowledge affects/is accessed by, individuals. Values are other things that can have both a psychological and a cultural existence. Dan On 11 Sep 2010, at 21:26, Matthew S. Dryer wrote: > > Dan etc, > > There have unfortunately been two sub-threads with the same subject heading "Re: > [FUNKNET] analysis: unhappiness". Even I have found the > Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have > pursued the other one only because it wasn't clear that others hadn't > misunderstood what I was trying to say. > > Unfortunately, the following comment from Dan seems to illustrate further that I > haven't made myself clear: > > "On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us)." > > BUT, it was Lise who urged that distinction on us. The whole point of my emails > has been to deny such a distinction, to argue that the only reality is the "in > the head" one. In fact, the gradual convergence of thinking in the Gibson-etc > subthread seems to reflect the idea that despite apparent differences, there is a > common underlying goal. > > Matthew > > On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: >> I think that Brian and Dick make excellent points. There are very good >> grammars written that could be improved by psycholinguistic experimentation >> and more quantitative approaches. But large sections of those grammars >> aren't going to change one bit (go-went) with quantitative tests and such >> tests would be completely counterproductive given the shortness of life and >> the vastness of the field linguist's tasks. >> Part of the problem is that linguistics is not simply a subdiscipline of >> psychology. Linguistics has its own objectives and those only occasionally >> overlap with psychology. The same for methods. >> On another note, I don't buy the 'in my head' 'out of my head' distinction >> either (that Matt seems to be urging upon us). We study different things >> and have different reasons for being satisfied with the results we >> achieve. >> I believe that we linguists are often complacent and fail to apply better >> methods. But of course that applies to all disciplines. >> In the meantime, checking corpora, collecting data as a result of careful >> interviews with native speakers, and the other aspects of the field >> linguist's task are vital parts of the linguist's task and much of this >> won't be improved by quantitative methods as we currently understand them. >> Maybe sometime. >> Dan >> >> P.S. In my original reference to Ted and Ev's paper, I said that they >> showed the danger of using intuitions. What I meant to say of using >> intuitions as standardly used by linguists. They convinced me that there is >> a lot to learn from quantitative methods. >> On 10 Sep 2010, at 19:40, Richard Hudson wrote: >> >>> Dear Ted and Ev, >>> Yes, I understand your view, but I think it's a >> psycholinguist's view. Your goal is to find general processes and >> principles that apply uniformly across individuals, so you have to use >> methods to check for generality. And (as you know) I admire the way you >> pursue that goal. But my goal, as a linguist, is different. I want to >> explore the structure of a language so that I can understand how all the >> bits fit together. Like you, I'm aiming to model cognition, but my focus is >> on items and structures, and I start from the assumption that these can and >> do vary across speakers.> >>> However, having said all that I do agree with >> you that linguists should all get used to collecting and using quantitative >> data; and with the help of Brian MacWhinney's typology we'd know what >> methods to use when. And I do agree with your points about bid/bidded: in >> cases like that, quantitative data would be at least a very good starting >> point for a proper investigation.> >>> Best wishes, Dick >>> >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> >>> On 10/09/2010 19:30, Ted Gibson >> wrote:>> Dear Dick: >>>> >>>> Perhaps we are talking at cross purposes. I >> don't understand what is confusing about what Ev Fedorenko and I are >> claiming. All we are saying is that if you have some testable claim >> involving a general hypothesis about a language, then you need to get >> quantitative data from unbiased sources to evaluate that claim. If you are >> interested in English past tense morphology, then depending on the claims >> that you might want to investigate, there are lots of ways to get relevant >> quantitative evidence. Corpus data will probably be useful. For very low >> frequency words, you can run experiments to test behavior with respect to >> such words.>> >>>> Your example of the past tense of >> "bid" is a fine such example. You can run an experiment like the >> one you suggested to find out what people think the past tense is. If you >> then found that 20/50 people responded "bidded" and 30/50 respond >> "bid", that is a lot of useful information. As you suggest in >> your discussion, this result wouldn't answer the question of how past tense >> is stored in each individual. This result would be ambiguous among several >> possible explanations. One possibility is that the probability distribution >> that is being discovered reflects different dialects, such that 2/5 of the >> population has one past tense, and 3/5 has another. Another possibility is >> that each person has a similar probability distribution in their heads, >> such that 2/5 of the time I respond one way, and 3/5 of the time I respond >> another. Further experiments would be necessary to answer between these and >> other possible theories (e.g., with repeated trials from the same person, >> carefully planned so that the participants don't notice that they are being >> asked multiple times). Without the quantitative evidence in the first >> place, there is no way to answer these kinds of questions.>> >>>> Regarding the past tense of "go", >> this would be useful as a baseline in an experiment involving the less >> frequent ones. So, yes, it would useful to gather quantitative evidence in >> such a case also, as baselines with respect to the more interesting cases >> for theories.>> >>>> The bottom line: if you have a >> generalization about a language that you wish to evaluate (such that you >> hypothesize that it is true across the speakers of the language), then you >> need quantitative evidence from multiple individuals, using an unbiased >> data collection method, to evaluate such a claim. The point about >> Mechanical Turk is that it is really *easy* to do this now, at least for >> languages like English.>> >>>> Best wishes, >>>> >>>> Ted Gibson & Ev Fedorenko >>>> >>>> On Sep 10, 2010, at 1:59 PM, Richard Hudson >> wrote:>> >>>>> Dear Ted, >>>>> Thanks for the very interesting comment, >> but are you REALLY saying that I shouldn't claim, for example, that the >> past tense of GO is "went" without first cross-checking with 50 >> native speakers?>>> >>>>> Isn't there a danger of missing the >> point that we all, as native speakers, spend our whole lives scanning other >> people's linguistic behaviour (language 'out there', E-language) and trying >> to explain it to ourselves in terms of a language system (language 'in >> here', I-language)? So every judgement we make is based on thousands or >> millions of observed exemplars, and reflects a unique experience of >> E-language filtered through a unique I-language.>>> >>>>> Given that view of language development, >> I don't see how quantitative data will help. Let's take a real uncertainty, >> such as the past tense of BID. If I want to say I did it, do I say "I >> bidded" or "I bid"? My judgement: I don't know. Ok, you get >> 50 people to oblige on Mechanical Turk, and 20 of them give >> "bidded" and 30 "bid". So what? Does that mean that the >> correct answer is "bidded"? Surely not. How is it better than my >> judgement? I agree you could record my speech and find how often I use each >> alternative; but the reason I don't know is precisely because it's a rare >> word, so in a sense quantitative data are irrelevant even there. What would >> solve the problem of subjectivity, of course, would be a machine for >> probing the bit of my mind (or even brain) that holds BID and its details; >> but I suspect that even that wouldn't move us much further forward than my >> original "don't know". (Incidentally I write as a fan of >> quantitative sociolinguistics, so I do accept that quantitative data are >> relevant to linguistic analysis in some areas, where the I-language >> phenomenon is frequent enough to produce usable data.)>>> >>>>> It seems to me that this discussion >> raises the really fundamental question of what kind of thing we think >> language is: social or individual. The problem isn't unique to linguistics >> of course; it's the same throughout the social sciences. But what's special >> about linguistics is that we deal in very fine details of culture (e.g. >> details of how a particular word is used or pronounced) so the differences >> between individuals really matter. I don't see that we're ever going to >> have anything better than judgements to go on, so what we need is a way to >> ensure that judgements are accurate reports of individual I-language. A >> rotten situation for a science, but I don't see how it can get >> better.>>> >>>>> Dick >>>>> >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> >>>>> On 10/09/2010 14:03, Ted Gibson >> wrote:>>>> Dear Dan, Dick: >>>>>> >>>>>> I would like to clarify some points >> that Dan Everett makes, in>>>> response to Dick Hudson. >>>>>> >>>>>> Ev Fedorenko and I have written a >> couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see >> references and links below) on what we>>>> think are weak methodological >> standards in syntax and semantics>>>> research over the past many years. >> The issue that we address is the>>>> prevalent method in syntax and >> semantics research, which involves>>>> obtaining a judgment of the >> acceptability of a sentence / meaning>>>> pair, typically by just the author >> of the paper, sometimes with>>>> feedback from colleagues. As we >> address in our papers, this>>>> methodology does not allow proper >> testing of scientific hypotheses>>>> because of (a) the small number of >> experimental participants>>>> (typically one); (b) the small >> number of experimental stimuli>>>> (typically one); (c) cognitive >> biases on the part of the researcher>>>> and participants; and (d) the effect >> of the preceding context (e.g.,>>>> other constructions the researcher >> may have been recently>>>> considering). (As Dan said, see >> Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & >> Fedorenko, in press; for similar>>>> points, but with not as strong a >> conclusion as ours).>>>> >>>>>> Three issues need to be separated >> here: (1) the use of intuitive>>>> judgments as a dependent measure in >> a language experiment; (2)>>>> potential cognitive biases on the >> part of experimental subjects and>>>> experimenters in language >> experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the >> dependent measure might be. The>>>> paper that Ev and I wrote addresses >> the last two issues, but does not>>>> go into depth on the first issue >> (the use of intuitions as a dependent>>>> measure in language experiments). >> Regarding this issue, we don't think>>>> that there is anything wrong with >> gathering intuitive judgments as a>>>> dependent measure, as long as the >> task is clear to the experimental>>>> participants. >>>>>> >>>>>> In the longer paper (Gibson & >> Fedorenko, in press) we respond to some>>>> arguments that have been given in >> support of continuing to use the>>>> traditional non-quantitative method >> in syntax / semantics research.>>>> One recent defense of the >> traditional method comes from Phillips>>>> (2008), who argues that no harm has >> come from the non-quantitative>>>> approach in syntax research thus >> far. Phillips argues that there are>>>> no cases in the literature where an >> incorrect intuitive judgment has>>>> become the basis for a widely >> accepted generalization or an important>>>> theoretical claim. He therefore >> concludes that there is no reason to>>>> adopt more rigorous data collection >> standards. We challenge Philips???>>>> conclusion by presenting three cases >> from the literature where a>>>> faulty intuition has led to >> incorrect generalizations and mistaken>>>> theorizing, plausibly due to >> cognitive biases on the part of the>>>> researchers. >>>>>> >>>>>> A second argument that is sometimes >> presented for the continued use of>>>> the traditional non-quantitative >> method is that it would be too>>>> inefficient to evaluate every >> syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For >> example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in >> their response to Gibson & Fedorenko>>>> (2010): ???It would cripple >> linguistic investigation if it were required>>>> that all judgments of > ambiguity and >> grammaticality be subject to>>>> statistically rigorous experiments >> on naive subjects, especially when>>>> investigating languages whose >> speakers are hard to access??? (Culicover>>>> & Jackendoff, 2010, p. 234). >> (Dick Hudson makes a similar point>>>> earlier in the discussion here.) >> Whereas we agree that in>>>> circumstances where gathering data >> is difficult, some evidence is>>>> better than no evidence, we do not >> agree that research would be slowed>>>> with respect to languages where >> experimental participants are easy to>>>> access, such as English. In >> contrast, we think that the opposite is>>>> true: the field???s progress >> is probably slowed by not doing>>>> quantitative research. >>>>>> Suppose that a typical syntax / >> semantics paper that lacks>>>> quantitative evidence includes >> judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 >> or more empirical claims. Even if>>>> most of the judgments from such a >> paper are correct or are on the>>>> right track, the problem is in >> knowing which judgments are correct.>>>> For example, suppose that 90% of the >> judgments from an arbitrary paper>>>> are correct (which is probably a >> high estimate). (Colin Phillips and>>>> some of his former students / >> postdocs have commented to us that, in>>>> their experience, quantitative >> acceptability judgment studies almost>>>> always validate the claim(s) in the >> literature. This is not our>>>> experience, however. Most >> experiments that we have run which attempt>>>> to test some syntactic / semantic >> hypothesis in the literature end up>>>> providing us with a pattern of data >> that had not been known before the>>>> experiment (e.g., Breen et al., in >> press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras >> & Gibson, submitted).) This means>>>> that in a paper with 50 empirical >> claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to >> choose 45 items from 50. That???s over>>>> two million different theories. By >> quantitatively evaluating the>>>> empirical claims, we reduce the >> uncertainty a great deal. To make>>>> progress, it is better to have >> theoretical claims supported by solid>>>> quantitative evidence, so that even >> if the interpretation of the data>>>> changes over time as new evidence >> becomes available ??? as is often the>>>> case in any field of science >> ??? the empirical pattern can be used as a>>>> basis for further >> theorizing.>>>> >>>>>> Furthermore, it is no longer >> expensive to run behavioral experiments,>>>> at least in English and other widely >> spoken languages. There now>>>> exists a marketplace interface >> ??? Amazon.com???s Mechanical Turk ??? which>>>> can be used for collecting >> behavioral data over the internet quickly>>>> and inexpensively. The cost of using >> an interface like this is>>>> minimal, and the time that it takes >> for the results to be returned is>>>> short. For example, currently on >> Mechanical Turk, a survey of>>>> approximately 50 items will be >> answered by 50 or more participants>>>> within a couple of hours, at a cost >> of approximately $1 per>>>> participant. Thus a survey can be >> completed within a day, at a cost of>>>> less than $50. (The hard work of >> designing the experiment, and>>>> constructing controlled materials >> remains of course.)>>>> >>>>>> Sorry to be so verbose. But I think >> that these methodological points>>>> are very important. >>>>>> >>>>>> Best wishes, >>>>>> >>>>>> Ted Gibson >>>>>> >>>>>> Gibson, E. & Fedorenko, E. (In >> press). The need for quantitative>>>> methods in syntax and semantics >> research. Language and Cognitive>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress >> LCP.pdf>>>> >>>>>> Gibson, E. & Fedorenko, E. >> (2010). Weak quantitative standards in>>>> linguistics research. Trends in >> Cognitive Science, 14, 233-234.>>>> > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 > TICS.pdf >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>> Dick, >>>>>>> >>>>>>> You raise an important issue >> here about methodology. I believe that>>>>> intuitions are a fine way to >> generate hypotheses and even to test>>>>> them - to a degree. But while it >> might not have been feasible for>>>>> Huddleston, Pullum, and the >> other contributors to the Cambridge>>>>> Grammar to conduct experiments >> on every point of the grammar,>>>>> experiments could have only made >> the grammar better. The use of>>>>> intuitions, corpora, and >> standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science >> Methodology) is vital for taking the>>>>> field forward and for providing >> the best support for different>>>>> analyses. Ted Gibson and Ev >> Fedorenko have written a very useful new>>>>> paper on this, showing serious >> shortcomings with intuitions as the>>>>> sole source of evidence, in >> their paper: "The need for quantitative>>>>> methods in syntax and semantics >> research".>>>>> >>>>>>> Carson Schutze and Wayne Cowart, >> among others, have also written>>>>> convincingly on this. >>>>>>> >>>>>>> It is one reason that a team >> from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from >> Brazil are beginning a third round of>>>>> experimental work among the >> Pirahas, since my own work on the syntax>>>>> was, like almost every other >> field researcher's, based on native>>>>> speaker intuitions and >> corpora.>>>>> >>>>>>> The discussion of methodologies >> reminds me of the initial reactions>>>>> to Greenberg's work on >> classifying the languages of the Americas. His>>>>> methods were strongly (and >> justifiably) criticized. However, I always>>>>> thought that his methods were a >> great way of generating hypotheses,>>>>> so long as they were ultimately >> put to the test of standard>>>>> historical linguistics methods. >> And the same seems true for use of>>>>> native-speaker >> intuitions.>>>>> >>>>>>> -- Dan >>>>>> >>>>>> >>>>>> >>>>>>>> We linguists can add a >> further layer of explanation to the>>>>>> judgements, but some >> judgements do seem to be more reliable than>>>>>> others. And if we have to >> wait for psycholinguistic evidence for>>>>>> every detailed analysis we >> make, our whole discipline will>>>>>> immediately grind to a halt. >> Like it or not, native speaker>>>>>> judgements are what put us >> linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing >> the Cambridge Grammar of the English>>>>>> Language (or the OED) >> without using native speaker judgements.>>>>>> >>>>>>>> Best wishes, Dick >> Hudson>>>> >>>>>> >>>>>> >>>> >>>> >>>> >>> >> >> >> >> >> > > From dan at daneverett.org Sun Sep 12 01:35:54 2010 From: dan at daneverett.org (Daniel Everett) Date: Sat, 11 Sep 2010 21:35:54 -0400 Subject: analysis: unhappiness In-Reply-To: Message-ID: To clarify, I actually *do* buy the distinction, not as a false dichotomy, not as something that every linguist/psychologist should be concerned with, but in the way I just outlined. Dan On 11 Sep 2010, at 21:32, Daniel Everett wrote: > Sorry, Matt, for being careless in my attribution of that distinction. I do believe that there is such a distinction, however, and that it is very important. If you are after careful and deep understanding of the nature of representations, then the 'in the head' focus is important. If you are more interested (always, occasionally, etc) in how discourses, sentences, words, phrases and so on are underwritten by/structured by or merely interact with cultural values, then there is a sense in which the focus is on language outside the head, in the community. > > Even cognition works this way, I think. For example, I know somethings about trees. But even greater knowledge, knowledge I can access, about trees is found in my culture. In studying knowledge, I might want to know what people know individually, or I might want to know about the knowledge of cultures, some of which will not be known/mastered by any one member of the culture, and how the cultural knowledge affects/is accessed by, individuals. Values are other things that can have both a psychological and a cultural existence. > > Dan > > > > > On 11 Sep 2010, at 21:26, Matthew S. Dryer wrote: > >> >> Dan etc, >> >> There have unfortunately been two sub-threads with the same subject heading "Re: >> [FUNKNET] analysis: unhappiness". Even I have found the >> Gibson-MacWhinney-Everett-Hudson subthread the more interesting one. I have >> pursued the other one only because it wasn't clear that others hadn't >> misunderstood what I was trying to say. >> >> Unfortunately, the following comment from Dan seems to illustrate further that I >> haven't made myself clear: >> >> "On another note, I don't buy the 'in my head' 'out of my head' distinction >> either (that Matt seems to be urging upon us)." >> >> BUT, it was Lise who urged that distinction on us. The whole point of my emails >> has been to deny such a distinction, to argue that the only reality is the "in >> the head" one. In fact, the gradual convergence of thinking in the Gibson-etc >> subthread seems to reflect the idea that despite apparent differences, there is a >> common underlying goal. >> >> Matthew >> >> On Fri 09/10/10 9:05 PM , Daniel Everett dan at daneverett.org sent: >>> I think that Brian and Dick make excellent points. There are very good >>> grammars written that could be improved by psycholinguistic experimentation >>> and more quantitative approaches. But large sections of those grammars >>> aren't going to change one bit (go-went) with quantitative tests and such >>> tests would be completely counterproductive given the shortness of life and >>> the vastness of the field linguist's tasks. >>> Part of the problem is that linguistics is not simply a subdiscipline of >>> psychology. Linguistics has its own objectives and those only occasionally >>> overlap with psychology. The same for methods. >>> On another note, I don't buy the 'in my head' 'out of my head' distinction >>> either (that Matt seems to be urging upon us). We study different things >>> and have different reasons for being satisfied with the results we >>> achieve. >>> I believe that we linguists are often complacent and fail to apply better >>> methods. But of course that applies to all disciplines. >>> In the meantime, checking corpora, collecting data as a result of careful >>> interviews with native speakers, and the other aspects of the field >>> linguist's task are vital parts of the linguist's task and much of this >>> won't be improved by quantitative methods as we currently understand them. >>> Maybe sometime. >>> Dan >>> >>> P.S. In my original reference to Ted and Ev's paper, I said that they >>> showed the danger of using intuitions. What I meant to say of using >>> intuitions as standardly used by linguists. They convinced me that there is >>> a lot to learn from quantitative methods. >>> On 10 Sep 2010, at 19:40, Richard Hudson wrote: >>> >>>> Dear Ted and Ev, >>>> Yes, I understand your view, but I think it's a >>> psycholinguist's view. Your goal is to find general processes and >>> principles that apply uniformly across individuals, so you have to use >>> methods to check for generality. And (as you know) I admire the way you >>> pursue that goal. But my goal, as a linguist, is different. I want to >>> explore the structure of a language so that I can understand how all the >>> bits fit together. Like you, I'm aiming to model cognition, but my focus is >>> on items and structures, and I start from the assumption that these can and >>> do vary across speakers.> >>>> However, having said all that I do agree with >>> you that linguists should all get used to collecting and using quantitative >>> data; and with the help of Brian MacWhinney's typology we'd know what >>> methods to use when. And I do agree with your points about bid/bidded: in >>> cases like that, quantitative data would be at least a very good starting >>> point for a proper investigation.> >>>> Best wishes, Dick >>>> >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm> >>>> On 10/09/2010 19:30, Ted Gibson >>> wrote:>> Dear Dick: >>>>> >>>>> Perhaps we are talking at cross purposes. I >>> don't understand what is confusing about what Ev Fedorenko and I are >>> claiming. All we are saying is that if you have some testable claim >>> involving a general hypothesis about a language, then you need to get >>> quantitative data from unbiased sources to evaluate that claim. If you are >>> interested in English past tense morphology, then depending on the claims >>> that you might want to investigate, there are lots of ways to get relevant >>> quantitative evidence. Corpus data will probably be useful. For very low >>> frequency words, you can run experiments to test behavior with respect to >>> such words.>> >>>>> Your example of the past tense of >>> "bid" is a fine such example. You can run an experiment like the >>> one you suggested to find out what people think the past tense is. If you >>> then found that 20/50 people responded "bidded" and 30/50 respond >>> "bid", that is a lot of useful information. As you suggest in >>> your discussion, this result wouldn't answer the question of how past tense >>> is stored in each individual. This result would be ambiguous among several >>> possible explanations. One possibility is that the probability distribution >>> that is being discovered reflects different dialects, such that 2/5 of the >>> population has one past tense, and 3/5 has another. Another possibility is >>> that each person has a similar probability distribution in their heads, >>> such that 2/5 of the time I respond one way, and 3/5 of the time I respond >>> another. Further experiments would be necessary to answer between these and >>> other possible theories (e.g., with repeated trials from the same person, >>> carefully planned so that the participants don't notice that they are being >>> asked multiple times). Without the quantitative evidence in the first >>> place, there is no way to answer these kinds of questions.>> >>>>> Regarding the past tense of "go", >>> this would be useful as a baseline in an experiment involving the less >>> frequent ones. So, yes, it would useful to gather quantitative evidence in >>> such a case also, as baselines with respect to the more interesting cases >>> for theories.>> >>>>> The bottom line: if you have a >>> generalization about a language that you wish to evaluate (such that you >>> hypothesize that it is true across the speakers of the language), then you >>> need quantitative evidence from multiple individuals, using an unbiased >>> data collection method, to evaluate such a claim. The point about >>> Mechanical Turk is that it is really *easy* to do this now, at least for >>> languages like English.>> >>>>> Best wishes, >>>>> >>>>> Ted Gibson & Ev Fedorenko >>>>> >>>>> On Sep 10, 2010, at 1:59 PM, Richard Hudson >>> wrote:>> >>>>>> Dear Ted, >>>>>> Thanks for the very interesting comment, >>> but are you REALLY saying that I shouldn't claim, for example, that the >>> past tense of GO is "went" without first cross-checking with 50 >>> native speakers?>>> >>>>>> Isn't there a danger of missing the >>> point that we all, as native speakers, spend our whole lives scanning other >>> people's linguistic behaviour (language 'out there', E-language) and trying >>> to explain it to ourselves in terms of a language system (language 'in >>> here', I-language)? So every judgement we make is based on thousands or >>> millions of observed exemplars, and reflects a unique experience of >>> E-language filtered through a unique I-language.>>> >>>>>> Given that view of language development, >>> I don't see how quantitative data will help. Let's take a real uncertainty, >>> such as the past tense of BID. If I want to say I did it, do I say "I >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get >>> 50 people to oblige on Mechanical Turk, and 20 of them give >>> "bidded" and 30 "bid". So what? Does that mean that the >>> correct answer is "bidded"? Surely not. How is it better than my >>> judgement? I agree you could record my speech and find how often I use each >>> alternative; but the reason I don't know is precisely because it's a rare >>> word, so in a sense quantitative data are irrelevant even there. What would >>> solve the problem of subjectivity, of course, would be a machine for >>> probing the bit of my mind (or even brain) that holds BID and its details; >>> but I suspect that even that wouldn't move us much further forward than my >>> original "don't know". (Incidentally I write as a fan of >>> quantitative sociolinguistics, so I do accept that quantitative data are >>> relevant to linguistic analysis in some areas, where the I-language >>> phenomenon is frequent enough to produce usable data.)>>> >>>>>> It seems to me that this discussion >>> raises the really fundamental question of what kind of thing we think >>> language is: social or individual. The problem isn't unique to linguistics >>> of course; it's the same throughout the social sciences. But what's special >>> about linguistics is that we deal in very fine details of culture (e.g. >>> details of how a particular word is used or pronounced) so the differences >>> between individuals really matter. I don't see that we're ever going to >>> have anything better than judgements to go on, so what we need is a way to >>> ensure that judgements are accurate reports of individual I-language. A >>> rotten situation for a science, but I don't see how it can get >>> better.>>> >>>>>> Dick >>>>>> >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm>>> >>>>>> On 10/09/2010 14:03, Ted Gibson >>> wrote:>>>> Dear Dan, Dick: >>>>>>> >>>>>>> I would like to clarify some points >>> that Dan Everett makes, in>>>> response to Dick Hudson. >>>>>>> >>>>>>> Ev Fedorenko and I have written a >>> couple of papers recently (Gibson &>>>> Fedorenko, 2010, in press, see >>> references and links below) on what we>>>> think are weak methodological >>> standards in syntax and semantics>>>> research over the past many years. >>> The issue that we address is the>>>> prevalent method in syntax and >>> semantics research, which involves>>>> obtaining a judgment of the >>> acceptability of a sentence / meaning>>>> pair, typically by just the author >>> of the paper, sometimes with>>>> feedback from colleagues. As we >>> address in our papers, this>>>> methodology does not allow proper >>> testing of scientific hypotheses>>>> because of (a) the small number of >>> experimental participants>>>> (typically one); (b) the small >>> number of experimental stimuli>>>> (typically one); (c) cognitive >>> biases on the part of the researcher>>>> and participants; and (d) the effect >>> of the preceding context (e.g.,>>>> other constructions the researcher >>> may have been recently>>>> considering). (As Dan said, see >>> Schutze, 1996; Cowart, 1997; and>>>> several others cited in Gibson & >>> Fedorenko, in press; for similar>>>> points, but with not as strong a >>> conclusion as ours).>>>> >>>>>>> Three issues need to be separated >>> here: (1) the use of intuitive>>>> judgments as a dependent measure in >>> a language experiment; (2)>>>> potential cognitive biases on the >>> part of experimental subjects and>>>> experimenters in language >>> experiments; and (3) the need for obtaining>>>> quantitative evidence, whatever the >>> dependent measure might be. The>>>> paper that Ev and I wrote addresses >>> the last two issues, but does not>>>> go into depth on the first issue >>> (the use of intuitions as a dependent>>>> measure in language experiments). >>> Regarding this issue, we don't think>>>> that there is anything wrong with >>> gathering intuitive judgments as a>>>> dependent measure, as long as the >>> task is clear to the experimental>>>> participants. >>>>>>> >>>>>>> In the longer paper (Gibson & >>> Fedorenko, in press) we respond to some>>>> arguments that have been given in >>> support of continuing to use the>>>> traditional non-quantitative method >>> in syntax / semantics research.>>>> One recent defense of the >>> traditional method comes from Phillips>>>> (2008), who argues that no harm has >>> come from the non-quantitative>>>> approach in syntax research thus >>> far. Phillips argues that there are>>>> no cases in the literature where an >>> incorrect intuitive judgment has>>>> become the basis for a widely >>> accepted generalization or an important>>>> theoretical claim. He therefore >>> concludes that there is no reason to>>>> adopt more rigorous data collection >>> standards. We challenge Philips???>>>> conclusion by presenting three cases >>> from the literature where a>>>> faulty intuition has led to >>> incorrect generalizations and mistaken>>>> theorizing, plausibly due to >>> cognitive biases on the part of the>>>> researchers. >>>>>>> >>>>>>> A second argument that is sometimes >>> presented for the continued use of>>>> the traditional non-quantitative >>> method is that it would be too>>>> inefficient to evaluate every >>> syntactic / semantic hypothesis or>>>> phenomenon quantitatively. For >>> example, Culicover & Jackendoff (2010)>>>> make this argument explicitly in >>> their response to Gibson & Fedorenko>>>> (2010): ???It would cripple >>> linguistic investigation if it were required>>>> that all judgments of >> ambiguity and >>> grammaticality be subject to>>>> statistically rigorous experiments >>> on naive subjects, especially when>>>> investigating languages whose >>> speakers are hard to access??? (Culicover>>>> & Jackendoff, 2010, p. 234). >>> (Dick Hudson makes a similar point>>>> earlier in the discussion here.) >>> Whereas we agree that in>>>> circumstances where gathering data >>> is difficult, some evidence is>>>> better than no evidence, we do not >>> agree that research would be slowed>>>> with respect to languages where >>> experimental participants are easy to>>>> access, such as English. In >>> contrast, we think that the opposite is>>>> true: the field???s progress >>> is probably slowed by not doing>>>> quantitative research. >>>>>>> Suppose that a typical syntax / >>> semantics paper that lacks>>>> quantitative evidence includes >>> judgments for 50 or more sentences />>>> meaning pairs, corresponding to 50 >>> or more empirical claims. Even if>>>> most of the judgments from such a >>> paper are correct or are on the>>>> right track, the problem is in >>> knowing which judgments are correct.>>>> For example, suppose that 90% of the >>> judgments from an arbitrary paper>>>> are correct (which is probably a >>> high estimate). (Colin Phillips and>>>> some of his former students / >>> postdocs have commented to us that, in>>>> their experience, quantitative >>> acceptability judgment studies almost>>>> always validate the claim(s) in the >>> literature. This is not our>>>> experience, however. Most >>> experiments that we have run which attempt>>>> to test some syntactic / semantic >>> hypothesis in the literature end up>>>> providing us with a pattern of data >>> that had not been known before the>>>> experiment (e.g., Breen et al., in >>> press; Fedorenko & Gibson, in>>>> press; Patel et al., 2009; Scontras >>> & Gibson, submitted).) This means>>>> that in a paper with 50 empirical >>> claims 45/50 are correct. But which>>>> 45? There are 2,118, 760 ways to >>> choose 45 items from 50. That???s over>>>> two million different theories. By >>> quantitatively evaluating the>>>> empirical claims, we reduce the >>> uncertainty a great deal. To make>>>> progress, it is better to have >>> theoretical claims supported by solid>>>> quantitative evidence, so that even >>> if the interpretation of the data>>>> changes over time as new evidence >>> becomes available ??? as is often the>>>> case in any field of science >>> ??? the empirical pattern can be used as a>>>> basis for further >>> theorizing.>>>> >>>>>>> Furthermore, it is no longer >>> expensive to run behavioral experiments,>>>> at least in English and other widely >>> spoken languages. There now>>>> exists a marketplace interface >>> ??? Amazon.com???s Mechanical Turk ??? which>>>> can be used for collecting >>> behavioral data over the internet quickly>>>> and inexpensively. The cost of using >>> an interface like this is>>>> minimal, and the time that it takes >>> for the results to be returned is>>>> short. For example, currently on >>> Mechanical Turk, a survey of>>>> approximately 50 items will be >>> answered by 50 or more participants>>>> within a couple of hours, at a cost >>> of approximately $1 per>>>> participant. Thus a survey can be >>> completed within a day, at a cost of>>>> less than $50. (The hard work of >>> designing the experiment, and>>>> constructing controlled materials >>> remains of course.)>>>> >>>>>>> Sorry to be so verbose. But I think >>> that these methodological points>>>> are very important. >>>>>>> >>>>>>> Best wishes, >>>>>>> >>>>>>> Ted Gibson >>>>>>> >>>>>>> Gibson, E. & Fedorenko, E. (In >>> press). The need for quantitative>>>> methods in syntax and semantics >>> research. Language and Cognitive>>>> Processes. >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson>>>> & Fedorenko InPress >>> LCP.pdf>>>> >>>>>>> Gibson, E. & Fedorenko, E. >>> (2010). Weak quantitative standards in>>>> linguistics research. Trends in >>> Cognitive Science, 14, 233-234.>>>> >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko>>>> 2010 >> TICS.pdf >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>>> Dick, >>>>>>>> >>>>>>>> You raise an important issue >>> here about methodology. I believe that>>>>> intuitions are a fine way to >>> generate hypotheses and even to test>>>>> them - to a degree. But while it >>> might not have been feasible for>>>>> Huddleston, Pullum, and the >>> other contributors to the Cambridge>>>>> Grammar to conduct experiments >>> on every point of the grammar,>>>>> experiments could have only made >>> the grammar better. The use of>>>>> intuitions, corpora, and >>> standard psycholinguistic experimentation>>>>> (indeed, Standard Social Science >>> Methodology) is vital for taking the>>>>> field forward and for providing >>> the best support for different>>>>> analyses. Ted Gibson and Ev >>> Fedorenko have written a very useful new>>>>> paper on this, showing serious >>> shortcomings with intuitions as the>>>>> sole source of evidence, in >>> their paper: "The need for quantitative>>>>> methods in syntax and semantics >>> research".>>>>> >>>>>>>> Carson Schutze and Wayne Cowart, >>> among others, have also written>>>>> convincingly on this. >>>>>>>> >>>>>>>> It is one reason that a team >>> from Stanford, MIT (Brain and Cognitive>>>>> Science), and researchers from >>> Brazil are beginning a third round of>>>>> experimental work among the >>> Pirahas, since my own work on the syntax>>>>> was, like almost every other >>> field researcher's, based on native>>>>> speaker intuitions and >>> corpora.>>>>> >>>>>>>> The discussion of methodologies >>> reminds me of the initial reactions>>>>> to Greenberg's work on >>> classifying the languages of the Americas. His>>>>> methods were strongly (and >>> justifiably) criticized. However, I always>>>>> thought that his methods were a >>> great way of generating hypotheses,>>>>> so long as they were ultimately >>> put to the test of standard>>>>> historical linguistics methods. >>> And the same seems true for use of>>>>> native-speaker >>> intuitions.>>>>> >>>>>>>> -- Dan >>>>>>> >>>>>>> >>>>>>> >>>>>>>>> We linguists can add a >>> further layer of explanation to the>>>>>> judgements, but some >>> judgements do seem to be more reliable than>>>>>> others. And if we have to >>> wait for psycholinguistic evidence for>>>>>> every detailed analysis we >>> make, our whole discipline will>>>>>> immediately grind to a halt. >>> Like it or not, native speaker>>>>>> judgements are what put us >>> linguists ahead of the rest in handling>>>>>> fine detail. Imagine writing >>> the Cambridge Grammar of the English>>>>>> Language (or the OED) >>> without using native speaker judgements.>>>>>> >>>>>>>>> Best wishes, Dick >>> Hudson>>>> >>>>>>> >>>>>>> >>>>> >>>>> >>>>> >>>> >>> >>> >>> >>> >>> >> >> > > From bischoff.st at gmail.com Sun Sep 12 02:30:20 2010 From: bischoff.st at gmail.com (s.t. bischoff) Date: Sat, 11 Sep 2010 22:30:20 -0400 Subject: FUNKNET Digest, Vol 84, Issue 10 In-Reply-To: Message-ID: Ted & Ev write... ____________ We have encountered a claim that the reason for different kinds of methods being used across the different fields of language study (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the research questions are different across these fields, and some methods may be better suited to ask some questions than others. Although the latter is likely true, the premise ? that the research questions are different across the fields ? is false... In our opinion then, the distinction between the fields of linguistics and psycho-/neuro-linguistics is purely along the lines of the kind of data that are used as evidence for or against theoretical hypotheses: typically non-quantitative data in linguistics vs. typically quantitative data in psycho-/neuro-linguistics. Given the superficial nature of this distinction, we think that there should be one field of language study where a wide range of dependent measures is used to investigate linguistic representations and computations. _____________ I would be curious to know how many on this listserv agree with the above...it seems to me that the assertion rests on a very narrow definition of "linguistics"...if language is viewed as a meso-oject and linguistics thus as a meso-science (perhaps it is not), it doesn't seem that the assertion necessarily holds...I'm not convinced that all linguistic inquiry is about computations and algorithms (despite my research being primarily in formal and computational linguistics)...nor am I convinced that the root of all linguistic inquiry is grounded in representations and computations (though I would be curious to know what "representations" refers to here as I could be misreading it)...in addition it seems that the object of study in neuro-linguistics and phsycholinguistics is the mind (possibly) and the brain (certainly) but for many linguists that doesn't seem to be the case and I do think this leads to different reseach questions...but then again I'm the one that started all this by asking the naive question about "unhappiness" in the first place. Cheers, Shannon On Sat, Sep 11, 2010 at 1:00 PM, wrote: > Send FUNKNET mailing list submissions to > funknet at mailman.rice.edu > > To subscribe or unsubscribe via the World Wide Web, visit > https://mailman.rice.edu/mailman/listinfo/funknet > or, via email, send a message with subject or body 'help' to > funknet-request at mailman.rice.edu > > You can reach the person managing the list at > funknet-owner at mailman.rice.edu > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of FUNKNET digest..." > > > Today's Topics: > > 1. Re: "Relative clauses" with no relativized element > (Suzanne Kemmer) > 2. Re: "Relative clauses" with no relativized element (Ron Kuzar) > 3. Re: "Relative clauses" with no relativized element (Tom Givon) > 4. Re: "Relative clauses" with no relativized element > (Giuliana Fiorentino) > 5. Re: "Relative clauses" with no relativized ele (E.G.) > 6. Re: "Relative clauses" with no relativized element > (Suzanne Kemmer) > 7. Re: "Relative clauses" with no relativized element (E.G.) > 8. Re: analysis: unhappiness (Richard Hudson) > 9. Re: analysis: unhappiness (Ted Gibson) > 10. Re: analysis: unhappiness (dryer at buffalo.edu) > 11. Re: analysis: unhappiness (A. Katz) > 12. Re: "Relative clauses" with no relativized ele (Ron Kuzar) > 13. Re: analysis: unhappiness (Brian MacWhinney) > 14. Re: analysis: unhappiness (dryer at buffalo.edu) > 15. Re: analysis: unhappiness (Richard Hudson) > 16. Re: analysis: unhappiness (Richard Hudson) > 17. Re: analysis: unhappiness (Lise Menn) > 18. Re: analysis: unhappiness (Lise Menn) > 19. Re: analysis: unhappiness (Daniel Everett) > 20. Re: "Relative clauses" with no relativized element > (Philippe De Brabanter) > 21. Re: analysis: unhappiness (Richard Hudson) > 22. Re: analysis: unhappiness (A. Katz) > 23. Re: FUNKNET] analysis: unhappiness (Chris Butler) > 24. Re: FUNKNET] analysis: unhappiness (A. Katz) > 25. Re: "Relative clauses" with no relativized ele (E.G.) > 26. Job Advertisement (Kristine Hildebrandt) > 27. Re: analysis: unhappiness (Ted Gibson) > 28. Re: analysis: unhappiness (Daniel Everett) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 10 Sep 2010 12:16:09 -0500 > From: Suzanne Kemmer > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: FUNKNET > Message-ID: > Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes > > In Generative Syntax these clauses were viewed as complement clauses > with an NP head, distinct > from relative clauses but having some parallels with them. I think it > was Joan Bresnan that > brought out the parallels and distinctions, maybe in her doctoral > dissertation . As I recall (but > my remembrance may be faulty), Bresnan named the > THAT element a COMP for complementizer. > > The term 'appositive' isn't very good because in traditional grammar > that is reserved for an 'UNrestrictive' relation of a noun and its > complement--an incidental description of > a head N's referent rather than a specification of which referent > ("the tree, a live oak, survived another 100 years or so"). > > In Cognitive Grammar nouns like claim, statement, idea, realization, > belief etc. are in almost all cases nominalizations of 'viewing > predicates' > (verbs like claim, believe, etc.) that introduce on-stage > predications 'viewed' by a conceptualizer (the person doing the > claiming, etc.). (the viewing > predicates are space builders in Fauconnier's mental spaces terminology) > > For the nominalizations of these predicates, the semantics of the > nouns intrinsically has an "e-site" or elaboration site > that allows for spelling out the content of the viewed predicate in > the form of a complement clause. The e-site > inherent to the semantics of the nouns is parallel to the e-site > inherent to the semantics of the corresponding verbs. > > There are a few cases I can think of of nouns that have 'viewing > predicate' e-sites but don't have corresponding verbs . > For example > the noun _view_ "The view that global climate change is > anthropogenic is widely held by scientists" > ( ' X views that (proposition)' is not possible, only 'X views Y > as ...' , with a restriction to equative or descriptive propositions). > Also _idea_----the verb has to be changed to something like 'believe' > to make a corresponding full predicate. > > I view (!) these nouns as semantically parallel in interesting ways to > picture nouns. The conceptualizer (viewer) in > both cases can designate the noun in a possessive phrase, but after > that the syntax diverges. > > --Suzanne > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > > And as another addition: the clauses that can only be introduced by > > *that* (with no role to > > play in the subordinate clause) may be seen as (subtypes of) > > complement clauses, > > expressing a proposition with the noun functioning as Complement > > Taking Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, > > etc. (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not > > relatives. Cf. constructions like > > "The claim is that X" (traditionally analysed as subject clauses), > > "I claim that X", "I put forward > > the claim that X", in which the relationship between the verb or > > noun and the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > >> To add to Joanne's comments: > >> > >> There are basically three ways to distinguish an appositive clause > >> (AC) from a relative clause (RC). > >> > >> 1) An AC can only be led by *that* while an RC can be led by other > >> pronouns. > >> > >> 2) The AC and the noun it modifies display an equative > >> relationship--one can say X > >> (denoted by the noun) is Y (presented by the appositive)--while an > >> RC often doesn't > >> (except, perhaps, when the relative clause is sentential). > > > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but > >> a relative pronoun > >> is always part of the clause in an RC. > >> > >> Rong Chen > >> > > > > > > > > ------------------------------ > > Message: 2 > Date: Fri, 10 Sep 2010 20:42:24 +0300 > From: Ron Kuzar > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: FUNKNET > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > A thorough discussion of the head nouns and their relation with their > complement clauses may be found in Hans-Joerg Schmid's book on Shell > Nouns (this is his term for the head nouns). > Ron Kuzar > --------- > On Fri, Sep 10, 2010 at 8:16 PM, Suzanne Kemmer wrote: > > > In Generative Syntax these clauses were viewed as complement clauses with > > an NP head, distinct > > from relative clauses but having some parallels with them. I think it > was > > Joan Bresnan that > > brought out the parallels and distinctions, maybe in her doctoral > > dissertation . As I recall (but > > my remembrance may be faulty), Bresnan named the > > THAT element a COMP for complementizer. > > > > The term 'appositive' isn't very good because in traditional grammar > > that is reserved for an 'UNrestrictive' relation of a noun and its > > complement--an incidental description of > > a head N's referent rather than a specification of which referent ("the > > tree, a live oak, survived another 100 years or so"). > > > > In Cognitive Grammar nouns like claim, statement, idea, realization, > > belief etc. are in almost all cases nominalizations of 'viewing > predicates' > > (verbs like claim, believe, etc.) that introduce on-stage predications > > 'viewed' by a conceptualizer (the person doing the claiming, etc.). (the > > viewing > > predicates are space builders in Fauconnier's mental spaces terminology) > > > > For the nominalizations of these predicates, the semantics of the nouns > > intrinsically has an "e-site" or elaboration site > > that allows for spelling out the content of the viewed predicate in the > > form of a complement clause. The e-site > > inherent to the semantics of the nouns is parallel to the e-site inherent > > to the semantics of the corresponding verbs. > > > > There are a few cases I can think of of nouns that have 'viewing > > predicate' e-sites but don't have corresponding verbs . > > For example > > the noun _view_ "The view that global climate change is anthropogenic > is > > widely held by scientists" > > ( ' X views that (proposition)' is not possible, only 'X views Y as ...' > , > > with a restriction to equative or descriptive propositions). > > Also _idea_----the verb has to be changed to something like 'believe' to > > make a corresponding full predicate. > > > > I view (!) these nouns as semantically parallel in interesting ways to > > picture nouns. The conceptualizer (viewer) in > > both cases can designate the noun in a possessive phrase, but after that > > the syntax diverges. > > > > --Suzanne > > > > On Sep 10, 2010, at 11:21 AM, Arie Verhagen wrote: > > > > And as another addition: the clauses that can only be introduced by > *that* > >> (with no role to > >> play in the subordinate clause) may be seen as (subtypes of) complement > >> clauses, > >> expressing a proposition with the noun functioning as Complement Taking > >> Predicate (CTP), > >> expressing a propostional attitude, epistemic/evaluative stance, etc. > >> (following analyses by > >> Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > >> constructions like > >> "The claim is that X" (traditionally analysed as subject clauses), "I > >> claim that X", "I put forward > >> the claim that X", in which the relationship between the verb or noun > and > >> the that-clause is > >> comparable to the one in "The claim that X". > >> > >> --Arie Verhagen > >> > >> ---------------- > >> Message from Rong Chen > >> 10 Sep 2010, 23:42 > >> > >> Subject: Re: [FUNKNET] "Relative clauses" with no relativi > >> > >> To add to Joanne's comments: > >>> > >>> There are basically three ways to distinguish an appositive clause > >>> (AC) from a relative clause (RC). > >>> > >>> 1) An AC can only be led by *that* while an RC can be led by other > >>> pronouns. > >>> > >>> 2) The AC and the noun it modifies display an equative > relationship--one > >>> can say X > >>> (denoted by the noun) is Y (presented by the appositive)--while an RC > >>> often doesn't > >>> (except, perhaps, when the relative clause is sentential). > >>> > >> > >> 3)--which Tom noted--*that* is not part of the clause in an AC; but a > >>> relative pronoun > >>> is always part of the clause in an RC. > >>> > >>> Rong Chen > >>> > >>> > >> > >> > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > ------------------------------ > > Message: 3 > Date: Fri, 10 Sep 2010 11:43:34 -0600 > From: Tom Givon > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: "Thomas E. Payne" > Cc: FUNKNET > Message-ID: <4C8A6E46.3040301 at uoregon.edu> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), > vol. I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this > construction described and analyzed as the product of nominalization of > clauses with verbs that take verbal complements ('know', 'think', 'say' > etc. The preceding section (6.6.3. "Complex NP's arising through > nominalization", p. 287) deals more generally with nominalizations. The > term "noun complements" was used in syntax classes at UCLA in the mid > 1960s, so certainly Joan Bresnan did not invent it. Best, TG > > =========== > > Thomas E. Payne wrote: > > Can anyone help me name the following structure in English, and maybe > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > > > ------------------------------ > > Message: 4 > Date: Fri, 10 Sep 2010 19:53:15 +0200 > From: "Giuliana Fiorentino" > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: "Thomas E. Payne" , "FUNKNET" > > Message-ID: <036386D3933D4CB99389FFB2576ADBE8 at giuliana> > Content-Type: text/plain; charset="iso-8859-1" > > Hi Tom, > clauses like: > > The importance of being Earnest > the fact of being late > the fact that you are late > the idea that world is round > etcetera > > are not relative clauses but can be considered among syntactic strategies > in order to nominalise events after a generic noun (working as a classifier > for nominalised events). > > Giuliana > > ----- Original Message ----- > From: Thomas E. Payne > To: FUNKNET > Sent: Friday, September 10, 2010 4:16 PM > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > Can anyone help me name the following structure in English, and maybe > point > me to some references? I do not find reference to this in the Cambridge > Grammar of the English Language or any other of my English grammar books. > But then, maybe I just don't know where to look. > > Here are two examples from a play: > > His protestations of devotion in the trial scene are, in our opinion, > genuine, as is his confession [that his affair with the Countess is > platonic]. > > The bracketed clause seems to modify "confession", though there is no > position for a confession in the clause itself. > > . . . forced hither with an impious black design [to have my innocence and > youth become the sacrifice of brutal violence]. > > Here the bracketed non-finite clause seems to modify "design." > > These are not all that rare. I'm reminded of examples like: > > "The claim [that my client is a murderer] is totally false." > > Are these relative clauses? If so what kind? Thanks for any help. > > Tom Payne > > > ------------------------------ > > Message: 5 > Date: Fri, 10 Sep 2010 20:54:06 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: Arie Verhagen , > funknet at mailman.rice.edu > Message-ID: > > > > Content-Type: text/plain; charset=ISO-8859-1 > > Hi all, > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > comparison can help what seems to be a purely theoretical question based on > a single language. The problem here is that English uses the same element > to > mark regular relatives and these "appositional" relatives. But if at least > one language encodes them by different means, then there's at least a good > case for seeing them as distinct functions. It's basically the same > principle that's used to decide whether to put a meaning on a semantic map. > So here are two languages that I know that encode them differently. > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > clause (ki), which differs from the relative clause marker (Se-), e.g. > > ha-hoda'a Se-kibalnu > the-announcment rel-we_got > "The announcement that we got." > > ha-hoda'a ki hitbatel ha-mifgaS > the-message CMP was_cancelled the-meeting > "The announcement that the meeting was cancelled." > > In Coptic, these clauses are marked by ce-, which marks complement clauses, > *inter alia*, but not relative clauses: > > ph-mewi ce- (complement clause) > 'the-thought that (we are angry)' > > ph-mewi ete- (relative clause) > 'the thought that (we used to think)' > > This seems to be a pretty clear indication that these are complement > clauses > rather than relatives. Even if one doesn't like the notion of nouns taking > complement clauses (and why not? nominalizations in some languages can take > accusative modifiers as well as genitives), it still probably isn't > incidental that the nominalizations are of verbs that take complement > clauses when finite. > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > looking at. > > Best, > Eitan > > > On 10 September 2010 19:21, Arie Verhagen > wrote: > > > And as another addition: the clauses that can only be introduced by > *that* > > (with no role to > > play in the subordinate clause) may be seen as (subtypes of) complement > > clauses, > > expressing a proposition with the noun functioning as Complement Taking > > Predicate (CTP), > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > (following analyses by > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > Cf. > > constructions like > > "The claim is that X" (traditionally analysed as subject clauses), "I > claim > > that X", "I put forward > > the claim that X", in which the relationship between the verb or noun and > > the that-clause is > > comparable to the one in "The claim that X". > > > > --Arie Verhagen > > > > ---------------- > > Message from Rong Chen > > 10 Sep 2010, 23:42 > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > To add to Joanne's comments: > > > > > > There are basically three ways to distinguish an appositive clause > > > (AC) from a relative clause (RC). > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > pronouns. > > > > > > 2) The AC and the noun it modifies display an equative > relationship--one > > can say X > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > often doesn't > > > (except, perhaps, when the relative clause is sentential). > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > relative pronoun > > > is always part of the clause in an RC. > > > > > > Rong Chen > > > > > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 6 > Date: Fri, 10 Sep 2010 12:55:30 -0500 > From: Suzanne Kemmer > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: Funknet > Message-ID: > Content-Type: text/plain; charset=us-ascii > > > > Talmy, yes Joan B did not invent noun complements nor > the term for them; but my recollection, in passing, was that she > named 'that' in such structures as "complementizer" (and I also > recall that she referred to 'that' relativizers with the same term, > while recognizing other differences between the two structures.) > > I may be wrong on that, but it's a different recollection, claim, or > whatever, > than the one you refer to. > > Not all the head nouns are nominalizations, but most are. > S. > > On Sep 10, 2010, at 12:43 PM, Tom Givon wrote: > > > > > > > Looking through my "English Grammar (Amsterdam: J. Benjamins, 1993), vol. > I, ch. 6, section 6.6.4. "Noun complements", p. 298, I find this > construction described and analyzed as the product of nominalization of > clauses with verbs that take verbal complements ('know', 'think', 'say' etc. > The preceding section (6.6.3. "Complex NP's arising through nominalization", > p. 287) deals more generally with nominalizations. The term "noun > complements" was used in syntax classes at UCLA in the mid 1960s, so > certainly Joan Bresnan did not invent it. Best, TG > > > > =========== > > > > Thomas E. Payne wrote: > >> Can anyone help me name the following structure in English, and maybe > point > >> me to some references? I do not find reference to this in the Cambridge > >> Grammar of the English Language or any other of my English grammar > books. > >> But then, maybe I just don't know where to look. > >> > >> Here are two examples from a play: > >> > >> His protestations of devotion in the trial scene are, in our opinion, > >> genuine, as is his confession [that his affair with the Countess is > >> platonic]. > >> > >> The bracketed clause seems to modify "confession", though there is no > >> position for a confession in the clause itself. > >> > >> . . . forced hither with an impious black design [to have my innocence > and > >> youth become the sacrifice of brutal violence]. > >> > >> Here the bracketed non-finite clause seems to modify "design." > >> > >> These are not all that rare. I'm reminded of examples like: > >> > >> "The claim [that my client is a murderer] is totally false." > >> > >> Are these relative clauses? If so what kind? Thanks for any help. > >> > >> Tom Payne > >> > >> > > > > > > > > ------------------------------ > > Message: 7 > Date: Fri, 10 Sep 2010 20:56:23 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: funknet at mailman.rice.edu > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Jespersen and his nexus-substantives should be mentioned (Philosophy of > Grammar, 1924). Also in his MEG and Analytic Syntax one could find > interesting discussions. > > Eitan > > > On 10 September 2010 20:53, Giuliana Fiorentino < > giuliana.fiorentino at unimol.it> wrote: > > > Hi Tom, > > clauses like: > > > > The importance of being Earnest > > the fact of being late > > the fact that you are late > > the idea that world is round > > etcetera > > > > are not relative clauses but can be considered among syntactic strategies > > in order to nominalise events after a generic noun (working as a > classifier > > for nominalised events). > > > > Giuliana > > > > ----- Original Message ----- > > From: Thomas E. Payne > > To: FUNKNET > > Sent: Friday, September 10, 2010 4:16 PM > > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > > > > Can anyone help me name the following structure in English, and maybe > > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar > books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 8 > Date: Fri, 10 Sep 2010 18:59:12 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Message-ID: <4C8A71F0.4050507 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Ted, > Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" > without first cross-checking with 50 native speakers? > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic > behaviour (language 'out there', E-language) and trying to explain it to > ourselves in terms of a language system (language 'in here', > I-language)? So every judgement we make is based on thousands or > millions of observed exemplars, and reflects a unique experience of > E-language filtered through a unique I-language. > > Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical > Turk, and 20 of them give "bidded" and 30 "bid". So what? Does that mean > that the correct answer is "bidded"? Surely not. How is it better than > my judgement? I agree you could record my speech and find how often I > use each alternative; but the reason I don't know is precisely because > it's a rare word, so in a sense quantitative data are irrelevant even > there. What would solve the problem of subjectivity, of course, would be > a machine for probing the bit of my mind (or even brain) that holds BID > and its details; but I suspect that even that wouldn't move us much > further forward than my original "don't know". (Incidentally I write as > a fan of quantitative sociolinguistics, so I do accept that quantitative > data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or > individual. The problem isn't unique to linguistics of course; it's the > same throughout the social sciences. But what's special about > linguistics is that we deal in very fine details of culture (e.g. > details of how a particular word is used or pronounced) so the > differences between individuals really matter. I don't see that we're > ever going to have anything better than judgements to go on, so what we > need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see > how it can get better. > > Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 14:03, Ted Gibson wrote: > > Dear Dan, Dick: > > > > I would like to clarify some points that Dan Everett makes, in > > response to Dick Hudson. > > > > Ev Fedorenko and I have written a couple of papers recently (Gibson & > > Fedorenko, 2010, in press, see references and links below) on what we > > think are weak methodological standards in syntax and semantics > > research over the past many years. The issue that we address is the > > prevalent method in syntax and semantics research, which involves > > obtaining a judgment of the acceptability of a sentence / meaning > > pair, typically by just the author of the paper, sometimes with > > feedback from colleagues. As we address in our papers, this > > methodology does not allow proper testing of scientific hypotheses > > because of (a) the small number of experimental participants > > (typically one); (b) the small number of experimental stimuli > > (typically one); (c) cognitive biases on the part of the researcher > > and participants; and (d) the effect of the preceding context (e.g., > > other constructions the researcher may have been recently > > considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > > several others cited in Gibson & Fedorenko, in press; for similar > > points, but with not as strong a conclusion as ours). > > > > Three issues need to be separated here: (1) the use of intuitive > > judgments as a dependent measure in a language experiment; (2) > > potential cognitive biases on the part of experimental subjects and > > experimenters in language experiments; and (3) the need for obtaining > > quantitative evidence, whatever the dependent measure might be. The > > paper that Ev and I wrote addresses the last two issues, but does not > > go into depth on the first issue (the use of intuitions as a dependent > > measure in language experiments). Regarding this issue, we don't think > > that there is anything wrong with gathering intuitive judgments as a > > dependent measure, as long as the task is clear to the experimental > > participants. > > > > In the longer paper (Gibson & Fedorenko, in press) we respond to some > > arguments that have been given in support of continuing to use the > > traditional non-quantitative method in syntax / semantics research. > > One recent defense of the traditional method comes from Phillips > > (2008), who argues that no harm has come from the non-quantitative > > approach in syntax research thus far. Phillips argues that there are > > no cases in the literature where an incorrect intuitive judgment has > > become the basis for a widely accepted generalization or an important > > theoretical claim. He therefore concludes that there is no reason to > > adopt more rigorous data collection standards. We challenge Philips? > > conclusion by presenting three cases from the literature where a > > faulty intuition has led to incorrect generalizations and mistaken > > theorizing, plausibly due to cognitive biases on the part of the > > researchers. > > > > A second argument that is sometimes presented for the continued use of > > the traditional non-quantitative method is that it would be too > > inefficient to evaluate every syntactic / semantic hypothesis or > > phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > > make this argument explicitly in their response to Gibson & Fedorenko > > (2010): ?It would cripple linguistic investigation if it were required > > that all judgments of ambiguity and grammaticality be subject to > > statistically rigorous experiments on naive subjects, especially when > > investigating languages whose speakers are hard to access? (Culicover > > & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > > earlier in the discussion here.) Whereas we agree that in > > circumstances where gathering data is difficult, some evidence is > > better than no evidence, we do not agree that research would be slowed > > with respect to languages where experimental participants are easy to > > access, such as English. In contrast, we think that the opposite is > > true: the field?s progress is probably slowed by not doing > > quantitative research. > > Suppose that a typical syntax / semantics paper that lacks > > quantitative evidence includes judgments for 50 or more sentences / > > meaning pairs, corresponding to 50 or more empirical claims. Even if > > most of the judgments from such a paper are correct or are on the > > right track, the problem is in knowing which judgments are correct. > > For example, suppose that 90% of the judgments from an arbitrary paper > > are correct (which is probably a high estimate). (Colin Phillips and > > some of his former students / postdocs have commented to us that, in > > their experience, quantitative acceptability judgment studies almost > > always validate the claim(s) in the literature. This is not our > > experience, however. Most experiments that we have run which attempt > > to test some syntactic / semantic hypothesis in the literature end up > > providing us with a pattern of data that had not been known before the > > experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > > press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > > that in a paper with 50 empirical claims 45/50 are correct. But which > > 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > > two million different theories. By quantitatively evaluating the > > empirical claims, we reduce the uncertainty a great deal. To make > > progress, it is better to have theoretical claims supported by solid > > quantitative evidence, so that even if the interpretation of the data > > changes over time as new evidence becomes available ? as is often the > > case in any field of science ? the empirical pattern can be used as a > > basis for further theorizing. > > > > Furthermore, it is no longer expensive to run behavioral experiments, > > at least in English and other widely spoken languages. There now > > exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > > can be used for collecting behavioral data over the internet quickly > > and inexpensively. The cost of using an interface like this is > > minimal, and the time that it takes for the results to be returned is > > short. For example, currently on Mechanical Turk, a survey of > > approximately 50 items will be answered by 50 or more participants > > within a couple of hours, at a cost of approximately $1 per > > participant. Thus a survey can be completed within a day, at a cost of > > less than $50. (The hard work of designing the experiment, and > > constructing controlled materials remains of course.) > > > > Sorry to be so verbose. But I think that these methodological points > > are very important. > > > > Best wishes, > > > > Ted Gibson > > > > Gibson, E. & Fedorenko, E. (In press). The need for quantitative > > methods in syntax and semantics research. Language and Cognitive > > Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > > & Fedorenko InPress LCP.pdf > > > > Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > > linguistics research. Trends in Cognitive Science, 14, 233-234. > > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > > 2010 TICS.pdf > > > > > > > > > >> Dick, > >> > >> You raise an important issue here about methodology. I believe that > >> intuitions are a fine way to generate hypotheses and even to test > >> them - to a degree. But while it might not have been feasible for > >> Huddleston, Pullum, and the other contributors to the Cambridge > >> Grammar to conduct experiments on every point of the grammar, > >> experiments could have only made the grammar better. The use of > >> intuitions, corpora, and standard psycholinguistic experimentation > >> (indeed, Standard Social Science Methodology) is vital for taking the > >> field forward and for providing the best support for different > >> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >> paper on this, showing serious shortcomings with intuitions as the > >> sole source of evidence, in their paper: "The need for quantitative > >> methods in syntax and semantics research". > >> > >> Carson Schutze and Wayne Cowart, among others, have also written > >> convincingly on this. > >> > >> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >> Science), and researchers from Brazil are beginning a third round of > >> experimental work among the Pirahas, since my own work on the syntax > >> was, like almost every other field researcher's, based on native > >> speaker intuitions and corpora. > >> > >> The discussion of methodologies reminds me of the initial reactions > >> to Greenberg's work on classifying the languages of the Americas. His > >> methods were strongly (and justifiably) criticized. However, I always > >> thought that his methods were a great way of generating hypotheses, > >> so long as they were ultimately put to the test of standard > >> historical linguistics methods. And the same seems true for use of > >> native-speaker intuitions. > >> > >> -- Dan > > > > > > > >>> We linguists can add a further layer of explanation to the > >>> judgements, but some judgements do seem to be more reliable than > >>> others. And if we have to wait for psycholinguistic evidence for > >>> every detailed analysis we make, our whole discipline will > >>> immediately grind to a halt. Like it or not, native speaker > >>> judgements are what put us linguists ahead of the rest in handling > >>> fine detail. Imagine writing the Cambridge Grammar of the English > >>> Language (or the OED) without using native speaker judgements. > >>> > >>> Best wishes, Dick Hudson > > > > > > > > > ------------------------------ > > Message: 9 > Date: Fri, 10 Sep 2010 14:30:16 -0400 > From: Ted Gibson > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Evelina Fedorenko , funknet > > Message-ID: > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dear Dick: > > Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are > saying is that if you have some testable claim involving a general > hypothesis about a language, then you need to get quantitative data > from unbiased sources to evaluate that claim. If you are interested > in English past tense morphology, then depending on the claims that > you might want to investigate, there are lots of ways to get relevant > quantitative evidence. Corpus data will probably be useful. For very > low frequency words, you can run experiments to test behavior with > respect to such words. > > Your example of the past tense of "bid" is a fine such example. You > can run an experiment like the one you suggested to find out what > people think the past tense is. If you then found that 20/50 people > responded "bidded" and 30/50 respond "bid", that is a lot of useful > information. As you suggest in your discussion, this result wouldn't > answer the question of how past tense is stored in each individual. > This result would be ambiguous among several possible explanations. > One possibility is that the probability distribution that is being > discovered reflects different dialects, such that 2/5 of the > population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution > in their heads, such that 2/5 of the time I respond one way, and 3/5 > of the time I respond another. Further experiments would be necessary > to answer between these and other possible theories (e.g., with > repeated trials from the same person, carefully planned so that the > participants don't notice that they are being asked multiple times). > Without the quantitative evidence in the first place, there is no way > to answer these kinds of questions. > > Regarding the past tense of "go", this would be useful as a baseline > in an experiment involving the less frequent ones. So, yes, it would > useful to gather quantitative evidence in such a case also, as > baselines with respect to the more interesting cases for theories. > > The bottom line: if you have a generalization about a language that > you wish to evaluate (such that you hypothesize that it is true across > the speakers of the language), then you need quantitative evidence > from multiple individuals, using an unbiased data collection method, > to evaluate such a claim. The point about Mechanical Turk is that it > is really *easy* to do this now, at least for languages like English. > > Best wishes, > > Ted Gibson & Ev Fedorenko > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > Dear Ted, > > Thanks for the very interesting comment, but are you REALLY saying > > that I shouldn't claim, for example, that the past tense of GO is > > "went" without first cross-checking with 50 native speakers? > > > > Isn't there a danger of missing the point that we all, as native > > speakers, spend our whole lives scanning other people's linguistic > > behaviour (language 'out there', E-language) and trying to explain > > it to ourselves in terms of a language system (language 'in here', I- > > language)? So every judgement we make is based on thousands or > > millions of observed exemplars, and reflects a unique experience of > > E-language filtered through a unique I-language. > > > > Given that view of language development, I don't see how > > quantitative data will help. Let's take a real uncertainty, such as > > the past tense of BID. If I want to say I did it, do I say "I > > bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > > people to oblige on Mechanical Turk, and 20 of them give "bidded" > > and 30 "bid". So what? Does that mean that the correct answer is > > "bidded"? Surely not. How is it better than my judgement? I agree > > you could record my speech and find how often I use each > > alternative; but the reason I don't know is precisely because it's a > > rare word, so in a sense quantitative data are irrelevant even > > there. What would solve the problem of subjectivity, of course, > > would be a machine for probing the bit of my mind (or even brain) > > that holds BID and its details; but I suspect that even that > > wouldn't move us much further forward than my original "don't know". > > (Incidentally I write as a fan of quantitative sociolinguistics, so > > I do accept that quantitative data are relevant to linguistic > > analysis in some areas, where the I-language phenomenon is frequent > > enough to produce usable data.) > > > > It seems to me that this discussion raises the really fundamental > > question of what kind of thing we think language is: social or > > individual. The problem isn't unique to linguistics of course; it's > > the same throughout the social sciences. But what's special about > > linguistics is that we deal in very fine details of culture (e.g. > > details of how a particular word is used or pronounced) so the > > differences between individuals really matter. I don't see that > > we're ever going to have anything better than judgements to go on, > > so what we need is a way to ensure that judgements are accurate > > reports of individual I-language. A rotten situation for a science, > > but I don't see how it can get better. > > > > Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 14:03, Ted Gibson wrote: > >> Dear Dan, Dick: > >> > >> I would like to clarify some points that Dan Everett makes, in > >> response to Dick Hudson. > >> > >> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >> Fedorenko, 2010, in press, see references and links below) on what we > >> think are weak methodological standards in syntax and semantics > >> research over the past many years. The issue that we address is the > >> prevalent method in syntax and semantics research, which involves > >> obtaining a judgment of the acceptability of a sentence / meaning > >> pair, typically by just the author of the paper, sometimes with > >> feedback from colleagues. As we address in our papers, this > >> methodology does not allow proper testing of scientific hypotheses > >> because of (a) the small number of experimental participants > >> (typically one); (b) the small number of experimental stimuli > >> (typically one); (c) cognitive biases on the part of the researcher > >> and participants; and (d) the effect of the preceding context (e.g., > >> other constructions the researcher may have been recently > >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >> several others cited in Gibson & Fedorenko, in press; for similar > >> points, but with not as strong a conclusion as ours). > >> > >> Three issues need to be separated here: (1) the use of intuitive > >> judgments as a dependent measure in a language experiment; (2) > >> potential cognitive biases on the part of experimental subjects and > >> experimenters in language experiments; and (3) the need for obtaining > >> quantitative evidence, whatever the dependent measure might be. The > >> paper that Ev and I wrote addresses the last two issues, but does not > >> go into depth on the first issue (the use of intuitions as a > >> dependent > >> measure in language experiments). Regarding this issue, we don't > >> think > >> that there is anything wrong with gathering intuitive judgments as a > >> dependent measure, as long as the task is clear to the experimental > >> participants. > >> > >> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >> arguments that have been given in support of continuing to use the > >> traditional non-quantitative method in syntax / semantics research. > >> One recent defense of the traditional method comes from Phillips > >> (2008), who argues that no harm has come from the non-quantitative > >> approach in syntax research thus far. Phillips argues that there are > >> no cases in the literature where an incorrect intuitive judgment has > >> become the basis for a widely accepted generalization or an important > >> theoretical claim. He therefore concludes that there is no reason to > >> adopt more rigorous data collection standards. We challenge Philips? > >> conclusion by presenting three cases from the literature where a > >> faulty intuition has led to incorrect generalizations and mistaken > >> theorizing, plausibly due to cognitive biases on the part of the > >> researchers. > >> > >> A second argument that is sometimes presented for the continued use > >> of > >> the traditional non-quantitative method is that it would be too > >> inefficient to evaluate every syntactic / semantic hypothesis or > >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >> make this argument explicitly in their response to Gibson & Fedorenko > >> (2010): ?It would cripple linguistic investigation if it were > >> required > >> that all judgments of ambiguity and grammaticality be subject to > >> statistically rigorous experiments on naive subjects, especially when > >> investigating languages whose speakers are hard to access? (Culicover > >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >> earlier in the discussion here.) Whereas we agree that in > >> circumstances where gathering data is difficult, some evidence is > >> better than no evidence, we do not agree that research would be > >> slowed > >> with respect to languages where experimental participants are easy to > >> access, such as English. In contrast, we think that the opposite is > >> true: the field?s progress is probably slowed by not doing > >> quantitative research. > >> Suppose that a typical syntax / semantics paper that lacks > >> quantitative evidence includes judgments for 50 or more sentences / > >> meaning pairs, corresponding to 50 or more empirical claims. Even if > >> most of the judgments from such a paper are correct or are on the > >> right track, the problem is in knowing which judgments are correct. > >> For example, suppose that 90% of the judgments from an arbitrary > >> paper > >> are correct (which is probably a high estimate). (Colin Phillips and > >> some of his former students / postdocs have commented to us that, in > >> their experience, quantitative acceptability judgment studies almost > >> always validate the claim(s) in the literature. This is not our > >> experience, however. Most experiments that we have run which attempt > >> to test some syntactic / semantic hypothesis in the literature end up > >> providing us with a pattern of data that had not been known before > >> the > >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >> that in a paper with 50 empirical claims 45/50 are correct. But which > >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >> two million different theories. By quantitatively evaluating the > >> empirical claims, we reduce the uncertainty a great deal. To make > >> progress, it is better to have theoretical claims supported by solid > >> quantitative evidence, so that even if the interpretation of the data > >> changes over time as new evidence becomes available ? as is often the > >> case in any field of science ? the empirical pattern can be used as a > >> basis for further theorizing. > >> > >> Furthermore, it is no longer expensive to run behavioral experiments, > >> at least in English and other widely spoken languages. There now > >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >> can be used for collecting behavioral data over the internet quickly > >> and inexpensively. The cost of using an interface like this is > >> minimal, and the time that it takes for the results to be returned is > >> short. For example, currently on Mechanical Turk, a survey of > >> approximately 50 items will be answered by 50 or more participants > >> within a couple of hours, at a cost of approximately $1 per > >> participant. Thus a survey can be completed within a day, at a cost > >> of > >> less than $50. (The hard work of designing the experiment, and > >> constructing controlled materials remains of course.) > >> > >> Sorry to be so verbose. But I think that these methodological points > >> are very important. > >> > >> Best wishes, > >> > >> Ted Gibson > >> > >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >> methods in syntax and semantics research. Language and Cognitive > >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >> & Fedorenko InPress LCP.pdf > >> > >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >> linguistics research. Trends in Cognitive Science, 14, 233-234. > >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >> Fedorenko > >> 2010 TICS.pdf > >> > >> > >> > >> > >>> Dick, > >>> > >>> You raise an important issue here about methodology. I believe that > >>> intuitions are a fine way to generate hypotheses and even to test > >>> them - to a degree. But while it might not have been feasible for > >>> Huddleston, Pullum, and the other contributors to the Cambridge > >>> Grammar to conduct experiments on every point of the grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, corpora, and standard psycholinguistic experimentation > >>> (indeed, Standard Social Science Methodology) is vital for taking > >>> the > >>> field forward and for providing the best support for different > >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>> paper on this, showing serious shortcomings with intuitions as the > >>> sole source of evidence, in their paper: "The need for quantitative > >>> methods in syntax and semantics research". > >>> > >>> Carson Schutze and Wayne Cowart, among others, have also written > >>> convincingly on this. > >>> > >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>> Science), and researchers from Brazil are beginning a third round of > >>> experimental work among the Pirahas, since my own work on the syntax > >>> was, like almost every other field researcher's, based on native > >>> speaker intuitions and corpora. > >>> > >>> The discussion of methodologies reminds me of the initial reactions > >>> to Greenberg's work on classifying the languages of the Americas. > >>> His > >>> methods were strongly (and justifiably) criticized. However, I > >>> always > >>> thought that his methods were a great way of generating hypotheses, > >>> so long as they were ultimately put to the test of standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions. > >>> > >>> -- Dan > >> > >> > >> > >>>> We linguists can add a further layer of explanation to the > >>>> judgements, but some judgements do seem to be more reliable than > >>>> others. And if we have to wait for psycholinguistic evidence for > >>>> every detailed analysis we make, our whole discipline will > >>>> immediately grind to a halt. Like it or not, native speaker > >>>> judgements are what put us linguists ahead of the rest in handling > >>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>> Language (or the OED) without using native speaker judgements. > >>>> > >>>> Best wishes, Dick Hudson > >> > >> > >> > > > > ------------------------------ > > Message: 10 > Date: Fri, 10 Sep 2010 14:51:45 -0400 > From: dryer at buffalo.edu > Subject: Re: [FUNKNET] analysis: unhappiness > To: Lise Menn , Funknet > > Cc: Richard Hudson > Message-ID: <2147483647.1284130304 at cast-dryerm2.caset.buffalo.edu> > Content-Type: text/plain; charset=us-ascii; format=flowed > > > The following sentence of Lise's > > "Since we don't in fact have such an agreed-on terminology, we have to be > quite careful in making clear what we are referring to when we talk about > 'the correct analysis' of a form like 'unhappiness'" > > suggests that there are two senses in which an analysis can be "the correct > analysis", one in terms of what is in people's heads, the other in terms of > what is "out there". > > There are a vast number of patterns "out there". The only distinction, on > my view, amongst this vast number of patterns, is between those that > reflect something inside people's heads and those that don't. But if that > is the case, then there is no coherent sense in which one can talk of "the > correct analysis" of what is "out there", except in terms of what is in > people's heads, and thus no second sense of "the correct analysis". The > patterns that don't correspond to things in people's heads fall into (at > least) two categories. There are those that are akin to constellations of > stars and, as with constellations, there is no reality to these patterns, > except in the minds of linguists. And there are those patterns which are > the fossil remains of what was in the heads of speakers of an earlier stage > of the language but which no longer are. These latter patterns are real, > and they are relevant to exlaining why the language is now the way it is, > but they are not relevant, I think many would agree, as to what is the > "correct analysis" of the language today. > > For this reason, I claim that the only sense in which an analysis can be > "the correct analysis" is in terms of what is inside of people's heads. > > Again, I recommend the work of Bruce Derwing for lengthy discussion of > these issues. > > Matthew > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > wrote: > > > I wish we had better terminology for keeping track of whether, at a > > given time, we are talking about the patterns that are 'out there' in > > the language and might possibly be apprehended (subconsciously) by a > > speaker, and when we are talking about the patterns that a particular > > speaker actually does apprehend, as indicated by experiments, from > > simple 'wug tests' up to brain wave and eye-gaze studies. And for > > distinguishing among the degrees of pattern apprehension that a person > > may have, from vague preferences detectable in reaction times or other > > behavior all the way up through clear metalinguistic insights. Dick > > Hudson's note reminding us of the Gleitman and Gleitman study is right > > on target. > > > > Since we don't in fact have such an agreed-on terminology, we have > to > > be quite careful in making clear what we are referring to when we talk > > about 'the correct analysis' of a form like 'unhappiness'. We know, but > > tend to forget - and tend to forget to tell our students! - that it's an > > empirical question as to whether the formal simplicity and coherence of > > description of forms 'out there' (e.g. lovely abstract morphophonemics) > > is any kind of approximation to the way knowledge of the same forms is > > organized in a particular person's head. If we remember that a very > > large proportion of what we know about our language is 'out there' when > > we are infants and has to be internalized through experience with the > > language (even if you believe in innate 'core language'), the variation > > in internal knowledge from one person to another is more understandable. > > > > We especially need to consider (and try to test) the possibility > that > > since > > the brain can make multiple cross-connections, multiple patterns are > > involved > > simultaneously in morphological and syntactic analyses. I suggest that > > that's the case with 'unhappiness' - and the linguistic analyses that I > > know about are not good at handling that kind of idea. > > > > Lise > > > > On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > > > >> > >> Two comments. > >> > >> First (elaborating perhaps on Dick Hudson's comment), I think there > >> is an > >> important distinction between low-level linguistic intuitions (like > >> whether a > >> word or sentence is well-formed or what it means) and higher-level > >> intuitions > >> (like what the structure of a word or sentence is). One can take > >> the position > >> that we need to account for the former (while recognizing that they > >> are not > >> always reliable) but not the latter. > >> > >> Second, the tension here is not only between evidence from speaker > >> intuitions > >> versus evidence from psycholinguistic experiments. There is also a > >> tension > >> between deciding on the correct analysis on the basis of a priori > >> simplicity > >> arguments versus deciding on the correct analysis on the basis of > >> psycholinguistic evidence (see Derwing 1973). The bracketing > >> paradox that Dan > >> referred to that arises with the word (semantics argues > >> for > >> [[un+happi] + er], morphology and phonology argues for [un + [happi > >> +er]] (the > >> comparative suffix can only be attached to adjectives containing one > >> or two > >> syllables) is only a paradox if one assumes that speakers adopt the > >> simplest > >> analysis. For example, if speakers adopt a more complex rule for > >> either of these > >> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >> trisyllabic > >> words beginning with un-), then the bracketing paradox disappears. > >> > >> Matthew > >> > >> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>> conscious judgements are only one kind of evidence that we need to > >>> take > >>> into account. I admire Carson Schutze's work (which I reviewed in > >>> fact), > >>> and of course I've been aware of complaints about judgements by > >>> people > >>> like Labov for decades. > >>> > >>> But you're missing my main point, which is that all judgements aren't > >>> equally reliable. If you want to know how /unhappiness/ is > >>> structured, > >>> ask a linguist, not a five-year old. And one of the by-products of > >>> education may be increased sensitivity to syntax - which is one of > >>> the > >>> many reasons why linguists need to pay more attention to education. > >>> > >>> Best wishes, Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>> Dick, > >>>> > >>>> You raise an important issue here about > >>> methodology. I believe that intuitions are a fine way to generate > >>> hypotheses and even to test them - to a degree. But while it might > >>> not have > >>> been feasible for Huddleston, Pullum, and the other contributors to > >>> the > >>> Cambridge Grammar to conduct experiments on every point of the > >>> grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, > >>> corpora, and standard psycholinguistic experimentation (indeed, > >>> Standard > >>> Social Science Methodology) is vital for taking the field forward > >>> and for > >>> providing the best support for different analyses. Ted Gibson and Ev > >>> Fedorenko have written a very useful new paper on this, showing > >>> serious > >>> shortcomings with intuitions as the sole source of evidence, in their > >>> paper: "The need for quantitative methods in syntax and semantics > >>> research".> > >>>> Carson Schutze and Wayne Cowart, among others, > >>> have also written convincingly on this.> > >>>> It is one reason that a team from Stanford, MIT > >>> (Brain and Cognitive Science), and researchers from Brazil are > >>> beginning a > >>> third round of experimental work among the Pirahas, since my own > >>> work on > >>> the syntax was, like almost every other field researcher's, based > >>> on native > >>> speaker intuitions and corpora.> > >>>> The discussion of methodologies reminds me of > >>> the initial reactions to Greenberg's work on classifying the > >>> languages of > >>> the Americas. His methods were strongly (and justifiably) criticized. > >>> However, I always thought that his methods were a great way of > >>> generating > >>> hypotheses, so long as they were ultimately put to the test of > >>> standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions.> > >>>> -- Dan > >>>> > >>>>> We linguists can add a further layer of > >>> explanation to the judgements, but some judgements do seem to be more > >>> reliable than others. And if we have to wait for psycholinguistic > >>> evidence > >>> for every detailed analysis we make, our whole discipline will > >>> immediately > >>> grind to a halt. Like it or not, native speaker judgements are what > >>> put us > >>> linguists ahead of the rest in handling fine detail. Imagine > >>> writing the > >>> Cambridge Grammar of the English Language (or the OED) without > >>> using native > >>> speaker judgements.>> > >>>>> Best wishes, Dick Hudson > >>>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> > >> > > > > Lise Menn Home Office: 303-444-4274 > > 1625 Mariposa Ave Fax: 303-413-0017 > > Boulder CO 80302 > > > > Professor Emerita of Linguistics > > Fellow, Institute of Cognitive Science > > University of Colorado > > > > Secretary, AAAS Section Z [Linguistics] > > > > Campus Mail Address: > > UCB 594, Institute of Cognitive Science > > > > Campus Physical Address: > > CINC 234 > > 1777 Exposition Ave, Boulder > > > > > > > > > > > > > > > ------------------------------ > > Message: 11 > Date: Fri, 10 Sep 2010 12:09:07 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > Matthew, > > Thanks for stating that, because I was almost beginning to imagine that > there was no essential disagreement, and that all of us agree that there > is more -- and less -- to language than what is found in people's heads. > > Your position is the one I am familiar with from the functionalist point > of view, and I was beginning to feel that it was underrepresented on > Funknet. > > Those of us who disagree with your stated position -- but are very > familiar with it -- are interested not just in psycholinguistics and how > people process language -- but also in the communicative function of > language as a system whereby information is transferred. Just as you and I > may not be aware of the way our emails are encoded and then decoded by the > computers that help us send emails back and forth, speakers may be > compeltely unaware of what language does in order to transmit information. > > After speakers have finished sending forth their linguistic output, it > matters not at all how they arrived at this output. Language processing > is separate from language in the same way that data processing is separate > from data. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > > > The following sentence of Lise's > > > > "Since we don't in fact have such an agreed-on terminology, we have to > be > > quite careful in making clear what we are referring to when we talk > about > > 'the correct analysis' of a form like 'unhappiness'" > > > > suggests that there are two senses in which an analysis can be "the > correct > > analysis", one in terms of what is in people's heads, the other in terms > of > > what is "out there". > > > > There are a vast number of patterns "out there". The only distinction, > on my > > view, amongst this vast number of patterns, is between those that reflect > > something inside people's heads and those that don't. But if that is the > > case, then there is no coherent sense in which one can talk of "the > correct > > analysis" of what is "out there", except in terms of what is in people's > > heads, and thus no second sense of "the correct analysis". The patterns > that > > don't correspond to things in people's heads fall into (at least) two > > categories. There are those that are akin to constellations of stars > and, as > > with constellations, there is no reality to these patterns, except in the > > minds of linguists. And there are those patterns which are the fossil > > remains of what was in the heads of speakers of an earlier stage of the > > language but which no longer are. These latter patterns are real, and > they > > are relevant to exlaining why the language is now the way it is, but they > are > > not relevant, I think many would agree, as to what is the "correct > analysis" > > of the language today. > > > > For this reason, I claim that the only sense in which an analysis can be > "the > > correct analysis" is in terms of what is inside of people's heads. > > > > Again, I recommend the work of Bruce Derwing for lengthy discussion of > these > > issues. > > > > Matthew > > > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > > wrote: > > > >> I wish we had better terminology for keeping track of whether, at a > >> given time, we are talking about the patterns that are 'out there' in > >> the language and might possibly be apprehended (subconsciously) by a > >> speaker, and when we are talking about the patterns that a particular > >> speaker actually does apprehend, as indicated by experiments, from > >> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >> distinguishing among the degrees of pattern apprehension that a person > >> may have, from vague preferences detectable in reaction times or other > >> behavior all the way up through clear metalinguistic insights. Dick > >> Hudson's note reminding us of the Gleitman and Gleitman study is right > >> on target. > >> > >> Since we don't in fact have such an agreed-on terminology, we have > to > >> be quite careful in making clear what we are referring to when we talk > >> about 'the correct analysis' of a form like 'unhappiness'. We know, > but > >> tend to forget - and tend to forget to tell our students! - that it's > an > >> empirical question as to whether the formal simplicity and coherence of > >> description of forms 'out there' (e.g. lovely abstract morphophonemics) > >> is any kind of approximation to the way knowledge of the same forms is > >> organized in a particular person's head. If we remember that a very > >> large proportion of what we know about our language is 'out there' when > >> we are infants and has to be internalized through experience with the > >> language (even if you believe in innate 'core language'), the variation > >> in internal knowledge from one person to another is more > understandable. > >> We especially need to consider (and try to test) the > >> possibility that > >> since > >> the brain can make multiple cross-connections, multiple patterns are > >> involved > >> simultaneously in morphological and syntactic analyses. I suggest that > >> that's the case with 'unhappiness' - and the linguistic analyses that I > >> know about are not good at handling that kind of idea. > >> > >> Lise > >> > >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> > >>> > >>> Two comments. > >>> > >>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>> is an > >>> important distinction between low-level linguistic intuitions (like > >>> whether a > >>> word or sentence is well-formed or what it means) and higher-level > >>> intuitions > >>> (like what the structure of a word or sentence is). One can take > >>> the position > >>> that we need to account for the former (while recognizing that they > >>> are not > >>> always reliable) but not the latter. > >>> > >>> Second, the tension here is not only between evidence from speaker > >>> intuitions > >>> versus evidence from psycholinguistic experiments. There is also a > >>> tension > >>> between deciding on the correct analysis on the basis of a priori > >>> simplicity > >>> arguments versus deciding on the correct analysis on the basis of > >>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>> paradox that Dan > >>> referred to that arises with the word (semantics argues > >>> for > >>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>> +er]] (the > >>> comparative suffix can only be attached to adjectives containing one > >>> or two > >>> syllables) is only a paradox if one assumes that speakers adopt the > >>> simplest > >>> analysis. For example, if speakers adopt a more complex rule for > >>> either of these > >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>> trisyllabic > >>> words beginning with un-), then the bracketing paradox disappears. > >>> > >>> Matthew > >>> > >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>>> conscious judgements are only one kind of evidence that we need to > >>>> take > >>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>> fact), > >>>> and of course I've been aware of complaints about judgements by > >>>> people > >>>> like Labov for decades. > >>>> > >>>> But you're missing my main point, which is that all judgements aren't > >>>> equally reliable. If you want to know how /unhappiness/ is > >>>> structured, > >>>> ask a linguist, not a five-year old. And one of the by-products of > >>>> education may be increased sensitivity to syntax - which is one of > >>>> the > >>>> many reasons why linguists need to pay more attention to education. > >>>> > >>>> Best wishes, Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about > >>>> methodology. I believe that intuitions are a fine way to generate > >>>> hypotheses and even to test them - to a degree. But while it might > >>>> not have > >>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>> the > >>>> Cambridge Grammar to conduct experiments on every point of the > >>>> grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, > >>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>> Standard > >>>> Social Science Methodology) is vital for taking the field forward > >>>> and for > >>>> providing the best support for different analyses. Ted Gibson and Ev > >>>> Fedorenko have written a very useful new paper on this, showing > >>>> serious > >>>> shortcomings with intuitions as the sole source of evidence, in their > >>>> paper: "The need for quantitative methods in syntax and semantics > >>>> research".> > >>>>> Carson Schutze and Wayne Cowart, among others, > >>>> have also written convincingly on this.> > >>>>> It is one reason that a team from Stanford, MIT > >>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>> beginning a > >>>> third round of experimental work among the Pirahas, since my own > >>>> work on > >>>> the syntax was, like almost every other field researcher's, based > >>>> on native > >>>> speaker intuitions and corpora.> > >>>>> The discussion of methodologies reminds me of > >>>> the initial reactions to Greenberg's work on classifying the > >>>> languages of > >>>> the Americas. His methods were strongly (and justifiably) criticized. > >>>> However, I always thought that his methods were a great way of > >>>> generating > >>>> hypotheses, so long as they were ultimately put to the test of > >>>> standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions.> > >>>>> -- Dan > >>>>> > >>>>>> We linguists can add a further layer of > >>>> explanation to the judgements, but some judgements do seem to be more > >>>> reliable than others. And if we have to wait for psycholinguistic > >>>> evidence > >>>> for every detailed analysis we make, our whole discipline will > >>>> immediately > >>>> grind to a halt. Like it or not, native speaker judgements are what > >>>> put us > >>>> linguists ahead of the rest in handling fine detail. Imagine > >>>> writing the > >>>> Cambridge Grammar of the English Language (or the OED) without > >>>> using native > >>>> speaker judgements.>> > >>>>>> Best wishes, Dick Hudson > >>>>>> > >>>>> > >>>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>> > >> > >> Lise Menn Home Office: 303-444-4274 > >> 1625 Mariposa Ave Fax: 303-413-0017 > >> Boulder CO 80302 > >> > >> Professor Emerita of Linguistics > >> Fellow, Institute of Cognitive Science > >> University of Colorado > >> > >> Secretary, AAAS Section Z [Linguistics] > >> > >> Campus Mail Address: > >> UCB 594, Institute of Cognitive Science > >> > >> Campus Physical Address: > >> CINC 234 > >> 1777 Exposition Ave, Boulder > >> > >> > >> > >> > > > > > > > > > > > > > ------------------------------ > > Message: 12 > Date: Fri, 10 Sep 2010 23:26:18 +0300 > From: Ron Kuzar > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: FUNKNET > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > The Modern Hebrew data supplied by Eitan are incomplete. > Hebrew distinguishes between locution (say, hear, think, etc.) and > situation (action, event, state, etc.). > What Eitan describes is only true with regard to nouns (and clauses) > expressing locution. 'Announcement' is indeed such a noun. > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > etc., whose denotatum is a situation, cannot be followed by ki, but only > by Se-, e.g.: > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > annoys me the-situation that-all went > 'I am upset about the situation that all have gone' > > On the other hand, the relative Se- may be replaced by the more > elegant and classical aSer, while the Se- of situation clauses may not. > Sorry about the invented example. I am overseas now. > All this has been described (with corpus data) in: > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut Ivrit > [Hebrew > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > The article is somewhat outdated and contains some inaccuracies I would > formulate differently today, but the basic distinction is valid in my > opinion. > Best, > Ron Kuzar > --------------- > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > Hi all, > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > comparison can help what seems to be a purely theoretical question based > on > > a single language. The problem here is that English uses the same element > > to > > mark regular relatives and these "appositional" relatives. But if at > least > > one language encodes them by different means, then there's at least a > good > > case for seeing them as distinct functions. It's basically the same > > principle that's used to decide whether to put a meaning on a semantic > map. > > So here are two languages that I know that encode them differently. > > > > In Modern Hebrew, these clauses can be encoded as a dedicated complement > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > ha-hoda'a Se-kibalnu > > the-announcment rel-we_got > > "The announcement that we got." > > > > ha-hoda'a ki hitbatel ha-mifgaS > > the-message CMP was_cancelled the-meeting > > "The announcement that the meeting was cancelled." > > > > In Coptic, these clauses are marked by ce-, which marks complement > clauses, > > *inter alia*, but not relative clauses: > > > > ph-mewi ce- (complement clause) > > 'the-thought that (we are angry)' > > > > ph-mewi ete- (relative clause) > > 'the thought that (we used to think)' > > > > This seems to be a pretty clear indication that these are complement > > clauses > > rather than relatives. Even if one doesn't like the notion of nouns > taking > > complement clauses (and why not? nominalizations in some languages can > take > > accusative modifiers as well as genitives), it still probably isn't > > incidental that the nominalizations are of verbs that take complement > > clauses when finite. > > > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > > looking at. > > > > Best, > > Eitan > > > > > > On 10 September 2010 19:21, Arie Verhagen > > wrote: > > > > > And as another addition: the clauses that can only be introduced by > > *that* > > > (with no role to > > > play in the subordinate clause) may be seen as (subtypes of) complement > > > clauses, > > > expressing a proposition with the noun functioning as Complement Taking > > > Predicate (CTP), > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > (following analyses by > > > Thompson, Diessel, Langacker, myself, and others), i.e. not relatives. > > Cf. > > > constructions like > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > claim > > > that X", "I put forward > > > the claim that X", in which the relationship between the verb or noun > and > > > the that-clause is > > > comparable to the one in "The claim that X". > > > > > > --Arie Verhagen > > > > > > ---------------- > > > Message from Rong Chen > > > 10 Sep 2010, 23:42 > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > To add to Joanne's comments: > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > (AC) from a relative clause (RC). > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > pronouns. > > > > > > > > 2) The AC and the noun it modifies display an equative > > relationship--one > > > can say X > > > > (denoted by the noun) is Y (presented by the appositive)--while an RC > > > often doesn't > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but a > > > relative pronoun > > > > is always part of the clause in an RC. > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > -- > > Eitan Grossman > > Martin Buber Society of Fellows > > Hebrew University of Jerusalem > > > > > > -- > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > ------------------------------ > > Message: 13 > Date: Fri, 10 Sep 2010 18:23:54 -0400 > From: Brian MacWhinney > Subject: Re: [FUNKNET] analysis: unhappiness > To: Funknet > Message-ID: > Content-Type: text/plain; charset=windows-1252 > > Dick and Ted, > I agree with parts of what each of you are saying. Which means that I > also disagree with other parts. In practice,, Gibson and Fedorenko, in > press, (which I downloaded and scanned) deals with no more than two or > three constructions. They mention the fact that people don't have problems > with sentences such as "Susan muttered him the news" despite claims that > verbs such as "mutter" cannot take the double object construction. They > also note that the claims from Jackendoff and Culicover about the > differences between the two sentences below are not supported by results > from the Mechanical Turk: > 1. Peter was trying to remember who carried what. > 2. Peter was trying to remember who carried what when. > These are interesting facts. If these sentences are supposed to be > different and people judge them to be similarly grammatical, then theories > based on the supposed differences should be reexamined. There are big > chunks of syntactic theory resting on shaky judgments about complex > sentences of this type. Getting some of this straight would be a big win, I > would say, particularly if linguists would pay attention to the results. > But I understand Dick's worry about how far Gibson and Fedorenko are > trying to push this. Neither their email nor their paper sets clear limits > on what we should be testing and we certainly don't want to waste time > checking out go-goed-went. So, Gibson and Fedorenko owe us those > clarifications. > But, Dick, you then move on to questioning data on bid-bidded. Here we > have a case of true variation in the population. I would love to know its > distribution. As a "fan of quantitative sociolinguistics" shouldn't you > too? > My take on this is that constructions are not created equal. The three > types mentioned here are probably just a start on an inventory of > evidentiary types. We need to correctly pair up appropriate methods with > each of the types. And we to make sure that people pay attention to the > results, once they are in > > --Brian MacWhinney > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > Dear Ted, > > Thanks for the very interesting comment, but are you REALLY saying that I > shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > > > > Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > > > > Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quantit > ative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > > > > It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > > > > Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 14:03, Ted Gibson wrote: > >> Dear Dan, Dick: > >> > >> I would like to clarify some points that Dan Everett makes, in > >> response to Dick Hudson. > >> > >> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >> Fedorenko, 2010, in press, see references and links below) on what we > >> think are weak methodological standards in syntax and semantics > >> research over the past many years. The issue that we address is the > >> prevalent method in syntax and semantics research, which involves > >> obtaining a judgment of the acceptability of a sentence / meaning > >> pair, typically by just the author of the paper, sometimes with > >> feedback from colleagues. As we address in our papers, this > >> methodology does not allow proper testing of scientific hypotheses > >> because of (a) the small number of experimental participants > >> (typically one); (b) the small number of experimental stimuli > >> (typically one); (c) cognitive biases on the part of the researcher > >> and participants; and (d) the effect of the preceding context (e.g., > >> other constructions the researcher may have been recently > >> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >> several others cited in Gibson & Fedorenko, in press; for similar > >> points, but with not as strong a conclusion as ours). > >> > >> Three issues need to be separated here: (1) the use of intuitive > >> judgments as a dependent measure in a language experiment; (2) > >> potential cognitive biases on the part of experimental subjects and > >> experimenters in language experiments; and (3) the need for obtaining > >> quantitative evidence, whatever the dependent measure might be. The > >> paper that Ev and I wrote addresses the last two issues, but does not > >> go into depth on the first issue (the use of intuitions as a dependent > >> measure in language experiments). Regarding this issue, we don't think > >> that there is anything wrong with gathering intuitive judgments as a > >> dependent measure, as long as the task is clear to the experimental > >> participants. > >> > >> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >> arguments that have been given in support of continuing to use the > >> traditional non-quantitative method in syntax / semantics research. > >> One recent defense of the traditional method comes from Phillips > >> (2008), who argues that no harm has come from the non-quantitative > >> approach in syntax research thus far. Phillips argues that there are > >> no cases in the literature where an incorrect intuitive judgment has > >> become the basis for a widely accepted generalization or an important > >> theoretical claim. He therefore concludes that there is no reason to > >> adopt more rigorous data collection standards. We challenge Philips? > >> conclusion by presenting three cases from the literature where a > >> faulty intuition has led to incorrect generalizations and mistaken > >> theorizing, plausibly due to cognitive biases on the part of the > >> researchers. > >> > >> A second argument that is sometimes presented for the continued use of > >> the traditional non-quantitative method is that it would be too > >> inefficient to evaluate every syntactic / semantic hypothesis or > >> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >> make this argument explicitly in their response to Gibson & Fedorenko > >> (2010): ?It would cripple linguistic investigation if it were required > >> that all judgments of ambiguity and grammaticality be subject to > >> statistically rigorous experiments on naive subjects, especially when > >> investigating languages whose speakers are hard to access? (Culicover > >> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >> earlier in the discussion here.) Whereas we agree that in > >> circumstances where gathering data is difficult, some evidence is > >> better than no evidence, we do not agree that research would be slowed > >> with respect to languages where experimental participants are easy to > >> access, such as English. In contrast, we think that the opposite is > >> true: the field?s progress is probably slowed by not doing > >> quantitative research. > >> Suppose that a typical syntax / semantics paper that lacks > >> quantitative evidence includes judgments for 50 or more sentences / > >> meaning pairs, corresponding to 50 or more empirical claims. Even if > >> most of the judgments from such a paper are correct or are on the > >> right track, the problem is in knowing which judgments are correct. > >> For example, suppose that 90% of the judgments from an arbitrary paper > >> are correct (which is probably a high estimate). (Colin Phillips and > >> some of his former students / postdocs have commented to us that, in > >> their experience, quantitative acceptability judgment studies almost > >> always validate the claim(s) in the literature. This is not our > >> experience, however. Most experiments that we have run which attempt > >> to test some syntactic / semantic hypothesis in the literature end up > >> providing us with a pattern of data that had not been known before the > >> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >> that in a paper with 50 empirical claims 45/50 are correct. But which > >> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >> two million different theories. By quantitatively evaluating the > >> empirical claims, we reduce the uncertainty a great deal. To make > >> progress, it is better to have theoretical claims supported by solid > >> quantitative evidence, so that even if the interpretation of the data > >> changes over time as new evidence becomes available ? as is often the > >> case in any field of science ? the empirical pattern can be used as a > >> basis for further theorizing. > >> > >> Furthermore, it is no longer expensive to run behavioral experiments, > >> at least in English and other widely spoken languages. There now > >> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >> can be used for collecting behavioral data over the internet quickly > >> and inexpensively. The cost of using an interface like this is > >> minimal, and the time that it takes for the results to be returned is > >> short. For example, currently on Mechanical Turk, a survey of > >> approximately 50 items will be answered by 50 or more participants > >> within a couple of hours, at a cost of approximately $1 per > >> participant. Thus a survey can be completed within a day, at a cost of > >> less than $50. (The hard work of designing the experiment, and > >> constructing controlled materials remains of course.) > >> > >> Sorry to be so verbose. But I think that these methodological points > >> are very important. > >> > >> Best wishes, > >> > >> Ted Gibson > >> > >> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >> methods in syntax and semantics research. Language and Cognitive > >> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >> & Fedorenko InPress LCP.pdf > >> > >> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >> linguistics research. Trends in Cognitive Science, 14, 233-234. > >> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > >> 2010 TICS.pdf > >> > >> > >> > >> > >>> Dick, > >>> > >>> You raise an important issue here about methodology. I believe that > >>> intuitions are a fine way to generate hypotheses and even to test > >>> them - to a degree. But while it might not have been feasible for > >>> Huddleston, Pullum, and the other contributors to the Cambridge > >>> Grammar to conduct experiments on every point of the grammar, > >>> experiments could have only made the grammar better. The use of > >>> intuitions, corpora, and standard psycholinguistic experimentation > >>> (indeed, Standard Social Science Methodology) is vital for taking the > >>> field forward and for providing the best support for different > >>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>> paper on this, showing serious shortcomings with intuitions as the > >>> sole source of evidence, in their paper: "The need for quantitative > >>> methods in syntax and semantics research". > >>> > >>> Carson Schutze and Wayne Cowart, among others, have also written > >>> convincingly on this. > >>> > >>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>> Science), and researchers from Brazil are beginning a third round of > >>> experimental work among the Pirahas, since my own work on the syntax > >>> was, like almost every other field researcher's, based on native > >>> speaker intuitions and corpora. > >>> > >>> The discussion of methodologies reminds me of the initial reactions > >>> to Greenberg's work on classifying the languages of the Americas. His > >>> methods were strongly (and justifiably) criticized. However, I always > >>> thought that his methods were a great way of generating hypotheses, > >>> so long as they were ultimately put to the test of standard > >>> historical linguistics methods. And the same seems true for use of > >>> native-speaker intuitions. > >>> > >>> -- Dan > >> > >> > >> > >>>> We linguists can add a further layer of explanation to the > >>>> judgements, but some judgements do seem to be more reliable than > >>>> others. And if we have to wait for psycholinguistic evidence for > >>>> every detailed analysis we make, our whole discipline will > >>>> immediately grind to a halt. Like it or not, native speaker > >>>> judgements are what put us linguists ahead of the rest in handling > >>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>> Language (or the OED) without using native speaker judgements. > >>>> > >>>> Best wishes, Dick Hudson > >> > >> > >> > > > > > > ------------------------------ > > Message: 14 > Date: Fri, 10 Sep 2010 18:33:49 -0400 > From: dryer at buffalo.edu > Subject: Re: [FUNKNET] analysis: unhappiness > To: "A. Katz" > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: <2147483647.1284143629 at cast-dryerm2.caset.buffalo.edu> > Content-Type: text/plain; charset=us-ascii; format=flowed > > > Aya, > > I actually agree with everything you say here. Personally, I am MORE > interested in the communicative function of language than I am in > psycholinguistics and how people process language. > > But none of that is relevant, I believe, to the very specific question of > what it means for an analysis to be correct. While one might conclude from > what I said that one ought to do psycholinguistics, that is not my > intention. Rather, my conclusion is that since I myself prefer not to do > psycholinguistics, I cannot really claim that the analyses I come up with > are "the correct" ones. And if it is really important to someone that they > identify "correct" analyses, then they ought to be doing psycholinguistics, > since there is no coherent notion of correct analysis outside of what is > inside of people's heads. > > Matthew > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > wrote: > > > Matthew, > > > > Thanks for stating that, because I was almost beginning to imagine that > > there was no essential disagreement, and that all of us agree that there > > is more -- and less -- to language than what is found in people's heads. > > > > Your position is the one I am familiar with from the functionalist point > > of view, and I was beginning to feel that it was underrepresented on > > Funknet. > > > > Those of us who disagree with your stated position -- but are very > > familiar with it -- are interested not just in psycholinguistics and how > > people process language -- but also in the communicative function of > > language as a system whereby information is transferred. Just as you and > > I may not be aware of the way our emails are encoded and then decoded by > > the computers that help us send emails back and forth, speakers may be > > compeltely unaware of what language does in order to transmit > information. > > > > After speakers have finished sending forth their linguistic output, it > > matters not at all how they arrived at this output. Language processing > > is separate from language in the same way that data processing is > > separate from data. > > > > Best, > > > > --Aya > > > > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > >> > >> The following sentence of Lise's > >> > >> "Since we don't in fact have such an agreed-on terminology, we have to > >> be quite careful in making clear what we are referring to when we talk > >> about 'the correct analysis' of a form like 'unhappiness'" > >> > >> suggests that there are two senses in which an analysis can be "the > >> correct analysis", one in terms of what is in people's heads, the other > >> in terms of what is "out there". > >> > >> There are a vast number of patterns "out there". The only distinction, > >> on my view, amongst this vast number of patterns, is between those that > >> reflect something inside people's heads and those that don't. But if > >> that is the case, then there is no coherent sense in which one can talk > >> of "the correct analysis" of what is "out there", except in terms of > >> what is in people's heads, and thus no second sense of "the correct > >> analysis". The patterns that don't correspond to things in people's > >> heads fall into (at least) two categories. There are those that are > >> akin to constellations of stars and, as with constellations, there is > >> no reality to these patterns, except in the minds of linguists. And > >> there are those patterns which are the fossil remains of what was in > >> the heads of speakers of an earlier stage of the language but which no > >> longer are. These latter patterns are real, and they are relevant to > >> exlaining why the language is now the way it is, but they are not > >> relevant, I think many would agree, as to what is the "correct analysis" > >> of the language today. > >> > >> For this reason, I claim that the only sense in which an analysis can be > >> "the correct analysis" is in terms of what is inside of people's heads. > >> > >> Again, I recommend the work of Bruce Derwing for lengthy discussion of > >> these issues. > >> > >> Matthew > >> > >> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > >> wrote: > >> > >>> I wish we had better terminology for keeping track of whether, at a > >>> given time, we are talking about the patterns that are 'out there' in > >>> the language and might possibly be apprehended (subconsciously) by a > >>> speaker, and when we are talking about the patterns that a particular > >>> speaker actually does apprehend, as indicated by experiments, from > >>> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >>> distinguishing among the degrees of pattern apprehension that a person > >>> may have, from vague preferences detectable in reaction times or other > >>> behavior all the way up through clear metalinguistic insights. Dick > >>> Hudson's note reminding us of the Gleitman and Gleitman study is right > >>> on target. > >>> > >>> Since we don't in fact have such an agreed-on terminology, we have > to > >>> be quite careful in making clear what we are referring to when we talk > >>> about 'the correct analysis' of a form like 'unhappiness'. We know, > >>> but tend to forget - and tend to forget to tell our students! - that > >>> it's an empirical question as to whether the formal simplicity and > >>> coherence of description of forms 'out there' (e.g. lovely abstract > >>> morphophonemics) is any kind of approximation to the way knowledge of > >>> the same forms is organized in a particular person's head. If we > >>> remember that a very large proportion of what we know about our > >>> language is 'out there' when we are infants and has to be internalized > >>> through experience with the language (even if you believe in innate > >>> 'core language'), the variation in internal knowledge from one person > >>> to another is more understandable. We especially need to consider (and > >>> try to test) the > >>> possibility that > >>> since > >>> the brain can make multiple cross-connections, multiple patterns are > >>> involved > >>> simultaneously in morphological and syntactic analyses. I suggest that > >>> that's the case with 'unhappiness' - and the linguistic analyses that > I > >>> know about are not good at handling that kind of idea. > >>> > >>> Lise > >>> > >>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >>> > >>>> > >>>> Two comments. > >>>> > >>>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>>> is an > >>>> important distinction between low-level linguistic intuitions (like > >>>> whether a > >>>> word or sentence is well-formed or what it means) and higher-level > >>>> intuitions > >>>> (like what the structure of a word or sentence is). One can take > >>>> the position > >>>> that we need to account for the former (while recognizing that they > >>>> are not > >>>> always reliable) but not the latter. > >>>> > >>>> Second, the tension here is not only between evidence from speaker > >>>> intuitions > >>>> versus evidence from psycholinguistic experiments. There is also a > >>>> tension > >>>> between deciding on the correct analysis on the basis of a priori > >>>> simplicity > >>>> arguments versus deciding on the correct analysis on the basis of > >>>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>>> paradox that Dan > >>>> referred to that arises with the word (semantics argues > >>>> for > >>>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>>> +er]] (the > >>>> comparative suffix can only be attached to adjectives containing one > >>>> or two > >>>> syllables) is only a paradox if one assumes that speakers adopt the > >>>> simplest > >>>> analysis. For example, if speakers adopt a more complex rule for > >>>> either of these > >>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>>> trisyllabic > >>>> words beginning with un-), then the bracketing paradox disappears. > >>>> > >>>> Matthew > >>>> > >>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree that > >>>>> conscious judgements are only one kind of evidence that we need to > >>>>> take > >>>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>>> fact), > >>>>> and of course I've been aware of complaints about judgements by > >>>>> people > >>>>> like Labov for decades. > >>>>> > >>>>> But you're missing my main point, which is that all judgements aren't > >>>>> equally reliable. If you want to know how /unhappiness/ is > >>>>> structured, > >>>>> ask a linguist, not a five-year old. And one of the by-products of > >>>>> education may be increased sensitivity to syntax - which is one of > >>>>> the > >>>>> many reasons why linguists need to pay more attention to education. > >>>>> > >>>>> Best wishes, Dick > >>>>> > >>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about > >>>>> methodology. I believe that intuitions are a fine way to generate > >>>>> hypotheses and even to test them - to a degree. But while it might > >>>>> not have > >>>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>>> the > >>>>> Cambridge Grammar to conduct experiments on every point of the > >>>>> grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, > >>>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>>> Standard > >>>>> Social Science Methodology) is vital for taking the field forward > >>>>> and for > >>>>> providing the best support for different analyses. Ted Gibson and Ev > >>>>> Fedorenko have written a very useful new paper on this, showing > >>>>> serious > >>>>> shortcomings with intuitions as the sole source of evidence, in their > >>>>> paper: "The need for quantitative methods in syntax and semantics > >>>>> research".> > >>>>>> Carson Schutze and Wayne Cowart, among others, > >>>>> have also written convincingly on this.> > >>>>>> It is one reason that a team from Stanford, MIT > >>>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>>> beginning a > >>>>> third round of experimental work among the Pirahas, since my own > >>>>> work on > >>>>> the syntax was, like almost every other field researcher's, based > >>>>> on native > >>>>> speaker intuitions and corpora.> > >>>>>> The discussion of methodologies reminds me of > >>>>> the initial reactions to Greenberg's work on classifying the > >>>>> languages of > >>>>> the Americas. His methods were strongly (and justifiably) criticized. > >>>>> However, I always thought that his methods were a great way of > >>>>> generating > >>>>> hypotheses, so long as they were ultimately put to the test of > >>>>> standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions.> > >>>>>> -- Dan > >>>>>> > >>>>>>> We linguists can add a further layer of > >>>>> explanation to the judgements, but some judgements do seem to be more > >>>>> reliable than others. And if we have to wait for psycholinguistic > >>>>> evidence > >>>>> for every detailed analysis we make, our whole discipline will > >>>>> immediately > >>>>> grind to a halt. Like it or not, native speaker judgements are what > >>>>> put us > >>>>> linguists ahead of the rest in handling fine detail. Imagine > >>>>> writing the > >>>>> Cambridge Grammar of the English Language (or the OED) without > >>>>> using native > >>>>> speaker judgements.>> > >>>>>>> Best wishes, Dick Hudson > >>>>>>> > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>> > >>> > >>> Lise Menn Home Office: 303-444-4274 > >>> 1625 Mariposa Ave Fax: 303-413-0017 > >>> Boulder CO 80302 > >>> > >>> Professor Emerita of Linguistics > >>> Fellow, Institute of Cognitive Science > >>> University of Colorado > >>> > >>> Secretary, AAAS Section Z [Linguistics] > >>> > >>> Campus Mail Address: > >>> UCB 594, Institute of Cognitive Science > >>> > >>> Campus Physical Address: > >>> CINC 234 > >>> 1777 Exposition Ave, Boulder > >>> > >>> > >>> > >>> > >> > >> > >> > >> > >> > > > > > > > > > > > ------------------------------ > > Message: 15 > Date: Sat, 11 Sep 2010 00:20:15 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet at mailman.rice.edu > Message-ID: <4C8ABD2F.8070902 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Brian, > What a helpful message! I think you're right: we need a typology of > cases, each needing a different range of methods, ranging from the > linguist's own judgements for really easy cases to more complicated > quantitative methods for more complicated ones. > > The trouble with our discipline is that for any community of N speakers, > and a language consisting of M 'items' (however you may choose to define > 'community' and 'item'), we have N*M datapoints that, in principle, all > need to be validated somehow. We might reduce the number by focusing on > one speaker, but then you can't use data from other speakers as evidence > for that speaker's language; or we might try to construct a 'typical' > speaker, but we don't know how to do that; or we might reduce the size > of the community by trying to find a 'dialect' (but dialects don't > really exist); or we might ignore most of the linguistic items and focus > on, say, the modal verbs - but then we miss their links to all the other > items. > > It's different for psycholinguists because they're only interested in > general processes, for which linguistic items are just evidence, not the > thing under investigation; but for us linguists, the fine detail is > everything because we're the people who explore the connections between > items. > > So I look forward to the day when your typology of cases will guide us > through a range of different methods to the appropriate ones for any > given item. > > Best wishes, Dick > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 23:23, Brian MacWhinney wrote: > > Dick and Ted, > > I agree with parts of what each of you are saying. Which means that > I also disagree with other parts. In practice,, Gibson and Fedorenko, in > press, (which I downloaded and scanned) deals with no more than two or > three constructions. They mention the fact that people don't have problems > with sentences such as "Susan muttered him the news" despite claims that > verbs such as "mutter" cannot take the double object construction. They > also note that the claims from Jackendoff and Culicover about the > differences between the two sentences below are not supported by results > from the Mechanical Turk: > > 1. Peter was trying to remember who carried what. > > 2. Peter was trying to remember who carried what when. > > These are interesting facts. If these sentences are supposed to be > different and people judge them to be similarly grammatical, then theories > based on the supposed differences should be reexamined. There are big > chunks of syntactic theory resting on shaky judgments about complex > sentences of this type. Getting some of this straight would be a big win, I > would say, particularly if linguists would pay attention to the results. > > But I understand Dick's worry about how far Gibson and Fedorenko > are trying to push this. Neither their email nor their paper sets clear > limits on what we should be testing and we certainly don't want to waste > time checking out go-goed-went. So, Gibson and Fedorenko owe us those > clarifications. > > But, Dick, you then move on to questioning data on bid-bidded. Here > we have a case of true variation in the population. I would love to know > its distribution. As a "fan of quantitative sociolinguistics" shouldn't you > too? > > My take on this is that constructions are not created equal. The > three types mentioned here are probably just a start on an inventory of > evidentiary types. We need to correctly pair up appropriate methods with > each of the types. And we to make sure that people pay attention to the > results, once they are in > > > > --Brian MacWhinney > > > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > >> Dear Ted, > >> Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > >> > >> Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > >> > >> Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quanti > tative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > >> > >> It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > >> > >> Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 14:03, Ted Gibson wrote: > >>> Dear Dan, Dick: > >>> > >>> I would like to clarify some points that Dan Everett makes, in > >>> response to Dick Hudson. > >>> > >>> Ev Fedorenko and I have written a couple of papers recently (Gibson& > >>> Fedorenko, 2010, in press, see references and links below) on what we > >>> think are weak methodological standards in syntax and semantics > >>> research over the past many years. The issue that we address is the > >>> prevalent method in syntax and semantics research, which involves > >>> obtaining a judgment of the acceptability of a sentence / meaning > >>> pair, typically by just the author of the paper, sometimes with > >>> feedback from colleagues. As we address in our papers, this > >>> methodology does not allow proper testing of scientific hypotheses > >>> because of (a) the small number of experimental participants > >>> (typically one); (b) the small number of experimental stimuli > >>> (typically one); (c) cognitive biases on the part of the researcher > >>> and participants; and (d) the effect of the preceding context (e.g., > >>> other constructions the researcher may have been recently > >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>> several others cited in Gibson& Fedorenko, in press; for similar > >>> points, but with not as strong a conclusion as ours). > >>> > >>> Three issues need to be separated here: (1) the use of intuitive > >>> judgments as a dependent measure in a language experiment; (2) > >>> potential cognitive biases on the part of experimental subjects and > >>> experimenters in language experiments; and (3) the need for obtaining > >>> quantitative evidence, whatever the dependent measure might be. The > >>> paper that Ev and I wrote addresses the last two issues, but does not > >>> go into depth on the first issue (the use of intuitions as a dependent > >>> measure in language experiments). Regarding this issue, we don't think > >>> that there is anything wrong with gathering intuitive judgments as a > >>> dependent measure, as long as the task is clear to the experimental > >>> participants. > >>> > >>> In the longer paper (Gibson& Fedorenko, in press) we respond to some > >>> arguments that have been given in support of continuing to use the > >>> traditional non-quantitative method in syntax / semantics research. > >>> One recent defense of the traditional method comes from Phillips > >>> (2008), who argues that no harm has come from the non-quantitative > >>> approach in syntax research thus far. Phillips argues that there are > >>> no cases in the literature where an incorrect intuitive judgment has > >>> become the basis for a widely accepted generalization or an important > >>> theoretical claim. He therefore concludes that there is no reason to > >>> adopt more rigorous data collection standards. We challenge Philips? > >>> conclusion by presenting three cases from the literature where a > >>> faulty intuition has led to incorrect generalizations and mistaken > >>> theorizing, plausibly due to cognitive biases on the part of the > >>> researchers. > >>> > >>> A second argument that is sometimes presented for the continued use of > >>> the traditional non-quantitative method is that it would be too > >>> inefficient to evaluate every syntactic / semantic hypothesis or > >>> phenomenon quantitatively. For example, Culicover& Jackendoff (2010) > >>> make this argument explicitly in their response to Gibson& Fedorenko > >>> (2010): ?It would cripple linguistic investigation if it were required > >>> that all judgments of ambiguity and grammaticality be subject to > >>> statistically rigorous experiments on naive subjects, especially when > >>> investigating languages whose speakers are hard to access? (Culicover > >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>> earlier in the discussion here.) Whereas we agree that in > >>> circumstances where gathering data is difficult, some evidence is > >>> better than no evidence, we do not agree that research would be slowed > >>> with respect to languages where experimental participants are easy to > >>> access, such as English. In contrast, we think that the opposite is > >>> true: the field?s progress is probably slowed by not doing > >>> quantitative research. > >>> Suppose that a typical syntax / semantics paper that lacks > >>> quantitative evidence includes judgments for 50 or more sentences / > >>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>> most of the judgments from such a paper are correct or are on the > >>> right track, the problem is in knowing which judgments are correct. > >>> For example, suppose that 90% of the judgments from an arbitrary paper > >>> are correct (which is probably a high estimate). (Colin Phillips and > >>> some of his former students / postdocs have commented to us that, in > >>> their experience, quantitative acceptability judgment studies almost > >>> always validate the claim(s) in the literature. This is not our > >>> experience, however. Most experiments that we have run which attempt > >>> to test some syntactic / semantic hypothesis in the literature end up > >>> providing us with a pattern of data that had not been known before the > >>> experiment (e.g., Breen et al., in press; Fedorenko& Gibson, in > >>> press; Patel et al., 2009; Scontras& Gibson, submitted).) This means > >>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>> two million different theories. By quantitatively evaluating the > >>> empirical claims, we reduce the uncertainty a great deal. To make > >>> progress, it is better to have theoretical claims supported by solid > >>> quantitative evidence, so that even if the interpretation of the data > >>> changes over time as new evidence becomes available ? as is often the > >>> case in any field of science ? the empirical pattern can be used as a > >>> basis for further theorizing. > >>> > >>> Furthermore, it is no longer expensive to run behavioral experiments, > >>> at least in English and other widely spoken languages. There now > >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>> can be used for collecting behavioral data over the internet quickly > >>> and inexpensively. The cost of using an interface like this is > >>> minimal, and the time that it takes for the results to be returned is > >>> short. For example, currently on Mechanical Turk, a survey of > >>> approximately 50 items will be answered by 50 or more participants > >>> within a couple of hours, at a cost of approximately $1 per > >>> participant. Thus a survey can be completed within a day, at a cost of > >>> less than $50. (The hard work of designing the experiment, and > >>> constructing controlled materials remains of course.) > >>> > >>> Sorry to be so verbose. But I think that these methodological points > >>> are very important. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson > >>> > >>> Gibson, E.& Fedorenko, E. (In press). The need for quantitative > >>> methods in syntax and semantics research. Language and Cognitive > >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>> & Fedorenko InPress LCP.pdf > >>> > >>> Gibson, E.& Fedorenko, E. (2010). Weak quantitative standards in > >>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson& Fedorenko > >>> 2010 TICS.pdf > >>> > >>> > >>> > >>> > >>>> Dick, > >>>> > >>>> You raise an important issue here about methodology. I believe that > >>>> intuitions are a fine way to generate hypotheses and even to test > >>>> them - to a degree. But while it might not have been feasible for > >>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>> Grammar to conduct experiments on every point of the grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>> field forward and for providing the best support for different > >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>> paper on this, showing serious shortcomings with intuitions as the > >>>> sole source of evidence, in their paper: "The need for quantitative > >>>> methods in syntax and semantics research". > >>>> > >>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>> convincingly on this. > >>>> > >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>> Science), and researchers from Brazil are beginning a third round of > >>>> experimental work among the Pirahas, since my own work on the syntax > >>>> was, like almost every other field researcher's, based on native > >>>> speaker intuitions and corpora. > >>>> > >>>> The discussion of methodologies reminds me of the initial reactions > >>>> to Greenberg's work on classifying the languages of the Americas. His > >>>> methods were strongly (and justifiably) criticized. However, I always > >>>> thought that his methods were a great way of generating hypotheses, > >>>> so long as they were ultimately put to the test of standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions. > >>>> > >>>> -- Dan > >>> > >>> > >>>>> We linguists can add a further layer of explanation to the > >>>>> judgements, but some judgements do seem to be more reliable than > >>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>> every detailed analysis we make, our whole discipline will > >>>>> immediately grind to a halt. Like it or not, native speaker > >>>>> judgements are what put us linguists ahead of the rest in handling > >>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>> Language (or the OED) without using native speaker judgements. > >>>>> > >>>>> Best wishes, Dick Hudson > >>> > >>> > > > > > > > > ------------------------------ > > Message: 16 > Date: Sat, 11 Sep 2010 00:40:06 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: Ted Gibson > Cc: Richard Hudson , Evelina Fedorenko > , funknet > Message-ID: <4C8AC1D6.3010203 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=windows-1252; format=flowed > > Dear Ted and Ev, > Yes, I understand your view, but I think it's a psycholinguist's view. > Your goal is to find general processes and principles that apply > uniformly across individuals, so you have to use methods to check for > generality. And (as you know) I admire the way you pursue that goal. But > my goal, as a linguist, is different. I want to explore the structure of > a language so that I can understand how all the bits fit together. Like > you, I'm aiming to model cognition, but my focus is on items and > structures, and I start from the assumption that these can and do vary > across speakers. > > However, having said all that I do agree with you that linguists should > all get used to collecting and using quantitative data; and with the > help of Brian MacWhinney's typology we'd know what methods to use when. > And I do agree with your points about bid/bidded: in cases like that, > quantitative data would be at least a very good starting point for a > proper investigation. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 10/09/2010 19:30, Ted Gibson wrote: > > Dear Dick: > > > > Perhaps we are talking at cross purposes. I don't understand what is > > confusing about what Ev Fedorenko and I are claiming. All we are > > saying is that if you have some testable claim involving a general > > hypothesis about a language, then you need to get quantitative data > > from unbiased sources to evaluate that claim. If you are interested in > > English past tense morphology, then depending on the claims that you > > might want to investigate, there are lots of ways to get relevant > > quantitative evidence. Corpus data will probably be useful. For very > > low frequency words, you can run experiments to test behavior with > > respect to such words. > > > > Your example of the past tense of "bid" is a fine such example. You > > can run an experiment like the one you suggested to find out what > > people think the past tense is. If you then found that 20/50 people > > responded "bidded" and 30/50 respond "bid", that is a lot of useful > > information. As you suggest in your discussion, this result wouldn't > > answer the question of how past tense is stored in each individual. > > This result would be ambiguous among several possible explanations. > > One possibility is that the probability distribution that is being > > discovered reflects different dialects, such that 2/5 of the > > population has one past tense, and 3/5 has another. Another > > possibility is that each person has a similar probability distribution > > in their heads, such that 2/5 of the time I respond one way, and 3/5 > > of the time I respond another. Further experiments would be necessary > > to answer between these and other possible theories (e.g., with > > repeated trials from the same person, carefully planned so that the > > participants don't notice that they are being asked multiple times). > > Without the quantitative evidence in the first place, there is no way > > to answer these kinds of questions. > > > > Regarding the past tense of "go", this would be useful as a baseline > > in an experiment involving the less frequent ones. So, yes, it would > > useful to gather quantitative evidence in such a case also, as > > baselines with respect to the more interesting cases for theories. > > > > The bottom line: if you have a generalization about a language that > > you wish to evaluate (such that you hypothesize that it is true across > > the speakers of the language), then you need quantitative evidence > > from multiple individuals, using an unbiased data collection method, > > to evaluate such a claim. The point about Mechanical Turk is that it > > is really *easy* to do this now, at least for languages like English. > > > > Best wishes, > > > > Ted Gibson & Ev Fedorenko > > > > On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > > > >> Dear Ted, > >> Thanks for the very interesting comment, but are you REALLY saying > >> that I shouldn't claim, for example, that the past tense of GO is > >> "went" without first cross-checking with 50 native speakers? > >> > >> Isn't there a danger of missing the point that we all, as native > >> speakers, spend our whole lives scanning other people's linguistic > >> behaviour (language 'out there', E-language) and trying to explain it > >> to ourselves in terms of a language system (language 'in here', > >> I-language)? So every judgement we make is based on thousands or > >> millions of observed exemplars, and reflects a unique experience of > >> E-language filtered through a unique I-language. > >> > >> Given that view of language development, I don't see how quantitative > >> data will help. Let's take a real uncertainty, such as the past tense > >> of BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > >> judgement: I don't know. Ok, you get 50 people to oblige on > >> Mechanical Turk, and 20 of them give "bidded" and 30 "bid". So what? > >> Does that mean that the correct answer is "bidded"? Surely not. How > >> is it better than my judgement? I agree you could record my speech > >> and find how often I use each alternative; but the reason I don't > >> know is precisely because it's a rare word, so in a sense > >> quantitative data are irrelevant even there. What would solve the > >> problem of subjectivity, of course, would be a machine for probing > >> the bit of my mind (or even brain) that holds BID and its details; > >> but I suspect that even that wouldn't move us much further forward > >> than my original "don't know". (Incidentally I write as a fan of > >> quantitative sociolinguistics, so I do accept that quantitative data > >> are relevant to linguistic analysis in some areas, where the > >> I-language phenomenon is frequent enough to produce usable data.) > >> > >> It seems to me that this discussion raises the really fundamental > >> question of what kind of thing we think language is: social or > >> individual. The problem isn't unique to linguistics of course; it's > >> the same throughout the social sciences. But what's special about > >> linguistics is that we deal in very fine details of culture (e.g. > >> details of how a particular word is used or pronounced) so the > >> differences between individuals really matter. I don't see that we're > >> ever going to have anything better than judgements to go on, so what > >> we need is a way to ensure that judgements are accurate reports of > >> individual I-language. A rotten situation for a science, but I don't > >> see how it can get better. > >> > >> Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 14:03, Ted Gibson wrote: > >>> Dear Dan, Dick: > >>> > >>> I would like to clarify some points that Dan Everett makes, in > >>> response to Dick Hudson. > >>> > >>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>> Fedorenko, 2010, in press, see references and links below) on what we > >>> think are weak methodological standards in syntax and semantics > >>> research over the past many years. The issue that we address is the > >>> prevalent method in syntax and semantics research, which involves > >>> obtaining a judgment of the acceptability of a sentence / meaning > >>> pair, typically by just the author of the paper, sometimes with > >>> feedback from colleagues. As we address in our papers, this > >>> methodology does not allow proper testing of scientific hypotheses > >>> because of (a) the small number of experimental participants > >>> (typically one); (b) the small number of experimental stimuli > >>> (typically one); (c) cognitive biases on the part of the researcher > >>> and participants; and (d) the effect of the preceding context (e.g., > >>> other constructions the researcher may have been recently > >>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>> several others cited in Gibson & Fedorenko, in press; for similar > >>> points, but with not as strong a conclusion as ours). > >>> > >>> Three issues need to be separated here: (1) the use of intuitive > >>> judgments as a dependent measure in a language experiment; (2) > >>> potential cognitive biases on the part of experimental subjects and > >>> experimenters in language experiments; and (3) the need for obtaining > >>> quantitative evidence, whatever the dependent measure might be. The > >>> paper that Ev and I wrote addresses the last two issues, but does not > >>> go into depth on the first issue (the use of intuitions as a dependent > >>> measure in language experiments). Regarding this issue, we don't think > >>> that there is anything wrong with gathering intuitive judgments as a > >>> dependent measure, as long as the task is clear to the experimental > >>> participants. > >>> > >>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>> arguments that have been given in support of continuing to use the > >>> traditional non-quantitative method in syntax / semantics research. > >>> One recent defense of the traditional method comes from Phillips > >>> (2008), who argues that no harm has come from the non-quantitative > >>> approach in syntax research thus far. Phillips argues that there are > >>> no cases in the literature where an incorrect intuitive judgment has > >>> become the basis for a widely accepted generalization or an important > >>> theoretical claim. He therefore concludes that there is no reason to > >>> adopt more rigorous data collection standards. We challenge Philips? > >>> conclusion by presenting three cases from the literature where a > >>> faulty intuition has led to incorrect generalizations and mistaken > >>> theorizing, plausibly due to cognitive biases on the part of the > >>> researchers. > >>> > >>> A second argument that is sometimes presented for the continued use of > >>> the traditional non-quantitative method is that it would be too > >>> inefficient to evaluate every syntactic / semantic hypothesis or > >>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>> make this argument explicitly in their response to Gibson & Fedorenko > >>> (2010): ?It would cripple linguistic investigation if it were required > >>> that all judgments of ambiguity and grammaticality be subject to > >>> statistically rigorous experiments on naive subjects, especially when > >>> investigating languages whose speakers are hard to access? (Culicover > >>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>> earlier in the discussion here.) Whereas we agree that in > >>> circumstances where gathering data is difficult, some evidence is > >>> better than no evidence, we do not agree that research would be slowed > >>> with respect to languages where experimental participants are easy to > >>> access, such as English. In contrast, we think that the opposite is > >>> true: the field?s progress is probably slowed by not doing > >>> quantitative research. > >>> Suppose that a typical syntax / semantics paper that lacks > >>> quantitative evidence includes judgments for 50 or more sentences / > >>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>> most of the judgments from such a paper are correct or are on the > >>> right track, the problem is in knowing which judgments are correct. > >>> For example, suppose that 90% of the judgments from an arbitrary paper > >>> are correct (which is probably a high estimate). (Colin Phillips and > >>> some of his former students / postdocs have commented to us that, in > >>> their experience, quantitative acceptability judgment studies almost > >>> always validate the claim(s) in the literature. This is not our > >>> experience, however. Most experiments that we have run which attempt > >>> to test some syntactic / semantic hypothesis in the literature end up > >>> providing us with a pattern of data that had not been known before the > >>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>> two million different theories. By quantitatively evaluating the > >>> empirical claims, we reduce the uncertainty a great deal. To make > >>> progress, it is better to have theoretical claims supported by solid > >>> quantitative evidence, so that even if the interpretation of the data > >>> changes over time as new evidence becomes available ? as is often the > >>> case in any field of science ? the empirical pattern can be used as a > >>> basis for further theorizing. > >>> > >>> Furthermore, it is no longer expensive to run behavioral experiments, > >>> at least in English and other widely spoken languages. There now > >>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>> can be used for collecting behavioral data over the internet quickly > >>> and inexpensively. The cost of using an interface like this is > >>> minimal, and the time that it takes for the results to be returned is > >>> short. For example, currently on Mechanical Turk, a survey of > >>> approximately 50 items will be answered by 50 or more participants > >>> within a couple of hours, at a cost of approximately $1 per > >>> participant. Thus a survey can be completed within a day, at a cost of > >>> less than $50. (The hard work of designing the experiment, and > >>> constructing controlled materials remains of course.) > >>> > >>> Sorry to be so verbose. But I think that these methodological points > >>> are very important. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson > >>> > >>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>> methods in syntax and semantics research. Language and Cognitive > >>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>> & Fedorenko InPress LCP.pdf > >>> > >>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & Fedorenko > >>> 2010 TICS.pdf > >>> > >>> > >>> > >>> > >>>> Dick, > >>>> > >>>> You raise an important issue here about methodology. I believe that > >>>> intuitions are a fine way to generate hypotheses and even to test > >>>> them - to a degree. But while it might not have been feasible for > >>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>> Grammar to conduct experiments on every point of the grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>> field forward and for providing the best support for different > >>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>> paper on this, showing serious shortcomings with intuitions as the > >>>> sole source of evidence, in their paper: "The need for quantitative > >>>> methods in syntax and semantics research". > >>>> > >>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>> convincingly on this. > >>>> > >>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>> Science), and researchers from Brazil are beginning a third round of > >>>> experimental work among the Pirahas, since my own work on the syntax > >>>> was, like almost every other field researcher's, based on native > >>>> speaker intuitions and corpora. > >>>> > >>>> The discussion of methodologies reminds me of the initial reactions > >>>> to Greenberg's work on classifying the languages of the Americas. His > >>>> methods were strongly (and justifiably) criticized. However, I always > >>>> thought that his methods were a great way of generating hypotheses, > >>>> so long as they were ultimately put to the test of standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions. > >>>> > >>>> -- Dan > >>> > >>> > >>> > >>>>> We linguists can add a further layer of explanation to the > >>>>> judgements, but some judgements do seem to be more reliable than > >>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>> every detailed analysis we make, our whole discipline will > >>>>> immediately grind to a halt. Like it or not, native speaker > >>>>> judgements are what put us linguists ahead of the rest in handling > >>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>> Language (or the OED) without using native speaker judgements. > >>>>> > >>>>> Best wishes, Dick Hudson > >>> > >>> > >>> > > > > > > > > > ------------------------------ > > Message: 17 > Date: Fri, 10 Sep 2010 18:40:47 -0600 > From: Lise Menn > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Richard Hudson , Funknet > > Message-ID: <91F0DEA7-53BE-405B-8087-88B7F1672447 at Colorado.EDU> > Content-Type: text/plain; charset=US-ASCII; format=flowed; > delsp=yes > > Matt, I have to disagree with you on the validity of describing what's > 'out there' (what Dick Hudson says is his interest, in his > contribution of 5:40:06 PM MDT today). We DO have to account for it > in order to understand how 'the language in speakers' heads' gets into > those heads in the first place. > In more detail: Each of us is immersed from (before) birth in a > sampling of utterances (and if we are literate, eventually also > written forms of the language). In order to understand how we really > create our internal representations of our language, we have to know > (or be able to estimate) something about the data our brains get as > input. There are at least better and worse descriptions of the > patterns in those data, and certainly there are wrong ones, though in > many cases - for example in the 'unhappiness' case - there are > probably conflicting right ones, rather than any single correct one. > (OT offers some help in thinking about this.) > > To take a concrete example, in order to account for the still- > unstable changes in English pronominal case marking in compound NP > objects of prepositions from a system based on syntactic case (He gave > the cookies to Mary and me) to a system apparently based partly on > whether the pronoun is next to the governing preposition (He gave the > cookies to Mary and I/ to me and Mary), you first have to do an > analysis of usage and figure out what the pattern is. And usage is > not in our heads (although it's the result of what's in our heads), > it's 'out there'. > > Even fossils and obscure patterns contribute to the redundancy of > the > language, making it more learnable and and helping to create the > resonances used by great poets and orators. (I admit to having > oversimplified in speaking as if there were always one 'correct' > analysis of the patterns 'out there' that might be (subconsciously) > discoverable by speakers. That's not true.) And because not all > speakers are equally sensitive to language patterns - again, the > Gleitman and Gleitman book is a terrific example - it's also an > oversimplification to talk about 'what is in speaker's heads' as if > the same thing is in everyone's head. (K.P. Mohanan has also published > on this.) At the lexical level, Danielle Cyr's examples (September 9, > 2010 8:38:59 PM MDT) further remind us that what's inside each > person's head changes over time. So we must also be careful not to > idealize "what's in people's heads" as if it were a single coherent > construct that we are trying to discover. It's not - it's more like a > complex mosaic that does not fit together perfectly. > > Lise > > On Sep 10, 2010, at 12:51 PM, dryer at buffalo.edu wrote: > > > > > The following sentence of Lise's > > > > "Since we don't in fact have such an agreed-on terminology, we have > > to be quite careful in making clear what we are referring to when > > we talk about 'the correct analysis' of a form like 'unhappiness'" > > > > suggests that there are two senses in which an analysis can be "the > > correct analysis", one in terms of what is in people's heads, the > > other in terms of what is "out there". > > > > There are a vast number of patterns "out there". The only > > distinction, on my view, amongst this vast number of patterns, is > > between those that reflect something inside people's heads and those > > that don't. But if that is the case, then there is no coherent > > sense in which one can talk of "the correct analysis" of what is > > "out there", except in terms of what is in people's heads, and thus > > no second sense of "the correct analysis". The patterns that don't > > correspond to things in people's heads fall into (at least) two > > categories. There are those that are akin to constellations of > > stars and, as with constellations, there is no reality to these > > patterns, except in the minds of linguists. And there are those > > patterns which are the fossil remains of what was in the heads of > > speakers of an earlier stage of the language but which no longer > > are. These latter patterns are real, and they are relevant to > > exlaining why the language is now the way it is, but they are not > > relevant, I think many would agree, as to what is the "correct > > analysis" of the language today. > > > > For this reason, I claim that the only sense in which an analysis > > can be "the correct analysis" is in terms of what is inside of > > people's heads. > > > > Again, I recommend the work of Bruce Derwing for lengthy discussion > > of these issues. > > > > Matthew > > > > --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > > > wrote: > > > >> I wish we had better terminology for keeping track of whether, at a > >> given time, we are talking about the patterns that are 'out there' in > >> the language and might possibly be apprehended (subconsciously) by a > >> speaker, and when we are talking about the patterns that a particular > >> speaker actually does apprehend, as indicated by experiments, from > >> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >> distinguishing among the degrees of pattern apprehension that a > >> person > >> may have, from vague preferences detectable in reaction times or > >> other > >> behavior all the way up through clear metalinguistic insights. Dick > >> Hudson's note reminding us of the Gleitman and Gleitman study is > >> right > >> on target. > >> > >> Since we don't in fact have such an agreed-on terminology, we have > >> to > >> be quite careful in making clear what we are referring to when we > >> talk > >> about 'the correct analysis' of a form like 'unhappiness'. We > >> know, but > >> tend to forget - and tend to forget to tell our students! - that > >> it's an > >> empirical question as to whether the formal simplicity and > >> coherence of > >> description of forms 'out there' (e.g. lovely abstract > >> morphophonemics) > >> is any kind of approximation to the way knowledge of the same > >> forms is > >> organized in a particular person's head. If we remember that a very > >> large proportion of what we know about our language is 'out there' > >> when > >> we are infants and has to be internalized through experience with > >> the > >> language (even if you believe in innate 'core language'), the > >> variation > >> in internal knowledge from one person to another is more > >> understandable. > >> > >> We especially need to consider (and try to test) the possibility > >> that > >> since > >> the brain can make multiple cross-connections, multiple patterns are > >> involved > >> simultaneously in morphological and syntactic analyses. I suggest > >> that > >> that's the case with 'unhappiness' - and the linguistic analyses > >> that I > >> know about are not good at handling that kind of idea. > >> > >> Lise > >> > >> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >> > >>> > >>> Two comments. > >>> > >>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>> is an > >>> important distinction between low-level linguistic intuitions (like > >>> whether a > >>> word or sentence is well-formed or what it means) and higher-level > >>> intuitions > >>> (like what the structure of a word or sentence is). One can take > >>> the position > >>> that we need to account for the former (while recognizing that they > >>> are not > >>> always reliable) but not the latter. > >>> > >>> Second, the tension here is not only between evidence from speaker > >>> intuitions > >>> versus evidence from psycholinguistic experiments. There is also a > >>> tension > >>> between deciding on the correct analysis on the basis of a priori > >>> simplicity > >>> arguments versus deciding on the correct analysis on the basis of > >>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>> paradox that Dan > >>> referred to that arises with the word (semantics argues > >>> for > >>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>> +er]] (the > >>> comparative suffix can only be attached to adjectives containing one > >>> or two > >>> syllables) is only a paradox if one assumes that speakers adopt the > >>> simplest > >>> analysis. For example, if speakers adopt a more complex rule for > >>> either of these > >>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>> trisyllabic > >>> words beginning with un-), then the bracketing paradox disappears. > >>> > >>> Matthew > >>> > >>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree > >>>> that > >>>> conscious judgements are only one kind of evidence that we need to > >>>> take > >>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>> fact), > >>>> and of course I've been aware of complaints about judgements by > >>>> people > >>>> like Labov for decades. > >>>> > >>>> But you're missing my main point, which is that all judgements > >>>> aren't > >>>> equally reliable. If you want to know how /unhappiness/ is > >>>> structured, > >>>> ask a linguist, not a five-year old. And one of the by-products of > >>>> education may be increased sensitivity to syntax - which is one of > >>>> the > >>>> many reasons why linguists need to pay more attention to education. > >>>> > >>>> Best wishes, Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about > >>>> methodology. I believe that intuitions are a fine way to generate > >>>> hypotheses and even to test them - to a degree. But while it might > >>>> not have > >>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>> the > >>>> Cambridge Grammar to conduct experiments on every point of the > >>>> grammar, > >>>> experiments could have only made the grammar better. The use of > >>>> intuitions, > >>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>> Standard > >>>> Social Science Methodology) is vital for taking the field forward > >>>> and for > >>>> providing the best support for different analyses. Ted Gibson and > >>>> Ev > >>>> Fedorenko have written a very useful new paper on this, showing > >>>> serious > >>>> shortcomings with intuitions as the sole source of evidence, in > >>>> their > >>>> paper: "The need for quantitative methods in syntax and semantics > >>>> research".> > >>>>> Carson Schutze and Wayne Cowart, among others, > >>>> have also written convincingly on this.> > >>>>> It is one reason that a team from Stanford, MIT > >>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>> beginning a > >>>> third round of experimental work among the Pirahas, since my own > >>>> work on > >>>> the syntax was, like almost every other field researcher's, based > >>>> on native > >>>> speaker intuitions and corpora.> > >>>>> The discussion of methodologies reminds me of > >>>> the initial reactions to Greenberg's work on classifying the > >>>> languages of > >>>> the Americas. His methods were strongly (and justifiably) > >>>> criticized. > >>>> However, I always thought that his methods were a great way of > >>>> generating > >>>> hypotheses, so long as they were ultimately put to the test of > >>>> standard > >>>> historical linguistics methods. And the same seems true for use of > >>>> native-speaker intuitions.> > >>>>> -- Dan > >>>>> > >>>>>> We linguists can add a further layer of > >>>> explanation to the judgements, but some judgements do seem to be > >>>> more > >>>> reliable than others. And if we have to wait for psycholinguistic > >>>> evidence > >>>> for every detailed analysis we make, our whole discipline will > >>>> immediately > >>>> grind to a halt. Like it or not, native speaker judgements are what > >>>> put us > >>>> linguists ahead of the rest in handling fine detail. Imagine > >>>> writing the > >>>> Cambridge Grammar of the English Language (or the OED) without > >>>> using native > >>>> speaker judgements.>> > >>>>>> Best wishes, Dick Hudson > >>>>>> > >>>>> > >>>>> > >>>> > >>>> > >>>> > >>>> > >>>> > >>> > >> > >> Lise Menn Home Office: 303-444-4274 > >> 1625 Mariposa Ave Fax: 303-413-0017 > >> Boulder CO 80302 > >> > >> Professor Emerita of Linguistics > >> Fellow, Institute of Cognitive Science > >> University of Colorado > >> > >> Secretary, AAAS Section Z [Linguistics] > >> > >> Campus Mail Address: > >> UCB 594, Institute of Cognitive Science > >> > >> Campus Physical Address: > >> CINC 234 > >> 1777 Exposition Ave, Boulder > >> > >> > >> > >> > > > > > > > > > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > > > ------------------------------ > > Message: 18 > Date: Fri, 10 Sep 2010 19:03:41 -0600 > From: Lise Menn > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Ted Gibson , Richard Hudson > , Evelina Fedorenko >, > funknet > Message-ID: <10638E94-DC8F-4181-8135-CE74F045079E at colorado.edu> > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dick, I think that what you say is true, but it is only (I hope) a > temporary description of the state of the art of psycholinguistics. > At least some of us would like our methods to become sensitive enough > to individual differences so that we can look at how 'the general > processes and principles' interact with the level of an individual > person's knowledge of particular constructions, to find out how much > each person knows of the patterns 'out there' in the language. Some > experimental methods are almost at that point already; they can > distinguish degrees of mastery of particular constructions of a > language among groups of second-language learners. > Have a look at > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > 2002. Overhearing a language during childhood. Psychological Science > 13.3, 238-243. > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding > on to childhood language memory. Cognition, 86(3), B53-B64. > > Lise > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a psycholinguist's > > view. Your goal is to find general processes and principles that > > apply uniformly across individuals, so you have to use methods to > > check for generality. And (as you know) I admire the way you pursue > > that goal. But my goal, as a linguist, is different. I want to > > explore the structure of a language so that I can understand how all > > the bits fit together. Like you, I'm aiming to model cognition, but > > my focus is on items and structures, and I start from the assumption > > that these can and do vary across speakers. > > > > However, having said all that I do agree with you that linguists > > should all get used to collecting and using quantitative data; and > > with the help of Brian MacWhinney's typology we'd know what methods > > to use when. And I do agree with your points about bid/bidded: in > > cases like that, quantitative data would be at least a very good > > starting point for a proper investigation. > > > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 19:30, Ted Gibson wrote: > >> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I don't understand what > >> is confusing about what Ev Fedorenko and I are claiming. All we are > >> saying is that if you have some testable claim involving a general > >> hypothesis about a language, then you need to get quantitative data > >> from unbiased sources to evaluate that claim. If you are interested > >> in English past tense morphology, then depending on the claims that > >> you might want to investigate, there are lots of ways to get > >> relevant quantitative evidence. Corpus data will probably be > >> useful. For very low frequency words, you can run experiments to > >> test behavior with respect to such words. > >> > >> Your example of the past tense of "bid" is a fine such example. You > >> can run an experiment like the one you suggested to find out what > >> people think the past tense is. If you then found that 20/50 people > >> responded "bidded" and 30/50 respond "bid", that is a lot of useful > >> information. As you suggest in your discussion, this result > >> wouldn't answer the question of how past tense is stored in each > >> individual. This result would be ambiguous among several possible > >> explanations. One possibility is that the probability distribution > >> that is being discovered reflects different dialects, such that 2/5 > >> of the population has one past tense, and 3/5 has another. Another > >> possibility is that each person has a similar probability > >> distribution in their heads, such that 2/5 of the time I respond > >> one way, and 3/5 of the time I respond another. Further experiments > >> would be necessary to answer between these and other possible > >> theories (e.g., with repeated trials from the same person, > >> carefully planned so that the participants don't notice that they > >> are being asked multiple times). Without the quantitative evidence > >> in the first place, there is no way to answer these kinds of > >> questions. > >> > >> Regarding the past tense of "go", this would be useful as a > >> baseline in an experiment involving the less frequent ones. So, > >> yes, it would useful to gather quantitative evidence in such a case > >> also, as baselines with respect to the more interesting cases for > >> theories. > >> > >> The bottom line: if you have a generalization about a language that > >> you wish to evaluate (such that you hypothesize that it is true > >> across the speakers of the language), then you need quantitative > >> evidence from multiple individuals, using an unbiased data > >> collection method, to evaluate such a claim. The point about > >> Mechanical Turk is that it is really *easy* to do this now, at > >> least for languages like English. > >> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> > >>> Dear Ted, > >>> Thanks for the very interesting comment, but are you REALLY saying > >>> that I shouldn't claim, for example, that the past tense of GO is > >>> "went" without first cross-checking with 50 native speakers? > >>> > >>> Isn't there a danger of missing the point that we all, as native > >>> speakers, spend our whole lives scanning other people's linguistic > >>> behaviour (language 'out there', E-language) and trying to explain > >>> it to ourselves in terms of a language system (language 'in here', > >>> I-language)? So every judgement we make is based on thousands or > >>> millions of observed exemplars, and reflects a unique experience > >>> of E-language filtered through a unique I-language. > >>> > >>> Given that view of language development, I don't see how > >>> quantitative data will help. Let's take a real uncertainty, such > >>> as the past tense of BID. If I want to say I did it, do I say "I > >>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>> and 30 "bid". So what? Does that mean that the correct answer is > >>> "bidded"? Surely not. How is it better than my judgement? I agree > >>> you could record my speech and find how often I use each > >>> alternative; but the reason I don't know is precisely because it's > >>> a rare word, so in a sense quantitative data are irrelevant even > >>> there. What would solve the problem of subjectivity, of course, > >>> would be a machine for probing the bit of my mind (or even brain) > >>> that holds BID and its details; but I suspect that even that > >>> wouldn't move us much further forward than my original "don't > >>> know". (Incidentally I write as a fan of quantitative > >>> sociolinguistics, so I do accept that quantitative data are > >>> relevant to linguistic analysis in some areas, where the I- > >>> language phenomenon is frequent enough to produce usable data.) > >>> > >>> It seems to me that this discussion raises the really fundamental > >>> question of what kind of thing we think language is: social or > >>> individual. The problem isn't unique to linguistics of course; > >>> it's the same throughout the social sciences. But what's special > >>> about linguistics is that we deal in very fine details of culture > >>> (e.g. details of how a particular word is used or pronounced) so > >>> the differences between individuals really matter. I don't see > >>> that we're ever going to have anything better than judgements to > >>> go on, so what we need is a way to ensure that judgements are > >>> accurate reports of individual I-language. A rotten situation for > >>> a science, but I don't see how it can get better. > >>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> > >>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points that Dan Everett makes, in > >>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a couple of papers recently > >>>> (Gibson & > >>>> Fedorenko, 2010, in press, see references and links below) on > >>>> what we > >>>> think are weak methodological standards in syntax and semantics > >>>> research over the past many years. The issue that we address is the > >>>> prevalent method in syntax and semantics research, which involves > >>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>> pair, typically by just the author of the paper, sometimes with > >>>> feedback from colleagues. As we address in our papers, this > >>>> methodology does not allow proper testing of scientific hypotheses > >>>> because of (a) the small number of experimental participants > >>>> (typically one); (b) the small number of experimental stimuli > >>>> (typically one); (c) cognitive biases on the part of the researcher > >>>> and participants; and (d) the effect of the preceding context > >>>> (e.g., > >>>> other constructions the researcher may have been recently > >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>> points, but with not as strong a conclusion as ours). > >>>> > >>>> Three issues need to be separated here: (1) the use of intuitive > >>>> judgments as a dependent measure in a language experiment; (2) > >>>> potential cognitive biases on the part of experimental subjects and > >>>> experimenters in language experiments; and (3) the need for > >>>> obtaining > >>>> quantitative evidence, whatever the dependent measure might be. The > >>>> paper that Ev and I wrote addresses the last two issues, but does > >>>> not > >>>> go into depth on the first issue (the use of intuitions as a > >>>> dependent > >>>> measure in language experiments). Regarding this issue, we don't > >>>> think > >>>> that there is anything wrong with gathering intuitive judgments > >>>> as a > >>>> dependent measure, as long as the task is clear to the experimental > >>>> participants. > >>>> > >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to > >>>> some > >>>> arguments that have been given in support of continuing to use the > >>>> traditional non-quantitative method in syntax / semantics research. > >>>> One recent defense of the traditional method comes from Phillips > >>>> (2008), who argues that no harm has come from the non-quantitative > >>>> approach in syntax research thus far. Phillips argues that there > >>>> are > >>>> no cases in the literature where an incorrect intuitive judgment > >>>> has > >>>> become the basis for a widely accepted generalization or an > >>>> important > >>>> theoretical claim. He therefore concludes that there is no reason > >>>> to > >>>> adopt more rigorous data collection standards. We challenge > >>>> Philips? > >>>> conclusion by presenting three cases from the literature where a > >>>> faulty intuition has led to incorrect generalizations and mistaken > >>>> theorizing, plausibly due to cognitive biases on the part of the > >>>> researchers. > >>>> > >>>> A second argument that is sometimes presented for the continued > >>>> use of > >>>> the traditional non-quantitative method is that it would be too > >>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>> phenomenon quantitatively. For example, Culicover & Jackendoff > >>>> (2010) > >>>> make this argument explicitly in their response to Gibson & > >>>> Fedorenko > >>>> (2010): ?It would cripple linguistic investigation if it were > >>>> required > >>>> that all judgments of ambiguity and grammaticality be subject to > >>>> statistically rigorous experiments on naive subjects, especially > >>>> when > >>>> investigating languages whose speakers are hard to > >>>> access? (Culicover > >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>> earlier in the discussion here.) Whereas we agree that in > >>>> circumstances where gathering data is difficult, some evidence is > >>>> better than no evidence, we do not agree that research would be > >>>> slowed > >>>> with respect to languages where experimental participants are > >>>> easy to > >>>> access, such as English. In contrast, we think that the opposite is > >>>> true: the field?s progress is probably slowed by not doing > >>>> quantitative research. > >>>> Suppose that a typical syntax / semantics paper that lacks > >>>> quantitative evidence includes judgments for 50 or more sentences / > >>>> meaning pairs, corresponding to 50 or more empirical claims. Even > >>>> if > >>>> most of the judgments from such a paper are correct or are on the > >>>> right track, the problem is in knowing which judgments are correct. > >>>> For example, suppose that 90% of the judgments from an arbitrary > >>>> paper > >>>> are correct (which is probably a high estimate). (Colin Phillips > >>>> and > >>>> some of his former students / postdocs have commented to us that, > >>>> in > >>>> their experience, quantitative acceptability judgment studies > >>>> almost > >>>> always validate the claim(s) in the literature. This is not our > >>>> experience, however. Most experiments that we have run which > >>>> attempt > >>>> to test some syntactic / semantic hypothesis in the literature > >>>> end up > >>>> providing us with a pattern of data that had not been known > >>>> before the > >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This > >>>> means > >>>> that in a paper with 50 empirical claims 45/50 are correct. But > >>>> which > >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s > >>>> over > >>>> two million different theories. By quantitatively evaluating the > >>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>> progress, it is better to have theoretical claims supported by > >>>> solid > >>>> quantitative evidence, so that even if the interpretation of the > >>>> data > >>>> changes over time as new evidence becomes available ? as is often > >>>> the > >>>> case in any field of science ? the empirical pattern can be used > >>>> as a > >>>> basis for further theorizing. > >>>> > >>>> Furthermore, it is no longer expensive to run behavioral > >>>> experiments, > >>>> at least in English and other widely spoken languages. There now > >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? > >>>> which > >>>> can be used for collecting behavioral data over the internet > >>>> quickly > >>>> and inexpensively. The cost of using an interface like this is > >>>> minimal, and the time that it takes for the results to be > >>>> returned is > >>>> short. For example, currently on Mechanical Turk, a survey of > >>>> approximately 50 items will be answered by 50 or more participants > >>>> within a couple of hours, at a cost of approximately $1 per > >>>> participant. Thus a survey can be completed within a day, at a > >>>> cost of > >>>> less than $50. (The hard work of designing the experiment, and > >>>> constructing controlled materials remains of course.) > >>>> > >>>> Sorry to be so verbose. But I think that these methodological > >>>> points > >>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>> methods in syntax and semantics research. Language and Cognitive > >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>> & Fedorenko InPress LCP.pdf > >>>> > >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >>>> Fedorenko > >>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about methodology. I believe > >>>>> that > >>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>> them - to a degree. But while it might not have been feasible for > >>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>> Grammar to conduct experiments on every point of the grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>> (indeed, Standard Social Science Methodology) is vital for > >>>>> taking the > >>>>> field forward and for providing the best support for different > >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful > >>>>> new > >>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>> sole source of evidence, in their paper: "The need for > >>>>> quantitative > >>>>> methods in syntax and semantics research". > >>>>> > >>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team from Stanford, MIT (Brain and > >>>>> Cognitive > >>>>> Science), and researchers from Brazil are beginning a third > >>>>> round of > >>>>> experimental work among the Pirahas, since my own work on the > >>>>> syntax > >>>>> was, like almost every other field researcher's, based on native > >>>>> speaker intuitions and corpora. > >>>>> > >>>>> The discussion of methodologies reminds me of the initial > >>>>> reactions > >>>>> to Greenberg's work on classifying the languages of the > >>>>> Americas. His > >>>>> methods were strongly (and justifiably) criticized. However, I > >>>>> always > >>>>> thought that his methods were a great way of generating > >>>>> hypotheses, > >>>>> so long as they were ultimately put to the test of standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions. > >>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a further layer of explanation to the > >>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>> every detailed analysis we make, our whole discipline will > >>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>> judgements are what put us linguists ahead of the rest in > >>>>>> handling > >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>> Language (or the OED) without using native speaker judgements. > >>>>>> > >>>>>> Best wishes, Dick Hudson > >>>> > >>>> > >>>> > >> > >> > >> > > Lise Menn Home Office: 303-444-4274 > 1625 Mariposa Ave Fax: 303-413-0017 > Boulder CO 80302 > > Professor Emerita of Linguistics > Fellow, Institute of Cognitive Science > University of Colorado > > Secretary, AAAS Section Z [Linguistics] > > Campus Mail Address: > UCB 594, Institute of Cognitive Science > > Campus Physical Address: > CINC 234 > 1777 Exposition Ave, Boulder > > > > > > ------------------------------ > > Message: 19 > Date: Fri, 10 Sep 2010 21:05:18 -0400 > From: Daniel Everett > Subject: Re: [FUNKNET] analysis: unhappiness > To: Richard Hudson > Cc: Ted Gibson , Richard Hudson > , Daniel Everett >, > Evelina Fedorenko , funknet > > Message-ID: > Content-Type: text/plain; charset=windows-1252 > > I think that Brian and Dick make excellent points. There are very good > grammars written that could be improved by psycholinguistic experimentation > and more quantitative approaches. But large sections of those grammars > aren't going to change one bit (go-went) with quantitative tests and such > tests would be completely counterproductive given the shortness of life and > the vastness of the field linguist's tasks. > > Part of the problem is that linguistics is not simply a subdiscipline of > psychology. Linguistics has its own objectives and those only occasionally > overlap with psychology. The same for methods. > > On another note, I don't buy the 'in my head' 'out of my head' distinction > either (that Matt seems to be urging upon us). We study different things and > have different reasons for being satisfied with the results we achieve. > > I believe that we linguists are often complacent and fail to apply better > methods. But of course that applies to all disciplines. > > In the meantime, checking corpora, collecting data as a result of careful > interviews with native speakers, and the other aspects of the field > linguist's task are vital parts of the linguist's task and much of this > won't be improved by quantitative methods as we currently understand them. > Maybe sometime. > > Dan > > P.S. In my original reference to Ted and Ev's paper, I said that they > showed the danger of using intuitions. What I meant to say of using > intuitions as standardly used by linguists. They convinced me that there is > a lot to learn from quantitative methods. > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > Dear Ted and Ev, > > Yes, I understand your view, but I think it's a psycholinguist's view. > Your goal is to find general processes and principles that apply uniformly > across individuals, so you have to use methods to check for generality. And > (as you know) I admire the way you pursue that goal. But my goal, as a > linguist, is different. I want to explore the structure of a language so > that I can understand how all the bits fit together. Like you, I'm aiming to > model cognition, but my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers. > > > > However, having said all that I do agree with you that linguists should > all get used to collecting and using quantitative data; and with the help of > Brian MacWhinney's typology we'd know what methods to use when. And I do > agree with your points about bid/bidded: in cases like that, quantitative > data would be at least a very good starting point for a proper > investigation. > > > > Best wishes, Dick > > > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > > > On 10/09/2010 19:30, Ted Gibson wrote: > >> Dear Dick: > >> > >> Perhaps we are talking at cross purposes. I don't understand what is > confusing about what Ev Fedorenko and I are claiming. All we are saying is > that if you have some testable claim involving a general hypothesis about a > language, then you need to get quantitative data from unbiased sources to > evaluate that claim. If you are interested in English past tense morphology, > then depending on the claims that you might want to investigate, there are > lots of ways to get relevant quantitative evidence. Corpus data will > probably be useful. For very low frequency words, you can run experiments to > test behavior with respect to such words. > >> > >> Your example of the past tense of "bid" is a fine such example. You can > run an experiment like the one you suggested to find out what people think > the past tense is. If you then found that 20/50 people responded "bidded" > and 30/50 respond "bid", that is a lot of useful information. As you suggest > in your discussion, this result wouldn't answer the question of how past > tense is stored in each individual. This result would be ambiguous among > several possible explanations. One possibility is that the probability > distribution that is being discovered reflects different dialects, such that > 2/5 of the population has one past tense, and 3/5 has another. Another > possibility is that each person has a similar probability distribution in > their heads, such that 2/5 of the time I respond one way, and 3/5 of the > time I respond another. Further experiments would be necessary to answer > between these and other possible theories (e.g., with repeated trials from > the same person, carefully pl > anned so that the participants don't notice that they are being asked > multiple times). Without the quantitative evidence in the first place, there > is no way to answer these kinds of questions. > >> > >> Regarding the past tense of "go", this would be useful as a baseline in > an experiment involving the less frequent ones. So, yes, it would useful to > gather quantitative evidence in such a case also, as baselines with respect > to the more interesting cases for theories. > >> > >> The bottom line: if you have a generalization about a language that you > wish to evaluate (such that you hypothesize that it is true across the > speakers of the language), then you need quantitative evidence from multiple > individuals, using an unbiased data collection method, to evaluate such a > claim. The point about Mechanical Turk is that it is really *easy* to do > this now, at least for languages like English. > >> > >> Best wishes, > >> > >> Ted Gibson & Ev Fedorenko > >> > >> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >> > >>> Dear Ted, > >>> Thanks for the very interesting comment, but are you REALLY saying that > I shouldn't claim, for example, that the past tense of GO is "went" without > first cross-checking with 50 native speakers? > >>> > >>> Isn't there a danger of missing the point that we all, as native > speakers, spend our whole lives scanning other people's linguistic behaviour > (language 'out there', E-language) and trying to explain it to ourselves in > terms of a language system (language 'in here', I-language)? So every > judgement we make is based on thousands or millions of observed exemplars, > and reflects a unique experience of E-language filtered through a unique > I-language. > >>> > >>> Given that view of language development, I don't see how quantitative > data will help. Let's take a real uncertainty, such as the past tense of > BID. If I want to say I did it, do I say "I bidded" or "I bid"? My > judgement: I don't know. Ok, you get 50 people to oblige on Mechanical Turk, > and 20 of them give "bidded" and 30 "bid". So what? Does that mean that the > correct answer is "bidded"? Surely not. How is it better than my judgement? > I agree you could record my speech and find how often I use each > alternative; but the reason I don't know is precisely because it's a rare > word, so in a sense quantitative data are irrelevant even there. What would > solve the problem of subjectivity, of course, would be a machine for probing > the bit of my mind (or even brain) that holds BID and its details; but I > suspect that even that wouldn't move us much further forward than my > original "don't know". (Incidentally I write as a fan of quantitative > sociolinguistics, so I do accept that quant > itative data are relevant to linguistic analysis in some areas, where the > I-language phenomenon is frequent enough to produce usable data.) > >>> > >>> It seems to me that this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual. > The problem isn't unique to linguistics of course; it's the same throughout > the social sciences. But what's special about linguistics is that we deal in > very fine details of culture (e.g. details of how a particular word is used > or pronounced) so the differences between individuals really matter. I don't > see that we're ever going to have anything better than judgements to go on, > so what we need is a way to ensure that judgements are accurate reports of > individual I-language. A rotten situation for a science, but I don't see how > it can get better. > >>> > >>> Dick > >>> > >>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>> > >>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>> Dear Dan, Dick: > >>>> > >>>> I would like to clarify some points that Dan Everett makes, in > >>>> response to Dick Hudson. > >>>> > >>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>>> Fedorenko, 2010, in press, see references and links below) on what we > >>>> think are weak methodological standards in syntax and semantics > >>>> research over the past many years. The issue that we address is the > >>>> prevalent method in syntax and semantics research, which involves > >>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>> pair, typically by just the author of the paper, sometimes with > >>>> feedback from colleagues. As we address in our papers, this > >>>> methodology does not allow proper testing of scientific hypotheses > >>>> because of (a) the small number of experimental participants > >>>> (typically one); (b) the small number of experimental stimuli > >>>> (typically one); (c) cognitive biases on the part of the researcher > >>>> and participants; and (d) the effect of the preceding context (e.g., > >>>> other constructions the researcher may have been recently > >>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>> points, but with not as strong a conclusion as ours). > >>>> > >>>> Three issues need to be separated here: (1) the use of intuitive > >>>> judgments as a dependent measure in a language experiment; (2) > >>>> potential cognitive biases on the part of experimental subjects and > >>>> experimenters in language experiments; and (3) the need for obtaining > >>>> quantitative evidence, whatever the dependent measure might be. The > >>>> paper that Ev and I wrote addresses the last two issues, but does not > >>>> go into depth on the first issue (the use of intuitions as a dependent > >>>> measure in language experiments). Regarding this issue, we don't think > >>>> that there is anything wrong with gathering intuitive judgments as a > >>>> dependent measure, as long as the task is clear to the experimental > >>>> participants. > >>>> > >>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>>> arguments that have been given in support of continuing to use the > >>>> traditional non-quantitative method in syntax / semantics research. > >>>> One recent defense of the traditional method comes from Phillips > >>>> (2008), who argues that no harm has come from the non-quantitative > >>>> approach in syntax research thus far. Phillips argues that there are > >>>> no cases in the literature where an incorrect intuitive judgment has > >>>> become the basis for a widely accepted generalization or an important > >>>> theoretical claim. He therefore concludes that there is no reason to > >>>> adopt more rigorous data collection standards. We challenge Philips? > >>>> conclusion by presenting three cases from the literature where a > >>>> faulty intuition has led to incorrect generalizations and mistaken > >>>> theorizing, plausibly due to cognitive biases on the part of the > >>>> researchers. > >>>> > >>>> A second argument that is sometimes presented for the continued use of > >>>> the traditional non-quantitative method is that it would be too > >>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>>> make this argument explicitly in their response to Gibson & Fedorenko > >>>> (2010): ?It would cripple linguistic investigation if it were required > >>>> that all judgments of ambiguity and grammaticality be subject to > >>>> statistically rigorous experiments on naive subjects, especially when > >>>> investigating languages whose speakers are hard to access? (Culicover > >>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>> earlier in the discussion here.) Whereas we agree that in > >>>> circumstances where gathering data is difficult, some evidence is > >>>> better than no evidence, we do not agree that research would be slowed > >>>> with respect to languages where experimental participants are easy to > >>>> access, such as English. In contrast, we think that the opposite is > >>>> true: the field?s progress is probably slowed by not doing > >>>> quantitative research. > >>>> Suppose that a typical syntax / semantics paper that lacks > >>>> quantitative evidence includes judgments for 50 or more sentences / > >>>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>>> most of the judgments from such a paper are correct or are on the > >>>> right track, the problem is in knowing which judgments are correct. > >>>> For example, suppose that 90% of the judgments from an arbitrary paper > >>>> are correct (which is probably a high estimate). (Colin Phillips and > >>>> some of his former students / postdocs have commented to us that, in > >>>> their experience, quantitative acceptability judgment studies almost > >>>> always validate the claim(s) in the literature. This is not our > >>>> experience, however. Most experiments that we have run which attempt > >>>> to test some syntactic / semantic hypothesis in the literature end up > >>>> providing us with a pattern of data that had not been known before the > >>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s over > >>>> two million different theories. By quantitatively evaluating the > >>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>> progress, it is better to have theoretical claims supported by solid > >>>> quantitative evidence, so that even if the interpretation of the data > >>>> changes over time as new evidence becomes available ? as is often the > >>>> case in any field of science ? the empirical pattern can be used as a > >>>> basis for further theorizing. > >>>> > >>>> Furthermore, it is no longer expensive to run behavioral experiments, > >>>> at least in English and other widely spoken languages. There now > >>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? which > >>>> can be used for collecting behavioral data over the internet quickly > >>>> and inexpensively. The cost of using an interface like this is > >>>> minimal, and the time that it takes for the results to be returned is > >>>> short. For example, currently on Mechanical Turk, a survey of > >>>> approximately 50 items will be answered by 50 or more participants > >>>> within a couple of hours, at a cost of approximately $1 per > >>>> participant. Thus a survey can be completed within a day, at a cost of > >>>> less than $50. (The hard work of designing the experiment, and > >>>> constructing controlled materials remains of course.) > >>>> > >>>> Sorry to be so verbose. But I think that these methodological points > >>>> are very important. > >>>> > >>>> Best wishes, > >>>> > >>>> Ted Gibson > >>>> > >>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>> methods in syntax and semantics research. Language and Cognitive > >>>> Processes. http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>> & Fedorenko InPress LCP.pdf > >>>> > >>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > Fedorenko > >>>> 2010 TICS.pdf > >>>> > >>>> > >>>> > >>>> > >>>>> Dick, > >>>>> > >>>>> You raise an important issue here about methodology. I believe that > >>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>> them - to a degree. But while it might not have been feasible for > >>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>> Grammar to conduct experiments on every point of the grammar, > >>>>> experiments could have only made the grammar better. The use of > >>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>> (indeed, Standard Social Science Methodology) is vital for taking the > >>>>> field forward and for providing the best support for different > >>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>> sole source of evidence, in their paper: "The need for quantitative > >>>>> methods in syntax and semantics research". > >>>>> > >>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>> convincingly on this. > >>>>> > >>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>>> Science), and researchers from Brazil are beginning a third round of > >>>>> experimental work among the Pirahas, since my own work on the syntax > >>>>> was, like almost every other field researcher's, based on native > >>>>> speaker intuitions and corpora. > >>>>> > >>>>> The discussion of methodologies reminds me of the initial reactions > >>>>> to Greenberg's work on classifying the languages of the Americas. His > >>>>> methods were strongly (and justifiably) criticized. However, I always > >>>>> thought that his methods were a great way of generating hypotheses, > >>>>> so long as they were ultimately put to the test of standard > >>>>> historical linguistics methods. And the same seems true for use of > >>>>> native-speaker intuitions. > >>>>> > >>>>> -- Dan > >>>> > >>>> > >>>> > >>>>>> We linguists can add a further layer of explanation to the > >>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>> every detailed analysis we make, our whole discipline will > >>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>> judgements are what put us linguists ahead of the rest in handling > >>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>> Language (or the OED) without using native speaker judgements. > >>>>>> > >>>>>> Best wishes, Dick Hudson > >>>> > >>>> > >>>> > >> > >> > >> > > > > > > ------------------------------ > > Message: 20 > Date: Sat, 11 Sep 2010 10:03:35 +0000 (GMT) > From: Philippe De Brabanter > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > To: funknet at mailman.rice.edu > Message-ID: <650247.87750.qm at web25504.mail.ukl.yahoo.com> > Content-Type: text/plain; charset=utf-8 > > Dear all, > > this is just to say that the Cambridge Grammar of the English Language, > like > several other grammars I'm aware of, does mention the sorts of > constructions Tom > is wondering about (esp. pp. 964-67). They're treated as noun complements, > whereas relative clauses usually function as modifiers of nouns. > H&P give a useful list of nouns licensing these complements (a list which > confirms Suzanne Kemmerer's point that these nouns do not always have a > verbal > counterpart taking a content clause as its complement ? H&P suggest that > the > most frequent of those licensing nouns is fact). > They also point out that content clauses can also sometimes function as > supplements (i.e. appositives), as in > > I'm inclined to favour your first suggestion, that we shelve the proposal > until > after the election. > > This confirms Suzanne's suggestion that we shouldn't say that the clausal > noun > complements are appositives. > > One last interesting point. On p. 967, H&P show that the licensor may > sometimes > be more than just a noun, with certain constructions like have + licensing > NP or > existential there + be facilitating (or being conditions for) the clausal > noun > complement: > > The present system has the disadvantage that it is inordinately > complicated. > vs. ? The disadvantage that it is inordinately complicated has been > overlooked. > > Probably an example like > > This principle may ground some optimism that the account can be usefully > pursued. (M. Sainsbury 2002: "Reference and anaphora", Mind & Language) > > also derives its acceptability from a construction rather than from just > optimism. > > Best, > > Philippe De Brabanter > Paris 4 - Sorbonne > > > > ________________________________ > From: E.G. > To: funknet at mailman.rice.edu > Sent: Fri, 10 September, 2010 19:56:23 > Subject: Re: [FUNKNET] "Relative clauses" with no relativized element > > Jespersen and his nexus-substantives should be mentioned (Philosophy of > Grammar, 1924). Also in his MEG and Analytic Syntax one could find > interesting discussions. > > Eitan > > > On 10 September 2010 20:53, Giuliana Fiorentino < > giuliana.fiorentino at unimol.it> wrote: > > > Hi Tom, > > clauses like: > > > > The importance of being Earnest > > the fact of being late > > the fact that you are late > > the idea that world is round > > etcetera > > > > are not relative clauses but can be considered among syntactic strategies > > in order to nominalise events after a generic noun (working as a > classifier > > for nominalised events). > > > > Giuliana > > > > ----- Original Message ----- > > From: Thomas E. Payne > > To: FUNKNET > > Sent: Friday, September 10, 2010 4:16 PM > > Subject: [FUNKNET] "Relative clauses" with no relativized element > > > > > > Can anyone help me name the following structure in English, and maybe > > point > > me to some references? I do not find reference to this in the Cambridge > > Grammar of the English Language or any other of my English grammar > books. > > But then, maybe I just don't know where to look. > > > > Here are two examples from a play: > > > > His protestations of devotion in the trial scene are, in our opinion, > > genuine, as is his confession [that his affair with the Countess is > > platonic]. > > > > The bracketed clause seems to modify "confession", though there is no > > position for a confession in the clause itself. > > > > . . . forced hither with an impious black design [to have my innocence > and > > youth become the sacrifice of brutal violence]. > > > > Here the bracketed non-finite clause seems to modify "design." > > > > These are not all that rare. I'm reminded of examples like: > > > > "The claim [that my client is a murderer] is totally false." > > > > Are these relative clauses? If so what kind? Thanks for any help. > > > > Tom Payne > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > > > > ------------------------------ > > Message: 21 > Date: Sat, 11 Sep 2010 12:03:56 +0100 > From: Richard Hudson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Cc: Jennifer Smith > Message-ID: <4C8B621C.6090609 at ling.ucl.ac.uk> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Dear Lise, > Many thanks for these references, which I've just looked at. As you say, > they do offer hope that we can apply the methods of psycholinguists to > the questions of linguists (e.g. how do individuals categorise Korean or > Spanish consonants?), and at the same time build a clearer understanding > of how these individuals' I-language is related to their E-language > (i.e. the language they've heard). That is encouraging, not least > because it breaks down what I see as a gulf between linguists and > psycholinguists. > > But even better, it complements very nicely some work in > sociolinguistics which I think you'd enjoy reading. It's by a young > Scottish sociolinguist called Jennifer Smith, who did a very careful > quantitative study (with two colleagues) of 24 3-year olds in a small > fishing town in the north of Scotland. She recorded each of them with > their mother, and then analysed two sociolinguistic variables that she'd > also analysed in adult speech: a phonological variable (pronunciation of > the /au/ vowel in "cow", and a morphosyntactic one, the use of -s on a > verb with a plural subject, e.g. "My trousers is falling doon." In both > cases usage is variable, so the analysis produces a percentage score for > each speaker (e.g. 5% of words with the /au/ vowel have a monophthong). > She then compared the children's scores with those of their mothers, and > found an astonishingly close match for the phonological variable but no > match at all for the morphosyntactic one. The reference is > > Jennifer Smith, Mercedes Durham & Liane Fortune, 2007. 'Mam, my > trousers is fa'in doon!' Community, caregiver and child in the > acquisition of variation in a Scottish dialect. (Language, Variation > and Change 19. 63-99) > > Once again we have evidence that output can be closely linked to input, > we have a nice quantitative method, and we see an example of the > 'global' analysis of language that we should all be striving for: one > which embraces both E-language (directly observable to the learner as > well as to the linguist) and I-language (only indirectly accessible to > both), and which tries to both describe and explain the relation between > the two. Neither E-language nor I-language is the 'real' language - > they're both part of it. And this global enterprise needs all the > methods we can muster. > > Best wishes, Dick > > Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > > On 11/09/2010 02:03, Lise Menn wrote: > > Dick, I think that what you say is true, but it is only (I hope) a > > temporary description of the state of the art of psycholinguistics. > > At least some of us would like our methods to become sensitive enough > > to individual differences so that we can look at how 'the general > > processes and principles' interact with the level of an individual > > person's knowledge of particular constructions, to find out how much > > each person knows of the patterns 'out there' in the language. Some > > experimental methods are almost at that point already; they can > > distinguish degrees of mastery of particular constructions of a > > language among groups of second-language learners. > > Have a look at > > Au, Terry Kit-fong, Leah M. Knightly, Sun-Ah Jun, and Janet S. Oh. > > 2002. Overhearing a language during childhood. /Psychological Science/ > > 13.3, 238-243. > > > > Oh, J. S., Jun, S.-A., Knightly, L. M., & Au, T. K. 2003. Holding on > > to childhood language memory. /Cognition, 86/(3), B53-B64. > > > > Lise > > > > On Sep 10, 2010, at 5:40 PM, Richard Hudson wrote: > > > >> Dear Ted and Ev, > >> Yes, I understand your view, but I think it's a psycholinguist's > >> view. Your goal is to find general processes and principles that > >> apply uniformly across individuals, so you have to use methods to > >> check for generality. And (as you know) I admire the way you pursue > >> that goal. But my goal, as a linguist, is different. I want to > >> explore the structure of a language so that I can understand how all > >> the bits fit together. Like you, I'm aiming to model cognition, but > >> my focus is on items and structures, and I start from the assumption > >> that these can and do vary across speakers. > >> > >> However, having said all that I do agree with you that linguists > >> should all get used to collecting and using quantitative data; and > >> with the help of Brian MacWhinney's typology we'd know what methods > >> to use when. And I do agree with your points about bid/bidded: in > >> cases like that, quantitative data would be at least a very good > >> starting point for a proper investigation. > >> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> > >> On 10/09/2010 19:30, Ted Gibson wrote: > >>> Dear Dick: > >>> > >>> Perhaps we are talking at cross purposes. I don't understand what is > >>> confusing about what Ev Fedorenko and I are claiming. All we are > >>> saying is that if you have some testable claim involving a general > >>> hypothesis about a language, then you need to get quantitative data > >>> from unbiased sources to evaluate that claim. If you are interested > >>> in English past tense morphology, then depending on the claims that > >>> you might want to investigate, there are lots of ways to get > >>> relevant quantitative evidence. Corpus data will probably be useful. > >>> For very low frequency words, you can run experiments to test > >>> behavior with respect to such words. > >>> > >>> Your example of the past tense of "bid" is a fine such example. You > >>> can run an experiment like the one you suggested to find out what > >>> people think the past tense is. If you then found that 20/50 people > >>> responded "bidded" and 30/50 respond "bid", that is a lot of useful > >>> information. As you suggest in your discussion, this result wouldn't > >>> answer the question of how past tense is stored in each individual. > >>> This result would be ambiguous among several possible explanations. > >>> One possibility is that the probability distribution that is being > >>> discovered reflects different dialects, such that 2/5 of the > >>> population has one past tense, and 3/5 has another. Another > >>> possibility is that each person has a similar probability > >>> distribution in their heads, such that 2/5 of the time I respond one > >>> way, and 3/5 of the time I respond another. Further experiments > >>> would be necessary to answer between these and other possible > >>> theories (e.g., with repeated trials from the same person, carefully > >>> planned so that the participants don't notice that they are being > >>> asked multiple times). Without the quantitative evidence in the > >>> first place, there is no way to answer these kinds of questions. > >>> > >>> Regarding the past tense of "go", this would be useful as a baseline > >>> in an experiment involving the less frequent ones. So, yes, it would > >>> useful to gather quantitative evidence in such a case also, as > >>> baselines with respect to the more interesting cases for theories. > >>> > >>> The bottom line: if you have a generalization about a language that > >>> you wish to evaluate (such that you hypothesize that it is true > >>> across the speakers of the language), then you need quantitative > >>> evidence from multiple individuals, using an unbiased data > >>> collection method, to evaluate such a claim. The point about > >>> Mechanical Turk is that it is really *easy* to do this now, at least > >>> for languages like English. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson & Ev Fedorenko > >>> > >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >>> > >>>> Dear Ted, > >>>> Thanks for the very interesting comment, but are you REALLY saying > >>>> that I shouldn't claim, for example, that the past tense of GO is > >>>> "went" without first cross-checking with 50 native speakers? > >>>> > >>>> Isn't there a danger of missing the point that we all, as native > >>>> speakers, spend our whole lives scanning other people's linguistic > >>>> behaviour (language 'out there', E-language) and trying to explain > >>>> it to ourselves in terms of a language system (language 'in here', > >>>> I-language)? So every judgement we make is based on thousands or > >>>> millions of observed exemplars, and reflects a unique experience of > >>>> E-language filtered through a unique I-language. > >>>> > >>>> Given that view of language development, I don't see how > >>>> quantitative data will help. Let's take a real uncertainty, such as > >>>> the past tense of BID. If I want to say I did it, do I say "I > >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>>> and 30 "bid". So what? Does that mean that the correct answer is > >>>> "bidded"? Surely not. How is it better than my judgement? I agree > >>>> you could record my speech and find how often I use each > >>>> alternative; but the reason I don't know is precisely because it's > >>>> a rare word, so in a sense quantitative data are irrelevant even > >>>> there. What would solve the problem of subjectivity, of course, > >>>> would be a machine for probing the bit of my mind (or even brain) > >>>> that holds BID and its details; but I suspect that even that > >>>> wouldn't move us much further forward than my original "don't > >>>> know". (Incidentally I write as a fan of quantitative > >>>> sociolinguistics, so I do accept that quantitative data are > >>>> relevant to linguistic analysis in some areas, where the I-language > >>>> phenomenon is frequent enough to produce usable data.) > >>>> > >>>> It seems to me that this discussion raises the really fundamental > >>>> question of what kind of thing we think language is: social or > >>>> individual. The problem isn't unique to linguistics of course; it's > >>>> the same throughout the social sciences. But what's special about > >>>> linguistics is that we deal in very fine details of culture (e.g. > >>>> details of how a particular word is used or pronounced) so the > >>>> differences between individuals really matter. I don't see that > >>>> we're ever going to have anything better than judgements to go on, > >>>> so what we need is a way to ensure that judgements are accurate > >>>> reports of individual I-language. A rotten situation for a science, > >>>> but I don't see how it can get better. > >>>> > >>>> Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> > >>>> > >>>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>>> Dear Dan, Dick: > >>>>> > >>>>> I would like to clarify some points that Dan Everett makes, in > >>>>> response to Dick Hudson. > >>>>> > >>>>> Ev Fedorenko and I have written a couple of papers recently (Gibson & > >>>>> Fedorenko, 2010, in press, see references and links below) on what we > >>>>> think are weak methodological standards in syntax and semantics > >>>>> research over the past many years. The issue that we address is the > >>>>> prevalent method in syntax and semantics research, which involves > >>>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>>> pair, typically by just the author of the paper, sometimes with > >>>>> feedback from colleagues. As we address in our papers, this > >>>>> methodology does not allow proper testing of scientific hypotheses > >>>>> because of (a) the small number of experimental participants > >>>>> (typically one); (b) the small number of experimental stimuli > >>>>> (typically one); (c) cognitive biases on the part of the researcher > >>>>> and participants; and (d) the effect of the preceding context (e.g., > >>>>> other constructions the researcher may have been recently > >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>>> points, but with not as strong a conclusion as ours). > >>>>> > >>>>> Three issues need to be separated here: (1) the use of intuitive > >>>>> judgments as a dependent measure in a language experiment; (2) > >>>>> potential cognitive biases on the part of experimental subjects and > >>>>> experimenters in language experiments; and (3) the need for obtaining > >>>>> quantitative evidence, whatever the dependent measure might be. The > >>>>> paper that Ev and I wrote addresses the last two issues, but does not > >>>>> go into depth on the first issue (the use of intuitions as a > dependent > >>>>> measure in language experiments). Regarding this issue, we don't > think > >>>>> that there is anything wrong with gathering intuitive judgments as a > >>>>> dependent measure, as long as the task is clear to the experimental > >>>>> participants. > >>>>> > >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to some > >>>>> arguments that have been given in support of continuing to use the > >>>>> traditional non-quantitative method in syntax / semantics research. > >>>>> One recent defense of the traditional method comes from Phillips > >>>>> (2008), who argues that no harm has come from the non-quantitative > >>>>> approach in syntax research thus far. Phillips argues that there are > >>>>> no cases in the literature where an incorrect intuitive judgment has > >>>>> become the basis for a widely accepted generalization or an important > >>>>> theoretical claim. He therefore concludes that there is no reason to > >>>>> adopt more rigorous data collection standards. We challenge Philips' > >>>>> conclusion by presenting three cases from the literature where a > >>>>> faulty intuition has led to incorrect generalizations and mistaken > >>>>> theorizing, plausibly due to cognitive biases on the part of the > >>>>> researchers. > >>>>> > >>>>> A second argument that is sometimes presented for the continued use > of > >>>>> the traditional non-quantitative method is that it would be too > >>>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff (2010) > >>>>> make this argument explicitly in their response to Gibson & Fedorenko > >>>>> (2010): "It would cripple linguistic investigation if it were > required > >>>>> that all judgments of ambiguity and grammaticality be subject to > >>>>> statistically rigorous experiments on naive subjects, especially when > >>>>> investigating languages whose speakers are hard to access" (Culicover > >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>>> earlier in the discussion here.) Whereas we agree that in > >>>>> circumstances where gathering data is difficult, some evidence is > >>>>> better than no evidence, we do not agree that research would be > slowed > >>>>> with respect to languages where experimental participants are easy to > >>>>> access, such as English. In contrast, we think that the opposite is > >>>>> true: the field's progress is probably slowed by not doing > >>>>> quantitative research. > >>>>> Suppose that a typical syntax / semantics paper that lacks > >>>>> quantitative evidence includes judgments for 50 or more sentences / > >>>>> meaning pairs, corresponding to 50 or more empirical claims. Even if > >>>>> most of the judgments from such a paper are correct or are on the > >>>>> right track, the problem is in knowing which judgments are correct. > >>>>> For example, suppose that 90% of the judgments from an arbitrary > paper > >>>>> are correct (which is probably a high estimate). (Colin Phillips and > >>>>> some of his former students / postdocs have commented to us that, in > >>>>> their experience, quantitative acceptability judgment studies almost > >>>>> always validate the claim(s) in the literature. This is not our > >>>>> experience, however. Most experiments that we have run which attempt > >>>>> to test some syntactic / semantic hypothesis in the literature end up > >>>>> providing us with a pattern of data that had not been known before > the > >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This means > >>>>> that in a paper with 50 empirical claims 45/50 are correct. But which > >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That's over > >>>>> two million different theories. By quantitatively evaluating the > >>>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>>> progress, it is better to have theoretical claims supported by solid > >>>>> quantitative evidence, so that even if the interpretation of the data > >>>>> changes over time as new evidence becomes available -- as is often > the > >>>>> case in any field of science -- the empirical pattern can be used as > a > >>>>> basis for further theorizing. > >>>>> > >>>>> Furthermore, it is no longer expensive to run behavioral experiments, > >>>>> at least in English and other widely spoken languages. There now > >>>>> exists a marketplace interface -- Amazon.com's Mechanical Turk -- > >>>>> which > >>>>> can be used for collecting behavioral data over the internet quickly > >>>>> and inexpensively. The cost of using an interface like this is > >>>>> minimal, and the time that it takes for the results to be returned is > >>>>> short. For example, currently on Mechanical Turk, a survey of > >>>>> approximately 50 items will be answered by 50 or more participants > >>>>> within a couple of hours, at a cost of approximately $1 per > >>>>> participant. Thus a survey can be completed within a day, at a cost > of > >>>>> less than $50. (The hard work of designing the experiment, and > >>>>> constructing controlled materials remains of course.) > >>>>> > >>>>> Sorry to be so verbose. But I think that these methodological points > >>>>> are very important. > >>>>> > >>>>> Best wishes, > >>>>> > >>>>> Ted Gibson > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>>> methods in syntax and semantics research. Language and Cognitive > >>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>>> & Fedorenko InPress LCP.pdf > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > Fedorenko > >>>>> 2010 TICS.pdf > >>>>> > >>>>> > >>>>> > >>>>> > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about methodology. I believe that > >>>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>>> them - to a degree. But while it might not have been feasible for > >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>>> Grammar to conduct experiments on every point of the grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, corpora, and standard psycholinguistic experimentation > >>>>>> (indeed, Standard Social Science Methodology) is vital for taking > the > >>>>>> field forward and for providing the best support for different > >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very useful new > >>>>>> paper on this, showing serious shortcomings with intuitions as the > >>>>>> sole source of evidence, in their paper: "The need for quantitative > >>>>>> methods in syntax and semantics research". > >>>>>> > >>>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>>> convincingly on this. > >>>>>> > >>>>>> It is one reason that a team from Stanford, MIT (Brain and Cognitive > >>>>>> Science), and researchers from Brazil are beginning a third round of > >>>>>> experimental work among the Pirahas, since my own work on the syntax > >>>>>> was, like almost every other field researcher's, based on native > >>>>>> speaker intuitions and corpora. > >>>>>> > >>>>>> The discussion of methodologies reminds me of the initial reactions > >>>>>> to Greenberg's work on classifying the languages of the Americas. > His > >>>>>> methods were strongly (and justifiably) criticized. However, I > always > >>>>>> thought that his methods were a great way of generating hypotheses, > >>>>>> so long as they were ultimately put to the test of standard > >>>>>> historical linguistics methods. And the same seems true for use of > >>>>>> native-speaker intuitions. > >>>>>> > >>>>>> -- Dan > >>>>> > >>>>> > >>>>> > >>>>>>> We linguists can add a further layer of explanation to the > >>>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>>> every detailed analysis we make, our whole discipline will > >>>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>>> judgements are what put us linguists ahead of the rest in handling > >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the English > >>>>>>> Language (or the OED) without using native speaker judgements. > >>>>>>> > >>>>>>> Best wishes, Dick Hudson > >>>>> > >>>>> > >>>>> > >>> > >>> > >>> > > > > Lise Menn Home Office: 303-444-4274 > > 1625 Mariposa Ave Fax: 303-413-0017 > > Boulder CO 80302 > > > > Professor Emerita of Linguistics > > Fellow, Institute of Cognitive Science > > University of Colorado > > > > Secretary, AAAS Section Z [Linguistics] > > > > Campus Mail Address: > > UCB 594, Institute of Cognitive Science > > > > Campus Physical Address: > > CINC 234 > > 1777 Exposition Ave, Boulder > > > > > > > > > ------------------------------ > > Message: 22 > Date: Sat, 11 Sep 2010 06:00:08 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] analysis: unhappiness > To: dryer at buffalo.edu > Cc: Lise Menn , Richard Hudson > , Funknet > Message-ID: > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > Matt, > > We appear to be fairly close in our approach, but I would have to add that > sometimes there isn't a unique "correct" analysis, because the language > allows equally for several different ones, and how any particular speaker > analyzes a phrase out of context says more about how their individual > brain is wired and less about the language. > > As an example, take the slogan the Coca-Cola company is currently using: > > "Open happiness". > > I first saw it on a cocktail napkin in flight. Reading the English slogan, > my first analysis was that "open" was an adjective modifying the noun > happiness, as opposed to say "closed happiness." > > That seemed weird, so I considered a few other possiblities. Maybe "open" > is a verb in the imperative, and "happiness" is a proper noun in the > vocative, as in "Open Sesame!" > > Then again, it could be that "Happiness" was just a proper noun in > objective case: as in "open America (to tourism)." > > Then I read the French translation on the napkin: "Ouvrez du bonheur." > > "Oh! So this means "open some happiness"!" I said to myself. > > English lexemes are underspecified for category, which is why we need > little words like "some" to disamnbiguate. That's how the language works. > But... all those different analyses could have been correct, given the > proper context, and experimenting even with a large population as to which > one they thought of first would tell you less about the language and more > about the people. > > The only analysis that seems a bit doubtful is the one suggested by the > translation. > > Best, > > --Aya > > > On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > > > > > Aya, > > > > I actually agree with everything you say here. Personally, I am MORE > > interested in the communicative function of language than I am in > > psycholinguistics and how people process language. > > > > But none of that is relevant, I believe, to the very specific question of > > what it means for an analysis to be correct. While one might conclude > from > > what I said that one ought to do psycholinguistics, that is not my > intention. > > Rather, my conclusion is that since I myself prefer not to do > > psycholinguistics, I cannot really claim that the analyses I come up with > are > > "the correct" ones. And if it is really important to someone that they > > identify "correct" analyses, then they ought to be doing > psycholinguistics, > > since there is no coherent notion of correct analysis outside of what is > > inside of people's heads. > > > > Matthew > > > > --On Friday, September 10, 2010 12:09 PM -0700 "A. Katz" > > > wrote: > > > >> Matthew, > >> > >> Thanks for stating that, because I was almost beginning to imagine that > >> there was no essential disagreement, and that all of us agree that there > >> is more -- and less -- to language than what is found in people's > heads. > >> > >> Your position is the one I am familiar with from the functionalist point > >> of view, and I was beginning to feel that it was underrepresented on > >> Funknet. > >> > >> Those of us who disagree with your stated position -- but are very > >> familiar with it -- are interested not just in psycholinguistics and how > >> people process language -- but also in the communicative function of > >> language as a system whereby information is transferred. Just as you and > >> I may not be aware of the way our emails are encoded and then decoded by > >> the computers that help us send emails back and forth, speakers may be > >> compeltely unaware of what language does in order to transmit > information. > >> > >> After speakers have finished sending forth their linguistic output, it > >> matters not at all how they arrived at this output. Language processing > >> is separate from language in the same way that data processing is > >> separate from data. > >> > >> Best, > >> > >> --Aya > >> > >> > >> On Fri, 10 Sep 2010, dryer at buffalo.edu wrote: > >> > >>> > >>> The following sentence of Lise's > >>> > >>> "Since we don't in fact have such an agreed-on terminology, we have to > >>> be quite careful in making clear what we are referring to when we talk > >>> about 'the correct analysis' of a form like 'unhappiness'" > >>> > >>> suggests that there are two senses in which an analysis can be "the > >>> correct analysis", one in terms of what is in people's heads, the > other > >>> in terms of what is "out there". > >>> > >>> There are a vast number of patterns "out there". The only distinction, > >>> on my view, amongst this vast number of patterns, is between those > that > >>> reflect something inside people's heads and those that don't. But if > >>> that is the case, then there is no coherent sense in which one can > talk > >>> of "the correct analysis" of what is "out there", except in terms of > >>> what is in people's heads, and thus no second sense of "the correct > >>> analysis". The patterns that don't correspond to things in people's > >>> heads fall into (at least) two categories. There are those that are > >>> akin to constellations of stars and, as with constellations, there is > >>> no reality to these patterns, except in the minds of linguists. And > >>> there are those patterns which are the fossil remains of what was in > >>> the heads of speakers of an earlier stage of the language but which no > >>> longer are. These latter patterns are real, and they are relevant to > >>> exlaining why the language is now the way it is, but they are not > >>> relevant, I think many would agree, as to what is the "correct > analysis" > >>> of the language today. > >>> > >>> For this reason, I claim that the only sense in which an analysis can > be > >>> "the correct analysis" is in terms of what is inside of people's > heads. > >>> > >>> Again, I recommend the work of Bruce Derwing for lengthy discussion of > >>> these issues. > >>> > >>> Matthew > >>> > >>> --On Thursday, September 9, 2010 5:26 PM -0600 Lise Menn > >>> wrote: > >>> > >>>> I wish we had better terminology for keeping track of whether, at a > >>>> given time, we are talking about the patterns that are 'out there' in > >>>> the language and might possibly be apprehended (subconsciously) by a > >>>> speaker, and when we are talking about the patterns that a particular > >>>> speaker actually does apprehend, as indicated by experiments, from > >>>> simple 'wug tests' up to brain wave and eye-gaze studies. And for > >>>> distinguishing among the degrees of pattern apprehension that a person > >>>> may have, from vague preferences detectable in reaction times or other > >>>> behavior all the way up through clear metalinguistic insights. Dick > >>>> Hudson's note reminding us of the Gleitman and Gleitman study is right > >>>> on target. > >>>> > >>>> Since we don't in fact have such an agreed-on terminology, we have > to > >>>> be quite careful in making clear what we are referring to when we talk > >>>> about 'the correct analysis' of a form like 'unhappiness'. We know, > >>>> but tend to forget - and tend to forget to tell our students! - that > >>>> it's an empirical question as to whether the formal simplicity and > >>>> coherence of description of forms 'out there' (e.g. lovely abstract > >>>> morphophonemics) is any kind of approximation to the way knowledge of > >>>> the same forms is organized in a particular person's head. If we > >>>> remember that a very large proportion of what we know about our > >>>> language is 'out there' when we are infants and has to be internalized > >>>> through experience with the language (even if you believe in innate > >>>> 'core language'), the variation in internal knowledge from one person > >>>> to another is more understandable. We especially need to consider (and > >>>> try to test) the > >>>> possibility that > >>>> since > >>>> the brain can make multiple cross-connections, multiple patterns are > >>>> involved > >>>> simultaneously in morphological and syntactic analyses. I suggest that > >>>> that's the case with 'unhappiness' - and the linguistic analyses that > I > >>>> know about are not good at handling that kind of idea. > >>>> > >>>> Lise > >>>> > >>>> On Sep 9, 2010, at 8:00 AM, Matthew S. Dryer wrote: > >>>> > >>>>> > >>>>> Two comments. > >>>>> > >>>>> First (elaborating perhaps on Dick Hudson's comment), I think there > >>>>> is an > >>>>> important distinction between low-level linguistic intuitions (like > >>>>> whether a > >>>>> word or sentence is well-formed or what it means) and higher-level > >>>>> intuitions > >>>>> (like what the structure of a word or sentence is). One can take > >>>>> the position > >>>>> that we need to account for the former (while recognizing that they > >>>>> are not > >>>>> always reliable) but not the latter. > >>>>> > >>>>> Second, the tension here is not only between evidence from speaker > >>>>> intuitions > >>>>> versus evidence from psycholinguistic experiments. There is also a > >>>>> tension > >>>>> between deciding on the correct analysis on the basis of a priori > >>>>> simplicity > >>>>> arguments versus deciding on the correct analysis on the basis of > >>>>> psycholinguistic evidence (see Derwing 1973). The bracketing > >>>>> paradox that Dan > >>>>> referred to that arises with the word (semantics argues > >>>>> for > >>>>> [[un+happi] + er], morphology and phonology argues for [un + [happi > >>>>> +er]] (the > >>>>> comparative suffix can only be attached to adjectives containing one > >>>>> or two > >>>>> syllables) is only a paradox if one assumes that speakers adopt the > >>>>> simplest > >>>>> analysis. For example, if speakers adopt a more complex rule for > >>>>> either of these > >>>>> (e.g. perhaps the rule for attaching -er can apply exceptionally to > >>>>> trisyllabic > >>>>> words beginning with un-), then the bracketing paradox disappears. > >>>>> > >>>>> Matthew > >>>>> > >>>>> On Thu 09/09/10 8:16 AM , Richard Hudson dick at ling.ucl.ac.uk sent: > >>>>>> Thanks Dan. I'm sure you're right, and I'd be the first to agree > that > >>>>>> conscious judgements are only one kind of evidence that we need to > >>>>>> take > >>>>>> into account. I admire Carson Schutze's work (which I reviewed in > >>>>>> fact), > >>>>>> and of course I've been aware of complaints about judgements by > >>>>>> people > >>>>>> like Labov for decades. > >>>>>> > >>>>>> But you're missing my main point, which is that all judgements > aren't > >>>>>> equally reliable. If you want to know how /unhappiness/ is > >>>>>> structured, > >>>>>> ask a linguist, not a five-year old. And one of the by-products of > >>>>>> education may be increased sensitivity to syntax - which is one of > >>>>>> the > >>>>>> many reasons why linguists need to pay more attention to education. > >>>>>> > >>>>>> Best wishes, Dick > >>>>>> > >>>>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>>>> On 09/09/2010 11:39, Daniel Everett wrote: > >>>>>>> Dick, > >>>>>>> > >>>>>>> You raise an important issue here about > >>>>>> methodology. I believe that intuitions are a fine way to generate > >>>>>> hypotheses and even to test them - to a degree. But while it might > >>>>>> not have > >>>>>> been feasible for Huddleston, Pullum, and the other contributors to > >>>>>> the > >>>>>> Cambridge Grammar to conduct experiments on every point of the > >>>>>> grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, > >>>>>> corpora, and standard psycholinguistic experimentation (indeed, > >>>>>> Standard > >>>>>> Social Science Methodology) is vital for taking the field forward > >>>>>> and for > >>>>>> providing the best support for different analyses. Ted Gibson and Ev > >>>>>> Fedorenko have written a very useful new paper on this, showing > >>>>>> serious > >>>>>> shortcomings with intuitions as the sole source of evidence, in > their > >>>>>> paper: "The need for quantitative methods in syntax and semantics > >>>>>> research".> > >>>>>>> Carson Schutze and Wayne Cowart, among others, > >>>>>> have also written convincingly on this.> > >>>>>>> It is one reason that a team from Stanford, MIT > >>>>>> (Brain and Cognitive Science), and researchers from Brazil are > >>>>>> beginning a > >>>>>> third round of experimental work among the Pirahas, since my own > >>>>>> work on > >>>>>> the syntax was, like almost every other field researcher's, based > >>>>>> on native > >>>>>> speaker intuitions and corpora.> > >>>>>>> The discussion of methodologies reminds me of > >>>>>> the initial reactions to Greenberg's work on classifying the > >>>>>> languages of > >>>>>> the Americas. His methods were strongly (and justifiably) > criticized. > >>>>>> However, I always thought that his methods were a great way of > >>>>>> generating > >>>>>> hypotheses, so long as they were ultimately put to the test of > >>>>>> standard > >>>>>> historical linguistics methods. And the same seems true for use of > >>>>>> native-speaker intuitions.> > >>>>>>> -- Dan > >>>>>>> > >>>>>>>> We linguists can add a further layer of > >>>>>> explanation to the judgements, but some judgements do seem to be > more > >>>>>> reliable than others. And if we have to wait for psycholinguistic > >>>>>> evidence > >>>>>> for every detailed analysis we make, our whole discipline will > >>>>>> immediately > >>>>>> grind to a halt. Like it or not, native speaker judgements are what > >>>>>> put us > >>>>>> linguists ahead of the rest in handling fine detail. Imagine > >>>>>> writing the > >>>>>> Cambridge Grammar of the English Language (or the OED) without > >>>>>> using native > >>>>>> speaker judgements.>> > >>>>>>>> Best wishes, Dick Hudson > >>>>>>>> > >>>>>>> > >>>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>>> > >>>>> > >>>> > >>>> Lise Menn Home Office: 303-444-4274 > >>>> 1625 Mariposa Ave Fax: 303-413-0017 > >>>> Boulder CO 80302 > >>>> > >>>> Professor Emerita of Linguistics > >>>> Fellow, Institute of Cognitive Science > >>>> University of Colorado > >>>> > >>>> Secretary, AAAS Section Z [Linguistics] > >>>> > >>>> Campus Mail Address: > >>>> UCB 594, Institute of Cognitive Science > >>>> > >>>> Campus Physical Address: > >>>> CINC 234 > >>>> 1777 Exposition Ave, Boulder > >>>> > >>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> > >> > >> > > > > > > > > > > > > > ------------------------------ > > Message: 23 > Date: Sat, 11 Sep 2010 12:17:29 +0100 > From: "Chris Butler" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > To: "FUNKNET" > Message-ID: <24693E13135D4D6AAD5A098FAA952480 at OwnerPC> > Content-Type: text/plain; charset="iso-8859-1" > > Dick's comment that "this discussion raises the really fundamental question > of what kind of thing we think language is: social or individual" is, it > seems to me, an important one, particularly for those of us who are > committed functionalists. My own view is that a truly functional model of > language would be one which aims to account for how human beings communicate > using language, or in other words tries to answer the question which was > posed by Simon Dik a long time ago now, but which was not tackled head-on in > his own work: "How does the natural language user work?' In trying to answer > this question we need to accept that language is BOTH social AND individual, > and we need to explore both aspects to get as complete a picture as possible > of how we communicate using language. We need to know BOTH how people create > and respond to meanings and express those meanings in forms during social > interaction AND the mechanisms which operate in the brains of individuals in > order to make > such interaction possible. Both are important parts of the answer to the > question 'How do we communicate using language?', though this particular > thread of the Funknet discussion has concentrated more on the second aspect, > and so will I. > > This doesn't mean that all the work linguists have done on "exploring the > structure of a language so that I can understand how all the bits fit > together" and "exploring the connections between items", as Dick puts it, is > useless - far from it. After all, the hypotheses that psycholinguists test > are based on ideas about what languages are like. But it does mean, in my > view, that ultimately we need to get evidence that the constructs and > analyses we propose are ones that are at least consistent with what we know > of the processes which go on when we use language. So I am with Matthew > when he says that for him, "the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads". Of > course, this doesn't imply that linguists should just give up their jobs > until such time as we know everything there is to know about language > processing. But it does mean that we need to collaborate with > psycholinguists, psychologists and neurologists, > as has also been pointed out by linguists such as Ray Jackendoff, Asif > Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more > with sociolinguists and sociologists, so that we can get a better handle on > the sociocultural aspects of how we communicate.] And it also means that > psycholinguists, for their part, need whenever possible to follow up tightly > controlled lab experiments with studies under more naturalistic conditions, > to avoid the criticism that what happens in artifical lab situations may not > happen in natural communicative conditions. > > I also agree with Dick when he says that "the differences between > individuals really matter", and with Lise when she points out that "we must > also be careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover". However, there > are surely processing mechanisms which are common to all language users by > virtue of the evolution of the language faculty and which constitute the > "general processes" which Dick says psycholinguists are interested in. > > On the issue of quantitative methodology, I'm sympathetic in general to Ted > and Ev's views, though it does seem sensible to prioritise cases in terms of > a hierarchy such as Brian suggests. One thing this means is that we should > be giving our university students of linguistics (and some of our > linguistics lecturers!) courses in quantitative aspects of linguistics that > introduce them to the use of at least some of the basic statistical methods > in language study, and I'm sure this is indeed going on in some enlightened > places. To those who suspect this can't be done with maths-shy students who > don't initially see the need for it, I offer my own experience, over quite a > long period, of teaching such courses to people with little or no prior > experience in quantitative techniques. For some years in the 1990s, I taught > such courses to all linguistics students in an institution where we had many > mature students who had come into university level studies with non-standard > qualificatio > ns, and were not well equipped for courses of this kind by their previous > experience. I'm glad to say that teaching the subject from their own > perspective as language students rather than that of the statistician, and > explaining the reasons for doing things in particular ways rather than just > presenting formulae, paid off in the end, so that most students were able to > appreciate the relevance of these courses and to turn in very creditable > projects showing an understanding of research design and competence in the > use of a range of basic statistical techniques. And I still find that bright > graduate students respond well to similar courses which incorporate some of > the rather more advanced techniques needed for many real research projects > in various areas of linguistics. But I may well be out of date with what is > now already happening in our fine institutions of higher education! > > Chris Butler > > > ------------------------------ > > Message: 24 > Date: Sat, 11 Sep 2010 07:18:29 -0700 (PDT) > From: "A. Katz" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > To: Chris Butler > Cc: FUNKNET > Message-ID: > Content-Type: text/plain; charset="us-ascii"; Format="flowed" > > The problem is that once we achieve the psycholinguistic goal and see what > is happening in each person's head, and we see that conflicting analyses > are the norm, rather than the exception, among normal language speakers, > then we will realize that the way language works to transmit information > is despite individual differences and not because of uniform processing > strategies. > > Even when all are agreed as to the meaning of an utterance, they do not > process it the same way. Which means that processing is seocndary to > information transmission. > > --Aya > > > > > On Sat, 11 Sep 2010, Chris Butler wrote: > > > Dick's comment that "this discussion raises the really fundamental > question of what kind of thing we think language is: social or individual" > is, it seems to me, an important one, particularly for those of us who are > committed functionalists. My own view is that a truly functional model of > language would be one which aims to account for how human beings communicate > using language, or in other words tries to answer the question which was > posed by Simon Dik a long time ago now, but which was not tackled head-on in > his own work: "How does the natural language user work?' In trying to answer > this question we need to accept that language is BOTH social AND individual, > and we need to explore both aspects to get as complete a picture as possible > of how we communicate using language. We need to know BOTH how people create > and respond to meanings and express those meanings in forms during social > interaction AND the mechanisms which operate in the brains of individuals in > order to ma > ke such interaction possible. Both are important parts of the answer to > the question 'How do we communicate using language?', though this particular > thread of the Funknet discussion has concentrated more on the second aspect, > and so will I. > > > > This doesn't mean that all the work linguists have done on "exploring the > structure of a language so that I can understand how all the bits fit > together" and "exploring the connections between items", as Dick puts it, is > useless - far from it. After all, the hypotheses that psycholinguists test > are based on ideas about what languages are like. But it does mean, in my > view, that ultimately we need to get evidence that the constructs and > analyses we propose are ones that are at least consistent with what we know > of the processes which go on when we use language. So I am with Matthew > when he says that for him, "the only sense in which an analysis can be "the > correct analysis" is in terms of what is inside of people's heads". Of > course, this doesn't imply that linguists should just give up their jobs > until such time as we know everything there is to know about language > processing. But it does mean that we need to collaborate with > psycholinguists, psychologists and neurologists > , as has also been pointed out by linguists such as Ray Jackendoff, Asif > Agha, Ewa Dabrowska and Jan Nuyts. [We also need to collaborate much more > with sociolinguists and sociologists, so that we can get a better handle on > the sociocultural aspects of how we communicate.] And it also means that > psycholinguists, for their part, need whenever possible to follow up tightly > controlled lab experiments with studies under more naturalistic conditions, > to avoid the criticism that what happens in artifical lab situations may not > happen in natural communicative conditions. > > > > I also agree with Dick when he says that "the differences between > individuals really matter", and with Lise when she points out that "we must > also be careful not to idealize "what's in people's heads" as if it were a > single coherent construct that we are trying to discover". However, there > are surely processing mechanisms which are common to all language users by > virtue of the evolution of the language faculty and which constitute the > "general processes" which Dick says psycholinguists are interested in. > > > > On the issue of quantitative methodology, I'm sympathetic in general to > Ted and Ev's views, though it does seem sensible to prioritise cases in > terms of a hierarchy such as Brian suggests. One thing this means is that we > should be giving our university students of linguistics (and some of our > linguistics lecturers!) courses in quantitative aspects of linguistics that > introduce them to the use of at least some of the basic statistical methods > in language study, and I'm sure this is indeed going on in some enlightened > places. To those who suspect this can't be done with maths-shy students who > don't initially see the need for it, I offer my own experience, over quite a > long period, of teaching such courses to people with little or no prior > experience in quantitative techniques. For some years in the 1990s, I taught > such courses to all linguistics students in an institution where we had many > mature students who had come into university level studies with non-standard > qualificat > ions, and were not well equipped for courses of this kind by their > previous experience. I'm glad to say that teaching the subject from their > own perspective as language students rather than that of the statistician, > and explaining the reasons for doing things in particular ways rather than > just presenting formulae, paid off in the end, so that most students were > able to appreciate the relevance of these courses and to turn in very > creditable projects showing an understanding of research design and > competence in the use of a range of basic statistical techniques. And I > still find that bright graduate students respond well to similar courses > which incorporate some of the rather more advanced techniques needed for > many real research projects in various areas of linguistics. But I may well > be out of date with what is now already happening in our fine institutions > of higher education! > > > > Chris Butler > > > > > > ------------------------------ > > Message: 25 > Date: Sat, 11 Sep 2010 17:54:58 +0300 > From: "E.G." > Subject: Re: [FUNKNET] "Relative clauses" with no relativized ele > To: funknet at mailman.rice.edu > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Hi all, > > I'd like to thank Ron for pointing out the incompleteness of what I wrote, > and for the reference to his article. I hope I didn't give the impression I > was trying to give a complete description of noun complementation in Modern > Hebrew in an email. > > What I *was* trying to say is that in some languages, unlike English, some > nominalizations (of utterance and cognition verbs, as Ron points out, but > also of some perception verbs too) can occur with a construction that is > explicitly and unmistakably marked as a complement clause. Moreover, the > nominalizations that take these explicit complement clauses are related to > verbs that can take the same type of complement clause. In such languages, > then, it's pretty clear that these instances involve complement clauses. It > doesn't mean that other noun complementation strategies don't exist for > other types of nouns. > > However, my main point was more general, albeit poorly expressed. It's that > we can turn to cross-linguistic comparison in order to try to reach > generalizations about how languages encode meaning. These generalizations > are useful, because they can be used to ask "why" questions. For example, > it's not really possible to ask *why* Hebrew has two distinct strategies > for > noun complementation, how *why* English has one. That's because it could > always be otherwise, and language change can alter the picture (and has!). > However, if we find that in languages that have two strategies, one is > limited to nominalizations of PCU verbs, then we have the beginnings of a > hierarchy that is amenable to functional explanation. > > Anyway, it seems that Thomas Payne's question has turned up a pretty > general > consensus that these constructions are complement clauses. > > Best wishes, > Eitan > > > > > > > On 10 September 2010 23:26, Ron Kuzar wrote: > > > The Modern Hebrew data supplied by Eitan are incomplete. > > Hebrew distinguishes between locution (say, hear, think, etc.) and > > situation (action, event, state, etc.). > > What Eitan describes is only true with regard to nouns (and clauses) > > expressing locution. 