Path: utzoo!attcan!uunet!zaphod.mps.ohio-state.edu!mips!wdl1.wdl.fac.com!wdl1!mikeb
From: mikeb@wdl31.wdl.fac.com (Michael H Bender)
Newsgroups: comp.ai.philosophy
Subject: Re: Intuition and intelligence (was emergent properties)
Message-ID: <MIKEB.90Oct26124238@wdl31.wdl.fac.com>
Date: 26 Oct 90 19:42:38 GMT
References: <3499@media-lab.MEDIA.MIT.EDU> <1990Oct3.183522.17076@riacs.edu>
	<3549@media-lab.MEDIA.MIT.EDU> <45348@apple.Apple.COM>
	<3560@media-lab.MEDIA.MIT.EDU> <4152@bingvaxu.cc.binghamton.edu>
	<15238@venera.isi.edu> <1990Oct23.165301.9813@riacs.edu> <BURLEY.90
Sender: root@wdl1.wdl.fac.com (SUPER USER)
Organization: Ford Aerospace, Western Development Laboratories
Lines: 90
In-Reply-To: fostel@eos.ncsu.edu's message of 24 Oct 90 20:18:51 GMT
Nntp-Posting-Host: wdl31

Gary Fostel writes:
     .....

     Is there a reason that thoughts which are not conscious are
     assumed to be thoughts which are not based on logic?  Certainly,
     from an introspective point of view, there is no way to be able 
     to tell if ones thoughts which are not conscious are following the
     principles of logic, but then again, I have no reason to feel that
     ANYONES thoughts, conscious or otherwise, are following such 
     processes since I do not have access to the set of observations
     they have of their own conscious activities.  Never-the-less, I
     am usually willing to agree that another persons thoughts are
     logical (at least at times :-) based on the observations I do
     have, e.g. things they say to me, actions they take etc.

     If I apply the same standard to my own thoughts which are not
     conscious, I would need to conclude that indeed these thoughts
     must at times be following sound logical processes.  The most
     striking example is probably chess, where the intense logical
     process of find a good move is, for most experienced players,
     not a conscious activity.  Many top level players have observed
     that they can tell quickly (not by conscious means) what the best
     move is but spend their clock time consciously verifying that
     the non-conscious process has not made an error.

     ....

It is my interpretation of the research that the majority of the
"reasoning" activity that goes on in the brain is pattern matching and that
only a very small part of this reasoning actually fits our concept of
"rational" thought. (I assume that you mean "rationality", not "logic"
which is really a formal, mathematical, system.) 

At one extreme let me point at the lower order functions (e.g., perception)
which are clearly pattern matching activities. However, it is also clear
that many high order functions (such as decision making) rely heavily on
pattern matching heuristics (let me point you to the research by A. Tversky
and D. Cahanman (spelling?) who did a number of classic studies in this
area.)

However, it appears that our brain (or to be more exact, the left side of
our brain) does have a wonderful ability to assign rationalizations (i.e.,
rational interpretations) of our pattern matching behavior to make it more
palatable to .... (our consciousness?). (I would refer you to the studies
that have been done on split-brain patients for evidence of this phenomenon.)

     One reaction is to go with the S/R, behaviorist, connectionist
     flow and say that this supposed logical process that is not
     conscious is not really a logical process at all. The best move
     is determined in some other way altogether.  Another reaction
     is to say that the process of logical thought is not inextricably
     bound to the process of conscious awareness.  A crude computational
     metaphore might be that the logical process is farmed out to another
     processor with capabilities quite similar to the one which supports
     the conscious thought process -- except of course that support 
     processor does not have what it takes to support conscious awareness
     of its working.

     ....

It is probable more accurate to imagine a number of "lower level"
processors which use pattern matching and other heuristics to generate data
for a high level processor. These low level processors are goal driven,
just like the high level processor. However, the high level processor has
only a limited ability to impact the goals of the lower level processors.
(As proof, try to not think about a pink elephant for 5 minutes -- you will
note that you cannot completely direct that part of your brain which brings
up the image.)

     There are several reasons why this would be an efficient design.
     The logical process might proceed more rapidly without the extra
     overhead of awareness, and it might be possible to have more than
     one such process concurrently working, all without awareness.  

     If memory serves me, this is one of the ideas woven into Minski's
     Society of Mind book.  Perhaps it is also the reason he and others
     remain skeptical and critical of connectionists who try to squeeze
     symbolic, logical thought processes into non-symbolic, non-logical
     models.  Connectionists may be too readily combining introspective
     evidence that some processes are divorced from conscious awareness,
     with the assumption that only conscious processes are logical and
     prematurely concluding that something else must be going on.  

I don't think you have to be a "connectionist" (I presume that its a
derogatory term) to believe that there is a fair amount of connectionist
type processing going on in the brain. I.e., I still doubt very much
whether consciousness is an emergent property from a connectionist
architecture.  

Mike Bender