OF COURSE YOUR DOG CAN THINK!

I wrote this small exploration of animal intelligence in
1994 and self published it on my own computer. I advertised
it in a few dog magazines and sold a dozen or so copies. I
also distributed a small number to friends and sent copies
to a few authors of books dealing with animal intelligence.
I am reprinting it now because I think it can shed some
useful light on an exciting new development. In June of
this year newspapers all over the United States and Europe
carried stories about, Rico, a German house dog able to
fetch any of 200 objects on command by name. "German Pet's
Vocabulary Stuns Scientists" was the sub heading in the
Washington Post of June 11, 2004. Rico can also, by a
process of elimination, apply a word he has never heard to a
toy he has never seen, and fetch the right toy.

My 1994 exploration of animal intellience makes clear three
important things about this dog's abilities.

First the ability of a dog to learn the names of 200 objects
and retrieve them on command, far from being something
unbelievable or astounding, is exactly what we should
expect. In the last section of this little book, THE
CHALLENGE OF ANIMAL MINDS, the reader will note that I
proposed an experiment which I predicted would produce
precisely the result which now "stuns scientists". I put it
this way:

"I specifically propose the further exploration of one major
area of animal intelligence. I have indicated my belief
that a great deal of our intelligence derives from our
capacity for meaningful perception. If this is the case
then one way to probe the magnitude of animal intelligence
would be to probe the limits of their ability see. How much
of our own visual continuum can animals discriminate? How
well can they remember such visual discriminations? To what
extent can they learn to associate what they see with
something else? What is the full range of conceptualization
that inheres in animal perception? Stanley Coren uses his
command "find glove" as an obedience exercise. If we
generalize this finding exercise, and let the animal find
things in plain view we could probe the limits of visual
discrimination, sound discrimination, thing memory and word
memory and powers of association in one extended experiment.
If continued over the years with much repetition (as we have
absorbed our millions of knowledge items) I suspect the
number of 'findees' would grow continuously.

"The essence of higher intelligence is generality. We might
be very much impressed with a certain feat of perception,
but if it turned out to have the specificity of a smoke
detector, we would know the intelligence involved was
minimal. On the other hand a feat of visual discrimination
on the part of a dog might be extremely mundane. But if the
dog has the ability to make thousands of such
discriminations then substantial intelligence must be
involved."

I also produced a four page paper titled "An Invitation to
Experiement" which I distributed to later purchasers of my
booklet. I enlarged upon the how's and why's of my proposed
experiment and added my prediction that "the number of
'findees' would grow continuously, frequently into the
hundreds and in some cases into the thousands."

The reader will find this "Invitation to Experiment" at the
end of this book.

The second thing which this book makes clear is that the
ability of a dog to retrieve hundreds of objects identified
by name could be more important for what it tells us about a
dog's capacity to see than about a capacity for learning
words, including the ability to generalize or form concepts
of categories of objects.

I might add that when we speak of conceptualization or
generalization in dogs we mean just when Darwin meant when
he wrote: "But when a dog sees another dog at a distance,
it is often clear that he perceives that it is a dog in the
abstract; for when he gets nearer his manner suddenly
changes if the other dog be a friend."

The third thing that this book makes clear is that we have
barely sratched the surface in testing animal intelligence.
As I put it in my section on THE CHALLENGE OF ANIMAL MINDS:
"What is usually left out of discussions of learning, human
or animal, is what I call the culture of learning. We have
made the point that people have millions of pieces of
knowledge that they are largely unaware of. Where did we
get that knowledge? I believe it is obtained from a
learning environment, that is culture.

"A culture, or subculture, is an environment in which the
individual is continuously exposed to the elements of that
culture. It is thus an environment of continuous, largely
unattended learning."

...

"It is this continuous unattended learning provided by the
child's culture that is so difficult to provide an animal."

As I put it in my write-up, "An Invitation to
Experiement": "Thus no test of an animal's ability to
learn which does not continue over the years can
remotely compare with enduring learning experiences
that have provided people with most of their cognitive
abilities. Thus no test has shown that animals cannot
learn to remember and distinguish hundreds or thousands
of items."
Here is an opportunity for a researcher to make a
unique and signficant contribution.

___This book contains the following chapters _______

1 EXPLORING ANIMAL MINDS 4
2 DARWIN OR DESCARTES? 4
3 WHAT IS INTELLIGENCE? 7
4 INTELLIGENCE AND THINKING 9
5 INTELLIGENCE AND INSTINCT 10
6 HUMAN INTELLIGENCE 11
7 ANIMAL MEMORY 13
8 INTELLIGENCE AND VISION 13
9 ANIMAL VISION 15
10 INTELLIGENCE AND REASON 17
11 ANIMAL REASON 18
12 ANIMAL SENTIMENT 19
13 CONCEPTUALIZATION 20
14 LANGUAGE 23
15 THE DIMENSIONS OF BIOLOGICAL INTELLIGENCE 25
16 CONSCIOUSNESS 27
17 WHY SO DEFENSIVE? 33
18 THE CHALLENGE OF ANIMAL MINDS 35
------------------------------
Here we have added
AN INVITATION TO EXPERIMENT 37

___Here follows my book as published in 1994________


OF COURSE YOUR DOG CAN THINK!

Nothing in modern science has been more
exciting than all the diverse discoveries
which have spawned a vast new respect for
animal intelligence. First we have a great
enlargement of the concept of intelligence
itself stemming from developments in brain
research, psychology, philosophy, linguistics
and computer science. It also derives from
wide ranging studies of animal behavior,
including the amazing breakthroughs in animal
language learning. And the excitement is no
way diminished by the fiery debates among
scientists and philosophers over the meaning
of these discoveries.

Ironically, one man seems to have had it all
figured out one hundred and twenty years ago.
Charles Darwin's observations on the animal
mind often sound like a report on the newest
discoveries. So I call upon this molder of
the modern mind to help tell the story and I
include in my appendix a chapter from his
monumental DESCENT of MAN published in 1871,
COMPARISON OF THE MENTAL POWERS OF MAN AND THE
LOWER ANIMALS.
(This is Chapter III from Darwin's book and it is
omitted from this printing. You can easily download it
from the Internet and if you don't have it we strongly
recommend that you do, because it is one of the best
things ever written about animal intelligence.)