'Announcement' is indeed such a noun. > > Words such as ba'ya 'problem', macav 'situation', or cara 'trouble', > > etc., whose denotatum is a situation, cannot be followed by ki, but only > > by Se-, e.g.: > > > > margiz oti ha-macav Se-kulam halxu (*ki kulam halxu) > > annoys me the-situation that-all went > > 'I am upset about the situation that all have gone' > > > > On the other hand, the relative Se- may be replaced by the more > > elegant and classical aSer, while the Se- of situation clauses may not. > > Sorry about the invented example. I am overseas now. > > All this has been described (with corpus data) in: > > > > Kuzar, Ron. 1993. Nominalization Clauses in Israeli Hebrew. Balshanut > Ivrit > > [Hebrew > > Linguistics] 36: 71-89 [unfortunately available only in Hebrew]. > > > > The article is somewhat outdated and contains some inaccuracies I would > > formulate differently today, but the basic distinction is valid in my > > opinion. > > Best, > > Ron Kuzar > > --------------- > > On Fri, Sep 10, 2010 at 8:54 PM, E.G. wrote: > > > > > Hi all, > > > > > > I'd agree with Arie Verhagen. But there's a way that cross-linguistic > > > comparison can help what seems to be a purely theoretical question > based > > on > > > a single language. The problem here is that English uses the same > element > > > to > > > mark regular relatives and these "appositional" relatives. But if at > > least > > > one language encodes them by different means, then there's at least a > > good > > > case for seeing them as distinct functions. It's basically the same > > > principle that's used to decide whether to put a meaning on a semantic > > map. > > > So here are two languages that I know that encode them differently. > > > > > > In Modern Hebrew, these clauses can be encoded as a dedicated > complement > > > clause (ki), which differs from the relative clause marker (Se-), e.g. > > > > > > ha-hoda'a Se-kibalnu > > > the-announcment rel-we_got > > > "The announcement that we got." > > > > > > ha-hoda'a ki hitbatel ha-mifgaS > > > the-message CMP was_cancelled the-meeting > > > "The announcement that the meeting was cancelled." > > > > > > In Coptic, these clauses are marked by ce-, which marks complement > > clauses, > > > *inter alia*, but not relative clauses: > > > > > > ph-mewi ce- (complement clause) > > > 'the-thought that (we are angry)' > > > > > > ph-mewi ete- (relative clause) > > > 'the thought that (we used to think)' > > > > > > This seems to be a pretty clear indication that these are complement > > > clauses > > > rather than relatives. Even if one doesn't like the notion of nouns > > taking > > > complement clauses (and why not? nominalizations in some languages can > > take > > > accusative modifiers as well as genitives), it still probably isn't > > > incidental that the nominalizations are of verbs that take complement > > > clauses when finite. > > > > > > As usual, the perspective in Talmy Giv?n's *Syntax* (vol. 2) is worth > > > looking at. > > > > > > Best, > > > Eitan > > > > > > > > > On 10 September 2010 19:21, Arie Verhagen > > > wrote: > > > > > > > And as another addition: the clauses that can only be introduced by > > > *that* > > > > (with no role to > > > > play in the subordinate clause) may be seen as (subtypes of) > complement > > > > clauses, > > > > expressing a proposition with the noun functioning as Complement > Taking > > > > Predicate (CTP), > > > > expressing a propostional attitude, epistemic/evaluative stance, etc. > > > > (following analyses by > > > > Thompson, Diessel, Langacker, myself, and others), i.e. not > relatives. > > > Cf. > > > > constructions like > > > > "The claim is that X" (traditionally analysed as subject clauses), "I > > > claim > > > > that X", "I put forward > > > > the claim that X", in which the relationship between the verb or noun > > and > > > > the that-clause is > > > > comparable to the one in "The claim that X". > > > > > > > > --Arie Verhagen > > > > > > > > ---------------- > > > > Message from Rong Chen > > > > 10 Sep 2010, 23:42 > > > > Subject: Re: [FUNKNET] "Relative clauses" with no relativi > > > > > > > > > To add to Joanne's comments: > > > > > > > > > > There are basically three ways to distinguish an appositive clause > > > > > (AC) from a relative clause (RC). > > > > > > > > > > 1) An AC can only be led by *that* while an RC can be led by other > > > > > pronouns. > > > > > > > > > > 2) The AC and the noun it modifies display an equative > > > relationship--one > > > > can say X > > > > > (denoted by the noun) is Y (presented by the appositive)--while an > RC > > > > often doesn't > > > > > (except, perhaps, when the relative clause is sentential). > > > > > > > > > 3)--which Tom noted--*that* is not part of the clause in an AC; but > a > > > > relative pronoun > > > > > is always part of the clause in an RC. > > > > > > > > > > Rong Chen > > > > > > > > > > > > > > > > > > > > > > -- > > > Eitan Grossman > > > Martin Buber Society of Fellows > > > Hebrew University of Jerusalem > > > > > > > > > > > -- > > =============================================== > > Dr. Ron Kuzar > > Address: Department of English Language and Literature > > University of Haifa > > IL-31905 Haifa, Israel > > Office: +972-4-824-9826, Fax: +972-4-824-9711 > > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > > Home fax: 153-77-481-9676 (only from Israel) > > Email: kuzar at research.haifa.ac.il > > Homepage: http://research.haifa.ac.il/~kuzar > > =============================================== > > > > > > -- > Eitan Grossman > Martin Buber Society of Fellows > Hebrew University of Jerusalem > > > ------------------------------ > > Message: 26 > Date: Sat, 11 Sep 2010 10:19:05 -0500 > From: Kristine Hildebrandt > Subject: [FUNKNET] Job Advertisement > To: funknet at mailman.rice.edu > Message-ID: > > > > Content-Type: text/plain; charset=windows-1252 > > Dear Funknetters: > > > Please distribute this job advertisement to interested colleagues and > doctoral students completing their degrees. > > > *HIRING UNIT*: *DEPARTMENT OF ENGLISH LANGUAGE AND LITERATURE* > > ** > > *TITLE/RANK*: Assistant Professor of English-Linguistics > > > *DESCRIPTION OF DUTIES*: The Department of English Language and > Literature > invites applications for a tenure-track position in general linguistics, > with secondary specialization in applied linguistics. The candidate will > teach courses in the MA TESL program, along with undergraduate courses in > linguistics, composition (ESL and regular), and some general education > courses. Academic year: 3/3 load. > > *TERMS OF APPOINTMENT*: Academic, tenure-track beginning August 16, 2011, > 100% appointment. > > *SOURCE OF FUNDS*: State > > > *SALARY RANGE*: commensurate with training and experience > > > *QUALIFICATIONS REQUIRED: *A Ph.D. in Linguistics, Applied Linguistics, or > related field required. If Ph.D. is not completed by the beginning of the > contract period, appointment will be at the rank of Instructor until all > degree requirements are fulfilled. A record of ESL and/or TESL experience > is desirable. > > > *CLOSING DATE FOR APPLICATIONS*: Position open until filled; completed > applications postmarked by November 15, 2010 will have priority. Possible > interviews at LSA in January 2011. > * ** > **SEND COVER LETTER, VITA, UNOFFICIAL TRANSCRIPT, STATEMENT OF TEACHING > PHILOSOPHY AND RESEARCH AGENDA, AND THREE LETTERS OF RECOMMENDATION TO:** > * > Linguistics Search Committee > Southern Illinois University Edwardsville > Department of English Language & Literature > Campus Box 1431 > Edwardsville, IL 62026-1431 > NOTE: Electronic applications will not be accepted for this position. > > > SIUE is a state university-benefits under state sponsored plans may not be > available to holders of F1 or J1 visas. Applicants may be subject to a > background check prior to an offer of employment. SIUE is an affirmative > action and equal opportunity employer. The SIUE ANNUAL SECURITY REPORT is > available on-line at: http://admin.siu.edu/studentrightto/. The report > contains safety and security information and crime statistics for the past > three (3) calendar years. This report is published in compliance with > Federal law, entitled *the ?Jeanne Clery Disclosure of Campus Security > Policy and Campus Crime Statistics Act.? * You may also access this report > through the SIUE Home Page: http://www.siue.edu under *Ready References, > Quick Links or Publications/Reports*. For those without computer access, a > paper copy of the report may be obtained from the Office of the Vice > Chancellor for Administration, Rendleman Hall, Room 2228. > > > -- > *Kristine A. Hildebrandt* > *Assistant Professor, Department of English Language & Literature > Southern Illinois University Edwardsville* > *Box 1431 > Edwardsville, IL 62026 U.S.A. > 618-650-3380 (office)* > *khildeb at siue.edu > http://www.siue.edu/~khildeb* > > > ------------------------------ > > Message: 27 > Date: Sat, 11 Sep 2010 11:45:27 -0400 > From: Ted Gibson > Subject: Re: [FUNKNET] analysis: unhappiness > To: funknet > Cc: Richard Hudson , Daniel Everett > , Evelina Fedorenko > Message-ID: <65E6B9DA-7FB1-4240-927B-C7141F6A55C9 at MIT.EDU> > Content-Type: text/plain; charset=WINDOWS-1252; format=flowed; > delsp=yes > > Dear Brian, Dick, Dan et al: > > Thanks for the discussion. Here are a few responses: > > 1. Brian: > > "But I understand Dick's worry about how far Gibson and Fedorenko are > trying to push this. Neither their email nor their paper sets clear > limits on what we should be testing and we certainly don't want to > waste time checking out go-goed-went. So, Gibson and Fedorenko owe > us those clarifications." > > The answer that we give to this question in Gibson & Fedorenko (in > press) is as follows (the final paragraph in the paper): > > "Finally, a question that is often put to us is whether it is > necessary to evaluate every empirical claim quantitatively. A major > problem with the fields of syntax and semantics is that many papers > include no quantitative evidence in support of their research > hypotheses. Because conducting experiments is now so easy to do with > the advent of Amazon.com?s Mechanical Turk, we recommend gathering > quantitative evidence for all empirical claims. However, it would > clearly be a vast improvement to the field for all research papers to > include at least some quantitative evidence evaluating their research > hypotheses." > > Another possible answer to this question is: the more important some > observation is, the better your evidence should be. If the > observation is a key reason for some important theoretical claim, then > there should be solid quantitative data supporting that observation. > > In practice, once a linguist starts gathering quantitative data, s/he > will realize (a) how easy it is to do; and (b) how beneficial the > methods are, with the consequence that these researchers will probably > do most or all of their work quantitatively in the future. > > > 2. Dick (by the way, thank you for the kind responses, and your > positive tone): > > "Your [the psycholinguists'] goal is to find general processes and > principles that apply uniformly across individuals, so you have to use > methods to check for generality." > > in contrast to "my focus is on items and structures, and I start from > the assumption that these can and do vary across speakers." > > Many cognitive psychologists / cognitive scientists (all the ones I > know at MIT for example) are interested in both cognitive > generalizations across people and ways in which people differ > cognitively. In fact, some methods (e.g., the individual differences > approach where co-variation of various behaviors / characteristics is > examined across individuals) have been specifically developed to study > differences among individuals. Both kinds of data are important for > understanding human cognition, including language. This applies to > language research directly: generalizations across people are > important, but so are individual differences. In either case, > quantitative data are necessary to evaluate research questions and > test hypotheses. > > On a related note, it is a mistake to characterize researchers with a > background in "psychology" or cognitive science as being interested in > "processing", and researchers with a background in "linguistics" as > being interested in "knowledge" or "representation / structure". Both > psychologists and linguists should be interested in *both* > representation and processing (and learning, for that matter). We > wrote a little about this confusion in Gibson & Fedorenko (in press), > which we include at the end of the message. > > This leads to something that Dan said: > > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > Both linguistics and psychology are big fields. We assume Dan is > referring to cognitive psychology / cognitive science here. (Of > course, there are sub-fields of psychology - e.g., personality > psychology or abnormal psychology - which are somewhat distinct from > linguistics, but those sub-fields are also distinct from cognitive > psychology.) It is true that historically linguistics is not treated > as a subfield of cognitive psychology / cognitive science. However, > key research questions in linguistics (i.e., the form of the knowledge > structures and algorithms underlying human language) are indeed a > subset of those investigated by cognitive psychologists / cognitive > scientists. We think that the biggest factor separating linguistics > from psychology is the methods used to explore the research questions, > rather than the research questions themselves. Consequently, we > would like to continue to see tighter connections among the fields of > psychology / cognitive science, linguistics, as well as other fields > like anthropology and computer science. > > Thanks to all for the interesting discussions. > > Ted & Ev > > We have encountered a claim that the reason for different kinds of > methods being used across the different fields of language study > (i.e., in linguistics vs. psycho-/neuro-linguistics) is that the > research questions are different across these fields, and some methods > may be better suited to ask some questions than others. Although the > latter is likely true, the premise ? that the research questions are > different across the fields ? is false. The typical claim is that > researchers in the field of linguistics are investigating linguistic > representations, and researchers in the fields of psycho-/neuro- > linguistics are investigating the computations that take place as > language is understood or produced. However, many researchers in the > fields of psycho-/neuro-linguistics are also interested in the nature > of the linguistic representations (at all levels; e.g., phonological > representations, lexical representations, syntactic representations, > etc.) [1]. By the same token, many researchers in the field of > linguistics are interested in the computations that take place in the > course of online comprehension or production. However, inferences ? > drawn from any dependent measure ? about either the linguistic > representations or computations are always indirect. And these > inferences are no more indirect in reading times or event-related > potentials, etc., than in acceptability judgments: across all > dependent measures we take some observable (e.g., a participant?s > rating on an acceptability judgment task or the time it took a > participant to read a sentence) and we try to infer something about > the underlying cognitive representations / processes. More generally, > methods in cognitive science are often used to jointly learn about > representations and computations, because inferences about > representations can inform questions about the computations, and vice > versa. For example, certain data structures can make a computation > more or less difficult to perform, and certain representations may > require assumptions about the algorithms being used. > > In our opinion then, the distinction between the fields of linguistics > and psycho-/neuro-linguistics is purely along the lines of the kind of > data that are used as evidence for or against theoretical hypotheses: > typically non-quantitative data in linguistics vs. typically > quantitative data in psycho-/neuro-linguistics. Given the superficial > nature of this distinction, we think that there should be one field of > language study where a wide range of dependent measures is used to > investigate linguistic representations and computations. > > > [1] In fact, some methods in cognitive science and cognitive > neuroscience were specifically developed to get at representational > questions (e.g., lexical / syntactic priming methods, neural > adaptation or multi-voxel pattern analyses in functional MRI). > > > On Sep 10, 2010, at 9:05 PM, Daniel Everett wrote: > > > I think that Brian and Dick make excellent points. There are very > > good grammars written that could be improved by psycholinguistic > > experimentation and more quantitative approaches. But large sections > > of those grammars aren't going to change one bit (go-went) with > > quantitative tests and such tests would be completely > > counterproductive given the shortness of life and the vastness of > > the field linguist's tasks. > > > > Part of the problem is that linguistics is not simply a > > subdiscipline of psychology. Linguistics has its own objectives and > > those only occasionally overlap with psychology. The same for methods. > > > > On another note, I don't buy the 'in my head' 'out of my head' > > distinction either (that Matt seems to be urging upon us). We study > > different things and have different reasons for being satisfied with > > the results we achieve. > > > > I believe that we linguists are often complacent and fail to apply > > better methods. But of course that applies to all disciplines. > > > > In the meantime, checking corpora, collecting data as a result of > > careful interviews with native speakers, and the other aspects of > > the field linguist's task are vital parts of the linguist's task and > > much of this won't be improved by quantitative methods as we > > currently understand them. Maybe sometime. > > > > Dan > > > > P.S. In my original reference to Ted and Ev's paper, I said that > > they showed the danger of using intuitions. What I meant to say of > > using intuitions as standardly used by linguists. They convinced me > > that there is a lot to learn from quantitative methods. > > > > On 10 Sep 2010, at 19:40, Richard Hudson wrote: > > > >> Dear Ted and Ev, > >> Yes, I understand your view, but I think it's a psycholinguist's > >> view. Your goal is to find general processes and principles that > >> apply uniformly across individuals, so you have to use methods to > >> check for generality. And (as you know) I admire the way you pursue > >> that goal. But my goal, as a linguist, is different. I want to > >> explore the structure of a language so that I can understand how > >> all the bits fit together. Like you, I'm aiming to model cognition, > >> but my focus is on items and structures, and I start from the > >> assumption that these can and do vary across speakers. > >> > >> However, having said all that I do agree with you that linguists > >> should all get used to collecting and using quantitative data; and > >> with the help of Brian MacWhinney's typology we'd know what methods > >> to use when. And I do agree with your points about bid/bidded: in > >> cases like that, quantitative data would be at least a very good > >> starting point for a proper investigation. > >> > >> Best wishes, Dick > >> > >> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >> > >> On 10/09/2010 19:30, Ted Gibson wrote: > >>> Dear Dick: > >>> > >>> Perhaps we are talking at cross purposes. I don't understand what > >>> is confusing about what Ev Fedorenko and I are claiming. All we > >>> are saying is that if you have some testable claim involving a > >>> general hypothesis about a language, then you need to get > >>> quantitative data from unbiased sources to evaluate that claim. If > >>> you are interested in English past tense morphology, then > >>> depending on the claims that you might want to investigate, there > >>> are lots of ways to get relevant quantitative evidence. Corpus > >>> data will probably be useful. For very low frequency words, you > >>> can run experiments to test behavior with respect to such words. > >>> > >>> Your example of the past tense of "bid" is a fine such example. > >>> You can run an experiment like the one you suggested to find out > >>> what people think the past tense is. If you then found that 20/50 > >>> people responded "bidded" and 30/50 respond "bid", that is a lot > >>> of useful information. As you suggest in your discussion, this > >>> result wouldn't answer the question of how past tense is stored in > >>> each individual. This result would be ambiguous among several > >>> possible explanations. One possibility is that the probability > >>> distribution that is being discovered reflects different dialects, > >>> such that 2/5 of the population has one past tense, and 3/5 has > >>> another. Another possibility is that each person has a similar > >>> probability distribution in their heads, such that 2/5 of the time > >>> I respond one way, and 3/5 of the time I respond another. Further > >>> experiments would be necessary to answer between these and other > >>> possible theories (e.g., with repeated trials from the same > >>> person, carefully planned so that the participants don't notice > >>> that they are being asked multiple times). Without the > >>> quantitative evidence in the first place, there is no way to > >>> answer these kinds of questions. > >>> > >>> Regarding the past tense of "go", this would be useful as a > >>> baseline in an experiment involving the less frequent ones. So, > >>> yes, it would useful to gather quantitative evidence in such a > >>> case also, as baselines with respect to the more interesting cases > >>> for theories. > >>> > >>> The bottom line: if you have a generalization about a language > >>> that you wish to evaluate (such that you hypothesize that it is > >>> true across the speakers of the language), then you need > >>> quantitative evidence from multiple individuals, using an unbiased > >>> data collection method, to evaluate such a claim. The point about > >>> Mechanical Turk is that it is really *easy* to do this now, at > >>> least for languages like English. > >>> > >>> Best wishes, > >>> > >>> Ted Gibson & Ev Fedorenko > >>> > >>> On Sep 10, 2010, at 1:59 PM, Richard Hudson wrote: > >>> > >>>> Dear Ted, > >>>> Thanks for the very interesting comment, but are you REALLY > >>>> saying that I shouldn't claim, for example, that the past tense > >>>> of GO is "went" without first cross-checking with 50 native > >>>> speakers? > >>>> > >>>> Isn't there a danger of missing the point that we all, as native > >>>> speakers, spend our whole lives scanning other people's > >>>> linguistic behaviour (language 'out there', E-language) and > >>>> trying to explain it to ourselves in terms of a language system > >>>> (language 'in here', I-language)? So every judgement we make is > >>>> based on thousands or millions of observed exemplars, and > >>>> reflects a unique experience of E-language filtered through a > >>>> unique I-language. > >>>> > >>>> Given that view of language development, I don't see how > >>>> quantitative data will help. Let's take a real uncertainty, such > >>>> as the past tense of BID. If I want to say I did it, do I say "I > >>>> bidded" or "I bid"? My judgement: I don't know. Ok, you get 50 > >>>> people to oblige on Mechanical Turk, and 20 of them give "bidded" > >>>> and 30 "bid". So what? Does that mean that the correct answer is > >>>> "bidded"? Surely not. How is it better than my judgement? I agree > >>>> you could record my speech and find how often I use each > >>>> alternative; but the reason I don't know is precisely because > >>>> it's a rare word, so in a sense quantitative data are irrelevant > >>>> even there. What would solve the problem of subjectivity, of > >>>> course, would be a machine for probing the bit of my mind (or > >>>> even brain) that holds BID and its details; but I suspect that > >>>> even that wouldn't move us much further forward than my original > >>>> "don't know". (Incidentally I write as a fan of quantitative > >>>> sociolinguistics, so I do accept that quantitative data are > >>>> relevant to linguistic analysis in some areas, where the I- > >>>> language phenomenon is frequent enough to produce usable data.) > >>>> > >>>> It seems to me that this discussion raises the really fundamental > >>>> question of what kind of thing we think language is: social or > >>>> individual. The problem isn't unique to linguistics of course; > >>>> it's the same throughout the social sciences. But what's special > >>>> about linguistics is that we deal in very fine details of culture > >>>> (e.g. details of how a particular word is used or pronounced) so > >>>> the differences between individuals really matter. I don't see > >>>> that we're ever going to have anything better than judgements to > >>>> go on, so what we need is a way to ensure that judgements are > >>>> accurate reports of individual I-language. A rotten situation for > >>>> a science, but I don't see how it can get better. > >>>> > >>>> Dick > >>>> > >>>> Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm > >>>> > >>>> On 10/09/2010 14:03, Ted Gibson wrote: > >>>>> Dear Dan, Dick: > >>>>> > >>>>> I would like to clarify some points that Dan Everett makes, in > >>>>> response to Dick Hudson. > >>>>> > >>>>> Ev Fedorenko and I have written a couple of papers recently > >>>>> (Gibson & > >>>>> Fedorenko, 2010, in press, see references and links below) on > >>>>> what we > >>>>> think are weak methodological standards in syntax and semantics > >>>>> research over the past many years. The issue that we address is > >>>>> the > >>>>> prevalent method in syntax and semantics research, which involves > >>>>> obtaining a judgment of the acceptability of a sentence / meaning > >>>>> pair, typically by just the author of the paper, sometimes with > >>>>> feedback from colleagues. As we address in our papers, this > >>>>> methodology does not allow proper testing of scientific hypotheses > >>>>> because of (a) the small number of experimental participants > >>>>> (typically one); (b) the small number of experimental stimuli > >>>>> (typically one); (c) cognitive biases on the part of the > >>>>> researcher > >>>>> and participants; and (d) the effect of the preceding context > >>>>> (e.g., > >>>>> other constructions the researcher may have been recently > >>>>> considering). (As Dan said, see Schutze, 1996; Cowart, 1997; and > >>>>> several others cited in Gibson & Fedorenko, in press; for similar > >>>>> points, but with not as strong a conclusion as ours). > >>>>> > >>>>> Three issues need to be separated here: (1) the use of intuitive > >>>>> judgments as a dependent measure in a language experiment; (2) > >>>>> potential cognitive biases on the part of experimental subjects > >>>>> and > >>>>> experimenters in language experiments; and (3) the need for > >>>>> obtaining > >>>>> quantitative evidence, whatever the dependent measure might be. > >>>>> The > >>>>> paper that Ev and I wrote addresses the last two issues, but > >>>>> does not > >>>>> go into depth on the first issue (the use of intuitions as a > >>>>> dependent > >>>>> measure in language experiments). Regarding this issue, we don't > >>>>> think > >>>>> that there is anything wrong with gathering intuitive judgments > >>>>> as a > >>>>> dependent measure, as long as the task is clear to the > >>>>> experimental > >>>>> participants. > >>>>> > >>>>> In the longer paper (Gibson & Fedorenko, in press) we respond to > >>>>> some > >>>>> arguments that have been given in support of continuing to use the > >>>>> traditional non-quantitative method in syntax / semantics > >>>>> research. > >>>>> One recent defense of the traditional method comes from Phillips > >>>>> (2008), who argues that no harm has come from the non-quantitative > >>>>> approach in syntax research thus far. Phillips argues that there > >>>>> are > >>>>> no cases in the literature where an incorrect intuitive judgment > >>>>> has > >>>>> become the basis for a widely accepted generalization or an > >>>>> important > >>>>> theoretical claim. He therefore concludes that there is no > >>>>> reason to > >>>>> adopt more rigorous data collection standards. We challenge > >>>>> Philips? > >>>>> conclusion by presenting three cases from the literature where a > >>>>> faulty intuition has led to incorrect generalizations and mistaken > >>>>> theorizing, plausibly due to cognitive biases on the part of the > >>>>> researchers. > >>>>> > >>>>> A second argument that is sometimes presented for the continued > >>>>> use of > >>>>> the traditional non-quantitative method is that it would be too > >>>>> inefficient to evaluate every syntactic / semantic hypothesis or > >>>>> phenomenon quantitatively. For example, Culicover & Jackendoff > >>>>> (2010) > >>>>> make this argument explicitly in their response to Gibson & > >>>>> Fedorenko > >>>>> (2010): ?It would cripple linguistic investigation if it were > >>>>> required > >>>>> that all judgments of ambiguity and grammaticality be subject to > >>>>> statistically rigorous experiments on naive subjects, especially > >>>>> when > >>>>> investigating languages whose speakers are hard to > >>>>> access? (Culicover > >>>>> & Jackendoff, 2010, p. 234). (Dick Hudson makes a similar point > >>>>> earlier in the discussion here.) Whereas we agree that in > >>>>> circumstances where gathering data is difficult, some evidence is > >>>>> better than no evidence, we do not agree that research would be > >>>>> slowed > >>>>> with respect to languages where experimental participants are > >>>>> easy to > >>>>> access, such as English. In contrast, we think that the opposite > >>>>> is > >>>>> true: the field?s progress is probably slowed by not doing > >>>>> quantitative research. > >>>>> Suppose that a typical syntax / semantics paper that lacks > >>>>> quantitative evidence includes judgments for 50 or more > >>>>> sentences / > >>>>> meaning pairs, corresponding to 50 or more empirical claims. > >>>>> Even if > >>>>> most of the judgments from such a paper are correct or are on the > >>>>> right track, the problem is in knowing which judgments are > >>>>> correct. > >>>>> For example, suppose that 90% of the judgments from an arbitrary > >>>>> paper > >>>>> are correct (which is probably a high estimate). (Colin Phillips > >>>>> and > >>>>> some of his former students / postdocs have commented to us > >>>>> that, in > >>>>> their experience, quantitative acceptability judgment studies > >>>>> almost > >>>>> always validate the claim(s) in the literature. This is not our > >>>>> experience, however. Most experiments that we have run which > >>>>> attempt > >>>>> to test some syntactic / semantic hypothesis in the literature > >>>>> end up > >>>>> providing us with a pattern of data that had not been known > >>>>> before the > >>>>> experiment (e.g., Breen et al., in press; Fedorenko & Gibson, in > >>>>> press; Patel et al., 2009; Scontras & Gibson, submitted).) This > >>>>> means > >>>>> that in a paper with 50 empirical claims 45/50 are correct. But > >>>>> which > >>>>> 45? There are 2,118, 760 ways to choose 45 items from 50. That?s > >>>>> over > >>>>> two million different theories. By quantitatively evaluating the > >>>>> empirical claims, we reduce the uncertainty a great deal. To make > >>>>> progress, it is better to have theoretical claims supported by > >>>>> solid > >>>>> quantitative evidence, so that even if the interpretation of the > >>>>> data > >>>>> changes over time as new evidence becomes available ? as is > >>>>> often the > >>>>> case in any field of science ? the empirical pattern can be used > >>>>> as a > >>>>> basis for further theorizing. > >>>>> > >>>>> Furthermore, it is no longer expensive to run behavioral > >>>>> experiments, > >>>>> at least in English and other widely spoken languages. There now > >>>>> exists a marketplace interface ? Amazon.com?s Mechanical Turk ? > >>>>> which > >>>>> can be used for collecting behavioral data over the internet > >>>>> quickly > >>>>> and inexpensively. The cost of using an interface like this is > >>>>> minimal, and the time that it takes for the results to be > >>>>> returned is > >>>>> short. For example, currently on Mechanical Turk, a survey of > >>>>> approximately 50 items will be answered by 50 or more participants > >>>>> within a couple of hours, at a cost of approximately $1 per > >>>>> participant. Thus a survey can be completed within a day, at a > >>>>> cost of > >>>>> less than $50. (The hard work of designing the experiment, and > >>>>> constructing controlled materials remains of course.) > >>>>> > >>>>> Sorry to be so verbose. But I think that these methodological > >>>>> points > >>>>> are very important. > >>>>> > >>>>> Best wishes, > >>>>> > >>>>> Ted Gibson > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (In press). The need for quantitative > >>>>> methods in syntax and semantics research. Language and Cognitive > >>>>> Processes. > http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson > >>>>> & Fedorenko InPress LCP.pdf > >>>>> > >>>>> Gibson, E. & Fedorenko, E. (2010). Weak quantitative standards in > >>>>> linguistics research. Trends in Cognitive Science, 14, 233-234. > >>>>> http://tedlab.mit.edu/tedlab_website/researchpapers/Gibson & > >>>>> Fedorenko > >>>>> 2010 TICS.pdf > >>>>> > >>>>> > >>>>> > >>>>> > >>>>>> Dick, > >>>>>> > >>>>>> You raise an important issue here about methodology. I believe > >>>>>> that > >>>>>> intuitions are a fine way to generate hypotheses and even to test > >>>>>> them - to a degree. But while it might not have been feasible for > >>>>>> Huddleston, Pullum, and the other contributors to the Cambridge > >>>>>> Grammar to conduct experiments on every point of the grammar, > >>>>>> experiments could have only made the grammar better. The use of > >>>>>> intuitions, corpora, and standard psycholinguistic > >>>>>> experimentation > >>>>>> (indeed, Standard Social Science Methodology) is vital for > >>>>>> taking the > >>>>>> field forward and for providing the best support for different > >>>>>> analyses. Ted Gibson and Ev Fedorenko have written a very > >>>>>> useful new > >>>>>> paper on this, showing serious shortcomings with intuitions as > >>>>>> the > >>>>>> sole source of evidence, in their paper: "The need for > >>>>>> quantitative > >>>>>> methods in syntax and semantics research". > >>>>>> > >>>>>> Carson Schutze and Wayne Cowart, among others, have also written > >>>>>> convincingly on this. > >>>>>> > >>>>>> It is one reason that a team from Stanford, MIT (Brain and > >>>>>> Cognitive > >>>>>> Science), and researchers from Brazil are beginning a third > >>>>>> round of > >>>>>> experimental work among the Pirahas, since my own work on the > >>>>>> syntax > >>>>>> was, like almost every other field researcher's, based on native > >>>>>> speaker intuitions and corpora. > >>>>>> > >>>>>> The discussion of methodologies reminds me of the initial > >>>>>> reactions > >>>>>> to Greenberg's work on classifying the languages of the > >>>>>> Americas. His > >>>>>> methods were strongly (and justifiably) criticized. However, I > >>>>>> always > >>>>>> thought that his methods were a great way of generating > >>>>>> hypotheses, > >>>>>> so long as they were ultimately put to the test of standard > >>>>>> historical linguistics methods. And the same seems true for use > >>>>>> of > >>>>>> native-speaker intuitions. > >>>>>> > >>>>>> -- Dan > >>>>> > >>>>> > >>>>> > >>>>>>> We linguists can add a further layer of explanation to the > >>>>>>> judgements, but some judgements do seem to be more reliable than > >>>>>>> others. And if we have to wait for psycholinguistic evidence for > >>>>>>> every detailed analysis we make, our whole discipline will > >>>>>>> immediately grind to a halt. Like it or not, native speaker > >>>>>>> judgements are what put us linguists ahead of the rest in > >>>>>>> handling > >>>>>>> fine detail. Imagine writing the Cambridge Grammar of the > >>>>>>> English > >>>>>>> Language (or the OED) without using native speaker judgements. > >>>>>>> > >>>>>>> Best wishes, Dick Hudson > >>>>> > >>>>> > >>>>> > >>> > >>> > >>> > >> > > > > > > ------------------------------ > > Message: 28 > Date: Sat, 11 Sep 2010 12:53:23 -0400 > From: Daniel Everett > Subject: Re: [FUNKNET] analysis: unhappiness > To: Ted Gibson > Cc: Richard Hudson , Daniel Everett > , Evelina Fedorenko , > funknet > > Message-ID: <0021EEE8-560C-4E0A-8A3B-9595384807D6 at daneverett.org> > Content-Type: text/plain; charset=us-ascii > > Ted, > > Let me clarify this: > > 3. Dan says: "linguistics is not simply a subdiscipline of psychology" > > > > Both linguistics and psychology are big fields. We assume Dan is > referring to cognitive psychology / cognitive science here. (Of course, > there are sub-fields of psychology - e.g., personality psychology or > abnormal psychology - which are somewhat distinct from linguistics, but > those sub-fields are also distinct from cognitive psychology.) It is true > that historically linguistics is not treated as a subfield of cognitive > psychology / cognitive science. However, key research questions in > linguistics (i.e., the form of the knowledge structures and algorithms > underlying human language) are indeed a subset of those investigated by > cognitive psychologists / cognitive scientists. We think that the biggest > factor separating linguistics from psychology is the methods used to explore > the research questions, rather than the research questions themselves. > Consequently, we would like to continue to see tighter connections among the > fields of psychology / cognitive science, lingu > istics, as well as other fields like anthropology and computer science. > > > Correct, I meant cognitive psychology, not, say, psychoanalysis. There are > definitely overlapping concerns. But my main concern about language is less > about representations and more about the cultural and sociological values > that lead to sentences and expressions in the corpus, rather than the mind. > I used to think that my main interest was representations in the mind. But I > find the psychology less interesting than the anthropology these days. > > But this is not an excuse to avoid quantitative methods. I believe that you > and Ev, and others, have made a convincing case for quantitative methods. > Quantitative methods in field research on syntax and semantics is vital. > > -- dan > > > > > End of FUNKNET Digest, Vol 84, Issue 10 > *************************************** > From alifarghaly at yahoo.com Sun Sep 12 06:20:16 2010 From: alifarghaly at yahoo.com (Ali Farghaly) Date: Sat, 11 Sep 2010 23:20:16 -0700 Subject: Ali Farghaly wants to stay in touch on LinkedIn Message-ID: LinkedIn ------------ I'd like to add you to my professional network on LinkedIn. - Ali Farghaly Ali Farghaly Founder at Language Applications and Services San Francisco Bay Area Confirm that you know Ali Farghaly https://www.linkedin.com/e/-tffddx-gdzimymn-3l/isd/1659499320/pcnZ4P7w/ -- (c) 2010, LinkedIn Corporation From abergs at uos.de Sun Sep 12 07:40:21 2010 From: abergs at uos.de (Alexander Bergs) Date: Sun, 12 Sep 2010 09:40:21 +0200 Subject: ICLCE 4 - 2nd call for papers Message-ID: 4th International Conference on the Linguistics of Contemporary English ICLCE-4 Date: 19-Jul-2011 - 23-Jul-2011 Location: Osnabrueck, Germany Contact Person: Alexander Bergs Meeting Email: iclce4 at uos.de Web Site: http://www.blogs.uos.de/iclce4 2nd Call for Papers The attention devoted to the linguistics of the English language has resulted in a broad body of work in diverse research traditions. The aim of the ICLCE conference is to encourage the cross-fertilisation of ideas between different frameworks and research traditions, all of which may address any aspect of the linguistics of English. Previous ICLCE conferences were held in Edinburgh (2005), Toulouse (2007) and London (2009) along the same lines. We aim for the Osnabrueck conference to build on the success of those events. Plenary Speakers Scott F. Kiesling (University of Pittsburgh) Daniel Schreier (University of Zurich) Peter Stockwell (University of Nottingham) Graeme Trousdale (University of Edinburgh) Jessica de Villiers (University of British Columbia) Rachel Walker (University of Southern California) Gert Webelhuth (Goethe Universit?t Frankfurt am Main) We invite papers on any topic concerning the linguistics of contemporary English. Workshop proposals are also welcome. We are using the EasyABS system delivered by LinguistList. If there are any problems or if you have any questions, please feel free to contact us at iclce4uos.de. Abstracts should be no longer than 350 words, preferably format A4 with 2.5 cm margins, single-spaced, Times New Roman 12 pt, and with normal character spacing. All examples and references in the abstract should be included, but it is enough, when referring to previous work, to cite 'Author (Date)' in the body of the abstract - you do not need to include the full reference. Please only use common phonetic fonts such as SIL. +++++++++++++ Univ.-Prof. Dr. Alexander Bergs, M.A. Chair of English Language and Linguistics Institut f?r Anglistik und Amerikanistik (IfAA) Fachbereich 7 -Universitaet Osnabrueck Neuer Graben 40 D-49069 Osnabrueck Germany Tel: +49 541 969 4255 Tel: +49 541 969 6042 (secy) Fax: +49 541 969 4738 http:/www.ifaa.uni-osnabrueck.de/bergs +++++++++++++ Univ.-Prof. Dr. Alexander Bergs, M.A. Chair of English Language and Linguistics Institut f?r Anglistik und Amerikanistik (IfAA) Fachbereich 7 -Universitaet Osnabrueck Neuer Graben 40 D-49069 Osnabrueck Germany Tel: +49 541 969 4255 Tel: +49 541 969 6042 (secy) Fax: +49 541 969 4738 http:/www.ifaa.uni-osnabrueck.de/bergs From cbutler at ntlworld.com Sun Sep 12 10:03:31 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Sun, 12 Sep 2010 11:03:31 +0100 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya, I think two different things are getting a bit mixed up here. I don't for a moment dispute that expressions are often susceptible to multiple interpretations, that these interpretations are guided by all kinds of contextual information, or that different people, or even the same person at different times, may end up selecting differently from the various interpretations. Your example 'Open happiness' in another contribution to this thread illustrates these points very well. My point, though, is that each of these different interpretations, as well as the selection of one (or more) as more likely in a particular context, is achieved through mechanisms in the interpreter's brain which evolved in the course of the phylogenetic development of language in the human species, and developed ontogenetically in that particular interpreter's brain. It is surely likely that those mechanisms are highly similar in different human beings, even though there may be differences in the detailed wiring in different brains. What I'm saying is that in order to answer the question 'How do we communicate using language?' or, if you prefer, 'How does the language user work?', we need to investigate what those mechanisms are, and this is what psycholinguists can help us with. In particular, as linguists, we are interested (well, some of us are, though clearly not all) in whether the constructs we posit in our theories of language have psychological validity in the sense that they correspond to ways in which aspects of language are represented in the brain. [As an aside, I do realise that there are linguists who strenuously resist what they see as a misguided emphasis on mental representation in the work of cognitive scientists, but it seems clear that language must be represented in some way in the brain in order that we can engage in the sociosemiotic acts of meaning making which are the primary focus for many of these critics.] Taking your 'Open happiness' examples again, I think we need answers to questions such as: What kind of representation does the human language processing system have for 'open' and for 'happiness'? Are the phonological, syntactic, semantic and pragmatic (for those who distinguish the semantic from the pragmatic) properties of these items (and we might want to add 'for this particular interpreter', though there must be similarities across interpreters for communication to be possible) represented in the same or different ways, in the same or different locations (or sets of distributed locations)? Indeed, are we right in thinking that these familiar levels of linguistic description must be differentiated, as such, in the human language processing system? Does the representation for 'open' distinguish between what we call verbal and adjectival uses of this item, and if so how? Or are syntactic analyses computed on the fly, using semantic and contextual clues, rather than the neural equivalent of pigeon holes corresponding to verbs and adjectives? Is 'happiness' represented as 'happy + ness', or in its entirety, or both? All these questions, and many many more, are relevant to the construction of a model of language which reflects how language users communicate (as, of course, are a whole set of other questions about the sociocultural aspects of communication). I am not a psycholinguist, but my all too superficial reading in the area suggests that psycholinguists have gone some of the way towards answering some of the questions we might want to ask, but that there is still a long way to go. Chris -------------------------------------------------- From: "A. Katz" Sent: Saturday, September 11, 2010 3:18 PM To: "Chris Butler" Cc: "FUNKNET" Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > The problem is that once we achieve the psycholinguistic goal and see what > is happening in each person's head, and we see that conflicting analyses > are the norm, rather than the exception, among normal language speakers, > then we will realize that the way language works to transmit information > is despite individual differences and not because of uniform processing > strategies. > > Even when all are agreed as to the meaning of an utterance, they do not > process it the same way. Which means that processing is seocndary to > information transmission. > > --Aya > > > > > On Sat, 11 Sep 2010, Chris Butler wrote: > >> Dick's comment that "this discussion raises the really fundamental >> question of what kind of thing we think language is: social or >> individual" is, it seems to me, an important one, particularly for those >> of us who are committed functionalists. My own view is that a truly >> functional model of language would be one which aims to account for how >> human beings communicate using language, or in other words tries to >> answer the question which was posed by Simon Dik a long time ago now, but >> which was not tackled head-on in his own work: "How does the natural >> language user work?' In trying to answer this question we need to accept >> that language is BOTH social AND individual, and we need to explore both >> aspects to get as complete a picture as possible of how we communicate >> using language. We need to know BOTH how people create and respond to >> meanings and express those meanings in forms during social interaction >> AND the mechanisms which operate in the brains of individuals in order to >> make such interaction possible. Both are important parts of the answer to >> the question 'How do we communicate using language?', though this >> particular thread of the Funknet discussion has concentrated more on the >> second aspect, and so will I. >> >> This doesn't mean that all the work linguists have done on "exploring the >> structure of a language so that I can understand how all the bits fit >> together" and "exploring the connections between items", as Dick puts it, >> is useless - far from it. After all, the hypotheses that psycholinguists >> test are based on ideas about what languages are like. But it does mean, >> in my view, that ultimately we need to get evidence that the constructs >> and analyses we propose are ones that are at least consistent with what >> we know of the processes which go on when we use language. So I am with >> Matthew when he says that for him, "the only sense in which an analysis >> can be "the correct analysis" is in terms of what is inside of people's >> heads". Of course, this doesn't imply that linguists should just give up >> their jobs until such time as we know everything there is to know about >> language processing. But it does mean that we need to collaborate with >> psycholinguists, psychologists and neurologists, as has also been pointed >> out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan >> Nuyts. [We also need to collaborate much more with sociolinguists and >> sociologists, so that we can get a better handle on the sociocultural >> aspects of how we communicate.] And it also means that psycholinguists, >> for their part, need whenever possible to follow up tightly controlled >> lab experiments with studies under more naturalistic conditions, to avoid >> the criticism that what happens in artifical lab situations may not >> happen in natural communicative conditions. >> >> I also agree with Dick when he says that "the differences between >> individuals really matter", and with Lise when she points out that "we >> must also be careful not to idealize "what's in people's heads" as if it >> were a single coherent construct that we are trying to discover". >> However, there are surely processing mechanisms which are common to all >> language users by virtue of the evolution of the language faculty and >> which constitute the "general processes" which Dick says psycholinguists >> are interested in. >> >> On the issue of quantitative methodology, I'm sympathetic in general to >> Ted and Ev's views, though it does seem sensible to prioritise cases in >> terms of a hierarchy such as Brian suggests. One thing this means is that >> we should be giving our university students of linguistics (and some of >> our linguistics lecturers!) courses in quantitative aspects of >> linguistics that introduce them to the use of at least some of the basic >> statistical methods in language study, and I'm sure this is indeed going >> on in some enlightened places. To those who suspect this can't be done >> with maths-shy students who don't initially see the need for it, I offer >> my own experience, over quite a long period, of teaching such courses to >> people with little or no prior experience in quantitative techniques. For >> some years in the 1990s, I taught such courses to all linguistics >> students in an institution where we had many mature students who had come >> into university level studies with non-standard qualifications, and were >> not well equipped for courses of this kind by their previous experience. >> I'm glad to say that teaching the subject from their own perspective as >> language students rather than that of the statistician, and explaining >> the reasons for doing things in particular ways rather than just >> presenting formulae, paid off in the end, so that most students were able >> to appreciate the relevance of these courses and to turn in very >> creditable projects showing an understanding of research design and >> competence in the use of a range of basic statistical techniques. And I >> still find that bright graduate students respond well to similar courses >> which incorporate some of the rather more advanced techniques needed for >> many real research projects in various areas of linguistics. But I may >> well be out of date with what is now already happening in our fine >> institutions of higher education! >> >> Chris Butler >> >> From amnfn at well.com Sun Sep 12 13:13:05 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 06:13:05 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <0EEE18DF122B4EDFBB744F0A83BB52AE@OwnerPC> Message-ID: Chris, Thanks for addressing this question. I understand that many, many linguists are quite properly and approriately interested in this ultimate question: "How does the language user work?" (I am also intetested in this question some of the time.) My point is that "how does language work?" is also a valid question, and a central one to the field of linguistics. These two questions are not at all the same. Let me be very explicit: My aim is to separate out the "human" from the "language". There are many good reasons to do so. For anyone working in computerized language processing or in non-human language studies, this is a significant point. It does not matter if a computerized language processing system even remotely simulates what humans do with language in their brains. It does matter whether it comes up with comparable or indistinguishable results. It does not matter whether a parrot, a dolphin or a chimpanzee is doing the same things inside the same modules in his brain as a human does. It does matter if the results are functionally equivalent. We need to make that distinction, between humans and their language, or we will be caught inside a circular definition with no way to break out or to prove anything, not about others and not about ourselves! --Aya http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies On Sun, 12 Sep 2010, Chris Butler wrote: > Aya, I think two different things are getting a bit mixed up here. > > I don't for a moment dispute that expressions are often susceptible to > multiple interpretations, that these interpretations are guided by all kinds > of contextual information, or that different people, or even the same person > at different times, may end up selecting differently from the various > interpretations. Your example 'Open happiness' in another contribution to > this thread illustrates these points very well. > > My point, though, is that each of these different interpretations, as well as > the selection of one (or more) as more likely in a particular context, is > achieved through mechanisms in the interpreter's brain which evolved in the > course of the phylogenetic development of language in the human species, and > developed ontogenetically in that particular interpreter's brain. It is > surely likely that those mechanisms are highly similar in different human > beings, even though there may be differences in the detailed wiring in > different brains. What I'm saying is that in order to answer the question > 'How do we communicate using language?' or, if you prefer, 'How does the > language user work?', we need to investigate what those mechanisms are, and > this is what psycholinguists can help us with. > > In particular, as linguists, we are interested (well, some of us are, though > clearly not all) in whether the constructs we posit in our theories of > language have psychological validity in the sense that they correspond to > ways in which aspects of language are represented in the brain. [As an aside, > I do realise that there are linguists who strenuously resist what they see as > a misguided emphasis on mental representation in the work of cognitive > scientists, but it seems clear that language must be represented in some way > in the brain in order that we can engage in the sociosemiotic acts of meaning > making which are the primary focus for many of these critics.] Taking your > 'Open happiness' examples again, I think we need answers to questions such > as: What kind of representation does the human language processing system > have for 'open' and for 'happiness'? Are the phonological, syntactic, > semantic and pragmatic (for those who distinguish the semantic from the > pragmatic) properties of these items (and we might want to add 'for this > particular interpreter', though there must be similarities across > interpreters for communication to be possible) represented in the same or > different ways, in the same or different locations (or sets of distributed > locations)? Indeed, are we right in thinking that these familiar levels of > linguistic description must be differentiated, as such, in the human language > processing system? Does the representation for 'open' distinguish between > what we call verbal and adjectival uses of this item, and if so how? Or are > syntactic analyses computed on the fly, using semantic and contextual clues, > rather than the neural equivalent of pigeon holes corresponding to verbs and > adjectives? Is 'happiness' represented as 'happy + ness', or in its entirety, > or both? All these questions, and many many more, are relevant to the > construction of a model of language which reflects how language users > communicate (as, of course, are a whole set of other questions about the > sociocultural aspects of communication). I am not a psycholinguist, but my > all too superficial reading in the area suggests that psycholinguists have > gone some of the way towards answering some of the questions we might want to > ask, but that there is still a long way to go. > > Chris > -------------------------------------------------- > From: "A. Katz" > Sent: Saturday, September 11, 2010 3:18 PM > To: "Chris Butler" > Cc: "FUNKNET" > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > >> The problem is that once we achieve the psycholinguistic goal and see what >> is happening in each person's head, and we see that conflicting analyses >> are the norm, rather than the exception, among normal language speakers, >> then we will realize that the way language works to transmit information >> is despite individual differences and not because of uniform processing >> strategies. >> >> Even when all are agreed as to the meaning of an utterance, they do not >> process it the same way. Which means that processing is seocndary to >> information transmission. >> >> --Aya >> >> >> >> >> On Sat, 11 Sep 2010, Chris Butler wrote: >> >>> Dick's comment that "this discussion raises the really fundamental >>> question of what kind of thing we think language is: social or individual" >>> is, it seems to me, an important one, particularly for those of us who are >>> committed functionalists. My own view is that a truly functional model of >>> language would be one which aims to account for how human beings >>> communicate using language, or in other words tries to answer the question >>> which was posed by Simon Dik a long time ago now, but which was not >>> tackled head-on in his own work: "How does the natural language user >>> work?' In trying to answer this question we need to accept that language >>> is BOTH social AND individual, and we need to explore both aspects to get >>> as complete a picture as possible of how we communicate using language. We >>> need to know BOTH how people create and respond to meanings and express >>> those meanings in forms during social interaction AND the mechanisms which >>> operate in the brains of individuals in order to make such interaction >>> possible. Both are important parts of the answer to the question 'How do >>> we communicate using language?', though this particular thread of the >>> Funknet discussion has concentrated more on the second aspect, and so will >>> I. >>> >>> This doesn't mean that all the work linguists have done on "exploring the >>> structure of a language so that I can understand how all the bits fit >>> together" and "exploring the connections between items", as Dick puts it, >>> is useless - far from it. After all, the hypotheses that psycholinguists >>> test are based on ideas about what languages are like. But it does mean, >>> in my view, that ultimately we need to get evidence that the constructs >>> and analyses we propose are ones that are at least consistent with what we >>> know of the processes which go on when we use language. So I am with >>> Matthew when he says that for him, "the only sense in which an analysis >>> can be "the correct analysis" is in terms of what is inside of people's >>> heads". Of course, this doesn't imply that linguists should just give up >>> their jobs until such time as we know everything there is to know about >>> language processing. But it does mean that we need to collaborate with >>> psycholinguists, psychologists and neurologists, as has also been pointed >>> out by linguists such as Ray Jackendoff, Asif Agha, Ewa Dabrowska and Jan >>> Nuyts. [We also need to collaborate much more with sociolinguists and >>> sociologists, so that we can get a better handle on the sociocultural >>> aspects of how we communicate.] And it also means that psycholinguists, >>> for their part, need whenever possible to follow up tightly controlled lab >>> experiments with studies under more naturalistic conditions, to avoid the >>> criticism that what happens in artifical lab situations may not happen in >>> natural communicative conditions. >>> >>> I also agree with Dick when he says that "the differences between >>> individuals really matter", and with Lise when she points out that "we >>> must also be careful not to idealize "what's in people's heads" as if it >>> were a single coherent construct that we are trying to discover". However, >>> there are surely processing mechanisms which are common to all language >>> users by virtue of the evolution of the language faculty and which >>> constitute the "general processes" which Dick says psycholinguists are >>> interested in. >>> >>> On the issue of quantitative methodology, I'm sympathetic in general to >>> Ted and Ev's views, though it does seem sensible to prioritise cases in >>> terms of a hierarchy such as Brian suggests. One thing this means is that >>> we should be giving our university students of linguistics (and some of >>> our linguistics lecturers!) courses in quantitative aspects of linguistics >>> that introduce them to the use of at least some of the basic statistical >>> methods in language study, and I'm sure this is indeed going on in some >>> enlightened places. To those who suspect this can't be done with maths-shy >>> students who don't initially see the need for it, I offer my own >>> experience, over quite a long period, of teaching such courses to people >>> with little or no prior experience in quantitative techniques. For some >>> years in the 1990s, I taught such courses to all linguistics students in >>> an institution where we had many mature students who had come into >>> university level studies with non-standard qualifications, and were not >>> well equipped for courses of this kind by their previous experience. I'm >>> glad to say that teaching the subject from their own perspective as >>> language students rather than that of the statistician, and explaining the >>> reasons for doing things in particular ways rather than just presenting >>> formulae, paid off in the end, so that most students were able to >>> appreciate the relevance of these courses and to turn in very creditable >>> projects showing an understanding of research design and competence in the >>> use of a range of basic statistical techniques. And I still find that >>> bright graduate students respond well to similar courses which incorporate >>> some of the rather more advanced techniques needed for many real research >>> projects in various areas of linguistics. But I may well be out of date >>> with what is now already happening in our fine institutions of higher >>> education! >>> >>> Chris Butler >>> >>> > > From tgivon at uoregon.edu Sun Sep 12 15:46:12 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Sun, 12 Sep 2010 09:46:12 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: I wonder whether asking "how does language work?" is really a meaningful question without asking "how does the language user work?" One of the worst legacies good ol' Noam stuck us with is his (really, Saussure's) distinction of competence ("language", "knowledge") vs. performance ("language user", "processing"). It purported to limit linguists to the armchair methods that study competence, and relegated to psychology the quantified, distributional/variationist methods that study behavior, processing and on-line communication. The first breach in this artificial methodological wall occurred, leastwise for some of us, when we discovered the intermediate method of quantified studies of text, interaction, and conversation. As an ex-biologist, I am forever puzzled by the methodological purism we sill seem to embrace in linguistics, in the face of the manifest complexity and connectivity of language (mind, brain, culture, sociality, evolution, etc.). In biology, another extremely complex science with multiple connections (chemistry, geology, paleontology, behavior, sociality, economics, evolution, etc.), ANY method is welcome so long as it does the job of furthering our understanding. And by understanding we mean ever-wider circles of connectivity. Best, TG ================ A. Katz wrote: > Chris, > > Thanks for addressing this question. I understand that many, many > linguists are quite properly and approriately interested in this > ultimate question: "How does the language user work?" (I am also > intetested in this question some of the time.) > > My point is that "how does language work?" is also a valid question, > and a central one to the field of linguistics. These two questions are > not at all the same. > > Let me be very explicit: My aim is to separate out the "human" from > the "language". There are many good reasons to do so. For anyone > working in computerized language processing or in non-human language > studies, this is a significant point. > > It does not matter if a computerized language processing system even > remotely simulates what humans do with language in their brains. It > does matter whether it comes up with comparable or indistinguishable > results. > > It does not matter whether a parrot, a dolphin or a chimpanzee is > doing the same things inside the same modules in his brain as a human > does. It does matter if the results are functionally equivalent. > > We need to make that distinction, between humans and their language, > or we will be caught inside a circular definition with no way to break > out or to prove anything, not about others and not about ourselves! > > --Aya > > http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies > > > On Sun, 12 Sep 2010, Chris Butler wrote: > >> Aya, I think two different things are getting a bit mixed up here. >> >> I don't for a moment dispute that expressions are often susceptible >> to multiple interpretations, that these interpretations are guided by >> all kinds of contextual information, or that different people, or >> even the same person at different times, may end up selecting >> differently from the various interpretations. Your example 'Open >> happiness' in another contribution to this thread illustrates these >> points very well. >> >> My point, though, is that each of these different interpretations, as >> well as the selection of one (or more) as more likely in a particular >> context, is achieved through mechanisms in the interpreter's brain >> which evolved in the course of the phylogenetic development of >> language in the human species, and developed ontogenetically in that >> particular interpreter's brain. It is surely likely that those >> mechanisms are highly similar in different human beings, even though >> there may be differences in the detailed wiring in different brains. >> What I'm saying is that in order to answer the question 'How do we >> communicate using language?' or, if you prefer, 'How does the >> language user work?', we need to investigate what those mechanisms >> are, and this is what psycholinguists can help us with. >> >> In particular, as linguists, we are interested (well, some of us are, >> though clearly not all) in whether the constructs we posit in our >> theories of language have psychological validity in the sense that >> they correspond to ways in which aspects of language are represented >> in the brain. [As an aside, I do realise that there are linguists who >> strenuously resist what they see as a misguided emphasis on mental >> representation in the work of cognitive scientists, but it seems >> clear that language must be represented in some way in the brain in >> order that we can engage in the sociosemiotic acts of meaning making >> which are the primary focus for many of these critics.] Taking your >> 'Open happiness' examples again, I think we need answers to questions >> such as: What kind of representation does the human language >> processing system have for 'open' and for 'happiness'? Are the >> phonological, syntactic, semantic and pragmatic (for those who >> distinguish the semantic from the pragmatic) properties of these >> items (and we might want to add 'for this particular interpreter', >> though there must be similarities across interpreters for >> communication to be possible) represented in the same or different >> ways, in the same or different locations (or sets of distributed >> locations)? Indeed, are we right in thinking that these familiar >> levels of linguistic description must be differentiated, as such, in >> the human language processing system? Does the representation for >> 'open' distinguish between what we call verbal and adjectival uses of >> this item, and if so how? Or are syntactic analyses computed on the >> fly, using semantic and contextual clues, rather than the neural >> equivalent of pigeon holes corresponding to verbs and adjectives? Is >> 'happiness' represented as 'happy + ness', or in its entirety, or >> both? All these questions, and many many more, are relevant to the >> construction of a model of language which reflects how language users >> communicate (as, of course, are a whole set of other questions about >> the sociocultural aspects of communication). I am not a >> psycholinguist, but my all too superficial reading in the area >> suggests that psycholinguists have gone some of the way towards >> answering some of the questions we might want to ask, but that there >> is still a long way to go. >> >> Chris >> -------------------------------------------------- >> From: "A. Katz" >> Sent: Saturday, September 11, 2010 3:18 PM >> To: "Chris Butler" >> Cc: "FUNKNET" >> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >> >>> The problem is that once we achieve the psycholinguistic goal and >>> see what >>> is happening in each person's head, and we see that conflicting >>> analyses >>> are the norm, rather than the exception, among normal language >>> speakers, >>> then we will realize that the way language works to transmit >>> information >>> is despite individual differences and not because of uniform processing >>> strategies. >>> >>> Even when all are agreed as to the meaning of an utterance, they do not >>> process it the same way. Which means that processing is seocndary to >>> information transmission. >>> >>> --Aya >>> >>> >>> >>> >>> On Sat, 11 Sep 2010, Chris Butler wrote: >>> >>>> Dick's comment that "this discussion raises the really fundamental >>>> question of what kind of thing we think language is: social or >>>> individual" is, it seems to me, an important one, particularly for >>>> those of us who are committed functionalists. My own view is that a >>>> truly functional model of language would be one which aims to >>>> account for how human beings communicate using language, or in >>>> other words tries to answer the question which was posed by Simon >>>> Dik a long time ago now, but which was not tackled head-on in his >>>> own work: "How does the natural language user work?' In trying to >>>> answer this question we need to accept that language is BOTH social >>>> AND individual, and we need to explore both aspects to get as >>>> complete a picture as possible of how we communicate using >>>> language. We need to know BOTH how people create and respond to >>>> meanings and express those meanings in forms during social >>>> interaction AND the mechanisms which operate in the brains of >>>> individuals in order to make such interaction possible. Both are >>>> important parts of the answer to the question 'How do we >>>> communicate using language?', though this particular thread of the >>>> Funknet discussion has concentrated more on the second aspect, and >>>> so will I. >>>> >>>> This doesn't mean that all the work linguists have done on >>>> "exploring the structure of a language so that I can understand how >>>> all the bits fit together" and "exploring the connections between >>>> items", as Dick puts it, is useless - far from it. After all, the >>>> hypotheses that psycholinguists test are based on ideas about what >>>> languages are like. But it does mean, in my view, that ultimately >>>> we need to get evidence that the constructs and analyses we propose >>>> are ones that are at least consistent with what we know of the >>>> processes which go on when we use language. So I am with Matthew >>>> when he says that for him, "the only sense in which an analysis can >>>> be "the correct analysis" is in terms of what is inside of people's >>>> heads". Of course, this doesn't imply that linguists should just >>>> give up their jobs until such time as we know everything there is >>>> to know about language processing. But it does mean that we need to >>>> collaborate with psycholinguists, psychologists and neurologists, >>>> as has also been pointed out by linguists such as Ray Jackendoff, >>>> Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>> collaborate much more with sociolinguists and sociologists, so that >>>> we can get a better handle on the sociocultural aspects of how we >>>> communicate.] And it also means that psycholinguists, for their >>>> part, need whenever possible to follow up tightly controlled lab >>>> experiments with studies under more naturalistic conditions, to >>>> avoid the criticism that what happens in artifical lab situations >>>> may not happen in natural communicative conditions. >>>> >>>> I also agree with Dick when he says that "the differences between >>>> individuals really matter", and with Lise when she points out that >>>> "we must also be careful not to idealize "what's in people's heads" >>>> as if it were a single coherent construct that we are trying to >>>> discover". However, there are surely processing mechanisms which >>>> are common to all language users by virtue of the evolution of the >>>> language faculty and which constitute the "general processes" which >>>> Dick says psycholinguists are interested in. >>>> >>>> On the issue of quantitative methodology, I'm sympathetic in >>>> general to Ted and Ev's views, though it does seem sensible to >>>> prioritise cases in terms of a hierarchy such as Brian suggests. >>>> One thing this means is that we should be giving our university >>>> students of linguistics (and some of our linguistics lecturers!) >>>> courses in quantitative aspects of linguistics that introduce them >>>> to the use of at least some of the basic statistical methods in >>>> language study, and I'm sure this is indeed going on in some >>>> enlightened places. To those who suspect this can't be done with >>>> maths-shy students who don't initially see the need for it, I offer >>>> my own experience, over quite a long period, of teaching such >>>> courses to people with little or no prior experience in >>>> quantitative techniques. For some years in the 1990s, I taught such >>>> courses to all linguistics students in an institution where we had >>>> many mature students who had come into university level studies >>>> with non-standard qualifications, and were not well equipped for >>>> courses of this kind by their previous experience. I'm glad to say >>>> that teaching the subject from their own perspective as language >>>> students rather than that of the statistician, and explaining the >>>> reasons for doing things in particular ways rather than just >>>> presenting formulae, paid off in the end, so that most students >>>> were able to appreciate the relevance of these courses and to turn >>>> in very creditable projects showing an understanding of research >>>> design and competence in the use of a range of basic statistical >>>> techniques. And I still find that bright graduate students respond >>>> well to similar courses which incorporate some of the rather more >>>> advanced techniques needed for many real research projects in >>>> various areas of linguistics. But I may well be out of date with >>>> what is now already happening in our fine institutions of higher >>>> education! >>>> >>>> Chris Butler >>>> >>>> >> >> From amnfn at well.com Sun Sep 12 16:19:46 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 09:19:46 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <4C8CF5C4.8070005@uoregon.edu> Message-ID: Tom, I understand the uncomfortable association with Chomsky that speaking of language apart from people can have. Competence versus performance, the way Chomsky used those terms, never made sense. But that's precisely because he associated "competence" with the brain and suggested that it was hard wired there -- when there was never any evidence of that. However, if we don't distinguish language from humans, and language processing from language data, then how are we going to judge artificial language-using devices as to their efficacy at producing and interpreting language? How are we going to determine whether and to what extent a non-human has acquired language? We aren't born with it. We don't embody it. It's a tool that we use to communicate. Other tools can be studied separately from the people who use them or created them. Why not language? --Aya On Sun, 12 Sep 2010, Tom Givon wrote: > > > > I wonder whether asking "how does language work?" is really a meaningful > question without asking "how does the language user work?" One of the worst > legacies good ol' Noam stuck us with is his (really, Saussure's) distinction > of competence ("language", "knowledge") vs. performance ("language user", > "processing"). It purported to limit linguists to the armchair methods that > study competence, and relegated to psychology the quantified, > distributional/variationist methods that study behavior, processing and > on-line communication. The first breach in this artificial methodological > wall occurred, leastwise for some of us, when we discovered the intermediate > method of quantified studies of text, interaction, and conversation. As an > ex-biologist, I am forever puzzled by the methodological purism we sill seem > to embrace in linguistics, in the face of the manifest complexity and > connectivity of language (mind, brain, culture, sociality, evolution, etc.). > In biology, another extremely complex science with multiple connections > (chemistry, geology, paleontology, behavior, sociality, economics, evolution, > etc.), ANY method is welcome so long as it does the job of furthering our > understanding. And by understanding we mean ever-wider circles of > connectivity. > > Best, TG > ================ > > > A. Katz wrote: >> Chris, >> >> Thanks for addressing this question. I understand that many, many linguists >> are quite properly and approriately interested in this ultimate question: >> "How does the language user work?" (I am also intetested in this question >> some of the time.) >> >> My point is that "how does language work?" is also a valid question, and a >> central one to the field of linguistics. These two questions are not at all >> the same. >> >> Let me be very explicit: My aim is to separate out the "human" from the >> "language". There are many good reasons to do so. For anyone working in >> computerized language processing or in non-human language studies, this is >> a significant point. >> >> It does not matter if a computerized language processing system even >> remotely simulates what humans do with language in their brains. It does >> matter whether it comes up with comparable or indistinguishable results. >> >> It does not matter whether a parrot, a dolphin or a chimpanzee is doing the >> same things inside the same modules in his brain as a human does. It does >> matter if the results are functionally equivalent. >> >> We need to make that distinction, between humans and their language, or we >> will be caught inside a circular definition with no way to break out or to >> prove anything, not about others and not about ourselves! >> >> --Aya >> >> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >> >> >> On Sun, 12 Sep 2010, Chris Butler wrote: >> >>> Aya, I think two different things are getting a bit mixed up here. >>> >>> I don't for a moment dispute that expressions are often susceptible to >>> multiple interpretations, that these interpretations are guided by all >>> kinds of contextual information, or that different people, or even the >>> same person at different times, may end up selecting differently from the >>> various interpretations. Your example 'Open happiness' in another >>> contribution to this thread illustrates these points very well. >>> >>> My point, though, is that each of these different interpretations, as well >>> as the selection of one (or more) as more likely in a particular context, >>> is achieved through mechanisms in the interpreter's brain which evolved in >>> the course of the phylogenetic development of language in the human >>> species, and developed ontogenetically in that particular interpreter's >>> brain. It is surely likely that those mechanisms are highly similar in >>> different human beings, even though there may be differences in the >>> detailed wiring in different brains. What I'm saying is that in order to >>> answer the question 'How do we communicate using language?' or, if you >>> prefer, 'How does the language user work?', we need to investigate what >>> those mechanisms are, and this is what psycholinguists can help us with. >>> >>> In particular, as linguists, we are interested (well, some of us are, >>> though clearly not all) in whether the constructs we posit in our theories >>> of language have psychological validity in the sense that they correspond >>> to ways in which aspects of language are represented in the brain. [As an >>> aside, I do realise that there are linguists who strenuously resist what >>> they see as a misguided emphasis on mental representation in the work of >>> cognitive scientists, but it seems clear that language must be represented >>> in some way in the brain in order that we can engage in the sociosemiotic >>> acts of meaning making which are the primary focus for many of these >>> critics.] Taking your 'Open happiness' examples again, I think we need >>> answers to questions such as: What kind of representation does the human >>> language processing system have for 'open' and for 'happiness'? Are the >>> phonological, syntactic, semantic and pragmatic (for those who distinguish >>> the semantic from the pragmatic) properties of these items (and we might >>> want to add 'for this particular interpreter', though there must be >>> similarities across interpreters for communication to be possible) >>> represented in the same or different ways, in the same or different >>> locations (or sets of distributed locations)? Indeed, are we right in >>> thinking that these familiar levels of linguistic description must be >>> differentiated, as such, in the human language processing system? Does the >>> representation for 'open' distinguish between what we call verbal and >>> adjectival uses of this item, and if so how? Or are syntactic analyses >>> computed on the fly, using semantic and contextual clues, rather than the >>> neural equivalent of pigeon holes corresponding to verbs and adjectives? >>> Is 'happiness' represented as 'happy + ness', or in its entirety, or both? >>> All these questions, and many many more, are relevant to the construction >>> of a model of language which reflects how language users communicate (as, >>> of course, are a whole set of other questions about the sociocultural >>> aspects of communication). I am not a psycholinguist, but my all too >>> superficial reading in the area suggests that psycholinguists have gone >>> some of the way towards answering some of the questions we might want to >>> ask, but that there is still a long way to go. >>> >>> Chris >>> -------------------------------------------------- >>> From: "A. Katz" >>> Sent: Saturday, September 11, 2010 3:18 PM >>> To: "Chris Butler" >>> Cc: "FUNKNET" >>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>> >>>> The problem is that once we achieve the psycholinguistic goal and see >>>> what >>>> is happening in each person's head, and we see that conflicting analyses >>>> are the norm, rather than the exception, among normal language speakers, >>>> then we will realize that the way language works to transmit information >>>> is despite individual differences and not because of uniform processing >>>> strategies. >>>> >>>> Even when all are agreed as to the meaning of an utterance, they do not >>>> process it the same way. Which means that processing is seocndary to >>>> information transmission. >>>> >>>> --Aya >>>> >>>> >>>> >>>> >>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>> >>>>> Dick's comment that "this discussion raises the really fundamental >>>>> question of what kind of thing we think language is: social or >>>>> individual" is, it seems to me, an important one, particularly for those >>>>> of us who are committed functionalists. My own view is that a truly >>>>> functional model of language would be one which aims to account for how >>>>> human beings communicate using language, or in other words tries to >>>>> answer the question which was posed by Simon Dik a long time ago now, >>>>> but which was not tackled head-on in his own work: "How does the natural >>>>> language user work?' In trying to answer this question we need to accept >>>>> that language is BOTH social AND individual, and we need to explore both >>>>> aspects to get as complete a picture as possible of how we communicate >>>>> using language. We need to know BOTH how people create and respond to >>>>> meanings and express those meanings in forms during social interaction >>>>> AND the mechanisms which operate in the brains of individuals in order >>>>> to make such interaction possible. Both are important parts of the >>>>> answer to the question 'How do we communicate using language?', though >>>>> this particular thread of the Funknet discussion has concentrated more >>>>> on the second aspect, and so will I. >>>>> >>>>> This doesn't mean that all the work linguists have done on "exploring >>>>> the structure of a language so that I can understand how all the bits >>>>> fit together" and "exploring the connections between items", as Dick >>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>> psycholinguists test are based on ideas about what languages are like. >>>>> But it does mean, in my view, that ultimately we need to get evidence >>>>> that the constructs and analyses we propose are ones that are at least >>>>> consistent with what we know of the processes which go on when we use >>>>> language. So I am with Matthew when he says that for him, "the only >>>>> sense in which an analysis can be "the correct analysis" is in terms of >>>>> what is inside of people's heads". Of course, this doesn't imply that >>>>> linguists should just give up their jobs until such time as we know >>>>> everything there is to know about language processing. But it does mean >>>>> that we need to collaborate with psycholinguists, psychologists and >>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>> collaborate much more with sociolinguists and sociologists, so that we >>>>> can get a better handle on the sociocultural aspects of how we >>>>> communicate.] And it also means that psycholinguists, for their part, >>>>> need whenever possible to follow up tightly controlled lab experiments >>>>> with studies under more naturalistic conditions, to avoid the criticism >>>>> that what happens in artifical lab situations may not happen in natural >>>>> communicative conditions. >>>>> >>>>> I also agree with Dick when he says that "the differences between >>>>> individuals really matter", and with Lise when she points out that "we >>>>> must also be careful not to idealize "what's in people's heads" as if it >>>>> were a single coherent construct that we are trying to discover". >>>>> However, there are surely processing mechanisms which are common to all >>>>> language users by virtue of the evolution of the language faculty and >>>>> which constitute the "general processes" which Dick says psycholinguists >>>>> are interested in. >>>>> >>>>> On the issue of quantitative methodology, I'm sympathetic in general to >>>>> Ted and Ev's views, though it does seem sensible to prioritise cases in >>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>> that we should be giving our university students of linguistics (and >>>>> some of our linguistics lecturers!) courses in quantitative aspects of >>>>> linguistics that introduce them to the use of at least some of the basic >>>>> statistical methods in language study, and I'm sure this is indeed going >>>>> on in some enlightened places. To those who suspect this can't be done >>>>> with maths-shy students who don't initially see the need for it, I offer >>>>> my own experience, over quite a long period, of teaching such courses to >>>>> people with little or no prior experience in quantitative techniques. >>>>> For some years in the 1990s, I taught such courses to all linguistics >>>>> students in an institution where we had many mature students who had >>>>> come into university level studies with non-standard qualifications, and >>>>> were not well equipped for courses of this kind by their previous >>>>> experience. I'm glad to say that teaching the subject from their own >>>>> perspective as language students rather than that of the statistician, >>>>> and explaining the reasons for doing things in particular ways rather >>>>> than just presenting formulae, paid off in the end, so that most >>>>> students were able to appreciate the relevance of these courses and to >>>>> turn in very creditable projects showing an understanding of research >>>>> design and competence in the use of a range of basic statistical >>>>> techniques. And I still find that bright graduate students respond well >>>>> to similar courses which incorporate some of the rather more advanced >>>>> techniques needed for many real research projects in various areas of >>>>> linguistics. But I may well be out of date with what is now already >>>>> happening in our fine institutions of higher education! >>>>> >>>>> Chris Butler >>>>> >>>>> >>> >>> > > From mark at polymathix.com Mon Sep 13 00:16:10 2010 From: mark at polymathix.com (Mark P. Line) Date: Sun, 12 Sep 2010 19:16:10 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya -- You said: "Other tools can be studied separately from the people who use them or created them. Why not language?" Although language can certainly be considered a tool, I think it's unlike other tools in several very significant respects. 1. Although we're not born with language, we can't avoid (pathologies excluded) acquiring it unless we're not exposed to it. To that extent, language is a biological phenomenon. A prototypical tool is not a biological phenomenon, so I'm not sure how valid any conclusions might be that are drawn from a premise of language-as-tool. 2. A tool is as a tool does: Anything is a tool that is being used as a tool, including dead wombats, broken screwdrivers or decks of playing cards. (Completing the imagined scenarios is left as an exercise for the reader...) So saying that language is a tool is only saying that language is used as a tool. Quite a few conclusions can be and have been drawn from the fact that language is used as a tool, but I would have to be convinced in detail that almost everything worth knowing about language is dependent on the premise of language-as-tool. 3. If language is a "tool" for (say) communicating ideas, then eating is a "tool" for reducing hunger. In both cases, I worry about the tool metaphor being stretched so far from the prototype that we're left with an out-and-out category fallacy. 4. More prototypical tools can be studied separately from those who use or create them because those tools are easily observed separately from those who use or create them. I don't think the same thing can be said of language -- very little about language can be observed apart from its use, so very little about language can be observed apart from its user(s). 