From mcginnis at ucalgary.ca Tue Feb 13 15:28:12 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 13 Feb 2001 08:28:12 -0700 Subject: Heidi Harley: Idioms Message-ID: Dear DMers -- I heard Jim Higgenbotham talk on the English progressive last Friday, and he touched on the well-known phenomenon in which Vendler's 'achievement' classes get a pre-event focus in the progressive: 'she's winning the race', 'he was dying for three weeks before the end,' 'she's reaching the top'. I niggled at him with respect to the true punctuality of 'croak' as opposed to the potential for long-drawn-outness of 'die' (*He was croaking for 3 weeks before the end). (I was trying to suggest that 'die' was really an accomplishment, so he didn't need to treat the whole batch of examples he had with 'die' significantly differently than his accomplishment examples, and that true acheivements, like 'croak', didn't have this progressive problem. Or something like that). (Here's why I'm aware of this fact, and where it's been relevant to my life as DMer: The distinction between the apparent non-punctuality of 'die' vs. the punctuality of 'kill' has been raised as an argument against decomposing 'kill' as 'cause to die' -- it's either not decomposable (if you're a lexicalist) or (if you're a DMer) it decomposes into something more like 'cause to croak' -- not that the specific identity of the root is important, necessarily, but it's clear, anyway, that we wouldn't expect a non-punctual root inside a punctual agentive transitive. Or something. Hmm -- now that I think of it, maybe that's _not_ the point, and I've been misunderstanding the argument all along -- you can certainly get a pre-event progressive reading of 'kill': 'Smoke from the factory next door is killing me.' Help!) Anyway, Jim said, that's not an argument against the existence of the acheivement class; what it is, he said, is an indication that 'idioms' (of the acheivement class) have the special property of not allowing the pre-event reading in the progressive. Non-idiomatic acheivements allow it, said he, idiomatic ones don't; cf. *he was croaking/kicking the bucket for 3 weeks before the end. So, I tried a) to think of a 'non-idiomatic' achievement verb that doesn't allow the pre-event reading with -ing (i.e. it's like 'win' or 'die' in that it's not an obvious idiom, but like 'croak' in that doesn't allow a progressive). b) to think of an 'idiomatic' acheivement verb or VP that DOES allow the pre-event focus with -ing (i.e. it's like 'croak' in being an 'idiom' but like 'win' or 'die' in allowing a progressive) Maddeningly, none sprang to mind in either category. There were some interesting things that showed up when I started running down a mental list of varying interpretations of 'take', when the meaning is apparently achievement-y: 'take a picture' allows a pre-event progressive he's taking a picture 'take the cake' doesn't: *he's taking the cake 'take a minute' doesn't (maybe? compare 'a while"): the analysis of the sample is taking ??a minute/??an hour/a while. 'take a beating' does: the Ravens are taking a beating (mid-event) 'take a powder' doesn't (I think?): *the jailbird is taking a powder 'take into account' does: 'he's taking it into account' (mid-event) 'take his temperature': 'he's taking his temperature' (mid-event) 'take off': The plane's taking off (pre-event?) 'take a left': he's taking a left (mid-event) 'take a break': he's taking a break (mid-event) 'take stock' : he's taking stock (mid-event) Anyway, that's just to give you a flavor for the kind of thing I'm woq ndering about. If there really *is* a difference between achievements and accomplishments, then presumably the examples that allow -ing are accomplishment idioms, and the 'ing' isn't pre-event, but mid-event (that seems true for the examples with 'take' above, except for 'take a picture'). The question is, is there really a distinciton between idioms and non-idioms in this regard? If so, that means that 'take a picture' is *not* idiomatic, and we've got some kind of test for idiomhood that distinguishes idioms from non-idioms, which would be REALLY WEIRD, given DM's general assumptions. (Especially for monomorphemic ones like 'croak' vs 'die'. Thoughts pls? :) hh --------------------------------------------------------------------- Heidi Harley (520) 626-3554 Department of Linguistics hharley at u.arizona.edu Douglass 200E Fax: (520) 626-9014 University of Arizona Tucson, AZ 85721 From mcginnis at ucalgary.ca Tue Feb 13 16:12:38 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 13 Feb 2001 09:12:38 -0700 Subject: Dan Everett: Word formation constraints Message-ID: Folks, How does DM handle the kinds of adjectives that Lieber discusses in several places, e.g. 'She's an I-don't -care-who-said-it-I'm-going-anyway kind of jungle explorer'? LFG treats these as idioms, which I think is highly implausible. They seem quite productive. A related question, revealing even more of my ignorance of DM, is whether DM allos for word-formation which involves neither vocabulary items (as immediate consituents of the word) or head-movement. Such a case might arise, for example, in certain kinds of sentential predicates (a large class of these is discussed under 'quotatives' in my grammar of Wari', published by Routledge). Thanks for any help, Dan Everett Research Professor of Phonetics and Phonology University of Manchester From mcginnis at ucalgary.ca Wed Feb 14 16:36:48 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 14 Feb 2001 09:36:48 -0700 Subject: Andrew Carstairs-McCarthy: Word formation constraints (reply to Dan Everett) Message-ID: >Folks, > >How does DM handle the kinds of adjectives that Lieber discusses in >several places, e.g. 'She's an I-don't -care-who-said-it-I'm-going-anyway >kind of jungle explorer'? LFG treats these as idioms, which I think is >highly implausible. They seem quite productive. I've pondered these quite a lot for the purpose of teaching morphology and also while writing a textbook on English word-formation (which should come out from Edinburgh University Press this year). It seems to me that one needs a separate category of (what I call) 'phrasal words': complex items that function syntactically as words, yet whose internal structure is that of a clause or phrase rather than a compound. This distinction between structure and function echoes DiSciullo and Williams's distinction between syntactic atoms and morphological objects. It provides a handy way of handling the difference as regards plural formation between e.g. _bother-in-law_ and _jack-in-the-box_. The former, with plural _brothers-in-law_, is not a (compound) word at all (despite the hyphens in the spelling!), but an idiomatic phrase (or phrasal idiom). The latter, with plural _jack-in-the-boxes_, is a phrasal word (and an idiom too). Being an idiom, as I see it, is a matter of whether (or to what extent) the meaning of the whole is predictable from that of its parts. ('To what extent' implies that idiomaticity is a matter of degree, which seems correct to me.) That's independent of whether the kind of structure that the parts enter into is syntactic or morphological. So I agree with Dan that it is unsatisfactory just to toss phrasal words into an 'idiom' category. His example illustrates a phrasal word that, unlike _jack-in-the-box_, is not an idiom. I don't know that there's much here that bears on DM in particular -- except this, perhaps: even though the way in which word-internal structure is typically represented in DM discussions is via labelled branching trees, nevertheless this sort of morphological tree structure should not be too hastily equated with syntactic structure. Andrew -- Andrew Carstairs-McCarthy Professor and Acting Head of Department Department of Linguistics, University of Canterbury, Private Bag 4800, Christchurch, New Zealand phone (work) +64-3-364 2211; (home) +64-3-355 5108 fax +64-3-364 2969 e-mail a.c-mcc at ling.canterbury.ac.nz http://www.ling.canterbury.ac.nz/adc-m.html From mcginnis at ucalgary.ca Thu Feb 15 15:52:39 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 15 Feb 2001 08:52:39 -0700 Subject: Dan Everett: Word formation constraints (reply to Andrew Carstairs-McCarthy) Message-ID: Andrew's answer is interesting. He says he doesn't see the relevance to DM of sentences-as-adjectives in English. The relevance to DM, as I see it, is knowing how DM could generate these, what Andrew calls, 'phrasal' words. If the only means of word-formation are Vocabulary Item insertion and Head Movement, then I cannot see how phrasal words can be formed.If that is correct, then DM would thus still be bound by a rather strong version of the 'Lexicalist Hypothesis'. I am betting that DM has no easy solution and that the problem will turn out to be the exocentricity of phrasal words. That is, the immediate constituent of a 'phrasal word' doesn't match the label of the category under which it occurs in the tree (a sentence in Adjective position, for example will provoke severe complications in explaining how it got there - although Andrew Carnie's thesis offers some suggestions, where Carnie allows phrases under X0s). The specific problem I have in mind comes from Wari' quotatives, in which the quotative phrase functions as the verb of the sentence. This is documented in my chapter in the Handbook of Morphology and also in the Everett & Kern grammar of Wari from Routledge. There is strong evidence in Wari, however, that not even Carnie's proposal (form a phrase and move it into X0) will work, since movement is fairly carefully marked in Wari, but quotatives show no signs of movement. I am sure that this is all too sketchy to make a great deal of. But I am hoping to finish a paper on this in the next few weeks. The problem seems to be the need to recognize exocentric constructions motivated directly by the semantics, both requirements hard to state in formal theories (like DM). Well, anyway, that is what the paper claims. More on that when it is available. Back to Andrew C-M's proposal. On the other hand, I am not sure how a new category of 'phrasal word' would help at all. The concept needs to be formalized. Perhaps this is done in Andrew's forthcoming book.I look forward to reading it. Dan Everett Research Professor of Phonetics and Phonology University of Manchester From mcginnis at ucalgary.ca Thu Feb 15 21:32:57 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 15 Feb 2001 14:32:57 -0700 Subject: Heidi Harley: Word formation constraints (reply to Dan Everett) Message-ID: Hi all -- well, I'll take a stab at it, without having thought about it very much. These are certainly thornily tricky constructions... pls consider the following to be just the naivest possible response. >The relevance to DM, as I see it, >is knowing how DM could generate these, what Andrew calls, 'phrasal' >words. If the only means of word-formation are Vocabulary Item insertion >and Head Movement, then I cannot see how phrasal words can be formed. why not allow the syntax to generate phrases and merge them in 'head' positions, with permission slips for such activity issued on a language-by-language, construction-by-construction basis (e.g. via varying what types of categories can bear certain types of features)? As Dan notes, this is essentially what Andrew Carnie argued for Irish nominal predicates in his thesis. The hard problem for such a proposal is figuring out how to constrain it, of course, but letting the syntax generate phrases seems natural enough -- it's what it does. the trick is letting it do it in certain head positions. Then, presumably, vocab. insertion will proceed as usual, and phrasal phonology will as well, and you end up with a syntactically complex item that is doing the job of a 'word'. Note that DM has an advantage if this is really a possible account of such constructions, since it's a Late Insertion theory: the syntax can do all the dirty work and you don't have to replicate overtly phrasal syntactic rules in the lexicon just to construct these complex phrase-words. >If >that is correct, then DM would thus still be bound by a rather strong >version of the 'Lexicalist Hypothesis'. if i'm saying a reasonable