Corpora: Chomsky and corpus linguistics

[J. Zavrel]

> [...] Based on *assumptions*  about
> what input children get in learning, Chomsky _then_ asks how the
> inferred structure of the black box might have gotten there, i.e. how
> language learning might have worked.

The problem with this approach could hardly have been summarized
more elegantly... this is precisely the methodology used in
generative linguistics, and the very reason it doesn't reveal much about
language (acquisition/use) in the human mind.

To me, this sounds like trying to deduce the logic of the
organisational principles of animal anatomy by just looking at
a picture book of zoo's, i.e. without any knowledge of evolutionary
processes and the ecological interactions that organisms have.
Well, that kind of thing has surely been done before as well.

This is not to say that generative linguistics has not produced
many neat insights, mini-theories, and cool ways to talk about the
taxonomy of this linguistic bestiary.
It's just so touching how little it has to say in terms of cognitive
processes, while at the same time being completely pre-occupied
by this strange concept of explanatory adequacy for the I-language.

Regards & sorry to waste your bandwidth with my 2c.

--Jakub

------------------------------------------------------------------------------
Jakub Zavrel, Textkernel B.V., Nieuwendammerkade 26, 1022 AB, Amsterdam, NL
------------------------------------------------------------------------------