Corpora: Chomsky/Harris - one more fun question.

[lgerber]

> Pete Whitelock wrote:
>
> Unfortunate it may be, but it is hardly surprising. Engineers who try to
> build useful artefacts might reasonably expect to build on a sound
> theoretical foundation provided by those who claim to be doing scientific
> study in the area concerned. As language engineers, or applied linguists, we
> must be profoundly disappointed by the abject failure of Chomskyan linguists
> to make the slightest useful contribution to human language technology.
>
> Instead of the support we need, we've been subjected to an ever
> proliferating array of inaudibilia, a complete lack of interest in real
> data, a standard of argumentation that is a million miles from what I'd call
> science, and condescension, if not hostility, towards those outside the
> initiated who -have- tried to do the science of language.
>
> Does anyone want to make a case for a contribution by the MIT school to any
> aspect of (human, as against programming) language technology or language
> teaching?

Since, as you observe, the generative school has taken little
interest in "real" data, we are not too likely to hear
their side of the story on this list, hence, we run the risk
of further entrenching our well established biases.  However
satisfying a bit of grousing is, especially when it is aimed
at a group that might be accused of collective arrogance and
insularity, I suspect that their (MIT/generative) disappointment
with us is no less bitter.  The fact that we don't care about
explanatory adequacy if it doesn't enable us to build better
systems is exactly parallel to their dismissal of our work
which doesn't even try to make a contribution within their
framework.

Laurie Gerber

University of Southern California