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 Thu Jan 16 17:19:24 EST 1992
Article 2613 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: Searle and the Chinese Room
Message-ID: <1992Jan10.005426.24694@bronze.ucs.indiana.edu>
Organization: Indiana University
References: <5815@skye.ed.ac.uk> <1991Dec12.193222.27298@bronze.ucs.indiana.edu> <5909@skye.ed.ac.uk>
Date: Fri, 10 Jan 92 00:54:26 GMT
Lines: 58

In article <5909@skye.ed.ac.uk> jeff@aiai.UUCP (Jeff Dalton) writes:

>We all know that we can follow a recipe to produce a cake.
>But it's hard to see how that's the same as instantiating
>(another word you've changed) the recipe, much less that it's
>the same as relying only on this instantiation to produce the
>cake.  We at least have to use the right materials, and it's
>the physical properties of the materials that ensure we get
>a cake in the end rather than a bowl of sludge.

The implementation of both programs and recipes (I don't use the word
"instantiation", as I don't think that computers instantiate programs,
they implement them) requires that the corresponding physical system has
certain physical properties, specified by the program/recipe.  For
recipes, these are physico-chemical properties; for programs, these
are causal properties.

>Nor is it clear that a recipe is treated as purely syntactic.
>It's necessary to know what words like "egg" and "flour" mean.
>They can't be treated as variables standing for arbitrary
>ingredients, as the analogy with a program would suggest, if
>what you want in the end is a cake.  In effect, there are
>two parts to a recipe: instructions for manipulating ingredients
>(this is the program), and an indication of what ingredients
>to use.  The indication is useless for persons (computers)
>who don't know the semantics, even if they implement (your
>word this time) the program.

Same point.  "Egg" certainly has to lead to eggs in the implementation,
and "S1->S2" has to lead to the appropriate causation in the
implementation.  There's nothing deep about getting this relation
right, though, as evidenced by the fact that it's straightforward in
principle to automate the implementation of both recipes and programs.

>A person following a cake recipe doesn't
>become crumbly; but if computational theories of mind are
>correct, something that employs the right program does become
>"minded".

No -- as any defender of the Systems reply will tell you, the person
following the program doesn't gain a mind any more than the person
following a recipe becomes a cake.

>Now, one might argue that a program could cause a mind to be
>constructed inside a computer, rather like a program might
>cause a computer that controlled a robot to make a cake,
>except that the "cake" would be inside the computer.  But
>if all it constructs is another program (and what else can
>it do if it's still going to be "syntax"?)

Who says it's still "syntax"?  It's real, full-blooded, physical
causation.  Syntax isn't sufficient for semantics, but the right
kind of causation might be.

-- 
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."


