syntax question.

sonya · Post by **sonya** » Tue Mar 14, 2006 9:29 am

...ehm. I'm puzzling over a syntax problem. I have no idea how to tackle it, no one I have spoken to really knows how, and when I spoke to my professor she replied nicely enough but deftly evaded an enlightening answer. So I thought maybe someone on the applied linguistics forum would be so kind as to know the answer and help..? Or know of a forum where people might?

The problem is how to structure a ditransitive sentence with a negative polarity item, such as, "We gave no one any money." Initial treatment of the structure of this sentence (including googling it) have the verb "give" and the two NPs as sisters.. However, that can't be true, as the two NPs can't c-command each other, and sisters do. ..So I'm arguing that the structure has to be heirarchical, in order for no one to c-command any.. but.. heirarchical how? I can't figure it out.

So is there something like a "Negative Polarity Item Phrase" (NPI P?)? How would it work? Is it simply the "no one" is a Neg P? (but Neg P's in English don't follow the verb, right?)

thanks in advance

lolwhites · Post by **lolwhites** » Tue Mar 14, 2006 12:01 pm

Forgive my ignorance as it's been years since I studied C-command, but surely there's no difference syntactically between I gave noone any money and I gave someone some money. I vaguely remember that ditransitivity posed a problem with C-command, but I don't see why the negative should be any harder to analyse.

abufletcher · Post by **abufletcher** » Tue Mar 14, 2006 8:59 pm

I'd like to add a discourse/used language perspective here and say that "I gave no one any money" is extremely odd and I believe would only have been produced in a very specific sort of situated context such as in the following (obviously imaginary) dialog:

A: Well, if you're going to be giving anyone any money it ought to be us!
B: I gave no one any money.

In my view, the extremely odd (low frequency) formulation of this turn-at-talk would only come about as a result of speaker B attempting to mirror the formation of the prior speaker's talk. I don't know what value there is to trying to posit some sort of syntactic tree diagram that would explain its structure in some decontextualized manner. This, of course, is a major challenge for (weakness of) approaches to formalist linguistics which attempt to explicate single (and imaginary) sentence-level units.

abufletcher · Post by **abufletcher** » Tue Mar 14, 2006 9:08 pm

duplicate post deleted

lolwhites · Post by **lolwhites** » Tue Mar 14, 2006 9:13 pm

I don't know what value there is to trying to posit some sort of syntactic tree diagram that would explain its structure in some decontextualized manner. This, of course, is a major challenge for (weakness of) approaches to formalist linguistics which attempt to explicate single (and imaginary) sentence-level units.

It's probably one of the reasons why Chomsky has always maintained that he can't see how his theories about language could have any application in the teaching of foreign languages.

abufletcher · Post by **abufletcher** » Tue Mar 14, 2006 11:36 pm

True, Chomsky only claimed to be modeling an abstract system of grammar. But I just don't see how any reasonable model/grammar of language can claim to exclude the baldly obvious fact that language is always produced within a context and that that context interacts at a fundamental level with the language produced.

The orthodox view is that some Chomskyan-style grammar (automonously located within "the mind") generates language and then sociolinguistic/pragmatic awareness ("the rest of the mind") molds and adapts those generated units to the context, e.g., the system generates "complete strings" which are subsequently subject to elipsis to fit some situation. I think approaches to linguistics that base their findings of examination of actually produced language (e.g. corpus linguistics, conversation analysis, interactional linguistics, etc.) paint a very different picture for how bits of langauge come to be produced-for-contexts.

So when I say I don't see the value for Chomskyan style analysis of decontextualized single utterances, I'm not talking about some pedegogic usefulness in the classroom but rather the real ability to say anything truly meaningful about the organization of the linguisitc resources implicated in human social interaction.

I realize this is a minority position within linguistics and even more so within language teaching circles, but that doesn't make it wrong.

wilderson · Post by **wilderson** » Tue Mar 14, 2006 11:49 pm

Sonya,

I would like to help you but I can't

...shamefully I came up with nothing after about 1 min. on the problem...so...my excuse for not knowing the answer or bothering to look it up is probably the same for why I chose applied linguistics rather than general: it just ain't my bag (perhaps I speak for a few of us?)

.

Um...you might consider asking your question at Linguist List.

sonya · Post by **sonya** » Wed Mar 15, 2006 3:03 am

That's an interesting point, Fletcher. I had a course last semester in a type of discourse analysis, and the professor mentioned Chomsky, who apparently ostracized all the generative semanticists at MIT and said they were diluting linguistics as a science. However, she argued, 1. It didn't, and 2. It was generative semantics that fulfilled Chomsky's talk of linguistics as a window into the mind.

What I took away from Lakoff's class was that generative grammar and generative semantics are seperate things, and we can use the former to help us empirically analyze things like discourse analysis. Not that I see how NPIs specifically allow us to do it, but I believe a grasp of syntax can.

NPIs aren't so odd and unusual, really. I can say, for example,

"Last night, I didn't sleep a wink"
"No one saw anything"
"You just don't understand anything"

"I gave no one any money" is just one grammatical example out of many for using NPIs to analyze, or provide insight into, the structure of ditransitives..

