From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!uunet!tdatirv!sarima Tue Nov 19 11:09:58 EST 1991
Article 1304 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!cs.utexas.edu!uunet!tdatirv!sarima
>From: sarima@tdatirv.UUCP (Stanley Friesen)
Newsgroups: comp.ai.philosophy
Subject: Re: Animal Intelligence vs Human Intelligence
Message-ID: <270@tdatirv.UUCP>
Date: 13 Nov 91 20:12:41 GMT
References: <3883@papaya.bbn.com> <37713@shamash.cdc.com> <1991Nov05.084137.29880km <37859@shamash.cdc.com> <5868@tamsun.tamu.edu> <37889@shamash.cdc.com>
Reply-To: sarima@tdatirv.UUCP (Stanley Friesen)
Organization: Teradata Corp., Irvine
Lines: 89


In some article map@svl.cdc.com  (Mark A. Peters) writes:
|In <3883@papaya.bbn.com> cbarber@bbn.com (Chris Barber) writes:
|>Wrong.  There is much opportunity for "error" in perception.  The brain (to
|>repeat an earlier posting) is NOT a tape recorder!  It does not perfectly
|>record all the information conveyed by the senses but consolidates it into a
|>more compact representation in doing so, some information must be lost and
|>this could include whether or not an apple or a dog or whaterver was
|>perceived.  
|
|But you are confusing the *conceptual* process of identification with
|perception.  We most certainly can err in our identification of the
|things we perceive, but this error is not attributable to the sensory-
|perceptual mechanism, it is an error on the conceptual level.  

No, Chris is *not* confusing concept and percept!
He is *not* talking just about mistaken identification.
See below for more detail.

|  However an organism's sensory-
|perceptual mechanism works, it *does* work in that way, and in no
|others, even if it is damaged or otherwise "abnormal."  Under a
|given set of circumstances, the sensory-perceptual mechanism will
|respond in a given way and always in that way, as dictated by the
|nature of the circumstances and the nature of the mechanism.

This is most certainly false for every component of the nervous system past
the individual sensory neurons.  *All* neural conections, even those mapping
the raw sensory input onto the processing areas of the CNS are capable of
*learning*, that is of conditional adaptation.  Since your 'percept' is
a level above the raw sensory data, it *cannot* be a simple mechanical
derivation of a given set of sensory inputs.  Any association or interaction
among seperate unitary signals is a trainable/learnable response.  This is
*basic* neurology.

The output of the eye is a complex function of the impinging light levels,
and it certainly does *not* retain all of the original information content,
it is already filtered.

The olfactory bulb and the olfactory cortex actually perform most of the
job of odor recognition directly on the immediate olfactory inputs.  And
this process *is* a learned one.  This was discussed in quite some detail
in a Scientific American article a few months ago.

By the time you get to something as complex as "apple", you are at least
*five* synaptic junctions away from the eye!  This is more preprocessing
than all the Cray's in the world could do in 10 times the time it takes
the human brain to do it.  And almost all of this is what we would call
subconscious.  And *every* *one* of those five+ levels performs a contingent
analysis, involving prior experience, surrounding context, and probably
emotional state.  In fact the main job of the primary and secondary visual
cortices is to integrate contextual and experiential information into a
cohesive whole.

I repeat, your 'percept' is not in any way automatic!
It is at best merely subconscious.

|>When in fact it is much more like this (if we accept your categories at face
|>value):
|
|>        Sensation -> Percept +> Concept -+
|>                        ^ ^  |     ^     |
|>                        | |  |     |     |
|>                        | +--+     +-----+
|>                        |                |
|>                        +----------------+
|
|>And this is an oversimplification itself.  
|
|Well, I'm not real sure what your symbols mean.  The view I'm advocating
|holds merely that concepts have their roots in percepts, and percepts
|have their roots in sensations.

And Chris is saying that percepts are formed by contingent feedback loops
involving both percepts and concepts.  They are rooted in an interaction
between concepts and sensations.

|BTW - The view I'm advocating isn't mine in the sense that it is
|      original with me.  I've made it "mine" by learning it and  
|      its validation - credit for originating it goes to Ayn Rand
|      and her philosophy of Objectivism.  
|
|      Any errors are wholly mine, however.

Not really, from what I have heard Ayn Rand has no idea how the brain works.
-- 
---------------
uunet!tdatirv!sarima				(Stanley Friesen)