thing, then it won't be so bound. >I am betting that DM has no easy >solution and that the problem will turn out to be the exocentricity of >phrasal words. That is, the immediate constituent of a 'phrasal word' >doesn't match the label of the category under which it occurs in the tree >(a sentence in Adjective position, for example will provoke severe >complications in explaining how it got there) right -- but i'm guessing that's a syntactic problem, not a morphological one. >The specific problem I have in mind comes from Wari' quotatives, in which >the quotative phrase functions as the verb of the sentence. This is >documented in my chapter in the Handbook of Morphology and also in the >Everett & Kern grammar of Wari from Routledge. There is strong evidence >in Wari, however, that not even Carnie's proposal (form a phrase and move >it into X0) will work, since movement is fairly carefully marked in Wari, >but quotatives show no signs of movement. I don't actually think that Carnie's proposal is that a phrase "moves into" an X0 slot and hence becomes an X0-- rather, the whole phrase, he says, _behaves_ like it's dominated by an X0 right from the get go (that is, the whole predicative NP bears verbal features). Then the whole thing head-moves to check those features, just like a regular V0 in Irish. But it's the fact that the nominal predicate can _bear_ such features in base position that allows the nominal predicate to head-move. so my guess is that movement isn't crucial: you ought to be able to generate a phrase that bears head features and have it merged into the syntactic structure just as if it was a word. i'll check with Andrew, who I don't think is a list member, and see if he can clarify a bit. >I am sure that this is all too sketchy to make a great deal of. But I am >hoping to finish a paper on this in the next few weeks. The problem seems >to be the need to recognize exocentric constructions motivated directly >by the semantics, both requirements hard to state in formal theories (like >DM). Well, anyway, that is what the paper claims. More on that when it is >available. looking forward to it! --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Fri Feb 16 17:58:53 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 10:58:53 -0700 Subject: Martha McGinnis: Idioms (reply to Heidi Harley) In-Reply-To: Message-ID: Heidi's right: if Higginbotham were correct that idiomatic VPs are aspectually different from non-idiomatic VPs, this would be a thorn in the side of DM. According to DM, both types of VPs are built in the syntax, so both should show the same 'compositional meaning.' Since aspectual meaning is clearly compositional, it should definitely allow the same possibilities for idioms and non-idioms. And, in fact, it does. I'm no aspect expert, but let's see if we can define the relevant verb classes. Accomplishments (1) and achievements (2) can be distinguished from activities (3) and states (4) by the fact that they're telic -- they allow endpoint-modification with "in an hour"-type adjuncts. (1) I climbed the mountain in an hour. ACCOMPLISHMENT (2) I recognized him in an instant. ACHIEVEMENT (3) *I climbed in an hour. ACTIVITY (4) *He was tall in an hour. STATE Achievements and accomplishments can be distinguished according to the interpretation they get with "for an hour"-type adjuncts. For accomplishments, the process preceding the endpoint is modified, as in (5). For achievements, if the modification is grammatical, it's not a pre-event modification but a result-modification, as in (6). (5) I climbed the mountain for an hour. ACCOMPLISHMENT (6) ?I recognized him for a moment. ACHIEVEMENT According to Vendler, accomplishments (7) are generally more compatible with the progressive in English than achievements (8), and activities (9) are more compatible with the progressive than states (10). ('Understand' is like 'recognize'... no doubt there are more.) (7) I'm climbing the mountain. (8) ??I'm recognizing him. (9) I'm climbing. (10)??He's being tall. However, there are some achievements which do allow the endpoint-modification (11), don't allow the pre-event modification interpretation with "for an hour"-type adjuncts (12), but are OK with the progressive (13). Let's call the first type of achievements Progless and the second type Progful. (11) He reached the top of the mountain in an hour (12) ??He reached the top of the mountain for an hour. (13) He's reaching the top of the mountain even as we speak. OK, so now we have a 5-way classification of VPs: states, activities, accomplishments, and progless and progful achievements. As expected, all classes contain idiomatic VPs as well as the non-idiomatic ones listed above. States: *in an hour, *progressive (14) He's the cat's pyjamas. (15) *He was the cat's pyjamas in an hour. (16)??He's being the cat's pyjamas. Activities: *in an hour, progressive (17) He pulls my leg constantly. (18) *He pulled my leg in an hour. (19) He's pulling my leg again! Accomplishments: in an hour, for an hour, progressive (20) He showed me the ropes in a couple of days. (21) He showed me the ropes for a couple of days. (22) He's showing me the ropes. Progful achievements: in an hour, *for an hour, progressive (23) I found my feet in a month. (24)??I found my feet for a month. (25) I'm still finding my feet. Others like this: cross the Rubicon, earn X's wings, make the grade, give up the ghost... Progless achievements: in an hour, *for an hour, *progressive (26) He croaked/kicked the bucket in three hours. (27) *He croaked/kicked the bucket for three hours. (28)??He's croaking/kicking the bucket. These are aspectually punctual, like their non-idiomatic counterparts, but they don't allow an iterative reading, for obvious reasons (a frog can croak again, a corpse can't). I think both the progressive and for-modification become OK if we talk about a group: (27)' The nearby animals were croaking for weeks following the Chernobyl meltdown. So there's no need for DM proponents to scurry around trying to find an explanation for why idioms aren't aspecually like non-idioms... they are. mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Fri Feb 16 17:59:47 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 10:59:47 -0700 Subject: Andrew Carstairs-McCarthy: Word formation constraints (reply to Dan Everett) Message-ID: Dan said, in reply to me: >Back to Andrew C-M's proposal [that e.g. _couldn't-care-less_ in _a >couldn't-care-less attitude_ should be classified as a phrasal word, >i.e. not some kind of compound but a phrase, yet functioning as a >word]. ... I am not sure how a new category of 'phrasal word' would >help at all. The concept needs to be formalized. Perhaps this is >done in Andrew's forthcoming book. I look forward to reading it. Well, I feel a bit sheepish now at having put in such a shameless plug for my forthcomingy book. It is an *introductory* book, and is aimed at English majors and trainee language teachers rather than budding linguists, so, although it is 'formal' in the sense that I aim for a clear and coherent set of definitions of technical terms, it is not theoretically adventurous. I merely use contrasts like _halfbrother_ vs _brother-in-law_ vs _jack-in-the-box_ to justify a distinction between compounds, phrases and phrasal words, this distinction being independent of that between idioms and non-idioms ... >The problem seems >to be the need to recognize exocentric constructions motivated directly by >the semantics, both requirements hard to state in formal theories (like >DM). Well, anyway, that is what the paper claims. More on that when it is >available. ... but I do venture into controversy, perhaps, when I discuss such expressions as _American history teacher_ with its two interpretations. Are there two constructions, each one motivated by its own semantics (as Dan might put it), or is there just one construction, with two interpretations nevertheless available? A traditional view is that there are two structures: (1) [[American history] teacher] 'teacher of Am hist' (2) [American [history teacher]] 'American teacher of hist' But if this is correct, then the structure at (1) ought also to be available for e.g. (3) and (4): (3) [[interesting history] teacher] 'teacher of interesting history' (4) [[suburban history] teacher] 'teacher of suburban history' But this seems incorrect. (3) and (4) can only mean 'interesting/suburban teacher of history'. That suggests to me that what makes the interpretation at (1) possible is not that the bracketing at (1) is made available by the syntax/morphology (with a N-bar inside a N -- a bit unwelcome!), but rather that the status of American history as an *institutionalized* subfield of history can force the interpretation at (1) on to the structure at (2). This analysis is heavily influenced by Andy Spencer's (in my view) excellent article on bracketing paradoxes in Language 1988. None of the structures at (1)-(4) is exocentric, under any plausible bracketing, so they may seem irrelevant to Dan's point. But I think they are relevant to the wider issue of the extent to which 'formal' grammatical (morphological or syntactic) structure determines interpretation, even with expressions such as (1) and (2) that most of us would hesitate to call idiomatic. -- Andrew Carstairs-McCarthy Professor and Acting Head of Department Department of Linguistics, University of Canterbury, Private Bag 4800, Christchurch, New Zealand phone (work) +64-3-364 2211; (home) +64-3-355 5108 fax +64-3-364 2969 e-mail a.c-mcc at ling.canterbury.ac.nz http://www.ling.canterbury.ac.nz/adc-m.html From mcginnis at ucalgary.ca Fri Feb 16 18:00:18 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 11:00:18 -0700 Subject: Alec Marantz: Word formation constraints (reply to Dan Everett) Message-ID: I must confess that I don't understand the connection between Dan's real worry -- what's the proper analysis of Wari' quotatives -- and DM (as Andrew C-M also points out). I think I agree with him that to the extent that there's a special problem, it's a problem for any theory (i.e., for what Dan is calling a "formal" theory, although I fail to see the difference here between "theory" and "formal theory"). There is a confusion about the role of node labels. I, for one, believe that the Chomskian program can do without node labels entirely. But, in any case, the first member of a compound noun in English can be complex (glass flower case, etc.) and it's clear that there is no category constraint on the first member of the compound. Thus, I believe, it would be hard for DM or any other theory to keep phrases out of this position. The problem is accounting for the interpretation on the one hand and the restriction on complements on the other. So, putting aside the "phrasal" cases like "keep off the grass sign," when the first member of the compound (or when an adjectival modifier in the same structural position) has a complement, the structure is bad (compare, "vest pocket kerchief" with, "*pocket of a vest kerchief" and recall, "*afraid of dogs man" vs., "man afraid of dogs"). When phrases like "keep off the grass" are treated as single units, they don't fall under the "no complement" generalization even though they contain complements. So in a clear sense they do behave like simple heads as opposed to complex heads (or phrases). Dan challenges us to find an account of this. Contrary to Dan, I believe that current "formal" theories are precisely addressing the relevant issues concerning the relation between the syntax and semantics of structures. That doesn't mean formal theories currently have an answer to the question Dan raises, although I agree with Heidi that Carnie's approach is closer to an answer than Dan seems to believe. marantz at mit.edu From mcginnis at ucalgary.ca Fri Feb 16 18:00:41 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 11:00:41 -0700 Subject: Dan Everett: Word formation constraints (reply to Heidi Harley) Message-ID: Heidi's solution to my query, that it is primarily a syntactic problem, may in fact be correct. However, this itself raises problems for the syntax. I will just mention what I think the problem is. It is not directly