Anyway, I don't know, if anyone's curious, I decided to draw two layers of VPs and have the verb and the negation move up one, and titled it "NPI movement." There's probably something wrong with it that I don't see, but it's something.

abufletcher · Post by **abufletcher** » Wed Mar 15, 2006 3:34 am

sonya wrote:Chomsky, who apparently ostracized all the generative semanticists at MIT and said they were diluting linguistics as a science. However, she argued, 1. It didn't, and 2. It was generative semantics that fulfilled Chomsky's talk of linguistics as a window into the mind.

I'm not sure I'd want to consider generative sematics as anything but a failed attempt to extend Chomskyan linguistics into the lexicon. Was this Geoff Lakoff or Robin Lakoff? Geoff was a leading proponent of generative semantics when the bottom dropped out of the field in the 80's and nearly everyone working in that area abandoned it. Geoff Lakoff moved off into linguistic metaphors. Robin Lakoff is more of a discourse analysis/sociolinguist looking at gender issues I believe. Anyway, if generative semantics did open a window into the mind is was a very small one with a very limited sense of mind.

What I took away from Lakoff's class was that generative grammar and generative semantics are seperate things, and we can use the former to help us empirically analyze things like discourse analysis.

As someone working in the variety of discourse analysis referred to as ethnomethodological conversation analysis, I personally don't see any way that Chomskyan generative linguistics can be useful in the analysis of actual "used language" i.e. language produce in-the-course-of-and-for interaction. Not unless theorists could come up with some way of making complex trees that would include multiple utterances by multiple speakers and allow for the fact that talk participants regularly alter the grammatical course of a turn-in-progess to accommodate the exigencies of the situation (including such things as the gaze direction of talk recipeints).

NPIs aren't so odd and unusual, really. I can say, for example,

"Last night, I didn't sleep a wink"
"No one saw anything"
"You just don't understand anything"

"I gave no one any money"

Interestingly, I find the first three of these to fit very well with my expectations about "used language" and would expect them all to have a reasonable frequency of occurrence in a large corpus. The "I gave no one any money" however, to my ear stands out as distinctly odd even if it may well share syntactic characteristics with the others. It just doesn't resonate culturally like the other three.

sonya · Post by **sonya** » Wed Mar 15, 2006 4:32 am

It wasn't Dr. George Lakoff. The way Dr. Robin Lakoff explained generative semantics (unless I misunderstood her, which is probable) was that it's considering the context of the speakers and the environment of their discourse while also using things like syntax and morphology to analyze what they say. This discussion of generative semantics was combined with discussion on language and gender and discourse analysis, so I came to understand them as connected.

All that said, excuse me for going off topic into the "used language" realm, which I obviously am wholly ignorant in. If you have any, I would appreciate feed-back on my syntax problem and the solution I came up for it.

much obliged,
sonya

abufletcher · Post by **abufletcher** » Wed Mar 15, 2006 6:08 am

George Lakoff! I knew Geoff just didn't sound right! Thanks.

abufletcher · Post by **abufletcher** » Wed Mar 15, 2006 6:20 am

Within conversation analysis there has been some (very limit) work done on "polarity" as it relates to the sort of response it makes expectable in the next speaking turn. For example, asking "Do you have any questions?" is apparently formated in such a way that the systematiclly preferred response would be to say "No." Normatively a bald negation of this type would be a dispreferred response following most Y/N questions and would therefore typically be done in a hedged manner:

A: Do you have a bicycle?
(.)
B: um...I used to. ((notice to complete absence of a negation token))

Just what linguistic factors affect this sort of polarity is not well understood.

lolwhites · Post by **lolwhites** » Wed Mar 15, 2006 12:48 pm

True, Chomsky only claimed to be modeling an abstract system of grammar. But I just don't see how any reasonable model/grammar of language can claim to exclude the baldly obvious fact that language is always produced within a context and that that context interacts at a fundamental level with the language produced

Well, there are branches of Mathematics and Theoretical Physics that appear to have little to do with the real world but it doesn't make the, wrong or not worthy of study. I think the same applies to Linguistics; what may now seem too abstract or a pointless endeavour may well turn out to have important implications later on.

abufletcher · Post by **abufletcher** » Wed Mar 15, 2006 1:38 pm

You miss my point. I am not dismissing linguistics because it is too theoretical and not "practical" enough. I'm suggesting that Chomskyans are simply barking up the wrong theoretical tree (diagram).

Other "theoretical" linguists working in the fields of corpus linguistics (Beebe), emergent grammars (P. Hopper), and interactional linguistics (Sandra Thompson) offer radically different perspectives on the nature of language than Chomskyans.

Still, no research enterprise is entirely without merit and I'm sure that Chomskyans have made their contribution as well. I just hate it when people treat Chomskyan views as the highest and purest (or only) form of linguistics.

abufletcher · Post by **abufletcher** » Wed Mar 15, 2006 2:49 pm

BTW, I think a lot of my ire comes from the fact that as a linguistics major years ago (BA, MA) all I was fed was a steady diet of Chomskyan linguistics with nary a hint that other approaches to the study of language existed.