[Corpora-List] Chomsky and Computational Lingusitics

[Terry]

Chomsky said that he wanted to generate all the grammatical sentences in the
language and none of the non-grammatical sentences. A correct defense of
Chomsky's generative project would then be one that says: We have not mapped
out 100% of the grammatical sentences. But we have mapped out 75% or 80%. We
also have good prospects over the next few years of making more progress
toward our ultimate goal. For his part, what Chomsky actually did in the
face of failure was to "assume" he was correct in his new Minimalist
framework and redefine the task as simply to generate a "core part" of the
grammar. Here the issue is: how much is the core? 20%? 10%? 

 

And did he succeed in this new minimalist task?

 

In the terms of Lakatos, Chomskyan generative grammar is a degenerative
research program.

 

Terry

 

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20070805/ecfe4ec3/attachment.htm>
-------------- next part --------------
_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora