From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!usc!sol.ctr.columbia.edu!bronze!chalmers Tue Apr  7 23:23:49 EDT 1992
Article 4884 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca comp.ai.philosophy:4884 sci.philosophy.tech:2500
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!usc!sol.ctr.columbia.edu!bronze!chalmers
>From: chalmers@bronze.ucs.indiana.edu (David Chalmers)
Newsgroups: comp.ai.philosophy,sci.philosophy.tech
Subject: Re: A rock implements every FSA
Message-ID: <1992Apr2.202642.25800@bronze.ucs.indiana.edu>
Date: 2 Apr 92 20:26:42 GMT
References: <1992Mar30.064140.8996@bronze.ucs.indiana.edu> <1992Mar30.231418.10488@husc3.harvard.edu> <45844@dime.cs.umass.edu>
Organization: Indiana University
Lines: 48

In article <45844@dime.cs.umass.edu> orourke@sophia.smith.edu (Joseph O'Rourke) writes:
>In article <1992Mar30.231418.10488@husc3.harvard.edu> 
>	zeleny@zariski.harvard.edu (Mikhail Zeleny) writes:
>
>>   ..., and printing and erasing what it is 
>>   supposed to print and erase when it is in a given state and scanning 
>>   a given symbol)
>>
>>Now, if the last anaphoric pronoun were meant to refer to the machine with
>>description D, the theorem proved earlier would have nothing to do with
>>this machine.  Since I know Putnam not to be an idiot, I concluded that it
>>referred to the object being modelled, and the machine with description D
>>was meant to satisfy the theorem's constraints.  
>
>	This is exactly how I interpreted it upon first reading, mentioning
>in an old post that D acts as if it "hallucinated" the input.  But David
>Chalmers questioned my interpretation, and after reading it several times
>now, I think that Putnam means D to have I/O that mirrors S's I/O.  I hesitate
>to take sides in this increasingly vitriolic exchange, but I think David
>is right on this point.  But you are right in that the theorem has nothing
>to do with D!
>	Only one point have we collectively established beyond a shadow
>of a doubt:  the "Discussion" in this Appendix is very poorly written!

Once again, strangely enough, I have to come to Putnam's defense:
although the appendix is vague in places, it seems to me that at this
point it is entirely clear.  Putnam is saying that *if* there is an FSA 
D that could produce the same I/O behaviour as S, then it turns out, by
similar reasoning to before, that S is *in fact* an implementation of D.

This seems crystal-clear to me.  It's entirely appropriate for the
pronoun to refer to D, in establishing that D is a machine that could
produce the same I/O behaviour as S.  And "showing" the equivalence of
D and S is then, as Putnam says, a straightforward application of the
methods of proof in the earlier theorem (not, I note, a direct
application of the theorem itself).

Of course clarity does not imply correctness, and the proof here fails
for the usual reasons -- i.e. because S doesn't satisfy the strong
conditionals required of an implementation of D.

(In fact S needn't even have the same behavioural *dispositions* as
D, so even a behaviourist like Ryle wouldn't think it equivalent.)

-- 
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."


