From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rutgers!micro-heart-of-gold.mit.edu!uw-beaver!cornell!batcomputer!munnari.oz.au!bunyip.cc.uq.oz.au!uqcspe!cs.uq.oz.au!bakker Tue Nov 19 11:09:16 EST 1991
Article 1225 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca comp.ai.philosophy:1225 sci.psychology:1139
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.csri.toronto.edu!rutgers!micro-heart-of-gold.mit.edu!uw-beaver!cornell!batcomputer!munnari.oz.au!bunyip.cc.uq.oz.au!uqcspe!cs.uq.oz.au!bakker
>From: bakker@cs.uq.oz.au (Paultje Bakker)
Newsgroups: comp.ai.philosophy,sci.psychology
Subject: SUMMARY of Responses : Animal vs Human Intelligence
Message-ID: <5010@uqcspe.cs.uq.oz.au>
Date: 7 Nov 91 05:04:30 GMT
Sender: news@cs.uq.oz.au
Reply-To: bakker@cs.uq.oz.au
Followup-To: comp.ai.philosophy
Organization: Depts. of Computer Science & Psychology, The University of Queensland, Brisbane, Australia
Lines: 365

SUMMARY of responses to the posted query:

"What is the basic difference between human intelligence and animal
intelligence? What is it that we can do that sets us apart from all
the other animals?
Is it the use of analogy, or our highly developed speech, or the ability
to reason logically, or something else? Is there a category of tasks
that only humans will ever be able to perform successfully?"

A big thank-you to the MANY people who responded!
>From the responses, it appears the main point of contention is still 
whether we are on a continuum with animals, or whether we stand 
(intellectually) far apart from them.

I've included excerpts from email and follow-up postings on the net;
less relevant matter (eg as occurred later in the discussion) has 
been omitted.

paul bakker
bakker@cs.uq.oz.au
---------------------------------
>From: PSYBM@lure.latrobe.edu.au (Catherine Dwyer)

The search for the definitive answer to what sets humans apart 
from animals has been going on for some time.

In the beginning, philosophers thought that simple things like
tool-use set man apart, but detailed ethological studies revealed that
many animals used tools.  So then it was claimed that language was
enough to define human from animal.  Again, this argument lost
credibility through the work of animal researchers who taught
chimpanzees in particular (and some gorillas) to use sign language.
Ultimately, the definitive difference between humans and animals was
thought to be self awareness, ie. we have it, they don't.  This view-
point also has been questioned by researchers using the mirror
recognition technique.  So basically, we STILL don't know exactly
where the true difference between animals and humans lies.  Most
philosophers (Kinget, 1975; Buss, 1988) believe that the difference
is qualitative, while others think the difference in animal intelligence
is only quantitative (Griffin, 1976).

Through my own reading, I have come to the conclusion that some
animals (chimps, orangutans, gorillas maybe) have a form of self
awareness that is similar to humans, but lacks a true reflexive
aspect.  That is, in order to be self aware an organism must have
a self concept, and this is only developed through interaction with
others in which the OTHER'S  concept of you is incorporated into
your world view (this is a highly simplistic version - see G.H Mead
for a thorough run down on this).  While this may be true of the
animals mentioned above, there is no evidence that they go the step
further that I believe humans do in that humans also incorporate
an IDEAL view-point of themselves, and this leads to second order
emotions like shame and embarrassment which so far, is difficult
to establish in animals.  The capacity for self awareness is based,
I believe, on the ability to manipulate symbolic representations.
Thus, while most animals are capable of symbolic representations, I
think the difference between humans (and those animals shown to be
capable of self awareness) and animals lies in our ability to
manipulate these representations - that is, second order symbolic
representation.

Some interesting reading:
The Ape's Reflexion - A.J Desmond (1979)
Intelligence in Ape and Man - D. Premack (1976)
The Question of Animal Awareness - D.R Griffin (1976)
Animal Mind-Human Mind - D.R Griffin (Ed.) (1982)
On Being Human - G.M Kinget (1975)
Anything by Gordon G. Gallup.
Chimpanzees, Symbols and the Reflexive Self - J. Meddin (1979) in
Social Psychology Quarterly, 42, 99-109.