relevant for DM but it is indirectly relevant to the degree that DM has an opinion about what syntactic theory its own apparatus follows. Assume some version of the MP. Andrew Carnie's solution in Irish will work in MP and DM and I think it is quite a reasonable one. However, it works in the MP because the syntax can form a phrase and then MOVE it into the appropriate and, this is crucial, preexisting (by Select) morphological position (and it can do this only in the MP,among formal theories, because other theories crucially distinguish bar-levels, certainly this is true of LFG and, I believe, it is still true for HPSG). However, in the absence of movement, the only way to embed a phrase into a morphological position is via MERGE. But Merge has problems here. Merge can only take place if a node is licensed for it (in the absence of base-generated trees, which are eliminated in the MP). The V node, so far as I can tell, is not really licensed by anything but the V itself (although I think some may have proposed having the Tense-node license/select it and John Bowers has argued for a Pr0 node - but that will still only work if the Pr0 node is lexical, I think). So, if there is no V, there doesn't seem to be any way to Merge a phrase into it (since the phrase is not in the Array and cannot be chosen by Select). That means that if one could show a case which needed to be analyzed as Merge into V (though I think, sigh, that it is probably impossible to conclude that the MP 'needs' to analyze anything in any particular way) that should be a problem for the MP and, thus, to DM if MP is its 'feeder syntax module'. But note that if one DOES have Merge into X0 categories (in spite of what I think the likelihood of MP being able to do this) it also provides a solution to the phrasal words/adjectives I brought up initially. Anyway, I have no doubt but that Heidi's proposal is the way to go about things wrt this type of phenomenon (which I believe to be fairly rich, cross-linguistically). The question is to find the best theory to do this in. -- Dan From mcginnis at ucalgary.ca Tue Feb 20 15:55:21 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 08:55:21 -0700 Subject: Heidi Harley: Idioms (reply to Martha McGinnis) Message-ID: Hi all -- Martha's excellent examples and discussion do go a long way to allaying my concerns. Especially the '??he's recognizing me' class of progless non-idioms is something that ought to confound Jim H., I think. However, I do have a concern with the examples of progful idioms that Martha gives, which will take a minute to lay out. The characterization of accomplishment verbs as taking a 'for' adverbial is incomplete -- the 'climb the mountain' accomplishments do certainly allow it, but there's some that canonically don't: #I built the house for an hour. #I ate the cookie for 5 minutes. and are usually remarked upon as only allowing the 'for' adverbials when the object is in 'partitive' case, or, in English, modified by a PP: I built on the house for an hour. I ate at the cookie for 5 mins. (The 'climb the mountain' example is like 'run the race': a canonical activity verb with an additional delimiting object argument -- and of course activity verbs allow 'for' modification. These 'build' and 'eat' verbs require objects, though.) So the question is, how do we distinguish between the 'build the house' accomplishments, which don't allow the 'for' modification, and the progful achievements? The difference between 'he's building the house' and 'he's winning the race' seems to be that the 'building the house' is mid-event -- you're partly completed building the house -- while 'winning the race' is pre-event: you're not partway through winning the race, rather, you're in the lead at the moment, but the instant of winning is at the finish line, and you can't have the race half-won, although you can have the house half- built. (don't say, what about a house-building race?) So to get back to the idiom question, are the examples that Martha gives as being progful acheivements, because they don't accept 'for'-adverbials -- are they like 'build the house' accomplishments or like 'win the race' achievements? >(23) I found my feet in a month. >(24)??I found my feet for a month. >(25) I'm still finding my feet. it seems to me that 'finding my feet' is in some sense 'mid-event' -- that is, you can be halfway along in finding your feet, unlike winning a race. If that's right, it's an idiom of the 'build' class, not of the 'win' class. Of the other ones that Martha suggests are progful achievements, 'cross the Rubicon' and 'earn X's wings' seem to me to be more like the 'build the house' accomplishments, not like the 'win the race' achievements. For me, however, 'make the grade' and 'give up the ghost...' are more acheivement-like, intuitively. But, maddeningly, 'give up the ghost' doesn't seem to go as well with progressives -- or insofar as it does, it doesn't mean 'pre-event', it means during the event, at the instant -- like 'croak' or 'kick the bucket': ?He's giving up the ghost. (*He was giving up the ghost for 2 hours before the end.) 'He's making the grade,' though, doesn't seem bad, and it does seem pre-event (I don't think you can have made the grade just halfway, right?) So I think this is, I hope, I hope, an example of a progful acheivement idiom. thoughts? Kate Kearns also sent me a note pointing out among other things that one of the properties of achievement progressives is that they don't accept modification by 'still', plus a couple of other things. Here's the relevant citation from Kate's note: "Anita Mittwoch's (1991) paper 'In defence of Vendler's achievements', Belgian Journal of Linguistics vol 6, pp. 71-85. She talks about the 'die/win the race' type, and says that their progressives aren't like real process progressives because they can't be modified by continuative 'still' or be in complements to 'stop, start, continue, keep' and so on." I still haven't gone to find the paper and check out her examples, but I thought the list might be interested. (The reason i wanted to suggest to Jim that there's no significant distance between accomplishments of the 'build' class and achievements is to cut through this whole tangled question; they behave the same w/r to time adverbials, and their interpretations with progressives are darn slippery. but for 'die' and 'win' and 'reach the top' the pre-event vs. mid-event distinction does seem fairly robust). Anyway, any thoughts? maybe i'm missing something obvious. and the judgments are kind of subtle. there might be better tests out there. Marf? prove me wrong, please! and thanks v. much for the input, hh --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Tue Feb 20 17:25:58 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 10:25:58 -0700 Subject: Martha McGinnis: Idioms (reply to Heidi Harley) In-Reply-To: Message-ID: >So the question is, how do we distinguish between >the 'build the house' accomplishments, which don't >allow the 'for' modification, and the progful >achievements? This sounds like a question for Higginbotham: wasn't he the one claiming that they should be distinguished? However, it looks like there's an answer: >"Anita Mittwoch's (1991) paper 'In defence of >Vendler's achievements', Belgian Journal of Linguistics >vol 6, pp. 71-85 ... talks about the 'die/win the race' type, >and says that their progressives aren't like real >process progressives because they can't be modified by >continuative 'still' or be in complements to 'stop, >start, continue, keep' and so on." OK, so now we have SIX verb classes -- call them states, activities, delimited activities, accomplishments, progful achievements, and progless achievements. The claim still holds, though: idiomatic and non-idiomatic VPs have the same range of aspectual possibilities. Let's focus on the classes I may have confounded earlier: Delimited activities: in an hour, for an hour, progressive, stop Ving Non-idiomatic: climb the mountain (1) He showed me the ropes in a couple of days. (2) He showed me the ropes for a couple of days. (3) He's showing me the ropes. (4) He stopped showing me the ropes when Sue arrived. Accomplishments: in an hour, *for an hour, progressive, stop Ving Non-idiomatic: build the house (5) I went to pieces in minutes. (6)??I went to pieces for minutes. (OK only with result-modification) (7) I'm going to pieces again. (8) I stopped going to pieces when Sue arrived. Others like this: make the grade, fall apart, lose one's marbles Progful achievements: in an hour, *for an hour, progressive, *stop Ving Non-idiomatic: win the race (9) I found my feet in a month. (10)??I found my feet for a month. (OK only with result-modification) (11) I'm still finding my feet. (12) *I stopped finding my feet when Sue arrived. Others like this: cross the Rubicon, bury the hatchet I'm afraid I have no insights at all regarding Heidi's Big Question -- i.e., whether these six should be primitive aspectual classes, or just six of many, or six permutations of a smaller number of primitive classes. But at least I think it's clear that idioms aren't aspectually 'special'. Cheers, Martha mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Tue Feb 20 21:09:46 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 14:09:46 -0700 Subject: Heidi Harley: Idioms (reply to Martha McGinnis) Message-ID: Hi all -- >This sounds like a question for Higginbotham: wasn't he the one >claiming that they should be distinguished? yes -- and he pointed to the 'pre-event' vs 'mid-event' readings of the progressive in each case as evidence that they are significantly different. The paper of Mittwoch's that Kate Kearns sent me the reference to, though, also claims that they must be distinguished, hence the title -- and given your results with her 'stop' test, let's assume she (and Jim) are right. >>"Anita Mittwoch's (1991) paper 'In defence of >>Vendler's achievements', Belgian Journal of Linguistics >>vol 6, pp. 71-85 ... talks about the 'die/win the race' type, >>and says that their progressives aren't like real >>process progressives because they can't be modified by >>continuative 'still' or be in complements to 'stop, >>start, continue, keep' and so on." >Accomplishments: in an hour, *for an hour, progressive, stop Ving >Non-idiomatic: build the house > >(5) I went to pieces in minutes. >(6)??I went to pieces for minutes. (OK only with result-modification) >(7) I'm going to pieces again. >(8) I stopped going to pieces when Sue arrived. > >Others like this: make the grade, fall apart, lose one's marbles > >Progful achievements: in an hour, *for an hour, progressive, *stop Ving >Non-idiomatic: win the race > >(9) I found my feet in a month. >(10)??I found my feet for a month. (OK only with result-modification) >(11) I'm still finding my feet. >(12) *I stopped finding my feet when Sue arrived. > >Others like this: cross the Rubicon, bury the hatchet These are great, although I'm definitely going to have to go check out Mittwoch, since "still" ought not to be compatible with real acheivements according to Kearns' short description. But the 'stop'-test looks like it works as advertised. >I'm afraid I have no insights at all regarding Heidi's Big Question >-- i.e., whether these six should be primitive aspectual classes, or >just six of many, or six permutations of a smaller number of >primitive classes. But at least I think it's clear that idioms >aren't aspectually 'special'. hooray, thanks. So, I think then that Jim's got a problem with progless acheivements; his account of the semantics of the progressive only allows for progful ones, and he can't point to 'idiomaticity' as artificially preventing progfulness. I'm not really hoping for an answer to the Big Question at this point; I do ultimately think/hope that most of these distinctions can be taken care of in the syntax, with the merest handful of 'primitive' distinctions at issue (cf. e.g my paper on denominal verbs and aktionsart in the 2nd MIT/Upenn book), but the main point at hand was whether or not there were such things as progful idioms -- if there really weren't, that'd be trouble! another interesting question is whether or not the 'idiomatic' item always