Copyright 1994 by Harold E. Wefald

*EXPLORING ANIMAL MINDS

An enduring friendship is the exploration of another mind.
The richest fruits are shared thoughts. Does your dog, your
cat, your horse, your parrot or any pet have a true capacity
to be your friend? Does your dog live in the same world as
you? Can you share experiences? Does your cat really know
who you are? Do animals have inner lives? Do they enjoy
their food? Or are they simply biological machines?

This is a question that has occupied some of the greatest
minds of the ages. Today new scientific discoveries are
heating up the debate. With remarkably little agreement.
Scientists can't even agree on whether or not animals are
conscious! Animal intelligence is heatedly debated. Where
some scientists find a strong capacity to reason in many
animals, other find little or none. The exciting discovery
of a capacity for language in animals is widely heralded and
widely ridiculed.

If world class thinkers can't agree about animal minds the
matter is obviously not simple. But it should be an
interesting pursuit, because you cannot explore the mind of
an animal without exploring your own. And some of the most
exciting discoveries of the century will illuminate the
search.

*DARWIN OR DESCARTES?

We can begin our search with the following question: Darwin
or Descartes? Yes or No? Rene Descartes who died in 1650
at the age of 54, invented analytic geometry and is often
considered the founder of modern philosophy. In his famous
"Discourse on Method" published in 1637, Descartes claims
that the most stupid human can use language, while no
animals can do so. From this he concludes: "And this
proves not only that the brutes have less reason than man,
but that they have none at all."

Charles Darwin, on the other hand, had tremendous respect
for animal intelligence. He laid out his theory of
evolution in THE ORIGIN OF SPECIES, published in 1859. He
discussed human evolution in THE DESCENT OF MAN, published
in l871. Many writers had objected to the idea that human
intelligence could have evolved from lower animals, since
the gulf between human and animal intelligence was so huge.

In THE DESCENT OF MAN Darwin answered these objections in
two chapters on the "Comparison of the Mental Powers of Man
and the Lower Animals", the first of which the reader will
find in the appendix. Here Darwin writes that his object is
to show "that there is no fundamental difference between man
and the higher mammals in their mental faculties". He is
speaking, of course, not of the extent or magnitude of these
faculties, but of their nature.

This is an exceptionally strong position. It contrasts with
writers of previous ages like Descartes, who thought of a
dog as little more than a biological machine. It contrasts
with many present day thinkers who tend to be either highly
sceptical or agnostic on the subject of animal minds.

Speaking of man and the higher animals, Darwin writes: "All
have the same senses, intuitions, and sensations,- similar
passions, affections, and emotions, even the more complex
ones, such as jealousy, suspicion, emulation, gratitude, and
magnanimity; they practise deceit and are revengeful; they
are sometimes susceptible to ridicule, and even have a sense
of humour; they feel wonder and curiosity; they possess the
same faculties of imitation, attention, deliberation,
choice, memory, imagination, the association of ideas, and
reason, though in very different degrees". This clearly
implies not only that animals have mental powers which they
apply to the challenges of the world around them, but also
that they have an inner mental life. The higher animals, in
Darwin's mind, are clearly conscious.

Daniel Dennett is perhaps America's current guru of
consciousness. He has written a fact laden, tightly
reasoned book called "Consciousness Explained" (Little
Brown, 1991), which has attracted much attention in the
cognitive science community. Ever since Donald Griffin and
Robert Galambos discovered that bats navigate by
echolocation people have been fascinated by the sensations
of bats. Speaking of how we might investigate the
consciousness of bats, Dennett writes: " ... but they
would show us only what bats are not conscious of, leaving
entirely open what, if anything, bats are conscious of."

Francis Crick is one of the molder's of modern science. He
shared a Nobel prize with James D. Watson for discovering
the molecular structure of DNA. Some people think of this
as the code of life which carries the blue print for babies
and birds and trees and flowers. He is now actively engaged
in brain research, which led to his new book, "The
Astonishing Hypothesis" (Scribners, 1994) . He puts it this
way: "I am less certain that a monkey is conscious than I
am that you are, but I can reasonably assume that a monkey
is not a total automaton - meaning a bit of machinery
showing somewhat complex behavior but completely lacking in
awareness."

The mystery of animal awareness is not an academic issue.
If we are wrong we could be tragically wrong. In his 1973
presidential address to the American Philosophical
Association (Thoughtless Brutes), N. Malcom said, "The
relationship between language and thought must be so close
that it is really senseless to conjecture that people may
NOT have thoughts, and also really senseless to conjecture
that animals MAY have thoughts." We learn from P.
Carruthers (Brute Experience, Journal of Philosophy, 1989)
what this could mean: "Indeed, since all the mental states
of brutes are nonconscious, their injuries are lacking in
even indirect moral concern ... many are now campaigning to
reduce the efficiency of modern farming methods because of
the pain caused to the animals involved. If the arguments
presented here have been sound, such activities are not only
morally unsupportable but morally objectionable."

Cognitive science has now been broadened to included
neuroscience, psychology, philosophy, artificial
intelligence, and cognitive ethology. Donald Griffin, often
called the dean of cognitive ethology, has summed up his
findings in his book "Animal Minds" (University of Chicago
Press, 1992). He shows a kind of Darwinian hospitality,
both to animal consciousness and the capacity of higher
animals to reason. He points out that many scientists
object to the study of animal consciousness because they
believe that animal mentality cannot be objectively
measured. Scientists, they say, should stick to the
observable. But these objections are being overcome and
more and more scientists are willing to talk about the
mental life of animals.

I think I can convince you that Darwin was right. Higher
animals have a rich mental life, and a great deal of
intelligence. Your dog, your cat, or your horse certainly
knows who you are. The emotions which you generate in your
pet are very real.

*WHAT IS INTELLIGENCE?

On the one hand we have the rich new findings of ethology,
the study of animal behavior. We have exciting new findings
in neuroscience and psychology. We have the sophisticated
new thinking of the philosophers. But most importantly we
have seen in the past half century a vast expansion of the
concept of intelligence itself which has been forced upon us
by our vast expenditure of resources in the effort to make
machines intelligent.

Simply put, many things we do which were not thought to
require intelligence, actually require a great deal of
intelligence. We therefore have a lot more intelligence
than we previously thought we did. But the same reasoning
applies to animals.

I became intensely interested in the problem of animal
intelligence after I was called upon to set up an artificial
intelligence laboratory for my company a few years ago.
Most of our work was for the Department of Defense, which is
tremendously interested in smart weapons. Artificial
intelligence is a branch of computer science that seeks to
enable computers to do "intelligent things". Many people
put it this way: we want to make computers do things that
would be considered intelligent if performed by people.
Everyone's favorite example of artificial intelligence is
chess. If a person can beat me at chess I have to consider
him intelligent so if a computer can beat me at chess, the
computer has intelligence. But I soon saw that much of the
intelligence we were so laboriously trying to put into our
machines was far surpassed by the capabilities of animals.
So I turned this AI principle around. If an action
demonstrates intelligence when performed by machines, then
it must demonstrate intelligence when performed by animals.

Note that the possession of intelligence does not
necessarily imply being intelligent. Intelligence is a
capacity which can come in large or small amounts. Being
intelligent implies that this capacity meets a certain
standard - usually a human standard. The operative word in
most of this discussion will be "intelligence" not
"intelligent". That is, I will demonstrate that animals
possess a great deal of intelligence, but I will not try to
measure the amount of intelligence against any human
standard.

With this in mind, let us look at two major goals of
artificial intelligence research. They are both extremely
difficult. Success has been extremely modest. Vast numbers
of animals accomplish both of these goals with ease. The
first is the so called autonomous vehicle. The second is
the recognition of visual scenes.

An autonomous vehicle would be able to navigate over a wide
variety of terrain and move to a chosen destination without
the aid of a human operator, either on board or remote. It
must avoid obstacles, and resume its course, having done so.
It should be obvious that just about every dog, cat, horse,
deer, rabbit, eagle, etc. is an autonomous agent, capable
of moving freely over wide areas of terrain without
assistance. They are always avoiding obstacles. The
ability of dogs to navigate their environment may or may not
make them intelligent, but it proves they possess a
tremendous amount of intelligence. We have put a tremendous
amount of intelligence into machines, at a cost of hundreds
of billions of dollars. Yet computer based autonomy, so
far, doesn't remotely approach that of dogs.

And what about visual recognition? If any company in the
world could spend ten billion dollars and develop a computer
that could recognize human faces like your dog and also
recognize all the other things you dog can recognize, it
would soon become a trillion dollar company. Computer
vision, limited as it is, requires more computer smarts than
just about anything else computers can do. The achievement
of computer vision comparable to the vision of higher
animals would represent the greatest leap of the power of
the machine in all of history. The ability of a dog to
recognize its master may not make that dog intelligent, but
it requires a truly vast amount of intelligence.

In pursuing the animal mind we must distinguish the question
of intelligence from the question of an inner mental life.
The concept of intelligence has been vastly extended in the
modern era to cover not just the capacities of biological
brains, but also machines and even abstract systems, such as
an economic system. Just as I have to admit that a machine
that can beat me at chess is more intelligent than one which
cannot, I also have to admit that an economic system which
responds to market forces in such a way as to facilitate the
prosperity of the public is more "intelligent" than one
which does not.

We can think of intelligence as the ability to process
information in such a way as to achieve certain goals. When
a market economy responds to market forces it is reacting to
information. It is thus processing information. When a
cruise missile flies along a street in Baghdad, turns a
corner, finds a target and blows it up, it is processing a
tremendous amount of information. It may not be
intelligent, but it is far more intelligent than a missile
which has to be guided to its target by a human operator.

When a dog finds its way home, or recognizes its master, or
catches a frisbee in the air, it is processing information -
a tremendous amount, as we shall see. A dog thus has a
great deal of intelligence. Whether it is intelligent
depends on the standard we set for being intelligent.

Information processing does not necessarily require
consciousness. Certainly if intelligence presupposed
consciousness, machines and abstract systems could not
possess intelligence.

We assume our friends and acquaintances have inner mental
lives because they tells us about their thoughts and
feelings using the same words that we apply to our own
thoughts and feelings. Because our friends are like us in
every other way we assume they are like us in the possession
of consciousness. An inductive leap, but an easy one to
make.

Dogs give us no report from within. And yet animals are
enough like us so that most of the human race makes the same
inductive leap. As Darwin put it: "the lower animals, like
man, manifestly feel pleasure and pain, happiness and
misery. Happiness is never better exhibited than by young
animals, such as puppies, kittens, lambs, etc., when playing
together, like our own children." Why then this massive
philosophical and scientific scepticism about animal
consciousness?

I think there is a way to resolve the issue. Let us
continue with our explorations of the animal mind, and then
we will take some of the most salient facts and put them
together in a powerful "thought experiment". I think this
thought experiment will resolve the issue and remove any
doubts about animal consciousness. But that all depends on
you. Along the way we will also pursue some little thought
experiments.

*INTELLIGENCE AND THINKING

We must also distinguish between intelligence and thinking.
Thinking requires intelligence, but intelligence does not
necessarily imply thinking. We will understand thinking to
mean the manipulation of internal images in such a way as to
achieve intelligent problem solving. By images we mean
anything that the memory presents to your awareness, words,
songs, noises, trees, houses, landscapes, faces, shapes,
colors, etc. By problem solving we do not mean solving
problems in logic, mathematics or science only. We mean
also the kind of problem solving you are doing all the time
without being aware of it. When you walk around a table to
get something on the other side, instead of simply smashing
through it, you are problem solving. We know this takes a
lot of intelligence because it is very difficult to teach a
computer based robot to do this. That is, we have to put an
awful lot of smarts in the computer to make it work.

Problem solving requires reason. Descartes believed that
animals "have none at all." Darwin on the other hand gives
us the following magnificent passage: "Of all the faculties
of the human mind, it will, I presume, be admitted that
Reason stands at the summit. Only a few persons now dispute
that animals possess some power of reasoning. Animals may
constantly be seen to pause, deliberate, and resolve. It is
a significant fact, that the more the habits of any
particular animal are studied by a naturalist, the more he
attributes to reason and the less to unlearnt instincts."

Before we proceed let me hasten to state that I in no way
wish to denigrate Descartes. He has made exceptional
contributions to the development of the modern mind. I cite
his ideas because he gives us such a clear statement of his
point of view. Darwin had the advantage of 200 years of
additional scientific development, and, of course, his own
intense study of the plant and animal kingdoms.

*INTELLIGENCE AND INSTINCT

Darwin shows us that we must distinguish between reason and
instinct. Must we also distinguish between intelligence and
instinct. Darwin writes: "The fewness and the comparative
simplicity of the instincts in the higher animals are
remarkable in contrast with those of the lower animals.
Cuvier maintained that instinct and intelligence stand in an
inverse ratio to each other ... But Pouchet, in an
interesting essay, has shewn that no such inverse ratio
really exists. ... In the vertebrate series, the least
intelligent members, namely fishes and amphibians, do not
possess complex instincts; and amongst mammals the animal
most remarkable for its instincts, namely the beaver, is
highly intelligent."