--------------------
>From: William Calvin <wcalvin@u.washington.edu>

      "Humans are separated from the apes by a number of augmented
mental abilities.  1) Multi-tasking mechanisms allow us to perform
parallel tasks simultaneously, as when a cook prepares a multicourse
meal.  2) Serial-sequential mechanisms permit grammatical language and
plan-ahead-for-tomorrow, plus such novelties as music and dance.  3)
Mechanisms for creativity seem to exist that evade the expectation that
most random changes in a standard behavior are likely to be detrimental
or even dangerous.  An explanation of higher intellectual function at a
neurophysiological level would need to account for these augmented and
novel behaviors, as well as addressing some questions that apply widely
among animals, e.g., 4) how new memory items are stored and retrieved;
5) how choices are made between alternative behaviors; and 6) how
categories, metaphors, and analogies are implemented.
      One approach to this multifaceted problem is to examine
darwinian evolutionary mechanisms, especially as exemplified by island
biogeography, for features that known brain mechanisms might mimic.
Darwinism is our baseline mechanism for creativity in nature; there may
be others, but we know that darwinism's environment-dependent
differential reproduction is capable of achieving new stable levels of
complexity, shaping up a new species within millennia and, within a
week or two, an antibody to fit a novel antigen.  Here I propose a
seconds-to-minutes neural mechanism that could exhibit the same wide
variations-on-a-theme as seen elsewhere in nature, shaped up by a
remembered environment, evolving via a pseudo-reproductive process in
a population of neural-encoded patterns to yield novel complexities such
as this sentence...."

(          to appear in a special issue of
           Seminars in the Neurosciences (November 1991)
	"Evolving Neural Functions" edited by W. H. Calvin & K. Graubard)

Two of my books also address this question:
THE CEREBRAL SYMPHONY:  Seashore Reflections of the Structure of
   Consciousness (New York, Bantam, 1989)
THE ASCENT OF MIND:  Ice Age Climates and the Evolution of
   Intelligence (Bantam 1990)

--------------------
>From: markman@aristotle.ils.nwu.edu (Art Markman)

The first chapter of Newell's 'Unified Theories of Cognition'
(Harvard U. Press, 1991) defines intelligence as the extent to which
a system approximates a knowledge-level system.  (I couldn't possibly
do justice to his explanation in a short space.  It's a good chapter
though).
At any rate, he links this to a human vs. animal intelligence debate by
noting that processing limitations on many animals keep them from
appropriately modelling a 'rational' (in the logical sense) system.  I
think that a number of systems like language and analogical reasoning
(to name but two) are designed to increase the probability that we will
achieve an optimal problem solution in finite time.

Anyhow, I suggest the first chapter of Newell's book for a discussion of
the issue.  John Anderson's 'The Adaptive Character of Thought'
discusses some related issues.

-----------------------
>From: orpheus@reed.edu (P. Hawthorne)

Noam Chomsky summarized animal intelligence experiments by concluding
that animals, so far, have only the cognition to feel and communicate
emotions along single dimensions scales. Warbling at a lower pitch, for
instance, means that they feel an emotion with less intensity than when
they warble a higher pitch.

By the same token, an animal could not do arithmetic or identify a visual
representation of a number to represent how many bananas there were on a
table. Chomsky cited one experiment in which apes were supposedly taught
to recognize numbers, but it took them a very, very long time to get a
couple of small numbers down and the numbers were shown in binary form.

I forget the name of the book, but I can ask my Siamese cat to get it 
out of our library for you if you want.

There is a song on the XTC album Skylarking that my cats really enjoy.
It has lots of bird sounds and sundry jungle noises. Artemis, a Siamese
cat, has walked up to a CD player, queued the first track with an arrow
button, set it to replay forever, and pressed play, then walked away to
sit down and listen. Kinda like the old RCA mascot.