has the same basic aktionsart as the literal reading of the verb or frame from which it's derived. it's not at first obviously true for 'croak', because you can only croak idiomatically once -- but as long as we say that THAT distinction between idiomatic croaking and literal croaking is part of the Encylcopedia, then both idiomatic croaking and literal croaking are punctual (instantaneous, in the relevant sense). So maybe (and hopefully, for a morphosyntactic theory of aktionsart) it is true. Anyway, that's probably a question for another day. thanks very much, martha, for your tremendous examples and discussion! best, hh >Cheers, >Martha > > >mcginnis at ucalgary.ca --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Wed Feb 21 16:02:11 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 09:02:11 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz) Message-ID: Folks, I have been traveling. Before I attempt an answer in more detail to Alec's posting, let me say what the difference is between a formal theory and a 'nonformal theory'. A formal theory is, like DM, MP, etc, a theory in which FORM drives the model, rather than meaning. That is, meaning is not directly causally implicated in the model. In this sense. DM, like MP, are structuralist theories. Generative semantics, on the other hand, would not have been a formal theory in this sense (many formalists seem to confuse 'formal' with 'explicit', well-defined, generative, etc. But that is a mistake, historically and etymologically. Structuralist and formalist/formal are synonyms in this, fairly wide spread and traditional, understanding (hence I am puzzled as to why it would be hard for Alec to see this). In any case, my 'real concern' is not Wari' quotatives. That is merely a very good example of the problem. My real concern is the problem that such examples raise. I have not yet seen in Andrew Carnie's thesis, for example (but I am working my way through it slowly, so if it is there, my apologies to Andrew), an attempt to ask the question of where the predicate node arises for the S to move into. It is just there in the phrase structure. And Alec makes no attempt to answer my question either. He merely restates Heidi's proposal, namely, that Andrew's thesis probably provides the answer. As I said in response to Heidi, this may in fact be true. But there are technical problems and I, as an outsider, am merely trying to see how DM would answer it. It is interesting that no one has yet tried to, not even in response to my last message. However, I have a long list of emails following this one, so perhaps someone will. Cheers, Dan Everett From mcginnis at ucalgary.ca Wed Feb 21 16:02:44 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 09:02:44 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz cont'd) Message-ID: Folks, One other thing about Alec's posting, since no one else tried to answer my technical question. Contrary to Alec's suggestion, there is no confusion about node labels, so far as I can see. I am quite able to understand the MP reasoning behind the suggestion that nodes have no labels. Since there is no base component and since nodes are not so labeled in the lexicon, ex hypothesi, there is no way to have labeled nodes in the syntax, at least until after Merge and the postsyntactic DM work. Na~o ha' problema. On the other hand, I must admit that I *am* confused about node origins when those nodes do not dominate anything. Take movement of a phrase into the head of TP, for example. Why is the head/phrase there to serve as a target for movement? Either a node is required for lexical reasons or it is not there. What requires T? If the node is required by some aspect of LF, then the relevant constraint is global (not unusual for MP, since, as I pointed out in my Lg review of Hornstein's LF book sometime ago, MP has all possible constraints: global and violable, global and inviolable, local and violable and local and inviolable). But globality seems to be an embarassment here. Surely GREED or something in that spirit is more desirable, i.e. a local lexical requirement. I am not claiming anything here about the plausibility or possibility of a structuralist analysis of the facts. I am simply saying that I do not know how MP (and hence, I believe ultimately DM) can have a node for a phrase to move into unless that target node is first itself lexically required, e.g. by a null item projecting it). I don't think that the solution is going to be via MERGE, but perhaps there is something along those lines I have missed. If people think that this is a syntactic question, not a morphological question, my response is "I thought DM said that morphological structure = syntax". -- Dan From mcginnis at ucalgary.ca Wed Feb 21 18:00:31 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 11:00:31 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz) In-Reply-To: Message-ID: OK, I've finally had a chance to read this thread carefully, and I find I have nothing to add to Heidi's remark that DM looks like a good theory to be adopting to try to account for these sentential adjuncts, since DM doesn't allow us to sweep their unusual 'atomic' behaviour under an all-purpose Lexicon. With luck the comparison with Wari' quotatives will shed some new light on the phenomenon. Nevertheless, I thought I would comment on the discussion of DM, MP, etc., in case another interpretation is helpful. >A formal theory is, like DM, MP, etc, a theory in >which FORM drives the model, rather than meaning. That is, meaning is not >directly causally implicated in the model. In this sense. DM, like MP, are >structuralist theories. There may be similarities between structuralism and generative linguistics (e.g. the focus on systematic properties of language, rather than on prescriptive quirks, metaphor, etc.), but there's an important distinction to be made here as well. DM and MP, like other generative theories, are concerned with systematic correspondences between sound and meaning. If I recall correctly, structuralists were convinced that language could be captured simply in terms of phonologically-based patterns, without reference to meaning. Thus structuralism treats different languages as internally coherent but fundamentally distinct systems. Generative linguistics treats them as different states of the language faculty, whose lexicons draw on a universal set of semantically based syntactic features (tense, aspect, definiteness, specificity...), as well as phonetically based phonological features. The heyday of structuralism is far enough back in history now that I know of few explicit arguments against it. One is Morris Halle's argument against the structuralist notion of 'phoneme'. I also seem to recall that Halle & Marantz argue against a structuralist approach to morphology in their first DM paper -- i.e. that 'position classes' based purely on morphological distribution don'tt capture the correct generalizations about Potawatomi morphology. >I have not yet seen in Andrew Carnie's thesis, for example >(but I am working my way through it slowly, so if it is there, my >apologies to Andrew), an attempt to ask the question of where the >predicate node arises for the S to move into. It is just there in the >phrase structure. I'm not 100% sure I understand the question, but the issue here might be the order of assembly of syntactic structures. It's not possible to assume that the syntax composes a single tree, step by step, moving upwards (or downwards, for that matter). For example, phrasal specifiers have to be pre-assembled in a subtree before they can merge with the main tree. >Since there >is no base component and since nodes are not so labeled in the lexicon, ex >hypothesi, there is no way to have labeled nodes in the syntax, at least >until after Merge and the postsyntactic DM work. I think there's some flexibility here. Virtually everyone assumes the lexicon contains items of different categories, which have different syntactic properties (both for Merge and Move). Since introducing Bare Phrase Structure, Chomsky has assumed that phrasal units also have category labels, whose X-bar level can be read off the structure. Alec suggests that the syntax can get away without making use of these category labels. This sounds to me like a new suggestion -- that labels need not be stipulated, but can be read off the structure. By contrast, BPS uses different types of node labels to distinguish between specifiers and adjuncts. >On the other hand, I must admit that I *am* confused about node origins >when those nodes do not dominate anything. Take movement of a phrase into >the head of TP, for example. Why is the head/phrase there to serve as a >target for movement? Either a node is required for lexical reasons or it >is not there. What requires T? Perhaps nothing, which is why we can just say "Gone fishin'." The notion that Merge has to be licensed by something also sounds to me like a new proposal. As far as I understand, Chomsky treats Merge as free: it either does or doesn't succeed in generating semantically interpretable structures. It's possible that arguments are forced to Merge by a local need to satisfy the theta-requirements of a predicate, but even if so, this would not entail that there is no unforced Merge. What empirical benefits do we reap by taking the position that Merge is always forced or licensed? >But globality seems to be an >embarassment here. Surely GREED or something in that spirit is more >desirable, i.e. a local lexical requirement. As you know, 'Suicidal Greed' is local, but Greed and Procrastinate are the classic global constraints. Am I allowed to say 'classic' about the 1990's yet? Cheers, Martha mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Thu Feb 22 15:35:34 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 22 Feb 2001 08:35:34 -0700 Subject: Dan Everett: Word formation constraints (reply to Martha McGinnis) Message-ID: Martha, Good attempts at answers. I enjoyed reading them. I am afraid that I recognize little of structuralism in your account of it however. Generative linguistics seems structuralist in exactly the same ways as, say, Tagmemics, except more so (because Tagmemics was not only the very first theory, long before Morris's arguments against the 'phoneme' - which, by the way, have been shown to be wrong: the major lesson of Lexical Phonology is that something like the 'taxonomic phoneme' is a necessity in linguistic theory, but Tagmemics also allowed semantics and pragmatics to play a causal role in syntax. Tagmemics is not that great a theory. But it is less structuralist than MP). All structuralists I have read (and met) said things like, "Of course languages are more alike than they are different or we couldn't have a theory of Language". Structuralists believe(d) in universals (Greenberg, not Chomsky, first drew our attention to them; structuralists are mainly responsible for our knowledge of 'exotic' languages, because, like Mead, Benedict, Sapir, and other students of Boas, they believed that there are universal patterns of language that need to first be documented and then explained). But the real hallmark of structuralism is that the only solutions that count are distributional and structural. And this has ALWAYS been true of generative linguistics. We all know that Noam, to take a random example, accepts x as a solution to y just in case x is stated in terms of structural relationships. That is basically the motivation for MP. As I recall from lectures back in the mid 80s Chomsky was getting fed up with the multiplication of entities beyond necessity. He wanted to get back to the basics, which MP is supposed to be. And the basics are trees and a lexicon. No indices. No meaning intrusions (e.g. 'affectedness', 'achievements' vs. 'accomplishments', etc.) These things are all to be derived. And derived by trees. That just IS structuralism, which Zellig Harris taught at Penn. As to the specific questions I raised. 