From newshub.ccs.yorku.ca!ists!torn!utgpu!news-server.csri.toronto.edu!rutgers!att!princeton!phoenix.Princeton.EDU!harnad Tue Jun 23 13:20:54 EDT 1992
Article 6280 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!torn!utgpu!news-server.csri.toronto.edu!rutgers!att!princeton!phoenix.Princeton.EDU!harnad
>From: harnad@phoenix.Princeton.EDU (Stevan Harnad)
Newsgroups: comp.ai.philosophy
Subject: Re: Transducers
Message-ID: <1992Jun17.132117.9273@Princeton.EDU>
Date: 17 Jun 92 13:21:17 GMT
References: <1992Jun10.203412.19158@news.Hawaii.Edu> <6980@pkmab.se>
Sender: news@Princeton.EDU (USENET News System)
Organization: Princeton University
Lines: 68
Originator: news@ernie.Princeton.EDU
Nntp-Posting-Host: phoenix.princeton.edu

In article <6980@pkmab.se> ske@pkmab.se (Kristoffer Eriksson) writes:
>
>As for the retinae cochleae and other parts of the human body being
>peripherals, there is a very easy way to define a border between them
>and the "central core": Define the grey stuff in the big hollow room
>in the skull as the "central core", and call it "brain". Define the
>rest as "peripherals". Perhaps not "mere" peripherals, but nevertheless
>peripherals. The interface consists of nerve fibres and blood streams,
>transmitting nerve pulses, hormones, nutrients, and possibly some other
>signals.
>
>I can define boundaries other ways too (wherever I like)

You can of course define anything any way that you like, but for
common discourse it helps if the same thing is meant by the words
of all the interlocutors. When I say "brain" I mean what
neuroanatomists and neurophysiologists mean by "brain" and the
what I mean by "retina" is literally part of that thing. So
when I say the brain cannot be just a computational core connected
to peripheral transducers, but rather the peripheral trasnducers
(at least in the case of the retina) are literally part of the
brain itself, I am referring to these anatomical and physiological
facts.

My arguments about transduction being an essential part of brain
function rather than an independent module were based in part on this
fact, in part on the fact that transduction is sufficient to immunize a
system against Searle's Argument, and in part on the essential role it
plays in giving a system TTT capacity. I made the further argument
that, having given several reasons why sensory transduction is an
essential part of brain function, there is no longer any reason to
believe that everything that goes on in between sensory and motor
transduction is merely computational either. A lot of it could be
analog all the way through (and it is in the service of this that I
pointed out that large parts of the brain are isomorphic ("tonotopic")
spatial analogs of, for example, the retina. I am not claiming that
some of those upstream analog transductions could not have been
accomplished by other means, e.g., by a Sparc (a lot of the brain's
parallel processing could also no doubt been done serially, in
principle), it's just that that's not the way the actual brain does it
(and there are probably good evolutionary reasons why not).

So let's keep the logical points seprate from the empirical ones.
Empirical point: Parts of the brain are doing transduction.
Logical point: Not all of the brain is a computational core.
Empirical point: Large parts of the brain are doing analog processing.
Logical point: Not everything between transducers and effectors is
               necessarily computational
Logical point: A TT-passing computer alone has no mind (Searle's Argument)
Logical point: The meanings of the symbols in a computer alone are
               ungrounded; they are parasitic on the interpretations
               we creatures with minds project on them
Logical point: A TTT-passing robot is immune to Searle's Argument
               and the meanings of its symbols are grounded in its capacity
               to discriminate, identify, and manipulate the objects,
	       events and states of affairs that they are
	       systematically interpretable as being about
Logical point: Take away the transducer function (and the rest of whatever
               analog function is going on in there) from a TTT-scale
               robot and (a) it is no longer able to pass the TTT,
                         (b) it is no longer grounded
                         (c) it is no longer immune to Searle
                         (d) it no longer has a mind
Empirical point: The real brain is none of (a) - (d).
-- 
Stevan Harnad  Department of Psychology  Princeton University
harnad@clarity.princeton.edu / harnad@pucc.bitnet / srh@flash.bellcore.com 
harnad@learning.siemens.com / harnad@elbereth.rutgers.edu / (609)-921-7771