5. Any proposal to study something as complex as language separately from its embodiment is suspicious to me, smacking of reductionism -- something up with which linguistics has had to put a tad much. Anything that puts language back into its human context would be a step forward. -- Mark Mark P. Line A. Katz wrote: > Tom, > > I understand the uncomfortable association with Chomsky that speaking of > language apart from people can have. Competence versus performance, the > way Chomsky used those terms, never made sense. But that's precisely > because > he associated "competence" with the brain and suggested that it was hard > wired there -- when there was never any evidence of that. > > However, if we don't distinguish language from humans, and language > processing from language data, then how are we going to judge artificial > language-using devices as to their efficacy at producing and interpreting > language? How are we going to determine whether and to what extent a > non-human has acquired language? > > We aren't born with it. We don't embody it. It's a tool that we use to > communicate. Other tools can be studied separately from the people who use > them or created them. Why not language? > > --Aya > > > > > On Sun, 12 Sep 2010, Tom Givon wrote: > >> >> >> >> I wonder whether asking "how does language work?" is really a meaningful >> question without asking "how does the language user work?" One of the >> worst >> legacies good ol' Noam stuck us with is his (really, Saussure's) >> distinction >> of competence ("language", "knowledge") vs. performance ("language >> user", >> "processing"). It purported to limit linguists to the armchair methods >> that >> study competence, and relegated to psychology the quantified, >> distributional/variationist methods that study behavior, processing and >> on-line communication. The first breach in this artificial >> methodological >> wall occurred, leastwise for some of us, when we discovered the >> intermediate >> method of quantified studies of text, interaction, and conversation. As >> an >> ex-biologist, I am forever puzzled by the methodological purism we sill >> seem >> to embrace in linguistics, in the face of the manifest complexity and >> connectivity of language (mind, brain, culture, sociality, evolution, >> etc.). >> In biology, another extremely complex science with multiple connections >> (chemistry, geology, paleontology, behavior, sociality, economics, >> evolution, >> etc.), ANY method is welcome so long as it does the job of furthering >> our >> understanding. And by understanding we mean ever-wider circles of >> connectivity. >> >> Best, TG >> ================ >> >> >> A. Katz wrote: >>> Chris, >>> >>> Thanks for addressing this question. I understand that many, many >>> linguists >>> are quite properly and approriately interested in this ultimate >>> question: >>> "How does the language user work?" (I am also intetested in this >>> question >>> some of the time.) >>> >>> My point is that "how does language work?" is also a valid question, >>> and a >>> central one to the field of linguistics. These two questions are not at >>> all >>> the same. >>> >>> Let me be very explicit: My aim is to separate out the "human" from the >>> "language". There are many good reasons to do so. For anyone working in >>> computerized language processing or in non-human language studies, this >>> is >>> a significant point. >>> >>> It does not matter if a computerized language processing system even >>> remotely simulates what humans do with language in their brains. It >>> does >>> matter whether it comes up with comparable or indistinguishable >>> results. >>> >>> It does not matter whether a parrot, a dolphin or a chimpanzee is doing >>> the >>> same things inside the same modules in his brain as a human does. It >>> does >>> matter if the results are functionally equivalent. >>> >>> We need to make that distinction, between humans and their language, or >>> we >>> will be caught inside a circular definition with no way to break out or >>> to >>> prove anything, not about others and not about ourselves! >>> >>> --Aya >>> >>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>> >>> >>> On Sun, 12 Sep 2010, Chris Butler wrote: >>> >>>> Aya, I think two different things are getting a bit mixed up here. >>>> >>>> I don't for a moment dispute that expressions are often susceptible to >>>> multiple interpretations, that these interpretations are guided by all >>>> kinds of contextual information, or that different people, or even the >>>> same person at different times, may end up selecting differently from >>>> the >>>> various interpretations. Your example 'Open happiness' in another >>>> contribution to this thread illustrates these points very well. >>>> >>>> My point, though, is that each of these different interpretations, as >>>> well >>>> as the selection of one (or more) as more likely in a particular >>>> context, >>>> is achieved through mechanisms in the interpreter's brain which >>>> evolved in >>>> the course of the phylogenetic development of language in the human >>>> species, and developed ontogenetically in that particular >>>> interpreter's >>>> brain. It is surely likely that those mechanisms are highly similar in >>>> different human beings, even though there may be differences in the >>>> detailed wiring in different brains. What I'm saying is that in order >>>> to >>>> answer the question 'How do we communicate using language?' or, if you >>>> prefer, 'How does the language user work?', we need to investigate >>>> what >>>> those mechanisms are, and this is what psycholinguists can help us >>>> with. >>>> >>>> In particular, as linguists, we are interested (well, some of us are, >>>> though clearly not all) in whether the constructs we posit in our >>>> theories >>>> of language have psychological validity in the sense that they >>>> correspond >>>> to ways in which aspects of language are represented in the brain. [As >>>> an >>>> aside, I do realise that there are linguists who strenuously resist >>>> what >>>> they see as a misguided emphasis on mental representation in the work >>>> of >>>> cognitive scientists, but it seems clear that language must be >>>> represented >>>> in some way in the brain in order that we can engage in the >>>> sociosemiotic >>>> acts of meaning making which are the primary focus for many of these >>>> critics.] Taking your 'Open happiness' examples again, I think we need >>>> answers to questions such as: What kind of representation does the >>>> human >>>> language processing system have for 'open' and for 'happiness'? Are >>>> the >>>> phonological, syntactic, semantic and pragmatic (for those who >>>> distinguish >>>> the semantic from the pragmatic) properties of these items (and we >>>> might >>>> want to add 'for this particular interpreter', though there must be >>>> similarities across interpreters for communication to be possible) >>>> represented in the same or different ways, in the same or different >>>> locations (or sets of distributed locations)? Indeed, are we right in >>>> thinking that these familiar levels of linguistic description must be >>>> differentiated, as such, in the human language processing system? Does >>>> the >>>> representation for 'open' distinguish between what we call verbal and >>>> adjectival uses of this item, and if so how? Or are syntactic analyses >>>> computed on the fly, using semantic and contextual clues, rather than >>>> the >>>> neural equivalent of pigeon holes corresponding to verbs and >>>> adjectives? >>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>> both? >>>> All these questions, and many many more, are relevant to the >>>> construction >>>> of a model of language which reflects how language users communicate >>>> (as, >>>> of course, are a whole set of other questions about the sociocultural >>>> aspects of communication). I am not a psycholinguist, but my all too >>>> superficial reading in the area suggests that psycholinguists have >>>> gone >>>> some of the way towards answering some of the questions we might want >>>> to >>>> ask, but that there is still a long way to go. >>>> >>>> Chris >>>> -------------------------------------------------- >>>> From: "A. Katz" >>>> Sent: Saturday, September 11, 2010 3:18 PM >>>> To: "Chris Butler" >>>> Cc: "FUNKNET" >>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>> >>>>> The problem is that once we achieve the psycholinguistic goal and see >>>>> what >>>>> is happening in each person's head, and we see that conflicting >>>>> analyses >>>>> are the norm, rather than the exception, among normal language >>>>> speakers, >>>>> then we will realize that the way language works to transmit >>>>> information >>>>> is despite individual differences and not because of uniform >>>>> processing >>>>> strategies. >>>>> >>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>> not >>>>> process it the same way. Which means that processing is seocndary to >>>>> information transmission. >>>>> >>>>> --Aya >>>>> >>>>> >>>>> >>>>> >>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>> >>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>> question of what kind of thing we think language is: social or >>>>>> individual" is, it seems to me, an important one, particularly for >>>>>> those >>>>>> of us who are committed functionalists. My own view is that a truly >>>>>> functional model of language would be one which aims to account for >>>>>> how >>>>>> human beings communicate using language, or in other words tries to >>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>> now, >>>>>> but which was not tackled head-on in his own work: "How does the >>>>>> natural >>>>>> language user work?' In trying to answer this question we need to >>>>>> accept >>>>>> that language is BOTH social AND individual, and we need to explore >>>>>> both >>>>>> aspects to get as complete a picture as possible of how we >>>>>> communicate >>>>>> using language. We need to know BOTH how people create and respond >>>>>> to >>>>>> meanings and express those meanings in forms during social >>>>>> interaction >>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>> order >>>>>> to make such interaction possible. Both are important parts of the >>>>>> answer to the question 'How do we communicate using language?', >>>>>> though >>>>>> this particular thread of the Funknet discussion has concentrated >>>>>> more >>>>>> on the second aspect, and so will I. >>>>>> >>>>>> This doesn't mean that all the work linguists have done on >>>>>> "exploring >>>>>> the structure of a language so that I can understand how all the >>>>>> bits >>>>>> fit together" and "exploring the connections between items", as Dick >>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>> psycholinguists test are based on ideas about what languages are >>>>>> like. >>>>>> But it does mean, in my view, that ultimately we need to get >>>>>> evidence >>>>>> that the constructs and analyses we propose are ones that are at >>>>>> least >>>>>> consistent with what we know of the processes which go on when we >>>>>> use >>>>>> language. So I am with Matthew when he says that for him, "the only >>>>>> sense in which an analysis can be "the correct analysis" is in terms >>>>>> of >>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>> that >>>>>> linguists should just give up their jobs until such time as we know >>>>>> everything there is to know about language processing. But it does >>>>>> mean >>>>>> that we need to collaborate with psycholinguists, psychologists and >>>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>>> collaborate much more with sociolinguists and sociologists, so that >>>>>> we >>>>>> can get a better handle on the sociocultural aspects of how we >>>>>> communicate.] And it also means that psycholinguists, for their >>>>>> part, >>>>>> need whenever possible to follow up tightly controlled lab >>>>>> experiments >>>>>> with studies under more naturalistic conditions, to avoid the >>>>>> criticism >>>>>> that what happens in artifical lab situations may not happen in >>>>>> natural >>>>>> communicative conditions. >>>>>> >>>>>> I also agree with Dick when he says that "the differences between >>>>>> individuals really matter", and with Lise when she points out that >>>>>> "we >>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>> if it >>>>>> were a single coherent construct that we are trying to discover". >>>>>> However, there are surely processing mechanisms which are common to >>>>>> all >>>>>> language users by virtue of the evolution of the language faculty >>>>>> and >>>>>> which constitute the "general processes" which Dick says >>>>>> psycholinguists >>>>>> are interested in. >>>>>> >>>>>> On the issue of quantitative methodology, I'm sympathetic in general >>>>>> to >>>>>> Ted and Ev's views, though it does seem sensible to prioritise cases >>>>>> in >>>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>>> that we should be giving our university students of linguistics (and >>>>>> some of our linguistics lecturers!) courses in quantitative aspects >>>>>> of >>>>>> linguistics that introduce them to the use of at least some of the >>>>>> basic >>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>> going >>>>>> on in some enlightened places. To those who suspect this can't be >>>>>> done >>>>>> with maths-shy students who don't initially see the need for it, I >>>>>> offer >>>>>> my own experience, over quite a long period, of teaching such >>>>>> courses to >>>>>> people with little or no prior experience in quantitative >>>>>> techniques. >>>>>> For some years in the 1990s, I taught such courses to all >>>>>> linguistics >>>>>> students in an institution where we had many mature students who had >>>>>> come into university level studies with non-standard qualifications, >>>>>> and >>>>>> were not well equipped for courses of this kind by their previous >>>>>> experience. I'm glad to say that teaching the subject from their own >>>>>> perspective as language students rather than that of the >>>>>> statistician, >>>>>> and explaining the reasons for doing things in particular ways >>>>>> rather >>>>>> than just presenting formulae, paid off in the end, so that most >>>>>> students were able to appreciate the relevance of these courses and >>>>>> to >>>>>> turn in very creditable projects showing an understanding of >>>>>> research >>>>>> design and competence in the use of a range of basic statistical >>>>>> techniques. And I still find that bright graduate students respond >>>>>> well >>>>>> to similar courses which incorporate some of the rather more >>>>>> advanced >>>>>> techniques needed for many real research projects in various areas >>>>>> of >>>>>> linguistics. But I may well be out of date with what is now already >>>>>> happening in our fine institutions of higher education! >>>>>> >>>>>> Chris Butler >>>>>> >>>>>> >>>> >>>> >> >> > > -- Mark Mark P. Line Bartlesville, OK From amnfn at well.com Mon Sep 13 00:54:09 2010 From: amnfn at well.com (A. Katz) Date: Sun, 12 Sep 2010 17:54:09 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <2a720add3be870cd8f39d93787201ee3.squirrel@sm.webmail.pair.com> Message-ID: Mark, Sorry, but "eating" is not a tool. It's a biological process. Eating cannot occur outside the biological context. A human being can avoid eating, but if so he starves. Feral children do not speak, but they eat, like everyone else, or they die. Eating does not have to be taught, there is no critical age of acquisition and it is not uniquely an artefact of human culture. If you are an animal, you eat. Eating cannot survive the death of the eater. Language can. Language can be transmitted over great distances by many different artificial methods. Someone's words may be heard long after he has died. They may be passed down verbatim from one generation to the next. Language survives its speakers. It can be re-established and re-used thousands of years after the last native speaker died. It can be generated by computers. It can be transmitted using light, sound, electrical pulses. It can be studied in the absence of speakers. The sensual modalities by which language can be transmitted are not themselbves language. Language is in the contrasts. It's a very abstract tool, not unlike morse code or music theory or mathematics. The fact that it's abstract is the biggest hurdle to understanding that language is a tool. But culture is abstract, too, and surely we can see that it's not a biologcally inherited property like eating or breathing or sleeping. You might as well say that a person cannot possibly avoid watching TV once he's exposed to it, as say the same about language. But people can survive just fine without television, and unless someone shows them how to make a TV set, most people will never figure out how to build one. The same goes for language. We're great at using it, not so great at generating it out of thin air with no ambient culture. --Aya On Sun, 12 Sep 2010, Mark P. Line wrote: > Aya -- > > You said: "Other tools can be studied separately from the people who use > them or created them. Why not language?" > > Although language can certainly be considered a tool, I think it's unlike > other tools in several very significant respects. > > 1. Although we're not born with language, we can't avoid (pathologies > excluded) acquiring it unless we're not exposed to it. To that extent, > language is a biological phenomenon. A prototypical tool is not a > biological phenomenon, so I'm not sure how valid any conclusions might be > that are drawn from a premise of language-as-tool. > > 2. A tool is as a tool does: Anything is a tool that is being used as a > tool, including dead wombats, broken screwdrivers or decks of playing > cards. (Completing the imagined scenarios is left as an exercise for the > reader...) So saying that language is a tool is only saying that language > is used as a tool. Quite a few conclusions can be and have been drawn from > the fact that language is used as a tool, but I would have to be convinced > in detail that almost everything worth knowing about language is dependent > on the premise of language-as-tool. > > 3. If language is a "tool" for (say) communicating ideas, then eating is a > "tool" for reducing hunger. In both cases, I worry about the tool metaphor > being stretched so far from the prototype that we're left with an > out-and-out category fallacy. > > 4. More prototypical tools can be studied separately from those who use or > create them because those tools are easily observed separately from those > who use or create them. I don't think the same thing can be said of > language -- very little about language can be observed apart from its use, > so very little about language can be observed apart from its user(s). > > 5. Any proposal to study something as complex as language separately from > its embodiment is suspicious to me, smacking of reductionism -- something > up with which linguistics has had to put a tad much. Anything that puts > language back into its human context would be a step forward. > > > -- Mark > > Mark P. Line > > > > A. Katz wrote: >> Tom, >> >> I understand the uncomfortable association with Chomsky that speaking of >> language apart from people can have. Competence versus performance, the >> way Chomsky used those terms, never made sense. But that's precisely >> because >> he associated "competence" with the brain and suggested that it was hard >> wired there -- when there was never any evidence of that. >> >> However, if we don't distinguish language from humans, and language >> processing from language data, then how are we going to judge artificial >> language-using devices as to their efficacy at producing and interpreting >> language? How are we going to determine whether and to what extent a >> non-human has acquired language? >> >> We aren't born with it. We don't embody it. It's a tool that we use to >> communicate. Other tools can be studied separately from the people who use >> them or created them. Why not language? >> >> --Aya >> >> >> >> >> On Sun, 12 Sep 2010, Tom Givon wrote: >> >>> >>> >>> >>> I wonder whether asking "how does language work?" is really a meaningful >>> question without asking "how does the language user work?" One of the >>> worst >>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>> distinction >>> of competence ("language", "knowledge") vs. performance ("language >>> user", >>> "processing"). It purported to limit linguists to the armchair methods >>> that >>> study competence, and relegated to psychology the quantified, >>> distributional/variationist methods that study behavior, processing and >>> on-line communication. The first breach in this artificial >>> methodological >>> wall occurred, leastwise for some of us, when we discovered the >>> intermediate >>> method of quantified studies of text, interaction, and conversation. As >>> an >>> ex-biologist, I am forever puzzled by the methodological purism we sill >>> seem >>> to embrace in linguistics, in the face of the manifest complexity and >>> connectivity of language (mind, brain, culture, sociality, evolution, >>> etc.). >>> In biology, another extremely complex science with multiple connections >>> (chemistry, geology, paleontology, behavior, sociality, economics, >>> evolution, >>> etc.), ANY method is welcome so long as it does the job of furthering >>> our >>> understanding. And by understanding we mean ever-wider circles of >>> connectivity. >>> >>> Best, TG >>> ================ >>> >>> >>> A. Katz wrote: >>>> Chris, >>>> >>>> Thanks for addressing this question. I understand that many, many >>>> linguists >>>> are quite properly and approriately interested in this ultimate >>>> question: >>>> "How does the language user work?" (I am also intetested in this >>>> question >>>> some of the time.) >>>> >>>> My point is that "how does language work?" is also a valid question, >>>> and a >>>> central one to the field of linguistics. These two questions are not at >>>> all >>>> the same. >>>> >>>> Let me be very explicit: My aim is to separate out the "human" from the >>>> "language". There are many good reasons to do so. For anyone working in >>>> computerized language processing or in non-human language studies, this >>>> is >>>> a significant point. >>>> >>>> It does not matter if a computerized language processing system even >>>> remotely simulates what humans do with language in their brains. It >>>> does >>>> matter whether it comes up with comparable or indistinguishable >>>> results. >>>> >>>> It does not matter whether a parrot, a dolphin or a chimpanzee is doing >>>> the >>>> same things inside the same modules in his brain as a human does. It >>>> does >>>> matter if the results are functionally equivalent. >>>> >>>> We need to make that distinction, between humans and their language, or >>>> we >>>> will be caught inside a circular definition with no way to break out or >>>> to >>>> prove anything, not about others and not about ourselves! >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>> >>>> >>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>> >>>>> Aya, I think two different things are getting a bit mixed up here. >>>>> >>>>> I don't for a moment dispute that expressions are often susceptible to >>>>> multiple interpretations, that these interpretations are guided by all >>>>> kinds of contextual information, or that different people, or even the >>>>> same person at different times, may end up selecting differently from >>>>> the >>>>> various interpretations. Your example 'Open happiness' in another >>>>> contribution to this thread illustrates these points very well. >>>>> >>>>> My point, though, is that each of these different interpretations, as >>>>> well >>>>> as the selection of one (or more) as more likely in a particular >>>>> context, >>>>> is achieved through mechanisms in the interpreter's brain which >>>>> evolved in >>>>> the course of the phylogenetic development of language in the human >>>>> species, and developed ontogenetically in that particular >>>>> interpreter's >>>>> brain. It is surely likely that those mechanisms are highly similar in >>>>> different human beings, even though there may be differences in the >>>>> detailed wiring in different brains. What I'm saying is that in order >>>>> to >>>>> answer the question 'How do we communicate using language?' or, if you >>>>> prefer, 'How does the language user work?', we need to investigate >>>>> what >>>>> those mechanisms are, and this is what psycholinguists can help us >>>>> with. >>>>> >>>>> In particular, as linguists, we are interested (well, some of us are, >>>>> though clearly not all) in whether the constructs we posit in our >>>>> theories >>>>> of language have psychological validity in the sense that they >>>>> correspond >>>>> to ways in which aspects of language are represented in the brain. [As >>>>> an >>>>> aside, I do realise that there are linguists who strenuously resist >>>>> what >>>>> they see as a misguided emphasis on mental representation in the work >>>>> of >>>>> cognitive scientists, but it seems clear that language must be >>>>> represented >>>>> in some way in the brain in order that we can engage in the >>>>> sociosemiotic >>>>> acts of meaning making which are the primary focus for many of these >>>>> critics.] Taking your 'Open happiness' examples again, I think we need >>>>> answers to questions such as: What kind of representation does the >>>>> human >>>>> language processing system have for 'open' and for 'happiness'? Are >>>>> the >>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>> distinguish >>>>> the semantic from the pragmatic) properties of these items (and we >>>>> might >>>>> want to add 'for this particular interpreter', though there must be >>>>> similarities across interpreters for communication to be possible) >>>>> represented in the same or different ways, in the same or different >>>>> locations (or sets of distributed locations)? Indeed, are we right in >>>>> thinking that these familiar levels of linguistic description must be >>>>> differentiated, as such, in the human language processing system? Does >>>>> the >>>>> representation for 'open' distinguish between what we call verbal and >>>>> adjectival uses of this item, and if so how? Or are syntactic analyses >>>>> computed on the fly, using semantic and contextual clues, rather than >>>>> the >>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>> adjectives? >>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>> both? >>>>> All these questions, and many many more, are relevant to the >>>>> construction >>>>> of a model of language which reflects how language users communicate >>>>> (as, >>>>> of course, are a whole set of other questions about the sociocultural >>>>> aspects of communication). I am not a psycholinguist, but my all too >>>>> superficial reading in the area suggests that psycholinguists have >>>>> gone >>>>> some of the way towards answering some of the questions we might want >>>>> to >>>>> ask, but that there is still a long way to go. >>>>> >>>>> Chris >>>>> -------------------------------------------------- >>>>> From: "A. Katz" >>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>> To: "Chris Butler" >>>>> Cc: "FUNKNET" >>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>> >>>>>> The problem is that once we achieve the psycholinguistic goal and see >>>>>> what >>>>>> is happening in each person's head, and we see that conflicting >>>>>> analyses >>>>>> are the norm, rather than the exception, among normal language >>>>>> speakers, >>>>>> then we will realize that the way language works to transmit >>>>>> information >>>>>> is despite individual differences and not because of uniform >>>>>> processing >>>>>> strategies. >>>>>> >>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>> not >>>>>> process it the same way. Which means that processing is seocndary to >>>>>> information transmission. >>>>>> >>>>>> --Aya >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>> >>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>> question of what kind of thing we think language is: social or >>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>> those >>>>>>> of us who are committed functionalists. My own view is that a truly >>>>>>> functional model of language would be one which aims to account for >>>>>>> how >>>>>>> human beings communicate using language, or in other words tries to >>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>> now, >>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>> natural >>>>>>> language user work?' In trying to answer this question we need to >>>>>>> accept >>>>>>> that language is BOTH social AND individual, and we need to explore >>>>>>> both >>>>>>> aspects to get as complete a picture as possible of how we >>>>>>> communicate >>>>>>> using language. We need to know BOTH how people create and respond >>>>>>> to >>>>>>> meanings and express those meanings in forms during social >>>>>>> interaction >>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>> order >>>>>>> to make such interaction possible. Both are important parts of the >>>>>>> answer to the question 'How do we communicate using language?', >>>>>>> though >>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>> more >>>>>>> on the second aspect, and so will I. >>>>>>> >>>>>>> This doesn't mean that all the work linguists have done on >>>>>>> "exploring >>>>>>> the structure of a language so that I can understand how all the >>>>>>> bits >>>>>>> fit together" and "exploring the connections between items", as Dick >>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>> like. >>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>> evidence >>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>> least >>>>>>> consistent with what we know of the processes which go on when we >>>>>>> use >>>>>>> language. So I am with Matthew when he says that for him, "the only >>>>>>> sense in which an analysis can be "the correct analysis" is in terms >>>>>>> of >>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>> that >>>>>>> linguists should just give up their jobs until such time as we know >>>>>>> everything there is to know about language processing. But it does >>>>>>> mean >>>>>>> that we need to collaborate with psycholinguists, psychologists and >>>>>>> neurologists, as has also been pointed out by linguists such as Ray >>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need to >>>>>>> collaborate much more with sociolinguists and sociologists, so that >>>>>>> we >>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>> part, >>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>> experiments >>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>> criticism >>>>>>> that what happens in artifical lab situations may not happen in >>>>>>> natural >>>>>>> communicative conditions. >>>>>>> >>>>>>> I also agree with Dick when he says that "the differences between >>>>>>> individuals really matter", and with Lise when she points out that >>>>>>> "we >>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>> if it >>>>>>> were a single coherent construct that we are trying to discover". >>>>>>> However, there are surely processing mechanisms which are common to >>>>>>> all >>>>>>> language users by virtue of the evolution of the language faculty >>>>>>> and >>>>>>> which constitute the "general processes" which Dick says >>>>>>> psycholinguists >>>>>>> are interested in. >>>>>>> >>>>>>> On the issue of quantitative methodology, I'm sympathetic in general >>>>>>> to >>>>>>> Ted and Ev's views, though it does seem sensible to prioritise cases >>>>>>> in >>>>>>> terms of a hierarchy such as Brian suggests. One thing this means is >>>>>>> that we should be giving our university students of linguistics (and >>>>>>> some of our linguistics lecturers!) courses in quantitative aspects >>>>>>> of >>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>> basic >>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>> going >>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>> done >>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>> offer >>>>>>> my own experience, over quite a long period, of teaching such >>>>>>> courses to >>>>>>> people with little or no prior experience in quantitative >>>>>>> techniques. >>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>> linguistics >>>>>>> students in an institution where we had many mature students who had >>>>>>> come into university level studies with non-standard qualifications, >>>>>>> and >>>>>>> were not well equipped for courses of this kind by their previous >>>>>>> experience. I'm glad to say that teaching the subject from their own >>>>>>> perspective as language students rather than that of the >>>>>>> statistician, >>>>>>> and explaining the reasons for doing things in particular ways >>>>>>> rather >>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>> students were able to appreciate the relevance of these courses and >>>>>>> to >>>>>>> turn in very creditable projects showing an understanding of >>>>>>> research >>>>>>> design and competence in the use of a range of basic statistical >>>>>>> techniques. And I still find that bright graduate students respond >>>>>>> well >>>>>>> to similar courses which incorporate some of the rather more >>>>>>> advanced >>>>>>> techniques needed for many real research projects in various areas >>>>>>> of >>>>>>> linguistics. But I may well be out of date with what is now already >>>>>>> happening in our fine institutions of higher education! >>>>>>> >>>>>>> Chris Butler >>>>>>> >>>>>>> >>>>> >>>>> >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From wilcox at unm.edu Mon Sep 13 02:11:14 2010 From: wilcox at unm.edu (Sherman Wilcox) Date: Sun, 12 Sep 2010 20:11:14 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: On Sep 12, 2010, at 6:54 PM, "A. Katz" wrote: > Sorry, but "eating" is not a tool. It's a biological process. So is language. Sherman Wilcox University of New Mexico From cbutler at ntlworld.com Mon Sep 13 09:50:07 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Mon, 13 Sep 2010 10:50:07 +0100 Subject: FUNKNET] analysis: unhappiness Message-ID: I sympathise with Tom's view that the question 'How does language work?' may not be meaningful, or may at least not be an appropriate question to ask, independently of the question 'How do language users work?' The complex human activity that we call language exists only through the capabilities and activities of its users, pace Aya's comment that "we do not embody it". Furthermore, there is a considerable body of opinion nowadays, not only among linguists interested in sociosemiotic aspects of language but also in the work of many cognitive linguists and of psycholinguists such as Ray Gibbs, which agrees with Mark's suspicion of "any proposal to study something as complex as language separately from its embodiment". One problem with a strict separation between language and humans, such as Aya would like us to accept, is that if we do this we deny ourselves natural explanations of many phenomena in human language which are otherwise somewhat mysterious. As Tom astutely observed more than 30 years ago, many structural properties of language are most insightfully accounted for in terms of a set of explanatory parameters many of which are concerned with the properties of the human language-processing organism, including general cognitive and perceptual structure, the specific structure of the neurological, acoustic, articulatory and other mechanisms, the ontogenetic development of the young child, etc. Aya is also concerned about the need to distinguish language processing from language data. But those of us who want to ask the question 'How does the language user work?' do make this distinction. The data are the input to language understanding and the output from language production, so the two are distinct. I understand Aya's point that we need some way of evaluating what computers and non-humans can do. But I think it is dangerous to put this in terms of "artifical language-using devices" and "to what extent a non-human has acquired language", as if there were some monolithic entity out there called language, which humans have and computers and non-humans may have to some extent. A more fruitful way to approach the situation, in my view, would be to study how and why what can be produced (and also understood) by a non-human animal or a computer is similar to and different from what a human being can produce and understand, looking at the systems in the round, including analyses not only of structures but also of the range of communicative and social functions they perform, but without assuming that what we are looking at is a single entity 'language', present to variable extents. Sorry if I haven't put this last bit clearly enough - I found it quite hard to formulate without making too many unwarranted assumptions! Chris From kuzar at research.haifa.ac.il Mon Sep 13 10:01:54 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Mon, 13 Sep 2010 12:01:54 +0200 Subject: Methodological MA workshop Message-ID: Posted to Funknet and Cogling. -------- Dear fellow linguists, I have been assigned to teach a methodological MA workshop in linguistics (officially titled "Approaches to research in linguistics", however it is in the Department of English, so mainly English linguistics). Presumably, students have taken "Style & composition" during their BA, where they have learned the basics of academic writing. It is a two semester course meeting once a week for 2 academic hours, so there will be some 26-28 classes. The seminar is intended to help students at the MA level in several areas: 1. Choosing a research domain, formulating a research question, choosing the appropriate methodology. 2. The different parts of a research paper, how to develop an argument (probably the hardest of all tasks), how to write the introduction. What's the difference between presenting the findings/results and discussing/interpreting them? 3. Corpora: what English corpora are are available, search strategy and search tools, regular expressions, organization and manipulation of data (e.g. in Excel). 4. The LLBA as a resource (believe it or not: some students have made it to the MA level without knowing about the LLBA). 5. Organizing bibliographical data and note-taking files (Refworks, NotaBene). 6. Preparing students for linguistic presentations by reading and discussing an article by that linguist beforehand. If any of you have taught a similar course, I'd be very grateful to hear from you. Are there any other points that should be covered? Do you have any practical suggestions for points 1. and 2. or any literature that might be consulted? As for point 3: does anybody know of a user friendly introduction (or crash-course) to regular expressions at a rudimentary level for non-computationally oriented students? Looking forward to your advice. Ron Kuzar =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From dan at daneverett.org Mon Sep 13 12:45:43 2010 From: dan at daneverett.org (Daniel Everett) Date: Mon, 13 Sep 2010 08:45:43 -0400 Subject: Language as a Tool Message-ID: Dear Mark, These are all excellent points. Clearly there is some biology that must underwrite language, or plants could speak. The question is, how much. Less than eating. More than wearing a tie, perhaps. I think that all of your points, however, are compatible with the idea that language is a tool, so long as we don't think that, as you say, this explains everything. It does, however, explain more than many have recognized. My new book on this, Cognitive Fire: Language as a Cultural Tool, is due out from Pantheon (US) and Profile (UK) sometime in 2011. Hopefully, I will have answered your questions. -- Dan On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: > Aya -- > > You said: "Other tools can be studied separately from the people who use > them or created them. Why not language?" > > Although language can certainly be considered a tool, I think it's unlike > other tools in several very significant respects. > > 1. Although we're not born with language, we can't avoid (pathologies > excluded) acquiring it unless we're not exposed to it. To that extent, > language is a biological phenomenon. A prototypical tool is not a > biological phenomenon, so I'm not sure how valid any conclusions might be > that are drawn from a premise of language-as-tool. > > 2. A tool is as a tool does: Anything is a tool that is being used as a > tool, including dead wombats, broken screwdrivers or decks of playing > cards. (Completing the imagined scenarios is left as an exercise for the > reader...) So saying that language is a tool is only saying that language > is used as a tool. Quite a few conclusions can be and have been drawn from > the fact that language is used as a tool, but I would have to be convinced > in detail that almost everything worth knowing about language is dependent > on the premise of language-as-tool. > > 3. If language is a "tool" for (say) communicating ideas, then eating is a > "tool" for reducing hunger. In both cases, I worry about the tool metaphor > being stretched so far from the prototype that we're left with an > out-and-out category fallacy. > > 4. More prototypical tools can be studied separately from those who use or > create them because those tools are easily observed separately from those > who use or create them. I don't think the same thing can be said of > language -- very little about language can be observed apart from its use, > so very little about language can be observed apart from its user(s). > > 5. Any proposal to study something as complex as language separately from > its embodiment is suspicious to me, smacking of reductionism -- something > up with which linguistics has had to put a tad much. Anything that puts > language back into its human context would be a step forward. > > > -- Mark > > Mark P. Line > > From mark at polymathix.com Mon Sep 13 12:46:45 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 07:46:45 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: Aya -- A. Katz wrote: > Mark, > > Sorry, but "eating" is not a tool. It's a biological process. > > Eating cannot occur outside the biological context. A human being can > avoid eating, but if so he starves. Feral children do not speak, but > they eat, like everyone else, or they die. Eating does not have to be > taught, there is no critical age of acquisition and it is not uniquely an > artefact of human culture. If you are an animal, you eat. Eating cannot > survive the death of the eater. > > Language can. I would say that not language, but the artifacts of language (texts, audio recordings) can survive the people who created them, because I try very hard not to reify the artifacts of language as "language". The (usually communicative) process I refer to as language cannot exist independently of its embodiment. That said, I don't care if the embodiment is human, computer, cetacean or non-human primate. -- Mark Mark P. Line > On Sun, 12 Sep 2010, Mark P. Line wrote: > >> Aya -- >> >> You said: "Other tools can be studied separately from the people who use >> them or created them. Why not language?" >> >> Although language can certainly be considered a tool, I think it's >> unlike >> other tools in several very significant respects. >> >> 1. Although we're not born with language, we can't avoid (pathologies >> excluded) acquiring it unless we're not exposed to it. To that extent, >> language is a biological phenomenon. A prototypical tool is not a >> biological phenomenon, so I'm not sure how valid any conclusions might >> be >> that are drawn from a premise of language-as-tool. >> >> 2. A tool is as a tool does: Anything is a tool that is being used as a >> tool, including dead wombats, broken screwdrivers or decks of playing >> cards. (Completing the imagined scenarios is left as an exercise for the >> reader...) So saying that language is a tool is only saying that >> language >> is used as a tool. Quite a few conclusions can be and have been drawn >> from >> the fact that language is used as a tool, but I would have to be >> convinced >> in detail that almost everything worth knowing about language is >> dependent >> on the premise of language-as-tool. >> >> 3. If language is a "tool" for (say) communicating ideas, then eating is >> a >> "tool" for reducing hunger. In both cases, I worry about the tool >> metaphor >> being stretched so far from the prototype that we're left with an >> out-and-out category fallacy. >> >> 4. More prototypical tools can be studied separately from those who use >> or >> create them because those tools are easily observed separately from >> those >> who use or create them. I don't think the same thing can be said of >> language -- very little about language can be observed apart from its >> use, >> so very little about language can be observed apart from its user(s). >> >> 5. Any proposal to study something as complex as language separately >> from >> its embodiment is suspicious to me, smacking of reductionism -- >> something >> up with which linguistics has had to put a tad much. Anything that puts >> language back into its human context would be a step forward. >> >> >> -- Mark >> >> Mark P. Line >> >> >> >> A. Katz wrote: >>> Tom, >>> >>> I understand the uncomfortable association with Chomsky that speaking >>> of >>> language apart from people can have. Competence versus performance, the >>> way Chomsky used those terms, never made sense. But that's precisely >>> because >>> he associated "competence" with the brain and suggested that it was >>> hard >>> wired there -- when there was never any evidence of that. >>> >>> However, if we don't distinguish language from humans, and language >>> processing from language data, then how are we going to judge >>> artificial >>> language-using devices as to their efficacy at producing and >>> interpreting >>> language? How are we going to determine whether and to what extent a >>> non-human has acquired language? >>> >>> We aren't born with it. We don't embody it. It's a tool that we use to >>> communicate. Other tools can be studied separately from the people who >>> use >>> them or created them. Why not language? >>> >>> --Aya >>> >>> >>> >>> >>> On Sun, 12 Sep 2010, Tom Givon wrote: >>> >>>> >>>> >>>> >>>> I wonder whether asking "how does language work?" is really a >>>> meaningful >>>> question without asking "how does the language user work?" One of >>>> the >>>> worst >>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>> distinction >>>> of competence ("language", "knowledge") vs. performance ("language >>>> user", >>>> "processing"). It purported to limit linguists to the armchair methods >>>> that >>>> study competence, and relegated to psychology the quantified, >>>> distributional/variationist methods that study behavior, processing >>>> and >>>> on-line communication. The first breach in this artificial >>>> methodological >>>> wall occurred, leastwise for some of us, when we discovered the >>>> intermediate >>>> method of quantified studies of text, interaction, and conversation. >>>> As >>>> an >>>> ex-biologist, I am forever puzzled by the methodological purism we >>>> sill >>>> seem >>>> to embrace in linguistics, in the face of the manifest complexity and >>>> connectivity of language (mind, brain, culture, sociality, evolution, >>>> etc.). >>>> In biology, another extremely complex science with multiple >>>> connections >>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>> evolution, >>>> etc.), ANY method is welcome so long as it does the job of furthering >>>> our >>>> understanding. And by understanding we mean ever-wider circles of >>>> connectivity. >>>> >>>> Best, TG >>>> ================ >>>> >>>> >>>> A. Katz wrote: >>>>> Chris, >>>>> >>>>> Thanks for addressing this question. I understand that many, many >>>>> linguists >>>>> are quite properly and approriately interested in this ultimate >>>>> question: >>>>> "How does the language user work?" (I am also intetested in this >>>>> question >>>>> some of the time.) >>>>> >>>>> My point is that "how does language work?" is also a valid question, >>>>> and a >>>>> central one to the field of linguistics. These two questions are not >>>>> at >>>>> all >>>>> the same. >>>>> >>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>> the >>>>> "language". There are many good reasons to do so. For anyone working >>>>> in >>>>> computerized language processing or in non-human language studies, >>>>> this >>>>> is >>>>> a significant point. >>>>> >>>>> It does not matter if a computerized language processing system even >>>>> remotely simulates what humans do with language in their brains. It >>>>> does >>>>> matter whether it comes up with comparable or indistinguishable >>>>> results. >>>>> >>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>> doing >>>>> the >>>>> same things inside the same modules in his brain as a human does. It >>>>> does >>>>> matter if the results are functionally equivalent. >>>>> >>>>> We need to make that distinction, between humans and their language, >>>>> or >>>>> we >>>>> will be caught inside a circular definition with no way to break out >>>>> or >>>>> to >>>>> prove anything, not about others and not about ourselves! >>>>> >>>>> --Aya >>>>> >>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>> >>>>> >>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>> >>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>> >>>>>> I don't for a moment dispute that expressions are often susceptible >>>>>> to >>>>>> multiple interpretations, that these interpretations are guided by >>>>>> all >>>>>> kinds of contextual information, or that different people, or even >>>>>> the >>>>>> same person at different times, may end up selecting differently >>>>>> from >>>>>> the >>>>>> various interpretations. Your example 'Open happiness' in another >>>>>> contribution to this thread illustrates these points very well. >>>>>> >>>>>> My point, though, is that each of these different interpretations, >>>>>> as >>>>>> well >>>>>> as the selection of one (or more) as more likely in a particular >>>>>> context, >>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>> evolved in >>>>>> the course of the phylogenetic development of language in the human >>>>>> species, and developed ontogenetically in that particular >>>>>> interpreter's >>>>>> brain. It is surely likely that those mechanisms are highly similar >>>>>> in >>>>>> different human beings, even though there may be differences in the >>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>> order >>>>>> to >>>>>> answer the question 'How do we communicate using language?' or, if >>>>>> you >>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>> what >>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>> with. >>>>>> >>>>>> In particular, as linguists, we are interested (well, some of us >>>>>> are, >>>>>> though clearly not all) in whether the constructs we posit in our >>>>>> theories >>>>>> of language have psychological validity in the sense that they >>>>>> correspond >>>>>> to ways in which aspects of language are represented in the brain. >>>>>> [As >>>>>> an >>>>>> aside, I do realise that there are linguists who strenuously resist >>>>>> what >>>>>> they see as a misguided emphasis on mental representation in the >>>>>> work >>>>>> of >>>>>> cognitive scientists, but it seems clear that language must be >>>>>> represented >>>>>> in some way in the brain in order that we can engage in the >>>>>> sociosemiotic >>>>>> acts of meaning making which are the primary focus for many of these >>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>> need >>>>>> answers to questions such as: What kind of representation does the >>>>>> human >>>>>> language processing system have for 'open' and for 'happiness'? Are >>>>>> the >>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>> distinguish >>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>> might >>>>>> want to add 'for this particular interpreter', though there must be >>>>>> similarities across interpreters for communication to be possible) >>>>>> represented in the same or different ways, in the same or different >>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>> in >>>>>> thinking that these familiar levels of linguistic description must >>>>>> be >>>>>> differentiated, as such, in the human language processing system? >>>>>> Does >>>>>> the >>>>>> representation for 'open' distinguish between what we call verbal >>>>>> and >>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>> analyses >>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>> than >>>>>> the >>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>> adjectives? >>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>>> both? >>>>>> All these questions, and many many more, are relevant to the >>>>>> construction >>>>>> of a model of language which reflects how language users communicate >>>>>> (as, >>>>>> of course, are a whole set of other questions about the >>>>>> sociocultural >>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>> too >>>>>> superficial reading in the area suggests that psycholinguists have >>>>>> gone >>>>>> some of the way towards answering some of the questions we might >>>>>> want >>>>>> to >>>>>> ask, but that there is still a long way to go. >>>>>> >>>>>> Chris >>>>>> -------------------------------------------------- >>>>>> From: "A. Katz" >>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>> To: "Chris Butler" >>>>>> Cc: "FUNKNET" >>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>> >>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>> see >>>>>>> what >>>>>>> is happening in each person's head, and we see that conflicting >>>>>>> analyses >>>>>>> are the norm, rather than the exception, among normal language >>>>>>> speakers, >>>>>>> then we will realize that the way language works to transmit >>>>>>> information >>>>>>> is despite individual differences and not because of uniform >>>>>>> processing >>>>>>> strategies. >>>>>>> >>>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>>> not >>>>>>> process it the same way. Which means that processing is seocndary >>>>>>> to >>>>>>> information transmission. >>>>>>> >>>>>>> --Aya >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>> >>>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>>> question of what kind of thing we think language is: social or >>>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>>> those >>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>> truly >>>>>>>> functional model of language would be one which aims to account >>>>>>>> for >>>>>>>> how >>>>>>>> human beings communicate using language, or in other words tries >>>>>>>> to >>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>> now, >>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>> natural >>>>>>>> language user work?' In trying to answer this question we need to >>>>>>>> accept >>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>> explore >>>>>>>> both >>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>> communicate >>>>>>>> using language. We need to know BOTH how people create and respond >>>>>>>> to >>>>>>>> meanings and express those meanings in forms during social >>>>>>>> interaction >>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>> order >>>>>>>> to make such interaction possible. Both are important parts of the >>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>> though >>>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>>> more >>>>>>>> on the second aspect, and so will I. >>>>>>>> >>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>> "exploring >>>>>>>> the structure of a language so that I can understand how all the >>>>>>>> bits >>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>> Dick >>>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>> like. >>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>> evidence >>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>> least >>>>>>>> consistent with what we know of the processes which go on when we >>>>>>>> use >>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>> only >>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>> terms >>>>>>>> of >>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>> that >>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>> know >>>>>>>> everything there is to know about language processing. But it does >>>>>>>> mean >>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>> and >>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>> Ray >>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need >>>>>>>> to >>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>> that >>>>>>>> we >>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>> part, >>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>> experiments >>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>> criticism >>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>> natural >>>>>>>> communicative conditions. >>>>>>>> >>>>>>>> I also agree with Dick when he says that "the differences between >>>>>>>> individuals really matter", and with Lise when she points out that >>>>>>>> "we >>>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>>> if it >>>>>>>> were a single coherent construct that we are trying to discover". >>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>> to >>>>>>>> all >>>>>>>> language users by virtue of the evolution of the language faculty >>>>>>>> and >>>>>>>> which constitute the "general processes" which Dick says >>>>>>>> psycholinguists >>>>>>>> are interested in. >>>>>>>> >>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>> general >>>>>>>> to >>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>> cases >>>>>>>> in >>>>>>>> terms of a hierarchy such as Brian suggests. One thing this means >>>>>>>> is >>>>>>>> that we should be giving our university students of linguistics >>>>>>>> (and >>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>> aspects >>>>>>>> of >>>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>>> basic >>>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>>> going >>>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>>> done >>>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>>> offer >>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>> courses to >>>>>>>> people with little or no prior experience in quantitative >>>>>>>> techniques. >>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>> linguistics >>>>>>>> students in an institution where we had many mature students who >>>>>>>> had >>>>>>>> come into university level studies with non-standard >>>>>>>> qualifications, >>>>>>>> and >>>>>>>> were not well equipped for courses of this kind by their previous >>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>> own >>>>>>>> perspective as language students rather than that of the >>>>>>>> statistician, >>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>> rather >>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>> and >>>>>>>> to >>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>> research >>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>> techniques. And I still find that bright graduate students respond >>>>>>>> well >>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>> advanced >>>>>>>> techniques needed for many real research projects in various areas >>>>>>>> of >>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>> already >>>>>>>> happening in our fine institutions of higher education! >>>>>>>> >>>>>>>> Chris Butler >>>>>>>> >>>>>>>> >>>>>> >>>>>> >>>> >>>> >>> >>> >> >> >> -- Mark >> >> Mark P. Line >> Bartlesville, OK >> >> > > -- Mark Mark P. Line Bartlesville, OK From mark at polymathix.com Mon Sep 13 12:59:47 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 07:59:47 -0500 Subject: Language as a Tool In-Reply-To: Message-ID: Thanks, Dan. I first got into linguistics through anthropology and never really lost that frame of reference, so I'm very much looking forward to your new book. Please rattle Pantheon's cage to get them to offer a Kindle edition of your book. :) -- Mark Mark P. Line Daniel Everett wrote: > > Dear Mark, > > These are all excellent points. Clearly there is some biology that must > underwrite language, or plants could speak. The question is, how much. > Less than eating. More than wearing a tie, perhaps. > > I think that all of your points, however, are compatible with the idea > that language is a tool, so long as we don't think that, as you say, this > explains everything. It does, however, explain more than many have > recognized. > > My new book on this, Cognitive Fire: Language as a Cultural Tool, is due > out from Pantheon (US) and Profile (UK) sometime in 2011. > > Hopefully, I will have answered your questions. > > -- Dan > > > On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: > >> Aya -- >> >> You said: "Other tools can be studied separately from the people who use >> them or created them. Why not language?" >> >> Although language can certainly be considered a tool, I think it's >> unlike >> other tools in several very significant respects. >> >> 1. Although we're not born with language, we can't avoid (pathologies >> excluded) acquiring it unless we're not exposed to it. To that extent, >> language is a biological phenomenon. A prototypical tool is not a >> biological phenomenon, so I'm not sure how valid any conclusions might >> be >> that are drawn from a premise of language-as-tool. >> >> 2. A tool is as a tool does: Anything is a tool that is being used as a >> tool, including dead wombats, broken screwdrivers or decks of playing >> cards. (Completing the imagined scenarios is left as an exercise for the >> reader...) So saying that language is a tool is only saying that >> language >> is used as a tool. Quite a few conclusions can be and have been drawn >> from >> the fact that language is used as a tool, but I would have to be >> convinced >> in detail that almost everything worth knowing about language is >> dependent >> on the premise of language-as-tool. >> >> 3. If language is a "tool" for (say) communicating ideas, then eating is >> a >> "tool" for reducing hunger. In both cases, I worry about the tool >> metaphor >> being stretched so far from the prototype that we're left with an >> out-and-out category fallacy. >> >> 4. More prototypical tools can be studied separately from those who use >> or >> create them because those tools are easily observed separately from >> those >> who use or create them. I don't think the same thing can be said of >> language -- very little about language can be observed apart from its >> use, >> so very little about language can be observed apart from its user(s). >> >> 5. Any proposal to study something as complex as language separately >> from >> its embodiment is suspicious to me, smacking of reductionism -- >> something >> up with which linguistics has had to put a tad much. Anything that puts >> language back into its human context would be a step forward. >> >> >> -- Mark >> >> Mark P. Line >> >> > > -- Mark Mark P. Line Bartlesville, OK From dlevere at ilstu.edu Mon Sep 13 13:07:38 2010 From: dlevere at ilstu.edu (Daniel Everett) Date: Mon, 13 Sep 2010 09:07:38 -0400 Subject: Language as a Tool In-Reply-To: Message-ID: Mark, Well, ahem, there *is* a kindle edition of Don't sleep there are snakes. There should also be one of Cognitive Fire. Dan http://www.amazon.com/Dont-Sleep-There-Snakes-ebook/dp/B0037Z8SMC/ref=tmm_kin_title_0?ie=UTF8&m=AG56TWVU5XWC2 On Sep 13, 2010, at 8:59 AM, Mark P. Line wrote: > Thanks, Dan. I first got into linguistics through anthropology and never > really lost that frame of reference, so I'm very much looking forward to > your new book. > > Please rattle Pantheon's cage to get them to offer a Kindle edition of > your book. :) > > -- Mark > > Mark P. Line > > > > Daniel Everett wrote: >> >> Dear Mark, >> >> These are all excellent points. Clearly there is some biology that must >> underwrite language, or plants could speak. The question is, how much. >> Less than eating. More than wearing a tie, perhaps. >> >> I think that all of your points, however, are compatible with the idea >> that language is a tool, so long as we don't think that, as you say, this >> explains everything. It does, however, explain more than many have >> recognized. >> >> My new book on this, Cognitive Fire: Language as a Cultural Tool, is due >> out from Pantheon (US) and Profile (UK) sometime in 2011. >> >> Hopefully, I will have answered your questions. >> >> -- Dan >> >> >> On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > From amnfn at well.com Mon Sep 13 13:17:28 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 06:17:28 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <149a0b7cd376d6c0f392a7d827ef7985.squirrel@sm.webmail.pair.com> Message-ID: Mark, Okay, I'm glad that you agree that a language user might be a machine or a non-human. That brings us much closer together. Nevertheless, there is also the fact that language can be preserved, but not used, for thousands of years, and then people start using it again. If it were an biological process, or even something like a live virus passed down from parent to child, then reviving it once it was dead would be impossible. But it's not. It's a tool. A tool that a group of people can stop using and making, but keep the blue prints for, and then make it again when they choose to. I wonder how many linguists realize that Hebrew was revived from a dead language. How many think it is just a slavic language with Hebrew lexemes? How many realize that the grammar, the triliteral roots, the derivational system, are all working, and that any changes in pronunciation are of little importance when the basic contrasts are preserved? Mark, if you can see that a language can be used by a non-human, then why not also acknowledge that it's a system for transferring information, and that the code can be preserved while in disuse? --Aya On Mon, 13 Sep 2010, Mark P. Line wrote: > Aya -- > > A. Katz wrote: >> Mark, >> >> Sorry, but "eating" is not a tool. It's a biological process. >> >> Eating cannot occur outside the biological context. A human being can >> avoid eating, but if so he starves. Feral children do not speak, but >> they eat, like everyone else, or they die. Eating does not have to be >> taught, there is no critical age of acquisition and it is not uniquely an >> artefact of human culture. If you are an animal, you eat. Eating cannot >> survive the death of the eater. >> >> Language can. > > I would say that not language, but the artifacts of language (texts, audio > recordings) can survive the people who created them, because I try very > hard not to reify the artifacts of language as "language". > > The (usually communicative) process I refer to as language cannot exist > independently of its embodiment. That said, I don't care if the embodiment > is human, computer, cetacean or non-human primate. > > > -- Mark > > Mark P. Line > > >> On Sun, 12 Sep 2010, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >>> >>> A. Katz wrote: >>>> Tom, >>>> >>>> I understand the uncomfortable association with Chomsky that speaking >>>> of >>>> language apart from people can have. Competence versus performance, the >>>> way Chomsky used those terms, never made sense. But that's precisely >>>> because >>>> he associated "competence" with the brain and suggested that it was >>>> hard >>>> wired there -- when there was never any evidence of that. >>>> >>>> However, if we don't distinguish language from humans, and language >>>> processing from language data, then how are we going to judge >>>> artificial >>>> language-using devices as to their efficacy at producing and >>>> interpreting >>>> language? How are we going to determine whether and to what extent a >>>> non-human has acquired language? >>>> >>>> We aren't born with it. We don't embody it. It's a tool that we use to >>>> communicate. Other tools can be studied separately from the people who >>>> use >>>> them or created them. Why not language? >>>> >>>> --Aya >>>> >>>> >>>> >>>> >>>> On Sun, 12 Sep 2010, Tom Givon wrote: >>>> >>>>> >>>>> >>>>> >>>>> I wonder whether asking "how does language work?" is really a >>>>> meaningful >>>>> question without asking "how does the language user work?" One of >>>>> the >>>>> worst >>>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>>> distinction >>>>> of competence ("language", "knowledge") vs. performance ("language >>>>> user", >>>>> "processing"). It purported to limit linguists to the armchair methods >>>>> that >>>>> study competence, and relegated to psychology the quantified, >>>>> distributional/variationist methods that study behavior, processing >>>>> and >>>>> on-line communication. The first breach in this artificial >>>>> methodological >>>>> wall occurred, leastwise for some of us, when we discovered the >>>>> intermediate >>>>> method of quantified studies of text, interaction, and conversation. >>>>> As >>>>> an >>>>> ex-biologist, I am forever puzzled by the methodological purism we >>>>> sill >>>>> seem >>>>> to embrace in linguistics, in the face of the manifest complexity and >>>>> connectivity of language (mind, brain, culture, sociality, evolution, >>>>> etc.). >>>>> In biology, another extremely complex science with multiple >>>>> connections >>>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>>> evolution, >>>>> etc.), ANY method is welcome so long as it does the job of furthering >>>>> our >>>>> understanding. And by understanding we mean ever-wider circles of >>>>> connectivity. >>>>> >>>>> Best, TG >>>>> ================ >>>>> >>>>> >>>>> A. Katz wrote: >>>>>> Chris, >>>>>> >>>>>> Thanks for addressing this question. I understand that many, many >>>>>> linguists >>>>>> are quite properly and approriately interested in this ultimate >>>>>> question: >>>>>> "How does the language user work?" (I am also intetested in this >>>>>> question >>>>>> some of the time.) >>>>>> >>>>>> My point is that "how does language work?" is also a valid question, >>>>>> and a >>>>>> central one to the field of linguistics. These two questions are not >>>>>> at >>>>>> all >>>>>> the same. >>>>>> >>>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>>> the >>>>>> "language". There are many good reasons to do so. For anyone working >>>>>> in >>>>>> computerized language processing or in non-human language studies, >>>>>> this >>>>>> is >>>>>> a significant point. >>>>>> >>>>>> It does not matter if a computerized language processing system even >>>>>> remotely simulates what humans do with language in their brains. It >>>>>> does >>>>>> matter whether it comes up with comparable or indistinguishable >>>>>> results. >>>>>> >>>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>>> doing >>>>>> the >>>>>> same things inside the same modules in his brain as a human does. It >>>>>> does >>>>>> matter if the results are functionally equivalent. >>>>>> >>>>>> We need to make that distinction, between humans and their language, >>>>>> or >>>>>> we >>>>>> will be caught inside a circular definition with no way to break out >>>>>> or >>>>>> to >>>>>> prove anything, not about others and not about ourselves! >>>>>> >>>>>> --Aya >>>>>> >>>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>>> >>>>>> >>>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>>> >>>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>>> >>>>>>> I don't for a moment dispute that expressions are often susceptible >>>>>>> to >>>>>>> multiple interpretations, that these interpretations are guided by >>>>>>> all >>>>>>> kinds of contextual information, or that different people, or even >>>>>>> the >>>>>>> same person at different times, may end up selecting differently >>>>>>> from >>>>>>> the >>>>>>> various interpretations. Your example 'Open happiness' in another >>>>>>> contribution to this thread illustrates these points very well. >>>>>>> >>>>>>> My point, though, is that each of these different interpretations, >>>>>>> as >>>>>>> well >>>>>>> as the selection of one (or more) as more likely in a particular >>>>>>> context, >>>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>>> evolved in >>>>>>> the course of the phylogenetic development of language in the human >>>>>>> species, and developed ontogenetically in that particular >>>>>>> interpreter's >>>>>>> brain. It is surely likely that those mechanisms are highly similar >>>>>>> in >>>>>>> different human beings, even though there may be differences in the >>>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>>> order >>>>>>> to >>>>>>> answer the question 'How do we communicate using language?' or, if >>>>>>> you >>>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>>> what >>>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>>> with. >>>>>>> >>>>>>> In particular, as linguists, we are interested (well, some of us >>>>>>> are, >>>>>>> though clearly not all) in whether the constructs we posit in our >>>>>>> theories >>>>>>> of language have psychological validity in the sense that they >>>>>>> correspond >>>>>>> to ways in which aspects of language are represented in the brain. >>>>>>> [As >>>>>>> an >>>>>>> aside, I do realise that there are linguists who strenuously resist >>>>>>> what >>>>>>> they see as a misguided emphasis on mental representation in the >>>>>>> work >>>>>>> of >>>>>>> cognitive scientists, but it seems clear that language must be >>>>>>> represented >>>>>>> in some way in the brain in order that we can engage in the >>>>>>> sociosemiotic >>>>>>> acts of meaning making which are the primary focus for many of these >>>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>>> need >>>>>>> answers to questions such as: What kind of representation does the >>>>>>> human >>>>>>> language processing system have for 'open' and for 'happiness'? Are >>>>>>> the >>>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>>> distinguish >>>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>>> might >>>>>>> want to add 'for this particular interpreter', though there must be >>>>>>> similarities across interpreters for communication to be possible) >>>>>>> represented in the same or different ways, in the same or different >>>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>>> in >>>>>>> thinking that these familiar levels of linguistic description must >>>>>>> be >>>>>>> differentiated, as such, in the human language processing system? >>>>>>> Does >>>>>>> the >>>>>>> representation for 'open' distinguish between what we call verbal >>>>>>> and >>>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>>> analyses >>>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>>> than >>>>>>> the >>>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>>> adjectives? >>>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, or >>>>>>> both? >>>>>>> All these questions, and many many more, are relevant to the >>>>>>> construction >>>>>>> of a model of language which reflects how language users communicate >>>>>>> (as, >>>>>>> of course, are a whole set of other questions about the >>>>>>> sociocultural >>>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>>> too >>>>>>> superficial reading in the area suggests that psycholinguists have >>>>>>> gone >>>>>>> some of the way towards answering some of the questions we might >>>>>>> want >>>>>>> to >>>>>>> ask, but that there is still a long way to go. >>>>>>> >>>>>>> Chris >>>>>>> -------------------------------------------------- >>>>>>> From: "A. Katz" >>>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>>> To: "Chris Butler" >>>>>>> Cc: "FUNKNET" >>>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>>> >>>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>>> see >>>>>>>> what >>>>>>>> is happening in each person's head, and we see that conflicting >>>>>>>> analyses >>>>>>>> are the norm, rather than the exception, among normal language >>>>>>>> speakers, >>>>>>>> then we will realize that the way language works to transmit >>>>>>>> information >>>>>>>> is despite individual differences and not because of uniform >>>>>>>> processing >>>>>>>> strategies. >>>>>>>> >>>>>>>> Even when all are agreed as to the meaning of an utterance, they do >>>>>>>> not >>>>>>>> process it the same way. Which means that processing is seocndary >>>>>>>> to >>>>>>>> information transmission. >>>>>>>> >>>>>>>> --Aya >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>>> >>>>>>>>> Dick's comment that "this discussion raises the really fundamental >>>>>>>>> question of what kind of thing we think language is: social or >>>>>>>>> individual" is, it seems to me, an important one, particularly for >>>>>>>>> those >>>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>>> truly >>>>>>>>> functional model of language would be one which aims to account >>>>>>>>> for >>>>>>>>> how >>>>>>>>> human beings communicate using language, or in other words tries >>>>>>>>> to >>>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>>> now, >>>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>>> natural >>>>>>>>> language user work?' In trying to answer this question we need to >>>>>>>>> accept >>>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>>> explore >>>>>>>>> both >>>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>>> communicate >>>>>>>>> using language. We need to know BOTH how people create and respond >>>>>>>>> to >>>>>>>>> meanings and express those meanings in forms during social >>>>>>>>> interaction >>>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>>> order >>>>>>>>> to make such interaction possible. Both are important parts of the >>>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>>> though >>>>>>>>> this particular thread of the Funknet discussion has concentrated >>>>>>>>> more >>>>>>>>> on the second aspect, and so will I. >>>>>>>>> >>>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>>> "exploring >>>>>>>>> the structure of a language so that I can understand how all the >>>>>>>>> bits >>>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>>> Dick >>>>>>>>> puts it, is useless - far from it. After all, the hypotheses that >>>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>>> like. >>>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>>> evidence >>>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>>> least >>>>>>>>> consistent with what we know of the processes which go on when we >>>>>>>>> use >>>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>>> only >>>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>>> terms >>>>>>>>> of >>>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>>> that >>>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>>> know >>>>>>>>> everything there is to know about language processing. But it does >>>>>>>>> mean >>>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>>> and >>>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>>> Ray >>>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also need >>>>>>>>> to >>>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>>> that >>>>>>>>> we >>>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>>> part, >>>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>>> experiments >>>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>>> criticism >>>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>>> natural >>>>>>>>> communicative conditions. >>>>>>>>> >>>>>>>>> I also agree with Dick when he says that "the differences between >>>>>>>>> individuals really matter", and with Lise when she points out that >>>>>>>>> "we >>>>>>>>> must also be careful not to idealize "what's in people's heads" as >>>>>>>>> if it >>>>>>>>> were a single coherent construct that we are trying to discover". >>>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>>> to >>>>>>>>> all >>>>>>>>> language users by virtue of the evolution of the language faculty >>>>>>>>> and >>>>>>>>> which constitute the "general processes" which Dick says >>>>>>>>> psycholinguists >>>>>>>>> are interested in. >>>>>>>>> >>>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>>> general >>>>>>>>> to >>>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>>> cases >>>>>>>>> in >>>>>>>>> terms of a hierarchy such as Brian suggests. One thing this means >>>>>>>>> is >>>>>>>>> that we should be giving our university students of linguistics >>>>>>>>> (and >>>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>>> aspects >>>>>>>>> of >>>>>>>>> linguistics that introduce them to the use of at least some of the >>>>>>>>> basic >>>>>>>>> statistical methods in language study, and I'm sure this is indeed >>>>>>>>> going >>>>>>>>> on in some enlightened places. To those who suspect this can't be >>>>>>>>> done >>>>>>>>> with maths-shy students who don't initially see the need for it, I >>>>>>>>> offer >>>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>>> courses to >>>>>>>>> people with little or no prior experience in quantitative >>>>>>>>> techniques. >>>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>>> linguistics >>>>>>>>> students in an institution where we had many mature students who >>>>>>>>> had >>>>>>>>> come into university level studies with non-standard >>>>>>>>> qualifications, >>>>>>>>> and >>>>>>>>> were not well equipped for courses of this kind by their previous >>>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>>> own >>>>>>>>> perspective as language students rather than that of the >>>>>>>>> statistician, >>>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>>> rather >>>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>>> and >>>>>>>>> to >>>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>>> research >>>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>>> techniques. And I still find that bright graduate students respond >>>>>>>>> well >>>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>>> advanced >>>>>>>>> techniques needed for many real research projects in various areas >>>>>>>>> of >>>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>>> already >>>>>>>>> happening in our fine institutions of higher education! >>>>>>>>> >>>>>>>>> Chris Butler >>>>>>>>> >>>>>>>>> >>>>>>> >>>>>>> >>>>> >>>>> >>>> >>>> >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> Bartlesville, OK >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From amnfn at well.com Mon Sep 13 13:19:36 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 06:19:36 -0700 Subject: Language as a Tool In-Reply-To: Message-ID: I also very much look forward to reading Daniel Everett's new book! --Aya On Mon, 13 Sep 2010, Mark P. Line wrote: > Thanks, Dan. I first got into linguistics through anthropology and never > really lost that frame of reference, so I'm very much looking forward to > your new book. > > Please rattle Pantheon's cage to get them to offer a Kindle edition of > your book. :) > > -- Mark > > Mark P. Line > > > > Daniel Everett wrote: >> >> Dear Mark, >> >> These are all excellent points. Clearly there is some biology that must >> underwrite language, or plants could speak. The question is, how much. >> Less than eating. More than wearing a tie, perhaps. >> >> I think that all of your points, however, are compatible with the idea >> that language is a tool, so long as we don't think that, as you say, this >> explains everything. It does, however, explain more than many have >> recognized. >> >> My new book on this, Cognitive Fire: Language as a Cultural Tool, is due >> out from Pantheon (US) and Profile (UK) sometime in 2011. >> >> Hopefully, I will have answered your questions. >> >> -- Dan >> >> >> On Sep 12, 2010, at 8:16 PM, Mark P. Line wrote: >> >>> Aya -- >>> >>> You said: "Other tools can be studied separately from the people who use >>> them or created them. Why not language?" >>> >>> Although language can certainly be considered a tool, I think it's >>> unlike >>> other tools in several very significant respects. >>> >>> 1. Although we're not born with language, we can't avoid (pathologies >>> excluded) acquiring it unless we're not exposed to it. To that extent, >>> language is a biological phenomenon. A prototypical tool is not a >>> biological phenomenon, so I'm not sure how valid any conclusions might >>> be >>> that are drawn from a premise of language-as-tool. >>> >>> 2. A tool is as a tool does: Anything is a tool that is being used as a >>> tool, including dead wombats, broken screwdrivers or decks of playing >>> cards. (Completing the imagined scenarios is left as an exercise for the >>> reader...) So saying that language is a tool is only saying that >>> language >>> is used as a tool. Quite a few conclusions can be and have been drawn >>> from >>> the fact that language is used as a tool, but I would have to be >>> convinced >>> in detail that almost everything worth knowing about language is >>> dependent >>> on the premise of language-as-tool. >>> >>> 3. If language is a "tool" for (say) communicating ideas, then eating is >>> a >>> "tool" for reducing hunger. In both cases, I worry about the tool >>> metaphor >>> being stretched so far from the prototype that we're left with an >>> out-and-out category fallacy. >>> >>> 4. More prototypical tools can be studied separately from those who use >>> or >>> create them because those tools are easily observed separately from >>> those >>> who use or create them. I don't think the same thing can be said of >>> language -- very little about language can be observed apart from its >>> use, >>> so very little about language can be observed apart from its user(s). >>> >>> 5. Any proposal to study something as complex as language separately >>> from >>> its embodiment is suspicious to me, smacking of reductionism -- >>> something >>> up with which linguistics has had to put a tad much. Anything that puts >>> language back into its human context would be a step forward. >>> >>> >>> -- Mark >>> >>> Mark P. Line >>> >>> >> >> > > > -- Mark > > Mark P. Line > Bartlesville, OK > > From geoffnathan at wayne.edu Mon Sep 13 13:36:40 2010 From: geoffnathan at wayne.edu (Geoff Nathan) Date: Mon, 13 Sep 2010 09:36:40 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: ----- Original Message ----- > From: "A. Katz" > You might as well say that a person cannot possibly avoid watching TV > once > he's exposed to it, as say the same about language. But people can > survive > just fine without television, and unless someone shows them how to > make a > TV set, most people will never figure out how to build one. The same > goes > for language. We're great at using it, not so great at generating it > out > of thin air with no ambient culture. > > --Aya Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage T?bingen 1994 or, for those, who, like me are Germanically-challenged, On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. Geoff Geoffrey S. Nathan Faculty Liaison, C&IT and Professor, Linguistics Program +1 (313) 577-1259 (C&IT) +1 (313) 577-8621 (English/Linguistics) From phonosemantics at earthlink.net Mon Sep 13 13:39:59 2010 From: phonosemantics at earthlink.net (jess tauber) Date: Mon, 13 Sep 2010 09:39:59 -0400 Subject: Language as a Tool Message-ID: Actually I'd say the true prototypical tools are body parts doing particular jobs- nails, teeth, hairs, wings, fins, legs etc. I doubt it escaped our early ancestors that their created or found tools had functions similar or identical to parts they had themselves or saw in other creatures. Even other lineages must have some inkling (as when a jay uses a thorn to prize a grub out of rotten wood, or a sea otter bashes shellfish with stones fished up from the sea floor). Not everything is pure instinct. Sometimes these parts come out- we lose teeth, nails, hair, birds feathers, and so on, and you can also yank them out of corpses, skeletons, etc., and the occasional unwilling live victim. We may find them loose. This sets the stage for alienability, and generalization perhaps? I can get a sharp tool from a sabertooth, or from the living rock if I knock politely. In languages with 'bipartite' structure, effector bodypart and instrument terms are often dealt with identically, and stand in contrast with pathway/position terms, which may have a mirror in the way the brain deals with tools and gait/posture related motion. Jess Tauber phonosemantics at earthlink.net From dan at daneverett.org Mon Sep 13 14:37:11 2010 From: dan at daneverett.org (Daniel Everett) Date: Mon, 13 Sep 2010 10:37:11 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <1143045877.645687.1284385000582.JavaMail.root@starship.merit.edu> Message-ID: Dear Geoff, These are all valid points, but none of them support nativism. One place to start is with Fiona Cowie's relatively recent book, What's Within (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) There are many cultural artifacts and values and types of knowledge that are never taught, but are all acquired just fine, from knowing to sit still in a canoe to how to use a bow and arrow. They are learned out of necessity and by observation, with no explicit instruction. In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev Fedorenko, in Cognition, support was offered for the idea that there are cognitive tools, number and counting being the examples we discussed. The reasoning chain for innatism often goes like this: 1. There is evidence that someone knows something that they were not taught. 