From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!think.com!yale.edu!cs.yale.edu!mcdermott-drew Mon Jan  6 10:30:03 EST 1992
Article 2440 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!think.com!yale.edu!cs.yale.edu!mcdermott-drew
>From: mcdermott-drew@CS.YALE.EDU (Drew McDermott)
Newsgroups: comp.ai.philosophy
Subject: Epiphenomenal semantics
Summary: Clear statement
Message-ID: <1991Dec30.150818.25714@cs.yale.edu>
Date: 30 Dec 91 15:08:18 GMT
References: <1991Dec15.023122.6582@husc3.harvard.edu> <1991Dec16.181202.526@cs.yale.edu> <41047@dime.cs.umass.edu>
Sender: news@cs.yale.edu (Usenet News)
Organization: Yale University Computer Science Dept., New Haven, CT 06520-2158
Lines: 32
Nntp-Posting-Host: aden.ai.cs.yale.edu


  In article <41047@dime.cs.umass.edu> orourke@sophia.smith..edu (Joseph O'Rourke) writes:
  >Drew McDermott wrote (in article <1991Dec11.170157.27053@cs.yale.edu>):
  >  
  >	>The bottom line is that semantics is epiphenomenal, [...]
  >
  >Dennett feels that the word "epiphenomenon" is used by different
  >writers with rather different meanings:  by philosophers to mean "has
  >no effect in the physical world whatsoever," and by psychologists to
  >mean "nonfunctional."

I meant it in the ironic-humorous sense, as in "Dan Quayle is
epiphenomenal."  But, to be precise, here is my claim:

It is the case that

 (a) People use symbols that refer to things
 (b) People can make semantic theories about what agents' symbols
     refer to

but the theories referred to in (b) play no role in the competence
described in (a).  That is, my ability to use a symbol is not
explained by my possession of a theory about what that symbol means.
There are no doubt marginal counterexamples (e.g., when I decipher an
unfamiliar word with familiar Latin roots), but in the normal case I
couldn't even state and use a theory about the meanings of symbols
without already being able to use the symbols that occur in the
theory.  A good semantic theory (if we had one) would *explain how*
one uses symbols to refer to things, but it wouldn't be *part of* that
use. 

                                             -- Drew McDermott


