From wcmann at JUNO.COM Sat Jul 3 20:43:53 1999 From: wcmann at JUNO.COM (William Mann) Date: Sat, 3 Jul 1999 16:43:53 -0400 Subject: discourse functions of property words Message-ID: Replying to Michael Dryer on discourse and predication: Also in 1988, Sandy Thompson and I published a paper defining Rhetorical Structure Theory, Text 8(3). It is an approach to assessing whole-text discourse structure, so far defined only for monologue. It makes extensive use of discourse relations, some of which link a text span (called the nucleus) to another (called a satellite.) Some of these discourse relations are sometimes signaled by discourse particles, which would certainly be included in �other types of words� in your inquiry. In various related papers, mostly published before that one, we talked about �relational propositions,� which are predications that arise when particular relations are found to be part of the discourse structure of a text. Such a predication will arise even if the presence of the relation is not signaled. The paper has no formal taxonomy, but it does include a number of distinctions which have been taken to be taxonomic. Exclusion of a taxonomy was deliberate on our part, because we did not have a particular taxonomy that we felt happy with. The distinctions included a volitional / nonvolitional contrast among causal relations, and a contrast between so called �subject matter� relations (e.g. Conditional) and �presentational� relations (e.g. Concessive.) There were multiple grounds for identifying the Elaboration relation. Some relations (e.g. Motivation) applied only where the nucleus span presented an action. Beyond the notion of belief, there was a broader notion called �positive regard� which included belief, intention to act and approval; it was used to avoid proliferation of relations whose definitions would otherwise be identical. All of this suggests that there is plenty in RST to taxonomize, and that different sorts of worthwhile taxonomies might be produced. Several research papers using RST have been critical of the irregularity of the set of relation definitions, the set of distinctions and the absence of a taxonomy. Particularly for purposes of formalization, sometimes as a means to programming text generation programs or analyzers of text, various modifications or replacements of the set of relations have been suggested. I hope that this gives you a useful lead into this particular corner of the literature. Bill Mann On Thu, 24 Jun 1999 20:18:26 -0400 Matthew S Dryer writes: >Sandy Thompson (1988) argues that property words tend to involve the >discourse function of "predication" in natural speech. Is anybody >aware >of any attempt to develop a finer taxonomy of discourse functions >associated with property words (or other types of words) that would >distinguish different subtypes of "predication"? > >Thanks, > >Matthew Dryer . . . . ..... end of Bill's message, apologies for the ad at the bottom. ..... From kosam at LIBR.UG.EDU.GH Thu Jul 8 07:16:17 1999 From: kosam at LIBR.UG.EDU.GH (Kweku Osam) Date: Thu, 8 Jul 1999 08:16:17 +0100 Subject: Multiple Choice Questions Message-ID: The Department of Linguistics at the University of Ghana is considering using the Multiple Choice Questions formart in examining Linguistics courses at the lower levels. We would like to know if there are Departments which have tried this method of examining in Linguistics and what their experience has been. We'd also appreciate sample questions in various areas of the discipline. Send all reactions directly to me. Thanks. Kweku Osam Department of Linguistics University of Ghana Legon GHANA kosam at libr.ug.edu.gh From pwd at RICE.EDU Sat Jul 17 01:31:49 1999 From: pwd at RICE.EDU (Philip W Davis) Date: Fri, 16 Jul 1999 20:31:49 -0500 Subject: Best Grammar(s) Message-ID: Best Grammar(s) of the Twentieth Century As we approach the quadruple witching of end-of-the-year, end-of-the-decade, end-of-the-century & end-of-the-millenium, register your nomination for the 'best' grammar(s) of the twentieth century. 'Best', of course, translates directly into 'favorite'. It is a subjective reaction, but we must all have our 'best'/'favorite' grammars. *These will be grammars made public/published after January 1, 1900. But they may be dissertations that were never distributed otherwise. *These will be 'book length', meaning that (however long) they will not be a part of a collection of 'sketches', although they may be monographs in a series. They will have a separate bibliographical identity, i.e., their own ISBN. *Please provide bibliographical information: Author(s). Date. Title. Place of Publication: Publisher. *These will be 'synchronic' grammars. *They will be 'grammars'. Although phonetics & phonology are normally components of a grammar, works confined to those areas, e.g. The Sound Pattern of English, will not be eligible. *You can nominate/vote for as many grammars as you wish. *You can only nominate/vote once for a given title. *You can return and nominate/vote again if a title, which had earlier slipped your mind, occurs to you. *Yes, if you insist, you can nominate/vote for your own grammars. *You can provide reasons for your selection(s), and those comments may appear in association with the respective titles. I _especially_ encourage this. *Only positive nominations/votes will be tallied and only neutral/positive remarks, repeated. Flames will be ignored. *If you wish your comments to remain anonymous, tell me and I will attribute them to 'Anon.' *When/If there are sufficient responses to generate a list, it will be available www.ruf.rice.edu/~pwd/index.html. *You have until December 31, 1999, 11:59pm. The list, if there is one, will stay up after that. *You may nominate/vote by leaving me e-mail at pwd at rice.edu. Use the subject line: 'Grammar(s)'. +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= Philip W. Davis e-mail: pwd at rice.edu Department of Linguistics MS23 tel: (713)527-6010 Rice University fax: (713)527-4718 6100 Main St. web: www.ruf.rice.edu/~pwd/index.html Houston, TX 77005 USA +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= From tpayne at OREGON.UOREGON.EDU Sat Jul 17 16:20:19 1999 From: tpayne at OREGON.UOREGON.EDU (Tom Payne) Date: Sat, 17 Jul 1999 09:20:19 -0700 Subject: New Linguistic Olympics puzzles Message-ID: Dear Funknet You may like to know that there are now seven new "puzzles" on the Linguistic Olympics website. These are the following: Samoan, Malay/Indonesian, Maasai, Swahili #2, Tamil, Yaqui, and Classical Nahuatl. This brings the total to 21 puzzles that are available on the site. For the past six months the site has been averaging about 250 "hits" per week. I answer between 5 and 10 messages a day from individuals who attempt to solve the puzzles. Several Junior, middle and high school teachers have let me know they are using the puzzles in their classes. Others have inquired about possibly organizing a "Linguistic Olympics" event at their school. So far I am not aware of anyone who has actually done this though. The main Linguistic Olympics webpage is http://darkwing.uoregon.edu/~tpayne/lingolym/lingolym.htm. There is also a report to the LSA on the 1998 Eugene Linguistic Olympics at http://darkwing.uoregon.edu/~tpayne/lingolym/LOreport.htm. The event was held again in 1999, but with substantial changes. If there is interest (and if I have time), I would like to report on that event as well. I encourage everyone to check out the site, and give me any comments, suggestions, etc. These are puzzles geared to American teenagers who have no knowledge of linguistics. Tom Payne From lrequena at DANYSOFT.COM.AR Sat Jul 17 17:52:27 1999 From: lrequena at DANYSOFT.COM.AR (Leonel Requena) Date: Sat, 17 Jul 1999 14:52:27 -0300 Subject: Twentieth Century Message-ID: Philip W. Davis wrote (16/07/99): >Best Grammar(s) of the Twentieth Century >As we approach the quadruple witching of end-of-the-year, >end-of-the-decade, end-of-the-century & end-of-the-millenium, register your >nomination for the 'best' grammar(s) of the twentieth century. > *You have until December 31, 1999, 11:59pm. The list, if there is one, > will stay up after that You should be aware that the end-of the-year is not, at the same time, the end-of-the-decade, end-of-the-century, etc., (all this at the end-of-next-year). If the best grammar of the century appears next year, what are you going to do? ¡That's unfair! Best wishes for the end-of-the-year and for the end-of-that-year. Leonel Requena (Not in a hurry). Buenos Aires -------------- next part -------------- An HTML attachment was scrubbed... URL: From cpensa at GUGU.USAL.ES Tue Jul 20 09:37:11 1999 From: cpensa at GUGU.USAL.ES (Carmen Pensado) Date: Tue, 20 Jul 1999 11:37:11 +0200 Subject: congress announcement Message-ID: El próximo XXIII Congreso Internacional de Lingüística y Filología Románica, reunión de la Société de Linguistique Romane, se celebrará en Salamanca (España) en Setiembre del 2001. > > La primera circular ha sido puesta en marcha. Pueden hallar una versión electrónica en: > > http://web.usal.es/~romanica/romanica.htm > From oesten at ling.su.se Fri Jul 30 13:58:36 1999 From: oesten at ling.su.se (Östen Dahl) Date: Fri, 30 Jul 1999 15:58:36 +0200 Subject: What is emergence anyway? Message-ID: What is emergence anyway? Recently, the terms "emergent" and "emergence" have become popular in linguistics and neighbouring fields. Here are two representative quotations: " Hopper suggested that the study of grammaticalization tended to undermine the assumption of preexistent a priori grammatical component that stood as a prerequisite for discourse and a precondition for communication, and he proposed instead that grammar was an emergent property of texts. "Structure" would then be an epiphenomenal by-product of discourse." (Hopper 1996, 231) "If you spend some time watching the checkout lines at a supermarket, you quickly find that the number of people queued up in each line is roughly the same There is no fixed rule governing this pattern. Instead, the rule that equalizes the number of shoppers in the various lines emerges from other basic facts about the goals and behavior of shoppers and supermarket managers. This simple idea of emergence through constraint satisfaction is currently being invoked as a central explanatory mechanism in many areas of cognitive science and neuroscience. the behaviors that we tend to characterize in terms of rules and symbols are in fact emergent patterns that arise from the interactions of other less complex or more stable underlying systems. I will refer to this new viewpoint on language learning and processing as “emergentism”." (MacWhinney 1999) The terms "emergent" and "emergence" however, have been around for quite some time. In Encyclopedia Britannica's article "Emergence", the term is said to be used in evolutionary theory in the sense of "the rise of a system that cannot be predicted or explained from antecedent conditions" EB then refers to the 19th century philosopher of science George Henry Lewes' distinction between resultants and emergents: "phenomena that are predictable from their constituent parts and those that are not". An example of a resultant would be a physical mixture of sand and talcum powder, while a chemical