'Gone fishing' is the answer I will take away from this discussion. Best, Dan From mcginnis at ucalgary.ca Tue Feb 13 15:28:12 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 13 Feb 2001 08:28:12 -0700 Subject: Heidi Harley: Idioms Message-ID: Dear DMers -- I heard Jim Higgenbotham talk on the English progressive last Friday, and he touched on the well-known phenomenon in which Vendler's 'achievement' classes get a pre-event focus in the progressive: 'she's winning the race', 'he was dying for three weeks before the end,' 'she's reaching the top'. I niggled at him with respect to the true punctuality of 'croak' as opposed to the potential for long-drawn-outness of 'die' (*He was croaking for 3 weeks before the end). (I was trying to suggest that 'die' was really an accomplishment, so he didn't need to treat the whole batch of examples he had with 'die' significantly differently than his accomplishment examples, and that true acheivements, like 'croak', didn't have this progressive problem. Or something like that). (Here's why I'm aware of this fact, and where it's been relevant to my life as DMer: The distinction between the apparent non-punctuality of 'die' vs. the punctuality of 'kill' has been raised as an argument against decomposing 'kill' as 'cause to die' -- it's either not decomposable (if you're a lexicalist) or (if you're a DMer) it decomposes into something more like 'cause to croak' -- not that the specific identity of the root is important, necessarily, but it's clear, anyway, that we wouldn't expect a non-punctual root inside a punctual agentive transitive. Or something. Hmm -- now that I think of it, maybe that's _not_ the point, and I've been misunderstanding the argument all along -- you can certainly get a pre-event progressive reading of 'kill': 'Smoke from the factory next door is killing me.' Help!) Anyway, Jim said, that's not an argument against the existence of the acheivement class; what it is, he said, is an indication that 'idioms' (of the acheivement class) have the special property of not allowing the pre-event reading in the progressive. Non-idiomatic acheivements allow it, said he, idiomatic ones don't; cf. *he was croaking/kicking the bucket for 3 weeks before the end. So, I tried a) to think of a 'non-idiomatic' achievement verb that doesn't allow the pre-event reading with -ing (i.e. it's like 'win' or 'die' in that it's not an obvious idiom, but like 'croak' in that doesn't allow a progressive). b) to think of an 'idiomatic' acheivement verb or VP that DOES allow the pre-event focus with -ing (i.e. it's like 'croak' in being an 'idiom' but like 'win' or 'die' in allowing a progressive) Maddeningly, none sprang to mind in either category. There were some interesting things that showed up when I started running down a mental list of varying interpretations of 'take', when the meaning is apparently achievement-y: 'take a picture' allows a pre-event progressive he's taking a picture 'take the cake' doesn't: *he's taking the cake 'take a minute' doesn't (maybe? compare 'a while"): the analysis of the sample is taking ??a minute/??an hour/a while. 'take a beating' does: the Ravens are taking a beating (mid-event) 'take a powder' doesn't (I think?): *the jailbird is taking a powder 'take into account' does: 'he's taking it into account' (mid-event) 'take his temperature': 'he's taking his temperature' (mid-event) 'take off': The plane's taking off (pre-event?) 'take a left': he's taking a left (mid-event) 'take a break': he's taking a break (mid-event) 'take stock' : he's taking stock (mid-event) Anyway, that's just to give you a flavor for the kind of thing I'm woq ndering about. If there really *is* a difference between achievements and accomplishments, then presumably the examples that allow -ing are accomplishment idioms, and the 'ing' isn't pre-event, but mid-event (that seems true for the examples with 'take' above, except for 'take a picture'). The question is, is there really a distinciton between idioms and non-idioms in this regard? If so, that means that 'take a picture' is *not* idiomatic, and we've got some kind of test for idiomhood that distinguishes idioms from non-idioms, which would be REALLY WEIRD, given DM's general assumptions. (Especially for monomorphemic ones like 'croak' vs 'die'. Thoughts pls? :) hh --------------------------------------------------------------------- Heidi Harley (520) 626-3554 Department of Linguistics hharley at u.arizona.edu Douglass 200E Fax: (520) 626-9014 University of Arizona Tucson, AZ 85721 From mcginnis at ucalgary.ca Tue Feb 13 16:12:38 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 13 Feb 2001 09:12:38 -0700 Subject: Dan Everett: Word formation constraints Message-ID: Folks, How does DM handle the kinds of adjectives that Lieber discusses in several places, e.g. 'She's an I-don't -care-who-said-it-I'm-going-anyway kind of jungle explorer'? LFG treats these as idioms, which I think is highly implausible. They seem quite productive. A related question, revealing even more of my ignorance of DM, is whether DM allos for word-formation which involves neither vocabulary items (as immediate consituents of the word) or head-movement. Such a case might arise, for example, in certain kinds of sentential predicates (a large class of these is discussed under 'quotatives' in my grammar of Wari', published by Routledge). Thanks for any help, Dan Everett Research Professor of Phonetics and Phonology University of Manchester From mcginnis at ucalgary.ca Wed Feb 14 16:36:48 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 14 Feb 2001 09:36:48 -0700 Subject: Andrew Carstairs-McCarthy: Word formation constraints (reply to Dan Everett) Message-ID: >Folks, > >How does DM handle the kinds of adjectives that Lieber discusses in >several places, e.g. 'She's an I-don't -care-who-said-it-I'm-going-anyway >kind of jungle explorer'? LFG treats these as idioms, which I think is >highly implausible. They seem quite productive. I've pondered these quite a lot for the purpose of teaching morphology and also while writing a textbook on English word-formation (which should come out from Edinburgh University Press this year). It seems to me that one needs a separate category of (what I call) 'phrasal words': complex items that function syntactically as words, yet whose internal structure is that of a clause or phrase rather than a compound. This distinction between structure and function echoes DiSciullo and Williams's distinction between syntactic atoms and morphological objects. It provides a handy way of handling the difference as regards plural formation between e.g. _bother-in-law_ and _jack-in-the-box_. The former, with plural _brothers-in-law_, is not a (compound) word at all (despite the hyphens in the spelling!), but an idiomatic phrase (or phrasal idiom). The latter, with plural _jack-in-the-boxes_, is a phrasal word (and an idiom too). Being an idiom, as I see it, is a matter of whether (or to what extent) the meaning of the whole is predictable from that of its parts. ('To what extent' implies that idiomaticity is a matter of degree, which seems correct to me.) That's independent of whether the kind of structure that the parts enter into is syntactic or morphological. So I agree with Dan that it is unsatisfactory just to toss phrasal words into an 'idiom' category. His example illustrates a phrasal word that, unlike _jack-in-the-box_, is not an idiom. I don't know that there's much here that bears on DM in particular -- except this, perhaps: even though the way in which word-internal structure is typically represented in DM discussions is via labelled branching trees, nevertheless this sort of morphological tree structure should not be too hastily equated with syntactic structure. Andrew -- Andrew Carstairs-McCarthy Professor and Acting Head of Department Department of Linguistics, University of Canterbury, Private Bag 4800, Christchurch, New Zealand phone (work) +64-3-364 2211; (home) +64-3-355 5108 fax +64-3-364 2969 e-mail a.c-mcc at ling.canterbury.ac.nz http://www.ling.canterbury.ac.nz/adc-m.html From mcginnis at ucalgary.ca Thu Feb 15 15:52:39 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 15 Feb 2001 08:52:39 -0700 Subject: Dan Everett: Word formation constraints (reply to Andrew Carstairs-McCarthy) Message-ID: Andrew's answer is interesting. He says he doesn't see the relevance to DM of sentences-as-adjectives in English. The relevance to DM, as I see it, is knowing how DM could generate these, what Andrew calls, 'phrasal' words. If the only means of word-formation are Vocabulary Item insertion and Head Movement, then I cannot see how phrasal words can be formed.If that is correct, then DM would thus still be bound by a rather strong version of the 'Lexicalist Hypothesis'. I am betting that DM has no easy solution and that the problem will turn out to be the exocentricity of phrasal words. That is, the immediate constituent of a 'phrasal word' doesn't match the label of the category under which it occurs in the tree (a sentence in Adjective position, for example will provoke severe complications in explaining how it got there - although Andrew Carnie's thesis offers some suggestions, where Carnie allows phrases under X0s). The specific problem I have in mind comes from Wari' quotatives, in which the quotative phrase functions as the verb of the sentence. This is documented in my chapter in the Handbook of Morphology and also in the Everett & Kern grammar of Wari from Routledge. There is strong evidence in Wari, however, that not even Carnie's proposal (form a phrase and move it into X0) will work, since movement is fairly carefully marked in Wari, but quotatives show no signs of movement. I am sure that this is all too sketchy to make a great deal of. But I am hoping to finish a paper on this in the next few weeks. The problem seems to be the need to recognize exocentric constructions motivated directly by the semantics, both requirements hard to state in formal theories (like DM). Well, anyway, that is what the paper claims. More on that when it is available. Back to Andrew C-M's proposal. On the other hand, I am not sure how a new category of 'phrasal word' would help at all. The concept needs to be formalized. Perhaps this is done in Andrew's forthcoming book.I look forward to reading it. Dan Everett Research Professor of Phonetics and Phonology University of Manchester From mcginnis at ucalgary.ca Thu Feb 15 21:32:57 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 15 Feb 2001 14:32:57 -0700 Subject: Heidi Harley: Word formation constraints (reply to Dan Everett) Message-ID: Hi all -- well, I'll take a stab at it, without having thought about it very much. These are certainly thornily tricky constructions... pls consider the following to be just the naivest possible response. >The relevance to DM, as I see it, >is knowing how DM could generate these, what Andrew calls, 'phrasal' >words. If the only means of word-formation are Vocabulary Item insertion >and Head Movement, then I cannot see how phrasal words can be formed. why not allow the syntax to generate phrases and merge them in 'head' positions, with permission slips for such activity issued on a language-by-language, construction-by-construction basis (e.g. via varying what types of categories can bear certain types of features)? As Dan notes, this is essentially what Andrew Carnie argued for Irish nominal predicates in his thesis. The hard problem for such a proposal is figuring out how to constrain it, of course, but letting the syntax generate phrases seems natural enough -- it's what it does. the trick is letting it do it in certain head positions. Then, presumably, vocab. insertion will proceed as usual, and phrasal phonology will as well, and you end up with a syntactically complex item that is doing the job of a 'word'. Note that DM has an advantage if this is really a possible account of such constructions, since it's a Late Insertion theory: the syntax can do all the dirty work and you don't have to replicate overtly phrasal syntactic rules in the lexicon just to construct these complex phrase-words. >If >that is correct, then DM would thus still be bound by a rather strong >version of the 'Lexicalist Hypothesis'. if i'm saying a reasonable thing, then it won't be so bound. >I am betting that DM has no easy >solution and that the problem will turn out to be the exocentricity