Today we know that instinct cannot be opposed to
intelligence, but actually requires intelligence. We know,
for example, that migrating birds don't sit down and figure
out where they should go for the winter, and how to get
there. However, if we should try to build an autonomous
airplane and have it do what birds do by instinct, we would
find that it would take a tremendous amount of computer
smarts to achieve it. We can think of instinct as handling
a problem which has been solved in the past history of the
species. The instinct represents the implementation of that
solution. And implementation is no pushover. It doesn't
require as much intelligence as reasoning anew each time the
bird should migrate, but a tremendous amount of information
processing must be built into the neural structure of the
possessor of that marvelous instinct.

And of course when we try to build autonomous machines, we
also try to solve all these problems before hand. That is
why some who labor to produce smart machines refer to their
objective as "artificial instinct". But it still requires a
massive amount of effective information processing capacity
and that means intelligence.

*HUMAN INTELLIGENCE

We have said that the exploration of the animal mind can
best proceed through the exploration of your own. The first
thing we want to do is to help you take the measure of the
awesome dimensions of your own intelligence. Then you will
know that even if the intelligence of your dog is only a
small fraction of your own it can still be very
considerable.

First let us look at your memory. Most of our cognitive
capacities rest upon a massive memory store that we are
largely unaware of deriving from our entire perceptual
history. I refer to this as "unattended knowledge." The
following experiment can clarify the basic idea. Fill a one
pound coffee can with pennies, and leave it sitting on your
desk with the cover on. Later casually ask a visitor to
hand it to you. Very likely your visitor will have to make
two attempts to pick up the coffee can. Why? Every time we
pick up anything, open a door, close a window, bite off a
piece of bread, brush our teeth, etc. we make an estimate
of the amount of force we should apply, usually unawares.
Based on what? Common sense?

It is actually based on a portion of the millions of pieces
of knowledge, continually being added to our memory store
(or subtracted from it). The reason we don't gouge our
scalp or drop our comb every time we comb our hair is not
just that we have learned to comb our hair, but we have
stored in our memory the feel of the proper forces for
picking up a comb and pulling it through our hair. Of
course, we can be tentative, and react to the resistance to
our cautious effort, but we would never get through the day
without the more usual quick and confident effort afforded
by that massive memory store.

Do you believe that? Do you really have millions of pieces
of knowledge stored in your memory? Here is a little
thought experiment that will show that you have a tremendous
amount of knowledge that you are unaware of:

In l982 an Air Florida jet took off from Washington National
with ice on the wings and slammed into the 14th Street
Bridge. Soon hundreds of motorists were standing by
helplessly as they watched most of the survivors thresh
around in the icy waters and slowly slip out of sight. Yet
almost every adult raised in the American culture had
information that likely could have saved every one of the
initial survivors in less than five minutes.

After the jet hit the Bridge, in the trunk of almost every
one of the hundreds of cars stopped at the scene was a spare
tire. When this life saving potential is mentioned the
first reaction of the average person is, "They'd sink!"
After all, our spare tire is actually a steel wheel with a
tire.

If someone had thought of his spare, and wondered if it
would sink, he could have dropped it in the water to see - a
small risk when life is at stake. He also could have
performed a little thought experiment.

Imagine yourself bending down lifting your spare onto your
inner wheel. Is it much effort? Now imagine yourself
lifting a firm but light plastic casing of exactly the same
size as your spare, but full of water? Is the effort
greater or less? I suspect you find this gives you a fairly
definitive answer.

Actually a spare tire will float like a balloon. Four or
more roped or belted together can make a raft. All these
hundreds of rafts lay unused while the Air Florida survivors
were drowning!

Suppose we ask some people to write a sample of every kind
of knowledge they possess. How many people would think of
the remembered feel of the weight of a familiar object as a
piece of knowledge? How many thousands of items can we lift
in our minds? Yet how clumsy we would be without this
ability. Imagine what would happen if you applied the same
force to everything you try to pick up!

It is fairly easy to demonstrate that we have millions of
pieces of knowledge, empirical knowledge, that we are rarely
aware of. We are usually unaware of the acquisition of this
knowledge and unaware of its application, which goes on
continually. Millions of pieces of knowledge? All stored
in your memory? Easy to demonstrate? Yes. Most assuredly.
For example, Steven Pinker in his widely acclaimed "The
Language Instinct" (Morrow, 1994) shows that when we count
linguistic items from the standpoint of cognition, rather
than formal rules, the average high school graduate knows
something like 60,000 words. That is a sixteenth of million
pieces right there. But pick up a dictionary and finger
words at random. For most of the words you know, how many
associated pieces of knowledge do you have?

How many items of any description are there in your house of
office. If you know anything at all about them, you have an
item of knowledge that must be managed by your brain. How
many untold thousands of bits and pieces of geographic
information do you have? What does it take to be "familiar"
with a city? How many acquaintances do you know something
about, anything at all? Celebrities? Historical figures?
Fictional characters?

*ANIMAL MEMORY

But what does this have to do with animals? First of all,
if you have millions of items of knowledge stored in your
brain, your dog must have thousands, maybe hundreds of
thousands. If you need thousands of little geographical
tidbits to move around you environment you dog does too.

Scientific observation seems to strongly confirm this line
of reasoning. Darwin puts it thus: "It is almost
superfluous to state that animals have excellent Memories
for persons and places. A baboon at the Cape of Good Hope,
as I have been informed by Sir Andrew Smith, recognized him
with joy after an absence of nine months. I had a dog who
was savage and averse to all strangers, and I purposely
tried his memory after an absence of five years and two
days. I went near the stable where he lived, and shouted to
him in my old manner; he shewed no joy, but instantly
followed me out walking, and obeyed me, exactly as if I had
parted with him only half an hour before. A train of old
associations, dormant during five years, had thus been
instantaneously awakened in his mind."

But we have been talking about thousands and millions of
pieces of information. Darwin here only talks of a couple
of items. Yet where scientists have been able to make
quantitative estimates the discoveries are sometimes
startling. Sara J. Shettleworth, in a remarkable article,
"Memory in Food-hoarding Birds (Scientific American, March
1983), reports on the excellent spatial memory of birds that
hide seed and later recover them. The evidence indicates
that one species can remember the location of thousands of
caches for as long as several months. This is undoubtedly a
highly specialized capability, but it certainly demonstrates
that the structure of the animal brain has the potential for
supporting a huge memory capacity.