(what a cat! :-) PB)

---------------------
>From: cbarber@bbn.com (Chris Barber)

I really do not see why it is necessary to examine animal VERSUS human
intelligence.  We are animals, and we also have the same basic learning
mechanisms that "lower" animals have.  Of course, we might have other
mechanisms in addition to the ones, say, dogs have.  But then again, it might
simply be a matter of computing power that allows us to do the obvious things
that seem to be lacking in other animals: abstract reasoning and language.  I
suspect that if we truly understood how the brain worked, we would probably be
able to observe at least the rudiments of our own "kind" of intelligence in
many mammals (perhaps not cockroaches).  My point is really that I think it is
more productive to think of "intelligence" as being more or less equivalent to
the process of adaptive learning rather than as a property of an organism.
Any organism which is capable of some kind of adaptive learning, is to some
extent, intelligent.  Considering our evolutionary history (I hope I am not
offending all you fundamentalist AI'ers out there), it seems likely that we
have built upon the learning capabilities of our ancestors.  It seems
reasonable to postulate the extra mechanism or ability which when added to
mechanisms inherited from our "lower" forms gives us the degree of
intelligence which "sets us apart" from other animals.  But in order to really
understand what this extra mechanism or whatever is, we probably have to
understand the foundation first. So instead, I would ask the following
question:

What about the way humans think and learn is the SAME as other animals?

--------------------
>From: map@u02.svl.cdc.com (Mark Peters)

There is a hierarchy of knowledge, going from most concrete to most
abstract:

  1) Sensation (automatic reaction of sense organ to stimulus)

  2) Percept (automatic integration of two or more sensations into a
     single mental unit, performed by the nervous system.  Yields
     awareness of entities as opposed to flux of disintegrated sensations)

  3) Concept (non-automatic integration of two or more percepts and/or
       other concepts into a single mental unit)

Percepts are formed automatically by the nervous system
in response to sensory input, while concepts can be formed only by
the deliberate focusing of the mind on the differences and similarities
between percepts (on the lowest level) and/or other concepts (on the
higher levels).  While virtually all life forms can reach the sensory level,
and many reach the perceptual level, only human beings can reach the
conceptual level - and that is the great dividing line between man and
other animals.

--------------------
>From: jason pouflis <s157958@cs.uq.oz.au>

BIPEDALISIM
Bipedalism allowed the hands to be freed from movement tasks,
The hands could be used for other purposes such as
TOOL USE
The use of tool increased the efficiency of food collection.
Along with omnivorism and agriculture human's dominance increased.
LANGUAGE
Some theories suggest that spoken language developed from
hand language/signals. The language ability allowed vast amounts of survival
information to be passed form generation to generation.

These 3 factors are the main markers (ignoring morphism)
to distinguish humans from others.

----------------------
>From: pnettlet@neumann.UNE.Edu.AU (Phillip Nettleton)

The main difference as I see it is sheer capacity (a rating of intelligence
is a rating of the capacity for intelligent activity). Human's have a huge
capacity for intelligence, mainly due to the size of the cortex. Even our
closest cousins (the great apes) have a large difference in capacity to
ourselves. Dolphins, on the other hand, may well have a similar capacity,
but as yet we can't communicate effectively with them.

> Is it the use of analogy, or our highly developed speech, or the ability
> to reason logically, or something else? Is there a category of tasks
> that only humans will ever be able to perform successfully?

A linguist would probably disagree with me (they have a much stricter
definition of language) but all high level creatures (down to insects)
have a capacity for language. We have a high capacity for analogy and
logic and animals have a lesser capacity (again with a large gap).

This large gap in capacity is probably of our own making, we probably
killed off our rival intellectual species.

-------------------
>From: noring@netcom.com (Jon Noring)

Humans can create, that is, to dream of things that do not yet exist
and then create those things with directed action.  The goal to land
a man on the moon was one such big dream.  Of course, one could argue
that some simians (apes, chimpanzees) can visualize the future and
then go into action to make it happen, but I believe that this is
extremely limited and is used only to fulfill basic needs (established
instinct).

Just my $0.02 worth.

----------------
>From: nagle@netcom.COM (John Nagle)

    "However, the evolutionary development of the brain appears to
be quantitative and not qualitative.  This is even true of the cerebral
cortex..."  [From "Evolution of the Brain", by Sir John Eccles, sec. 3.1]

     From a hardware perspective, all the mammals seem to have similar
brain architecture.  Measured by DNA differences, dissection, time to evolve,
or electron microscopy, [see Braitenberg's "Anatomy of the Cortex"] the
mammals seem to differ mostly in the sizes of the various brain components.

     This leads one to expect that if we can build anything with the
intelligence of a low-end mammal (a mouse, say) scaling it up will
make it smarter.  A mouse has about 1gm of brain mass.  A human has about
1000g of brain mass.  If you believe Moravec (in "Mind Children") 1gm
of brain mass equates to about 1000 to 10,000 MIPS.  Machines in that
range exist now, but don't achieve mouse-level intelligence.  It's
worth noting, though, that as autonomous vehicles approach that
computational range, they start to work, as the CMU NavLab efforts demonstrate.
This may indicate that we may be entering the computational range needed
to deal with the real world with an animal level of competence.

---------------------
>From: rick@xing.unocal.com (Richard Ottolini)

Humans have the extra brain brownpower to control time-- that is contemplate
memories and imagine/plan the future.    I hypothsize imagining is selective
remembering.  I spend most of my mental activity
activity not in the immediate world of sensation, but in artificial
worlds constructed by culture and communication.
I would image that being an animal feels much like having amnesia.
You'll evoke memories in response to immediate sensation, but you can't
suspend much in consciousness for contemplation.

----------------------
>From: kla!zardoz@sun.com (Phillip Wayne)

There is no basic difference between human and animal intelligence.
The primary difference seems to be one of complexity of task, and
even that is often blurred. I live with a Macaw who has learning to
pick the lock (!!) on his cage. This is a most complex behaviour,
apparently learned from simply watching.

Dogs, cats, and apes can often out perform young humans in many
tasks. There is no *qualitative* difference here. Only a quantitative
one.

To judge from Washoe (the only example I have really familiar with)
humans do not have a monopoly on logical reason or use of analogy. A
better question to ask might be "What is the entire set of human capa-
bility? Is it unique?"

At least to the second question, I would be quite tempted to answer
"No.".

-----------------------------
>From: map@u02.svl.cdc.com (Mark Peters)

The Washoe experiment (and many others) was thoroughly debunked by a
behaviorist by the name of Herb Tares (sp?).  Tares put a chimp
named Nim Chimpsky thru the same kind of training given to Washoe,
but he also videotaped every interaction between Nim and humans.

Upon analyzing the videos, it was discovered that a large majority of
the signs that Nim gave were revealed to him (unknowingly) by the
humans, fractions of a second before he signed.  Tares (who prior to
the experiment thought apes had language) concluded that Nim was
using signs as a perceptual-level mechanism to get what he wanted -
no different in principle from a dog shaking hands for a treat.

Tares further concluded that Nim had no understanding whatsoever of the
meaning of the signs he used without prompting - he merely associated
one perceptual level thing (a sign) with another (e.g., apple, bird).
In short, all Nim needed to do what he did was a functioning sensory
apparatus, memory, and the ability to associate one object with another -
nothing even remotely close to a conceptual faculty was required.

Shortly after Tares' book came out, the money for ape language
research dried up, and research in the field dropped nearly to zero.
Today, poor Washoe has to depend on charity to stay alive.

--------------------------
>From: todd@juno.elcom.nitech.ac.jp (Todd Law)

I think the basic difference between human and animal intelligence
is that humans are largely able to abstract a wide variety of inputs,
attach symbols to these inputs (often, but not always in the form
of words), and juxtapose these symbols in a variety of other
arrangements.  Of course, some evidence of this is seen in higher
level animals, which I see mostly as being at the low end of the
abstraction continuum.

Some people may not like this, but we are symbolic machines,
regardless of the various emotions (symbols themselves!) present
in all of us associated with these symbols.

--------------------------------



--
PaulBakker ------------------------------------- email:bakker@cs.uq.oz.au
Depts.ofComputerScience/Psychology,UniversityofQueensland,Qld4072,Australia

"My CPU is a neural network processor - a learning machine" - Arnie, T2


