[Corpora-List] Chomsky and computational linguistics

[Rob Freeman]

On 8/1/07, Mike Maxwell <maxwell at umiacs.umd.edu> wrote:
>
> ...I don't think most generativists would argue against the idea
> that usage is part of language.  They (and I) would just say that
> structure is what usage has to deal with, just like twenty amino acids
> and 64 base pairs are what cells have to deal with.


I agree. I'm not arguing against "structure" at all.

I think the debate in linguistics has for too long been polarized into this
structure vs. function/meaning thing. That is one lasting bad result of
Chomsky's strong personality.

In fact a rejection of structure was just one response to observational
insufficiency (as Universal Grammar was another.) If we can explain the
"observational insufficiency" of language structure in other ways there is
no need for us to make a big distinction any more between structural
descriptions and functional or cognitive descriptions.

As I say, Functional and Cognitive linguistics are in many ways much
stranger bedfellows.

>     If they all fit the observations thus far, how would we choose among
> >     them?
> >
> > It is not a problem in practice. Take the one which best answers the
> > question you want answered at any given moment. Is "black" in the same
> > class as "strong"? Check if your context is "coffee" or "cloud".
>
> I would call that an issue of semantics, and that it's possible to tease
> apart syntax and semantics...


I'm sure that's how Generativism deals with this kind of thing. But once
Generativism has redefined it out of syntax and satisfactorily washed its
hands of the dirty thing, where does that leave poor old semantics?

Anyway, I wasn't presenting this as a problem, but a solution, a solution
for how to select between grammars. Notice it solved the problem you
presented. I'd rather not redefine my solution out of syntax, thanks. If it
turns out not to be a problem, but actually a solution, can keep it in
syntax? :-)

If you're wondering what syntax has to do with the case of 'black', then
> ask what "context" means.
>
> Coffee drunk in Ethiopia is black.
> Which coffee is better served black?
> Clouds with silver linings are never as big as those which are black.


I wasn't wondering! I certainly believe we can characterize the properties
of "black" as syntactic. Indeed, I think that is a good thing to do.

But I must have missed something in your examples. What are you
demonstrating?

-Rob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20070801/e6c869c8/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora