From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!bronze!chalmers Fri Jan 31 10:27:06 EST 1992
Article 3277 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!bronze!chalmers
>From: chalmers@bronze.ucs.indiana.edu (David Chalmers)
Subject: Re: Virtual Person?
Message-ID: <1992Jan30.000852.11991@bronze.ucs.indiana.edu>
Organization: Indiana University
References: <1992Jan24.171454.7033@aisb.ed.ac.uk> <1992Jan29.004822.23755@bronze.ucs.indiana.edu> <1992Jan29.192858.25598@aisb.ed.ac.uk>
Date: Thu, 30 Jan 92 00:08:52 GMT
Lines: 39

In article <1992Jan29.192858.25598@aisb.ed.ac.uk> jeff@aiai.ed.ac.uk (Jeff Dalton) writes:

>>Only a trivial few of the symbols will be Chinese characters -- the
>>peripheral ones for input/output.
>
>I don't think that claim can be supported from the text.  Rather than
>type in more quotes, I'll just repeat what I quoted in the message
>you're answering.  It's from the 2nd Reith Lecture:

>If it's not clear already, it should be clear from the rest of the
>paragraph that it's not just the I/O symbols that are Chinese in this
>version of the argument.

Look, this is Searle's problem, not mine.  If he wants his argument
to apply to all computer programs that produce Chinese-speaking
behaviour, then he has to consider programs that don't have Chinese
symbols internally.  If on the other hand he only wants to limit it
to programs with this form, he's welcome to, but then AI has a vast
escape hole.  I imagine that when pressed, Searle would take the
first option.  But one of the points of my article on this topic was
that Searle plays on our intuitions by implicitly or explicitly
introducing the assumption that the symbols shoved around in the room
would have to be Chinese symbols, or other putatively meaning-bearing
entities, when in fact AI is committed to no such thing.

>However, my main point was not how many of the symbols wre Chinese but
>that at least some of the time the issue for Searle is that the
>symbols are manipulated only on the basis of their formal properties
>(such as their shape) and that they can never acquire meaning.

Right, and neurons are sensitive to other neurons solely on the basis
of properties such as firing patterns, so they can never acquire
meaning.  The point is that neither neurons in the brain nor symbols
in the room need be the locus of meaning.

-- 
Dave Chalmers                            (dave@cogsci.indiana.edu)      
Center for Research on Concepts and Cognition, Indiana University.
"It is not the least charm of a theory that it is refutable."


