Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!Germany.EU.net!EU.net!uknet!bcc.ac.uk!link-1.ts.bcc.ac.uk!sslyjim
From: sslyjim@ucl.ac.uk (Mr Jim Tyson)
Subject: Re: Universal Grammar
Sender: news@ucl.ac.uk (Usenet News System)
Message-ID: <1994Dec5.104617.57283@ucl.ac.uk>
Date: Mon, 5 Dec 1994 10:46:17 GMT
References: <jon-0112941057550001@hfmac323.uio.no> <3bljr1$ajp@agate.berkeley.edu> <jon-0312940940110001@hfmac323.uio.no>
Organization: Bloomsbury Computing Consortium
Lines: 30

In article <jon-0312940940110001@hfmac323.uio.no> jon@babel.ifl.uib.no (Jon Hareide Aarbakke) writes:
>In article <3bljr1$ajp@agate.berkeley.edu>, Patrick Hall
><pathall@uclink.berkeley.edu> wrote:
>
[deleted for brevity]
>> 
>> Could you explain a little more about what a connectionist machine is? 
>Or maybe some suggestions for introductory essays, etc?
>> 
>> Patrick Hall
>

Try 

Zeidenberg, M, 1990, _Neural Network Models in Artificial Intelligence_,
New York: Ellis Horwood

Very accessible and with nice introductions to various implementations using
(quasi) connectionist techniques, several of which are for natural langauge
processing.  But then be sure to take a look at 

Asher, 199?, _The Chomskyan Turn_, Oxford: Blackwell

(sorry for the incomplete reference, the book is at homed).

which will give you some idea of why some people are still symbol processing
fans.  Especially relevant is the chapter by Pylyshyn.

Jim