of >phrasal words. That is, the immediate constituent of a 'phrasal word' >doesn't match the label of the category under which it occurs in the tree >(a sentence in Adjective position, for example will provoke severe >complications in explaining how it got there) right -- but i'm guessing that's a syntactic problem, not a morphological one. >The specific problem I have in mind comes from Wari' quotatives, in which >the quotative phrase functions as the verb of the sentence. This is >documented in my chapter in the Handbook of Morphology and also in the >Everett & Kern grammar of Wari from Routledge. There is strong evidence >in Wari, however, that not even Carnie's proposal (form a phrase and move >it into X0) will work, since movement is fairly carefully marked in Wari, >but quotatives show no signs of movement. I don't actually think that Carnie's proposal is that a phrase "moves into" an X0 slot and hence becomes an X0-- rather, the whole phrase, he says, _behaves_ like it's dominated by an X0 right from the get go (that is, the whole predicative NP bears verbal features). Then the whole thing head-moves to check those features, just like a regular V0 in Irish. But it's the fact that the nominal predicate can _bear_ such features in base position that allows the nominal predicate to head-move. so my guess is that movement isn't crucial: you ought to be able to generate a phrase that bears head features and have it merged into the syntactic structure just as if it was a word. i'll check with Andrew, who I don't think is a list member, and see if he can clarify a bit. >I am sure that this is all too sketchy to make a great deal of. But I am >hoping to finish a paper on this in the next few weeks. The problem seems >to be the need to recognize exocentric constructions motivated directly >by the semantics, both requirements hard to state in formal theories (like >DM). Well, anyway, that is what the paper claims. More on that when it is >available. looking forward to it! --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Fri Feb 16 17:58:53 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 10:58:53 -0700 Subject: Martha McGinnis: Idioms (reply to Heidi Harley) In-Reply-To: Message-ID: Heidi's right: if Higginbotham were correct that idiomatic VPs are aspectually different from non-idiomatic VPs, this would be a thorn in the side of DM. According to DM, both types of VPs are built in the syntax, so both should show the same 'compositional meaning.' Since aspectual meaning is clearly compositional, it should definitely allow the same possibilities for idioms and non-idioms. And, in fact, it does. I'm no aspect expert, but let's see if we can define the relevant verb classes. Accomplishments (1) and achievements (2) can be distinguished from activities (3) and states (4) by the fact that they're telic -- they allow endpoint-modification with "in an hour"-type adjuncts. (1) I climbed the mountain in an hour. ACCOMPLISHMENT (2) I recognized him in an instant. ACHIEVEMENT (3) *I climbed in an hour. ACTIVITY (4) *He was tall in an hour. STATE Achievements and accomplishments can be distinguished according to the interpretation they get with "for an hour"-type adjuncts. For accomplishments, the process preceding the endpoint is modified, as in (5). For achievements, if the modification is grammatical, it's not a pre-event modification but a result-modification, as in (6). (5) I climbed the mountain for an hour. ACCOMPLISHMENT (6) ?I recognized him for a moment. ACHIEVEMENT According to Vendler, accomplishments (7) are generally more compatible with the progressive in English than achievements (8), and activities (9) are more compatible with the progressive than states (10). ('Understand' is like 'recognize'... no doubt there are more.) (7) I'm climbing the mountain. (8) ??I'm recognizing him. (9) I'm climbing. (10)??He's being tall. However, there are some achievements which do allow the endpoint-modification (11), don't allow the pre-event modification interpretation with "for an hour"-type adjuncts (12), but are OK with the progressive (13). Let's call the first type of achievements Progless and the second type Progful. (11) He reached the top of the mountain in an hour (12) ??He reached the top of the mountain for an hour. (13) He's reaching the top of the mountain even as we speak. OK, so now we have a 5-way classification of VPs: states, activities, accomplishments, and progless and progful achievements. As expected, all classes contain idiomatic VPs as well as the non-idiomatic ones listed above. States: *in an hour, *progressive (14) He's the cat's pyjamas. (15) *He was the cat's pyjamas in an hour. (16)??He's being the cat's pyjamas. Activities: *in an hour, progressive (17) He pulls my leg constantly. (18) *He pulled my leg in an hour. (19) He's pulling my leg again! Accomplishments: in an hour, for an hour, progressive (20) He showed me the ropes in a couple of days. (21) He showed me the ropes for a couple of days. (22) He's showing me the ropes. Progful achievements: in an hour, *for an hour, progressive (23) I found my feet in a month. (24)??I found my feet for a month. (25) I'm still finding my feet. Others like this: cross the Rubicon, earn X's wings, make the grade, give up the ghost... Progless achievements: in an hour, *for an hour, *progressive (26) He croaked/kicked the bucket in three hours. (27) *He croaked/kicked the bucket for three hours. (28)??He's croaking/kicking the bucket. These are aspectually punctual, like their non-idiomatic counterparts, but they don't allow an iterative reading, for obvious reasons (a frog can croak again, a corpse can't). I think both the progressive and for-modification become OK if we talk about a group: (27)' The nearby animals were croaking for weeks following the Chernobyl meltdown. So there's no need for DM proponents to scurry around trying to find an explanation for why idioms aren't aspecually like non-idioms... they are. mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Fri Feb 16 17:59:47 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 10:59:47 -0700 Subject: Andrew Carstairs-McCarthy: Word formation constraints (reply to Dan Everett) Message-ID: Dan said, in reply to me: >Back to Andrew C-M's proposal [that e.g. _couldn't-care-less_ in _a >couldn't-care-less attitude_ should be classified as a phrasal word, >i.e. not some kind of compound but a phrase, yet functioning as a >word]. ... I am not sure how a new category of 'phrasal word' would >help at all. The concept needs to be formalized. Perhaps this is >done in Andrew's forthcoming book. I look forward to reading it. Well, I feel a bit sheepish now at having put in such a shameless plug for my forthcomingy book. It is an *introductory* book, and is aimed at English majors and trainee language teachers rather than budding linguists, so, although it is 'formal' in the sense that I aim for a clear and coherent set of definitions of technical terms, it is not theoretically adventurous. I merely use contrasts like _halfbrother_ vs _brother-in-law_ vs _jack-in-the-box_ to justify a distinction between compounds, phrases and phrasal words, this distinction being independent of that between idioms and non-idioms ... >The problem seems >to be the need to recognize exocentric constructions motivated directly by >the semantics, both requirements hard to state in formal theories (like >DM). Well, anyway, that is what the paper claims. More on that when it is >available. ... but I do venture into controversy, perhaps, when I discuss such expressions as _American history teacher_ with its two interpretations. Are there two constructions, each one motivated by its own semantics (as Dan might put it), or is there just one construction, with two interpretations nevertheless available? A traditional view is that there are two structures: (1) [[American history] teacher] 'teacher of Am hist' (2) [American [history teacher]] 'American teacher of hist' But if this is correct, then the structure at (1) ought also to be available for e.g. (3) and (4): (3) [[interesting history] teacher] 'teacher of interesting history' (4) [[suburban history] teacher] 'teacher of suburban history' But this seems incorrect. (3) and (4) can only mean 'interesting/suburban teacher of history'. That suggests to me that what makes the interpretation at (1) possible is not that the bracketing at (1) is made available by the syntax/morphology (with a N-bar inside a N -- a bit unwelcome!), but rather that the status of American history as an *institutionalized* subfield of history can force the interpretation at (1) on to the structure at (2). This analysis is heavily influenced by Andy Spencer's (in my view) excellent article on bracketing paradoxes in Language 1988. None of the structures at (1)-(4) is exocentric, under any plausible bracketing, so they may seem irrelevant to Dan's point. But I think they are relevant to the wider issue of the extent to which 'formal' grammatical (morphological or syntactic) structure determines interpretation, even with expressions such as (1) and (2) that most of us would hesitate to call idiomatic. -- Andrew Carstairs-McCarthy Professor and Acting Head of Department Department of Linguistics, University of Canterbury, Private Bag 4800, Christchurch, New Zealand phone (work) +64-3-364 2211; (home) +64-3-355 5108 fax +64-3-364 2969 e-mail a.c-mcc at ling.canterbury.ac.nz http://www.ling.canterbury.ac.nz/adc-m.html From mcginnis at ucalgary.ca Fri Feb 16 18:00:18 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 11:00:18 -0700 Subject: Alec Marantz: Word formation constraints (reply to Dan Everett) Message-ID: I must confess that I don't understand the connection between Dan's real worry -- what's the proper analysis of Wari' quotatives -- and DM (as Andrew C-M also points out). I think I agree with him that to the extent that there's a special problem, it's a problem for any theory (i.e., for what Dan is calling a "formal" theory, although I fail to see the difference here between "theory" and "formal theory"). There is a confusion about the role of node labels. I, for one, believe that the Chomskian program can do without node labels entirely. But, in any case, the first member of a compound noun in English can be complex (glass flower case, etc.) and it's clear that there is no category constraint on the first member of the compound. Thus, I believe, it would be hard for DM or any other theory to keep phrases out of this position. The problem is accounting for the interpretation on the one hand and the restriction on complements on the other. So, putting aside the "phrasal" cases like "keep off the grass sign," when the first member of the compound (or when an adjectival modifier in the same structural position) has a complement, the structure is bad (compare, "vest pocket kerchief" with, "*pocket of a vest kerchief" and recall, "*afraid of dogs man" vs., "man afraid of dogs"). When phrases like "keep off the grass" are treated as single units, they don't fall under the "no complement" generalization even though they contain complements. So in a clear sense they do behave like simple heads as opposed to complex heads (or phrases). Dan challenges us to find an account of this. Contrary to Dan, I believe that current "formal" theories are precisely addressing the relevant issues concerning the relation between the syntax and semantics of structures. That doesn't mean formal theories currently have an answer to the question Dan raises, although I agree with Heidi that Carnie's approach is closer to an answer than Dan seems to believe. marantz at mit.edu From mcginnis at ucalgary.ca Fri Feb 16 18:00:41 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Fri, 16 Feb 2001 11:00:41 -0700 Subject: Dan Everett: Word formation constraints (reply to Heidi Harley) Message-ID: Heidi's solution to my query, that it is primarily a syntactic problem, may in fact be correct. However, this itself raises problems for the syntax. I will just mention what I think the problem is. It is not directly relevant for DM but it is indirectly relevant to the degree that DM has an opinion about what syntactic theory its own apparatus