From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!bronze!chalmers Wed Apr 22 12:03:50 EDT 1992
Article 5128 of comp.ai.philosophy:
Newsgroups: comp.ai.philosophy
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!qt.cs.utexas.edu!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!bronze!chalmers
>From: chalmers@bronze.ucs.indiana.edu (David Chalmers)
Subject: Re: What counts as the "Right" functional organization?
Message-ID: <1992Apr16.232832.24427@bronze.ucs.indiana.edu>
Organization: Indiana University
References: <1992Apr5.210553.11966@psych.toronto.edu> <1992Apr14.064526.16723@bronze.ucs.indiana.edu> <1992Apr14.142239.7807@psych.toronto.edu>
Date: Thu, 16 Apr 92 23:28:32 GMT
Lines: 64

In article <1992Apr14.142239.7807@psych.toronto.edu> christo@psych.toronto.edu (Christopher Green) writes:

>Okay, then you'd better start specifying what counts as the "right" functional
>organization lest you slip into a Davidsonian circle.  Actually, snarkiness
>aside, I'm honestly interested in what might count as right and not right.
>Presumably, recourse to human behavior is out of bounds. Presumably, phrases
>such as "sufficiently complex", or anything else countaining relative terms
>such as "sufficient", "enough", "necessary", etc., are all off bounds as
>well. What sorts of criteria are you going to use to explicate "right"? I
>would think, though I have trouble imagining how it will be done, that you
>will have to divide functional architectures into something like "natural
>kinds", and then argue that some of these are "intelligent".
>Ready, set, go!

Well, the usual position is that belief is a theoretical term,
characterized as a state that plays a certain causal role in
psychological functioning.  This requires an underlying theory,
with a specification of the role that belief plays in that theory.
Once this is done, then the "right" functional organization
falls out (i.e., systems that satisfy the theory).

Different functionalists hold that the theory should be specified
in different ways.  The so-called "psychofunctionalists", or
empirical functionalists, think that it's a matter for empirical
scientific investigation, and that the scientists shouldn't be
second-guessed.  The analytic functionalists think that on the
contrary, it's mostly a matter of conceptual analysis -- a belief
is something that interacts with other cognitive states in certain
ways (e.g. the belief/desire/action triangle), is caused by 
appropriate stimuli, leads to appropriate kinds of behaviour, etc.
The details of the conceptual analysis aren't usually carried
through in detail, partly because the concept of belief seems to
have very blurry edges; but see Michael Loar's "Mind and Meaning"
for an attempt in that direction.

Personally, I lean toward analytic functionalism for concepts
like belief.  On the other hand, I don't think that qualia
can be dealt with by conceptual analysis alone -- there's
something irreducibly contingent going on, so it's an "empirical"
question what kinds of system will have qualia; the only trouble
is that it's awfully hard to find data.  So one is in the
position of doing speculative theory-building from limited
observations.  I think there are good reasons to believe that
qualia arise from functional organization, but the question
of which functional organizations suffice seems irremediably
underdetermined by the available evidence.  As you know, I think
that it's not implausible that organization as simple as that
of a thermostat suffices for qualia, but that's not a position
that a functionalist is committed to.

I recommend David Lewis's "Psychophysical and theoretical
identification" for an outline of how psychological terms can
be characterized by their role in a theory, and Ned Block's
"Functionalism" for more on this theme, with elaboration
on the empirical and analytic versions of functionalism.
(Both are in Block's anthology _Readings in the Philosophy of
Psychology_, volume 1.)  These articles should also be good
antidotes for the chap who thinks that functionalism is all
about computing mathematical functions.

-- 
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."