compound such as salt, which looks nothing like sodium or chlorine, would exemplify an emergent. The EB article enumerates a number of stages in biological evolution "at which fundamentally new forms have appeared", such the orgin of life, the origin of nucleus-bearing protozoa, and the rise of sentient beings with nervous systems, and goes on: "Each of these new modes of life, though grounded in the physicochemical and biochemical conditions of the previous and simpler stage, is intelligible only in terms of its own ordering principle. These are thus cases of emergence." Another presumably authoritative source is the website of the journal Emergence: A Journal of Complexity Issues in Organizations and Management (http://emergence.org). It declares that the idea of emergence is used in the study of complex systems to "indicate the arising of patterns, structures, or properties that do not seem adequately explained by referring only to the system's pre-existing components and their interaction." and says that emergence is particularly important when "- when the organization of the system, i.e., its global order, appears to be more salient and of a different kind than the components alone; - when the components can be replaced without an accompanying decommissioning of the whole system; - when the new global patterns or properties are radically novel with respect to the pre-existing components; thus, the emergent patterns seem to be unpredictable and nondeducible from the components as well as irreducible to those components." A reader who tries to reconcile what is said about emergence in all these quotations feels rather bewildered. In the older tradition, still represented by EB and the journal Emergence, "emergence" stands for new and interesting higher-order structures that are not reducible to the lower-order ones. MacWhinney and Hopper, on the contrary, seem to use the same term precisely for the opposite: seemingly complex systems that are in fact derivable from -- "epiphenomenal by-products" of -- other simpler systems. One may ask how such a radical shift in meaning may have occurred. It seems that we can find the seed of the conflict in the original notion of emergence. On one hand, the target has new and interesting properties that cannot be described in terms of the source, on the other, there is presumably some kind of causal chain that leads from the source to the target. The essence of the notion of a self-organizing system seems to be precisely the fact that unexpected things happen as it were by themselves. Depending on whether one is more fascinated by the novel or the predictable component in this process, one may come to see different and seemingly contradictory aspects of "emergence" as criterial. Hopefully, we will eventually be able to see both sides of the phenomena at the same time. REFERENCES Hopper, Paul J. 1996. Some recent trends in grammaticalization. Annual Review of Anthropology. 25:217-36. MacWhinney, Brian. 1999. Emergent Language. In M. Darnell, E. Moravcsik, F. Newmeyer, M. Noonan, and K. Wheatley, eds. Functionalism and Formalism in Linguistics. New York: John Benjamins. ----------------------- Östen Dahl oesten at ling.su.se http://www.ling.su.se/staff/oesten/oesten.htm From W.Schulze at LRZ.UNI-MUENCHEN.DE Fri Jul 30 15:37:34 1999 From: W.Schulze at LRZ.UNI-MUENCHEN.DE (Wolfgang Schulze) Date: Fri, 30 Jul 1999 17:37:34 +0200 Subject: What is emergence anyway? Message-ID: Dear all, I think, all linguists (from the camp of functionalism/derivationalism) should be gratefull to Östen that he has raised the issue of 'emergence'. This problem with this term is - I think - just the same one as that related to other 'semi-TT' (terminus technicus) terms introduced in linguistics. 'Emergence' (together with its derivatives) clearly had its source in early 19th century 'philosophy of nature'. >>From there, it became popularized and adopted by other scientific (as well as non-scientific) domains (cf. the well-documented but hardly reflected history of the term 'pragmatics'). Hence, 'emergence' acquired a relatively strong peripheric extension with respect to its semantics which resulted in a relatively vague use of the term. In fact - I think - everybody has some notion of 'emergence', though (s)he would have difficulties to define the term when used in everyday (more or less) scientific discourse. The re-specification of this term by e.g. Paul Hopper or Brian MacWhinney for linguistics naturally retains some idiosyncratic reflexes of the early (informal) use of the term (refering esp. to its broad semantic periphery). Östen has reconstructed the prototypical (or central) semantics of 'emergence' etc. in a very convincing way to which only little is to add. NB: The Latin basis 'e-mergo:' means 'rise, come up' etc., based on 'mergo:' 'hide, cover, dive'.... (a good IE root, cf. Walde/Hoffmann LEW II:76p.). However, since, 'emergence' has a semi-TT reading in linguistics now, it would be good to always state the theoretical frame work or context in which this term is embedded. In my own frame work ('Grammar of Scenes and Scenarios (Schulze 1998, see http://www.lrz-muenchen.de/~wschulze/pkk_1abs.htm); Schulze 1999 (see http://www.lrz-muenchen.de/~wschulze/cog_typ.htm)) for instance, I claim that 'emergence' is a cognitive procedure that denotes the 'dynamic semantics' that are activated during the structural coupling of a polycentric 'system' (or parts of it). By this is meant that the interaction of source domains (organized in form of a polycentric network) establishes an activation level that is characterized by the production (or emergence) of some kind of specific behavior. This behavior can be either somehow related to the behavior/properties of one or more source domains, or totally different from the behavior/properties of the source domains. The more the behavior of the resutlting 'higher level domain' is different from that of its source domains the more likely it is that its behavior is construed as a cognitive 'reality' (that is as a idealized cognitive model) by human experience. The main point is that contrary to the source domains, 'higher level domains' emerging therefrom do not have 'properties' etc. prima facie. Only IF they are construed (or experienced) as 'domains' they are interpreted as having (autonomous) properties of their own. 'Language' - in my thinking - is a construction that refers to the experience of emergent activities of underlying, structurally coupled cognitive centers (which themselves again are partly cognitive hypotheses resulting from the experience of emergence). Within grammar, the same is true for many structures that we are used to call 'categories'. For instance, in GSS I claim that subjecthood is the construction of emergent activities exerted by information flow (word order), NP semantics, case marking (if present), agreement (if present) and much more. In some languages, the experience of such emergent activities is secondarily 'grammaticalized' or construed as a 'category'. I mention this example only to stress that - according to my opinion - 'emergence' (if applied to human cognition) refers to both the BEHAVIOR of a polycentric 'system' and to its (ritualized) interpretation as a cognitive hypothesis. A language system does not HAVE a category 'subject' per se, but parts of the system may behave in a way that leads to the 'assumption' (or hypothesis, or ICM) that there should be something like a category 'subject'. The problem is that linguists sometimes believe in such 'categories' without having checked whether they really ARE (in a substantial sense) or whether they are nothing but (more or less) grammaticalized experience of emergence. References: Schulze, Wolfgang 1998. Person, Klasse, Kongruenz - Fragmente einer Kategorialtypologie des einfachen Satzes in den ostkaukasischen Sprachen. Vol. I (in two parts): Die Grundlagen. München / Newcastle: LINCOM Europa (esp. chapters I, III, IV). Schulze, Wolfgang 1999. Cognitive Linguistics meet Typology: The Architecture of a "Grammar of Scenes and Scenarios" (ms., place of publication to be announced). Wolfgang -- [Note: My email address has been modified: Please use W.Schulze at lrz.uni-muenchen.de!] ___________________________________ | Prof. Dr. Wolfgang Schulze | Institut fuer Allgemeine und Indogermanische Sprachwissenschaft | Ludwig-Maximilians-Universitaet Muenchen | Geschwister-Scholl-Platz 1 | D-80539 Muenchen | Tel: +89-21802486 (secr.) | +89-21805343 (office) NEW ! NEW ! | Fax: +89-21805345 | Email: W.Schulze at lrz.uni-muenchen.de | | http://www.lrz-muenchen.de/~wschulze/ _____________________________________________________ From tgivon at OREGON.UOREGON.EDU Fri Jul 30 17:30:31 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 10:30:31 -0700 Subject: What is emergence anyway? Message-ID: RE: EMERGENCE The problem is not how to reconstruct the "proto semantics" of theterm. Neithe Brian nor Paul (nor anyone else, for that matter) should be bound by such historical reconstruction. Technical meanings arise from older informal usages, but are not forever bound to them by the original, metaphoric, umbilical cord. ("Radial categories" are a DIACHRONIC, not synchronic, phonomenon). As I see it, a much more vexing problem is: WHAT EMPIRICAL CLAIMS ARE ENTAILED BY THE USE OF THIS TERM? At the very least, we are in desperate need of some reliable procedure by which a "strong statistical tendency" that IS the result of a dedicated neuro-cognitive structure (an 'automated', 'routinized' system), can be differentiated from an equally strong statistical tengency that is NOT; i.e., from a "tendency" that is purely "epiphenomenal" or "emergent" (in the sense Osten attributes to Paul & Brian). As far as I know, neither Paul nor Brian have ever bothered to supply such criteria--or even ideas about how to look for such criteria. But this remains is a critical issue, because (a) many behavioral/statistical generalizations may indeed be epiphenomenal, the consequence of complex, indirect chains of dependencies. Such "tendencies" thus presumably don't have any structure(s)that motivate them DIRECTLY; tho some of the complex factors that give them rise may indeed have their own dedicated structures. (b) many behaviors that DO have strong neuro-cognitive dedicated structures do not reach the level of 100% generativity (re- gularity, predictability), but remain at the "merely statistical" level (say ca. 90%-95% predictability). So, the fact that a behavioral sub-system has some residual variability in it, by itself predicts nothing about whether the system is either "epiphenomenal" or "routinized/automated" (by the latter I now mean "having a dedicated neuro-cognitive structure"). And the question cannot, in principle, be resolved by citing linguistic-communicative (behavioral) facts alone. One has to look for other criteria. Of course, linguistic facts can give us STRONG HINTS. A behavior that reaches 90%-plus predictability is quite often a good candidate for having become automated (thus having "emerged", "grammaticalized" with a dedicated neuro-cognitively structure). Such cases are so well known in the psychology of automaticity elsewhere (mnemonics, piano-playing, kynesiology, etc.). But hints are only hints, and it would be nice if we could tap into some more direct criteria from "having a dedicated mental structure" that 'supports' grammaticalized behavior. It would be nice if, at the very least, we could begin to shift the discussion in that general direction. Cheers, TG ================= Wolfgang Schulze wrote: > > Dear all, > > I think, all linguists (from the camp of functionalism/derivationalism) > should be gratefull to Östen that he has raised the issue of > 'emergence'. This problem with this term is - I think - just the same > one as that related to other 'semi-TT' (terminus technicus) terms > introduced in linguistics. 