*INTELLIGENCE AND VISION

Probably the greatest mental capacity which we share with
higher animals is vision. When we see the easiest part is
done by the eyes. The hard part is done by the brain. A
truly massive amount of information is continually streaming
into our brains through our eyes, requiring a tremendous
amount of information processing capacity for us to make
sense of it. While the vision of higher animals is likely
much more specialized than our own, their capacity to see
has enough generality for them to recognize a very wide
range of objects.

If you have a personal computer you are probably aware of
the fact that the letters and images displayed on your
screen are composed of pixels, that is little dots. If the
dot is black and white it represents one bit of information.
Anything which has two states, on or off, black or white,
one or zero, can be thought of as a binary digit, that is a
bit. What you may not be aware of is that in order to
create the image on your screen some logical processing
capacity must manipulate each pixel every time an image is
displayed. If your screen could display any of 16 colors
per pixel then the pixel would represent four bits of
information. Combinations of pixels give huge numbers of
colors.

Your own visual field can be analyzed into pixels. Your
brain works so incredibly well that you don't realize it but
it is vastly easier to make a machine which can display an
image than one which recognizes an image.

Look at the scene in front of you and think of the
information streaming into your brain through your eyes.
Consider the tiniest speck you can see. That much area of
your visual scene represents at least one bit of
information. In your mind's eye extend that speck across
your visual field to get a line of thousands of bits. Then
extend that line upward and downward in your visual field to
get an area of tens of millions of bits. This is tens of
millions of points of light.

Your visual power is concentrated in the center of your
visual field which weakens as you move outward from the
center to the periphery. Yet the total mass of visual
information streaming into your brain at every glance is
awesome. Your ability to see smooth motion demonstrates
that you process your visual field at least 30 times a
second. Just to see your brain must process something on
the order of a billion bits of information in a second.

When you see, the lens of your eye focuses light reflected
off of trees, people, books etc. on a set of sensors in the
back of your eye. Each sensor is like a tiny camera that
can record a single speck, a single bit. Combinations give
you colors. There are over one hundred million sensors in
each eye. This information is then fed to the brain through
the optic nerve which can transmit about a million signals
simultaneously.

Out of all these millions of one's and zeros the brain must
construct the world you see, and must also pull in from the
untold billions of bits of information stored in your brain
all the information that will give meaning to what you see
and allow you to recognize words, faces, trees, houses,
cars, friends, enemies etc. Of course while this is going
on your brain must also give meaning to the millions of bits
of information that make up the sounds, smells, tastes and
feelings that you are also experiencing.

Every single bit of information going into your brain
requires the transmission of at least one signal, and every
signal requires the operation of at least one gate. A gate
is a device which allows one signal to control the passage
of other signals. Computers have millions of them. You
have billions of them.

*ANIMAL VISION

Now everything we have just said about your vision also
applies to your dog. Your dog may not see all the colors
you do. Maybe its vision is not as rich. Perhaps it does
not have the fine resolution you do. That is the image
which the dog sees may be rougher. The visual panorama of a
dog is certainly not so fully enriched by the memory store.

And yet everything that your brain must do to enable you to
recognize your dog, the brain of your dog must do to enable
it to recognize you. Where your brain processes billions of
bits of information, the brain of your dog must process at
least hundreds of millions of bits. If your dog does not
share the total richness of your visual experience, it does
not live in visual poverty either. A seeing eye dog is also
a world constructing dog.

Darwin recognized this: "The same principle seems to come
into play with vision, as the eye prefers symmetry or
figures with some regular recurrence. Patterns of this kind
are employed by even the lowest savages as ornaments; and
they have been developed through sexual selection for the
adornment of some male animals. Whether we can or not give
any reason for the pleasure thus derived from vision and
hearing, yet man and many of the lower animals are alike
pleased by the same colours, graceful shading and forms, and
the same sounds."

Now that we have a feel for the massive amount of
information processing involved in vision we are in a better
position to appreciate the true magnitude of your memory
capacity. For much of what you remember involves images.
To the British philosopher, David Hume, the recalled image
is a pale copy of the original perception. We have no way
of precisely computing the total amount of information in
the images which you form from remembered experience. But
it is far more than the information needed to record
thousands of words and notes. This recalled image provides
a sense of place and space along with recognition.

I suspect that you can not only recall vast numbers of
spatial images, rich in information, but you also frequently
use your powers of logic to reconstruct these images in such
a way as to take a fairly complex journey in you mind. Make
a short grocery list. Then in your mind walk through your
usual grocery store and pick the items off the shelves. In
how many stores, houses and other buildings can you walk
about in your mind. Along how many streets, roads, highways
and byways can you journey with your mental resources?
However imperfectly the journey may be pursued, the capacity
to do so at all reflects an awesome ability to store,
retrieve and logically manipulate information.

These journeys of the mind are actually part of a larger
capability, the ability not just to remember but to actually
replay many of our experiences, and to rearrange and
manipulate them in our continual effort to anticipate and
plan for the future. A very large part of our life is that
inner mental life which is only made possible by massive
mental resources. As we continually rework the events of
our life, as we think about problems and people and places
of our experience, just past and long past, we are actually
working with billions of bits of information. And we are
doing so with billions of logic gates, continually passing
or stopping millions and millions of signals under the
control of millions and millions of signals. We are sorting
out our lives with the same machinery of intelligence that
Newton and Einstein used to unlock the secrets of the
universe.

And yet much of this capacity we must share with higher
animals. For they must move around their environment much
as we do. Unless one is wholly driven by instinct, or by
someone else's direction, you cannot take a journey with
your body without first taking a journey with your mind. If
we need to form an image of our destination or points along
the way, so must your dog. All the visual capacities we
have been talking about are involved in hunting, and this
must the origin of much of our visual capacities.

This should not be surprising. The vertebrate brain has
been developing for over 500 million years. Our vision
represents a huge portion of our intelligence. It includes
not just the ability to recognize, remember and recall
visual scenes, but to actually construct the world we see.
It would certainly seem improbable that most of this
developed just in the last one percent of the neural
evolution of phylum Chordata.