follows. Assume some version of the MP. Andrew Carnie's solution in Irish will work in MP and DM and I think it is quite a reasonable one. However, it works in the MP because the syntax can form a phrase and then MOVE it into the appropriate and, this is crucial, preexisting (by Select) morphological position (and it can do this only in the MP,among formal theories, because other theories crucially distinguish bar-levels, certainly this is true of LFG and, I believe, it is still true for HPSG). However, in the absence of movement, the only way to embed a phrase into a morphological position is via MERGE. But Merge has problems here. Merge can only take place if a node is licensed for it (in the absence of base-generated trees, which are eliminated in the MP). The V node, so far as I can tell, is not really licensed by anything but the V itself (although I think some may have proposed having the Tense-node license/select it and John Bowers has argued for a Pr0 node - but that will still only work if the Pr0 node is lexical, I think). So, if there is no V, there doesn't seem to be any way to Merge a phrase into it (since the phrase is not in the Array and cannot be chosen by Select). That means that if one could show a case which needed to be analyzed as Merge into V (though I think, sigh, that it is probably impossible to conclude that the MP 'needs' to analyze anything in any particular way) that should be a problem for the MP and, thus, to DM if MP is its 'feeder syntax module'. But note that if one DOES have Merge into X0 categories (in spite of what I think the likelihood of MP being able to do this) it also provides a solution to the phrasal words/adjectives I brought up initially. Anyway, I have no doubt but that Heidi's proposal is the way to go about things wrt this type of phenomenon (which I believe to be fairly rich, cross-linguistically). The question is to find the best theory to do this in. -- Dan From mcginnis at ucalgary.ca Tue Feb 20 15:55:21 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 08:55:21 -0700 Subject: Heidi Harley: Idioms (reply to Martha McGinnis) Message-ID: Hi all -- Martha's excellent examples and discussion do go a long way to allaying my concerns. Especially the '??he's recognizing me' class of progless non-idioms is something that ought to confound Jim H., I think. However, I do have a concern with the examples of progful idioms that Martha gives, which will take a minute to lay out. The characterization of accomplishment verbs as taking a 'for' adverbial is incomplete -- the 'climb the mountain' accomplishments do certainly allow it, but there's some that canonically don't: #I built the house for an hour. #I ate the cookie for 5 minutes. and are usually remarked upon as only allowing the 'for' adverbials when the object is in 'partitive' case, or, in English, modified by a PP: I built on the house for an hour. I ate at the cookie for 5 mins. (The 'climb the mountain' example is like 'run the race': a canonical activity verb with an additional delimiting object argument -- and of course activity verbs allow 'for' modification. These 'build' and 'eat' verbs require objects, though.) So the question is, how do we distinguish between the 'build the house' accomplishments, which don't allow the 'for' modification, and the progful achievements? The difference between 'he's building the house' and 'he's winning the race' seems to be that the 'building the house' is mid-event -- you're partly completed building the house -- while 'winning the race' is pre-event: you're not partway through winning the race, rather, you're in the lead at the moment, but the instant of winning is at the finish line, and you can't have the race half-won, although you can have the house half- built. (don't say, what about a house-building race?) So to get back to the idiom question, are the examples that Martha gives as being progful acheivements, because they don't accept 'for'-adverbials -- are they like 'build the house' accomplishments or like 'win the race' achievements? >(23) I found my feet in a month. >(24)??I found my feet for a month. >(25) I'm still finding my feet. it seems to me that 'finding my feet' is in some sense 'mid-event' -- that is, you can be halfway along in finding your feet, unlike winning a race. If that's right, it's an idiom of the 'build' class, not of the 'win' class. Of the other ones that Martha suggests are progful achievements, 'cross the Rubicon' and 'earn X's wings' seem to me to be more like the 'build the house' accomplishments, not like the 'win the race' achievements. For me, however, 'make the grade' and 'give up the ghost...' are more acheivement-like, intuitively. But, maddeningly, 'give up the ghost' doesn't seem to go as well with progressives -- or insofar as it does, it doesn't mean 'pre-event', it means during the event, at the instant -- like 'croak' or 'kick the bucket': ?He's giving up the ghost. (*He was giving up the ghost for 2 hours before the end.) 'He's making the grade,' though, doesn't seem bad, and it does seem pre-event (I don't think you can have made the grade just halfway, right?) So I think this is, I hope, I hope, an example of a progful acheivement idiom. thoughts? Kate Kearns also sent me a note pointing out among other things that one of the properties of achievement progressives is that they don't accept modification by 'still', plus a couple of other things. Here's the relevant citation from Kate's note: "Anita Mittwoch's (1991) paper 'In defence of Vendler's achievements', Belgian Journal of Linguistics vol 6, pp. 71-85. She talks about the 'die/win the race' type, and says that their progressives aren't like real process progressives because they can't be modified by continuative 'still' or be in complements to 'stop, start, continue, keep' and so on." I still haven't gone to find the paper and check out her examples, but I thought the list might be interested. (The reason i wanted to suggest to Jim that there's no significant distance between accomplishments of the 'build' class and achievements is to cut through this whole tangled question; they behave the same w/r to time adverbials, and their interpretations with progressives are darn slippery. but for 'die' and 'win' and 'reach the top' the pre-event vs. mid-event distinction does seem fairly robust). Anyway, any thoughts? maybe i'm missing something obvious. and the judgments are kind of subtle. there might be better tests out there. Marf? prove me wrong, please! and thanks v. much for the input, hh --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Tue Feb 20 17:25:58 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 10:25:58 -0700 Subject: Martha McGinnis: Idioms (reply to Heidi Harley) In-Reply-To: Message-ID: >So the question is, how do we distinguish between >the 'build the house' accomplishments, which don't >allow the 'for' modification, and the progful >achievements? This sounds like a question for Higginbotham: wasn't he the one claiming that they should be distinguished? However, it looks like there's an answer: >"Anita Mittwoch's (1991) paper 'In defence of >Vendler's achievements', Belgian Journal of Linguistics >vol 6, pp. 71-85 ... talks about the 'die/win the race' type, >and says that their progressives aren't like real >process progressives because they can't be modified by >continuative 'still' or be in complements to 'stop, >start, continue, keep' and so on." OK, so now we have SIX verb classes -- call them states, activities, delimited activities, accomplishments, progful achievements, and progless achievements. The claim still holds, though: idiomatic and non-idiomatic VPs have the same range of aspectual possibilities. Let's focus on the classes I may have confounded earlier: Delimited activities: in an hour, for an hour, progressive, stop Ving Non-idiomatic: climb the mountain (1) He showed me the ropes in a couple of days. (2) He showed me the ropes for a couple of days. (3) He's showing me the ropes. (4) He stopped showing me the ropes when Sue arrived. Accomplishments: in an hour, *for an hour, progressive, stop Ving Non-idiomatic: build the house (5) I went to pieces in minutes. (6)??I went to pieces for minutes. (OK only with result-modification) (7) I'm going to pieces again. (8) I stopped going to pieces when Sue arrived. Others like this: make the grade, fall apart, lose one's marbles Progful achievements: in an hour, *for an hour, progressive, *stop Ving Non-idiomatic: win the race (9) I found my feet in a month. (10)??I found my feet for a month. (OK only with result-modification) (11) I'm still finding my feet. (12) *I stopped finding my feet when Sue arrived. Others like this: cross the Rubicon, bury the hatchet I'm afraid I have no insights at all regarding Heidi's Big Question -- i.e., whether these six should be primitive aspectual classes, or just six of many, or six permutations of a smaller number of primitive classes. But at least I think it's clear that idioms aren't aspectually 'special'. Cheers, Martha mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Tue Feb 20 21:09:46 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Tue, 20 Feb 2001 14:09:46 -0700 Subject: Heidi Harley: Idioms (reply to Martha McGinnis) Message-ID: Hi all -- >This sounds like a question for Higginbotham: wasn't he the one >claiming that they should be distinguished? yes -- and he pointed to the 'pre-event' vs 'mid-event' readings of the progressive in each case as evidence that they are significantly different. The paper of Mittwoch's that Kate Kearns sent me the reference to, though, also claims that they must be distinguished, hence the title -- and given your results with her 'stop' test, let's assume she (and Jim) are right. >>"Anita Mittwoch's (1991) paper 'In defence of >>Vendler's achievements', Belgian Journal of Linguistics >>vol 6, pp. 71-85 ... talks about the 'die/win the race' type, >>and says that their progressives aren't like real >>process progressives because they can't be modified by >>continuative 'still' or be in complements to 'stop, >>start, continue, keep' and so on." >Accomplishments: in an hour, *for an hour, progressive, stop Ving >Non-idiomatic: build the house > >(5) I went to pieces in minutes. >(6)??I went to pieces for minutes. (OK only with result-modification) >(7) I'm going to pieces again. >(8) I stopped going to pieces when Sue arrived. > >Others like this: make the grade, fall apart, lose one's marbles > >Progful achievements: in an hour, *for an hour, progressive, *stop Ving >Non-idiomatic: win the race > >(9) I found my feet in a month. >(10)??I found my feet for a month. (OK only with result-modification) >(11) I'm still finding my feet. >(12) *I stopped finding my feet when Sue arrived. > >Others like this: cross the Rubicon, bury the hatchet These are great, although I'm definitely going to have to go check out Mittwoch, since "still" ought not to be compatible with real acheivements according to Kearns' short description. But the 'stop'-test looks like it works as advertised. >I'm afraid I have no insights at all regarding Heidi's Big Question >-- i.e., whether these six should be primitive aspectual classes, or >just six of many, or six permutations of a smaller number of >primitive classes. But at least I think it's clear that idioms >aren't aspectually 'special'. hooray, thanks. So, I think then that Jim's got a problem with progless acheivements; his account of the semantics of the progressive only allows for progful ones, and he can't point to 'idiomaticity' as artificially preventing progfulness. I'm not really hoping for an answer to the Big Question at this point; I do ultimately think/hope that most of these distinctions can be taken care of in the syntax, with the merest handful of 'primitive' distinctions at issue (cf. e.g my paper on denominal verbs and aktionsart in the 2nd MIT/Upenn book), but the main point at hand was whether or not there were such things as progful idioms -- if there really weren't, that'd be trouble! another interesting question is whether or not the 'idiomatic' item always has the same basic aktionsart as the literal reading of the verb or frame from which it's derived. it's