2. I cannot think of how they learned it. 3. Therefore, they didn't learn it. 4. Therefore, it is innate. Both 3 and 4 are non-sequiturs. I don't want to get into a big discussion of this here. And I talk about this quite a bit in the book. So I will try to resist the temptation to respond to further postings. So I find your concerns both quite understandable, but not insurmountable or even particularly difficult problems for understanding language as a cognitive tool. -- Dan On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > ----- Original Message ----- >> From: "A. Katz" > >> You might as well say that a person cannot possibly avoid watching TV >> once >> he's exposed to it, as say the same about language. But people can >> survive >> just fine without television, and unless someone shows them how to >> make a >> TV set, most people will never figure out how to build one. The same >> goes >> for language. We're great at using it, not so great at generating it >> out >> of thin air with no ambient culture. >> >> --Aya > > Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. > This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. > All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. > I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. > I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book > > Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage T?bingen 1994 > > or, for those, who, like me are Germanically-challenged, > > On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. > > Geoff > > > Geoffrey S. Nathan > Faculty Liaison, C&IT > and Professor, Linguistics Program > +1 (313) 577-1259 (C&IT) > +1 (313) 577-8621 (English/Linguistics) > From vanvalin at buffalo.edu Mon Sep 13 14:46:06 2010 From: vanvalin at buffalo.edu (Robert Van Valin) Date: Mon, 13 Sep 2010 08:46:06 -0600 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <911A69BB-4302-4257-8452-B91B7CF61EB6@daneverett.org> Message-ID: Dan Slobin refers to the argument that Dan outlined as 'the argument from the poverty of the imagination'. Robert Van Valin On Sep 13, 2010, at 8:37 AM, Daniel Everett wrote: > Dear Geoff, > > These are all valid points, but none of them support nativism. > > One place to start is with Fiona Cowie's relatively recent book, What's Within (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) > > There are many cultural artifacts and values and types of knowledge that are never taught, but are all acquired just fine, from knowing to sit still in a canoe to how to use a bow and arrow. They are learned out of necessity and by observation, with no explicit instruction. > > In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev Fedorenko, in Cognition, support was offered for the idea that there are cognitive tools, number and counting being the examples we discussed. > > The reasoning chain for innatism often goes like this: > > 1. There is evidence that someone knows something that they were not taught. > > 2. I cannot think of how they learned it. > > 3. Therefore, they didn't learn it. > > 4. Therefore, it is innate. > > Both 3 and 4 are non-sequiturs. I don't want to get into a big discussion of this here. And I talk about this quite a bit in the book. So I will try to resist the temptation to respond to further postings. > > So I find your concerns both quite understandable, but not insurmountable or even particularly difficult problems for understanding language as a cognitive tool. > > -- Dan > > > > > > > On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > >> ----- Original Message ----- >>> From: "A. Katz" >> >>> You might as well say that a person cannot possibly avoid watching TV >>> once >>> he's exposed to it, as say the same about language. But people can >>> survive >>> just fine without television, and unless someone shows them how to >>> make a >>> TV set, most people will never figure out how to build one. The same >>> goes >>> for language. We're great at using it, not so great at generating it >>> out >>> of thin air with no ambient culture. >>> >>> --Aya >> >> Those who are familiar with my work know that I'm anything but a Chomskyan, but I'm sorry, there's an enormous difference between the acquisition of language and the acquisition of the knowledge necessary to build a television set (or the brodcasting and recording technology behind it). As generativists have pointed out since the early sixties, nobody is explicitly taught language, yet we all acquire it. Conversely many are intensively taught elementary physics, engineering etc. and DON'T aquire it. This is a difference in kind, not in degree. >> This is not to say that culture is taught either (of course nobody learns in school the correct distance to stand apart from an interlocutor, or how many milliseconds of silence in a conversation constitutes a 'pregnant pause'), but these are different kinds of knowledge from academic knowledge explicitly taught in some cultures and not in others. >> All cultures have correct social distance rules, syntactic structures and other tacitly acquired knowledge, but not all cultures learn physics, or which mushrooms are edible and which fatal. >> I'm looking forward to reading Dan's book too, but I find 'tool' an inappropriate metaphor for a cultural artifact that is never explicitly 'taught', is learned without effort in all cultures regardless of level of technology and is never improved by explicit experimentation or accidental innovation--there will never be the linguistic equivalent of a 'better mousetrap'. >> I prefer Rudi Keller's view that language is an object of the 'third kind'--an artifact that is neither wholly natural nor man-made, but that develops as a spontaneous order, without being designed, and with 'improvements' developing in different directions from the intentions of the developers. See his book >> >> Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage T?bingen 1994 >> >> or, for those, who, like me are Germanically-challenged, >> >> On Language Change: The Invisible Hand in Language, Routledge 1995, translated by Brigitte Nerlich. >> >> Geoff >> >> >> Geoffrey S. Nathan >> Faculty Liaison, C&IT >> and Professor, Linguistics Program >> +1 (313) 577-1259 (C&IT) >> +1 (313) 577-8621 (English/Linguistics) >> > > From geoffnathan at wayne.edu Mon Sep 13 15:39:03 2010 From: geoffnathan at wayne.edu (Geoff Nathan) Date: Mon, 13 Sep 2010 11:39:03 -0400 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <911A69BB-4302-4257-8452-B91B7CF61EB6@daneverett.org> Message-ID: I did not consider my argument to be one supporting Chomsky's version of nativism, in which linguistic knowledge is independent of (and prior to) experience. Rather I was arguing that the acquisition of language was quite different from the acquisition of a carpenter's skills in that language is not a consciously designed object, unlike what I would think of as a prototypical tool, such as a hammer or a pair of glasses. The structure of language is constrained, not by our genes, but by the 'equipment' that is used to produce, perceive and understand it. I am arguing here for the 'natural' view in 'Natural' Phonology. Stampe and Donegan, in their seminal paper 'The Study of Natural Phonology' argued for a different kind of innateness, in which universals arise from the interaction of the nature of the materials used with the general cognitive processes of the user. This is most clear in phonology, of course, where phonological structures are (partly) determined by the anatomy and physiology of the vocal tract, the physics of sound production and the psychology of perception. We make the sounds that our mouths make it easy to produce clearly (fortitions), and in the way that it's easy to produce them (lenitions), and we hear what other people say by assuming they have the same vocal apparatus and we hear what we would have intended should we have said that. But folks like Karen van Hoek have shown that syntactic island constraints have a similar basis in processing limitations, and others have also found extrinsic causes for other universal constraints on morphology and syntax. So language is certainly learned, but not in the same way that arithmetic is learned--it's more like how riding a bicycle is learned, through interaction with the physical stuff, and additionally, guided by our categorization and perceptual systems. Donegan, Patricia J., and David Stampe, 1979. ?The Study of Natural Phonology,? in Current Approaches to Phonological Theory., ed. by Dan Dinnsen. Bloomington: Indiana University Press. Geoffrey S. Nathan Faculty Liaison, C&IT and Professor, Linguistics Program +1 (313) 577-1259 (C&IT) +1 (313) 577-8621 (English/Linguistics) ----- Original Message ----- > From: "Daniel Everett" > To: "Geoff Nathan" > Cc: "Funknet Funknet" > Sent: Monday, September 13, 2010 10:37:11 AM > Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness > Dear Geoff, > > These are all valid points, but none of them support nativism. > > One place to start is with Fiona Cowie's relatively recent book, > What's Within > (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) > > There are many cultural artifacts and values and types of knowledge > that are never taught, but are all acquired just fine, from knowing to > sit still in a canoe to how to use a bow and arrow. They are learned > out of necessity and by observation, with no explicit instruction. > > In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev > Fedorenko, in Cognition, support was offered for the idea that there > are cognitive tools, number and counting being the examples we > discussed. > > The reasoning chain for innatism often goes like this: > > 1. There is evidence that someone knows something that they were not > taught. > > 2. I cannot think of how they learned it. > > 3. Therefore, they didn't learn it. > > 4. Therefore, it is innate. > > Both 3 and 4 are non-sequiturs. I don't want to get into a big > discussion of this here. And I talk about this quite a bit in the > book. So I will try to resist the temptation to respond to further > postings. > > So I find your concerns both quite understandable, but not > insurmountable or even particularly difficult problems for > understanding language as a cognitive tool. > > -- Dan > > > > > > > On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: > > > ----- Original Message ----- > >> From: "A. Katz" > > > >> You might as well say that a person cannot possibly avoid watching > >> TV > >> once > >> he's exposed to it, as say the same about language. But people can > >> survive > >> just fine without television, and unless someone shows them how to > >> make a > >> TV set, most people will never figure out how to build one. The > >> same > >> goes > >> for language. We're great at using it, not so great at generating > >> it > >> out > >> of thin air with no ambient culture. > >> > >> --Aya > > > > Those who are familiar with my work know that I'm anything but a > > Chomskyan, but I'm sorry, there's an enormous difference between the > > acquisition of language and the acquisition of the knowledge > > necessary to build a television set (or the brodcasting and > > recording technology behind it). As generativists have pointed out > > since the early sixties, nobody is explicitly taught language, yet > > we all acquire it. Conversely many are intensively taught elementary > > physics, engineering etc. and DON'T aquire it. This is a difference > > in kind, not in degree. > > This is not to say that culture is taught either (of course nobody > > learns in school the correct distance to stand apart from an > > interlocutor, or how many milliseconds of silence in a conversation > > constitutes a 'pregnant pause'), but these are different kinds of > > knowledge from academic knowledge explicitly taught in some cultures > > and not in others. > > All cultures have correct social distance rules, syntactic > > structures and other tacitly acquired knowledge, but not all > > cultures learn physics, or which mushrooms are edible and which > > fatal. > > I'm looking forward to reading Dan's book too, but I find 'tool' an > > inappropriate metaphor for a cultural artifact that is never > > explicitly 'taught', is learned without effort in all cultures > > regardless of level of technology and is never improved by explicit > > experimentation or accidental innovation--there will never be the > > linguistic equivalent of a 'better mousetrap'. > > I prefer Rudi Keller's view that language is an object of the 'third > > kind'--an artifact that is neither wholly natural nor man-made, but > > that develops as a spontaneous order, without being designed, and > > with 'improvements' developing in different directions from the > > intentions of the developers. See his book > > > > Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage > > T?bingen 1994 > > > > or, for those, who, like me are Germanically-challenged, > > > > On Language Change: The Invisible Hand in Language, Routledge 1995, > > translated by Brigitte Nerlich. > > > > Geoff > > > > > > Geoffrey S. Nathan > > Faculty Liaison, C&IT > > and Professor, Linguistics Program > > +1 (313) 577-1259 (C&IT) > > +1 (313) 577-8621 (English/Linguistics) > > From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 17:49:39 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 19:49:39 +0200 Subject: unhappiness debate-psycholinguistic research methods Message-ID: It seems that we are getting caught in an escalation of intellectual firepower, when a look at the research literature could focus more specifically on the issue Shannon raised. First of all, in reference to Tom's suggestions, it is not merely semantic priming that would be utilized, but a variety of priming techniques. Most important would be that the priming be masked, so that the subject would be unaware of the prime or of the nature of the real experimental question so as not to bias the results. The use of psycholinguistics as a research method in this case is not only not incompatible with a review of the literature of theoretical proposals from English diachronic phonologists and morphologists, it actually could be a necessity. To adequately structure a psycholinguistic experiment, information regarding the history and the age of the affixes could be required as background information. For example, if the bracketing paradox has as its origin that -ness and un- are on the same morpho-phonological level at some diachronic stage (the English of older speakers-who would not use the un-prefix on nouns), but younger speakers having been exposed to modern advertising language that uses it (Remember the "un-Cola"?) tend to think "un + happiness" the psycholinguist should separate subjects by age and use it is a factor within the statistical analysis. Not doing so might give inconclusive results. The differences between older and younger speakers would demonstrate a mini-stage in the historical process of affix integration. How does psycholinguistic research with agglutinating languages better inform us? We need to look at experiments with Hungarian, Turkish, Hungarian and Finnish-a special issue of Language and Cognitive Processes at the link below is a good summary of the state of the art as of 2008, and an earlier book by Baayen & Schreuder (2003) Morphological Structure in Language Processing is also informative. http://www.cognitivepsychologyarena.com/books/Advances-in-Morphological- Processing-isbn9781841698670 Ken Forster advised me, when I was consulting him some years ago about possibly applying this line of experimentation to my African language research, to study the literature on Finnish. At the time it seemed somewhat inconclusive, but I intend to keep checking back. There are people working in Hungarian, another vowel harmony language, so it looks like there will eventually be some convergence of research methods in the future. In fact, Vannerst and Boland (1999) already successfully tested the presence of morpho-phonological levels in English, explicitly invoking the lexical phonology model of English. It can be accessed at this link: http://www.ling.ohio-state.edu/~skang/371/DownloadablePapers/VannestBola nd99.pdf Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From amnfn at well.com Mon Sep 13 17:50:08 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 10:50:08 -0700 Subject: FUNKNET] analysis: unhappiness In-Reply-To: <416964165.651968.1284392343336.JavaMail.root@starship.merit.edu> Message-ID: Geoff, Let's say we accept your bicycle analogy. Learning language is like learning to ride a bicycle. Children in bicycle riding cultures pick it up fairly quickly, barring any physical disability. Children in cultures where bicycles are not present do not pick it up at all. Very few people could design a bicycle, even in cultures where bicycles abound. They usually buy their bicycles ready made. If they make adgjustments to the frame, they are fairly minor. --Aya On Mon, 13 Sep 2010, Geoff Nathan wrote: > I did not consider my argument to be one supporting Chomsky's version of nativism, in which linguistic knowledge is independent of (and prior to) experience. Rather I was arguing that the acquisition of language was quite different from the acquisition of a carpenter's skills in that language is not a consciously designed object, unlike what I would think of as a prototypical tool, such as a hammer or a pair of glasses. The structure of language is constrained, not by our genes, but by the 'equipment' that is used to produce, perceive and understand it. I am arguing here for the 'natural' view in 'Natural' Phonology. > > Stampe and Donegan, in their seminal paper 'The Study of Natural Phonology' argued for a different kind of innateness, in which universals arise from the interaction of the nature of the materials used with the general cognitive processes of the user. This is most clear in phonology, of course, where phonological structures are (partly) determined by the anatomy and physiology of the vocal tract, the physics of sound production and the psychology of perception. We make the sounds that our mouths make it easy to produce clearly (fortitions), and in the way that it's easy to produce them (lenitions), and we hear what other people say by assuming they have the same vocal apparatus and we hear what we would have intended should we have said that. But folks like Karen van Hoek have shown that syntactic island constraints have a similar basis in processing limitations, and others have also found extrinsic causes for other universal constraints on morphology and syntax. > > So language is certainly learned, but not in the same way that arithmetic is learned--it's more like how riding a bicycle is learned, through interaction with the physical stuff, and additionally, guided by our categorization and perceptual systems. > > Donegan, Patricia J., and David Stampe, 1979. ?The Study of Natural Phonology,? in Current Approaches to Phonological Theory., ed. by Dan Dinnsen. Bloomington: Indiana University Press. > > > Geoffrey S. Nathan > Faculty Liaison, C&IT > and Professor, Linguistics Program > +1 (313) 577-1259 (C&IT) > +1 (313) 577-8621 (English/Linguistics) > > ----- Original Message ----- >> From: "Daniel Everett" >> To: "Geoff Nathan" >> Cc: "Funknet Funknet" >> Sent: Monday, September 13, 2010 10:37:11 AM >> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >> Dear Geoff, >> >> These are all valid points, but none of them support nativism. >> >> One place to start is with Fiona Cowie's relatively recent book, >> What's Within >> (http://www.oup.com/us/catalog/general/subject/Philosophy/Mind/~~/dmlldz11c2EmY2k9OTc4MDE5NTE1OTc4Mw==) >> >> There are many cultural artifacts and values and types of knowledge >> that are never taught, but are all acquired just fine, from knowing to >> sit still in a canoe to how to use a bow and arrow. They are learned >> out of necessity and by observation, with no explicit instruction. >> >> In a paper with Mike Frank (Stanford Pscyhology), Ted Gibson and Ev >> Fedorenko, in Cognition, support was offered for the idea that there >> are cognitive tools, number and counting being the examples we >> discussed. >> >> The reasoning chain for innatism often goes like this: >> >> 1. There is evidence that someone knows something that they were not >> taught. >> >> 2. I cannot think of how they learned it. >> >> 3. Therefore, they didn't learn it. >> >> 4. Therefore, it is innate. >> >> Both 3 and 4 are non-sequiturs. I don't want to get into a big >> discussion of this here. And I talk about this quite a bit in the >> book. So I will try to resist the temptation to respond to further >> postings. >> >> So I find your concerns both quite understandable, but not >> insurmountable or even particularly difficult problems for >> understanding language as a cognitive tool. >> >> -- Dan >> >> >> >> >> >> >> On Sep 13, 2010, at 9:36 AM, Geoff Nathan wrote: >> >>> ----- Original Message ----- >>>> From: "A. Katz" >>> >>>> You might as well say that a person cannot possibly avoid watching >>>> TV >>>> once >>>> he's exposed to it, as say the same about language. But people can >>>> survive >>>> just fine without television, and unless someone shows them how to >>>> make a >>>> TV set, most people will never figure out how to build one. The >>>> same >>>> goes >>>> for language. We're great at using it, not so great at generating >>>> it >>>> out >>>> of thin air with no ambient culture. >>>> >>>> --Aya >>> >>> Those who are familiar with my work know that I'm anything but a >>> Chomskyan, but I'm sorry, there's an enormous difference between the >>> acquisition of language and the acquisition of the knowledge >>> necessary to build a television set (or the brodcasting and >>> recording technology behind it). As generativists have pointed out >>> since the early sixties, nobody is explicitly taught language, yet >>> we all acquire it. Conversely many are intensively taught elementary >>> physics, engineering etc. and DON'T aquire it. This is a difference >>> in kind, not in degree. >>> This is not to say that culture is taught either (of course nobody >>> learns in school the correct distance to stand apart from an >>> interlocutor, or how many milliseconds of silence in a conversation >>> constitutes a 'pregnant pause'), but these are different kinds of >>> knowledge from academic knowledge explicitly taught in some cultures >>> and not in others. >>> All cultures have correct social distance rules, syntactic >>> structures and other tacitly acquired knowledge, but not all >>> cultures learn physics, or which mushrooms are edible and which >>> fatal. >>> I'm looking forward to reading Dan's book too, but I find 'tool' an >>> inappropriate metaphor for a cultural artifact that is never >>> explicitly 'taught', is learned without effort in all cultures >>> regardless of level of technology and is never improved by explicit >>> experimentation or accidental innovation--there will never be the >>> linguistic equivalent of a 'better mousetrap'. >>> I prefer Rudi Keller's view that language is an object of the 'third >>> kind'--an artifact that is neither wholly natural nor man-made, but >>> that develops as a spontaneous order, without being designed, and >>> with 'improvements' developing in different directions from the >>> intentions of the developers. See his book >>> >>> Sprachwandel. Von der unsichtbaren Hand in der Sprache. 2. Auflage >>> T?bingen 1994 >>> >>> or, for those, who, like me are Germanically-challenged, >>> >>> On Language Change: The Invisible Hand in Language, Routledge 1995, >>> translated by Brigitte Nerlich. >>> >>> Geoff >>> >>> >>> Geoffrey S. Nathan >>> Faculty Liaison, C&IT >>> and Professor, Linguistics Program >>> +1 (313) 577-1259 (C&IT) >>> +1 (313) 577-8621 (English/Linguistics) >>> > > From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 18:02:25 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 20:02:25 +0200 Subject: Query of interest to Funknetters Message-ID: Given the debate, I should think that this query on the Linguist List should be of interest. ===========================Directory============================== 1) Date: 09-Sep-2010 From: Edouard Machery < machery at pitt.edu > Subject: Request for Participation for Innateness Study -------------------------Message 1 ---------------------------------- Date: Mon, 13 Sep 2010 11:41:07 From: Edouard Machery [machery at pitt.edu] Subject: Request for Participation for Innateness Study E-mail this message to a friend: http://linguistlist.org/issues/emailmessage/verification.cfm?iss=21-3616 .html&submissionid=2647414&topicid=8&msgnumber=1 Dear colleagues, Paul Griffiths (Sydney), Joshua Knobe (Yale), Stefan Linquist (Guelph), Edouard Machery (Pittsburgh), Richard Samuels (OSU), and Karola Stotz (Sydney) are collecting data about scientists' use of the notion of innateness in several disciplines. We are particularly interested in how linguists use this notion. We would like to invite you to take part to a short web study (at most 10 minutes). You will be asked to read a few vignettes and to judge whether some traits are innate. You will also be asked to fill out a brief demographic questionnaire. This survey is entirely anonymous, and the study has be approved by Guelph IRB, OSU IRB, Pittsburgh IRB, Sydney IRB, and Yale IRB. To participate to this study, follow the following link: http://yale.qualtrics.com/SE/?SID=SV_5yTK6uiKUc9HPy4 We would also appreciate if you could forward this e-mail to your colleagues and graduate students! Many thanks for your help!! Edouard Machery Linguistic Field(s): Cognitive Science Pragmatics Semantics Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From erin.shay at colorado.edu Mon Sep 13 18:46:56 2010 From: erin.shay at colorado.edu (Erin Shay) Date: Mon, 13 Sep 2010 12:46:56 -0600 Subject: Unsubscribe Message-ID: Would you mind taking me off the subscription list, please? Thank you. Erin Erin Shay, Ph.D. Research Ass't Professor Dept. of Linguistics, Box 295 University of Colorado Boulder, CO 80309-0295 USA Phone: 303-882-3786 From d.f.lesley-neuman.2 at hum.leidenuniv.nl Mon Sep 13 20:14:13 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Mon, 13 Sep 2010 22:14:13 +0200 Subject: unhappiness Message-ID: In response to Craig's observation: This is important. Now we would need to look at the history of the formation of -ness nouns to provide some clues as to the interactions of the two processes for a proper design of the study. If the observations of our list participants is correct, a study controlling age and other social factors might provide some clues regarding some type of historical change in progress, or a recycling of older meanings, further underscoring the importance of cross-fertilization between the fields of historical linguistics and psychology. Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BV Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From mark at polymathix.com Mon Sep 13 22:04:31 2010 From: mark at polymathix.com (Mark P. Line) Date: Mon, 13 Sep 2010 17:04:31 -0500 Subject: FUNKNET] analysis: unhappiness In-Reply-To: Message-ID: A. Katz wrote: > Mark, > > Okay, I'm glad that you agree that a language user might be a machine or a > non-human. That brings us much closer together. Nevertheless, there is > also the fact that language can be preserved, but not used, for > thousands of years, and then people start using it again. Again, I would say that artifacts of language can be preserved separately from language users, but not language itself. To my way of thinking, language is the processes by which such artifacts are created and used (when artifacts are involved at all). Those processes are synonymous with language use and they cannot be divorced from their embodiment, the language user. > If it were an biological process, or even something like a live virus > passed down from parent to child, then reviving it once it was dead would > be > impossible. But it's not. It's a tool. A tool that a group of people can > stop using and making, but keep the blue prints for, and then make it > again when they choose to. > > I wonder how many linguists realize that Hebrew was revived from a dead > language. I would assume most of us do. But I would also assume that most of us realize that Modern Hebrew was engineered to capture the available data in and about the dead language, not the language itself. There's no way to revive a dead language because all of its embodiments are, like, dead. To my mind, Modern Hebrew was created by tracing over a palimpsest. > How many think it is just a slavic language with Hebrew lexemes? I can't imagine any linguist who would think that. I might be wrong, I guess. > How many realize that the grammar, the triliteral roots, the derivational > system, are all working, and that any changes in pronunciation are of > little importance when the basic contrasts are preserved? This is all Ling 101 stuff, where I'm from, since Hebrew is the most culturally salient Semitic language in these parts, and Semitic is an important part of typological space with its triliteral and quadriliteral roots. Some people's mileage might vary, I guess. > Mark, if you can see that a language can be used by a non-human, then why > not also acknowledge that it's a system for transferring information, and > that the code can be preserved while in disuse? First, I accept that language is a tool in a far-from-prototypical sense (it's not a wrench, but it's not eating, either). As I said, a tool is as a tool does, and since language can be used as a tool, it's a tool, trivially, when it's being used as a tool. My point has been that not everything worth knowing about language profits from the insight that language can be a tool, and that that insight is not enough to justify the study of language separately from any of its observed or potential embodiments (and that, in fact, I know of nothing that might justify a reductionist approach of that order). Second, though, and more importantly, I most emphatically do not believe that ***THE*** code of ***A*** language is preserved by any finite volume of preserved text, for at least two related reasons: 1. There's no such thing as THE code of A language, so the whole endeavor suffers from presupposition failure. Every embodiment (like you, and me) has its own code. Communication ensues more or less effectively when a speaker and hearer are using codes that overlap to a sufficient degree, all the way up and down the chain from phonology to ontological commitment. 2. Even if we stipulated some "common code" that was shared (perfectly, or within some variationist framework) by some community of speakers, no amount of preserved text and/or grammatical description would suffice to preserve that common code. I would also note that, the farther you go from the artifacts of language to the code we postulate as existing in the minds of speaker/hearers, the less likely you're going to understand anything about it without taking the exigencies of the embodiment into account. My favorite example of this is the ease with which a language can be constructed (constructed as in Esperanto and Klingon) in which center embedding is rampant, the ease with which software parsers and generators can be built for such languages, and the abject difficulties any human being will have speaking or understanding any such language. -- Mark > On Mon, 13 Sep 2010, Mark P. Line wrote: > >> Aya -- >> >> A. Katz wrote: >>> Mark, >>> >>> Sorry, but "eating" is not a tool. It's a biological process. >>> >>> Eating cannot occur outside the biological context. A human being can >>> avoid eating, but if so he starves. Feral children do not speak, but >>> they eat, like everyone else, or they die. Eating does not have to be >>> taught, there is no critical age of acquisition and it is not uniquely >>> an >>> artefact of human culture. If you are an animal, you eat. Eating cannot >>> survive the death of the eater. >>> >>> Language can. >> >> I would say that not language, but the artifacts of language (texts, >> audio >> recordings) can survive the people who created them, because I try very >> hard not to reify the artifacts of language as "language". >> >> The (usually communicative) process I refer to as language cannot exist >> independently of its embodiment. That said, I don't care if the >> embodiment >> is human, computer, cetacean or non-human primate. >> >> >> -- Mark >> >> Mark P. Line >> >> >>> On Sun, 12 Sep 2010, Mark P. Line wrote: >>> >>>> Aya -- >>>> >>>> You said: "Other tools can be studied separately from the people who >>>> use >>>> them or created them. Why not language?" >>>> >>>> Although language can certainly be considered a tool, I think it's >>>> unlike >>>> other tools in several very significant respects. >>>> >>>> 1. Although we're not born with language, we can't avoid (pathologies >>>> excluded) acquiring it unless we're not exposed to it. To that extent, >>>> language is a biological phenomenon. A prototypical tool is not a >>>> biological phenomenon, so I'm not sure how valid any conclusions might >>>> be >>>> that are drawn from a premise of language-as-tool. >>>> >>>> 2. A tool is as a tool does: Anything is a tool that is being used as >>>> a >>>> tool, including dead wombats, broken screwdrivers or decks of playing >>>> cards. (Completing the imagined scenarios is left as an exercise for >>>> the >>>> reader...) So saying that language is a tool is only saying that >>>> language >>>> is used as a tool. Quite a few conclusions can be and have been drawn >>>> from >>>> the fact that language is used as a tool, but I would have to be >>>> convinced >>>> in detail that almost everything worth knowing about language is >>>> dependent >>>> on the premise of language-as-tool. >>>> >>>> 3. If language is a "tool" for (say) communicating ideas, then eating >>>> is >>>> a >>>> "tool" for reducing hunger. In both cases, I worry about the tool >>>> metaphor >>>> being stretched so far from the prototype that we're left with an >>>> out-and-out category fallacy. >>>> >>>> 4. More prototypical tools can be studied separately from those who >>>> use >>>> or >>>> create them because those tools are easily observed separately from >>>> those >>>> who use or create them. I don't think the same thing can be said of >>>> language -- very little about language can be observed apart from its >>>> use, >>>> so very little about language can be observed apart from its user(s). >>>> >>>> 5. Any proposal to study something as complex as language separately >>>> from >>>> its embodiment is suspicious to me, smacking of reductionism -- >>>> something >>>> up with which linguistics has had to put a tad much. Anything that >>>> puts >>>> language back into its human context would be a step forward. >>>> >>>> >>>> -- Mark >>>> >>>> Mark P. Line >>>> >>>> >>>> >>>> A. Katz wrote: >>>>> Tom, >>>>> >>>>> I understand the uncomfortable association with Chomsky that speaking >>>>> of >>>>> language apart from people can have. Competence versus performance, >>>>> the >>>>> way Chomsky used those terms, never made sense. But that's precisely >>>>> because >>>>> he associated "competence" with the brain and suggested that it was >>>>> hard >>>>> wired there -- when there was never any evidence of that. >>>>> >>>>> However, if we don't distinguish language from humans, and language >>>>> processing from language data, then how are we going to judge >>>>> artificial >>>>> language-using devices as to their efficacy at producing and >>>>> interpreting >>>>> language? How are we going to determine whether and to what extent a >>>>> non-human has acquired language? >>>>> >>>>> We aren't born with it. We don't embody it. It's a tool that we use >>>>> to >>>>> communicate. Other tools can be studied separately from the people >>>>> who >>>>> use >>>>> them or created them. Why not language? >>>>> >>>>> --Aya >>>>> >>>>> >>>>> >>>>> >>>>> On Sun, 12 Sep 2010, Tom Givon wrote: >>>>> >>>>>> >>>>>> >>>>>> >>>>>> I wonder whether asking "how does language work?" is really a >>>>>> meaningful >>>>>> question without asking "how does the language user work?" One of >>>>>> the >>>>>> worst >>>>>> legacies good ol' Noam stuck us with is his (really, Saussure's) >>>>>> distinction >>>>>> of competence ("language", "knowledge") vs. performance ("language >>>>>> user", >>>>>> "processing"). It purported to limit linguists to the armchair >>>>>> methods >>>>>> that >>>>>> study competence, and relegated to psychology the quantified, >>>>>> distributional/variationist methods that study behavior, processing >>>>>> and >>>>>> on-line communication. The first breach in this artificial >>>>>> methodological >>>>>> wall occurred, leastwise for some of us, when we discovered the >>>>>> intermediate >>>>>> method of quantified studies of text, interaction, and conversation. >>>>>> As >>>>>> an >>>>>> ex-biologist, I am forever puzzled by the methodological purism we >>>>>> sill >>>>>> seem >>>>>> to embrace in linguistics, in the face of the manifest complexity >>>>>> and >>>>>> connectivity of language (mind, brain, culture, sociality, >>>>>> evolution, >>>>>> etc.). >>>>>> In biology, another extremely complex science with multiple >>>>>> connections >>>>>> (chemistry, geology, paleontology, behavior, sociality, economics, >>>>>> evolution, >>>>>> etc.), ANY method is welcome so long as it does the job of >>>>>> furthering >>>>>> our >>>>>> understanding. And by understanding we mean ever-wider circles of >>>>>> connectivity. >>>>>> >>>>>> Best, TG >>>>>> ================ >>>>>> >>>>>> >>>>>> A. Katz wrote: >>>>>>> Chris, >>>>>>> >>>>>>> Thanks for addressing this question. I understand that many, many >>>>>>> linguists >>>>>>> are quite properly and approriately interested in this ultimate >>>>>>> question: >>>>>>> "How does the language user work?" (I am also intetested in this >>>>>>> question >>>>>>> some of the time.) >>>>>>> >>>>>>> My point is that "how does language work?" is also a valid >>>>>>> question, >>>>>>> and a >>>>>>> central one to the field of linguistics. These two questions are >>>>>>> not >>>>>>> at >>>>>>> all >>>>>>> the same. >>>>>>> >>>>>>> Let me be very explicit: My aim is to separate out the "human" from >>>>>>> the >>>>>>> "language". There are many good reasons to do so. For anyone >>>>>>> working >>>>>>> in >>>>>>> computerized language processing or in non-human language studies, >>>>>>> this >>>>>>> is >>>>>>> a significant point. >>>>>>> >>>>>>> It does not matter if a computerized language processing system >>>>>>> even >>>>>>> remotely simulates what humans do with language in their brains. It >>>>>>> does >>>>>>> matter whether it comes up with comparable or indistinguishable >>>>>>> results. >>>>>>> >>>>>>> It does not matter whether a parrot, a dolphin or a chimpanzee is >>>>>>> doing >>>>>>> the >>>>>>> same things inside the same modules in his brain as a human does. >>>>>>> It >>>>>>> does >>>>>>> matter if the results are functionally equivalent. >>>>>>> >>>>>>> We need to make that distinction, between humans and their >>>>>>> language, >>>>>>> or >>>>>>> we >>>>>>> will be caught inside a circular definition with no way to break >>>>>>> out >>>>>>> or >>>>>>> to >>>>>>> prove anything, not about others and not about ourselves! >>>>>>> >>>>>>> --Aya >>>>>>> >>>>>>> http://hubpages.com/hub/What-Constitutes-Proof-in-Ape-Language-Studies >>>>>>> >>>>>>> >>>>>>> On Sun, 12 Sep 2010, Chris Butler wrote: >>>>>>> >>>>>>>> Aya, I think two different things are getting a bit mixed up here. >>>>>>>> >>>>>>>> I don't for a moment dispute that expressions are often >>>>>>>> susceptible >>>>>>>> to >>>>>>>> multiple interpretations, that these interpretations are guided by >>>>>>>> all >>>>>>>> kinds of contextual information, or that different people, or even >>>>>>>> the >>>>>>>> same person at different times, may end up selecting differently >>>>>>>> from >>>>>>>> the >>>>>>>> various interpretations. Your example 'Open happiness' in another >>>>>>>> contribution to this thread illustrates these points very well. >>>>>>>> >>>>>>>> My point, though, is that each of these different interpretations, >>>>>>>> as >>>>>>>> well >>>>>>>> as the selection of one (or more) as more likely in a particular >>>>>>>> context, >>>>>>>> is achieved through mechanisms in the interpreter's brain which >>>>>>>> evolved in >>>>>>>> the course of the phylogenetic development of language in the >>>>>>>> human >>>>>>>> species, and developed ontogenetically in that particular >>>>>>>> interpreter's >>>>>>>> brain. It is surely likely that those mechanisms are highly >>>>>>>> similar >>>>>>>> in >>>>>>>> different human beings, even though there may be differences in >>>>>>>> the >>>>>>>> detailed wiring in different brains. What I'm saying is that in >>>>>>>> order >>>>>>>> to >>>>>>>> answer the question 'How do we communicate using language?' or, if >>>>>>>> you >>>>>>>> prefer, 'How does the language user work?', we need to investigate >>>>>>>> what >>>>>>>> those mechanisms are, and this is what psycholinguists can help us >>>>>>>> with. >>>>>>>> >>>>>>>> In particular, as linguists, we are interested (well, some of us >>>>>>>> are, >>>>>>>> though clearly not all) in whether the constructs we posit in our >>>>>>>> theories >>>>>>>> of language have psychological validity in the sense that they >>>>>>>> correspond >>>>>>>> to ways in which aspects of language are represented in the brain. >>>>>>>> [As >>>>>>>> an >>>>>>>> aside, I do realise that there are linguists who strenuously >>>>>>>> resist >>>>>>>> what >>>>>>>> they see as a misguided emphasis on mental representation in the >>>>>>>> work >>>>>>>> of >>>>>>>> cognitive scientists, but it seems clear that language must be >>>>>>>> represented >>>>>>>> in some way in the brain in order that we can engage in the >>>>>>>> sociosemiotic >>>>>>>> acts of meaning making which are the primary focus for many of >>>>>>>> these >>>>>>>> critics.] Taking your 'Open happiness' examples again, I think we >>>>>>>> need >>>>>>>> answers to questions such as: What kind of representation does the >>>>>>>> human >>>>>>>> language processing system have for 'open' and for 'happiness'? >>>>>>>> Are >>>>>>>> the >>>>>>>> phonological, syntactic, semantic and pragmatic (for those who >>>>>>>> distinguish >>>>>>>> the semantic from the pragmatic) properties of these items (and we >>>>>>>> might >>>>>>>> want to add 'for this particular interpreter', though there must >>>>>>>> be >>>>>>>> similarities across interpreters for communication to be possible) >>>>>>>> represented in the same or different ways, in the same or >>>>>>>> different >>>>>>>> locations (or sets of distributed locations)? Indeed, are we right >>>>>>>> in >>>>>>>> thinking that these familiar levels of linguistic description must >>>>>>>> be >>>>>>>> differentiated, as such, in the human language processing system? >>>>>>>> Does >>>>>>>> the >>>>>>>> representation for 'open' distinguish between what we call verbal >>>>>>>> and >>>>>>>> adjectival uses of this item, and if so how? Or are syntactic >>>>>>>> analyses >>>>>>>> computed on the fly, using semantic and contextual clues, rather >>>>>>>> than >>>>>>>> the >>>>>>>> neural equivalent of pigeon holes corresponding to verbs and >>>>>>>> adjectives? >>>>>>>> Is 'happiness' represented as 'happy + ness', or in its entirety, >>>>>>>> or >>>>>>>> both? >>>>>>>> All these questions, and many many more, are relevant to the >>>>>>>> construction >>>>>>>> of a model of language which reflects how language users >>>>>>>> communicate >>>>>>>> (as, >>>>>>>> of course, are a whole set of other questions about the >>>>>>>> sociocultural >>>>>>>> aspects of communication). I am not a psycholinguist, but my all >>>>>>>> too >>>>>>>> superficial reading in the area suggests that psycholinguists have >>>>>>>> gone >>>>>>>> some of the way towards answering some of the questions we might >>>>>>>> want >>>>>>>> to >>>>>>>> ask, but that there is still a long way to go. >>>>>>>> >>>>>>>> Chris >>>>>>>> -------------------------------------------------- >>>>>>>> From: "A. Katz" >>>>>>>> Sent: Saturday, September 11, 2010 3:18 PM >>>>>>>> To: "Chris Butler" >>>>>>>> Cc: "FUNKNET" >>>>>>>> Subject: Re: [FUNKNET] FUNKNET] analysis: unhappiness >>>>>>>> >>>>>>>>> The problem is that once we achieve the psycholinguistic goal and >>>>>>>>> see >>>>>>>>> what >>>>>>>>> is happening in each person's head, and we see that conflicting >>>>>>>>> analyses >>>>>>>>> are the norm, rather than the exception, among normal language >>>>>>>>> speakers, >>>>>>>>> then we will realize that the way language works to transmit >>>>>>>>> information >>>>>>>>> is despite individual differences and not because of uniform >>>>>>>>> processing >>>>>>>>> strategies. >>>>>>>>> >>>>>>>>> Even when all are agreed as to the meaning of an utterance, they >>>>>>>>> do >>>>>>>>> not >>>>>>>>> process it the same way. Which means that processing is seocndary >>>>>>>>> to >>>>>>>>> information transmission. >>>>>>>>> >>>>>>>>> --Aya >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> On Sat, 11 Sep 2010, Chris Butler wrote: >>>>>>>>> >>>>>>>>>> Dick's comment that "this discussion raises the really >>>>>>>>>> fundamental >>>>>>>>>> question of what kind of thing we think language is: social or >>>>>>>>>> individual" is, it seems to me, an important one, particularly >>>>>>>>>> for >>>>>>>>>> those >>>>>>>>>> of us who are committed functionalists. My own view is that a >>>>>>>>>> truly >>>>>>>>>> functional model of language would be one which aims to account >>>>>>>>>> for >>>>>>>>>> how >>>>>>>>>> human beings communicate using language, or in other words tries >>>>>>>>>> to >>>>>>>>>> answer the question which was posed by Simon Dik a long time ago >>>>>>>>>> now, >>>>>>>>>> but which was not tackled head-on in his own work: "How does the >>>>>>>>>> natural >>>>>>>>>> language user work?' In trying to answer this question we need >>>>>>>>>> to >>>>>>>>>> accept >>>>>>>>>> that language is BOTH social AND individual, and we need to >>>>>>>>>> explore >>>>>>>>>> both >>>>>>>>>> aspects to get as complete a picture as possible of how we >>>>>>>>>> communicate >>>>>>>>>> using language. We need to know BOTH how people create and >>>>>>>>>> respond >>>>>>>>>> to >>>>>>>>>> meanings and express those meanings in forms during social >>>>>>>>>> interaction >>>>>>>>>> AND the mechanisms which operate in the brains of individuals in >>>>>>>>>> order >>>>>>>>>> to make such interaction possible. Both are important parts of >>>>>>>>>> the >>>>>>>>>> answer to the question 'How do we communicate using language?', >>>>>>>>>> though >>>>>>>>>> this particular thread of the Funknet discussion has >>>>>>>>>> concentrated >>>>>>>>>> more >>>>>>>>>> on the second aspect, and so will I. >>>>>>>>>> >>>>>>>>>> This doesn't mean that all the work linguists have done on >>>>>>>>>> "exploring >>>>>>>>>> the structure of a language so that I can understand how all the >>>>>>>>>> bits >>>>>>>>>> fit together" and "exploring the connections between items", as >>>>>>>>>> Dick >>>>>>>>>> puts it, is useless - far from it. After all, the hypotheses >>>>>>>>>> that >>>>>>>>>> psycholinguists test are based on ideas about what languages are >>>>>>>>>> like. >>>>>>>>>> But it does mean, in my view, that ultimately we need to get >>>>>>>>>> evidence >>>>>>>>>> that the constructs and analyses we propose are ones that are at >>>>>>>>>> least >>>>>>>>>> consistent with what we know of the processes which go on when >>>>>>>>>> we >>>>>>>>>> use >>>>>>>>>> language. So I am with Matthew when he says that for him, "the >>>>>>>>>> only >>>>>>>>>> sense in which an analysis can be "the correct analysis" is in >>>>>>>>>> terms >>>>>>>>>> of >>>>>>>>>> what is inside of people's heads". Of course, this doesn't imply >>>>>>>>>> that >>>>>>>>>> linguists should just give up their jobs until such time as we >>>>>>>>>> know >>>>>>>>>> everything there is to know about language processing. But it >>>>>>>>>> does >>>>>>>>>> mean >>>>>>>>>> that we need to collaborate with psycholinguists, psychologists >>>>>>>>>> and >>>>>>>>>> neurologists, as has also been pointed out by linguists such as >>>>>>>>>> Ray >>>>>>>>>> Jackendoff, Asif Agha, Ewa Dabrowska and Jan Nuyts. [We also >>>>>>>>>> need >>>>>>>>>> to >>>>>>>>>> collaborate much more with sociolinguists and sociologists, so >>>>>>>>>> that >>>>>>>>>> we >>>>>>>>>> can get a better handle on the sociocultural aspects of how we >>>>>>>>>> communicate.] And it also means that psycholinguists, for their >>>>>>>>>> part, >>>>>>>>>> need whenever possible to follow up tightly controlled lab >>>>>>>>>> experiments >>>>>>>>>> with studies under more naturalistic conditions, to avoid the >>>>>>>>>> criticism >>>>>>>>>> that what happens in artifical lab situations may not happen in >>>>>>>>>> natural >>>>>>>>>> communicative conditions. >>>>>>>>>> >>>>>>>>>> I also agree with Dick when he says that "the differences >>>>>>>>>> between >>>>>>>>>> individuals really matter", and with Lise when she points out >>>>>>>>>> that >>>>>>>>>> "we >>>>>>>>>> must also be careful not to idealize "what's in people's heads" >>>>>>>>>> as >>>>>>>>>> if it >>>>>>>>>> were a single coherent construct that we are trying to >>>>>>>>>> discover". >>>>>>>>>> However, there are surely processing mechanisms which are common >>>>>>>>>> to >>>>>>>>>> all >>>>>>>>>> language users by virtue of the evolution of the language >>>>>>>>>> faculty >>>>>>>>>> and >>>>>>>>>> which constitute the "general processes" which Dick says >>>>>>>>>> psycholinguists >>>>>>>>>> are interested in. >>>>>>>>>> >>>>>>>>>> On the issue of quantitative methodology, I'm sympathetic in >>>>>>>>>> general >>>>>>>>>> to >>>>>>>>>> Ted and Ev's views, though it does seem sensible to prioritise >>>>>>>>>> cases >>>>>>>>>> in >>>>>>>>>> terms of a hierarchy such as Brian suggests. One thing this >>>>>>>>>> means >>>>>>>>>> is >>>>>>>>>> that we should be giving our university students of linguistics >>>>>>>>>> (and >>>>>>>>>> some of our linguistics lecturers!) courses in quantitative >>>>>>>>>> aspects >>>>>>>>>> of >>>>>>>>>> linguistics that introduce them to the use of at least some of >>>>>>>>>> the >>>>>>>>>> basic >>>>>>>>>> statistical methods in language study, and I'm sure this is >>>>>>>>>> indeed >>>>>>>>>> going >>>>>>>>>> on in some enlightened places. To those who suspect this can't >>>>>>>>>> be >>>>>>>>>> done >>>>>>>>>> with maths-shy students who don't initially see the need for it, >>>>>>>>>> I >>>>>>>>>> offer >>>>>>>>>> my own experience, over quite a long period, of teaching such >>>>>>>>>> courses to >>>>>>>>>> people with little or no prior experience in quantitative >>>>>>>>>> techniques. >>>>>>>>>> For some years in the 1990s, I taught such courses to all >>>>>>>>>> linguistics >>>>>>>>>> students in an institution where we had many mature students who >>>>>>>>>> had >>>>>>>>>> come into university level studies with non-standard >>>>>>>>>> qualifications, >>>>>>>>>> and >>>>>>>>>> were not well equipped for courses of this kind by their >>>>>>>>>> previous >>>>>>>>>> experience. I'm glad to say that teaching the subject from their >>>>>>>>>> own >>>>>>>>>> perspective as language students rather than that of the >>>>>>>>>> statistician, >>>>>>>>>> and explaining the reasons for doing things in particular ways >>>>>>>>>> rather >>>>>>>>>> than just presenting formulae, paid off in the end, so that most >>>>>>>>>> students were able to appreciate the relevance of these courses >>>>>>>>>> and >>>>>>>>>> to >>>>>>>>>> turn in very creditable projects showing an understanding of >>>>>>>>>> research >>>>>>>>>> design and competence in the use of a range of basic statistical >>>>>>>>>> techniques. And I still find that bright graduate students >>>>>>>>>> respond >>>>>>>>>> well >>>>>>>>>> to similar courses which incorporate some of the rather more >>>>>>>>>> advanced >>>>>>>>>> techniques needed for many real research projects in various >>>>>>>>>> areas >>>>>>>>>> of >>>>>>>>>> linguistics. But I may well be out of date with what is now >>>>>>>>>> already >>>>>>>>>> happening in our fine institutions of higher education! >>>>>>>>>> >>>>>>>>>> Chris Butler >>>>>>>>>> >>>>>>>>>> >>>>>>>> >>>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>> >>>> >>>> -- Mark >>>> >>>> Mark P. Line >>>> Bartlesville, OK >>>> >>>> >>> >>> >> >> >> -- Mark >> >> Mark P. Line >> Bartlesville, OK >> >> > > -- Mark Mark P. Line Bartlesville, OK From amnfn at well.com Mon Sep 13 23:33:06 2010 From: amnfn at well.com (A. Katz) Date: Mon, 13 Sep 2010 16:33:06 -0700 Subject: Language as a Tool In-Reply-To: <24374061.1284385200100.JavaMail.root@wamui-june.atl.sa.earthlink.net> Message-ID: Jess, Your observation about body parts serving as prototypical tools is apt. I would add that teeth serving as cutting tools came first, and that their use in making dental consonants was discovered later. But we have other tools now to help us represent dental consonants, including but not limited to the writing system and speech synthesizers. Parrots can make sounds that pass for dental consonants, though they have no teeth. Chimpanzees have teeth, but cannot make those sounds. The point? There is more than one object that can serve as a tool to produce language contrasts. Some are embodied, and some are not. Anatomical similarity isn't everything. --Aya On Mon, 13 Sep 2010, jess tauber wrote: > Actually I'd say the true prototypical tools are body parts doing particular jobs- nails, teeth, hairs, wings, fins, legs etc. I doubt it escaped our early ancestors that their created or found tools had functions similar or identical to parts they had themselves or saw in other creatures. Even other lineages must have some inkling (as when a jay uses a thorn to prize a grub out of rotten wood, or a sea otter bashes shellfish with stones fished up from the sea floor). Not everything is pure instinct. > > Sometimes these parts come out- we lose teeth, nails, hair, birds feathers, and so on, and you can also yank them out of corpses, skeletons, etc., and the occasional unwilling live victim. We may find them loose. This sets the stage for alienability, and generalization perhaps? I can get a sharp tool from a sabertooth, or from the living rock if I knock politely. > > In languages with 'bipartite' structure, effector bodypart and instrument terms are often dealt with identically, and stand in contrast with pathway/position terms, which may have a mirror in the way the brain deals with tools and gait/posture related motion. > > Jess Tauber > phonosemantics at earthlink.net > > From LGarrison at gc.cuny.edu Mon Sep 20 15:11:21 2010 From: LGarrison at gc.cuny.edu (Garrison, Leigh) Date: Mon, 20 Sep 2010 11:11:21 -0400 Subject: Columbia School Linguistics Conference Announcement Message-ID: 10th International Columbia School Conference on the Interaction of Linguistic Form and Meaning with Human Behavior Conference theme: Grammatical analysis and the discovery of meaning October 9-11, 2010 Rutgers University New Brunswick, New Jersey Invited speakers: Flora Klein-Andreu (Stony Brook University) Linguistics for non-linguists Andrea Tyler (Georgetown University) Connecting Spatial Particles and Aspect Markers: Applying the Principled Polysemy Model to Russian za List of presenters: Tanya Karoli Christensen (Copenhagen University) Sign combinations in context: Imperatives and modal particles in Danish Ellen Contini-Morava (University of Virginia) The meaning(s?) of non-animate deictic markers in Swahili Joseph Davis (The City College ? CUNY) Diver?s Latin voice and case Bob de Jonge (University of Groningen) Phonology as Human Behaviour revisited: The case of Romance languages Thomas Eccardt (Independent scholar) Pitch and aperture: Two articulatory scalars in comparison Richard Epstein (Rutgers University) Some discourse uses of the distal demonstrative determiner in Beowulf Elena Even-Simkin, Yishai Tobin (Ben-Gurion University of the Negev) Internal vowel alternation as a phonological-semantic sign system in English according to the sign-oriented theory of the Columbia School Alan Huffman (The Graduate Center ? CUNY) The phonological motivation for Verner's Law and Grimm's Law Robert Kirsner (University of California, Los Angeles) Minimal units, their context, and the insufficiency of conceptual metaphor: Revisiting the Dutch dismissive idiom ho maar ?fuhgeddaboudit, of course not!? Robert Leonard (Hofstra University) Linguistic meaning, pragmatics and context: Semantic analysis of evidence in a double homicide trial seeking to weigh intent Lin Lin (University of California, Los Angeles) Rethinking of the Chinese demonstratives in the Columbia School framework Carol Moder (Oklahoma State University) Dirty hands, dirty work: Usage-based noun modification Ricardo Otheguy (The Graduate Center ? CUNY) A report on current research on Spanish in New York Wallis Reid (Rutgers University) English verb number: Syntactic or semantic? Hidemi Riggs (Soka University of America) The structure of Japanese conditionals in Modern Japanese: A grammatical account from a functional linguistics perspective Nancy Stern (The City College ? CUNY) Ourself, themself, and more: The communicative function of Number in -self pronouns Lavi Wolf, Yishai Tobin (Ben-Gurion University of the Negev) Phonological proclivities across languages according to the theory of Phonology as Human Behavior The Columbia School is a group of linguists developing the theoretical framework origi?nally established by the late William Diver. Language is seen as a symbolic tool whose structure is shaped both by its communicative function and by the characteristics of its human users. Grammatical analyses account for the distribution of linguistic forms as an interaction between linguistic meaning and pragmatic and functional factors such as inference, ease of processing, and iconicity. Phonological analyses explain the syntag?matic and paradigmatic distribution of phonological units within signals, also drawing on both communicative function and human physiological and psychological characteristics. * * * * * * * * The support of The Columbia School Linguistic Society is gratefully acknowledged www.csling.org * * * * * * * * Selected Columbia School bibliography: Contini-Morava, Ellen, Robert S. Kirsner, and Betsy Rodriguez-Bachiller (eds.). 2005. Cognitive and Communicative Approaches to Linguistic Analysis. Amsterdam/Philadelphia: John Benjamins. Contini-Morava, Ellen, and Barbara Sussman Goldberg (eds.). 1995. Meaning as Explanation: Advances in Linguistic Sign Theory. Berlin: Mouton de Gruyter. Davis, Joseph, Radmila Gorup, and Nancy Stern (eds.). 2006. Advances in Functional Linguistics: Columbia School beyond its origins. Amsterdam/Philadelphia: John Benjamins. Huffman, Alan. 1997. The Categories of Grammar: French lui and le. Amsterdam/Philadelphia: John Benjamins. Huffman, Alan. 2001. ?The Linguistics of William Diver and the Columbia School.? WORD 52:1, 29-68. Reid, Wallis. 1991. Verb and Noun Number in English: A Functional Explanation. London: Longman. Reid, Wallis, Ricardo Otheguy, and Nancy Stern (eds.). 2002. Signal, Meaning, and Message: Perspectives on Sign-Based Linguistics. Amsterdam/Philadelphia: John Benjamins. Tobin, Yishai. 1997. Phonology as Human Behavior: Theoretical Implications and Clinical Applications. Durham: Duke U Press. * * * * * * * * For more information, please contact Joseph Davis at jdavis at ccny.cuny.edu From yutamb at mail.ru Mon Sep 20 19:54:12 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Tue, 21 Sep 2010 02:54:12 +0700 Subject: References to your own articles Message-ID: Dear Funknet colleagues, selfreferences are prohibited. But what can I do if nobody writes anything in my narrow field of research? If I omit references to my own articles, then there'd be no references at all. What can you advise me? Is it not noble to refer to your own opus? Looking forward to hearing from you soon to yutamb at mail.ru Bewell, Yuri Tambovtsev From yutamb at mail.ru Wed Sep 22 11:23:33 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Wed, 22 Sep 2010 18:23:33 +0700 Subject: Language variation Message-ID: Dear Funknet colleagues, I study language variation on phonetic and semantic level. I mean I research how the speech souns vary in different languages from the point of view of their occurrence. That is, how many particular phonemes occur in this or that world language. I studied the occurrence of phonemes in the speech chain of some 300 world languages. However, in every language speech sound chain the element vary. I study their variation with the help of the coefficient of variation. I wonder if I could compare my results to some other results of this sort. Who studies the variation of the occurrence speech sounds in different languages now? Or is it not in fashion any more? It was very fashionable in the 1960-1970. The other question is if it is all right to use the coefficient of variation to study the occurrence of speech sounds? Looking forward to your advice to yutamb at mail.ru Yours sincerely Yuri Tambovtsev From francisco.ruizdemendoza at unirioja.es Wed Sep 22 13:26:59 2010 From: francisco.ruizdemendoza at unirioja.es (=?iso-8859-1?Q?=22Francisco_Jos=E9_Ruiz_De_Mendoza_Ib=E1=F1ez=22?=) Date: Wed, 22 Sep 2010 15:26:59 +0200 Subject: 44th Annual Meeting of the Societas Linguistica Europaea Sept 8-11, 2011 University of La Rioja Message-ID: Dear list members, The first call for papers for the 44th Annual Meeting of the Societas Linguistica Europaea, to be held at the University of La Rioja, September 8-11, 2011 can be found at: http://sle2011.cilap.es/call-for-papers Best regards, Francisco Ruiz de Mendoza =========================== Francisco Ruiz de Mendoza www.cilap.es www.lexicom.es From djh514 at york.ac.uk Wed Sep 22 17:23:29 2010 From: djh514 at york.ac.uk (Damien Hall) Date: Wed, 22 Sep 2010 18:23:29 +0100 Subject: Language variation In-Reply-To: Message-ID: Dear Yuri > I research how the speech souns vary in different > languages from the point of view of their occurrence. That is, how many > particular phonemes occur in this or that world language. I studied the > occurrence of phonemes in the speech chain of some 300 world languages. > However, in every language speech sound chain the element vary. I study > their variation with the help of the coefficient of variation. I wonder > if I could compare my results to some other results of this sort. Who > studies the variation of the occurrence speech sounds in different > languages now? If you can get hold of it, you should have a look at work like that of Ian Maddieson http://en.wikipedia.org/wiki/Ian_Maddieson and Peter Ladefoged (particularly his 1996 book with Ian Maddieson and his 2001 sole-authored volume): http://en.wikipedia.org/wiki/Peter_Ladefoged#Bibliography > The other question is if it is all right to use the > coefficient of variation to study the occurrence of speech sounds? In order to know what to say about this, I'd need to know more about how you calculate the coefficient. I might not be able to say more even then, but that would be a start! Sorry not to be more helpful, but good luck! Best wishes Damien Hall -- Damien Hall University of York Department of Language and Linguistic Science Heslington YORK YO10 5DD UK Tel. (office) +44 (0)1904 432665 (mobile) +44 (0)771 853 5634 Fax +44 (0)1904 432673 http://www.york.ac.uk/res/aiseb http://www.york.ac.uk/depts/lang/people/pages/hall.htm DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm From yasshiraijp at gmail.com Wed Sep 22 18:00:43 2010 From: yasshiraijp at gmail.com (Yasuhiro Shirai) Date: Wed, 22 Sep 2010 14:00:43 -0400 Subject: Hispanic Linguistics Position (open-rank), University of Pittsburgh In-Reply-To: Message-ID: Hispanic Linguistics The University of Pittsburgh, Department of Linguistics, invites applications for a tenure stream professor (open rank) position in Hispanic Linguistics, pending budgetary approval. Applicants *must* have a solid training in a core area of linguistics. Competence in one or more of the following areas will be an advantage: second language acquisition, discourse analysis, bilingualism, language policy, and bilingual education. Candidates should send a CV (including a list of funded research if applicable), a statement of research and teaching interests, copies of 2 reprints or other written work, teaching evaluations (if available), and the names of three references. Hard copies of reference letters should be sent *directly* to the search committee. Send materials to: Search Committee, Hispanic Linguistics, Dept. of Linguistics, 2816 CL, University of Pittsburgh, Pittsburgh, PA 15260; (412) 624 5900; Fax (412) 624 6130. E-mail inquiries should be directed to Professor Yasuhiro Shirai, Chair, at yshirai at pitt.edu. The web page for the department is http://www.linguistics.pitt.edu/*.* To ensure full consideration, complete applications should be received by December 1, 2010. Preliminary interviews will be held at the LSA meeting in Pittsburgh in January 2011. The University of Pittsburgh is an Affirmative Action, Equal Opportunity Employer. Women and members of minority groups under-represented in academia are especially encouraged to apply. Yasuhiro Shirai Professor and Chair Department of Linguistics 2801 Cathedral of Learning University of Pittsburgh Pittsburgh, PA 15260 Tel: 412-624-5933 URL: http://www.linguistics.pitt.edu/people/faculty/shirai.htm From d.f.lesley-neuman.2 at hum.leidenuniv.nl Wed Sep 22 18:55:01 2010 From: d.f.lesley-neuman.2 at hum.leidenuniv.nl (Lesley-Neuman, D.F.) Date: Wed, 22 Sep 2010 20:55:01 +0200 Subject: Variation of speech sounds Message-ID: I know of work that has been done on distinctive features rather than on phonemes: Jeff Mielke (2008) The Emergence of Distinctive Features. Oxford University Press. In this Mielke describes distinctive features as constituting a complex system. If this is indeed so, then like other phenomena in the physical and social sciences, it the system would obey a Zipfian distribution in which the frequency of any one distinctive feature would be inversely proportional to its rank in the frequency table. I do not know if anyone has checked out this hypothesis. Other work was done by the late George Clements. Clements (2009) The Role of Features in Phonological Inventories in Raimy & Cairns (2009). Based on a survey of 451 inventories he proposes principles Feature Bounding and Feature Economy. Unfortunately, I have found from looking at the historical evolution of Kwa languages in Stewart (1971) and those noted in the Nilotic literature that the principle of feature economy does not pan out with regards to historical changes in inventories common to African languages with vowel harmony. San Duanmu at the University of Michigan has been examining Clements proposals in other languages; he and his student Huili Zhang have upheld the feature economy principle in an analysis of the historical evolution of Chinese. This paper was presented last fall at the Michigan Linguistic Society, the program of which can be accessed here: http://ling.lsa.umich.edu/mls2009/program/MLS_Program_Posted.pdf I know that Dr. Duanmu has a very strong interest in this area and would probably gladly participate in a fruitful exchange on this issue. Diane Lesley-Neuman PhD Researcher Leiden University Centre for Linguistics/ Languages and Cultures of Africa Van Wijkplaats 4 Office 103A 2311 BX Leiden The Netherlands Email: d.f.lesley-neuman.2 at hum.leidenuniv.nl Telephone: +31 71 527-1663 From jkaplan at mail.sdsu.edu Thu Sep 23 22:25:26 2010 From: jkaplan at mail.sdsu.edu (Jeffrey P. Kaplan) Date: Thu, 23 Sep 2010 15:25:26 -0700 Subject: job posting Message-ID: Hi, How can one post a job ad on FUNKnet? Is it as simple as just sending you an e-mail containing the ad copy? Is there a charge? Thanks, Jeff Kaplan -- Jeffrey P. Kaplan Professor of linguistics Dept. of Linguistics & Asian/Middle Eastern Languages San Diego State University San Diego, CA 92182-7727 619-594-5879 fax 619-594-4877 http://www-rohan.sdsu.edu/~jeff315/ This e-mail is a natural product. The slight variations in spelling and grammar enhance its individual character and beauty and in no way are to be considered flaws or defects. From jkaplan at mail.sdsu.edu Fri Sep 24 01:53:00 2010 From: jkaplan at mail.sdsu.edu (Jeffrey P. Kaplan) Date: Thu, 23 Sep 2010 18:53:00 -0700 Subject: Sociolinguistics opening Message-ID: The Department of Linguistics and Asian/Middle Eastern Languages at San Diego State University seeks a tenure-track assistant professor with a specialization in sociolinguistics, to begin in fall 2011. Areas of particular interest include, but are not limited to, variation theory and language contact. Preference will be shown to candidates interested in carrying out field work in the linguistically and culturally diverse speech communities of San Diego. The successful candidate will show a promising research agenda, demonstrated excellence in teaching, and the potential to attract students from across the university. Candidates should send a letter of application, CV, documentation of teaching excellence, and sample publications; and should arrange for three letters of recommendation to be sent. All materials should be sent as e-mail attachments, if possible (MS WORD or .pdf files), to Dr. Jeffrey P. Kaplan, chair, Sociolinguistics Search Committee, at jkaplan at mail.sdsu.edu, or in hard copy form to Dept. of Linguistics & Asian/Middle Eastern Languages, San Diego State University, San Diego, CA 92182-7727. Screening will begin on Nov. 1, 2010 and will continue until the position is filled. San Diego State University is an equal opportunity employer and does not discriminate against persons on the basis of race, religion, national origin, sexual orientation, gender, gender identity and expression, marital status, age, disability, pregnancy, medical condition, or covered veteran status. -- Jeffrey P. Kaplan Professor of linguistics Dept. of Linguistics & Asian/Middle Eastern Languages San Diego State University San Diego, CA 92182-7727 619-594-5879 fax 619-594-4877 http://www-rohan.sdsu.edu/~jeff315/ This e-mail is a natural product. The slight variations in spelling and grammar enhance its individual character and beauty and in no way are to be considered flaws or defects. From rdbusser at gmail.com Fri Sep 24 05:40:45 2010 From: rdbusser at gmail.com (Rik) Date: Fri, 24 Sep 2010 15:40:45 +1000 Subject: Research Centre for Linguistic Typology Podcast Message-ID: (Apologies for cross-posting) The Research Centre for Linguistic Typology at La Trobe University is proud to announce its own podcast on iTunesU: http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewiTunesUInstitution?id= 380435656 (Alternatively, go to iTunesU in the iTunes store and search for "Research Centre for Linguistic Typology") In the future, we aim at making all talks given at RCLT freely available to the public through this channel, as long as speakers give permission. At the moment, three are available for download: - Prof. Randy LaPolla, "Preserving Languages of the World" - Prof. Christian Lehmann, "Distance Iconicity" - Prof. Christian Lehmann, "The Function of Numeral Classifiers" Some of them might be of interest to the members of Funknet. Rik ------------------------------ Rik De Busser Research Centre for Linguistic Typology La Trobe University, Bundoora 3086 VIC, Australia T: +61 3 9479 6413 www.rdbusser.com From Florence.Chenu at univ-lyon2.fr Fri Sep 24 20:18:02 2010 From: Florence.Chenu at univ-lyon2.fr (Florence Chenu (FC)) Date: Fri, 24 Sep 2010 22:18:02 +0200 Subject: [conference] 4th International conference of the French Cognitive Linguistics Association (AFLiCo IV; 24-27 May 2011; Lyon, France) Message-ID: AFLiCo IV Fourth International Conference of the French Cognitive Linguistics Association, Lyon, France, 24th-27th May 2011 (French version follows) CONFERENCE THEME of AFLiCo IV The theme of the 2011 conference is: 'Cognitive Linguistics and Typology: Language diversity, variation and change '. This conference aims to bring together linguists engaged in cognitively-oriented research with those working in a functional-typological framework on cross-linguistic variation and on language description. The emphasis will be on (1) language diversity of both spoken and signed languages; (2) inter- and intra-linguistic variation; (3) language change. The conference will bring together linguists working with various methodological approaches and using various kinds of spontaneous and elicited data, including spoken and written corpora, fieldwork data, and experimental data. Proposals are invited for workshops/thematic sessions, for general session papers, and for posters, on topics related to the theme, and on topics in Cognitive Linguistics generally. Papers that report empirically-grounded research on less-studied languages and on typologically, genetically and areally diverse languages will be particularly welcome. Topics include, but are not limited to: - methods and data in cognitive linguistics and in language typology and description - convergence and divergence between cognitive linguistics and functional-typological linguistics - studies from a cognitive and/or typological perspective in phonetics, phonology, morphosyntax, semantics and pragmatics - language variation within and across languages, both spoken and signed - language change from a cognitive and/or typological perspective - language acquisition - studies and advances in construction grammar - language and gesture in cross-linguistic perspective LANGUAGES OF THE CONFERENCE The languages of the conference are French and English. ORAL PRESENTATIONS AND POSTERS Proposals are invited for 30-minute slots (20-minute presentation plus question time) in the general sessions and for posters (A1 size). WORKSHOPS, INCLUDING THEMATIC SESSIONS Proposals are invited for half-day or full-day workshops/thematic sessions. Each workshop proposal should contain the following information: - the names and contact details of two workshop organizers - the title of the proposed workshop - an overview of the topic and aims of the workshop (up to 2 pages) - an indication of the desired schedule (number of slots: 4, 6 or 10; half day or full day; number and nature of presentations, discussions, round tables, etc. that the workshop will comprise). Note that, within a workshop, each presentation, discussion or round table will occupy one 30-minute slot in parallel with one general session slot. - an abstract (consistent with the indications below under 'Submission procedure') for each proposed 30-minute presentation Workshop proposals will be refereed in the same way as general session and poster proposals. SUBMISSION PROCEDURE Proposals should be submitted online following the instructions to be found at the following address: http://www.ddl.ish-lyon.cnrs.fr/colloques/AFLICO_IV/ Author information (name, affiliation, email address) will be required on the submission website. An author may submit a maximum of two abstracts, of which at least one must be co-authored. In the case of co-authored abstracts, the first-named author will be the contact person. Abstracts will be anonymously reviewed and notification of acceptance will be sent out from 25th February 2011. The anonymous abstracts must be in 12 point Times or Times New Roman font, formatted for A4 or US Letter size paper with margins of 2.5 cm or 1 inch. The maximum length for the text of the abstract is one page; a second page may be used only for figures, glossed examples and bibliographical references. Deadline for general session papers: 22nd December 2010 Deadline for workshops/thematic sessions: 18th December 2010 ******************************************************* ******************************************************* AFLiCo IV Quatri?me Colloque International de l?Association Fran?aise de Linguistique Cognitive Lyon, France, 24-27 Mai 2011 TH?ME DU COLLOQUE AFLiCo IV ?Linguistique cognitive et typologie : diversit? des langues, variation et changement?. L?objectif de ce colloque est de r?unir des linguistes travaillant dans le domaine de la linguistique cognitive et/ou dans le domaine de la linguistique fonctionnelle-typologique sur la variation inter-linguistique et la description des langues. L?accent du colloque sera mis sur (1) la diversit? des syst?mes linguistiques aussi bien oraux que sign?s, (2) la variation qui s?op?re sur les plans inter- et intra- linguistiques et (3) les changements des syst?mes linguistiques. Dans cette perspective, le colloque rassemblera des chercheurs qui travaillent sur des terrains linguistiques vari?s, qui abordent leur objet d??tude dans une perspective synchronique et/ou diachronique et qui utilisent diff?rentes m?thodes et diff?rents types de donn?es telles que des donn?es spontan?es ou ?licit?es, y compris orales ou ?crites, des donn?es de terrain ou encore des donn?es exp?rimentales. Nous attendons des propositions de sessions th?matiques, des propositions de pr?sentations orales de sessions g?n?rales et de posters sur des probl?matiques en lien avec le th?me du colloque et dans le domaine de la linguistique cognitive en g?n?ral. Les propositions portant sur des langues moins bien d?crites et des langues qui varient du point de vue typologique, g?n?tique et ar?al seront particuli?rement appr?ci?es. Les th?matiques incluent, mais ne se limitent pas aux suivantes : - m?thodes et donn?es en linguistique cognitive, typologie et description des langues ; - convergence et divergence entre linguistique cognitive et linguistique fonctionnelle-typologique ; - ?tudes men?es dans une perspective cognitive et/ou typologique dans les domaines de la phon?tique, phonologie, morphosyntaxe, s?mantique et pragmatique ; - variation inter- et intra-linguistique dans les langues parl?es et les langues sign?es ; - changements linguistiques dans une perspective cognitive et/ou typologique ; - acquisition du langage ; - recherches et avanc?es dans le domaine de la grammaire des constructions ; - langue et geste dans une perspective inter-linguistique. LANGUES OFFICIELLES DU COLLOQUE Les deux langues du colloque sont le fran?ais et l?anglais. COMMUNICATIONS ET POSTERS Nous invitons des propositions de communication aux sessions g?n?rales de 30 minutes (20 minutes de pr?sentation et 10 minutes de questions) et des propositions de posters (format A1). ATELIERS ET SESSIONS TH?MATIQUES Nous accueillons des propositions d?une demi-journ?e ou d?une journ?e enti?re pour des ateliers et/ou sessions th?matiques. Ces ateliers/sessions th?matiques doivent ?tre propos?s par deux organisateurs. Chaque proposition doit inclure les informations suivantes : - les noms et les coordonn?es des deux organisateurs - le titre de la session - une pr?sentation du th?me et des objectifs de la session (2 pages maximum) - une pr?cision concernant le temps souhait? (nombre de cr?neaux horaires : 4, 6 ou 10 ; une journ?e ou une journ?e enti?re ; nombre et nature des pr?sentations, discussions, tables rondes, etc.). - un r?sum? d?une page pour chaque pr?sentation (une deuxi?me page peut ?tre utilis?e pour des figures, exemples glos?s et r?f?rences bibliographiques) Les propositions d?ateliers et/ou de sessions th?matiques seront soumises ? la m?me proc?dure d??valuation que les propositions pour les sessions g?n?rales et les posters. La notification d?acceptation sera envoy?e aux deux organisateurs ? partir du 25 f?vrier 2011. SOUMISSION DES PROPOSITIONS Les propositions seront soumises en ligne suivant les instructions indiqu?es ? l?adresse suivante : http://www.ddl.ish-lyon.cnrs.fr/colloques/AFLICO_IV/ Un auteur ne peut soumettre que deux propositions de communication dont une au moins devrait ?tre en co-auteur. Les informations concernant l?auteur (nom, affiliation, adresse email) seront requises lors de la soumission en ligne mais les propositions seront ?valu?es de fa?on anonyme. Dans le cas des propositions en co-auteur le premier auteur sera la personne r?f?rente/contact. Les propositions seront examin?es de fa?on anonyme par 2 membres experts du comit? scientifique. La notification d?acceptation sera envoy?e aux auteurs ? partir du 25 f?vrier 2011. Les propositions ne devront pas d?passer une page. Une deuxi?me page peut ?tre utilis?e pour des figures, exemples glos?s et r?f?rences bibliographiques. Format des propositions : papier A4, marges 2,5 cm, police Times ou Times New Roman. Date limite pour les sessions g?n?rales : 22 d?cembre 2010 Date limite pour les sessions th?matiques : 18 d?cembre 2010 From mg246 at cornell.edu Sat Sep 25 22:17:01 2010 From: mg246 at cornell.edu (monica gonzalez-marquez) Date: Sun, 26 Sep 2010 00:17:01 +0200 Subject: CFP: EMCL 5.1 - Freiburg, Germany Message-ID: Empirical Methods in Cognitive Linguistics 5.1 -- Freiburg **** March 6 -- 11, 2011 https://sites.google.com/site/emcl5freiburg/ --------------------Application deadline: December 15, 2010---------------------- We invite applications for the, 5th Empirical Methods in Cognitive Linguistics workshop, to be held in, Freiburg, Germany, March 6 -- 11, 2011 The goal of EMCL is to facilitate dialogue among language researchers with different methodological backgrounds, i.e. theorists, experimentalists, corpus linguists, etc. We do this by creating an environment where specialists learn from each other by developing a research project together where their various skills are combined. Intended audience: Language researchers with an embodiment, situated cognition and/or cognitive linguistics background. No prior experimental or corpus training is required though an understanding of the theoretical issues is necessary. Participants can be at different early stages in their careers, i.e. graduate students, post-grads, post-docs, junior faculty, etc. Format: During the course of a week, participants will join one of 5 hands-on mini-labs. Each mini-lab will be responsible for completing a joint research project. A select group of students (max. 8 per group for a total of 40***) will be invited to participate. Each group will work with two researchers who will guide the group in selecting an idea for the group to investigate, structuring and organizing a research project, and carrying it out. The session will end with the presentation of findings and a general discussion. Topics to be covered include, - Deciding on a research topic - Transforming the research topic into a research question - Developing experimental hypotheses and designing an experiment - Data collection - Statistical analysis and interpretation - Presentation of findings to an audience Workshop Faculty Group 1: Rolf Zwaan University of Rotterdam Interests: The relationship between cognition and systems of perception, action, and emotion, Language comprehension, Embodied cognition http://www.brain-cognition.eu/index.html?personal.php?id=Zwaan Alan Cienki, Vrije Universiteit Interests: cognitive linguistics, pragmatics, spoken language, gesture, political discourse, contrastive linguistics http://www.let.vu.nl/en/about-the-faculty/academic-staff/staff-listed-alphabetically/staff-a-d/dr-a-cienki/index.asp Group 2: Kenny Coventry, Northumbria University Interests: language and perception, spatial language, embodiment, decision making http://kenny.coventry.googlepages.com/home Katharina Rohlfing, Bielefeld University Interests: emergentist semantics, early literacy, human-machine interaction, rhetoric and communication https://www.techfak.uni-bielefeld.de/~rohlfing/website/data/index.html http://www.cit-ec.de/es Group 3: Lars Konieczny, University of Freiburg Interests: Theoretical, Empirical, and Computational Psycholinguistics, Eye-movements research, Reading, Spoken language comprehension in the Visual-World-paradigm, Spatial reasoning and wayfinding, Cognitive modeling (ACT-R, Connectionist Modeling), Embodied Cognition http://portal.uni-freiburg.de/cognition/Members/konieczny Michele Feist, University of Louisiana at Lafayette Interests: lexical semantics, spatial language, psycholinguistics, acquisition of semantics, language and cognition http://www.ucs.louisiana.edu/~mif8232/ Group 4: Seana Coulson, University of California, San Diego Interests: Conceptual Blending, Joke Comprehension, Metaphor, Analogical Reasoning, Concept Combination, Sentence Processing http://www.cogsci.ucsd.edu/~coulson/ Panos Athanasopoulos, Bangor University Interests: Bilingualism and Cognition, Language and Thought, Emotion, Language Acquisition, conceptual development http://www.bangor.ac.uk/linguistics/about/panos.php.en Group 5: Pia Knoeferle, Bielefeld University Interests: Influences of visual context on real-time language comprehension, Picture-sentence verification, Models and processing accounts of situated language comprehension http://wwwhomes.uni-bielefeld.de/pknoeferle/Staff_PK/PK.html Anatol Stefanowitsch, University of Hamburg Interests: Encoding of motion events, Second language research, Construction Grammar, Quantitative Corpus Linguistics, Metaphor, Negative evidence http://www1.uni-hamburg.de/stefanowitsch/ Accommodation: Accommodation at walking distance to the university will be arranged for all student participants. Cost will be EUR20 per night. (We "may" receive funds to cover student accommodation, in which case all applicants will be notified.) Participation Fee: EUR125 **, payable by bank transfer or upon arrival by prior arrangement. (This fee helps cover the costs of organization and faculty travel.) Application: To apply, please send the following by December 15, 2010. All materials must be submitted electronically to emcl5.freiburg (at) googlemail.com PLEASE WRITE 'APPLICATION' IN THE SUBJECT LINE. 1. A maximum of two (2) pages, (1000 words), describing, - your background, - your reasons for wanting to participate, - the research group you would like to work in and why. Please include in this section a brief description of your research interests. All topics listed above must be addressed. Incomplete applications will not be reviewed. 2. A copy of your curriculum vitae. The application deadline is December 15, 2010 Accepted applicants will be notified on or before January 15, 2011 This workshop is supported by: the FRIAS at Freiburg University http://www.frias.uni-freiburg.de/ the Research training group (GRK DFG 1624/1) Frequency effects in language http://frequenz.uni-freiburg.de/abstract&language=de and the DFG (pending). www.dfg.de ** 2 (two) tuition scholarships will be awarded by lottery to students traveling from Eastern Europe and 3rd world countries. Please state in your application whether you would like to be included in the lottery. *** Please note: Attendance is strictly limited to invited participants. No exceptions will be made so as to preserve pedagogical integrity. **** EMCL 5.2 will be held in Chicago, USA, June 2011 with a different set of faculty. That notice will follow in January, 2011. --- EMCL 5 Organizing Committee: Monica Gonzalez-Marquez, Chair, Cornell University Martin Hilpert, University of Freiburg Pia Knoeferle, Bielefeld University Lars Konieczny, University of Freiburg From hallowel at ohio.edu Mon Sep 27 18:34:19 2010 From: hallowel at ohio.edu (Hallowell, Brooke) Date: Mon, 27 Sep 2010 14:34:19 -0400 Subject: indexing word difficulty Message-ID: Dear colleagues: We are working on deriving a list of "easy" and "difficult" words for an auditory comprehension study in American English. As we're unaware of databases of words indexed by "difficulty," we are, as many have done, approximating an index of difficulty by using multiple indices, including word length, familiarity, imageability, and frequency. Do know of a more index of word difficulty that more explicitly taps how hard it is to understand a word? Thanks very much for any suggestions. Brooke Hallowell Brooke Hallowell, Ph.D., CCC-SLP, F-ASHA President, Council of Academic Programs in Communication Sciences and Disorders Professor and Director, Neurolinguistics Laboratory Communication Sciences and Disorders School of Rehabilitation and Communication Sciences College of Health Sciences and Professions W237 Grover Center Ohio University Athens, OH 45701 USA From crm5 at rice.edu Mon Sep 27 22:55:32 2010 From: crm5 at rice.edu (crm5 at rice.edu) Date: Mon, 27 Sep 2010 17:55:32 -0500 Subject: Call for papers: Rice Working Papers in Linguistics Vol 3 Message-ID: Dear Funknet members, The Rice Working Papers in Linguistics is currently soliciting submissions for its third volume (you can see published volumes at http://owling.blogs.rice.edu/rwpl-vol-1/ and http://owling.blogs.rice.edu/rwpl-vol-2/). The deadline is November 15th. Please see the guidelines below and consider submitting your work to rwpl at rice.edu. Carlos Molina-Vital Rice Working Papers in Linguistics EIC *** Rice Working Papers in Linguistics, Volume 3 Deadline: **November 15th, 2010** The Rice Linguistics Society (RLS) solicits submissions from all subfields of linguistics (with the exception of ESL/TESOL and related areas of applied linguistics) for publication in the Rice Working Papers in Linguistics. Students and post-docs are strongly encouraged to submit. We especially welcome submissions in line with our department's focus on functional, usage-based approaches to language study using empirical data, including but not limited to the following: -cognitive/functional linguistics -typology and language universals -field studies in less commonly researched languages -sociolinguistics, including sociophonetics -phonetics and speech processing -laboratory phonology -forensic linguistics -corpus linguistics -discourse -neurolinguistics -psycholinguistics and language processing -language change and grammaticalization Submitted papers must meet the following minimum style requirements: -recommended length 15-25 pages (normally 5000-8000 words); significantly longer or shorter papers will be considered on a case-by-base basis (contact the editorial board) -For comprehensive details on format (such as font, margins, examples, references, etc.) please refer to the RWPL template available on the Style sheet link at http://owling.blogs.rice.edu/ -submit an abstract (maximum 500 words), including 3-5 keywords, as a separate Word file -submit two copies (in addition to your abstract): (1) one copy in Word (2003 or 2007) (2) in addition to the Word submission, you must send a PDF version to ensure fonts are preserved RLS accepts only electronic submissions for the working papers. These must be sent to rwpl at rice.edu, and the body of the e-mail should include: -title of paper -name of author(s) -affiliation -address -phone number -contact e-mail address The deadline for receipt of submissions is **November 15th, 2010**. Questions regarding the submissions process or style requirements may be addressed to the editorial board at rwpl at rice.edu. ******* From yutamb at mail.ru Wed Sep 29 09:01:31 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Wed, 29 Sep 2010 16:01:31 +0700 Subject: Variation of different functions of gerund and participle Message-ID: Dear Funknet List members, do you use the coefficient of variation and Chi-square to study functioning of phonemes or prepositions in language? In fact, the application of coefficient of variation and Chi-square to investigate the variation of linguistic elements in language may stop endless debates about language variation problems. I use them all right. They proved quite useful. With their help I also studied variation of phomenes and groups of phonemes (labials, velars, sonorants, fricatives, etc.) across languages. Usually they were used to study the variation of phonemes in texts. I did that as well. The degree of variation of different functions of gerund and participle allowed us to distinquish between authors. It contributed much to the theory of authorship. I wonder if you read my publications? Who is researching in the same area? I call this area phonostatistical typology. Looking forward to hearing from you soon to yutamb at mail.ru Yours sincerely Yuri Tambovtsev, Novosibirsk, Russia From yutamb at mail.ru Wed Sep 29 20:10:52 2010 From: yutamb at mail.ru (Yuri Tambovtsev) Date: Thu, 30 Sep 2010 03:10:52 +0700 Subject: Different authors use different functions differently Message-ID: Dear Funknet colleagues, if corpora of English texts is not one unity but a conglomerate, how can we use statistic criteria? I started the discussion what tools to use for the analysis because many of linguists do not use all these complex statistical packets correctly. The other thing is who knows what and how the data are being analysed in them. We have a sort of a "black box" which has the entrance in which you put your data and the outcome where you receive your results. You must be quite sure that the data are analysed correctly. The more simple criteria you use, the better. This is why I stopped using all the stat. packets and began using very simple criteria like the coefficient of variation, the Chi-square and the t-test. At least, I am sure about the outcome. More often than not, it is advisable to try you data manually than to use the statpack. Looking forward to hearing from you if you agree with me that using statpacks may give you strange results. The fact that so many scholar spoke on the list about how to use the statpacks showed that it is a burning question for linguists and other researchers who are not specialists in mathematics. Yours sincerely Yuri Tambovtsev, yutamb at mail.ru From tgivon at uoregon.edu Thu Sep 30 06:03:25 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Thu, 30 Sep 2010 00:03:25 -0600 Subject: A poet slips Message-ID: A GREAT POET CAN STILL GET IT WRONG I came to know of Salman Masalha ("Arabs, speak Hebrew!", Haaretz/English, International Herald Tribune 9-27-10) accidentally by stumbling a on his truly great quote: "All fixed identities are imposed from the outside. Whoever has a clear identity knows it can assume multiple forms". In the context of Palestine/Israel, what a breath of fresh, rare clarity. Still, like the rest of us mortals, a great poet can on occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got only one third of the story right. At first, he was led astray be the academic researchers he cited, who claimed that the lagging reading skills of Israeli-Arab students is correlated to lagging R-hemisphere activity, then explained this neurological lag by suggesting that the Arab script requires more contextual analysis. But it is the R-hemisphere of the human cortex that is more context oriented, less automated. If Arab-reading students required more contextual labor, it should have been registered as a higher R-hemisphere activity, not lower. Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among many others, have established beyond reasonable doubt that written words in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere module on the boundary of the occipital and temporal lobes (just past Brodman's Area 19), along the ventral visual object-recognition 'stream' that flows from the back to the front of the L-cortex. And the L-cortex is in general responsible for the more automated--less context-dependent--processing of language (as well as visual, motor and other skills). The visual word-recognition module is, in turn, recruited from the pre-existing visual object-recognition ventral-stream module. A considerable amount of life-time practice and repetition is required to affect this late-cultural adaptation. The human brain is not (yet) genetically configured at birth for visual-word recognition, only for visual-object recognition. A similar cultural adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) in the L-pareita lobe, an area originally configured for analysis of object-location in space. Mr. Masalha then, on his own, points out to a more plausible right answer: Arab students, in Israel as well as all over the Arab world, are not taught literacy in their native language (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 years or more. That is, in a foreign language. The discrepancy would be just as great if Israeli kids were taught their Hebrew literacy first in the language of Genesis; or if French students were taught literacy first in the language of La Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or English-speaking kids in the language of Beowolf. As far as my frail guessing powers go, remedying the situation would be much easier by combining two well-known verities of second language acquisition: (a) Teach them both early, together--'co-ordinated bilingualism'. And (b), teach literacy first in the student's spoken native language; only then gradually 'stretch' it to more literary genres. This method, bhy the way, was suggested in the late 1930's by no other than L. Bloomfield, in a book outlining a 'phonics-first ' literacy program for native English speakers. Rather than depriving Israeli-Arab students of literacy in their own--equally glorious--native language, just teach them smart. For his last culprit, the presumed--tho hardly unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized prejudices of the educated classes. While readily endorsing his aesthetic sentiments about modern media, I would still like to point out that the 'vulgar' genre is much closer to the Arab students' spoken native language, and if anything should facilitate the easier initial acquisition of native-language literacy. Respectuosamente, ma'-salaam, T. Giv?n From john at research.haifa.ac.il Thu Sep 30 08:06:57 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 10:06:57 +0200 Subject: A poet slips In-Reply-To: <4CA4282D.6080303@uoregon.edu> Message-ID: Tom, I've been trying for several years to get Israeli Arabs and Jews to seriously consider the possibility of educating Israeli Arabs in a written version of their spoken language, as you suggest, at least through 3rd grade. I can send you some things I've written on this topic, with a lot of cross-linguistic data. But thus far, it isn't working. As with most aspects of 'the situation' here, politically active Arabs think that the solution is to pursue the same self-destructive strategy which they've been following for almost a century but with even more vigor and steadfastness (e.g. Masalha points the blame at the Arab media because they do not 'provide the linguistic richness of formal Arabic') while Jews are basically content to let the Arabs stew in their own juices so that they can reap the benefits. There may, however, be hope for the future in the form of the radically increased usage of written forms of colloquial Arabic dialects in electronic media such as Facebook, blogs, emails, etc., by Arabic speakers below the age of 30, which will soon come to seriously threaten the status of classical Arabic in the same way that the invention of the printing press overturned the linguistic hierarchy in Western Europe by overthrowing Latin. I'm working on this too. We'll see what happens. I realize that Masalha as well as the researchers he referred to don't necessarily know what they're talking about regarding the brain. But the general point is still potentially significant--that the connected script which Arabic uses as well as the multiple forms which many Arabic letters have may be a significant obstacle to literacy, however this may be related to processes in the brain. Best wishes, John Quoting Tom Givon : > > > A GREAT POET CAN STILL GET IT WRONG > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > Haaretz/English, International Herald Tribune 9-27-10) accidentally > by stumbling a on his truly great quote: > "All fixed identities are imposed from the outside. > Whoever has a clear identity knows it can assume > multiple forms". > In the context of Palestine/Israel, what a breath of fresh, rare > clarity. Still, like the rest of us mortals, a great poet can on > occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > intended--surely got only one third of the story right. At first, he was > led astray be the academic researchers he cited, who claimed that the > lagging reading skills of Israeli-Arab students is correlated to > lagging R-hemisphere activity, then explained this neurological lag > by suggesting that the Arab script requires more contextual analysis. > But it is the R-hemisphere of the human cortex that is more context > oriented, less automated. If Arab-reading students required more > contextual labor, it should have been registered as a higher > R-hemisphere activity, not lower. > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > among many others, have established beyond reasonable doubt that > written words in all languages (English, Mandarin, Hindi, Amharic, > Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > module on the boundary of the occipital and temporal lobes (just past > Brodman's Area 19), along the ventral visual object-recognition > 'stream' that flows from the back to the front of the L-cortex. And > the L-cortex is in general responsible for the more automated--less > context-dependent--processing of language (as well as visual, motor and > other skills). The visual word-recognition module is, in turn, > recruited from the pre-existing visual object-recognition > ventral-stream module. A considerable amount of life-time practice and > repetition is required to affect this late-cultural adaptation. The > human brain is not (yet) genetically configured at birth for > visual-word recognition, only for visual-object recognition. A similar > cultural adaptation, this one for math, has been shown for (Dahaene & > Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > in the L-pareita lobe, an area originally configured for analysis of > object-location in space. > Mr. Masalha then, on his own, points out to a more > plausible right answer: Arab students, in Israel as well as all over the > Arab world, are not taught literacy in their native language > (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > instrument harking back 1,400 years or more. That is, in a foreign > language. The discrepancy would be just as great if Israeli kids were > taught their Hebrew literacy first in the language of Genesis; or if > French students were taught literacy first in the language of La > Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or > English-speaking kids in the language of Beowolf. As far as my frail > guessing powers go, remedying the situation would be much easier by > combining two well-known verities of second language acquisition: (a) > Teach them both early, together--'co-ordinated bilingualism'. And (b), > teach literacy first in the student's spoken native language; only then > gradually 'stretch' it to more literary genres. This method, bhy the > way, was suggested in the late 1930's by no other than L. Bloomfield, > in a book outlining a 'phonics-first ' literacy program for native > English speakers. Rather than depriving Israeli-Arab students of > literacy in their own--equally glorious--native language, just teach > them smart. > For his last culprit, the presumed--tho hardly > unique--vulgarity of Arab media, Mr. Masalha lapses into well > recognized prejudices of the educated classes. While readily endorsing > his aesthetic sentiments about modern media, I would still like to point > out that the 'vulgar' genre is much closer to the Arab students' > spoken native language, and if anything should facilitate the easier > initial acquisition of native-language literacy. Respectuosamente, > ma'-salaam, > > > T. Giv?n > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From tgivon at uoregon.edu Thu Sep 30 08:25:21 2010 From: tgivon at uoregon.edu (Tom Givon) Date: Thu, 30 Sep 2010 02:25:21 -0600 Subject: A poet slips In-Reply-To: <1285834017.4ca445214a9c7@webmail.haifa.ac.il> Message-ID: When I learned to read Arabic on the kibbutz (Maabarot) as a child, our textbook was of written COLLOQUIAL Falastini Arabic. We never got to Classical (after 1949 things changed...). That book may still exist, you might track it down. It was easy, a cinch really. I concluded it was really just Hebrew with a few trivial transformations in Phonology & Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat on a few sessions of a faculty study group at UO who were trying to learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, it was like pullin' teeth. But Leonard Bloomfield said it already, and well, in 1939 (or was it 1943?). Cheers, TG ============== john at research.haifa.ac.il wrote: > Tom, > I've been trying for several years to get Israeli Arabs and Jews to seriously > consider the possibility of educating Israeli Arabs in a written version of > their spoken language, as you suggest, at least through 3rd grade. I can send > you some things I've written on this topic, with a lot of cross-linguistic > data. But thus far, it isn't working. As with most aspects of 'the situation' > here, politically active Arabs think that the solution is to pursue the same > self-destructive strategy which they've been following for almost a century but > with even more vigor and steadfastness (e.g. Masalha points the blame at the > Arab media because they do not 'provide the linguistic richness of formal > Arabic') while Jews are basically content to let the Arabs stew in their own > juices so that they can reap the benefits. There may, however, be hope for the > future in the form of the radically increased usage of written forms of > colloquial Arabic dialects in electronic media such as Facebook, blogs, emails, > etc., by Arabic speakers below the age of 30, which will soon come to seriously > threaten the status of classical Arabic in the same way that the invention of > the printing press overturned the linguistic hierarchy in Western Europe by > overthrowing Latin. I'm working on this too. We'll see what happens. > > I realize that Masalha as well as the researchers he referred to don't > necessarily know what they're talking about regarding the brain. But the > general point is still potentially significant--that the connected script which > Arabic uses as well as the multiple forms which many Arabic letters have may be > a significant obstacle to literacy, however this may be related to processes in > the brain. > Best wishes, > John > > > > Quoting Tom Givon : > > >> A GREAT POET CAN STILL GET IT WRONG >> >> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >> Haaretz/English, International Herald Tribune 9-27-10) accidentally >> by stumbling a on his truly great quote: >> "All fixed identities are imposed from the outside. >> Whoever has a clear identity knows it can assume >> multiple forms". >> In the context of Palestine/Israel, what a breath of fresh, rare >> clarity. Still, like the rest of us mortals, a great poet can on >> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >> intended--surely got only one third of the story right. At first, he was >> led astray be the academic researchers he cited, who claimed that the >> lagging reading skills of Israeli-Arab students is correlated to >> lagging R-hemisphere activity, then explained this neurological lag >> by suggesting that the Arab script requires more contextual analysis. >> But it is the R-hemisphere of the human cortex that is more context >> oriented, less automated. If Arab-reading students required more >> contextual labor, it should have been registered as a higher >> R-hemisphere activity, not lower. >> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >> among many others, have established beyond reasonable doubt that >> written words in all languages (English, Mandarin, Hindi, Amharic, >> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >> module on the boundary of the occipital and temporal lobes (just past >> Brodman's Area 19), along the ventral visual object-recognition >> 'stream' that flows from the back to the front of the L-cortex. And >> the L-cortex is in general responsible for the more automated--less >> context-dependent--processing of language (as well as visual, motor and >> other skills). The visual word-recognition module is, in turn, >> recruited from the pre-existing visual object-recognition >> ventral-stream module. A considerable amount of life-time practice and >> repetition is required to affect this late-cultural adaptation. The >> human brain is not (yet) genetically configured at birth for >> visual-word recognition, only for visual-object recognition. A similar >> cultural adaptation, this one for math, has been shown for (Dahaene & >> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >> in the L-pareita lobe, an area originally configured for analysis of >> object-location in space. >> Mr. Masalha then, on his own, points out to a more >> plausible right answer: Arab students, in Israel as well as all over the >> Arab world, are not taught literacy in their native language >> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >> instrument harking back 1,400 years or more. That is, in a foreign >> language. The discrepancy would be just as great if Israeli kids were >> taught their Hebrew literacy first in the language of Genesis; or if >> French students were taught literacy first in the language of La >> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >> English-speaking kids in the language of Beowolf. As far as my frail >> guessing powers go, remedying the situation would be much easier by >> combining two well-known verities of second language acquisition: (a) >> Teach them both early, together--'co-ordinated bilingualism'. And (b), >> teach literacy first in the student's spoken native language; only then >> gradually 'stretch' it to more literary genres. This method, bhy the >> way, was suggested in the late 1930's by no other than L. Bloomfield, >> in a book outlining a 'phonics-first ' literacy program for native >> English speakers. Rather than depriving Israeli-Arab students of >> literacy in their own--equally glorious--native language, just teach >> them smart. >> For his last culprit, the presumed--tho hardly >> unique--vulgarity of Arab media, Mr. Masalha lapses into well >> recognized prejudices of the educated classes. While readily endorsing >> his aesthetic sentiments about modern media, I would still like to point >> out that the 'vulgar' genre is much closer to the Arab students' >> spoken native language, and if anything should facilitate the easier >> initial acquisition of native-language literacy. Respectuosamente, >> ma'-salaam, >> >> >> T. Giv?n >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > From john at research.haifa.ac.il Thu Sep 30 08:40:23 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 10:40:23 +0200 Subject: A poet slips In-Reply-To: <4CA44971.2050105@uoregon.edu> Message-ID: Learning the classical language is like pulling teeth for Arabs kids also. They just can't publicly say it because that would make them bad Arabs. It's just ridiculous. I have an even better plan for the Jewish kids--teach them to use the written version of the spoken language which Arab kids are using for Facebook. That's how kids make friends these days anyway. Best wishes, John Quoting Tom Givon : > > > When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > textbook was of written COLLOQUIAL Falastini Arabic. We never got to > Classical (after 1949 things changed...). That book may still exist, > you might track it down. It was easy, a cinch really. I concluded it was > really just Hebrew with a few trivial transformations in Phonology & > Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > on a few sessions of a faculty study group at UO who were trying to > learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > it was like pullin' teeth. But Leonard Bloomfield said it already, and > well, in 1939 (or was it 1943?). > Cheers, TG > > ============== > > > john at research.haifa.ac.il wrote: > > Tom, > > I've been trying for several years to get Israeli Arabs and Jews to > seriously > > consider the possibility of educating Israeli Arabs in a written version of > > their spoken language, as you suggest, at least through 3rd grade. I can > send > > you some things I've written on this topic, with a lot of cross-linguistic > > data. But thus far, it isn't working. As with most aspects of 'the > situation' > > here, politically active Arabs think that the solution is to pursue the > same > > self-destructive strategy which they've been following for almost a century > but > > with even more vigor and steadfastness (e.g. Masalha points the blame at > the > > Arab media because they do not 'provide the linguistic richness of formal > > Arabic') while Jews are basically content to let the Arabs stew in their > own > > juices so that they can reap the benefits. There may, however, be hope for > the > > future in the form of the radically increased usage of written forms of > > colloquial Arabic dialects in electronic media such as Facebook, blogs, > emails, > > etc., by Arabic speakers below the age of 30, which will soon come to > seriously > > threaten the status of classical Arabic in the same way that the invention > of > > the printing press overturned the linguistic hierarchy in Western Europe by > > overthrowing Latin. I'm working on this too. We'll see what happens. > > > > I realize that Masalha as well as the researchers he referred to don't > > necessarily know what they're talking about regarding the brain. But the > > general point is still potentially significant--that the connected script > which > > Arabic uses as well as the multiple forms which many Arabic letters have > may be > > a significant obstacle to literacy, however this may be related to > processes in > > the brain. > > Best wishes, > > John > > > > > > > > Quoting Tom Givon : > > > > > >> A GREAT POET CAN STILL GET IT WRONG > >> > >> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >> by stumbling a on his truly great quote: > >> "All fixed identities are imposed from the outside. > >> Whoever has a clear identity knows it can assume > >> multiple forms". > >> In the context of Palestine/Israel, what a breath of fresh, rare > >> clarity. Still, like the rest of us mortals, a great poet can on > >> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >> intended--surely got only one third of the story right. At first, he was > >> led astray be the academic researchers he cited, who claimed that the > >> lagging reading skills of Israeli-Arab students is correlated to > >> lagging R-hemisphere activity, then explained this neurological lag > >> by suggesting that the Arab script requires more contextual analysis. > >> But it is the R-hemisphere of the human cortex that is more context > >> oriented, less automated. If Arab-reading students required more > >> contextual labor, it should have been registered as a higher > >> R-hemisphere activity, not lower. > >> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >> among many others, have established beyond reasonable doubt that > >> written words in all languages (English, Mandarin, Hindi, Amharic, > >> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >> module on the boundary of the occipital and temporal lobes (just past > >> Brodman's Area 19), along the ventral visual object-recognition > >> 'stream' that flows from the back to the front of the L-cortex. And > >> the L-cortex is in general responsible for the more automated--less > >> context-dependent--processing of language (as well as visual, motor and > >> other skills). The visual word-recognition module is, in turn, > >> recruited from the pre-existing visual object-recognition > >> ventral-stream module. A considerable amount of life-time practice and > >> repetition is required to affect this late-cultural adaptation. The > >> human brain is not (yet) genetically configured at birth for > >> visual-word recognition, only for visual-object recognition. A similar > >> cultural adaptation, this one for math, has been shown for (Dahaene & > >> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >> in the L-pareita lobe, an area originally configured for analysis of > >> object-location in space. > >> Mr. Masalha then, on his own, points out to a more > >> plausible right answer: Arab students, in Israel as well as all over the > >> Arab world, are not taught literacy in their native language > >> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >> instrument harking back 1,400 years or more. That is, in a foreign > >> language. The discrepancy would be just as great if Israeli kids were > >> taught their Hebrew literacy first in the language of Genesis; or if > >> French students were taught literacy first in the language of La > >> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or > >> English-speaking kids in the language of Beowolf. As far as my frail > >> guessing powers go, remedying the situation would be much easier by > >> combining two well-known verities of second language acquisition: (a) > >> Teach them both early, together--'co-ordinated bilingualism'. And (b), > >> teach literacy first in the student's spoken native language; only then > >> gradually 'stretch' it to more literary genres. This method, bhy the > >> way, was suggested in the late 1930's by no other than L. Bloomfield, > >> in a book outlining a 'phonics-first ' literacy program for native > >> English speakers. Rather than depriving Israeli-Arab students of > >> literacy in their own--equally glorious--native language, just teach > >> them smart. > >> For his last culprit, the presumed--tho hardly > >> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >> recognized prejudices of the educated classes. While readily endorsing > >> his aesthetic sentiments about modern media, I would still like to point > >> out that the 'vulgar' genre is much closer to the Arab students' > >> spoken native language, and if anything should facilitate the easier > >> initial acquisition of native-language literacy. Respectuosamente, > >> ma'-salaam, > >> > >> > >> T. Giv?n > >> > >> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From cbutler at ntlworld.com Thu Sep 30 09:51:34 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Thu, 30 Sep 2010 10:51:34 +0100 Subject: Different authors use different functions differently In-Reply-To: <174B792CEB644021811F83E8342A0A97@ngufa28a6c2639> Message-ID: Dear Yuri, Statistical packages are indeed sometimes quite difficult for linguists and other researchers who are not statisticians. But they are extremely useful - sometimes essential - and it is worth getting to know how to use at least one of them, for anyone whose work is primarily quantitative. One of the most important aspects of using statistics in linguistic research is knowing the basis for the various measures and tests that are available, and what criteria you can use in order to decide which measure or test is appropriate for your data. This is something you need to know, whether you are using a statistical package such as SPSS or doing the calculations manually. However, although it is not too hard to do the calculations manually for simple tests such as chi square or the t-test, it is much harder for more advanced tests, and almost impossible for some techniques such as various kinds of multivariate analysis on large data sets, which require the use of computing power. Manual analysis is error-prone, and this is avoided by using a computer package. On the other hand, the statistical packages often give you a lot of output, some of it very important, other parts of it less so, so that you have to know how to interpret the output. There are some good books around to help researchers to get to grips with the standard packages. For SPSS I would recommend the 3rd edition of Andy Field's 'Discovering Statistics Using SPSS' (Sage Publications, 2009), which is comprehensive and detailed, but well explained and even entertaining. There is also the statistical programming language R, which is being used by an increasing number of adventurous quantitative linguists. For this, there are at least two recent books: Stefan Gries' 'Statistics for Linguists with R: A Practical Introduction' (Routledge [Taylor and Francis Group], 2009), and Harald Baayen's 'Analyzing Linguistic Data. A Practical Introduction to Statistics Using R' (Cambridge University Press 2008), which is rather difficult in parts, but useful nevertheless. R can be downloaded freely from http://www.r-project.org/, and free packages with some of the functionality of the standard ones can be found (e.g. PSPP from http://www.gnu.org/software/pspp/, which I have not tried myself), and may be useful for researchers who do not have access, through their institutions, to the (expensive) standard packages. There are also books which are introductions to statistics for linguists, but are not linked to specific packages. If you are interested in details of these, please get in touch with me directly. Best wishes, Chris Butler -------------------------------------------------- From: "Yuri Tambovtsev" Sent: Wednesday, September 29, 2010 9:10 PM To: Subject: [FUNKNET] Different authors use different functions differently > Dear Funknet colleagues, if corpora of English texts is not one unity but > a conglomerate, how can we use statistic criteria? I started the > discussion what tools to use for the analysis because many of linguists do > not use all these complex statistical packets correctly. The other thing > is who knows what and how the data are being analysed in them. We have a > sort of a "black box" which has the entrance in which you put your data > and the outcome where you receive your results. You must be quite sure > that the data are analysed correctly. The more simple criteria you use, > the better. This is why I stopped using all the stat. packets and began > using very simple criteria like the coefficient of variation, the > Chi-square and the t-test. At least, I am sure about the outcome. More > often than not, it is advisable to try you data manually than to use the > statpack. Looking forward to hearing from you if you agree with me that > using statpacks may give you strange results. The fact that so many > scholar spoke on the list about how to use the statpacks showed that it is > a burning question for linguists and other researchers who are not > specialists in mathematics. Yours sincerely Yuri Tambovtsev, > yutamb at mail.ru From haspelmath at eva.mpg.de Thu Sep 30 11:09:29 2010 From: haspelmath at eva.mpg.de (Martin Haspelmath) Date: Thu, 30 Sep 2010 13:09:29 +0200 Subject: A poet slips In-Reply-To: <1285836023.4ca44cf7aeb68@webmail.haifa.ac.il> Message-ID: Could it be that linguists, who tend to be more interested in structural forms, underestimate the social value of a particular form of a language? I'm a native speaker of a language (Common German of Germany) that didn't have native speakers 150 years ago. It had a lot of social prestige, so the southern dialects have been becoming more and more similar to it, and in northern Germany, Low German (of which my father was still a semi-speaker) has been abandoned in favour of the school language (which happened to be the sacred language of protestantism). Nowadays there are millions of native speakers oft this artificial language, whose front rounded ? and ? umlaut vowels were kept alive by the spelling (the vernaculars lost them centuries ago). From a linguist's point of view, this is a deplorable development, but the speakers sometimes view things differently. Martin john at research.haifa.ac.il wrote: > Learning the classical language is like pulling teeth for Arabs kids also. They > just can't publicly say it because that would make them bad Arabs. It's just > ridiculous. I have an even better plan for the Jewish kids--teach them to use > the written version of the spoken language which Arab kids are using for > Facebook. That's how kids make friends these days anyway. > Best wishes, > John > > > > > > Quoting Tom Givon : > > >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >> Classical (after 1949 things changed...). That book may still exist, >> you might track it down. It was easy, a cinch really. I concluded it was >> really just Hebrew with a few trivial transformations in Phonology & >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >> on a few sessions of a faculty study group at UO who were trying to >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >> it was like pullin' teeth. But Leonard Bloomfield said it already, and >> well, in 1939 (or was it 1943?). >> Cheers, TG >> >> ============== >> >> >> john at research.haifa.ac.il wrote: >> >>> Tom, >>> I've been trying for several years to get Israeli Arabs and Jews to >>> >> seriously >> >>> consider the possibility of educating Israeli Arabs in a written version of >>> their spoken language, as you suggest, at least through 3rd grade. I can >>> >> send >> >>> you some things I've written on this topic, with a lot of cross-linguistic >>> data. But thus far, it isn't working. As with most aspects of 'the >>> >> situation' >> >>> here, politically active Arabs think that the solution is to pursue the >>> >> same >> >>> self-destructive strategy which they've been following for almost a century >>> >> but >> >>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>> >> the >> >>> Arab media because they do not 'provide the linguistic richness of formal >>> Arabic') while Jews are basically content to let the Arabs stew in their >>> >> own >> >>> juices so that they can reap the benefits. There may, however, be hope for >>> >> the >> >>> future in the form of the radically increased usage of written forms of >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>> >> emails, >> >>> etc., by Arabic speakers below the age of 30, which will soon come to >>> >> seriously >> >>> threaten the status of classical Arabic in the same way that the invention >>> >> of >> >>> the printing press overturned the linguistic hierarchy in Western Europe by >>> overthrowing Latin. I'm working on this too. We'll see what happens. >>> >>> I realize that Masalha as well as the researchers he referred to don't >>> necessarily know what they're talking about regarding the brain. But the >>> general point is still potentially significant--that the connected script >>> >> which >> >>> Arabic uses as well as the multiple forms which many Arabic letters have >>> >> may be >> >>> a significant obstacle to literacy, however this may be related to >>> >> processes in >> >>> the brain. >>> Best wishes, >>> John >>> >>> >>> >>> Quoting Tom Givon : >>> >>> >>> >>>> A GREAT POET CAN STILL GET IT WRONG >>>> >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>> by stumbling a on his truly great quote: >>>> "All fixed identities are imposed from the outside. >>>> Whoever has a clear identity knows it can assume >>>> multiple forms". >>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>> clarity. Still, like the rest of us mortals, a great poet can on >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>> intended--surely got only one third of the story right. At first, he was >>>> led astray be the academic researchers he cited, who claimed that the >>>> lagging reading skills of Israeli-Arab students is correlated to >>>> lagging R-hemisphere activity, then explained this neurological lag >>>> by suggesting that the Arab script requires more contextual analysis. >>>> But it is the R-hemisphere of the human cortex that is more context >>>> oriented, less automated. If Arab-reading students required more >>>> contextual labor, it should have been registered as a higher >>>> R-hemisphere activity, not lower. >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>> among many others, have established beyond reasonable doubt that >>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>> module on the boundary of the occipital and temporal lobes (just past >>>> Brodman's Area 19), along the ventral visual object-recognition >>>> 'stream' that flows from the back to the front of the L-cortex. And >>>> the L-cortex is in general responsible for the more automated--less >>>> context-dependent--processing of language (as well as visual, motor and >>>> other skills). The visual word-recognition module is, in turn, >>>> recruited from the pre-existing visual object-recognition >>>> ventral-stream module. A considerable amount of life-time practice and >>>> repetition is required to affect this late-cultural adaptation. The >>>> human brain is not (yet) genetically configured at birth for >>>> visual-word recognition, only for visual-object recognition. A similar >>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>> in the L-pareita lobe, an area originally configured for analysis of >>>> object-location in space. >>>> Mr. Masalha then, on his own, points out to a more >>>> plausible right answer: Arab students, in Israel as well as all over the >>>> Arab world, are not taught literacy in their native language >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>> instrument harking back 1,400 years or more. That is, in a foreign >>>> language. The discrepancy would be just as great if Israeli kids were >>>> taught their Hebrew literacy first in the language of Genesis; or if >>>> French students were taught literacy first in the language of La >>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>> English-speaking kids in the language of Beowolf. As far as my frail >>>> guessing powers go, remedying the situation would be much easier by >>>> combining two well-known verities of second language acquisition: (a) >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>> teach literacy first in the student's spoken native language; only then >>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>> in a book outlining a 'phonics-first ' literacy program for native >>>> English speakers. Rather than depriving Israeli-Arab students of >>>> literacy in their own--equally glorious--native language, just teach >>>> them smart. >>>> For his last culprit, the presumed--tho hardly >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>> recognized prejudices of the educated classes. While readily endorsing >>>> his aesthetic sentiments about modern media, I would still like to point >>>> out that the 'vulgar' genre is much closer to the Arab students' >>>> spoken native language, and if anything should facilitate the easier >>>> initial acquisition of native-language literacy. Respectuosamente, >>>> ma'-salaam, >>>> >>>> >>>> T. Giv?n >>>> >>>> >>>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > > -- Martin Haspelmath (haspelmath at eva.mpg.de) Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 From cbutler at ntlworld.com Thu Sep 30 12:06:15 2010 From: cbutler at ntlworld.com (Chris Butler) Date: Thu, 30 Sep 2010 13:06:15 +0100 Subject: Books on statistics for linguistics Message-ID: Shanley Allen has asked me to post to the list some references to general books on statistics for linguistics. Here is an annotated selection, which I hope will be useful. Other list members will no doubt be able to suggest other books which they have found helpful. Chris Butler ****************************** Butler, C. S. (1985) Statistics in Linguistics. Oxford: Blackwell. [Covers basic techniques only; explains the maths on the way; many examples. Last chapter hopelessly out of date 25 years on! Now out of print, but available via the internet at http://www.uwe.ac.uk/hlss/llas/statistics-in-linguistics/bkindex.shtml] Hatch, E. & A. Lazaraton (1991) The Research Manual: Design and Statistics for Applied Linguistics. Boston: Heinle & Heinle. [Fairly high level treatment, with many useful examples.] Johnson, K. (2008) Quantitative Methods in Linguistics. Oxford, Malden MA (USA) and Carlton, Victoria (Australia): Blackwell Publishing. [A fairly advanced treatment of a wide range of statistical techniques for the study of language, with statistical routines in the programming environment 'R'. Deals with the basics rather quickly, but also covers regression methods in detail. Oakes, M. P. (1998) Statistics for Corpus Linguistics. Edinburgh Textbooks in Linguistics. Edinburgh: Edinburgh University Press. [One chapter covers basic techniques rather briefly. Good discussion of multivariate methods. Also deals with other statistically-related procedures relevant to corpus linguistics, such as word tagging.] Rasinger, S. M. (2008) Quantitative Research in Linguistics: An Introduction. London: Continuum. [An introduction to quantitative data and its analysis, including basic statistical techniques.] Woods, A., P. Fletcher & A. Hughes (1986) Statistics in Language Studies. Cambridge: Cambridge University Press. [A more comprehensive treatment than Butler 1985, and at a considerably higher level, so correspondingly more difficult to read. Covers multivariate techniques, regression, etc in some detail. Very good examples.] From amnfn at well.com Thu Sep 30 12:42:11 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:42:11 -0700 Subject: A poet slips In-Reply-To: <4CA4282D.6080303@uoregon.edu> Message-ID: Tom, I feel the need to point out a few things: 1) There are no Arab students in the Israeli primary schools. There are those whose native language is Arabic, but they are not ethnically Arabs. Israeli citizens who do not identify as Israeli used to think of themselves as fellahin and now often identify as Palestinian -- precisely because they were not allowed to go to an integrated school where the only language spoken is Hebrew. They are ethnically Judeans who never went away. 2) The language of Genesis is the language that was revived by Zionists in the 19th century. I learned to read Hebrew in the language of Genesis. (And yes, I was born in Israel.) My father learned to read Hebrew in the language of Genesis. He was born in Poland. It was his first language, and the first language he was literate in. His parents were Zionists who learned to read Hebrew in the language of the OT. If it were not for people who learned to read Hebrew in the language of Genesis, there would be no native Hebrew speakers in the world today. Right cortex or left is a personal matter, depending on how you are wired. Left handers and right handers do it differently. But native speaker or oursider has everything to do with how you are treated and which school you go to. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, Tom Givon wrote: > > > A GREAT POET CAN STILL GET IT WRONG > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > Haaretz/English, International Herald Tribune 9-27-10) accidentally by > stumbling a on his truly great quote: > "All fixed identities are imposed from the outside. > Whoever has a clear identity knows it can assume multiple > forms". > In the context of Palestine/Israel, what a breath of fresh, rare clarity. > Still, like the rest of us mortals, a great poet can on occasion get it > wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got > only one third of the story right. At first, he was led astray be the > academic researchers he cited, who claimed that the lagging reading skills of > Israeli-Arab students is correlated to lagging R-hemisphere activity, > then explained this neurological lag by suggesting that the Arab script > requires more contextual analysis. But it is the R-hemisphere of the human > cortex that is more context oriented, less automated. If Arab-reading > students required more contextual labor, it should have been registered as a > higher R-hemisphere activity, not lower. > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among > many others, have established beyond reasonable doubt that written words > in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) > are decoded automatically in an L-hemisphere module on the boundary of the > occipital and temporal lobes (just past Brodman's Area 19), along the ventral > visual object-recognition 'stream' that flows from the back to the front of > the L-cortex. And the L-cortex is in general responsible for the more > automated--less context-dependent--processing of language (as well as > visual, motor and other skills). The visual word-recognition module is, in > turn, recruited from the pre-existing visual object-recognition > ventral-stream module. A considerable amount of life-time practice and > repetition is required to affect this late-cultural adaptation. The human > brain is not (yet) genetically configured at birth for visual-word > recognition, only for visual-object recognition. A similar cultural > adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see > recent article in The New Yorker by Oliver Sachs) > in the L-pareita lobe, an area originally configured for analysis of > object-location in space. > Mr. Masalha then, on his own, points out to a more plausible > right answer: Arab students, in Israel as well as all over the Arab world, > are not taught literacy in their native language (Falastini, Maghrebi, > Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 > years or more. That is, in a foreign language. The discrepancy would be just > as great if Israeli kids were taught their Hebrew literacy first in the > language of Genesis; or if French students were taught literacy first in > the language of La Chan?on de Roland, Guilhome de Machaut, or Chr? tien de > Troyes. Or English-speaking kids in the language of Beowolf. As far as my > frail guessing powers go, remedying the situation would be much easier by > combining two well-known verities of second language acquisition: (a) Teach > them both early, together--'co-ordinated bilingualism'. And (b), teach > literacy first in the student's spoken native language; only then gradually > 'stretch' it to more literary genres. This method, bhy the way, was > suggested in the late 1930's by no other than L. Bloomfield, in a book > outlining a 'phonics-first ' literacy program for native English speakers. > Rather than depriving Israeli-Arab students of literacy in their own--equally > glorious--native language, just teach them smart. > For his last culprit, the presumed--tho hardly > unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized > prejudices of the educated classes. While readily endorsing his aesthetic > sentiments about modern media, I would still like to point out that the > 'vulgar' genre is much closer to the Arab students' spoken native language, > and if anything should facilitate the easier initial acquisition of > native-language literacy. Respectuosamente, ma'-salaam, > > T. > Giv?n > > From amnfn at well.com Thu Sep 30 12:51:34 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:51:34 -0700 Subject: A poet slips In-Reply-To: <1285836023.4ca44cf7aeb68@webmail.haifa.ac.il> Message-ID: John, These people you speak of are not Arabs. Some of them are Moslem and they read the Quran in the original. Some of them are not Moslem. All of them speak a local dialect of Arabic. Ask them sometimes if they think they are Arabs. Trying to turn every dialect into a separate language with a separate writing system is a way to try to disunite people. But a common language, however differently it is pronounced, unites disparate people. Australians and Cockneys and Indians and Americans speak sometimes mutually unintelligible versions of English. Using the same writing system and the same classic texts unites them. Instead of telling people they should magnify every difference, why not offer to share your language with them? Hebrew could be a uniting factor if spoken in all Israeli schools. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Learning the classical language is like pulling teeth for Arabs kids also. They > just can't publicly say it because that would make them bad Arabs. It's just > ridiculous. I have an even better plan for the Jewish kids--teach them to use > the written version of the spoken language which Arab kids are using for > Facebook. That's how kids make friends these days anyway. > Best wishes, > John > > > > > > Quoting Tom Givon : > >> >> >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >> Classical (after 1949 things changed...). That book may still exist, >> you might track it down. It was easy, a cinch really. I concluded it was >> really just Hebrew with a few trivial transformations in Phonology & >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >> on a few sessions of a faculty study group at UO who were trying to >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >> it was like pullin' teeth. But Leonard Bloomfield said it already, and >> well, in 1939 (or was it 1943?). >> Cheers, TG >> >> ============== >> >> >> john at research.haifa.ac.il wrote: >>> Tom, >>> I've been trying for several years to get Israeli Arabs and Jews to >> seriously >>> consider the possibility of educating Israeli Arabs in a written version of >>> their spoken language, as you suggest, at least through 3rd grade. I can >> send >>> you some things I've written on this topic, with a lot of cross-linguistic >>> data. But thus far, it isn't working. As with most aspects of 'the >> situation' >>> here, politically active Arabs think that the solution is to pursue the >> same >>> self-destructive strategy which they've been following for almost a century >> but >>> with even more vigor and steadfastness (e.g. Masalha points the blame at >> the >>> Arab media because they do not 'provide the linguistic richness of formal >>> Arabic') while Jews are basically content to let the Arabs stew in their >> own >>> juices so that they can reap the benefits. There may, however, be hope for >> the >>> future in the form of the radically increased usage of written forms of >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >> emails, >>> etc., by Arabic speakers below the age of 30, which will soon come to >> seriously >>> threaten the status of classical Arabic in the same way that the invention >> of >>> the printing press overturned the linguistic hierarchy in Western Europe by >>> overthrowing Latin. I'm working on this too. We'll see what happens. >>> >>> I realize that Masalha as well as the researchers he referred to don't >>> necessarily know what they're talking about regarding the brain. But the >>> general point is still potentially significant--that the connected script >> which >>> Arabic uses as well as the multiple forms which many Arabic letters have >> may be >>> a significant obstacle to literacy, however this may be related to >> processes in >>> the brain. >>> Best wishes, >>> John >>> >>> >>> >>> Quoting Tom Givon : >>> >>> >>>> A GREAT POET CAN STILL GET IT WRONG >>>> >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>> by stumbling a on his truly great quote: >>>> "All fixed identities are imposed from the outside. >>>> Whoever has a clear identity knows it can assume >>>> multiple forms". >>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>> clarity. Still, like the rest of us mortals, a great poet can on >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>> intended--surely got only one third of the story right. At first, he was >>>> led astray be the academic researchers he cited, who claimed that the >>>> lagging reading skills of Israeli-Arab students is correlated to >>>> lagging R-hemisphere activity, then explained this neurological lag >>>> by suggesting that the Arab script requires more contextual analysis. >>>> But it is the R-hemisphere of the human cortex that is more context >>>> oriented, less automated. If Arab-reading students required more >>>> contextual labor, it should have been registered as a higher >>>> R-hemisphere activity, not lower. >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>> among many others, have established beyond reasonable doubt that >>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>> module on the boundary of the occipital and temporal lobes (just past >>>> Brodman's Area 19), along the ventral visual object-recognition >>>> 'stream' that flows from the back to the front of the L-cortex. And >>>> the L-cortex is in general responsible for the more automated--less >>>> context-dependent--processing of language (as well as visual, motor and >>>> other skills). The visual word-recognition module is, in turn, >>>> recruited from the pre-existing visual object-recognition >>>> ventral-stream module. A considerable amount of life-time practice and >>>> repetition is required to affect this late-cultural adaptation. The >>>> human brain is not (yet) genetically configured at birth for >>>> visual-word recognition, only for visual-object recognition. A similar >>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>> in the L-pareita lobe, an area originally configured for analysis of >>>> object-location in space. >>>> Mr. Masalha then, on his own, points out to a more >>>> plausible right answer: Arab students, in Israel as well as all over the >>>> Arab world, are not taught literacy in their native language >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>> instrument harking back 1,400 years or more. That is, in a foreign >>>> language. The discrepancy would be just as great if Israeli kids were >>>> taught their Hebrew literacy first in the language of Genesis; or if >>>> French students were taught literacy first in the language of La >>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>> English-speaking kids in the language of Beowolf. As far as my frail >>>> guessing powers go, remedying the situation would be much easier by >>>> combining two well-known verities of second language acquisition: (a) >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>> teach literacy first in the student's spoken native language; only then >>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>> in a book outlining a 'phonics-first ' literacy program for native >>>> English speakers. Rather than depriving Israeli-Arab students of >>>> literacy in their own--equally glorious--native language, just teach >>>> them smart. >>>> For his last culprit, the presumed--tho hardly >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>> recognized prejudices of the educated classes. While readily endorsing >>>> his aesthetic sentiments about modern media, I would still like to point >>>> out that the 'vulgar' genre is much closer to the Arab students' >>>> spoken native language, and if anything should facilitate the easier >>>> initial acquisition of native-language literacy. Respectuosamente, >>>> ma'-salaam, >>>> >>>> >>>> T. Giv?n >>>> >>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 12:55:16 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 05:55:16 -0700 Subject: A poet slips In-Reply-To: <4CA46FE9.2000008@eva.mpg.de> Message-ID: Martin, Language is a tool. People want to communicate and fit in. Ultimately, that's more important than the ornamental differences between one tool and another. As functionalists, we should be cognizant of that. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, Martin Haspelmath wrote: > Could it be that linguists, who tend to be more interested in structural > forms, underestimate the social value of a particular form of a language? > > I'm a native speaker of a language (Common German of Germany) that didn't > have native speakers 150 years ago. It had a lot of social prestige, so the > southern dialects have been becoming more and more similar to it, and in > northern Germany, Low German (of which my father was still a semi-speaker) > has been abandoned in favour of the school language (which happened to be the > sacred language of protestantism). Nowadays there are millions of native > speakers oft this artificial language, whose front rounded ? and ? umlaut > vowels were kept alive by the spelling (the vernaculars lost them centuries > ago). > > From a linguist's point of view, this is a deplorable development, but the > speakers sometimes view things differently. > > Martin > > john at research.haifa.ac.il wrote: >> Learning the classical language is like pulling teeth for Arabs kids also. >> They >> just can't publicly say it because that would make them bad Arabs. It's >> just >> ridiculous. I have an even better plan for the Jewish kids--teach them to >> use >> the written version of the spoken language which Arab kids are using for >> Facebook. That's how kids make friends these days anyway. >> Best wishes, >> John >> >> >> >> >> >> Quoting Tom Givon : >> >> >>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>> Classical (after 1949 things changed...). That book may still exist, >>> you might track it down. It was easy, a cinch really. I concluded it was >>> really just Hebrew with a few trivial transformations in Phonology & >>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>> on a few sessions of a faculty study group at UO who were trying to >>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>> well, in 1939 (or was it 1943?). >>> Cheers, TG >>> >>> ============== >>> >>> >>> john at research.haifa.ac.il wrote: >>> >>>> Tom, >>>> I've been trying for several years to get Israeli Arabs and Jews to >>>> >>> seriously >>> >>>> consider the possibility of educating Israeli Arabs in a written version >>>> of >>>> their spoken language, as you suggest, at least through 3rd grade. I can >>>> >>> send >>> >>>> you some things I've written on this topic, with a lot of >>>> cross-linguistic >>>> data. But thus far, it isn't working. As with most aspects of 'the >>>> >>> situation' >>> >>>> here, politically active Arabs think that the solution is to pursue the >>>> >>> same >>> >>>> self-destructive strategy which they've been following for almost a >>>> century >>>> >>> but >>> >>>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>>> >>> the >>> >>>> Arab media because they do not 'provide the linguistic richness of formal >>>> Arabic') while Jews are basically content to let the Arabs stew in their >>>> >>> own >>> >>>> juices so that they can reap the benefits. There may, however, be hope >>>> for >>>> >>> the >>> >>>> future in the form of the radically increased usage of written forms of >>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>> >>> emails, >>> >>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>> >>> seriously >>> >>>> threaten the status of classical Arabic in the same way that the >>>> invention >>>> >>> of >>> >>>> the printing press overturned the linguistic hierarchy in Western Europe >>>> by >>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>> >>>> I realize that Masalha as well as the researchers he referred to don't >>>> necessarily know what they're talking about regarding the brain. But the >>>> general point is still potentially significant--that the connected script >>>> >>> which >>> >>>> Arabic uses as well as the multiple forms which many Arabic letters have >>>> >>> may be >>> >>>> a significant obstacle to literacy, however this may be related to >>>> >>> processes in >>> >>>> the brain. >>>> Best wishes, >>>> John >>>> >>>> >>>> >>>> Quoting Tom Givon : >>>> >>>> >>>> >>>>> A GREAT POET CAN STILL GET IT >>>>> WRONG >>>>> >>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>> by stumbling a on his truly great quote: >>>>> "All fixed identities are imposed from the outside. >>>>> Whoever has a clear identity knows it can assume >>>>> multiple forms". >>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>> intended--surely got only one third of the story right. At first, he was >>>>> led astray be the academic researchers he cited, who claimed that the >>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>> by suggesting that the Arab script requires more contextual analysis. >>>>> But it is the R-hemisphere of the human cortex that is more context >>>>> oriented, less automated. If Arab-reading students required more >>>>> contextual labor, it should have been registered as a higher >>>>> R-hemisphere activity, not lower. >>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>> among many others, have established beyond reasonable doubt that >>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>> module on the boundary of the occipital and temporal lobes (just past >>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>> the L-cortex is in general responsible for the more automated--less >>>>> context-dependent--processing of language (as well as visual, motor and >>>>> other skills). The visual word-recognition module is, in turn, >>>>> recruited from the pre-existing visual object-recognition >>>>> ventral-stream module. A considerable amount of life-time practice and >>>>> repetition is required to affect this late-cultural adaptation. The >>>>> human brain is not (yet) genetically configured at birth for >>>>> visual-word recognition, only for visual-object recognition. A similar >>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>> object-location in space. >>>>> Mr. Masalha then, on his own, points out to a more >>>>> plausible right answer: Arab students, in Israel as well as all over the >>>>> Arab world, are not taught literacy in their native language >>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>> language. The discrepancy would be just as great if Israeli kids were >>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>> French students were taught literacy first in the language of La >>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>> guessing powers go, remedying the situation would be much easier by >>>>> combining two well-known verities of second language acquisition: (a) >>>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>>> teach literacy first in the student's spoken native language; only then >>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>> literacy in their own--equally glorious--native language, just teach >>>>> them smart. >>>>> For his last culprit, the presumed--tho hardly >>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>> recognized prejudices of the educated classes. While readily endorsing >>>>> his aesthetic sentiments about modern media, I would still like to point >>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>> spoken native language, and if anything should facilitate the easier >>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>> ma'-salaam, >>>>> >>>>> >>>>> T. Giv?n >>>>> >>>>> >>>>> >>>> >>>> >>>> ------------------------------------------------------------------------ >>>> This message was sent using IMP, the Webmail Program of Haifa University >>>> >>>> >>> >> >> >> >> >> ------------------------------------------------------------------------ >> This message was sent using IMP, the Webmail Program of Haifa University >> >> >> > > > -- > Martin Haspelmath (haspelmath at eva.mpg.de) > Max-Planck-Institut fuer evolutionaere Anthropologie, Deutscher Platz 6 > D-04103 Leipzig Tel. (MPI) +49-341-3550 307, (priv.) +49-341-980 1616 > > > > > > From mariel at post.tau.ac.il Thu Sep 30 14:36:24 2010 From: mariel at post.tau.ac.il (Mira Ariel) Date: Thu, 30 Sep 2010 06:36:24 -0800 Subject: Learning to read Hebrew Message-ID: Hi, I'm surprised to hear that Aya Katz learnt to read Hebrew using Biblical texts. Growing up in Israel even before Aya, we leant to read and write from Modern Hebrew texts, albeit, somewhat formal, written Hebrew. We never started to read the Bible till we knew how to read (2nd grade), and at the beginning, the Biblical text was a simplified one ("Bible for children"). So, no doubt one CAN learn Hebrew from Biblical texts, but we don't impose it on all kids in Israel. Best, Mira _____________ Mira Ariel http://www.tau.ac.il/~mariel From amnfn at well.com Thu Sep 30 14:27:30 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 07:27:30 -0700 Subject: Learning to read Hebrew In-Reply-To: <001001cb60ac$dd8ae5a0$98a0b0e0$@tau.ac.il> Message-ID: Mira, My story is in fact a little unusual. While I grew up in Israel, I went to first and second grade in the United States. I had picked up a little reading when I was five, but as you can imagine, I was rusty by the time we returned to Israel and I was expected to go into third grade. My grandfather, who was then the Rector of the University of Tel-Aviv, brought me a Bible and made me start reading Genesis out loud to him. He made me do this until he was satisfied that my reading was good enough for third grade. In third grade, we read the old testament in its unexpurgated and uncut form. I believe we started on Judges, because I had missed out on some of the earlier books. I would like to emphasize that none of my classmates needed a translation from Biblical to "modern" Hebrew. I know that in academic circles much is made of the difference, but believe me Biblical texts are intelligible to modern day readers. The difference in more like that between Shakespeare and today's English. It is not like the difference between Chaucer and modern English. There's a lot of politics that goes on in academic circles to make mountains out of molehills where minute differences between different versions of the same language are concerned. However, if one insists on modern day school children unable to understand the old language, not making them read it in school is the best first step. We know what we use. --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, Mira Ariel wrote: > Hi, > > > > I'm surprised to hear that Aya Katz learnt to read Hebrew using Biblical > texts. Growing up in Israel even before Aya, we leant to read and write from > Modern Hebrew texts, albeit, somewhat formal, written Hebrew. We never > started to read the Bible till we knew how to read (2nd grade), and at the > beginning, the Biblical text was a simplified one ("Bible for children"). > So, no doubt one CAN learn Hebrew from Biblical texts, but we don't impose > it on all kids in Israel. > > > > Best, > > > > Mira > > > > _____________ > > Mira Ariel > > http://www.tau.ac.il/~mariel > > > > From kuzar at research.haifa.ac.il Thu Sep 30 14:25:21 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Thu, 30 Sep 2010 16:25:21 +0200 Subject: A poet slips In-Reply-To: Message-ID: I see that all (ex)Israelis on this list are throwing in their 2Cs. Here are mine. One of the slippery things about identity is that it is dynamic, multiple, and defies one-word definitions. Israeli Arabs/Palestinians are a sector of Israel's citizenry. I even hesitate to say that they are a community, but even if they are, they are as diversified among themselves, as any other citizenry nowadays. Just because they happen to be Arabs (people of the Orient) doesn't mean you can easily draw generalizing cultural conclusions any more than you could do this about SAE nations. I am saying all this, because I hear an Orientalistic tone in both John's and Aya's postings, who just simply know and are so quick to define Arab/Palestinian/Arabic speaking citizens of Israel as either this or that. Some Israeli Arabs put more emphasis on their general ethnic common origin, others more on their local Palestinian side of their identity, others yet are first and foremost Muslims, or Christians, some slip in and out of this or that identity every five years, and some couldn't care less and would in fact become Hebrew speaking Israeli's, if the Israeli system had been more forthcoming. Masalha himself is from Druze origin, a minority that has been persecuted by Muslim majority for centuries. This resulted in massive support of Zionism by the Druze and they go to the army and fight the Arabs without hesitation. Except that this is also a massive generalization. Israel's Jewish ethnocentric laws and practices don't even fully embrace the Druze population with great love, a fact which brought some of them to identify with and to view themselves as Palestinian and to struggle for the common Palestinian cause. Add to this the fact that any two communities living together produce hybridity, and you will find all sorts of creative (and wonderful) hybrid people who combine Jewish and Arab identities in their personality. Masalha himself is from Druze origin, anti Zionist, but he is also a poet, writing in both Hebrew and Arabic, and also a translator both ways. This is only a shallow introduction to the complexities of identity for those of you who are not so well informed in this discussion. The question of language is a very complex one, precisely because the educational system has to devise a model that would be equally beneficial for a widely diversified community. Assuming that there is an Arab speaking population in Israel (which we don't have to define culturally for the sake of this discussion) and assuming that many of them support the idea that Arabic should be maintained and should be the first schooling language for these people, we may indeed observe that Arab children are faced with two languages, their native "dialect" and the language of broad cultural exchange in the Arab world today. If one day the native tongues are going to become official languages of Egypt, Syria, Jordan, Morocco, Yemen, etc. I am sure that the Palestinians will follow suit, and the Israeli Arabs will be a part of this process. It is not reasonable to expect Israeli Arabs to be the pioneers here. Now, Modern Standard Arabic is not Classical (Quran) Arabic, it is a modern language that has been revived parallel to the revival of Hebrew, except that the revival of Hebrew included the spoken language and the revival of Arabic was only in the written language. This is however a full-fledged communication system, in which anything can be written, from engineering to modern philosophy. Another complicating factor is the fact that MSA and local dialects are not a dichotomy but rather a scale. The is the more formal form, which is a simplified version of Classical Arabic, but then there is also a simplified version of this language that is used without the case endings, and sometimes in the SVO order of the spoken language, and so on. While it is true that the Internet and Facebook have brought about new ways of communicating in some koine forms of the spoken dialects, it is not true that the whole population shares this mode of communication (only the more privileged ones) and you certainly cannot write books on philosophy and engineering in these language forms. So let's be a little more careful about giving sweeping simple answers to the complex situation of learner of Arabic as a first language. In fact, the Arabs are not the first ones to have to face such a complex linguistic situation. I would suggest that this issue be approached with more cultural sensitivity and attempts to to operate within the given complex situation. Believe me, with a good schooling system that is well funded and has a good educational leadership (which is not the case in Israel at large, and is much worse in the Arab sector), the left and right hemispheres of the brain will adjust just fine, and will learn what they need to do to deal with all the complexities, despite the "difficult" cursive alphabet, despite the right-to-left writing direction, and despite all other "objective" difficulties. Ron Kuzar ---------------- On Thu, 30 Sep 2010 05:42:11 -0700 (PDT) "A. Katz" wrote: > Tom, > > I feel the need to point out a few things: > > 1) There are no Arab students in the Israeli primary schools. There are > those whose native language is Arabic, but they are not ethnically Arabs. > Israeli citizens who do not identify as Israeli used to think of themselves as > fellahin and now often identify as Palestinian -- precisely because they > were not allowed to go to an integrated school where the only language > spoken is Hebrew. They are ethnically Judeans who never went away. > > 2) The language of Genesis is the language that was revived by Zionists in > the 19th century. I learned to read Hebrew in the language of Genesis. > (And yes, I was born in Israel.) My father learned to read Hebrew in the > language of Genesis. He was born in Poland. It was his first language, > and the first language he was literate in. His parents were Zionists > who learned to read Hebrew in the language of the OT. If it were not for > people who learned to read Hebrew in the language of Genesis, there would be no > native Hebrew speakers in the world today. > > Right cortex or left is a personal matter, depending on how you are wired. > Left handers and right handers do it differently. But native speaker or > oursider has everything to do with how you are treated and which school > you go to. > > > --Aya > > http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > On Thu, 30 Sep 2010, Tom Givon wrote: > > > > > > > A GREAT POET CAN STILL GET IT WRONG > > > > I came to know of Salman Masalha ("Arabs, speak Hebrew!", > > Haaretz/English, International Herald Tribune 9-27-10) accidentally by > > stumbling a on his truly great quote: > > "All fixed identities are imposed from the outside. > > Whoever has a clear identity knows it can assume multiple > > forms". > > In the context of Palestine/Israel, what a breath of fresh, rare clarity. > > Still, like the rest of us mortals, a great poet can on occasion get it > > wrong too, and Mr. Masalha--may he be forgiven; pun intended--surely got > > only one third of the story right. At first, he was led astray be the > > academic researchers he cited, who claimed that the lagging reading skills of > > Israeli-Arab students is correlated to lagging R-hemisphere activity, > > then explained this neurological lag by suggesting that the Arab script > > requires more contextual analysis. But it is the R-hemisphere of the human > > cortex that is more context oriented, less automated. If Arab-reading > > students required more contextual labor, it should have been registered as a > > higher R-hemisphere activity, not lower. > > Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, among > > many others, have established beyond reasonable doubt that written words > > in all languages (English, Mandarin, Hindi, Amharic, Hebrew, Arabic, etc.) > > are decoded automatically in an L-hemisphere module on the boundary of the > > occipital and temporal lobes (just past Brodman's Area 19), along the ventral > > visual object-recognition 'stream' that flows from the back to the front of > > the L-cortex. And the L-cortex is in general responsible for the more > > automated--less context-dependent--processing of language (as well as > > visual, motor and other skills). The visual word-recognition module is, in > > turn, recruited from the pre-existing visual object-recognition > > ventral-stream module. A considerable amount of life-time practice and > > repetition is required to affect this late-cultural adaptation. The human > > brain is not (yet) genetically configured at birth for visual-word > > recognition, only for visual-object recognition. A similar cultural > > adaptation, this one for math, has been shown for (Dahaene & Cohen, 2007; see > > recent article in The New Yorker by Oliver Sachs) > > in the L-pareita lobe, an area originally configured for analysis of > > object-location in space. > > Mr. Masalha then, on his own, points out to a more plausible > > right answer: Arab students, in Israel as well as all over the Arab world, > > are not taught literacy in their native language (Falastini, Maghrebi, > > Masri, Yemeni, etc.), but in a frozen literary instrument harking back 1,400 > > years or more. That is, in a foreign language. The discrepancy would be just > > as great if Israeli kids were taught their Hebrew literacy first in the > > language of Genesis; or if French students were taught literacy first in > > the language of La Chan?on de Roland, Guilhome de Machaut, or Chr? tien de > > Troyes. Or English-speaking kids in the language of Beowolf. As far as my > > frail guessing powers go, remedying the situation would be much easier by > > combining two well-known verities of second language acquisition: (a) Teach > > them both early, together--'co-ordinated bilingualism'. And (b), teach > > literacy first in the student's spoken native language; only then gradually > > 'stretch' it to more literary genres. This method, bhy the way, was > > suggested in the late 1930's by no other than L. Bloomfield, in a book > > outlining a 'phonics-first ' literacy program for native English speakers. > > Rather than depriving Israeli-Arab students of literacy in their own--equally > > glorious--native language, just teach them smart. > > For his last culprit, the presumed--tho hardly > > unique--vulgarity of Arab media, Mr. Masalha lapses into well recognized > > prejudices of the educated classes. While readily endorsing his aesthetic > > sentiments about modern media, I would still like to point out that the > > 'vulgar' genre is much closer to the Arab students' spoken native language, > > and if anything should facilitate the easier initial acquisition of > > native-language literacy. Respectuosamente, ma'-salaam, > > > > T. > > Giv?n > > > > =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From john at research.haifa.ac.il Thu Sep 30 14:43:30 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 16:43:30 +0200 Subject: A poet slips In-Reply-To: Message-ID: Aya, I ask them constantly if they think they are Arabs (60% of my students are Arabic speakers and this is a central topic of many classes that I teach) and almost without exception they do. The only exceptions are some Druze (mostly males) and most Maronites. I am aware that 100 years ago Arabic speakers living in this area did not consider themselves to be Arabs (except for the Bedouins), but the situation has completely changed. It is totally unrealistic to put Israeli Arabs as a group into Hebrew-speaking schools. No one wants it, not the Jews and not the Arabs. There are a tiny number of Arabs who for one reason or another send their children to Hebrew-speaking schools (for example the author Sayed Kashua) but this is insignificant. I don't know what the situation was when you were a child, and I don't know about the situation among religious Jews, but secular Jewish children do not learn to read by reading Genesis today. They learn to read with secular texts in 1st grade and go on to the Bible in 2nd grade, but far from learning to read by reading the Bible, teachers have to explain what's written in the Bible to the students. NO ONE can understand it without an explanation from the teacher or their parents. Ron--I agree with everything you say except that I believe that the radical difference between spoken and written Arabic definitely is a serious problem for literacy. I would not be so quick to dismiss the effect of the Arabic writing system--we had a conference on this topic in Haifa in May and the general consensus of people who sounded like they knew what they were talking about was that it is a significant problem-- but they hadn't done cross-linguistic research and this isn't something I know enough about to have strong opinions one way or the other. Also, the usage of written colloquial Arabic is basically universal among all Israeli Arabs under the age of 25 who have a cellular phone. This is definitely not just a privileged section of the population, it's most people of the relevant age group. John Quoting "A. Katz" : > John, > > These people you speak of are not Arabs. Some of them are Moslem and they > read the Quran in the original. Some of them are not Moslem. All of them > speak a local dialect of Arabic. Ask them sometimes if they think they are > Arabs. > > Trying to turn every dialect into a separate language with a separate > writing system is a way to try to disunite people. But a common language, > however differently it is pronounced, unites disparate people. Australians > and Cockneys and Indians and Americans speak sometimes mutually > unintelligible versions of English. Using the same writing system and > the same classic texts unites them. > > Instead of telling people they should magnify every difference, why not > offer to share your language with them? Hebrew could be a uniting factor > if spoken in all Israeli schools. > > --Aya > > http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Learning the classical language is like pulling teeth for Arabs kids also. > They > > just can't publicly say it because that would make them bad Arabs. It's > just > > ridiculous. I have an even better plan for the Jewish kids--teach them to > use > > the written version of the spoken language which Arab kids are using for > > Facebook. That's how kids make friends these days anyway. > > Best wishes, > > John > > > > > > > > > > > > Quoting Tom Givon : > > > >> > >> > >> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > >> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >> Classical (after 1949 things changed...). That book may still exist, > >> you might track it down. It was easy, a cinch really. I concluded it was > >> really just Hebrew with a few trivial transformations in Phonology & > >> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > >> on a few sessions of a faculty study group at UO who were trying to > >> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > >> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >> well, in 1939 (or was it 1943?). > >> Cheers, TG > >> > >> ============== > >> > >> > >> john at research.haifa.ac.il wrote: > >>> Tom, > >>> I've been trying for several years to get Israeli Arabs and Jews to > >> seriously > >>> consider the possibility of educating Israeli Arabs in a written version > of > >>> their spoken language, as you suggest, at least through 3rd grade. I can > >> send > >>> you some things I've written on this topic, with a lot of > cross-linguistic > >>> data. But thus far, it isn't working. As with most aspects of 'the > >> situation' > >>> here, politically active Arabs think that the solution is to pursue the > >> same > >>> self-destructive strategy which they've been following for almost a > century > >> but > >>> with even more vigor and steadfastness (e.g. Masalha points the blame at > >> the > >>> Arab media because they do not 'provide the linguistic richness of formal > >>> Arabic') while Jews are basically content to let the Arabs stew in their > >> own > >>> juices so that they can reap the benefits. There may, however, be hope > for > >> the > >>> future in the form of the radically increased usage of written forms of > >>> colloquial Arabic dialects in electronic media such as Facebook, blogs, > >> emails, > >>> etc., by Arabic speakers below the age of 30, which will soon come to > >> seriously > >>> threaten the status of classical Arabic in the same way that the > invention > >> of > >>> the printing press overturned the linguistic hierarchy in Western Europe > by > >>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>> > >>> I realize that Masalha as well as the researchers he referred to don't > >>> necessarily know what they're talking about regarding the brain. But the > >>> general point is still potentially significant--that the connected script > >> which > >>> Arabic uses as well as the multiple forms which many Arabic letters have > >> may be > >>> a significant obstacle to literacy, however this may be related to > >> processes in > >>> the brain. > >>> Best wishes, > >>> John > >>> > >>> > >>> > >>> Quoting Tom Givon : > >>> > >>> > >>>> A GREAT POET CAN STILL GET IT > WRONG > >>>> > >>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >>>> by stumbling a on his truly great quote: > >>>> "All fixed identities are imposed from the outside. > >>>> Whoever has a clear identity knows it can assume > >>>> multiple forms". > >>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>> intended--surely got only one third of the story right. At first, he was > >>>> led astray be the academic researchers he cited, who claimed that the > >>>> lagging reading skills of Israeli-Arab students is correlated to > >>>> lagging R-hemisphere activity, then explained this neurological lag > >>>> by suggesting that the Arab script requires more contextual analysis. > >>>> But it is the R-hemisphere of the human cortex that is more context > >>>> oriented, less automated. If Arab-reading students required more > >>>> contextual labor, it should have been registered as a higher > >>>> R-hemisphere activity, not lower. > >>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>> among many others, have established beyond reasonable doubt that > >>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>> module on the boundary of the occipital and temporal lobes (just past > >>>> Brodman's Area 19), along the ventral visual object-recognition > >>>> 'stream' that flows from the back to the front of the L-cortex. And > >>>> the L-cortex is in general responsible for the more automated--less > >>>> context-dependent--processing of language (as well as visual, motor and > >>>> other skills). The visual word-recognition module is, in turn, > >>>> recruited from the pre-existing visual object-recognition > >>>> ventral-stream module. A considerable amount of life-time practice and > >>>> repetition is required to affect this late-cultural adaptation. The > >>>> human brain is not (yet) genetically configured at birth for > >>>> visual-word recognition, only for visual-object recognition. A similar > >>>> cultural adaptation, this one for math, has been shown for (Dahaene & > >>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>> in the L-pareita lobe, an area originally configured for analysis of > >>>> object-location in space. > >>>> Mr. Masalha then, on his own, points out to a more > >>>> plausible right answer: Arab students, in Israel as well as all over the > >>>> Arab world, are not taught literacy in their native language > >>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>> language. The discrepancy would be just as great if Israeli kids were > >>>> taught their Hebrew literacy first in the language of Genesis; or if > >>>> French students were taught literacy first in the language of La > >>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or > >>>> English-speaking kids in the language of Beowolf. As far as my frail > >>>> guessing powers go, remedying the situation would be much easier by > >>>> combining two well-known verities of second language acquisition: (a) > >>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), > >>>> teach literacy first in the student's spoken native language; only then > >>>> gradually 'stretch' it to more literary genres. This method, bhy the > >>>> way, was suggested in the late 1930's by no other than L. Bloomfield, > >>>> in a book outlining a 'phonics-first ' literacy program for native > >>>> English speakers. Rather than depriving Israeli-Arab students of > >>>> literacy in their own--equally glorious--native language, just teach > >>>> them smart. > >>>> For his last culprit, the presumed--tho hardly > >>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>> recognized prejudices of the educated classes. While readily endorsing > >>>> his aesthetic sentiments about modern media, I would still like to point > >>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>> spoken native language, and if anything should facilitate the easier > >>>> initial acquisition of native-language literacy. Respectuosamente, > >>>> ma'-salaam, > >>>> > >>>> > >>>> T. Giv?n > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >> > >> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From amnfn at well.com Thu Sep 30 14:51:29 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 07:51:29 -0700 Subject: A poet slips In-Reply-To: <20100930162521.B22D.BA0BAB47@research.haifa.ac.il> Message-ID: Ron, Most of what you said made sense to me. I am not familiar with the revival of Arabic and would be happy if you posted a link for all us Funknetters on this topic. It makes sense to see different versions of the langauge as gradient. One thing you said, though, struck me as weird. "Despite the right-to-left direction...." You're not seriously suggesting that RTL or LTR makes a difference in ease of reading, are you? --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz From grvsmth at panix.com Thu Sep 30 15:01:07 2010 From: grvsmth at panix.com (Angus B. Grieve-Smith) Date: Thu, 30 Sep 2010 11:01:07 -0400 Subject: A poet slips In-Reply-To: <1285857810.4ca4a212286fb@webmail.haifa.ac.il> Message-ID: On Thu, September 30, 2010 10:43 am, john at research.haifa.ac.il wrote: > Also, the usage of written colloquial > Arabic is basically universal among all Israeli Arabs under the age of 25 > who have a cellular phone. > This is definitely not just a privileged section of the population, it's > most people of the relevant age group. That's very interesting. What alphabet do they use? -- -Angus B. Grieve-Smith grvsmth at panix.com From john at research.haifa.ac.il Thu Sep 30 15:06:52 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 17:06:52 +0200 Subject: A poet slips In-Reply-To: Message-ID: For learning to read for the first time, obviously it doesn't make a difference. But once you've learned to read in one direction, it's tough to start reading in the opposite direction, particularly if you've started late in life. I know from personal experience. John Quoting "A. Katz" : > Ron, > > Most of what you said made sense to me. I am not familiar with the revival > of Arabic and would be happy if you posted a link for all us Funknetters > on this topic. It makes sense to see different versions of the langauge as > gradient. > > One thing you said, though, struck me as weird. "Despite the right-to-left > direction...." You're not seriously suggesting that RTL or LTR makes a > difference in ease of reading, are you? > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From kuzar at research.haifa.ac.il Thu Sep 30 15:14:04 2010 From: kuzar at research.haifa.ac.il (Ron Kuzar) Date: Thu, 30 Sep 2010 17:14:04 +0200 Subject: A poet slips In-Reply-To: Message-ID: The most informative source, albeit a bit old, is: Blau, Joshua. 1981. The renaissance of modern Hebrew and modern standard Arabic : parallels and differences in the revival of two semitic languages http://books.google.com/books?id=EwbvrNRcaNIC&printsec=frontcover&dq=blau+revival+arabic&source=bl&ots=P_1nIzaFDb&sig=zCfhko8v9UY8rbZC3-OKp_P3kq4&hl=en&ei=pqikTM3NOYiPswbJoYWiCA&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBIQ6AEwAA#v=onepage&q&f=false Directionality has been suggested in scholarly circles as a problem. I am not an expert on this issue, and I do NOT believe it could be true. Ron On Thu, 30 Sep 2010 07:51:29 -0700 (PDT) "A. Katz" wrote: > Ron, > > Most of what you said made sense to me. I am not familiar with the revival > of Arabic and would be happy if you posted a link for all us Funknetters > on this topic. It makes sense to see different versions of the langauge as > gradient. > > One thing you said, though, struck me as weird. "Despite the right-to-left > direction...." You're not seriously suggesting that RTL or LTR makes a > difference in ease of reading, are you? > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > =============================================== Dr. Ron Kuzar Address: Department of English Language and Literature University of Haifa IL-31905 Haifa, Israel Office: +972-4-824-9826, Fax: +972-4-824-9711 Home: +972-77-481-9676, Mobile: +972-54-481-9676 Home fax: 153-77-481-9676 (only from Israel) Email: kuzar at research.haifa.ac.il Homepage: http://research.haifa.ac.il/~kuzar =============================================== From amnfn at well.com Thu Sep 30 15:18:51 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:18:51 -0700 Subject: A poet slips In-Reply-To: <1285857810.4ca4a212286fb@webmail.haifa.ac.il> Message-ID: John, Well, at least I'm glad you realize that one hundred years ago, and possibly even more recently, they did not think of themselves as Arabs. I still wonder whether among themselves they say "Arabs". Isn't there another word? If any of them tried to emigrate to Saudi Arabia, I doubt very much they'd be called "Arabs." For that matter, what do Jordanians call non-Jewish Israelis? The biggest obstacle to allowing for integration in the schools is not the Palestinian population -- it's those among the Israelis who adopt a "Jews first" attitude. --Aya http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Aya, > I ask them constantly if they think they are Arabs (60% of my students > are Arabic speakers and this is a central topic of many classes that I teach) > and almost without exception they do. The only exceptions are some Druze > (mostly males) and most Maronites. I am aware that 100 years ago Arabic > speakers living in this area did not consider themselves to be Arabs (except > for the Bedouins), but the situation has completely changed. > > It is totally unrealistic to put Israeli Arabs as a group into Hebrew-speaking > schools. No one wants it, not the Jews and not the Arabs. There are a tiny > number of Arabs who for one reason or another send their children to > Hebrew-speaking schools (for example the author Sayed Kashua) but this > is insignificant. > > I don't know what the situation was when you were a child, and I don't know > about the situation among religious Jews, but secular Jewish children do not > learn to read by reading Genesis today. They learn to read with secular texts > in 1st grade and go on to the Bible in 2nd grade, but far from learning to > read by reading the Bible, teachers have to explain what's written in the > Bible to the students. NO ONE can understand it without an explanation from the > teacher or their parents. > > Ron--I agree with everything you say except that I believe that the radical > difference between spoken and written Arabic definitely is a serious problem > for literacy. I would not be so quick to dismiss the effect of the Arabic > writing system--we had a conference on this topic in Haifa in May and the > general consensus of people who sounded like they knew what they were talking > about was that it is a significant problem-- but they hadn't done > cross-linguistic research and this isn't something I know enough about to have > strong opinions one way or the other. Also, the usage of written colloquial > Arabic is basically universal among all Israeli Arabs under the age of 25 who > have a cellular phone. > This is definitely not just a privileged section of the population, it's most > people of the relevant age group. > > John > > > > > > Quoting "A. Katz" : > >> John, >> >> These people you speak of are not Arabs. Some of them are Moslem and they >> read the Quran in the original. Some of them are not Moslem. All of them >> speak a local dialect of Arabic. Ask them sometimes if they think they are >> Arabs. >> >> Trying to turn every dialect into a separate language with a separate >> writing system is a way to try to disunite people. But a common language, >> however differently it is pronounced, unites disparate people. Australians >> and Cockneys and Indians and Americans speak sometimes mutually >> unintelligible versions of English. Using the same writing system and >> the same classic texts unites them. >> >> Instead of telling people they should magnify every difference, why not >> offer to share your language with them? Hebrew could be a uniting factor >> if spoken in all Israeli schools. >> >> --Aya >> >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Learning the classical language is like pulling teeth for Arabs kids also. >> They >>> just can't publicly say it because that would make them bad Arabs. It's >> just >>> ridiculous. I have an even better plan for the Jewish kids--teach them to >> use >>> the written version of the spoken language which Arab kids are using for >>> Facebook. That's how kids make friends these days anyway. >>> Best wishes, >>> John >>> >>> >>> >>> >>> >>> Quoting Tom Givon : >>> >>>> >>>> >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>> Classical (after 1949 things changed...). That book may still exist, >>>> you might track it down. It was easy, a cinch really. I concluded it was >>>> really just Hebrew with a few trivial transformations in Phonology & >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>>> on a few sessions of a faculty study group at UO who were trying to >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>> well, in 1939 (or was it 1943?). >>>> Cheers, TG >>>> >>>> ============== >>>> >>>> >>>> john at research.haifa.ac.il wrote: >>>>> Tom, >>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>> seriously >>>>> consider the possibility of educating Israeli Arabs in a written version >> of >>>>> their spoken language, as you suggest, at least through 3rd grade. I can >>>> send >>>>> you some things I've written on this topic, with a lot of >> cross-linguistic >>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>> situation' >>>>> here, politically active Arabs think that the solution is to pursue the >>>> same >>>>> self-destructive strategy which they've been following for almost a >> century >>>> but >>>>> with even more vigor and steadfastness (e.g. Masalha points the blame at >>>> the >>>>> Arab media because they do not 'provide the linguistic richness of formal >>>>> Arabic') while Jews are basically content to let the Arabs stew in their >>>> own >>>>> juices so that they can reap the benefits. There may, however, be hope >> for >>>> the >>>>> future in the form of the radically increased usage of written forms of >>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>> emails, >>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>> seriously >>>>> threaten the status of classical Arabic in the same way that the >> invention >>>> of >>>>> the printing press overturned the linguistic hierarchy in Western Europe >> by >>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>> >>>>> I realize that Masalha as well as the researchers he referred to don't >>>>> necessarily know what they're talking about regarding the brain. But the >>>>> general point is still potentially significant--that the connected script >>>> which >>>>> Arabic uses as well as the multiple forms which many Arabic letters have >>>> may be >>>>> a significant obstacle to literacy, however this may be related to >>>> processes in >>>>> the brain. >>>>> Best wishes, >>>>> John >>>>> >>>>> >>>>> >>>>> Quoting Tom Givon : >>>>> >>>>> >>>>>> A GREAT POET CAN STILL GET IT >> WRONG >>>>>> >>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>>> by stumbling a on his truly great quote: >>>>>> "All fixed identities are imposed from the outside. >>>>>> Whoever has a clear identity knows it can assume >>>>>> multiple forms". >>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>> intended--surely got only one third of the story right. At first, he was >>>>>> led astray be the academic researchers he cited, who claimed that the >>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>>> by suggesting that the Arab script requires more contextual analysis. >>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>> oriented, less automated. If Arab-reading students required more >>>>>> contextual labor, it should have been registered as a higher >>>>>> R-hemisphere activity, not lower. >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>> among many others, have established beyond reasonable doubt that >>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>> module on the boundary of the occipital and temporal lobes (just past >>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>>> the L-cortex is in general responsible for the more automated--less >>>>>> context-dependent--processing of language (as well as visual, motor and >>>>>> other skills). The visual word-recognition module is, in turn, >>>>>> recruited from the pre-existing visual object-recognition >>>>>> ventral-stream module. A considerable amount of life-time practice and >>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>> human brain is not (yet) genetically configured at birth for >>>>>> visual-word recognition, only for visual-object recognition. A similar >>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>> object-location in space. >>>>>> Mr. Masalha then, on his own, points out to a more >>>>>> plausible right answer: Arab students, in Israel as well as all over the >>>>>> Arab world, are not taught literacy in their native language >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>> language. The discrepancy would be just as great if Israeli kids were >>>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>>> French students were taught literacy first in the language of La >>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>>> guessing powers go, remedying the situation would be much easier by >>>>>> combining two well-known verities of second language acquisition: (a) >>>>>> Teach them both early, together--'co-ordinated bilingualism'. And (b), >>>>>> teach literacy first in the student's spoken native language; only then >>>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>>> way, was suggested in the late 1930's by no other than L. Bloomfield, >>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>> literacy in their own--equally glorious--native language, just teach >>>>>> them smart. >>>>>> For his last culprit, the presumed--tho hardly >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>> recognized prejudices of the educated classes. While readily endorsing >>>>>> his aesthetic sentiments about modern media, I would still like to point >>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>> spoken native language, and if anything should facilitate the easier >>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>> ma'-salaam, >>>>>> >>>>>> >>>>>> T. Giv?n >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>> >>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 15:23:09 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:23:09 -0700 Subject: A poet slips In-Reply-To: <1285859212.4ca4a78cc9912@webmail.haifa.ac.il> Message-ID: Well, everything is hard if you start too late. But if you start almost simultaneously in both directions in different languages, then, in my experience, it's not a problem. --Aya On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > For learning to read for the first time, obviously it doesn't make a difference. > But once you've learned to read in one direction, it's tough to start reading > in the opposite direction, particularly if you've started late in life. I > know from personal experience. > John > > > > > > Quoting "A. Katz" : > >> Ron, >> >> Most of what you said made sense to me. I am not familiar with the revival >> of Arabic and would be happy if you posted a link for all us Funknetters >> on this topic. It makes sense to see different versions of the langauge as >> gradient. >> >> One thing you said, though, struck me as weird. "Despite the right-to-left >> direction...." You're not seriously suggesting that RTL or LTR makes a >> difference in ease of reading, are you? >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> >> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 15:24:14 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 08:24:14 -0700 Subject: A poet slips In-Reply-To: <20100930171403.B234.BA0BAB47@research.haifa.ac.il> Message-ID: Ron, Thanks for the link! --Aya On Thu, 30 Sep 2010, Ron Kuzar wrote: > The most informative source, albeit a bit old, is: > Blau, Joshua. 