Darwin strongly believed that animals have the capacity to
form mental images: "The Imagination is one of the highest
prerogatives of man. ... Dreaming gives us the best notion
of this power; ... As dogs, cats, horses, and probably all
the higher animals, even birds have vivid dreams, and this
is shewn by their movements and the sounds uttered, we must
admit that they possess some power of imagination."

*INTELLIGENCE AND REASON

For many people, thinking implies reasoning. Reasoning
requires language. Animals do not have the capacity to use
language. Therefore, they conclude, animals cannot think.
This raises two questions. Can any intelligent creature
think without language? Do animals have any capacity for
language?

Many educators claim that their primary objective is to
teach their students to think. This would seem to imply
that thinking doesn't come naturally to human beings.
Obviously no animal is going to learn to think in school. I
don't think many people learn to think in school, either.
Hopefully their imagination will be stimulated, and they
will learn to apply intellectual discipline to their
thought, that is logic. But people are thinking all the
time, and don't need to go to school to think.

We all have a tremendous capacity for logical thinking, much
of which tends to be embedded in our memory store. Most
people are quite puzzled if told that the following three
assertions are equivalent:

A IMPLIES B EITHER B OR NOT A NOT (A AND NOT B)

Yet they have no problem with the following:

IF YOU WANT TO SUCCEED YOU HAVE TO TRY EITHER YOU TRY OR YOU
DON'T SUCCEED YOU CANNOT SUCCEED WITHOUT TRYING

Of course we make a lot of mistakes in reasoning, but that
reflects mainly two things. One is the complexity of
reality, including the vast amount of new knowledge that
impacts the validity of our assertions. The other is the
simple fact that we usually end the search for truth too
quickly. We are too easily satisfied with the first
thoughts that come to mind on most subjects.

If we think of reasoning as manipulating images to
anticipate the future, then we are always reasoning. Since
a lot of those images are non verbal, we actually do a great
deal of reasoning without language. We usually don't
realize it because most of our reasoning is so easy for us
that we are unaware that we have been doing it. If I take a
short cut across vacant lot, I have perhaps manipulated an
image of a rectangle, and concluded that a path along the
diagonal will save me time. More likely I just remember
than such a path will save me time. Yet however
facilitated, I am using my mental resources to solve a
problem

*ANIMAL REASON

Darwin cautions that we must consider the circumstances
before we decide if an action was "due to instinct, or to
reason, or to the mere association of ideas." He then gives
several examples of animal problem solving, where novelty
seems to preclude instinct.

For example: "The promptings of reason, after very short
experience, are well shewn by the following actions of
American monkeys, which stand low in their order. Rengger,
a most careful observer, states that when he first gave eggs
to his monkeys in Paraguay, they smashed them, and thus lost
much of their contents; afterwards they gently hit one end
against some hard body, and picked off the bits of shell
with their fingers. After cutting themselves only once with
any sharp tool, they would not touch it again, or would
handle it with the greatest caution. Lumps of sugar were
often given them wrapped up in paper; and Rengger sometimes
put a live wasp in the paper, so that in hastily unfolding
it they got stung; after this had once happened, they always
first held the packet to their ears to detect any movement
within."

To some these might seem like such utterly trivial examples
as to not qualify as reason at all. But again our struggles
to impart intelligence to machines tell us that these
examples are very significant. We know that we would have
to put an exceptional amount of logic and knowledge into a
computer to get it to deal successfully with this amount of
novelty.

Donald Griffin in his "Animal Minds" gives numerous examples
of adaptive problem solving. "In a city park in Japan, a
hungry green-backed heron picks up a twig, breaks it into
small pieces, and carries one to the edge of a pond, where
she drops it in the water. At first it drifts away, but she
picks it up and brings it back. She watches the floating
twig intently until small minnows swim up to it, and she
then seizes one by a rapid thrusting jab with her long sharp
bill."

He tells of African chimpanzees who select suitable branches
and then break off twigs to make probes which they poke into
termite nests, providing a termite snack when the twig is
withdrawn. When food is placed at the top of a pole,
captive beavers will modify their lodge and dam building
behavior and pile material around the pole until they can
reach the food. Some beavers in the wild will cut holes in
their own damns in the winter in order to lower the water
level so they can swim around under the ice without holding
their breath. It is hard to see how instinct could
accommodate this much novelty.

Darwin gives an example of animal tool usage that even a man
might brag about. He relates the account of the German
naturalist Brehm (Illustriertes Thierleben, 1864) who once
took part in an attack with firearms on a troop of baboons
in the pass of Mensa in Abyssinia. The baboons responded by
rolling stones down the mountain, some as large as a man's
head. In fact they rolled so many stones that the attackers
had to beat a hasty retreat. Not only that but, for a time,
the pass was actually impassable by the caravan. Darwin
adds, "It deserves notice that these baboons thus acted in
concert."

*ANIMAL SENTIMENT

Darwin gives numerous examples of emotional states that we
usually think of as peculiarly human, such as the desire for
revenge. "Sir Andrew Smith, a zoologist whose scrupulous
accuracy was known to many persons, told me the following
story of which he was himself an eyewitness; at the Cape of
Good Hope an officer had often plagued a certain baboon, and
the animal, seeing him approaching one Sunday for parade,
poured water into a hole and hastily made some thick mud,
which he skilfully dashed over the officer as he passed by,
to the amusement of many bystanders. For long afterwards
the baboon rejoiced and triumphed whenever he saw his
victim". Could we call that emotional problem solving? Or
is it reasoned emotion solving?

Revenge, of course, is often thought of as a moral
obligation, though fortunately not by all. Do animals have
moral sentiments? Darwin wrote, "I FULLY subscribe to the
judgment of those writers who maintain that of all the
differences between man and the lower animals, the moral
sense or conscience is by far the most important." Also, "A
moral being is one who is capable of comparing his past and
future actions or motives, and of approving or disapproving
of them. We have no reason to suppose that any of the lower
animals have this capacity." However he also writes,
"Besides love and sympathy, animals exhibit other qualities
connected with the social instincts, which in us would be
called moral; and I agree with Agassiz that dogs possess
something very like a conscience."

Speaking of these social instincts he writes, "Such animals
are always ready to utter the danger-signal, to defend the
community, and to give aid to their fellows in accordance
with their habits; they feel at all times, without the
stimulus of any special passion or desire, some degree of
love and sympathy for them; they are unhappy if long
separated from them, and always happy to be again in their
company." He gives numerous examples of admirable animal
behavior which follows therefrom.