not at first obviously true for 'croak', because you can only croak idiomatically once -- but as long as we say that THAT distinction between idiomatic croaking and literal croaking is part of the Encylcopedia, then both idiomatic croaking and literal croaking are punctual (instantaneous, in the relevant sense). So maybe (and hopefully, for a morphosyntactic theory of aktionsart) it is true. Anyway, that's probably a question for another day. thanks very much, martha, for your tremendous examples and discussion! best, hh >Cheers, >Martha > > >mcginnis at ucalgary.ca --------------------------------------------------------------------- Heidi Harley Department of Linguistics Douglass 200E University of Arizona Tucson, AZ 85721 Ph: (520) 626-3554 Fax: (520) 626-9014 hharley at u.arizona.edu From mcginnis at ucalgary.ca Wed Feb 21 16:02:11 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 09:02:11 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz) Message-ID: Folks, I have been traveling. Before I attempt an answer in more detail to Alec's posting, let me say what the difference is between a formal theory and a 'nonformal theory'. A formal theory is, like DM, MP, etc, a theory in which FORM drives the model, rather than meaning. That is, meaning is not directly causally implicated in the model. In this sense. DM, like MP, are structuralist theories. Generative semantics, on the other hand, would not have been a formal theory in this sense (many formalists seem to confuse 'formal' with 'explicit', well-defined, generative, etc. But that is a mistake, historically and etymologically. Structuralist and formalist/formal are synonyms in this, fairly wide spread and traditional, understanding (hence I am puzzled as to why it would be hard for Alec to see this). In any case, my 'real concern' is not Wari' quotatives. That is merely a very good example of the problem. My real concern is the problem that such examples raise. I have not yet seen in Andrew Carnie's thesis, for example (but I am working my way through it slowly, so if it is there, my apologies to Andrew), an attempt to ask the question of where the predicate node arises for the S to move into. It is just there in the phrase structure. And Alec makes no attempt to answer my question either. He merely restates Heidi's proposal, namely, that Andrew's thesis probably provides the answer. As I said in response to Heidi, this may in fact be true. But there are technical problems and I, as an outsider, am merely trying to see how DM would answer it. It is interesting that no one has yet tried to, not even in response to my last message. However, I have a long list of emails following this one, so perhaps someone will. Cheers, Dan Everett From mcginnis at ucalgary.ca Wed Feb 21 16:02:44 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 09:02:44 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz cont'd) Message-ID: Folks, One other thing about Alec's posting, since no one else tried to answer my technical question. Contrary to Alec's suggestion, there is no confusion about node labels, so far as I can see. I am quite able to understand the MP reasoning behind the suggestion that nodes have no labels. Since there is no base component and since nodes are not so labeled in the lexicon, ex hypothesi, there is no way to have labeled nodes in the syntax, at least until after Merge and the postsyntactic DM work. Na~o ha' problema. On the other hand, I must admit that I *am* confused about node origins when those nodes do not dominate anything. Take movement of a phrase into the head of TP, for example. Why is the head/phrase there to serve as a target for movement? Either a node is required for lexical reasons or it is not there. What requires T? If the node is required by some aspect of LF, then the relevant constraint is global (not unusual for MP, since, as I pointed out in my Lg review of Hornstein's LF book sometime ago, MP has all possible constraints: global and violable, global and inviolable, local and violable and local and inviolable). But globality seems to be an embarassment here. Surely GREED or something in that spirit is more desirable, i.e. a local lexical requirement. I am not claiming anything here about the plausibility or possibility of a structuralist analysis of the facts. I am simply saying that I do not know how MP (and hence, I believe ultimately DM) can have a node for a phrase to move into unless that target node is first itself lexically required, e.g. by a null item projecting it). I don't think that the solution is going to be via MERGE, but perhaps there is something along those lines I have missed. If people think that this is a syntactic question, not a morphological question, my response is "I thought DM said that morphological structure = syntax". -- Dan From mcginnis at ucalgary.ca Wed Feb 21 18:00:31 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Wed, 21 Feb 2001 11:00:31 -0700 Subject: Dan Everett: Word formation constraints (reply to Alec Marantz) In-Reply-To: Message-ID: OK, I've finally had a chance to read this thread carefully, and I find I have nothing to add to Heidi's remark that DM looks like a good theory to be adopting to try to account for these sentential adjuncts, since DM doesn't allow us to sweep their unusual 'atomic' behaviour under an all-purpose Lexicon. With luck the comparison with Wari' quotatives will shed some new light on the phenomenon. Nevertheless, I thought I would comment on the discussion of DM, MP, etc., in case another interpretation is helpful. >A formal theory is, like DM, MP, etc, a theory in >which FORM drives the model, rather than meaning. That is, meaning is not >directly causally implicated in the model. In this sense. DM, like MP, are >structuralist theories. There may be similarities between structuralism and generative linguistics (e.g. the focus on systematic properties of language, rather than on prescriptive quirks, metaphor, etc.), but there's an important distinction to be made here as well. DM and MP, like other generative theories, are concerned with systematic correspondences between sound and meaning. If I recall correctly, structuralists were convinced that language could be captured simply in terms of phonologically-based patterns, without reference to meaning. Thus structuralism treats different languages as internally coherent but fundamentally distinct systems. Generative linguistics treats them as different states of the language faculty, whose lexicons draw on a universal set of semantically based syntactic features (tense, aspect, definiteness, specificity...), as well as phonetically based phonological features. The heyday of structuralism is far enough back in history now that I know of few explicit arguments against it. One is Morris Halle's argument against the structuralist notion of 'phoneme'. I also seem to recall that Halle & Marantz argue against a structuralist approach to morphology in their first DM paper -- i.e. that 'position classes' based purely on morphological distribution don'tt capture the correct generalizations about Potawatomi morphology. >I have not yet seen in Andrew Carnie's thesis, for example >(but I am working my way through it slowly, so if it is there, my >apologies to Andrew), an attempt to ask the question of where the >predicate node arises for the S to move into. It is just there in the >phrase structure. I'm not 100% sure I understand the question, but the issue here might be the order of assembly of syntactic structures. It's not possible to assume that the syntax composes a single tree, step by step, moving upwards (or downwards, for that matter). For example, phrasal specifiers have to be pre-assembled in a subtree before they can merge with the main tree. >Since there >is no base component and since nodes are not so labeled in the lexicon, ex >hypothesi, there is no way to have labeled nodes in the syntax, at least >until after Merge and the postsyntactic DM work. I think there's some flexibility here. Virtually everyone assumes the lexicon contains items of different categories, which have different syntactic properties (both for Merge and Move). Since introducing Bare Phrase Structure, Chomsky has assumed that phrasal units also have category labels, whose X-bar level can be read off the structure. Alec suggests that the syntax can get away without making use of these category labels. This sounds to me like a new suggestion -- that labels need not be stipulated, but can be read off the structure. By contrast, BPS uses different types of node labels to distinguish between specifiers and adjuncts. >On the other hand, I must admit that I *am* confused about node origins >when those nodes do not dominate anything. Take movement of a phrase into >the head of TP, for example. Why is the head/phrase there to serve as a >target for movement? Either a node is required for lexical reasons or it >is not there. What requires T? Perhaps nothing, which is why we can just say "Gone fishin'." The notion that Merge has to be licensed by something also sounds to me like a new proposal. As far as I understand, Chomsky treats Merge as free: it either does or doesn't succeed in generating semantically interpretable structures. It's possible that arguments are forced to Merge by a local need to satisfy the theta-requirements of a predicate, but even if so, this would not entail that there is no unforced Merge. What empirical benefits do we reap by taking the position that Merge is always forced or licensed? >But globality seems to be an >embarassment here. Surely GREED or something in that spirit is more >desirable, i.e. a local lexical requirement. As you know, 'Suicidal Greed' is local, but Greed and Procrastinate are the classic global constraints. Am I allowed to say 'classic' about the 1990's yet? Cheers, Martha mcginnis at ucalgary.ca From mcginnis at ucalgary.ca Thu Feb 22 15:35:34 2001 From: mcginnis at ucalgary.ca (Martha McGinnis) Date: Thu, 22 Feb 2001 08:35:34 -0700 Subject: Dan Everett: Word formation constraints (reply to Martha McGinnis) Message-ID: Martha, Good attempts at answers. I enjoyed reading them. I am afraid that I recognize little of structuralism in your account of it however. Generative linguistics seems structuralist in exactly the same ways as, say, Tagmemics, except more so (because Tagmemics was not only the very first theory, long before Morris's arguments against the 'phoneme' - which, by the way, have been shown to be wrong: the major lesson of Lexical Phonology is that something like the 'taxonomic phoneme' is a necessity in linguistic theory, but Tagmemics also allowed semantics and pragmatics to play a causal role in syntax. Tagmemics is not that great a theory. But it is less structuralist than MP). All structuralists I have read (and met) said things like, "Of course languages are more alike than they are different or we couldn't have a theory of Language". Structuralists believe(d) in universals (Greenberg, not Chomsky, first drew our attention to them; structuralists are mainly responsible for our knowledge of 'exotic' languages, because, like Mead, Benedict, Sapir, and other students of Boas, they believed that there are universal patterns of language that need to first be documented and then explained). But the real hallmark of structuralism is that the only solutions that count are distributional and structural. And this has ALWAYS been true of generative linguistics. We all know that Noam, to take a random example, accepts x as a solution to y just in case x is stated in terms of structural relationships. That is basically the motivation for MP. As I recall from lectures back in the mid 80s Chomsky was getting fed up with the multiplication of entities beyond necessity. He wanted to get back to the basics, which MP is supposed to be. And the basics are trees and a lexicon. No indices. No meaning intrusions (e.g. 'affectedness', 'achievements' vs. 'accomplishments', etc.) These things are all to be derived. And derived by trees. That just IS structuralism, which Zellig Harris taught at Penn. As to the specific questions I raised. 'Gone fishing' is the answer I will take away from this discussion. Best, Dan