'Emergence' (together with its derivatives) > clearly had its source in early 19th century 'philosophy of nature'. > >From there, it became popularized and adopted by other scientific (as > well as non-scientific) domains (cf. the well-documented but hardly > reflected history of the term 'pragmatics'). Hence, 'emergence' acquired > a relatively strong peripheric extension with respect to its semantics > which resulted in a relatively vague use of the term. > > In fact - I think - everybody has some notion of 'emergence', though > (s)he would have difficulties to define the term when used in everyday > (more or less) scientific discourse. The re-specification of this term > by e.g. Paul Hopper or Brian MacWhinney for linguistics naturally > retains some idiosyncratic reflexes of the early (informal) use of the > term (refering esp. to its broad semantic periphery). Östen has > reconstructed the prototypical (or central) semantics of 'emergence' > etc. in a very convincing way to which only little is to add. > > NB: The Latin basis 'e-mergo:' means 'rise, come up' etc., based on > 'mergo:' 'hide, cover, dive'.... (a good IE root, cf. Walde/Hoffmann LEW > II:76p.). > > However, since, 'emergence' has a semi-TT reading in linguistics now, it > would be good to always state the theoretical frame work or context in > which this term is embedded. In my own frame work ('Grammar of Scenes > and Scenarios (Schulze 1998, see > http://www.lrz-muenchen.de/~wschulze/pkk_1abs.htm); Schulze 1999 (see > http://www.lrz-muenchen.de/~wschulze/cog_typ.htm)) for instance, I claim > that 'emergence' is a cognitive procedure that denotes the 'dynamic > semantics' that are activated during the structural coupling of a > polycentric 'system' (or parts of it). By this is meant that the > interaction of source domains (organized in form of a polycentric > network) establishes an activation level that is characterized by the > production (or emergence) of some kind of specific behavior. This > behavior can be either somehow related to the behavior/properties of one > or more source domains, or totally different from the > behavior/properties of the source domains. The more the behavior of the > resutlting 'higher level domain' is different from that of its source > domains the more likely it is that its behavior is construed as a > cognitive 'reality' (that is as a idealized cognitive model) by human > experience. The main point is that contrary to the source domains, > 'higher level domains' emerging therefrom do not have 'properties' etc. > prima facie. Only IF they are construed (or experienced) as 'domains' > they are interpreted as having (autonomous) properties of their own. > > 'Language' - in my thinking - is a construction that refers to the > experience of emergent activities of underlying, structurally coupled > cognitive centers (which themselves again are partly cognitive > hypotheses resulting from the experience of emergence). Within grammar, > the same is true for many structures that we are used to call > 'categories'. For instance, in GSS I claim that subjecthood is the > construction of emergent activities exerted by information flow (word > order), NP semantics, case marking (if present), agreement (if present) > and much more. In some languages, the experience of such emergent > activities is secondarily 'grammaticalized' or construed as a > 'category'. I mention this example only to stress that - according to my > opinion - 'emergence' (if applied to human cognition) refers to both the > BEHAVIOR of a polycentric 'system' and to its (ritualized) > interpretation as a cognitive hypothesis. A language system does not > HAVE a category 'subject' per se, but parts of the system may behave in > a way that leads to the 'assumption' (or hypothesis, or ICM) that there > should be something like a category 'subject'. > > The problem is that linguists sometimes believe in such 'categories' > without having checked whether they really ARE (in a substantial sense) > or whether they are nothing but (more or less) grammaticalized > experience of emergence. > > References: > > Schulze, Wolfgang 1998. Person, Klasse, Kongruenz - Fragmente einer > Kategorialtypologie des einfachen Satzes in den ostkaukasischen > Sprachen. Vol. I (in two parts): Die Grundlagen. München / Newcastle: > LINCOM Europa (esp. chapters I, III, IV). > > Schulze, Wolfgang 1999. Cognitive Linguistics meet Typology: The > Architecture of a "Grammar of Scenes and > Scenarios" (ms., place of publication to be announced). > > Wolfgang > > -- > [Note: My email address has been modified: Please use > W.Schulze at lrz.uni-muenchen.de!] > ___________________________________ > | Prof. Dr. Wolfgang Schulze > | Institut fuer Allgemeine und Indogermanische Sprachwissenschaft > | Ludwig-Maximilians-Universitaet Muenchen > | Geschwister-Scholl-Platz 1 > | D-80539 Muenchen > | Tel: +89-21802486 (secr.) > | +89-21805343 (office) NEW ! NEW ! > | Fax: +89-21805345 > | Email: W.Schulze at lrz.uni-muenchen.de > | > | > http://www.lrz-muenchen.de/~wschulze/ > _____________________________________________________ From bates at CRL.UCSD.EDU Fri Jul 30 18:46:17 1999 From: bates at CRL.UCSD.EDU (Elizabeth Bates) Date: Fri, 30 Jul 1999 11:46:17 -0700 Subject: What is emergence anyway? In-Reply-To: <37A1E137.85E30833@oregon.uoregon.edu> Message-ID: Talmy is right that it is important to find empirical tests that can distinguish between grammatical outcomes that have a dedicated neurocognitive structure and those that have instead emerged across the course of learning and communication, supported by neurocognitive mechanisms that have "still kept their day jobs" (i.e. mechanisms that continue to do non-linguistic work for which they evolved long before grammar ever appeared). It's not an easy matter, but there is evidence on the point. Here are just a few relevant bits: 1) Children who have early damage to the putative language organs in the left hemisphere go on to acquire lexical and grammatical abilities that are perfectly normal (i.e. they are not aphasic, although as a group brain-damaged children tend to perform in the low-normal range, corresponding to an approximate average loss of 5 - 7 IQ points). Interesting, if these children are studied after 5 years of age, there seems to be no evidence whatsoever for differences as a function of early left- vs. right-hemisphere damage. This conclusion now appears to be on very solid ground, as several recent and comprehensive reviews of the literature have shown (by our research group; by Vargha-Khadem and colleagues; by Eisele and Aram). There were some inklings in the 1970's that this might not be the case (e.g. papers by Dennis & Whitaker, Woods & Teuber), but if you go back and read those carefully you will see that none of these studies were able to compare left- and right-hemisphere-damaged children directly. Either they looked only at children with left-hemisphere damage vs. controls, or they compared LHD and RHD to separate control groups without ever conducting a direct "head-on" comparison of the sort that would be required to conclude that one hemisphere is irreversibly superior to the other. At the very least, this conclusion suggests that there is no bounded and lateralized "organ" in the left hemisphere that is innately and irreversibly dedicated to grammar, and furthermore, parts of the brain that would ordinarily not be used to mediate grammatical processing can be encouraged to do so, quite successfully. 2) Recent neural imaging studies of normal adults have shown that the putative language areas of the left hemisphere are also involved in the mediation of non-linguistic processes. For example, studies from many different laboratories have shown that *ALL* of the candidate areas in and around the "Broca complex" are active in the planning and/or covert execution of non-linguistic mouth movements and/or non-symbolic movements of the hand. In other words, these areas may be involved in the processing of grammar, but they do a lot of other things as well, a point that is quite relevant to the notion of a "dedicated" neuro-cognitive system. Similar findings are beginning to pile up for the posterior language areas as well. 3) Another outcome of recent imaging studies is the now-widespread finding that many different parts of the brain are activated during language tasks, in patterns that vary depending on task demands, strategies, the expertise and experience of an individual subject. It may turn out that it makes no more sense to ask "where is language in the brain" than it does to ask "where is the dance in the dancer?" To be sure, it is the human brain (and only the human brain) that does language, but the case is not looking very good right now for the idea that it does so with a dedicated and well-defined language organ that has evolved to do language and language alone. -liz bates From tgivon at OREGON.UOREGON.EDU Fri Jul 30 19:14:27 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 12:14:27 -0700 Subject: What is emergence anyway? Message-ID: Liz addressed an important part of the question--whether there are or are not language-specific dedicated structure. I have no strong preconceptions on the subject (tho some weak biases...). But the question of "emergence" in the Hopper/MacWhinney sense (if I understand it) remains even if the structures are not language-specific: Automaticity during lifetime learning, through whatever neural structure, is the hallmark of all skilled acquisition. Grammar is a skilled performance PAR EXCELLENCE. So when it "emerges", does it acquire the (at least partial) rigidity & categoriality characteristics that have been shown in the acquisition of all other complex, rhythmic-hierarchic skilled behaviours? Here, I obviously have a VERY strong bias. But regardless of what ones bias is, we've got to find non-linguistic criteria for seeing whether the PROCESS of "emergence" does lead to the STATE of "having emerged", or not. I myself find the "not" option rather bizarre, given what is known about the expenditure of neuro-cognitive resources during development (both ontogenetic & phylogenetic). But one way or another, the issue needs to be resolved through making empirical predictions & then seeing if they pan out. TG Elizabeth Bates wrote: > > Talmy is right that it is important to find empirical tests that can > distinguish between grammatical outcomes that have a dedicated > neurocognitive structure and those that have instead emerged > across the course of learning and communication, supported by > neurocognitive mechanisms that have "still kept their day jobs" > (i.e. mechanisms that continue to do non-linguistic work