From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!news.cs.indiana.edu!mips!darwin.sura.net!gatech!mcnc!aurs01!throop Tue Jun  9 10:06:08 EDT 1992
Article 6018 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!news.cs.indiana.edu!mips!darwin.sura.net!gatech!mcnc!aurs01!throop
>From: throop@aurs01.UUCP (Wayne Throop)
Newsgroups: comp.ai.philosophy
Subject: Re: Hypothesis: I am a Transducer
Summary: Computationalism = Homuncularism ... NOT
Keywords: computation, transduction, homunculus, sensorimotor physiology
Message-ID: <60757@aurs01.UUCP>
Date: 1 Jun 92 21:40:32 GMT
References: <1992May31.145204.16357@Princeton.EDU>
Sender: news@aurs01.UUCP
Lines: 62

> harnad@phoenix.Princeton.EDU (Stevan Harnad)
> Although it is
> an over-simplification, consider my hypothesis to be that you ARE a
> transducer. If that hypothesis is correct, then there may be many
> different ways to implement you -- namely, all the different ways of
> implementing a transducer with your capabilities (TTT) -- but among
> those ways is definitely NOT one in which instead of a transducer there
> is a computer simulation of a transducer (a "virtual" transducer).

I'd be more interested in WHY a transducer can't be implemented
as a process which is a realization of a symbol system.  I see no
adequate support for such a position.

> I might add that there is some real homuncular thinking involved in the
> persistent misunderstanding of my hypothesis. People keep reverting to
> the rival computational hypothesis (which I have tried to show is
> false, Harnad 1989, 1990, 1991, 1992) in which you are a computational
> core, with the transducers simply carrying information TO it ("you"),
> as our senses do to "us."

I can see how my questions can be interpreted in this way (that is,
that they assume some sort of homnucular background).  But that is
a misinterpretation of what I was getting at.

First, I wasn't "reverting" to a computational hypothesis.  I was
attempting to see how and why the TTT rules OUT a computationally
implemented testee.  As near as I can tell, it does not.  If it
DOES rule it out, I'd be pleased to have it explained to me how.
(And I'm grateful for the ftp-able references to pursue.)

Given that I'm trying to fit a computationally implemented testee into
the TTT framework, there is a natural tendency to wonder about
redrawing the entity/environment boundary at a layer "inside" the
TTT-satisfying acouterments.   But this is no more a "homnucular" bias
than than the before-after case of an amputee.  Certainly the patient
after surgery isn't a "homunuculus" of the patient before.  But some
interesting questions about identity result in either case.

Just as Searle can be accused of "homnucular" thinking because his
thought pump contains a homunuculus-like operator which he uses to try
to make his point, so it is with the computational-derived thought
pumps that try to make the reverse point.  Yet I think it unfair to
characterize *either* scenario in this way.

> [...] I am beginning to think that that perseverative
> computationalism is just homuncularism (an unhelpful form of dualism).

Not at all.  The computational entity in computationalism contains no
homnuculus.  The computation can be teased apart in just the way it is
supposed that a non-computational "transducer/effector" system can be
teased apart.  Neither decomposition leaves a homnucular core.  It is
only when the (it seems to me false) wall is built between computation
and "analog-ness" or "transduction" that a decomposition leaves behind
an inner homnuculus.

But whether you strip away a computational outer core (the CR) or a
"transductional" outer core (the computational TTT testee), neither
"transductionalism" or "computationalism" is genuinely homnucular.  The
fault is in not pursuing the decomposition across the artifically
raised boundary.

Wayne Throop       ...!mcnc!aurgate!throop