There were the crows feeding two or three of their blind
companions. There was an old and completely blind pelican,
found on a salt lake in Utah, which was very fat, and "must
have been well fed for a long time by his companions".
There was the little American monkey, who was dreadfully
afraid of the great baboon which attacked his keeper. Yet
he rushed to the rescue, screaming and biting until the
keeper escaped, quite possibly saving the man's life.

Elizabeth Marshall Thomas in her fascinating book, "The
Hidden Life of Dogs" (Houghton Mifflin, 1993) raises and
answers Darwin's question. "Do dogs have morals? Bingo
did." Bingo was a male pug who had developed great affection
for a young Husky name Maria. Yet when Maria lunged at a
parakeet, and went tearing around the cage trying to catch
the frantic birds inside, Bingo put love aside. He slammed
into Maria and then stood guard over the cage. The behavior
of a dozen dogs observed by Ms. Thomas over the years
confirms many of Darwin's observations. On the one hand
there are the strong and sometimes puzzling problem solving
abilities of her dogs. On the other hand there are the
complex social interactions that have to be the product of
complex intelligence. Of special interest is the manner in
which her new dogs learned how to behave from her other
dogs.

*CONCEPTUALIZATION

Both Darwin and Griffin give examples of animal
conceptualization. What do we mean by this? Darwin notes
that many people deny that animals have the power of
abstraction, or of forming general concepts. He then notes,
"But when a dog sees another dog at a distance, it is often
clear that he perceives that it is a dog in the abstract;
for when he gets nearer his whole manner suddenly changes if
the other dog be a friend."

Griffin describes experiments with "concept learning" in
pigeons, where food is obtain by pecking at photographs
projected on small screens. That is reinforcement came not
from pecking at specific pictures but rather at pictures
which belonged to various categories. Positive pictures
signaled the availability of food. Negative ones did not.
The pigeons learned to distinguish photos showing wide
varieties of people from those containing no people. They
also learned to distinguish oak leaves from other leaves,
scenes with and without trees, scenes with and without
bodies of water, etc. Typically they would peck at 70 to 80
percent of the positive pictures and 20 to 30 percent of the
negative ones. Pigeons were also trained to follow "an
abstract relational rule" by pecking at patterns in which
one object was placed inside rather than outside of a closed
figure. No one claims that this in any way compares with
the richness of human conceptualization. But it certainly
demonstrates information processing at a very high level of
complexity.

Otto Koehler and colleagues in Germany (Griffin, Animal
Minds) believe they have demonstrated that birds can develop
the concept of numbers as high as seven. They assembled
various vessels with different numbers of spots on the lid.
Though the spots varied in size, shape and position, a well
trained raven could reliably select the pot with the right
number of spots, when the number could vary from one to
seven.

The following considerations might lead to a certain
scepticism about the examples of conceptualization just
cited. For example one researcher recently received
nationwide attention with the announcement that babies do
arithmetic. Babies look longer at something that surprises
them. By this test babies are surprised when a visual image
is presented to them with a number of items increased or
decreased by some small number.

When I read this I thought of a baby whose Mother might be
turned into a cyclops, with the baby subtracting the eye of
the new face from two eyes of the remembered face, and
deducing that there was reason to be surprised. The neural
machinery of the baby certainly does arithmetic, summations
of synaptic strengths by the billions, as we shall see.
From this results perception to which number imparts quality
(and distinctive numbers of human features a very
distinctive quality).

In other words one might object that the above examples
represent nothing more than the capacity for discrimination
that is inherent in meaningful perception. Speaking of
number, if we saw a cow with six legs we wouldn't have to
count them to know that this was no ordinary cow. Speaking
of relational rules, if we saw a man whose nose was outside
of his face we wouldn't have to undertake a process of
logical analysis to be astonished. The process of
perception would handle it.

To which I answer: precisely! When we look at the
incredible difficulties we have with computer vision we
realize that the amount of logic necessary to achieve human
vision is so great that if the brain could be "rewired" for
specific problems like a super autistic there would be
enough logical capacity to prove all the theorems of
Principia Mathematica in a tenth of a second. The logic of
perception represents a huge chunk of our intelligence and
much of that logic underlies thought in general.

I also believe that a capacity for symbol processing is
inherent in the capacity for meaningful complex perception.
It relates to what I call the integrating function of
memory. Think of all the tree related experience that was
essentially disconnected until my brain integrated the
leaves, trunk, and branches into a single image. The
perception of a tree is meaningful only because the tree has
come to stand for tree related experience beyond what is
immediately perceived. Without the attachment of the
related experience the image of the tree is nothing but a
blob of shapes and colors. So it is with symbols.

Stanley Coren in his insightful and fact filled book, "The
Intelligence of Dogs, Canine Consciousness and Capabilities"
(Free Press, 1994) tells of a dog that could count. Up to
five lures would be thrown into high grass, followed by the
command "Poco, fetch". Each time a lure was retrieved
another fetch command was issued. After all lures were
retrieved Poco refused follow the fetch command but would
simply bark and sit down. I believe that here we have a
capacity for discrimination that is inherent in the capacity
to remember. Poco knew there were no more lures because he
remembered each toss. At least he remembered the last one
because he went for it first. In terms of information
processing Poco was doing more than counting.

My research leads me to believe that perception, memory and
reason are three aspects of a more fundamental process, the
essence of which is association. Association also now
appears to be a part of the primitive bedrock process that
triggers memory formation on the most fundamental level of
brain processing, that is, the synaptic connection. We
speak of our associative memory to distinguish it from more
static type of memory found in computers, but we could also
speak of our associative perception and our associative
reason. Any animal with the capacity for meaningful
perception of complex objects has already made a massive
leap in the direction of reason. Which is perhaps one
reason why half a billion years of finding food and escaping
from predators eventually led us to the stars.

*LANGUAGE

We almost certainly do problem solving that doesn't involve
language and the evidence is very strong that animals do
too. Can animals reason with language? Do they have any
ability to use language?

Darwin was not at all bashful in ascribing a capacity for
language to animals. "The habitual use of articulate
language is, however, peculiar to man; but he uses, in
common with the lower animals, inarticulate cries to express
his meaning, aided by gestures and the movements of the
muscles of the face. ... That which distinguishes man from
the lower animals is not the understanding of articulate
sounds, for, as every one knows, dogs understand many words
and sentences. In this respect they are at the same stage
of development as infants, between the ages of ten and
twelve months, who understand many words and short
sentences, but cannot yet utter a single word. It is not
the mere articulation which is our distinguishing character,
for parrots and other birds possess this power. Nor is it
the mere capacity of connecting definite sounds with
definite ideas; for it is certain that some parrots, which
have been taught to speak, connect unerringly words with
things, and persons with events."