1981. The renaissance of modern Hebrew and modern standard > Arabic : parallels and differences in the revival of two semitic languages > http://books.google.com/books?id=EwbvrNRcaNIC&printsec=frontcover&dq=blau+revival+arabic&source=bl&ots=P_1nIzaFDb&sig=zCfhko8v9UY8rbZC3-OKp_P3kq4&hl=en&ei=pqikTM3NOYiPswbJoYWiCA&sa=X&oi=book_result&ct=result&resnum=1&ved=0CBIQ6AEwAA#v=onepage&q&f=false > Directionality has been suggested in scholarly circles as a problem. I > am not an expert on this issue, and I do NOT believe it could be true. > Ron > > > > On Thu, 30 Sep 2010 07:51:29 -0700 (PDT) > "A. Katz" wrote: > >> Ron, >> >> Most of what you said made sense to me. I am not familiar with the revival >> of Arabic and would be happy if you posted a link for all us Funknetters >> on this topic. It makes sense to see different versions of the langauge as >> gradient. >> >> One thing you said, though, struck me as weird. "Despite the right-to-left >> direction...." You're not seriously suggesting that RTL or LTR makes a >> difference in ease of reading, are you? >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> > > =============================================== > Dr. Ron Kuzar > Address: Department of English Language and Literature > University of Haifa > IL-31905 Haifa, Israel > Office: +972-4-824-9826, Fax: +972-4-824-9711 > Home: +972-77-481-9676, Mobile: +972-54-481-9676 > Home fax: 153-77-481-9676 (only from Israel) > Email: kuzar at research.haifa.ac.il > Homepage: http://research.haifa.ac.il/~kuzar > =============================================== > > > From john at research.haifa.ac.il Thu Sep 30 15:26:50 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 17:26:50 +0200 Subject: A poet slips In-Reply-To: Message-ID: Yes, actually I thought more and there is some truth to what you said. It seems that at a psychological level Arabs in other countries don't consider them to be 'real Arabs'. A student of mine who went to the Egyptian national museum in Cairo saw that there were radically different entry fees posted for 'Arabs' and 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in fact he was a Druze who had served in the Israeli Army, but they had no way of knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker asked to see his passport and when he saw that it was from Israel informed him that he was not an Arab and would have to pay the higher 'non-Arab' fee. I have told this story to many of my Israeli Arab students and not a single one has even expressed surprise--they completely expect such treatment. When I suggest that this means that they are not regarded as being real Arabs by Arabs from other countries, they admit that this seems to be true--but they still regard themselves as being Arabs. I then sum up the situation by saying that the people called 'Israeli Arabs' are not considered to be Israelis by Israelis and they are not considered to be Arabs by Arabs. It's complicated... With regard to the obstacles to integration, it really is true that neither side wants integration if this threatens their distinctive identities. In no sense can it be said that the Jewish side rejects assimilation more than the Arab side. I remember one class in which a Christian Arab woman reported temporarily sending her child to a Jewish kindergarten because it was the most convenient one--but when the child came home one day and insisted on lighting candles on Friday evening, the parents realized that this was too much and they needed another solution. Everyone in the class, Jews and Arabs, recognized this--except for an American visiting scholar who said 'What's the problem, it's cute!'. Everyone looked at him like he'd landed from outer space. The difference is more that the Arab side wants EQUALITY, but (contrary to what Americans generally think) that isn't the same thing as integration. John Quoting "A. Katz" : > John, > > Well, at least I'm glad you realize that one hundred years ago, and > possibly even more recently, they did not think of themselves as Arabs. > I still wonder whether among themselves they say "Arabs". Isn't there > another word? If any of them tried to emigrate to Saudi Arabia, I doubt > very much they'd be called "Arabs." For that matter, what do Jordanians > call non-Jewish Israelis? > > The biggest obstacle to allowing for integration in the schools is not the > Palestinian population -- it's those among the Israelis who adopt a "Jews > first" attitude. > > --Aya > > http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Aya, > > I ask them constantly if they think they are Arabs (60% of my students > > are Arabic speakers and this is a central topic of many classes that I > teach) > > and almost without exception they do. The only exceptions are some Druze > > (mostly males) and most Maronites. I am aware that 100 years ago Arabic > > speakers living in this area did not consider themselves to be Arabs > (except > > for the Bedouins), but the situation has completely changed. > > > > It is totally unrealistic to put Israeli Arabs as a group into > Hebrew-speaking > > schools. No one wants it, not the Jews and not the Arabs. There are a tiny > > number of Arabs who for one reason or another send their children to > > Hebrew-speaking schools (for example the author Sayed Kashua) but this > > is insignificant. > > > > I don't know what the situation was when you were a child, and I don't know > > about the situation among religious Jews, but secular Jewish children do > not > > learn to read by reading Genesis today. They learn to read with secular > texts > > in 1st grade and go on to the Bible in 2nd grade, but far from learning to > > read by reading the Bible, teachers have to explain what's written in the > > Bible to the students. NO ONE can understand it without an explanation from > the > > teacher or their parents. > > > > Ron--I agree with everything you say except that I believe that the radical > > difference between spoken and written Arabic definitely is a serious > problem > > for literacy. I would not be so quick to dismiss the effect of the Arabic > > writing system--we had a conference on this topic in Haifa in May and the > > general consensus of people who sounded like they knew what they were > talking > > about was that it is a significant problem-- but they hadn't done > > cross-linguistic research and this isn't something I know enough about to > have > > strong opinions one way or the other. Also, the usage of written colloquial > > Arabic is basically universal among all Israeli Arabs under the age of 25 > who > > have a cellular phone. > > This is definitely not just a privileged section of the population, it's > most > > people of the relevant age group. > > > > John > > > > > > > > > > > > Quoting "A. Katz" : > > > >> John, > >> > >> These people you speak of are not Arabs. Some of them are Moslem and they > >> read the Quran in the original. Some of them are not Moslem. All of them > >> speak a local dialect of Arabic. Ask them sometimes if they think they are > >> Arabs. > >> > >> Trying to turn every dialect into a separate language with a separate > >> writing system is a way to try to disunite people. But a common language, > >> however differently it is pronounced, unites disparate people. Australians > >> and Cockneys and Indians and Americans speak sometimes mutually > >> unintelligible versions of English. Using the same writing system and > >> the same classic texts unites them. > >> > >> Instead of telling people they should magnify every difference, why not > >> offer to share your language with them? Hebrew could be a uniting factor > >> if spoken in all Israeli schools. > >> > >> --Aya > >> > >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > >> > >> > >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >> > >>> Learning the classical language is like pulling teeth for Arabs kids > also. > >> They > >>> just can't publicly say it because that would make them bad Arabs. It's > >> just > >>> ridiculous. I have an even better plan for the Jewish kids--teach them to > >> use > >>> the written version of the spoken language which Arab kids are using for > >>> Facebook. That's how kids make friends these days anyway. > >>> Best wishes, > >>> John > >>> > >>> > >>> > >>> > >>> > >>> Quoting Tom Givon : > >>> > >>>> > >>>> > >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our > >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >>>> Classical (after 1949 things changed...). That book may still exist, > >>>> you might track it down. It was easy, a cinch really. I concluded it was > >>>> really just Hebrew with a few trivial transformations in Phonology & > >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat > >>>> on a few sessions of a faculty study group at UO who were trying to > >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, > >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >>>> well, in 1939 (or was it 1943?). > >>>> Cheers, TG > >>>> > >>>> ============== > >>>> > >>>> > >>>> john at research.haifa.ac.il wrote: > >>>>> Tom, > >>>>> I've been trying for several years to get Israeli Arabs and Jews to > >>>> seriously > >>>>> consider the possibility of educating Israeli Arabs in a written > version > >> of > >>>>> their spoken language, as you suggest, at least through 3rd grade. I > can > >>>> send > >>>>> you some things I've written on this topic, with a lot of > >> cross-linguistic > >>>>> data. But thus far, it isn't working. As with most aspects of 'the > >>>> situation' > >>>>> here, politically active Arabs think that the solution is to pursue the > >>>> same > >>>>> self-destructive strategy which they've been following for almost a > >> century > >>>> but > >>>>> with even more vigor and steadfastness (e.g. Masalha points the blame > at > >>>> the > >>>>> Arab media because they do not 'provide the linguistic richness of > formal > >>>>> Arabic') while Jews are basically content to let the Arabs stew in > their > >>>> own > >>>>> juices so that they can reap the benefits. There may, however, be hope > >> for > >>>> the > >>>>> future in the form of the radically increased usage of written forms of > >>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, > >>>> emails, > >>>>> etc., by Arabic speakers below the age of 30, which will soon come to > >>>> seriously > >>>>> threaten the status of classical Arabic in the same way that the > >> invention > >>>> of > >>>>> the printing press overturned the linguistic hierarchy in Western > Europe > >> by > >>>>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>>>> > >>>>> I realize that Masalha as well as the researchers he referred to don't > >>>>> necessarily know what they're talking about regarding the brain. But > the > >>>>> general point is still potentially significant--that the connected > script > >>>> which > >>>>> Arabic uses as well as the multiple forms which many Arabic letters > have > >>>> may be > >>>>> a significant obstacle to literacy, however this may be related to > >>>> processes in > >>>>> the brain. > >>>>> Best wishes, > >>>>> John > >>>>> > >>>>> > >>>>> > >>>>> Quoting Tom Givon : > >>>>> > >>>>> > >>>>>> A GREAT POET CAN STILL GET IT > >> WRONG > >>>>>> > >>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally > >>>>>> by stumbling a on his truly great quote: > >>>>>> "All fixed identities are imposed from the outside. > >>>>>> Whoever has a clear identity knows it can assume > >>>>>> multiple forms". > >>>>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>>>> intended--surely got only one third of the story right. At first, he > was > >>>>>> led astray be the academic researchers he cited, who claimed that the > >>>>>> lagging reading skills of Israeli-Arab students is correlated to > >>>>>> lagging R-hemisphere activity, then explained this neurological lag > >>>>>> by suggesting that the Arab script requires more contextual analysis. > >>>>>> But it is the R-hemisphere of the human cortex that is more context > >>>>>> oriented, less automated. If Arab-reading students required more > >>>>>> contextual labor, it should have been registered as a higher > >>>>>> R-hemisphere activity, not lower. > >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>>>> among many others, have established beyond reasonable doubt that > >>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>>>> module on the boundary of the occipital and temporal lobes (just past > >>>>>> Brodman's Area 19), along the ventral visual object-recognition > >>>>>> 'stream' that flows from the back to the front of the L-cortex. And > >>>>>> the L-cortex is in general responsible for the more automated--less > >>>>>> context-dependent--processing of language (as well as visual, motor > and > >>>>>> other skills). The visual word-recognition module is, in turn, > >>>>>> recruited from the pre-existing visual object-recognition > >>>>>> ventral-stream module. A considerable amount of life-time practice > and > >>>>>> repetition is required to affect this late-cultural adaptation. The > >>>>>> human brain is not (yet) genetically configured at birth for > >>>>>> visual-word recognition, only for visual-object recognition. A similar > >>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & > >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>>>> in the L-pareita lobe, an area originally configured for analysis of > >>>>>> object-location in space. > >>>>>> Mr. Masalha then, on his own, points out to a more > >>>>>> plausible right answer: Arab students, in Israel as well as all over > the > >>>>>> Arab world, are not taught literacy in their native language > >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>>>> language. The discrepancy would be just as great if Israeli kids > were > >>>>>> taught their Hebrew literacy first in the language of Genesis; or if > >>>>>> French students were taught literacy first in the language of La > >>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or > >>>>>> English-speaking kids in the language of Beowolf. As far as my frail > >>>>>> guessing powers go, remedying the situation would be much easier by > >>>>>> combining two well-known verities of second language acquisition: (a) > >>>>>> Teach them both early, together--'co-ordinated bilingualism'. And > (b), > >>>>>> teach literacy first in the student's spoken native language; only > then > >>>>>> gradually 'stretch' it to more literary genres. This method, bhy the > >>>>>> way, was suggested in the late 1930's by no other than L. > Bloomfield, > >>>>>> in a book outlining a 'phonics-first ' literacy program for native > >>>>>> English speakers. Rather than depriving Israeli-Arab students of > >>>>>> literacy in their own--equally glorious--native language, just teach > >>>>>> them smart. > >>>>>> For his last culprit, the presumed--tho hardly > >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>>>> recognized prejudices of the educated classes. While readily > endorsing > >>>>>> his aesthetic sentiments about modern media, I would still like to > point > >>>>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>>>> spoken native language, and if anything should facilitate the easier > >>>>>> initial acquisition of native-language literacy. Respectuosamente, > >>>>>> ma'-salaam, > >>>>>> > >>>>>> > >>>>>> T. Giv?n > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > ------------------------------------------------------------------------ > >>>>> This message was sent using IMP, the Webmail Program of Haifa > University > >>>>> > >>>> > >>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >>> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From rcameron at uic.edu Thu Sep 30 15:45:48 2010 From: rcameron at uic.edu (Cameron, Richard) Date: Thu, 30 Sep 2010 10:45:48 -0500 Subject: A poet slips In-Reply-To: <1285860410.4ca4ac3acc01a@webmail.haifa.ac.il> Message-ID: John's point about "that the people called 'Israeli Arabs' are not considered to be Israelis by Israelis and they are not considered to be Arabs by Arabs." may (I say, may) have some curious parallels among Puerto Ricans and Mexicans in the States in their relationships with those countries as they will claim that back home they are called Americanos but in the States they are not. The old "Ni de aqui, ni de alla." Not from here, not from there. - Richard Cameron On Thu, September 30, 2010 10:26 am, john at research.haifa.ac.il wrote: > Yes, actually I thought more and there is some truth to what you said. It > seems > that at a psychological level Arabs in other countries don't consider them > to > be 'real Arabs'. A student of mine who went to the Egyptian national > museum in > Cairo saw that there were radically different entry fees posted for > 'Arabs' and > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in > fact > he was a Druze who had served in the Israeli Army, but they had no way of > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker > asked > to see his passport and when he saw that > it was from Israel informed him that he was not an Arab and would have to > pay > the higher 'non-Arab' fee. I have told this story to many of my Israeli > Arab > students and not a single one has even expressed surprise--they completely > expect such treatment. When I suggest that this means that they are not > regarded as being real Arabs by Arabs from other countries, they admit > that > this seems to be true--but they still regard themselves as being Arabs. I > then > sum up the situation by saying that the people called 'Israeli Arabs' are > not > considered to be Israelis by Israelis and they are not considered to be > Arabs > by Arabs. > It's complicated... > > With regard to the obstacles to integration, it really is true that > neither side > wants integration if this threatens their distinctive identities. In no > sense > can it be said that the Jewish side rejects assimilation more than the > Arab > side. I remember one class in which a Christian Arab woman reported > temporarily > sending her child to a Jewish kindergarten because it was the most > convenient > one--but when the child came home one day and insisted on lighting candles > on > Friday evening, the parents realized that this was too much and they > needed > another solution. Everyone in the class, Jews and Arabs, recognized > this--except for an American visiting scholar who said 'What's the > problem, > it's cute!'. Everyone looked at him like he'd landed from outer space. The > difference is more that the Arab side wants EQUALITY, but (contrary to > what > Americans generally think) that isn't the same thing as integration. > > John > > > > Quoting "A. Katz" : > >> John, >> >> Well, at least I'm glad you realize that one hundred years ago, and >> possibly even more recently, they did not think of themselves as Arabs. >> I still wonder whether among themselves they say "Arabs". Isn't there >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >> very much they'd be called "Arabs." For that matter, what do Jordanians >> call non-Jewish Israelis? >> >> The biggest obstacle to allowing for integration in the schools is not >> the >> Palestinian population -- it's those among the Israelis who adopt a >> "Jews >> first" attitude. >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >> > Aya, >> > I ask them constantly if they think they are Arabs (60% of my students >> > are Arabic speakers and this is a central topic of many classes that I >> teach) >> > and almost without exception they do. The only exceptions are some >> Druze >> > (mostly males) and most Maronites. I am aware that 100 years ago >> Arabic >> > speakers living in this area did not consider themselves to be Arabs >> (except >> > for the Bedouins), but the situation has completely changed. >> > >> > It is totally unrealistic to put Israeli Arabs as a group into >> Hebrew-speaking >> > schools. No one wants it, not the Jews and not the Arabs. There are a >> tiny >> > number of Arabs who for one reason or another send their children to >> > Hebrew-speaking schools (for example the author Sayed Kashua) but this >> > is insignificant. >> > >> > I don't know what the situation was when you were a child, and I don't >> know >> > about the situation among religious Jews, but secular Jewish children >> do >> not >> > learn to read by reading Genesis today. They learn to read with >> secular >> texts >> > in 1st grade and go on to the Bible in 2nd grade, but far from >> learning to >> > read by reading the Bible, teachers have to explain what's written in >> the >> > Bible to the students. NO ONE can understand it without an explanation >> from >> the >> > teacher or their parents. >> > >> > Ron--I agree with everything you say except that I believe that the >> radical >> > difference between spoken and written Arabic definitely is a serious >> problem >> > for literacy. I would not be so quick to dismiss the effect of the >> Arabic >> > writing system--we had a conference on this topic in Haifa in May and >> the >> > general consensus of people who sounded like they knew what they were >> talking >> > about was that it is a significant problem-- but they hadn't done >> > cross-linguistic research and this isn't something I know enough about >> to >> have >> > strong opinions one way or the other. Also, the usage of written >> colloquial >> > Arabic is basically universal among all Israeli Arabs under the age of >> 25 >> who >> > have a cellular phone. >> > This is definitely not just a privileged section of the population, >> it's >> most >> > people of the relevant age group. >> > >> > John >> > >> > >> > >> > >> > >> > Quoting "A. Katz" : >> > >> >> John, >> >> >> >> These people you speak of are not Arabs. Some of them are Moslem and >> they >> >> read the Quran in the original. Some of them are not Moslem. All of >> them >> >> speak a local dialect of Arabic. Ask them sometimes if they think >> they are >> >> Arabs. >> >> >> >> Trying to turn every dialect into a separate language with a separate >> >> writing system is a way to try to disunite people. But a common >> language, >> >> however differently it is pronounced, unites disparate people. >> Australians >> >> and Cockneys and Indians and Americans speak sometimes mutually >> >> unintelligible versions of English. Using the same writing system and >> >> the same classic texts unites them. >> >> >> >> Instead of telling people they should magnify every difference, why >> not >> >> offer to share your language with them? Hebrew could be a uniting >> factor >> >> if spoken in all Israeli schools. >> >> >> >> --Aya >> >> >> >> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >> >> >> >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >> >> >>> Learning the classical language is like pulling teeth for Arabs kids >> also. >> >> They >> >>> just can't publicly say it because that would make them bad Arabs. >> It's >> >> just >> >>> ridiculous. I have an even better plan for the Jewish kids--teach >> them to >> >> use >> >>> the written version of the spoken language which Arab kids are using >> for >> >>> Facebook. That's how kids make friends these days anyway. >> >>> Best wishes, >> >>> John >> >>> >> >>> >> >>> >> >>> >> >>> >> >>> Quoting Tom Givon : >> >>> >> >>>> >> >>>> >> >>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, >> our >> >>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got >> to >> >>>> Classical (after 1949 things changed...). That book may still >> exist, >> >>>> you might track it down. It was easy, a cinch really. I concluded >> it was >> >>>> really just Hebrew with a few trivial transformations in Phonology >> & >> >>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I >> sat >> >>>> on a few sessions of a faculty study group at UO who were trying to >> >>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. >> Boy, >> >>>> it was like pullin' teeth. But Leonard Bloomfield said it already, >> and >> >>>> well, in 1939 (or was it 1943?). >> >>>> Cheers, TG >> >>>> >> >>>> ============== >> >>>> >> >>>> >> >>>> john at research.haifa.ac.il wrote: >> >>>>> Tom, >> >>>>> I've been trying for several years to get Israeli Arabs and Jews >> to >> >>>> seriously >> >>>>> consider the possibility of educating Israeli Arabs in a written >> version >> >> of >> >>>>> their spoken language, as you suggest, at least through 3rd grade. >> I >> can >> >>>> send >> >>>>> you some things I've written on this topic, with a lot of >> >> cross-linguistic >> >>>>> data. But thus far, it isn't working. As with most aspects of 'the >> >>>> situation' >> >>>>> here, politically active Arabs think that the solution is to >> pursue the >> >>>> same >> >>>>> self-destructive strategy which they've been following for almost >> a >> >> century >> >>>> but >> >>>>> with even more vigor and steadfastness (e.g. Masalha points the >> blame >> at >> >>>> the >> >>>>> Arab media because they do not 'provide the linguistic richness of >> formal >> >>>>> Arabic') while Jews are basically content to let the Arabs stew in >> their >> >>>> own >> >>>>> juices so that they can reap the benefits. There may, however, be >> hope >> >> for >> >>>> the >> >>>>> future in the form of the radically increased usage of written >> forms of >> >>>>> colloquial Arabic dialects in electronic media such as Facebook, >> blogs, >> >>>> emails, >> >>>>> etc., by Arabic speakers below the age of 30, which will soon come >> to >> >>>> seriously >> >>>>> threaten the status of classical Arabic in the same way that the >> >> invention >> >>>> of >> >>>>> the printing press overturned the linguistic hierarchy in Western >> Europe >> >> by >> >>>>> overthrowing Latin. I'm working on this too. We'll see what >> happens. >> >>>>> >> >>>>> I realize that Masalha as well as the researchers he referred to >> don't >> >>>>> necessarily know what they're talking about regarding the brain. >> But >> the >> >>>>> general point is still potentially significant--that the connected >> script >> >>>> which >> >>>>> Arabic uses as well as the multiple forms which many Arabic >> letters >> have >> >>>> may be >> >>>>> a significant obstacle to literacy, however this may be related to >> >>>> processes in >> >>>>> the brain. >> >>>>> Best wishes, >> >>>>> John >> >>>>> >> >>>>> >> >>>>> >> >>>>> Quoting Tom Givon : >> >>>>> >> >>>>> >> >>>>>> A GREAT POET CAN STILL GET >> IT >> >> WRONG >> >>>>>> >> >>>>>> I came to know of Salman Masalha ("Arabs, speak >> Hebrew!", >> >>>>>> Haaretz/English, International Herald Tribune 9-27-10) >> accidentally >> >>>>>> by stumbling a on his truly great quote: >> >>>>>> "All fixed identities are imposed from the >> outside. >> >>>>>> Whoever has a clear identity knows it can >> assume >> >>>>>> multiple forms". >> >>>>>> In the context of Palestine/Israel, what a breath of fresh, >> rare >> >>>>>> clarity. Still, like the rest of us mortals, a great poet can >> on >> >>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; >> pun >> >>>>>> intended--surely got only one third of the story right. At first, >> he >> was >> >>>>>> led astray be the academic researchers he cited, who claimed that >> the >> >>>>>> lagging reading skills of Israeli-Arab students is correlated >> to >> >>>>>> lagging R-hemisphere activity, then explained this >> neurological lag >> >>>>>> by suggesting that the Arab script requires more contextual >> analysis. >> >>>>>> But it is the R-hemisphere of the human cortex that is more >> context >> >>>>>> oriented, less automated. If Arab-reading students required more >> >>>>>> contextual labor, it should have been registered as a higher >> >>>>>> R-hemisphere activity, not lower. >> >>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. >> Dahane, >> >>>>>> among many others, have established beyond reasonable doubt >> that >> >>>>>> written words in all languages (English, Mandarin, Hindi, >> Amharic, >> >>>>>> Hebrew, Arabic, etc.) are decoded automatically in an >> L-hemisphere >> >>>>>> module on the boundary of the occipital and temporal lobes (just >> past >> >>>>>> Brodman's Area 19), along the ventral visual object-recognition >> >>>>>> 'stream' that flows from the back to the front of the L-cortex. >> And >> >>>>>> the L-cortex is in general responsible for the more >> automated--less >> >>>>>> context-dependent--processing of language (as well as visual, >> motor >> and >> >>>>>> other skills). The visual word-recognition module is, in turn, >> >>>>>> recruited from the pre-existing visual object-recognition >> >>>>>> ventral-stream module. A considerable amount of life-time >> practice >> and >> >>>>>> repetition is required to affect this late-cultural adaptation. >> The >> >>>>>> human brain is not (yet) genetically configured at birth for >> >>>>>> visual-word recognition, only for visual-object recognition. A >> similar >> >>>>>> cultural adaptation, this one for math, has been shown for >> (Dahaene & >> >>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver >> Sachs) >> >>>>>> in the L-pareita lobe, an area originally configured for analysis >> of >> >>>>>> object-location in space. >> >>>>>> Mr. Masalha then, on his own, points out to a more >> >>>>>> plausible right answer: Arab students, in Israel as well as all >> over >> the >> >>>>>> Arab world, are not taught literacy in their native language >> >>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen >> literary >> >>>>>> instrument harking back 1,400 years or more. That is, in a >> foreign >> >>>>>> language. The discrepancy would be just as great if Israeli >> kids >> were >> >>>>>> taught their Hebrew literacy first in the language of Genesis; >> or if >> >>>>>> French students were taught literacy first in the language of >> La >> >>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. >> Or >> >>>>>> English-speaking kids in the language of Beowolf. As far as my >> frail >> >>>>>> guessing powers go, remedying the situation would be much easier >> by >> >>>>>> combining two well-known verities of second language acquisition: >> (a) >> >>>>>> Teach them both early, together--'co-ordinated bilingualism'. >> And >> (b), >> >>>>>> teach literacy first in the student's spoken native language; >> only >> then >> >>>>>> gradually 'stretch' it to more literary genres. This method, bhy >> the >> >>>>>> way, was suggested in the late 1930's by no other than L. >> Bloomfield, >> >>>>>> in a book outlining a 'phonics-first ' literacy program for >> native >> >>>>>> English speakers. Rather than depriving Israeli-Arab students of >> >>>>>> literacy in their own--equally glorious--native language, just >> teach >> >>>>>> them smart. >> >>>>>> For his last culprit, the presumed--tho hardly >> >>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >> >>>>>> recognized prejudices of the educated classes. While readily >> endorsing >> >>>>>> his aesthetic sentiments about modern media, I would still like >> to >> point >> >>>>>> out that the 'vulgar' genre is much closer to the Arab students' >> >>>>>> spoken native language, and if anything should facilitate the >> easier >> >>>>>> initial acquisition of native-language literacy. >> Respectuosamente, >> >>>>>> ma'-salaam, >> >>>>>> >> >>>>>> >> >>>>>> T. Giv?n >> >>>>>> >> >>>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> >>>>> >> ------------------------------------------------------------------------ >> >>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >> >>>>> >> >>>> >> >>>> >> >>> >> >>> >> >>> >> >>> >> >>> ------------------------------------------------------------------------ >> >>> This message was sent using IMP, the Webmail Program of Haifa >> University >> >>> >> >>> >> > >> > >> > >> > >> > ------------------------------------------------------------------------ >> > This message was sent using IMP, the Webmail Program of Haifa >> University >> > >> > > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From amnfn at well.com Thu Sep 30 16:10:45 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 09:10:45 -0700 Subject: A poet slips In-Reply-To: <1285860410.4ca4ac3acc01a@webmail.haifa.ac.il> Message-ID: John, Thanks for sharing the story about the Egyptian national museum. It's ironic that this would happen there, of all places, because ethnically most Egyptians are not Arab, either. Yes, it is complicated. Everybody wants to belong somewhere, and the reason this particular group of people are denied this opportunity is through a weird fluke of history. To Arabs, they are Judeans or Palestinians. To Jews, they are Arabs. The story about the Kindergarten is very sad. Why does religion have to be part of the curriculum in a public school? Separation of Church and State would go a long way toward integration and ultimately peace. --Aya But, I agree, it is complicated. On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Yes, actually I thought more and there is some truth to what you said. It seems > that at a psychological level Arabs in other countries don't consider them to > be 'real Arabs'. A student of mine who went to the Egyptian national museum in > Cairo saw that there were radically different entry fees posted for 'Arabs' and > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in fact > he was a Druze who had served in the Israeli Army, but they had no way of > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker asked > to see his passport and when he saw that > it was from Israel informed him that he was not an Arab and would have to pay > the higher 'non-Arab' fee. I have told this story to many of my Israeli Arab > students and not a single one has even expressed surprise--they completely > expect such treatment. When I suggest that this means that they are not > regarded as being real Arabs by Arabs from other countries, they admit that > this seems to be true--but they still regard themselves as being Arabs. I then > sum up the situation by saying that the people called 'Israeli Arabs' are not > considered to be Israelis by Israelis and they are not considered to be Arabs > by Arabs. > It's complicated... > > With regard to the obstacles to integration, it really is true that neither side > wants integration if this threatens their distinctive identities. In no sense > can it be said that the Jewish side rejects assimilation more than the Arab > side. I remember one class in which a Christian Arab woman reported temporarily > sending her child to a Jewish kindergarten because it was the most convenient > one--but when the child came home one day and insisted on lighting candles on > Friday evening, the parents realized that this was too much and they needed > another solution. Everyone in the class, Jews and Arabs, recognized > this--except for an American visiting scholar who said 'What's the problem, > it's cute!'. Everyone looked at him like he'd landed from outer space. The > difference is more that the Arab side wants EQUALITY, but (contrary to what > Americans generally think) that isn't the same thing as integration. > > John > > > > Quoting "A. Katz" : > >> John, >> >> Well, at least I'm glad you realize that one hundred years ago, and >> possibly even more recently, they did not think of themselves as Arabs. >> I still wonder whether among themselves they say "Arabs". Isn't there >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >> very much they'd be called "Arabs." For that matter, what do Jordanians >> call non-Jewish Israelis? >> >> The biggest obstacle to allowing for integration in the schools is not the >> Palestinian population -- it's those among the Israelis who adopt a "Jews >> first" attitude. >> >> --Aya >> >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >> >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Aya, >>> I ask them constantly if they think they are Arabs (60% of my students >>> are Arabic speakers and this is a central topic of many classes that I >> teach) >>> and almost without exception they do. The only exceptions are some Druze >>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic >>> speakers living in this area did not consider themselves to be Arabs >> (except >>> for the Bedouins), but the situation has completely changed. >>> >>> It is totally unrealistic to put Israeli Arabs as a group into >> Hebrew-speaking >>> schools. No one wants it, not the Jews and not the Arabs. There are a tiny >>> number of Arabs who for one reason or another send their children to >>> Hebrew-speaking schools (for example the author Sayed Kashua) but this >>> is insignificant. >>> >>> I don't know what the situation was when you were a child, and I don't know >>> about the situation among religious Jews, but secular Jewish children do >> not >>> learn to read by reading Genesis today. They learn to read with secular >> texts >>> in 1st grade and go on to the Bible in 2nd grade, but far from learning to >>> read by reading the Bible, teachers have to explain what's written in the >>> Bible to the students. NO ONE can understand it without an explanation from >> the >>> teacher or their parents. >>> >>> Ron--I agree with everything you say except that I believe that the radical >>> difference between spoken and written Arabic definitely is a serious >> problem >>> for literacy. I would not be so quick to dismiss the effect of the Arabic >>> writing system--we had a conference on this topic in Haifa in May and the >>> general consensus of people who sounded like they knew what they were >> talking >>> about was that it is a significant problem-- but they hadn't done >>> cross-linguistic research and this isn't something I know enough about to >> have >>> strong opinions one way or the other. Also, the usage of written colloquial >>> Arabic is basically universal among all Israeli Arabs under the age of 25 >> who >>> have a cellular phone. >>> This is definitely not just a privileged section of the population, it's >> most >>> people of the relevant age group. >>> >>> John >>> >>> >>> >>> >>> >>> Quoting "A. Katz" : >>> >>>> John, >>>> >>>> These people you speak of are not Arabs. Some of them are Moslem and they >>>> read the Quran in the original. Some of them are not Moslem. All of them >>>> speak a local dialect of Arabic. Ask them sometimes if they think they are >>>> Arabs. >>>> >>>> Trying to turn every dialect into a separate language with a separate >>>> writing system is a way to try to disunite people. But a common language, >>>> however differently it is pronounced, unites disparate people. Australians >>>> and Cockneys and Indians and Americans speak sometimes mutually >>>> unintelligible versions of English. Using the same writing system and >>>> the same classic texts unites them. >>>> >>>> Instead of telling people they should magnify every difference, why not >>>> offer to share your language with them? Hebrew could be a uniting factor >>>> if spoken in all Israeli schools. >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >>>> >>>> >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>> >>>>> Learning the classical language is like pulling teeth for Arabs kids >> also. >>>> They >>>>> just can't publicly say it because that would make them bad Arabs. It's >>>> just >>>>> ridiculous. I have an even better plan for the Jewish kids--teach them to >>>> use >>>>> the written version of the spoken language which Arab kids are using for >>>>> Facebook. That's how kids make friends these days anyway. >>>>> Best wishes, >>>>> John >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Quoting Tom Givon : >>>>> >>>>>> >>>>>> >>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, our >>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>>>> Classical (after 1949 things changed...). That book may still exist, >>>>>> you might track it down. It was easy, a cinch really. I concluded it was >>>>>> really just Hebrew with a few trivial transformations in Phonology & >>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I sat >>>>>> on a few sessions of a faculty study group at UO who were trying to >>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. Boy, >>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>>>> well, in 1939 (or was it 1943?). >>>>>> Cheers, TG >>>>>> >>>>>> ============== >>>>>> >>>>>> >>>>>> john at research.haifa.ac.il wrote: >>>>>>> Tom, >>>>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>>>> seriously >>>>>>> consider the possibility of educating Israeli Arabs in a written >> version >>>> of >>>>>>> their spoken language, as you suggest, at least through 3rd grade. I >> can >>>>>> send >>>>>>> you some things I've written on this topic, with a lot of >>>> cross-linguistic >>>>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>>>> situation' >>>>>>> here, politically active Arabs think that the solution is to pursue the >>>>>> same >>>>>>> self-destructive strategy which they've been following for almost a >>>> century >>>>>> but >>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame >> at >>>>>> the >>>>>>> Arab media because they do not 'provide the linguistic richness of >> formal >>>>>>> Arabic') while Jews are basically content to let the Arabs stew in >> their >>>>>> own >>>>>>> juices so that they can reap the benefits. There may, however, be hope >>>> for >>>>>> the >>>>>>> future in the form of the radically increased usage of written forms of >>>>>>> colloquial Arabic dialects in electronic media such as Facebook, blogs, >>>>>> emails, >>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>>>> seriously >>>>>>> threaten the status of classical Arabic in the same way that the >>>> invention >>>>>> of >>>>>>> the printing press overturned the linguistic hierarchy in Western >> Europe >>>> by >>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>>>> >>>>>>> I realize that Masalha as well as the researchers he referred to don't >>>>>>> necessarily know what they're talking about regarding the brain. But >> the >>>>>>> general point is still potentially significant--that the connected >> script >>>>>> which >>>>>>> Arabic uses as well as the multiple forms which many Arabic letters >> have >>>>>> may be >>>>>>> a significant obstacle to literacy, however this may be related to >>>>>> processes in >>>>>>> the brain. >>>>>>> Best wishes, >>>>>>> John >>>>>>> >>>>>>> >>>>>>> >>>>>>> Quoting Tom Givon : >>>>>>> >>>>>>> >>>>>>>> A GREAT POET CAN STILL GET IT >>>> WRONG >>>>>>>> >>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) accidentally >>>>>>>> by stumbling a on his truly great quote: >>>>>>>> "All fixed identities are imposed from the outside. >>>>>>>> Whoever has a clear identity knows it can assume >>>>>>>> multiple forms". >>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>>>> intended--surely got only one third of the story right. At first, he >> was >>>>>>>> led astray be the academic researchers he cited, who claimed that the >>>>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>>>> lagging R-hemisphere activity, then explained this neurological lag >>>>>>>> by suggesting that the Arab script requires more contextual analysis. >>>>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>>>> oriented, less automated. If Arab-reading students required more >>>>>>>> contextual labor, it should have been registered as a higher >>>>>>>> R-hemisphere activity, not lower. >>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>>>> among many others, have established beyond reasonable doubt that >>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>>>> module on the boundary of the occipital and temporal lobes (just past >>>>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>>>> 'stream' that flows from the back to the front of the L-cortex. And >>>>>>>> the L-cortex is in general responsible for the more automated--less >>>>>>>> context-dependent--processing of language (as well as visual, motor >> and >>>>>>>> other skills). The visual word-recognition module is, in turn, >>>>>>>> recruited from the pre-existing visual object-recognition >>>>>>>> ventral-stream module. A considerable amount of life-time practice >> and >>>>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>>>> human brain is not (yet) genetically configured at birth for >>>>>>>> visual-word recognition, only for visual-object recognition. A similar >>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene & >>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>>>> object-location in space. >>>>>>>> Mr. Masalha then, on his own, points out to a more >>>>>>>> plausible right answer: Arab students, in Israel as well as all over >> the >>>>>>>> Arab world, are not taught literacy in their native language >>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>>>> language. The discrepancy would be just as great if Israeli kids >> were >>>>>>>> taught their Hebrew literacy first in the language of Genesis; or if >>>>>>>> French students were taught literacy first in the language of La >>>>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>>>>>> English-speaking kids in the language of Beowolf. As far as my frail >>>>>>>> guessing powers go, remedying the situation would be much easier by >>>>>>>> combining two well-known verities of second language acquisition: (a) >>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And >> (b), >>>>>>>> teach literacy first in the student's spoken native language; only >> then >>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy the >>>>>>>> way, was suggested in the late 1930's by no other than L. >> Bloomfield, >>>>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>>>> literacy in their own--equally glorious--native language, just teach >>>>>>>> them smart. >>>>>>>> For his last culprit, the presumed--tho hardly >>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>>>> recognized prejudices of the educated classes. While readily >> endorsing >>>>>>>> his aesthetic sentiments about modern media, I would still like to >> point >>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>>>> spoken native language, and if anything should facilitate the easier >>>>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>>>> ma'-salaam, >>>>>>>> >>>>>>>> >>>>>>>> T. Giv?n >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >> ------------------------------------------------------------------------ >>>>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >>>>>>> >>>>>> >>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From john at research.haifa.ac.il Thu Sep 30 16:26:14 2010 From: john at research.haifa.ac.il (john at research.haifa.ac.il) Date: Thu, 30 Sep 2010 18:26:14 +0200 Subject: A poet slips In-Reply-To: Message-ID: Stop thinking like an American and thinking that other places should be like the wonderful United States. No one here thinks the Kindergarten situation is sad except for American tourists. On the other hand, the situation of Israeli Arabs is sad, and there is pretty widespread recognition of this here. Richard--yes, this situation isn't so unusual. Russians Israelis say that they aren't thought of as Israelis here or as Russians in Russia. John Quoting "A. Katz" : > John, > > Thanks for sharing the story about the Egyptian national museum. It's > ironic that this would happen there, of all places, because ethnically > most Egyptians are not Arab, either. > > Yes, it is complicated. Everybody wants to belong somewhere, and the > reason this particular group of people are denied this opportunity is > through a weird fluke of history. To Arabs, they are Judeans or > Palestinians. To Jews, they are Arabs. > > The story about the Kindergarten is very sad. Why does religion have to be > part of the curriculum in a public school? Separation of Church and State > would go a long way toward integration and ultimately peace. > > --Aya > > > But, I agree, it is complicated. > > On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > > > Yes, actually I thought more and there is some truth to what you said. It > seems > > that at a psychological level Arabs in other countries don't consider them > to > > be 'real Arabs'. A student of mine who went to the Egyptian national museum > in > > Cairo saw that there were radically different entry fees posted for 'Arabs' > and > > 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in > fact > > he was a Druze who had served in the Israeli Army, but they had no way of > > knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker > asked > > to see his passport and when he saw that > > it was from Israel informed him that he was not an Arab and would have to > pay > > the higher 'non-Arab' fee. I have told this story to many of my Israeli > Arab > > students and not a single one has even expressed surprise--they completely > > expect such treatment. When I suggest that this means that they are not > > regarded as being real Arabs by Arabs from other countries, they admit that > > this seems to be true--but they still regard themselves as being Arabs. I > then > > sum up the situation by saying that the people called 'Israeli Arabs' are > not > > considered to be Israelis by Israelis and they are not considered to be > Arabs > > by Arabs. > > It's complicated... > > > > With regard to the obstacles to integration, it really is true that neither > side > > wants integration if this threatens their distinctive identities. In no > sense > > can it be said that the Jewish side rejects assimilation more than the Arab > > side. I remember one class in which a Christian Arab woman reported > temporarily > > sending her child to a Jewish kindergarten because it was the most > convenient > > one--but when the child came home one day and insisted on lighting candles > on > > Friday evening, the parents realized that this was too much and they needed > > another solution. Everyone in the class, Jews and Arabs, recognized > > this--except for an American visiting scholar who said 'What's the problem, > > it's cute!'. Everyone looked at him like he'd landed from outer space. The > > difference is more that the Arab side wants EQUALITY, but (contrary to what > > Americans generally think) that isn't the same thing as integration. > > > > John > > > > > > > > Quoting "A. Katz" : > > > >> John, > >> > >> Well, at least I'm glad you realize that one hundred years ago, and > >> possibly even more recently, they did not think of themselves as Arabs. > >> I still wonder whether among themselves they say "Arabs". Isn't there > >> another word? If any of them tried to emigrate to Saudi Arabia, I doubt > >> very much they'd be called "Arabs." For that matter, what do Jordanians > >> call non-Jewish Israelis? > >> > >> The biggest obstacle to allowing for integration in the schools is not the > >> Palestinian population -- it's those among the Israelis who adopt a "Jews > >> first" attitude. > >> > >> --Aya > >> > >> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz > >> > >> > >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >> > >>> Aya, > >>> I ask them constantly if they think they are Arabs (60% of my students > >>> are Arabic speakers and this is a central topic of many classes that I > >> teach) > >>> and almost without exception they do. The only exceptions are some Druze > >>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic > >>> speakers living in this area did not consider themselves to be Arabs > >> (except > >>> for the Bedouins), but the situation has completely changed. > >>> > >>> It is totally unrealistic to put Israeli Arabs as a group into > >> Hebrew-speaking > >>> schools. No one wants it, not the Jews and not the Arabs. There are a > tiny > >>> number of Arabs who for one reason or another send their children to > >>> Hebrew-speaking schools (for example the author Sayed Kashua) but this > >>> is insignificant. > >>> > >>> I don't know what the situation was when you were a child, and I don't > know > >>> about the situation among religious Jews, but secular Jewish children do > >> not > >>> learn to read by reading Genesis today. They learn to read with secular > >> texts > >>> in 1st grade and go on to the Bible in 2nd grade, but far from learning > to > >>> read by reading the Bible, teachers have to explain what's written in the > >>> Bible to the students. NO ONE can understand it without an explanation > from > >> the > >>> teacher or their parents. > >>> > >>> Ron--I agree with everything you say except that I believe that the > radical > >>> difference between spoken and written Arabic definitely is a serious > >> problem > >>> for literacy. I would not be so quick to dismiss the effect of the Arabic > >>> writing system--we had a conference on this topic in Haifa in May and the > >>> general consensus of people who sounded like they knew what they were > >> talking > >>> about was that it is a significant problem-- but they hadn't done > >>> cross-linguistic research and this isn't something I know enough about to > >> have > >>> strong opinions one way or the other. Also, the usage of written > colloquial > >>> Arabic is basically universal among all Israeli Arabs under the age of 25 > >> who > >>> have a cellular phone. > >>> This is definitely not just a privileged section of the population, it's > >> most > >>> people of the relevant age group. > >>> > >>> John > >>> > >>> > >>> > >>> > >>> > >>> Quoting "A. Katz" : > >>> > >>>> John, > >>>> > >>>> These people you speak of are not Arabs. Some of them are Moslem and > they > >>>> read the Quran in the original. Some of them are not Moslem. All of them > >>>> speak a local dialect of Arabic. Ask them sometimes if they think they > are > >>>> Arabs. > >>>> > >>>> Trying to turn every dialect into a separate language with a separate > >>>> writing system is a way to try to disunite people. But a common > language, > >>>> however differently it is pronounced, unites disparate people. > Australians > >>>> and Cockneys and Indians and Americans speak sometimes mutually > >>>> unintelligible versions of English. Using the same writing system and > >>>> the same classic texts unites them. > >>>> > >>>> Instead of telling people they should magnify every difference, why not > >>>> offer to share your language with them? Hebrew could be a uniting factor > >>>> if spoken in all Israeli schools. > >>>> > >>>> --Aya > >>>> > >>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation > >>>> > >>>> > >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > >>>> > >>>>> Learning the classical language is like pulling teeth for Arabs kids > >> also. > >>>> They > >>>>> just can't publicly say it because that would make them bad Arabs. It's > >>>> just > >>>>> ridiculous. I have an even better plan for the Jewish kids--teach them > to > >>>> use > >>>>> the written version of the spoken language which Arab kids are using > for > >>>>> Facebook. That's how kids make friends these days anyway. > >>>>> Best wishes, > >>>>> John > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> Quoting Tom Givon : > >>>>> > >>>>>> > >>>>>> > >>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, > our > >>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to > >>>>>> Classical (after 1949 things changed...). That book may still exist, > >>>>>> you might track it down. It was easy, a cinch really. I concluded it > was > >>>>>> really just Hebrew with a few trivial transformations in Phonology & > >>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I > sat > >>>>>> on a few sessions of a faculty study group at UO who were trying to > >>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. > Boy, > >>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and > >>>>>> well, in 1939 (or was it 1943?). > >>>>>> Cheers, TG > >>>>>> > >>>>>> ============== > >>>>>> > >>>>>> > >>>>>> john at research.haifa.ac.il wrote: > >>>>>>> Tom, > >>>>>>> I've been trying for several years to get Israeli Arabs and Jews to > >>>>>> seriously > >>>>>>> consider the possibility of educating Israeli Arabs in a written > >> version > >>>> of > >>>>>>> their spoken language, as you suggest, at least through 3rd grade. I > >> can > >>>>>> send > >>>>>>> you some things I've written on this topic, with a lot of > >>>> cross-linguistic > >>>>>>> data. But thus far, it isn't working. As with most aspects of 'the > >>>>>> situation' > >>>>>>> here, politically active Arabs think that the solution is to pursue > the > >>>>>> same > >>>>>>> self-destructive strategy which they've been following for almost a > >>>> century > >>>>>> but > >>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame > >> at > >>>>>> the > >>>>>>> Arab media because they do not 'provide the linguistic richness of > >> formal > >>>>>>> Arabic') while Jews are basically content to let the Arabs stew in > >> their > >>>>>> own > >>>>>>> juices so that they can reap the benefits. There may, however, be > hope > >>>> for > >>>>>> the > >>>>>>> future in the form of the radically increased usage of written forms > of > >>>>>>> colloquial Arabic dialects in electronic media such as Facebook, > blogs, > >>>>>> emails, > >>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to > >>>>>> seriously > >>>>>>> threaten the status of classical Arabic in the same way that the > >>>> invention > >>>>>> of > >>>>>>> the printing press overturned the linguistic hierarchy in Western > >> Europe > >>>> by > >>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. > >>>>>>> > >>>>>>> I realize that Masalha as well as the researchers he referred to > don't > >>>>>>> necessarily know what they're talking about regarding the brain. But > >> the > >>>>>>> general point is still potentially significant--that the connected > >> script > >>>>>> which > >>>>>>> Arabic uses as well as the multiple forms which many Arabic letters > >> have > >>>>>> may be > >>>>>>> a significant obstacle to literacy, however this may be related to > >>>>>> processes in > >>>>>>> the brain. > >>>>>>> Best wishes, > >>>>>>> John > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> Quoting Tom Givon : > >>>>>>> > >>>>>>> > >>>>>>>> A GREAT POET CAN STILL GET IT > >>>> WRONG > >>>>>>>> > >>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", > >>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) > accidentally > >>>>>>>> by stumbling a on his truly great quote: > >>>>>>>> "All fixed identities are imposed from the outside. > >>>>>>>> Whoever has a clear identity knows it can assume > >>>>>>>> multiple forms". > >>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare > >>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on > >>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun > >>>>>>>> intended--surely got only one third of the story right. At first, he > >> was > >>>>>>>> led astray be the academic researchers he cited, who claimed that > the > >>>>>>>> lagging reading skills of Israeli-Arab students is correlated to > >>>>>>>> lagging R-hemisphere activity, then explained this neurological > lag > >>>>>>>> by suggesting that the Arab script requires more contextual > analysis. > >>>>>>>> But it is the R-hemisphere of the human cortex that is more context > >>>>>>>> oriented, less automated. If Arab-reading students required more > >>>>>>>> contextual labor, it should have been registered as a higher > >>>>>>>> R-hemisphere activity, not lower. > >>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, > >>>>>>>> among many others, have established beyond reasonable doubt > that > >>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, > >>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere > >>>>>>>> module on the boundary of the occipital and temporal lobes (just > past > >>>>>>>> Brodman's Area 19), along the ventral visual object-recognition > >>>>>>>> 'stream' that flows from the back to the front of the L-cortex. > And > >>>>>>>> the L-cortex is in general responsible for the more automated--less > >>>>>>>> context-dependent--processing of language (as well as visual, motor > >> and > >>>>>>>> other skills). The visual word-recognition module is, in turn, > >>>>>>>> recruited from the pre-existing visual object-recognition > >>>>>>>> ventral-stream module. A considerable amount of life-time practice > >> and > >>>>>>>> repetition is required to affect this late-cultural adaptation. The > >>>>>>>> human brain is not (yet) genetically configured at birth for > >>>>>>>> visual-word recognition, only for visual-object recognition. A > similar > >>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene > & > >>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) > >>>>>>>> in the L-pareita lobe, an area originally configured for analysis of > >>>>>>>> object-location in space. > >>>>>>>> Mr. Masalha then, on his own, points out to a more > >>>>>>>> plausible right answer: Arab students, in Israel as well as all over > >> the > >>>>>>>> Arab world, are not taught literacy in their native language > >>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary > >>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign > >>>>>>>> language. The discrepancy would be just as great if Israeli kids > >> were > >>>>>>>> taught their Hebrew literacy first in the language of Genesis; or > if > >>>>>>>> French students were taught literacy first in the language of La > >>>>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or > >>>>>>>> English-speaking kids in the language of Beowolf. As far as my > frail > >>>>>>>> guessing powers go, remedying the situation would be much easier by > >>>>>>>> combining two well-known verities of second language acquisition: > (a) > >>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And > >> (b), > >>>>>>>> teach literacy first in the student's spoken native language; only > >> then > >>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy > the > >>>>>>>> way, was suggested in the late 1930's by no other than L. > >> Bloomfield, > >>>>>>>> in a book outlining a 'phonics-first ' literacy program for native > >>>>>>>> English speakers. Rather than depriving Israeli-Arab students of > >>>>>>>> literacy in their own--equally glorious--native language, just teach > >>>>>>>> them smart. > >>>>>>>> For his last culprit, the presumed--tho hardly > >>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well > >>>>>>>> recognized prejudices of the educated classes. While readily > >> endorsing > >>>>>>>> his aesthetic sentiments about modern media, I would still like to > >> point > >>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' > >>>>>>>> spoken native language, and if anything should facilitate the > easier > >>>>>>>> initial acquisition of native-language literacy. Respectuosamente, > >>>>>>>> ma'-salaam, > >>>>>>>> > >>>>>>>> > >>>>>>>> T. Giv?n > >>>>>>>> > >>>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >>>>>>> > >> ------------------------------------------------------------------------ > >>>>>>> This message was sent using IMP, the Webmail Program of Haifa > >> University > >>>>>>> > >>>>>> > >>>>>> > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> > ------------------------------------------------------------------------ > >>>>> This message was sent using IMP, the Webmail Program of Haifa > University > >>>>> > >>>>> > >>> > >>> > >>> > >>> > >>> ------------------------------------------------------------------------ > >>> This message was sent using IMP, the Webmail Program of Haifa University > >>> > >>> > > > > > > > > > > ------------------------------------------------------------------------ > > This message was sent using IMP, the Webmail Program of Haifa University > > > > ------------------------------------------------------------------------ This message was sent using IMP, the Webmail Program of Haifa University From amnfn at well.com Thu Sep 30 16:50:12 2010 From: amnfn at well.com (A. Katz) Date: Thu, 30 Sep 2010 09:50:12 -0700 Subject: A poet slips In-Reply-To: <1285863974.4ca4ba269d9ed@webmail.haifa.ac.il> Message-ID: John, You may think of me as an American, and correctly so, but the reason I am an American is that my father left Israel because separation of Church and State was not the law. Many Israelis whose ancestors came from the European diaspora are also not members of the Jewish religion and do not want that forced on their children. I know of one family that left because the State was trying to force them to circumcize their son. Their children are no longer Israeli. But they could have been! Wanting freedom of religion is not an exclusively American desire. It's universal. Your Kindergarten example showed that children do assimilate easily in the public schools, but that religion is a touchy subject for everybody. Even if you don't care about secular Israelis, please keep in mind that not all Palestinians who speak Arabic are Moslem. There is already a split between religious Islamic fundamentalists and others among the Palestinians, and many lives have been lost because of religious issues. The solution to the problem has to be secular. --Aya http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: > Stop thinking like an American and thinking that other places should be like the > wonderful United States. No one here thinks the Kindergarten situation is sad > except for American tourists. > > On the other hand, the situation of Israeli Arabs is sad, and there is pretty > widespread recognition of this here. > > Richard--yes, this situation isn't so unusual. Russians Israelis say that they > aren't thought of as Israelis here or as Russians in Russia. > John > > > Quoting "A. Katz" : > >> John, >> >> Thanks for sharing the story about the Egyptian national museum. It's >> ironic that this would happen there, of all places, because ethnically >> most Egyptians are not Arab, either. >> >> Yes, it is complicated. Everybody wants to belong somewhere, and the >> reason this particular group of people are denied this opportunity is >> through a weird fluke of history. To Arabs, they are Judeans or >> Palestinians. To Jews, they are Arabs. >> >> The story about the Kindergarten is very sad. Why does religion have to be >> part of the curriculum in a public school? Separation of Church and State >> would go a long way toward integration and ultimately peace. >> >> --Aya >> >> >> But, I agree, it is complicated. >> >> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >> >>> Yes, actually I thought more and there is some truth to what you said. It >> seems >>> that at a psychological level Arabs in other countries don't consider them >> to >>> be 'real Arabs'. A student of mine who went to the Egyptian national museum >> in >>> Cairo saw that there were radically different entry fees posted for 'Arabs' >> and >>> 'non-Arabs'. He explained in his Levantine accent that he was an Arab (in >> fact >>> he was a Druze who had served in the Israeli Army, but they had no way of >>> knowing that) and attempted to pay the lower 'Arab' fee. The tickettaker >> asked >>> to see his passport and when he saw that >>> it was from Israel informed him that he was not an Arab and would have to >> pay >>> the higher 'non-Arab' fee. I have told this story to many of my Israeli >> Arab >>> students and not a single one has even expressed surprise--they completely >>> expect such treatment. When I suggest that this means that they are not >>> regarded as being real Arabs by Arabs from other countries, they admit that >>> this seems to be true--but they still regard themselves as being Arabs. I >> then >>> sum up the situation by saying that the people called 'Israeli Arabs' are >> not >>> considered to be Israelis by Israelis and they are not considered to be >> Arabs >>> by Arabs. >>> It's complicated... >>> >>> With regard to the obstacles to integration, it really is true that neither >> side >>> wants integration if this threatens their distinctive identities. In no >> sense >>> can it be said that the Jewish side rejects assimilation more than the Arab >>> side. I remember one class in which a Christian Arab woman reported >> temporarily >>> sending her child to a Jewish kindergarten because it was the most >> convenient >>> one--but when the child came home one day and insisted on lighting candles >> on >>> Friday evening, the parents realized that this was too much and they needed >>> another solution. Everyone in the class, Jews and Arabs, recognized >>> this--except for an American visiting scholar who said 'What's the problem, >>> it's cute!'. Everyone looked at him like he'd landed from outer space. The >>> difference is more that the Arab side wants EQUALITY, but (contrary to what >>> Americans generally think) that isn't the same thing as integration. >>> >>> John >>> >>> >>> >>> Quoting "A. Katz" : >>> >>>> John, >>>> >>>> Well, at least I'm glad you realize that one hundred years ago, and >>>> possibly even more recently, they did not think of themselves as Arabs. >>>> I still wonder whether among themselves they say "Arabs". Isn't there >>>> another word? If any of them tried to emigrate to Saudi Arabia, I doubt >>>> very much they'd be called "Arabs." For that matter, what do Jordanians >>>> call non-Jewish Israelis? >>>> >>>> The biggest obstacle to allowing for integration in the schools is not the >>>> Palestinian population -- it's those among the Israelis who adopt a "Jews >>>> first" attitude. >>>> >>>> --Aya >>>> >>>> http://hubpages.com/hub/My-Grandfathers-Voice-Recordings-of-Benzion-Katz >>>> >>>> >>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>> >>>>> Aya, >>>>> I ask them constantly if they think they are Arabs (60% of my students >>>>> are Arabic speakers and this is a central topic of many classes that I >>>> teach) >>>>> and almost without exception they do. The only exceptions are some Druze >>>>> (mostly males) and most Maronites. I am aware that 100 years ago Arabic >>>>> speakers living in this area did not consider themselves to be Arabs >>>> (except >>>>> for the Bedouins), but the situation has completely changed. >>>>> >>>>> It is totally unrealistic to put Israeli Arabs as a group into >>>> Hebrew-speaking >>>>> schools. No one wants it, not the Jews and not the Arabs. There are a >> tiny >>>>> number of Arabs who for one reason or another send their children to >>>>> Hebrew-speaking schools (for example the author Sayed Kashua) but this >>>>> is insignificant. >>>>> >>>>> I don't know what the situation was when you were a child, and I don't >> know >>>>> about the situation among religious Jews, but secular Jewish children do >>>> not >>>>> learn to read by reading Genesis today. They learn to read with secular >>>> texts >>>>> in 1st grade and go on to the Bible in 2nd grade, but far from learning >> to >>>>> read by reading the Bible, teachers have to explain what's written in the >>>>> Bible to the students. NO ONE can understand it without an explanation >> from >>>> the >>>>> teacher or their parents. >>>>> >>>>> Ron--I agree with everything you say except that I believe that the >> radical >>>>> difference between spoken and written Arabic definitely is a serious >>>> problem >>>>> for literacy. I would not be so quick to dismiss the effect of the Arabic >>>>> writing system--we had a conference on this topic in Haifa in May and the >>>>> general consensus of people who sounded like they knew what they were >>>> talking >>>>> about was that it is a significant problem-- but they hadn't done >>>>> cross-linguistic research and this isn't something I know enough about to >>>> have >>>>> strong opinions one way or the other. Also, the usage of written >> colloquial >>>>> Arabic is basically universal among all Israeli Arabs under the age of 25 >>>> who >>>>> have a cellular phone. >>>>> This is definitely not just a privileged section of the population, it's >>>> most >>>>> people of the relevant age group. >>>>> >>>>> John >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> Quoting "A. Katz" : >>>>> >>>>>> John, >>>>>> >>>>>> These people you speak of are not Arabs. Some of them are Moslem and >> they >>>>>> read the Quran in the original. Some of them are not Moslem. All of them >>>>>> speak a local dialect of Arabic. Ask them sometimes if they think they >> are >>>>>> Arabs. >>>>>> >>>>>> Trying to turn every dialect into a separate language with a separate >>>>>> writing system is a way to try to disunite people. But a common >> language, >>>>>> however differently it is pronounced, unites disparate people. >> Australians >>>>>> and Cockneys and Indians and Americans speak sometimes mutually >>>>>> unintelligible versions of English. Using the same writing system and >>>>>> the same classic texts unites them. >>>>>> >>>>>> Instead of telling people they should magnify every difference, why not >>>>>> offer to share your language with them? Hebrew could be a uniting factor >>>>>> if spoken in all Israeli schools. >>>>>> >>>>>> --Aya >>>>>> >>>>>> http://hubpages.com/hub/ISRAEL-The-Two-Halves-of-the-Nation >>>>>> >>>>>> >>>>>> On Thu, 30 Sep 2010, john at research.haifa.ac.il wrote: >>>>>> >>>>>>> Learning the classical language is like pulling teeth for Arabs kids >>>> also. >>>>>> They >>>>>>> just can't publicly say it because that would make them bad Arabs. It's >>>>>> just >>>>>>> ridiculous. I have an even better plan for the Jewish kids--teach them >> to >>>>>> use >>>>>>> the written version of the spoken language which Arab kids are using >> for >>>>>>> Facebook. That's how kids make friends these days anyway. >>>>>>> Best wishes, >>>>>>> John >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> Quoting Tom Givon : >>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> When I learned to read Arabic on the kibbutz (Maabarot) as a child, >> our >>>>>>>> textbook was of written COLLOQUIAL Falastini Arabic. We never got to >>>>>>>> Classical (after 1949 things changed...). That book may still exist, >>>>>>>> you might track it down. It was easy, a cinch really. I concluded it >> was >>>>>>>> really just Hebrew with a few trivial transformations in Phonology & >>>>>>>> Grammar. (I was 7 years old & a bit naive then). Then 7 years ago I >> sat >>>>>>>> on a few sessions of a faculty study group at UO who were trying to >>>>>>>> learn Arabic (post 9/11...)--from a Classical Koranic grammar book. >> Boy, >>>>>>>> it was like pullin' teeth. But Leonard Bloomfield said it already, and >>>>>>>> well, in 1939 (or was it 1943?). >>>>>>>> Cheers, TG >>>>>>>> >>>>>>>> ============== >>>>>>>> >>>>>>>> >>>>>>>> john at research.haifa.ac.il wrote: >>>>>>>>> Tom, >>>>>>>>> I've been trying for several years to get Israeli Arabs and Jews to >>>>>>>> seriously >>>>>>>>> consider the possibility of educating Israeli Arabs in a written >>>> version >>>>>> of >>>>>>>>> their spoken language, as you suggest, at least through 3rd grade. I >>>> can >>>>>>>> send >>>>>>>>> you some things I've written on this topic, with a lot of >>>>>> cross-linguistic >>>>>>>>> data. But thus far, it isn't working. As with most aspects of 'the >>>>>>>> situation' >>>>>>>>> here, politically active Arabs think that the solution is to pursue >> the >>>>>>>> same >>>>>>>>> self-destructive strategy which they've been following for almost a >>>>>> century >>>>>>>> but >>>>>>>>> with even more vigor and steadfastness (e.g. Masalha points the blame >>>> at >>>>>>>> the >>>>>>>>> Arab media because they do not 'provide the linguistic richness of >>>> formal >>>>>>>>> Arabic') while Jews are basically content to let the Arabs stew in >>>> their >>>>>>>> own >>>>>>>>> juices so that they can reap the benefits. There may, however, be >> hope >>>>>> for >>>>>>>> the >>>>>>>>> future in the form of the radically increased usage of written forms >> of >>>>>>>>> colloquial Arabic dialects in electronic media such as Facebook, >> blogs, >>>>>>>> emails, >>>>>>>>> etc., by Arabic speakers below the age of 30, which will soon come to >>>>>>>> seriously >>>>>>>>> threaten the status of classical Arabic in the same way that the >>>>>> invention >>>>>>>> of >>>>>>>>> the printing press overturned the linguistic hierarchy in Western >>>> Europe >>>>>> by >>>>>>>>> overthrowing Latin. I'm working on this too. We'll see what happens. >>>>>>>>> >>>>>>>>> I realize that Masalha as well as the researchers he referred to >> don't >>>>>>>>> necessarily know what they're talking about regarding the brain. But >>>> the >>>>>>>>> general point is still potentially significant--that the connected >>>> script >>>>>>>> which >>>>>>>>> Arabic uses as well as the multiple forms which many Arabic letters >>>> have >>>>>>>> may be >>>>>>>>> a significant obstacle to literacy, however this may be related to >>>>>>>> processes in >>>>>>>>> the brain. >>>>>>>>> Best wishes, >>>>>>>>> John >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> Quoting Tom Givon : >>>>>>>>> >>>>>>>>> >>>>>>>>>> A GREAT POET CAN STILL GET IT >>>>>> WRONG >>>>>>>>>> >>>>>>>>>> I came to know of Salman Masalha ("Arabs, speak Hebrew!", >>>>>>>>>> Haaretz/English, International Herald Tribune 9-27-10) >> accidentally >>>>>>>>>> by stumbling a on his truly great quote: >>>>>>>>>> "All fixed identities are imposed from the outside. >>>>>>>>>> Whoever has a clear identity knows it can assume >>>>>>>>>> multiple forms". >>>>>>>>>> In the context of Palestine/Israel, what a breath of fresh, rare >>>>>>>>>> clarity. Still, like the rest of us mortals, a great poet can on >>>>>>>>>> occasion get it wrong too, and Mr. Masalha--may he be forgiven; pun >>>>>>>>>> intended--surely got only one third of the story right. At first, he >>>> was >>>>>>>>>> led astray be the academic researchers he cited, who claimed that >> the >>>>>>>>>> lagging reading skills of Israeli-Arab students is correlated to >>>>>>>>>> lagging R-hemisphere activity, then explained this neurological >> lag >>>>>>>>>> by suggesting that the Arab script requires more contextual >> analysis. >>>>>>>>>> But it is the R-hemisphere of the human cortex that is more context >>>>>>>>>> oriented, less automated. If Arab-reading students required more >>>>>>>>>> contextual labor, it should have been registered as a higher >>>>>>>>>> R-hemisphere activity, not lower. >>>>>>>>>> Works by M. Posner, S. Petersen, M. Raichle and S. Dahane, >>>>>>>>>> among many others, have established beyond reasonable doubt >> that >>>>>>>>>> written words in all languages (English, Mandarin, Hindi, Amharic, >>>>>>>>>> Hebrew, Arabic, etc.) are decoded automatically in an L-hemisphere >>>>>>>>>> module on the boundary of the occipital and temporal lobes (just >> past >>>>>>>>>> Brodman's Area 19), along the ventral visual object-recognition >>>>>>>>>> 'stream' that flows from the back to the front of the L-cortex. >> And >>>>>>>>>> the L-cortex is in general responsible for the more automated--less >>>>>>>>>> context-dependent--processing of language (as well as visual, motor >>>> and >>>>>>>>>> other skills). The visual word-recognition module is, in turn, >>>>>>>>>> recruited from the pre-existing visual object-recognition >>>>>>>>>> ventral-stream module. A considerable amount of life-time practice >>>> and >>>>>>>>>> repetition is required to affect this late-cultural adaptation. The >>>>>>>>>> human brain is not (yet) genetically configured at birth for >>>>>>>>>> visual-word recognition, only for visual-object recognition. A >> similar >>>>>>>>>> cultural adaptation, this one for math, has been shown for (Dahaene >> & >>>>>>>>>> Cohen, 2007; see recent article in The New Yorker by Oliver Sachs) >>>>>>>>>> in the L-pareita lobe, an area originally configured for analysis of >>>>>>>>>> object-location in space. >>>>>>>>>> Mr. Masalha then, on his own, points out to a more >>>>>>>>>> plausible right answer: Arab students, in Israel as well as all over >>>> the >>>>>>>>>> Arab world, are not taught literacy in their native language >>>>>>>>>> (Falastini, Maghrebi, Masri, Yemeni, etc.), but in a frozen literary >>>>>>>>>> instrument harking back 1,400 years or more. That is, in a foreign >>>>>>>>>> language. The discrepancy would be just as great if Israeli kids >>>> were >>>>>>>>>> taught their Hebrew literacy first in the language of Genesis; or >> if >>>>>>>>>> French students were taught literacy first in the language of La >>>>>>>>>> Chan?on de Roland, Guilhome de Machaut, or Chr? tien de Troyes. Or >>>>>>>>>> English-speaking kids in the language of Beowolf. As far as my >> frail >>>>>>>>>> guessing powers go, remedying the situation would be much easier by >>>>>>>>>> combining two well-known verities of second language acquisition: >> (a) >>>>>>>>>> Teach them both early, together--'co-ordinated bilingualism'. And >>>> (b), >>>>>>>>>> teach literacy first in the student's spoken native language; only >>>> then >>>>>>>>>> gradually 'stretch' it to more literary genres. This method, bhy >> the >>>>>>>>>> way, was suggested in the late 1930's by no other than L. >>>> Bloomfield, >>>>>>>>>> in a book outlining a 'phonics-first ' literacy program for native >>>>>>>>>> English speakers. Rather than depriving Israeli-Arab students of >>>>>>>>>> literacy in their own--equally glorious--native language, just teach >>>>>>>>>> them smart. >>>>>>>>>> For his last culprit, the presumed--tho hardly >>>>>>>>>> unique--vulgarity of Arab media, Mr. Masalha lapses into well >>>>>>>>>> recognized prejudices of the educated classes. While readily >>>> endorsing >>>>>>>>>> his aesthetic sentiments about modern media, I would still like to >>>> point >>>>>>>>>> out that the 'vulgar' genre is much closer to the Arab students' >>>>>>>>>> spoken native language, and if anything should facilitate the >> easier >>>>>>>>>> initial acquisition of native-language literacy. Respectuosamente, >>>>>>>>>> ma'-salaam, >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> T. Giv?n >>>>>>>>>> >>>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>> ------------------------------------------------------------------------ >>>>>>>>> This message was sent using IMP, the Webmail Program of Haifa >>>> University >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >>>>>>> >> ------------------------------------------------------------------------ >>>>>>> This message was sent using IMP, the Webmail Program of Haifa >> University >>>>>>> >>>>>>> >>>>> >>>>> >>>>> >>>>> >>>>> ------------------------------------------------------------------------ >>>>> This message was sent using IMP, the Webmail Program of Haifa University >>>>> >>>>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------ >>> This message was sent using IMP, the Webmail Program of Haifa University >>> >>> > > > > > ------------------------------------------------------------------------ > This message was sent using IMP, the Webmail Program of Haifa University > > From Jordan.Zlatev at ling.lu.se Thu Sep 30 22:45:56 2010 From: Jordan.Zlatev at ling.lu.se (Jordan Zlatev) Date: Fri, 1 Oct 2010 00:45:56 +0200 Subject: CfP Evolition at ICLC 11, 2011 Xian Message-ID: Call for Papers for a Theme Session at the ICLC 11, July 11-17, 2011, Xi'an, China Language Evolution: Biological, Cultural and Bio-Cultural Organizers: Arie Verhagen (Leiden University, email Arie.Verhagen at hum.leidenuniv.nl) Jordan Zlatev (Lund University, email Jordan.Zlatev at ling.lu.se) Oct. 10, 2010: Deadline for submitting presentation proposals to the session organizers (Title, and mini-abstract of at most three lines) Nov. 15, 2010: Deadline for submitting full abstracts to ICLC 11 (http://www.iclc11.org) Evolution of language is emerging as a prominent interdisciplinary field of research, bringing together linguists, biologists, psychologists, anthropologists, literary scholars, semioticians and cognitive scientists. A key challenge is to unify theories of language and cognition with the theory of biological evolution: language has evidently evolved in the human lineage, but it has some very special features that make it ?one of the most significant and interesting evolutionary events [?] during the entire history of life on Earth? (Fitch 2010: 1): it is a ?cheap?, honest, flexible, and powerful system of communication (and thought) that no other animal species appears to have and that must have required some special circumstances to have evolved. A central issue of debate is the relation between biological and cultural factors involved in the evolution of Homo Sapiens on the hand, and of modern languages on the other. It is undeniable that mechanisms and principles beyond those involved in biological evolution play a role in cultural evolution, but there is considerable room for variation and even disagreement on what these mechanisms are, how big a role each of them plays, how they interact with each other and with biological factors, and ?very fundamentally? to what extent an overall account, incorporating such mechanisms, can still be considered Darwinian. At one end, the view is found that both the biological evolution of the brain (and body) necessary for language, and the historical changes that have given rise to the 6000+ languages spoken today are explainable in terms of standard evolutionary theory, based on the dynamics of variation, selection and replication combined (e.g. Croft 2000). Others hold that even biological evolution, and much more so cultural evolution of language and other institutional practices involves principles such as ?autopoiesis? and ?qualitative emergence? that intrinsically lead to complexification (e.g. Kull 2009). Whether such processes are fundamentally different from (if somewhat analogous to) Darwinian processes is an additional issue. Yet another one is the extent to which genetic evolution per se, cultural evolution per se and/or the interaction of biological and cultural processes (among which gene-culture co-evolution and niche-construction) account for modern human cognition, language, and culture in general. Different positions and discussion of the nature and consequences of mechanisms and principles being proposed can be found in publications by researchers from different disciplinary backgrounds, e.g. Deacon (1997), Levinson (2000), Laland & Brown (2002), Odling-Smee et al. (2003), Heine and Kuteva (2007), Brier (2008), Sinha (2010). We invite contributions addressing linguistic issues, based on theoretical argumentation and empirical evidence, that contribute to this debate. Some examples (non-exhaustive!) of possible topics include: ? Construal: how does the capacity to construe the same object of conceptualization in different ways fit into an evolutionary perspective (biological and/or cultural)? ? Grammaticalization and semantic change as cultural evolution. ? Grammatical and lexical systems as adaptations, to cultural and/or biological environments. ? Co-evolution of cognition (e.g. intersubjectivity, shared intentionality), language, and culture; cultural/linguistic niche-construction. ? Co-evolution of communicative and collaborative practices and linguistic systems. - The roles of narrative, writing and symbolic artifacts in the cultural evolution of language. References Brier, S?ren (2008). Biosemiotics: Why Information is not Enough. Toronto: University of Toronto Press. Croft, William (2000), Explaining Language Change. An Evolutionary Approach. London, etc.: Longman. Deacon, Terrence W. (1997), The Symbolic Species: The Coevolution of Language and the Brain. New York: W.W. Norton. Fitch, Tecumseh W. (2010), The Evolution of Language. Cambridge: Cambridge University Press. Heine, Bernd and Tania Kuteva (2007), The Genesis of Grammar. A Reconstruction. Oxford: Oxford University Press. Kull, Kalevi, (2009) Vegetative, animal, and propositional semiosis: The semiotic threshold zones. Cognitive Semiotics #4. Laland, Kevin H. & Gillian R. Brown (2002), Sense and Nonsense. Evolutionary Perspectives on Human Behaviour. Oxford: Oxford University Press. Levinson, Stephen C. (2000), Language as nature and language as art. In: R. Hide, J. Mittelstrass, W. Singer (eds.), Changing concepts of nature at the turn of the millennium. Vatican City: Pontifical Academy of Sciences, 257?287. Odling-Smee, F.J., Laland, K.N. and Feldman, M.W.: 2003.,Niche Construction. The Neglected Process in Evolution. Princeton: Princeton University Press. Sinha, C. (2010). Language as a biocultural niche and social institution. In V. Evans, & S. Pourcel (eds.), New Directions in Cognitive Linguistics. Amsterdam: John Benjamins.