for which > they evolved long before grammar ever appeared). It's not an easy > matter, but there is evidence on the point. Here are just a few > relevant bits: > > 1) Children who have early damage to the putative language organs > in the left hemisphere go on to acquire lexical and grammatical > abilities that are perfectly normal (i.e. they are not aphasic, > although as a group brain-damaged children tend to perform in > the low-normal range, corresponding to an approximate average > loss of 5 - 7 IQ points). Interesting, if these children are > studied after 5 years of age, there seems to be no evidence > whatsoever for differences as a function of early left- vs. > right-hemisphere damage. This conclusion now appears to be > on very solid ground, as several recent and comprehensive > reviews of the literature have shown (by our research group; > by Vargha-Khadem and colleagues; by Eisele and Aram). There > were some inklings in the 1970's that this might not be the > case (e.g. papers by Dennis & Whitaker, Woods & Teuber), but if you go > back and read those carefully you will see that none of these studies > were able to compare left- and right-hemisphere-damaged children > directly. Either they looked only at children with left-hemisphere > damage vs. controls, or they compared LHD and RHD to separate > control groups without ever conducting a direct "head-on" comparison > of the sort that would be required to conclude that one hemisphere > is irreversibly superior to the other. At the very least, this > conclusion suggests that there is no bounded and lateralized "organ" > in the left hemisphere that is innately and irreversibly dedicated > to grammar, and furthermore, parts of the brain that would ordinarily not > be used to mediate grammatical processing can be encouraged to do so, quite > successfully. > > 2) Recent neural imaging studies of normal adults have shown that > the putative language areas of the left hemisphere are also involved > in the mediation of non-linguistic processes. For example, studies > from many different laboratories have shown that *ALL* of the > candidate areas in and around the "Broca complex" are active in the > planning and/or covert execution of non-linguistic mouth movements > and/or non-symbolic movements of the hand. In other words, these > areas may be involved in the processing of grammar, but they do a lot > of other things as well, a point that is quite relevant to the > notion of a "dedicated" neuro-cognitive system. Similar findings > are beginning to pile up for the posterior language areas as well. > > 3) Another outcome of recent imaging studies is the now-widespread > finding that many different parts of the brain are activated during > language tasks, in patterns that vary depending on task demands, > strategies, the expertise and experience of an individual subject. > It may turn out that it makes no more sense to ask "where is language > in the brain" than it does to ask "where is the dance in the dancer?" > To be sure, it is the human brain (and only the human brain) that > does language, but the case is not looking very good right now for > the idea that it does so with a dedicated and well-defined language > organ that has evolved to do language and language alone. -liz bates From jtang at COGSCI.BERKELEY.EDU Fri Jul 30 19:56:31 1999 From: jtang at COGSCI.BERKELEY.EDU (Joyce Tang Boyland) Date: Fri, 30 Jul 1999 12:56:31 -0700 Subject: automatized or neurocognitively dedicated? Message-ID: I think some clarification is in order on what people mean by "dedicated neurocognitive structure". I can imagine interpretations that fit what Liz Bates is saying, and I can imagine others that fit what Tom Givon is saying, but I do not think that they are using the term in the same sense. Compare: Bates writes: Talmy is right that it is important to find empirical tests that can DISTINGUISH between grammatical outcomes that have a dedicated neurocognitive structure and those that have instead emerged across the course of learning and communication, supported by neurocognitive mechanisms that have "still kept their day jobs" (i.e. mechanisms that continue to do non-linguistic work for which they evolved long before grammar ever appeared). [emphasis mine --JTB] My reading of Givon is that he is *not* in fact making that distinction. Givon writes: a behavioral subsystem ...[might be] ... either "epiphenomenal" or "routinized/automated" (by the LATTER I now mean "having a dedicated neuro-cognitive structure"). [emphasis mine --JTB] and: [some behaviors become] ... automated (thus having "emerged", "grammaticalized" with a dedicated neuro-cognitive structure). The critical fact is that routinization/automatization is a prime instance of the various neurocognitive mechanisms that have "kept their day jobs". So the issue is that I think we need to decide whether we mean the kinds of neurocognitive structures that exist *regardless* of routinization through experience, or the kinds of neurocognitive structures that are the *result* of general mechanisms such as routinization through experience? Joyce Tang Boyland Alverno College Milwaukee, WI 53234-3922 From macw at CMU.EDU Fri Jul 30 22:37:46 1999 From: macw at CMU.EDU (Brian MacWhinney) Date: Fri, 30 Jul 1999 18:37:46 -0400 Subject: emergence Message-ID: Dear FunkNet, It was a busy day, or I would have chimed in earlier regarding this useful discussion of emergence. I found Östen Dahl's note useful. In particular, I very much agreed with the idea that one tends to evaluate emergence differently according to whether one emphasizes the novelty of the emergent system or the degree to which its structure can be partially predicted from its predecessors. Wolfgang Schulze's claim that "subjecthood is the construction of emergent activities exerted by information flow (word order), NP semantics, case marking (if present), agreement (if present) and much more" is certainly a correct application of the notion of emergence to language, as Liz Bates and I argued in a series of papers in the 1980s. In those papers, we distinguished three levels of emergence: diachronic, acquisitional, and synchronic (sentence processing). More recently, I have found it useful to think in terms of five levels of emergence: 1. Evolutionary 2. Embryological 3. Developmental 4. On-line 5. Diachronic It seems to me that it is crucial to view emergence in terms of these varieties and to keep in mind the fact that the time scales for these varieties of emergence differ vastly. Simply listing these varieties of emergence does not answer Tom Givon's question. He wants to know which type of emergence is operative in a given behavior. To do that, we have to dig into the guts of specific models. For example, in their chapter in "The Emergence of Language" (Lawrence Erlbaum, 1999, B. MacWhinney (Ed.)), Prahlad Gupta and Gary Dell show that "a good lexicon" has words that share neighbors in the rime, but which differ in the onset. (CAT, BAT, SAT, RAT, MAT, FAT ..) They then proceed to show how this aspect of lexical structure arises directly from the Dell Speech Error model -- a data-driven connectionist model of speech production. What Tom wants to know, if I read him correctly, is whether the automated phonotactic processes that Gupta and Dell have studied are supported by simple general purpose learning mechanism or instead by some more dedicated neural hardware. The likely answer to this is that general purpose learning mechanisms are themselves based on specific neural adaptations. For example, Kohonen self-organizing feature maps demonstrate behaviors that emerge from the neural preference for lateral inhibition and short connections. These models can help explain the neurological grounding of the emergence of the preference for word onsets during development. Once models are formulated in this way, we start to realize that behaviors arise from an interaction of sources of emergence on varying time scales. Although automaticity is typically a fact of developmental emergence, it builds on embryologically-emergent structures and can be modified by on-line processes. The moral: the importance of emergence becomes clear once we begin to think in terms of models of specific behaviors. If we don't think in terms of concrete models, it seems like semi-TT gobbledy-gook. --Brian MacWhinney From tgivon at OREGON.UOREGON.EDU Sat Jul 31 00:10:31 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 17:10:31 -0700 Subject: automatized or neurocognitively dedicated? Message-ID: I think Joyce is misinterpreting me. All I said is that there are two separate issues. And that they are distinct. One is about evolution, and Liz has certainly expressed her position about this. If I elected not to touch upon it at this point, it is not because I subscribe to Liz's (what seems to me...) rather extreme position (in essence, that language is too recent to have any genetic consequences; while memory & attention are old enough & have genetic consequences). The issue I did touch upon exists regardless of how you resolve the evolutionary/genetic issue. Whether the underlying mechanisms are language-specific or not, "emergence" during lifetime skilled learning either does or doesn't have neuro-cognitive consequences (the creation of automated structured during skill acquisition). My bias on this is obviously strong, as I have stated. I would like to find empirical means of resolving this. And I am still waiting either Brian or Paul to address the SECOND issue. TG Joyce Tang Boyland wrote: > > I think some clarification is in order on what people mean > by "dedicated neurocognitive structure". I can imagine > interpretations that fit what Liz Bates is saying, and I can > imagine others that fit what Tom Givon is saying, but I do not > think that they are using the term in the same sense. > Compare: > > Bates writes: > Talmy is right that it is important to find empirical tests that can > DISTINGUISH between grammatical outcomes that have a dedicated > neurocognitive structure and those that have instead emerged > across the course of learning and communication, supported by > neurocognitive mechanisms that have "still kept their day jobs" > (i.e. mechanisms that continue to do non-linguistic work for which > they evolved long before grammar ever appeared). [emphasis mine --JTB] > > My reading of Givon is that he is *not* in fact making that distinction. > > Givon writes: > a behavioral subsystem ...