Cognitive Science Preprint

[Prahlad.Gupta@K.GP.CS.CMU.EDU]

FTP-host: reports.adm.cs.cmu.edu (128.2.218.42) 
FTP-filename: /1993/CMU-CS-93-146.ps



The following article will appear in the Cognitive Science journal.  A preprint
of the paper is available as CMU Computer Science Technical Report No.
CMU-CS-93-146, in electronic as well as hard-copy form.  Information follows
about electronic retrieval (free), as well as ordering hard copies (for a small
charge).

Comments on the paper are invited.


Note: A preliminary and substantially different version of this paper
was announced in the neuroprose electronic archive in
December-January 1991-92 as the file gupta.stress.ps.Z (which is no
longer available).



-- Prahlad 

------------------------------------------------------------------------------
      Connectionist Models and Linguistic Theory: Investigations of Stress
                        Systems in Language

                Prahlad Gupta and David S. Touretzky
                      Carnegie Mellon University


                   (To appear in Cognitive Science)


                            Abstract

We question the widespread assumption that linguistic theory should
guide the formulation of mechanistic accounts of human language
processing.  We develop a pseudo-linguistic theory for the domain of
linguistic stress, based on observation of the learning behavior of a
perceptron exposed to a variety of stress patterns.  There are
significant similarities between our analysis of perceptron stress
learning and metrical phonology, the linguistic theory of human
stress.  Both approaches attempt to identify salient characteristics
of the stress systems under examination without reference to the
workings of the underlying processor.  Our theory and computer
simulations exhibit some strikingly suggestive correspondences with
metrical theory.  We show, however, that our high-level
pseudo-linguistic account bears no causal relation to processing in
the perceptron, and provides little insight into the nature of this
processing.  Because of the persuasive similarities between the nature
of our theory and linguistic theorizing, we suggest that linguistic
theory may be in much the same position.  Contrary to the usual
assumption, it may not provide useful guidance in attempts to identify
processing mechanisms underlying human language.
------------------------------------------------------------------------------
             INSTRUCTIONS FOR ELECTRONIC RETRIEVAL VIA ANONYMOUS FTP

unix> ftp reports.adm.cs.cmu.edu      # or ftp 128.2.218.42
Connected to reports.adm.cs.cmu.edu.
220 REPORTS.ADM.CS.CMU.EDU FTP server (Version 4.105 of 10-Jul-90 12:08) ready.
Name (reports.adm.cs.cmu.edu:prahlad): anonymous
331 Guest login ok, send username at node as password.
Password: <your-user-id at site>         # you must include the "@" 
230-Filenames can not begin with "/.." .
    Other than that, everything is ok.
230 User anon logged in.
ftp> cd 1993
250 Directory path set to 1993.
ftp> get CMU-CS-93-146.ps
200 PORT command successful.
150 Opening data connection for CMU-CS-93-146.ps (128.2.248.83,1073) (591324 byt
es).
226 Transfer complete.
local: CMU-CS-93-146.ps remote: CMU-CS-93-146.ps
600021 bytes received in 10 seconds (57 Kbytes/s)
ftp> quit
unix> lpr -P<your-PostScript-printer-name> CMU-CS-93-146.ps  
                      # or however you print PostScript files
------------------------------------------------------------------------------
                        ORDERING A HARD COPY 

                  (The TR No. is CMU-CS-93-146)

Contact:          Computer Science Documentation
                  School of Computer Science
                  Carnegie Mellon University
                  Pittsburgh,  Pennsylvania 15213, USA

Phone:            (412) 268-2596
Internet:         reports at cs.cmu.edu

------------------------------------------------------------------------------