Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site umcp-cs.UUCP
Path: utzoo!decvax!decwrl!pyramid!pesnta!amd!amdcad!lll-crg!seismo!umcp-cs!mangoe
From: mangoe@umcp-cs.UUCP (Charley Wingate)
Newsgroups: net.philosophy
Subject: Re: mind vs. brain
Message-ID: <2274@umcp-cs.UUCP>
Date: Wed, 20-Nov-85 10:24:38 EST
Article-I.D.: umcp-cs.2274
Posted: Wed Nov 20 10:24:38 1985
Date-Received: Sat, 23-Nov-85 04:20:53 EST
References: <1794@watdcsu.UUCP> <2005@umcp-cs.UUCP> <1884@watdcsu.UUCP>
Distribution: na
Organization: U of Maryland, Computer Science Dept., College Park, MD
Lines: 125

In article <1884@watdcsu.UUCP> dmcanzi@watdcsu.UUCP (David Canzi) writes:

>>But I think another analogy fits at lest as well, if not better:
>>    brain:mind::hardware:software
>>or
>>    brain:mind::hardware:process
>>and people DO talk about the dichotomies implied by the latter pair. 

>There's an important difference between the brain:mind::hardware:software
>analogy and the brain:mind::hardware:process analogy that you cite as
>if they were the same thing.  A process is performed by a computer.
>Software, on the other hand, is a *description* of a process for the
>computer to perform.  The computer performs processes by interpreting
>(ie. running) software.

The reason why I gave both analogies is not that I think they are identical,
but that that on or the other was The Analogy.  And while I'll agree to the
distinction between software and process, I think the last sentence should
be essentially reversed.  Processes within the computer are the dynamic
realization of the software, which is essentially a static thing.

>There is no such thing as software.  We can rephrase any statement
>about software to eliminate all references to software as a thing.  For
>example, "this software is portable" can mean something like "Existing
>equipment can detect the magnetic fields near the surface of this disk,
>and magnetize other disks so that the other disks can cause different
>computers to behave similarly."

But these two statements are in fact NOT equivalent; one can replace the
second statement with "This series of symbols will cause the same behavior
regardless of what system it is entered into."  This bears the the same
relationship with the first statement as the orignal second statement did,
namely, that of instantiation.  But the immediate meaning of the this new
statement is not identical to the orignal second statement.

There is in fact a certain level of abstraction on which the notion of
software (or programs, or a number of constructs of equivalent abstraction)
is necessary.  Consider the following statement:

  "This software to implement a relational database is written in Pascal."

Now if you attempt to avoid the word "software" and talk about patterns on
a disk, you will be making an untruth.  The patterns on the disk are
irrelevant except as a representation, and the most immediate representation
is that the disk's file system, not of "software".  So you have to back away
from patterns on a disk and abstract them to files.  Then again, at some
point the software was at least partially written on paper, and there is
therefore another level of abstraction necessary: that of abstract
characters.  But not all patterns of characters are software.  Both of the
two statements (about what the software does and what language it is written
in) refer only to software; therefore, these have to be rewritten to
eliminate such references too.  But the rub is that once you get rid of the
semantics of Pascal it ceases to be a Programming Language-- and you can't
get rid of the abstract operational implications of a "relational database"
at all.  Its entire content is abstracted to the user level!

It seems to me therefore that software IS a necessary concept with respect
to computers.  It describes the realization of programmer intent in the form
of what to the programmer are abstract operations (i.e. they are not
patterns of electrical charge or the like) but which the computer can (by an
appropriate representation and transformation) actually execute and produce
the desired (abstract) behavior.

It should be quite clear from this that I now think that software is
therefore an inappropriate concept with respect to the mind and brain,
because it presupposes the realization of this level of abstraction (even if
we punt on the clear intentionality implied here by assigning it to
"evolutionary processes").

>Some statements containing the word "mind" can be rephrased to eliminate
>the use of any noun equivalent to "mind".  (For example: "That's what's
>on my mind" can be rephrased as "That's what I'm thinking about.")  Some
>statements containing the word "mind" can't be rephrased this way.

Well, it seems to me that commonly there are two different levels on which
the concept of mind is used.  On the lower level, the mind is simply the
dynamic state of the brain.  If one realizes that one is working at this
level, then the phrase and the word can be exchanged with impunity.  The
other level, that of mental abstractions, is much more troublesome.  Both of
the statements above occur at this level, and it is quite important NOT to
confuse this with the lower level.

>Some people use arguments in which the word "mind" plays an important
>part to prove some point.  If attempting to rephrase such an argument
>to remove the use of the word "mind" as a noun destroys the argument's
>validity, then the argument was dependent on the unstated assumption
>that "the" mind is a "thing".  If I don't accept this unstated
>assumption, then I am not forced to accept the conclusion.

>For example, John Searle's attempt to prove that one must be a dualist
>to believe that "strong AI" is possible:

>    Unless you believe that the mind is separable from the brain both
>    conceptually and empirically -- dualism in a strong form -- you
>    cannot hope to reproduce the mental by writing and running programs
>    since programs must be independent of brains or any other
>    particular forms of instantation.

>The phrase "the mind is separable from the brain" can be rephrased as
>"some other system can be programmed to behave like the brain", a
>description which doesn't resemble dualism at all.

No, that's NOT what his statement says.  If you want to get rid of the mind
there, you have to rephrase it as

 "For all brains X, some other system can be induced to produce the same
  abstract behavior as brain X."

It should be apparent that this is so close to the notion of software that I
described above that it is essentially the same as the statement

 "Software can be described for the brain."

(Written in NEUROSYS, of course :-)

It's certainly indisputable that strong AI asserts the last.  It's also
evident that if one relaxes the constraints and forgets about the *abstract*
behavior, one can readily model the brain simply by bulding a machine
comprised of a network of eloctronically simulated neurons, connected
together with appropriate sensors and mechanical manipulators.  But such a
machine isn't what is currently desired; the *abstract* behavior is
*exactly* what AI people are setting out to model.  So they certainly DO
believe in this dualism!

Charley Wingate