[might be] ... either "epiphenomenal" > or "routinized/automated" (by the LATTER I now mean > "having a dedicated neuro-cognitive structure"). [emphasis mine --JTB] > and: > [some behaviors become] ... automated (thus having "emerged", > "grammaticalized" with a dedicated neuro-cognitive structure). > > The critical fact is that routinization/automatization > is a prime instance of the various neurocognitive mechanisms > that have "kept their day jobs". > So the issue is that I think we need to decide > whether we mean the kinds of neurocognitive structures that > exist *regardless* of routinization through experience, or > the kinds of neurocognitive structures that are the *result* of > general mechanisms such as routinization through experience? > > Joyce Tang Boyland > Alverno College > Milwaukee, WI 53234-3922 From wcmann at JUNO.COM Sat Jul 3 20:43:53 1999 From: wcmann at JUNO.COM (William Mann) Date: Sat, 3 Jul 1999 16:43:53 -0400 Subject: discourse functions of property words Message-ID: Replying to Michael Dryer on discourse and predication: Also in 1988, Sandy Thompson and I published a paper defining Rhetorical Structure Theory, Text 8(3). It is an approach to assessing whole-text discourse structure, so far defined only for monologue. It makes extensive use of discourse relations, some of which link a text span (called the nucleus) to another (called a satellite.) Some of these discourse relations are sometimes signaled by discourse particles, which would certainly be included in ?other types of words? in your inquiry. In various related papers, mostly published before that one, we talked about ?relational propositions,? which are predications that arise when particular relations are found to be part of the discourse structure of a text. Such a predication will arise even if the presence of the relation is not signaled. The paper has no formal taxonomy, but it does include a number of distinctions which have been taken to be taxonomic. Exclusion of a taxonomy was deliberate on our part, because we did not have a particular taxonomy that we felt happy with. The distinctions included a volitional / nonvolitional contrast among causal relations, and a contrast between so called ?subject matter? relations (e.g. Conditional) and ?presentational? relations (e.g. Concessive.) There were multiple grounds for identifying the Elaboration relation. Some relations (e.g. Motivation) applied only where the nucleus span presented an action. Beyond the notion of belief, there was a broader notion called ?positive regard? which included belief, intention to act and approval; it was used to avoid proliferation of relations whose definitions would otherwise be identical. All of this suggests that there is plenty in RST to taxonomize, and that different sorts of worthwhile taxonomies might be produced. Several research papers using RST have been critical of the irregularity of the set of relation definitions, the set of distinctions and the absence of a taxonomy. Particularly for purposes of formalization, sometimes as a means to programming text generation programs or analyzers of text, various modifications or replacements of the set of relations have been suggested. I hope that this gives you a useful lead into this particular corner of the literature. Bill Mann On Thu, 24 Jun 1999 20:18:26 -0400 Matthew S Dryer writes: >Sandy Thompson (1988) argues that property words tend to involve the >discourse function of "predication" in natural speech. Is anybody >aware >of any attempt to develop a finer taxonomy of discourse functions >associated with property words (or other types of words) that would >distinguish different subtypes of "predication"? > >Thanks, > >Matthew Dryer . . . . ..... end of Bill's message, apologies for the ad at the bottom. ..... From kosam at LIBR.UG.EDU.GH Thu Jul 8 07:16:17 1999 From: kosam at LIBR.UG.EDU.GH (Kweku Osam) Date: Thu, 8 Jul 1999 08:16:17 +0100 Subject: Multiple Choice Questions Message-ID: The Department of Linguistics at the University of Ghana is considering using the Multiple Choice Questions formart in examining Linguistics courses at the lower levels. We would like to know if there are Departments which have tried this method of examining in Linguistics and what their experience has been. We'd also appreciate sample questions in various areas of the discipline. Send all reactions directly to me. Thanks. Kweku Osam Department of Linguistics University of Ghana Legon GHANA kosam at libr.ug.edu.gh From pwd at RICE.EDU Sat Jul 17 01:31:49 1999 From: pwd at RICE.EDU (Philip W Davis) Date: Fri, 16 Jul 1999 20:31:49 -0500 Subject: Best Grammar(s) Message-ID: Best Grammar(s) of the Twentieth Century As we approach the quadruple witching of end-of-the-year, end-of-the-decade, end-of-the-century & end-of-the-millenium, register your nomination for the 'best' grammar(s) of the twentieth century. 'Best', of course, translates directly into 'favorite'. It is a subjective reaction, but we must all have our 'best'/'favorite' grammars. *These will be grammars made public/published after January 1, 1900. But they may be dissertations that were never distributed otherwise. *These will be 'book length', meaning that (however long) they will not be a part of a collection of 'sketches', although they may be monographs in a series. They will have a separate bibliographical identity, i.e., their own ISBN. *Please provide bibliographical information: Author(s). Date. Title. Place of Publication: Publisher. *These will be 'synchronic' grammars. *They will be 'grammars'. Although phonetics & phonology are normally components of a grammar, works confined to those areas, e.g. The Sound Pattern of English, will not be eligible. *You can nominate/vote for as many grammars as you wish. *You can only nominate/vote once for a given title. *You can return and nominate/vote again if a title, which had earlier slipped your mind, occurs to you. *Yes, if you insist, you can nominate/vote for your own grammars. *You can provide reasons for your selection(s), and those comments may appear in association with the respective titles. I _especially_ encourage this. *Only positive nominations/votes will be tallied and only neutral/positive remarks, repeated. Flames will be ignored. *If you wish your comments to remain anonymous, tell me and I will attribute them to 'Anon.' *When/If there are sufficient responses to generate a list, it will be available www.ruf.rice.edu/~pwd/index.html. *You have until December 31, 1999, 11:59pm. The list, if there is one, will stay up after that. *You may nominate/vote by leaving me e-mail at pwd at rice.edu. Use the subject line: 'Grammar(s)'. +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= Philip W. Davis e-mail: pwd at rice.edu Department of Linguistics MS23 tel: (713)527-6010 Rice University fax: (713)527-4718 6100 Main St. web: www.ruf.rice.edu/~pwd/index.html Houston, TX 77005 USA +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= From tpayne at OREGON.UOREGON.EDU Sat Jul 17 16:20:19 1999 From: tpayne at OREGON.UOREGON.EDU (Tom Payne) Date: Sat, 17 Jul 1999 09:20:19 -0700 Subject: New Linguistic Olympics puzzles Message-ID: Dear Funknet You may like to know that there are now seven new "puzzles" on the Linguistic Olympics website. These are the following: Samoan, Malay/Indonesian, Maasai, Swahili #2, Tamil, Yaqui, and Classical Nahuatl. This brings the total to 21 puzzles that are available on the site. For the past six months the site has been averaging about 250 "hits" per week. I answer between 5 and 10 messages a day from individuals who attempt to solve the puzzles. Several Junior, middle and high school teachers have let me know they are using the puzzles in their classes. Others have inquired about possibly organizing a "Linguistic Olympics" event at their school. So far I am not aware of anyone who has actually done this though. The main Linguistic Olympics webpage is http://darkwing.uoregon.edu/~tpayne/lingolym/lingolym.htm. There is also a report to the LSA on the 1998 Eugene Linguistic Olympics at http://darkwing.uoregon.edu/~tpayne/lingolym/LOreport.htm. The event was held again in 1999, but with substantial changes. If there is interest (and if I have time), I would like to report on that event as well. I encourage everyone to check out the site, and give me any comments, suggestions, etc. These are puzzles geared to American teenagers who have no knowledge of linguistics. Tom Payne From lrequena at DANYSOFT.COM.AR Sat Jul 17 17:52:27 1999 From: lrequena at DANYSOFT.COM.AR (Leonel Requena) Date: Sat, 17 Jul 1999 14:52:27 -0300 Subject: Twentieth Century Message-ID: Philip W. Davis wrote (16/07/99): >Best Grammar(s) of the Twentieth Century >As we approach the quadruple witching of end-of-the-year, >end-of-the-decade, end-of-the-century & end-of-the-millenium, register your >nomination for the 'best' grammar(s) of the twentieth century. > *You have until December 31, 1999, 11:59pm. The list, if there is one, > will stay up after that You should be aware that the end-of the-year is not, at the same time, the end-of-the-decade, end-of-the-century, etc., (all this at the end-of-next-year). If the best grammar of the century appears next year, what are you going to do? ?That's unfair! Best wishes for the end-of-the-year and for the end-of-that-year. Leonel Requena (Not in a hurry). Buenos Aires -------------- next part -------------- An HTML attachment was scrubbed... URL: From cpensa at GUGU.USAL.ES Tue Jul 20 09:37:11 1999 From: cpensa at GUGU.USAL.ES (Carmen Pensado) Date: Tue, 20 Jul 1999 11:37:11 +0200 Subject: congress announcement Message-ID: El pr?ximo XXIII Congreso Internacional de Ling??stica y Filolog?a Rom?nica, reuni?n de la Soci?t? de Linguistique Romane, se celebrar? en Salamanca (Espa?a) en Setiembre del 2001. > > La primera circular ha sido puesta en marcha. Pueden hallar una versi?n electr?nica en: > > http://web.usal.es/~romanica/romanica.htm > From oesten at ling.su.se Fri Jul 30 13:58:36 1999 From: oesten at ling.su.se (Östen Dahl) Date: Fri, 30 Jul 1999 15:58:36 +0200 Subject: What is emergence anyway? Message-ID: What is emergence anyway? Recently, the terms "emergent" and "emergence" have become popular in linguistics and neighbouring fields. Here are two representative quotations: " Hopper suggested that the study of grammaticalization tended to undermine the assumption of preexistent a priori grammatical component that stood as a prerequisite for discourse and a precondition for communication, and he proposed instead that grammar was an emergent property of texts. "Structure" would then be an epiphenomenal by-product of discourse." (Hopper 1996, 231) "If you spend some time watching the checkout lines at a supermarket, you quickly find that the number of people queued up in each line is roughly the same There is no fixed rule governing this pattern. Instead, the rule that equalizes the number of shoppers in the various lines emerges from other basic facts about the goals and behavior of shoppers and supermarket managers. This simple idea of emergence through constraint satisfaction is currently being invoked as a central explanatory mechanism in many areas of cognitive science and neuroscience. the behaviors that we tend to characterize in terms of rules and symbols are in fact emergent patterns that arise from the interactions of other less complex or more stable underlying systems. I will refer to this new viewpoint on language learning and processing as ?emergentism?." (MacWhinney 1999) The terms "emergent" and "emergence" however, have been around for quite some time. In Encyclopedia Britannica's article "Emergence", the term is said to be used in evolutionary theory in the sense of "the rise of a system that cannot be predicted or explained from antecedent conditions" EB then refers to the 19th century philosopher of science George Henry Lewes' distinction between resultants and emergents: "phenomena that are predictable from their constituent parts and those that are not". An example of a resultant would be a physical mixture of sand and talcum powder, while a chemical compound such