Newsgroups: comp.ai.philosophy
Path: cantaloupe.srv.cs.cmu.edu!fs7.ece.cmu.edu!kinky.eng.gtefsd.com!europa.eng.gtefsd.com!howland.reston.ans.net!gatech!bloom-beacon.mit.edu!news.media.mit.edu!minsky
From: minsky@media.mit.edu (Marvin Minsky)
Subject: Re: What's innate? (Was Re: Artificial Neural Networks and Cognition
Message-ID: <1995Jan26.224354.401@news.media.mit.edu>
Sender: news@news.media.mit.edu (USENET News System)
Cc: minsky
Organization: MIT Media Laboratory
References: <3g6js6$fug@mp.cs.niu.edu> <1995Jan26.150315.1420@il.us.swissbank.com> <3g8sru$jsn@mp.cs.niu.edu>
Date: Thu, 26 Jan 1995 22:43:54 GMT
Lines: 25

In article <3g8sru$jsn@mp.cs.niu.edu> rickert@cs.niu.edu (Neil Rickert) writes:

>In spite of the claims of the 'poverty of stimulus' argument, we are
>exposed to an enormous volume of language, and for those of us who
>become bookworms, this is particularly so.  At the time I was into
>mathematics, and perhaps that was why it was so clear to me that
>there was no mathematical structure to be found in natural language
>syntax.  I never could have attempted to develop a syntactical theory
>of language since it was always so obvious that any such theory was
>doomed to fail.  But one can be wrong in these intuitions, so I
>applaud Chomsky for trying.

Bah.  You cannot assess the "poverty" unless you have an idea of the
learning scheme.  Didn't Mitchell Marcus show that surprisingly few
examples were needed, if you have the right sort of representations
and learning algorithms?

Furthermore, isn't it obvious that the "constraints of Universal
Grammar" could just as well emerge from the limitations of what the
learning algorithm could discover, rather than the limitations of the
ultimate language processor?

I suspect this whole subject is an artifact that came from Chomsky's
initial separation of "competence" from, I forget what the other is.