as salt, which looks nothing like sodium or chlorine, would exemplify an emergent. The EB article enumerates a number of stages in biological evolution "at which fundamentally new forms have appeared", such the orgin of life, the origin of nucleus-bearing protozoa, and the rise of sentient beings with nervous systems, and goes on: "Each of these new modes of life, though grounded in the physicochemical and biochemical conditions of the previous and simpler stage, is intelligible only in terms of its own ordering principle. These are thus cases of emergence." Another presumably authoritative source is the website of the journal Emergence: A Journal of Complexity Issues in Organizations and Management (http://emergence.org). It declares that the idea of emergence is used in the study of complex systems to "indicate the arising of patterns, structures, or properties that do not seem adequately explained by referring only to the system's pre-existing components and their interaction." and says that emergence is particularly important when "- when the organization of the system, i.e., its global order, appears to be more salient and of a different kind than the components alone; - when the components can be replaced without an accompanying decommissioning of the whole system; - when the new global patterns or properties are radically novel with respect to the pre-existing components; thus, the emergent patterns seem to be unpredictable and nondeducible from the components as well as irreducible to those components." A reader who tries to reconcile what is said about emergence in all these quotations feels rather bewildered. In the older tradition, still represented by EB and the journal Emergence, "emergence" stands for new and interesting higher-order structures that are not reducible to the lower-order ones. MacWhinney and Hopper, on the contrary, seem to use the same term precisely for the opposite: seemingly complex systems that are in fact derivable from -- "epiphenomenal by-products" of -- other simpler systems. One may ask how such a radical shift in meaning may have occurred. It seems that we can find the seed of the conflict in the original notion of emergence. On one hand, the target has new and interesting properties that cannot be described in terms of the source, on the other, there is presumably some kind of causal chain that leads from the source to the target. The essence of the notion of a self-organizing system seems to be precisely the fact that unexpected things happen as it were by themselves. Depending on whether one is more fascinated by the novel or the predictable component in this process, one may come to see different and seemingly contradictory aspects of "emergence" as criterial. Hopefully, we will eventually be able to see both sides of the phenomena at the same time. REFERENCES Hopper, Paul J. 1996. Some recent trends in grammaticalization. Annual Review of Anthropology. 25:217-36. MacWhinney, Brian. 1999. Emergent Language. In M. Darnell, E. Moravcsik, F. Newmeyer, M. Noonan, and K. Wheatley, eds. Functionalism and Formalism in Linguistics. New York: John Benjamins. ----------------------- ?sten Dahl oesten at ling.su.se http://www.ling.su.se/staff/oesten/oesten.htm From W.Schulze at LRZ.UNI-MUENCHEN.DE Fri Jul 30 15:37:34 1999 From: W.Schulze at LRZ.UNI-MUENCHEN.DE (Wolfgang Schulze) Date: Fri, 30 Jul 1999 17:37:34 +0200 Subject: What is emergence anyway? Message-ID: Dear all, I think, all linguists (from the camp of functionalism/derivationalism) should be gratefull to ?sten that he has raised the issue of 'emergence'. This problem with this term is - I think - just the same one as that related to other 'semi-TT' (terminus technicus) terms introduced in linguistics. 'Emergence' (together with its derivatives) clearly had its source in early 19th century 'philosophy of nature'. >>From there, it became popularized and adopted by other scientific (as well as non-scientific) domains (cf. the well-documented but hardly reflected history of the term 'pragmatics'). Hence, 'emergence' acquired a relatively strong peripheric extension with respect to its semantics which resulted in a relatively vague use of the term. In fact - I think - everybody has some notion of 'emergence', though (s)he would have difficulties to define the term when used in everyday (more or less) scientific discourse. The re-specification of this term by e.g. Paul Hopper or Brian MacWhinney for linguistics naturally retains some idiosyncratic reflexes of the early (informal) use of the term (refering esp. to its broad semantic periphery). ?sten has reconstructed the prototypical (or central) semantics of 'emergence' etc. in a very convincing way to which only little is to add. NB: The Latin basis 'e-mergo:' means 'rise, come up' etc., based on 'mergo:' 'hide, cover, dive'.... (a good IE root, cf. Walde/Hoffmann LEW II:76p.). However, since, 'emergence' has a semi-TT reading in linguistics now, it would be good to always state the theoretical frame work or context in which this term is embedded. In my own frame work ('Grammar of Scenes and Scenarios (Schulze 1998, see http://www.lrz-muenchen.de/~wschulze/pkk_1abs.htm); Schulze 1999 (see http://www.lrz-muenchen.de/~wschulze/cog_typ.htm)) for instance, I claim that 'emergence' is a cognitive procedure that denotes the 'dynamic semantics' that are activated during the structural coupling of a polycentric 'system' (or parts of it). By this is meant that the interaction of source domains (organized in form of a polycentric network) establishes an activation level that is characterized by the production (or emergence) of some kind of specific behavior. This behavior can be either somehow related to the behavior/properties of one or more source domains, or totally different from the behavior/properties of the source domains. The more the behavior of the resutlting 'higher level domain' is different from that of its source domains the more likely it is that its behavior is construed as a cognitive 'reality' (that is as a idealized cognitive model) by human experience. The main point is that contrary to the source domains, 'higher level domains' emerging therefrom do not have 'properties' etc. prima facie. Only IF they are construed (or experienced) as 'domains' they are interpreted as having (autonomous) properties of their own. 'Language' - in my thinking - is a construction that refers to the experience of emergent activities of underlying, structurally coupled cognitive centers (which themselves again are partly cognitive hypotheses resulting from the experience of emergence). Within grammar, the same is true for many structures that we are used to call 'categories'. For instance, in GSS I claim that subjecthood is the construction of emergent activities exerted by information flow (word order), NP semantics, case marking (if present), agreement (if present) and much more. In some languages, the experience of such emergent activities is secondarily 'grammaticalized' or construed as a 'category'. I mention this example only to stress that - according to my opinion - 'emergence' (if applied to human cognition) refers to both the BEHAVIOR of a polycentric 'system' and to its (ritualized) interpretation as a cognitive hypothesis. A language system does not HAVE a category 'subject' per se, but parts of the system may behave in a way that leads to the 'assumption' (or hypothesis, or ICM) that there should be something like a category 'subject'. The problem is that linguists sometimes believe in such 'categories' without having checked whether they really ARE (in a substantial sense) or whether they are nothing but (more or less) grammaticalized experience of emergence. References: Schulze, Wolfgang 1998. Person, Klasse, Kongruenz - Fragmente einer Kategorialtypologie des einfachen Satzes in den ostkaukasischen Sprachen. Vol. I (in two parts): Die Grundlagen. M?nchen / Newcastle: LINCOM Europa (esp. chapters I, III, IV). Schulze, Wolfgang 1999. Cognitive Linguistics meet Typology: The Architecture of a "Grammar of Scenes and Scenarios" (ms., place of publication to be announced). Wolfgang -- [Note: My email address has been modified: Please use W.Schulze at lrz.uni-muenchen.de!] ___________________________________ | Prof. Dr. Wolfgang Schulze | Institut fuer Allgemeine und Indogermanische Sprachwissenschaft | Ludwig-Maximilians-Universitaet Muenchen | Geschwister-Scholl-Platz 1 | D-80539 Muenchen | Tel: +89-21802486 (secr.) | +89-21805343 (office) NEW ! NEW ! | Fax: +89-21805345 | Email: W.Schulze at lrz.uni-muenchen.de | | http://www.lrz-muenchen.de/~wschulze/ _____________________________________________________ From tgivon at OREGON.UOREGON.EDU Fri Jul 30 17:30:31 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 10:30:31 -0700 Subject: What is emergence anyway? Message-ID: RE: EMERGENCE The problem is not how to reconstruct the "proto semantics" of theterm. Neithe Brian nor Paul (nor anyone else, for that matter) should be bound by such historical reconstruction. Technical meanings arise from older informal usages, but are not forever bound to them by the original, metaphoric, umbilical cord. ("Radial categories" are a DIACHRONIC, not synchronic, phonomenon). As I see it, a much more vexing problem is: WHAT EMPIRICAL CLAIMS ARE ENTAILED BY THE USE OF THIS TERM? At the very least, we are in desperate need of some reliable procedure by which a "strong statistical tendency" that IS the result of a dedicated neuro-cognitive structure (an 'automated', 'routinized' system), can be differentiated from an equally strong statistical tengency that is NOT; i.e., from a "tendency" that is purely "epiphenomenal" or "emergent" (in the sense Osten attributes to Paul & Brian). As far as I know, neither Paul nor Brian have ever bothered to supply such criteria--or even ideas about how to look for such criteria. But this remains is a critical issue, because (a) many behavioral/statistical generalizations may indeed be epiphenomenal, the consequence of complex, indirect chains of dependencies. Such "tendencies" thus presumably don't have any structure(s)that motivate them DIRECTLY; tho some of the complex factors that give them rise may indeed have their own dedicated structures. (b) many behaviors that DO have strong neuro-cognitive dedicated structures do not reach the level of 100% generativity (re- gularity, predictability), but remain at the "merely statistical" level (say ca. 90%-95% predictability). So, the fact that a behavioral sub-system has some residual variability in it, by itself predicts nothing about whether the system is either "epiphenomenal" or "routinized/automated" (by the latter I now mean "having a dedicated neuro-cognitive structure"). And the question cannot, in principle, be resolved by citing linguistic-communicative (behavioral) facts alone. One has to look for other criteria. Of course, linguistic facts can give us STRONG HINTS. A behavior that reaches 90%-plus predictability is quite often a good candidate for having become automated (thus having "emerged", "grammaticalized" with a dedicated neuro-cognitively structure). Such cases are so well known in the psychology of automaticity elsewhere (mnemonics, piano-playing, kynesiology, etc.). But hints are only hints, and it would be nice if we could tap into some more direct criteria from "having a dedicated mental structure" that 'supports' grammaticalized behavior. It would be nice if, at the very least, we could begin to shift the discussion in that general direction. Cheers, TG ================= Wolfgang Schulze wrote: > > Dear all, > > I think, all linguists (from the camp of functionalism/derivationalism) > should be gratefull to ?sten that he has raised the issue of > 'emergence'. This problem with this term is - I think - just the same > one as that related to other 'semi-TT' (terminus technicus) terms > introduced in linguistics. 'Emergence' (together with its derivatives) > clearly had its source in early 19th century 'philosophy of nature'. > >From there, it became popularized and adopted by other scientific (as > well as non-scientific) domains (cf. the well-documented but hardly > reflected history of the term 'pragmatics'). Hence, 'emergence' acquired > a relatively strong peripheric extension with respect to its semantics > which resulted in a relatively vague use of the term. > > In fact - I think - everybody has some notion of 'emergence', though > (s)he would have difficulties to define the term when used in everyday > (more or less) scientific discourse. The re-specification of this term > by e.g. Paul Hopper or Brian MacWhinney for linguistics naturally > retains some idiosyncratic reflexes of the early (informal) use of the > term (refering esp. to its broad semantic periphery). ?sten has > reconstructed the prototypical (or central) semantics of 'emergence' > etc. in a very convincing way to which only little is to add. > > NB: The Latin basis 'e-mergo:' means 'rise, come up' etc., based on > 'mergo:' 'hide, cover, dive'.... (a good IE root, cf. Walde/Hoffmann LEW > II:76p.). > > However, since, 'emergence' has a semi-TT reading in linguistics now, it > would be good to always state the theoretical frame work or context in > which this term is embedded. In my own frame work ('Grammar of Scenes > and Scenarios (Schulze 1998, see > http://www.lrz-muenchen.de/~wschulze/pkk_1abs.htm); Schulze 1999 (see > http://www.lrz-muenchen.de/~wschulze/cog_typ.htm)) for instance, I claim > that 'emergence' is a cognitive procedure that denotes the 'dynamic > semantics' that are activated during the structural coupling of a > polycentric 'system' (or parts of it). By this is meant that the > interaction of source domains (organized in form of a polycentric > network) establishes an activation level that is characterized by the > production (or emergence) of some kind of specific behavior. This > behavior can be either somehow related to the behavior/properties of one > or more source domains, or totally different from the > behavior/properties of the source domains. The more the behavior of the > resutlting 'higher level domain' is different from that of its source > domains the more likely it is that its behavior is construed as a > cognitive 'reality' (that is as a idealized cognitive model) by human > experience. The main point is that contrary to the source domains, > 'higher level domains' emerging therefrom do not have 'properties' etc. > prima facie. Only IF they are construed (or experienced) as 'domains' > they are interpreted as having (autonomous) properties of their own. > > 'Language' - in my thinking - is a construction that refers to the > experience of emergent activities of underlying, structurally coupled > cognitive centers (which themselves again are partly cognitive > hypotheses resulting from the experience of emergence). Within grammar, > the same is true for many structures that we are used to call > 'categories'. For instance, in GSS I claim that subjecthood is the > construction of emergent activities exerted by information flow (word > order), NP semantics, case marking (if present), agreement (if present) > and much more. In some languages, the experience of such emergent > activities is secondarily 'grammaticalized' or construed as a > 'category'. I mention this example only to stress that - according to my > opinion - 'emergence' (if applied to human cognition) refers to both the > BEHAVIOR of a polycentric 'system' and to its (ritualized) > interpretation as a cognitive hypothesis. A language system does not > HAVE a category 'subject' per se, but parts of the system may behave in > a way that leads to the 'assumption' (or hypothesis, or ICM) that there > should be something like a category 'subject'. > > The problem is that linguists sometimes believe in such 'categories' > without having checked whether they really ARE (in a substantial sense) > or whether they are nothing but (more or less) grammaticalized > experience of emergence. > > References: > > Schulze, Wolfgang 1998. Person, Klasse, Kongruenz - Fragmente einer > Kategorialtypologie des einfachen Satzes in den ostkaukasischen > Sprachen. Vol. I (in two parts): Die Grundlagen. M?nchen / Newcastle: > LINCOM Europa (esp. chapters I, III, IV). > > Schulze, Wolfgang 1999. Cognitive Linguistics meet Typology: The > Architecture of a "Grammar of Scenes and > Scenarios" (ms., place of publication to be announced). > > Wolfgang > > -- > [Note: My email address has been modified: Please use > W.Schulze at lrz.uni-muenchen.de!] > ___________________________________ > | Prof. Dr. Wolfgang Schulze > | Institut fuer Allgemeine und Indogermanische Sprachwissenschaft > | Ludwig-Maximilians-Universitaet Muenchen > | Geschwister-Scholl-Platz 1 > | D-80539 Muenchen > | Tel: +89-21802486 (secr.) > | +89-21805343 (office) NEW ! NEW ! > | Fax: +89-21805345 > | Email: W.Schulze at lrz.uni-muenchen.de > | > | > http://www.lrz-muenchen.de/~wschulze/ > _____________________________________________________ From bates at CRL.UCSD.EDU Fri Jul 30 18:46:17 1999 From: bates at CRL.UCSD.EDU (Elizabeth Bates) Date: Fri, 30 Jul 1999 11:46:17 -0700 Subject: What is emergence anyway? In-Reply-To: <37A1E137.85E30833@oregon.uoregon.edu> Message-ID: Talmy is right that it is important to find empirical tests that can distinguish between grammatical outcomes that have a dedicated neurocognitive structure and those that have instead emerged across the course of learning and communication, supported by neurocognitive mechanisms that have "still kept their day jobs" (i.e. mechanisms that continue to do non-linguistic work for which they evolved long before grammar ever appeared). It's not an easy matter, but there is evidence on the point. Here are just a few relevant bits: 1) Children who have early damage to the putative language organs in the left hemisphere go on to acquire lexical and grammatical abilities that are perfectly normal (i.e. they are not aphasic, although as a group brain-damaged children tend to perform in the low-normal range, corresponding to an approximate average loss of 5 - 7 IQ points). Interesting, if these children are studied after 5 years of age, there seems to be no evidence whatsoever for differences as a function of early left- vs. right-hemisphere damage. This conclusion now appears to be on very solid ground, as several recent and comprehensive reviews of the literature have shown (by our research group; by Vargha-Khadem and colleagues; by Eisele and Aram). There were some inklings in the 1970's that this might not be the case (e.g. papers by Dennis & Whitaker, Woods & Teuber), but if you go back and read those carefully you will see that none of these studies were able to compare left- and right-hemisphere-damaged children directly. Either they looked only at children with left-hemisphere damage vs. controls, or they compared LHD and RHD to separate control groups without ever conducting a direct "head-on" comparison of the sort that would be required to conclude that one hemisphere is irreversibly superior to the other. At the very least, this conclusion suggests that there is no bounded and lateralized "organ" in the left hemisphere that is innately and irreversibly dedicated to grammar, and furthermore, parts of the brain that would ordinarily not be used to mediate grammatical processing can be encouraged to do so, quite successfully. 2) Recent neural imaging studies of normal adults have shown that the putative language areas of the left hemisphere are also involved in the mediation of non-linguistic processes. For example, studies from many different laboratories have shown that *ALL* of the candidate areas in and around the "Broca complex" are active in the planning and/or covert execution of non-linguistic mouth movements and/or non-symbolic movements of the hand. In other words, these areas may be involved in the processing of grammar, but they do a lot of other things as well, a point that is quite relevant to the notion of a "dedicated" neuro-cognitive system. Similar findings are beginning to pile up for the posterior language areas as well. 3) Another outcome of recent imaging studies is the now-widespread finding that many different parts of the brain are activated during language tasks, in patterns that vary depending on task demands, strategies, the expertise and experience of an individual subject. It may turn out that it makes no more sense to ask "where is language in the brain" than it does to ask "where is the dance in the dancer?" To be sure, it is the human brain (and only the human brain) that does language, but the case is not looking very good right now for the idea that it does so with a dedicated and well-defined language organ that has evolved to do language and language alone. -liz bates From tgivon at OREGON.UOREGON.EDU Fri Jul 30 19:14:27 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 12:14:27 -0700 Subject: What is emergence anyway? Message-ID: Liz addressed an important part of the question--whether there are or are not language-specific dedicated structure. I have no strong preconceptions on the subject (tho some weak biases...). But the question of "emergence" in the Hopper/MacWhinney sense (if I understand it) remains even if the structures are not language-specific: Automaticity during lifetime learning, through whatever neural structure, is the hallmark of all skilled acquisition. Grammar is a skilled performance PAR EXCELLENCE. So when it "emerges", does it acquire the (at least partial) rigidity & categoriality characteristics that have been shown in the acquisition of all other complex, rhythmic-hierarchic skilled behaviours? Here, I obviously have a VERY strong bias. But regardless of what ones bias is, we've got to find non-linguistic criteria for seeing whether the PROCESS of "emergence" does lead to the STATE of "having emerged", or not. I myself find the "not" option rather bizarre, given what is known about the expenditure of neuro-cognitive resources during development (both ontogenetic & phylogenetic). But one way or another, the issue needs to be resolved through making empirical predictions & then seeing if they pan out. TG Elizabeth Bates wrote: > > Talmy is right that it is important to find empirical tests that can > distinguish between grammatical outcomes that have a dedicated > neurocognitive structure and those that have instead emerged > across the course of learning and communication, supported by > neurocognitive mechanisms that have "still kept their day jobs" > (i.e. mechanisms that continue to do non-linguistic work for which > they evolved long before grammar ever appeared). It's not an easy > matter, but there is evidence on the point. Here are just a few > relevant bits: > > 1) Children who have early damage to the putative language organs > in the left hemisphere go on to acquire lexical and grammatical > abilities that are perfectly normal (i.e. they are not aphasic, > although as a group brain-damaged children tend to perform in > the low-normal range, corresponding to an approximate average > loss of 5 - 7 IQ points). Interesting, if these children are > studied after 5 years of age, there seems to be no evidence > whatsoever for differences as a function of early left- vs. > right-hemisphere damage. This conclusion now appears to be > on very solid ground, as several recent and comprehensive > reviews of the literature have shown (by our research group; > by Vargha-Khadem and colleagues; by Eisele and Aram). There > were some inklings in the 1970's that this might not be the > case (e.g. papers by Dennis & Whitaker, Woods & Teuber), but if you go > back and read those carefully you will see that none of these studies > were able to compare left- and right-hemisphere-damaged children > directly. Either they looked only at children with left-hemisphere > damage vs. controls, or they compared LHD and RHD to separate > control groups without ever conducting a direct "head-on" comparison > of the sort that would be required to conclude that one hemisphere > is irreversibly superior to the other. At the very least, this > conclusion suggests that there is no bounded and lateralized "organ" > in the left hemisphere that is innately and irreversibly dedicated > to grammar, and furthermore, parts of the brain that would ordinarily not > be used to mediate grammatical processing can be encouraged to do so, quite > successfully. > > 2) Recent neural imaging studies of normal adults have shown that > the putative language areas of the left hemisphere are also involved > in the mediation of non-linguistic processes. For example, studies > from many different laboratories have shown that *ALL* of the > candidate areas in and around the "Broca complex" are active in the > planning and/or covert execution of non-linguistic mouth movements > and/or non-symbolic movements of the hand. In other words, these > areas may be involved in the processing of grammar, but they do a lot > of other things as well, a point that is quite relevant to the > notion of a "dedicated" neuro-cognitive system. Similar findings > are beginning to pile up for the posterior language areas as well. > > 3) Another outcome of recent imaging studies is the now-widespread > finding that many different parts of the brain are activated during > language tasks, in patterns that vary depending on task demands, > strategies, the expertise and experience of an individual subject. > It may turn out that it makes no more sense to ask "where is language > in the brain" than it does to ask "where is the dance in the dancer?" > To be sure, it is the human brain (and only the human brain) that > does language, but the case is not looking very good right now for > the idea that it does so with a dedicated and well-defined language > organ that has evolved to do language and language alone. -liz bates From jtang at COGSCI.BERKELEY.EDU Fri Jul 30 19:56:31 1999 From: jtang at COGSCI.BERKELEY.EDU (Joyce Tang Boyland) Date: Fri, 30 Jul 1999 12:56:31 -0700 Subject: automatized or neurocognitively dedicated? Message-ID: I think some clarification is in order on what people mean by "dedicated neurocognitive structure". I can imagine interpretations that fit what Liz Bates is saying, and I can imagine others that fit what Tom Givon is saying, but I do not think that they are using the term in the same sense. Compare: Bates writes: Talmy is right that it is important to find empirical tests that can DISTINGUISH between grammatical outcomes that have a dedicated neurocognitive structure and those that have instead emerged across the course of learning and communication, supported by neurocognitive mechanisms that have "still kept their day jobs" (i.e. mechanisms that continue to do non-linguistic work for which they evolved long before grammar ever appeared). [emphasis mine --JTB] My reading of Givon is that he is *not* in fact making that distinction. Givon writes: a behavioral subsystem ...[might be] ... either "epiphenomenal" or "routinized/automated" (by the LATTER I now mean "having a dedicated neuro-cognitive structure"). [emphasis mine --JTB] and: [some behaviors become] ... automated (thus having "emerged", "grammaticalized" with a dedicated neuro-cognitive structure). The critical fact is that routinization/automatization is a prime instance of the various neurocognitive mechanisms that have "kept their day jobs". So the issue is that I think we need to decide whether we mean the kinds of neurocognitive structures that exist *regardless* of routinization through experience, or the kinds of neurocognitive structures that are the *result* of general mechanisms such as routinization through experience? Joyce Tang Boyland Alverno College Milwaukee, WI 53234-3922 From macw at CMU.EDU Fri Jul 30 22:37:46 1999 From: macw at CMU.EDU (Brian MacWhinney) Date: Fri, 30 Jul 1999 18:37:46 -0400 Subject: emergence Message-ID: Dear FunkNet, It was a busy day, or I would have chimed in earlier regarding this useful discussion of emergence. I found ?sten Dahl's note useful. In particular, I very much agreed with the idea that one tends to evaluate emergence differently according to whether one emphasizes the novelty of the emergent system or the degree to which its structure can be partially predicted from its predecessors. Wolfgang Schulze's claim that "subjecthood is the construction of emergent activities exerted by information flow (word order), NP semantics, case marking (if present), agreement (if present) and much more" is certainly a correct application of the notion of emergence to language, as Liz Bates and I argued in a series of papers in the 1980s. In those papers, we distinguished three levels of emergence: diachronic, acquisitional, and synchronic (sentence processing). More recently, I have found it useful to think in terms of five levels of emergence: 1. Evolutionary 2. Embryological 3. Developmental 4. On-line 5. Diachronic It seems to me that it is crucial to view emergence in terms of these varieties and to keep in mind the fact that the time scales for these varieties of emergence differ vastly. Simply listing these varieties of emergence does not answer Tom Givon's question. He wants to know which type of emergence is operative in a given behavior. To do that, we have to dig into the guts of specific models. For example, in their chapter in "The Emergence of Language" (Lawrence Erlbaum, 1999, B. MacWhinney (Ed.)), Prahlad Gupta and Gary Dell show that "a good lexicon" has words that share neighbors in the rime, but which differ in the onset. (CAT, BAT, SAT, RAT, MAT, FAT ..) They then proceed to show how this aspect of lexical structure arises directly from the Dell Speech Error model -- a data-driven connectionist model of speech production. What Tom wants to know, if I read him correctly, is whether the automated phonotactic processes that Gupta and Dell have studied are supported by simple general purpose learning mechanism or instead by some more dedicated neural hardware. The likely answer to this is that general purpose learning mechanisms are themselves based on specific neural adaptations. For example, Kohonen self-organizing feature maps demonstrate behaviors that emerge from the neural preference for lateral inhibition and short connections. These models can help explain the neurological grounding of the emergence of the preference for word onsets during development. Once models are formulated in this way, we start to realize that behaviors arise from an interaction of sources of emergence on varying time scales. Although automaticity is typically a fact of developmental emergence, it builds on embryologically-emergent structures and can be modified by on-line processes. The moral: the importance of emergence becomes clear once we begin to think in terms of models of specific behaviors. If we don't think in terms of concrete models, it seems like semi-TT gobbledy-gook. --Brian MacWhinney From tgivon at OREGON.UOREGON.EDU Sat Jul 31 00:10:31 1999 From: tgivon at OREGON.UOREGON.EDU (Tom Givon) Date: Fri, 30 Jul 1999 17:10:31 -0700 Subject: automatized or neurocognitively dedicated? Message-ID: I think Joyce is misinterpreting me. All I said is that there are two separate issues. And that they are distinct. One is about evolution, and Liz has certainly expressed her position about this. If I elected not to touch upon it at this point, it is not because I subscribe to Liz's (what seems to me...) rather extreme position (in essence, that language is too recent to have any genetic consequences; while memory & attention are old enough & have genetic consequences). The issue I did touch upon exists regardless of how you resolve the evolutionary/genetic issue. Whether the underlying mechanisms are language-specific or not, "emergence" during lifetime skilled learning either does or doesn't have neuro-cognitive consequences (the creation of automated structured during skill acquisition). My bias on this is obviously strong, as I have stated. I would like to find empirical means of resolving this. And I am still waiting either Brian or Paul to address the SECOND issue. TG Joyce Tang Boyland wrote: > > I think some clarification is in order on what people mean > by "dedicated neurocognitive structure". I can imagine > interpretations that fit what Liz Bates is saying, and I can > imagine others that fit what Tom Givon is saying, but I do not > think that they are using the term in the same sense. > Compare: > > Bates writes: > Talmy is right that it is important to find empirical tests that can > DISTINGUISH between grammatical outcomes that have a dedicated > neurocognitive structure and those that have instead emerged > across the course of learning and communication, supported by > neurocognitive mechanisms that have "still kept their day jobs" > (i.e. mechanisms that continue to do non-linguistic work for which > they evolved long before grammar ever appeared). [emphasis mine --JTB] > > My reading of Givon is that he is *not* in fact making that distinction. > > Givon writes: > a behavioral subsystem ...[might be] ... either "epiphenomenal" > or "routinized/automated" (by the LATTER I now mean > "having a dedicated neuro-cognitive structure"). [emphasis mine --JTB] > and: > [some behaviors become] ... automated (thus having "emerged", > "grammaticalized" with a dedicated neuro-cognitive structure). > > The critical fact is that routinization/automatization > is a prime instance of the various neurocognitive mechanisms > that have "kept their day jobs". > So the issue is that I think we need to decide > whether we mean the kinds of neurocognitive structures that > exist *regardless* of routinization through experience, or > the kinds of neurocognitive structures that are the *result* of > general mechanisms such as routinization through experience? > > Joyce Tang Boyland > Alverno College > Milwaukee, WI 53234-3922