He then goes on to relate the case of an African parrot,
which invariably called certain persons by their names, and
said "good morning" to everyone at breakfast and "good
night" when they left at night. Also "He scolded violently
a strange dog which came into the room through the open
window; and he scolded another parrot (saying "you naughty
polly") which had got out of its cage, and was eating apples
on the kitchen table." The passage of time has not been
unkind to these observations. Irene Pepperberg has been
working with a parrot name Alex. When a toy is picked up
from a crowded tray, and Alex is asked "What toy?" he
answers "Block". Alex can also say what a toy is made of,
what color, and whether it is bigger or smaller than another
object. He can even say how many pieces of green wool or
blue blocks lie amid the clutter (Eugene Linden, "Can
Animals Think?", Cover Story, Time, March 22, 1993. Much of
this article sounds like it was written by Darwin's ghost,
eager to tell us of new discoveries that confirm his
observations made six score years ago. For example Darwin's
claim that animals have a sense of humor is confirmed by the
story of Newton, a golden retriever who would pretend he
couldn't find the frisbee until the owner came to help,
whereupon Newton would grab it and run away with it.)

Stanley Coren (The Intelligence of Dogs) makes the sensible
point that "Descartes seemed to forget: the earliest stages
of language development involve language comprehension
rather than language production." He then lists over sixty
commands to which his dogs respond, such as "give me a paw"
or "open your mouth". The list does not include words
producing untrained responses, such as "bath" which causes
one dog to go and hide and another to await the inevitable.

In addition to the memory of symbols and their association
with meaning, animal response to human speech demonstrates
another form of highly complex intelligent information
processing, namely, the identification of phonemes. We know
how difficult this is. With most computer based speech
recognition systems users have to teach them each word for
each voice. When a parrot learns to parrot a word or your
dog learns to respond to a command, they are not just doing
sound perception but very complex sound abstraction.

Coren has also identified numerous instances of what he
calls "dog-productive language", that is sounds, signs and
behaviors which dogs use to impart meaning. These include
eleven patterns of barking, five kinds of growls, plus other
vocalizations, such as whimpering, baying, howling and
moaning, plus numerous signals and gestures. While many
would debate linguistic significance of these components
they certainly involve the association of symbols with
meaning.

A great deal of excitement has been occasioned by recent
language experiments with chimpanzees. The most famous
involved teaching them American Sign Language. In other
experiments the chimpanzees press buttons with symbols on a
computer console or point to them on a portable tablet.
They have generated much debate and much emotion.
Researchers report that vocabularies of hundreds of words
were achieved, and that novel sentences demonstrated a grasp
of syntax.

Colin Blakemore in his "Mechanics of Mind" (Cambridge
University Press, 1977) describes a conversation which a
reporter from the New York Times had with a chimpanzee named
Lucy. He held up a key and asked, 'What's this?' 'Key',
Lucy replied. Then he picked up a comb; 'What's this?'
'Comb', answered Lucy, as she took it from his hands and
proceeded to comb his hair. She stopped. 'Comb me', she
pleaded. 'O.K.', he said, and combed her. 'Lucy', you want
to go outside?', he suggested. She thought for a while and
replied, 'Outside, no.' Want food; apple.'

Blakemore comments, "She is a member of a small and elite
group of apes who are, unknowingly, excavating the
foundations on which man has built the myth of his
biological uniqueness."

Steven Pinker, who strongly supports Noam Chomsky's theory
of a unique innate human capacity for language (The Language
Instinct) disagrees. He says that with years of intensive
training the average length of the chimp's sentence remains
constant, compared to that of a child, which "shoots off
like a rocket". He gives as typical sentences from a
language trained chimp such examples as, "Nim eat Nim eat."
"Drink eat me Nim." "Tickle me Nim play." "Me banana you
banana me you give." Not that much different from Lucy, but
in context Lucy makes sense.

Darwin wouldn't be disappointed with that. His own examples
of animal language involved little more than the ability to
connect "words with things". Animals don't think our
thoughts. They don't reason about economics and nuclear
physics. They don't draw logical inferences from chains of
sentences. But they can make use of symbols in dealing with
the world. We must assume that animal problem solving is
overwhelmingly nonverbal. But we must also acknowledge that
it is very real.

*THE DIMENSIONS OF BIOLOGICAL INTELLIGENCE

We should not be surprised if higher animals have a useful
capacity to reason about the challenges they face. We
should be surprised if the don't. They have brains that are
very much like ours. And if ours represent the pinnacle of
500 million years of vertebrate evolution, some of theirs
represent the pinnacle of about 490 million years. If we
depict the brains of the vertebrates in a rough evolutionary
order, going from fish through frogs, turtles, birds, and
then mammals like the opossum, rabbit, cat or dog, macaque,
and chimpanzee to human, they start out as a few bumps and
come to look progressively more like ours. When we get to
cats and dogs they definitely look like a brain ought to
look. We see the fissures formed by the folded cerebral
cortex with a hemispheric shape, and the obvious cerebellum
and brain stem below. The brain of a chimpanzee looks
almost like a human brain, but half size.

As Darwin put it, "The lower animals differ from man solely
in his almost infinitely larger power of associating
together the most diversified sounds and ideas; and this
obviously depends on the high development of his mental
powers." In other words it is a difference in degree (albeit
a large one), rather than a difference in kind.

Our basic point is this: the extent of animal intelligence
is only a fraction of the extent of human intelligence. But
when we make an objective measure of human intelligence, in
terms of the amount of information that we process, and the
amount of logical processing that we do, both purposefully
and unawares, the extent is so huge that even that fraction
can be very significant.

Let us now take a look at one awesome measure of the extent
of human intelligence, the size of the neural net. This
might gives us some clues.

You have, by common estimate, about a hundred billion
neurons. A neuron is a nerve cell. Like the logic gate in
the computer it transmits a signal under the control of
other signals. Many of these other signals are transmitted
to a neuron from sensory cells in your eyes, ears, skin,
nose or tongue, but most often they are transmitted from
other neurons. These signals connect with other cells at
junctions called synapses.

There are something like ten thousand of these synapses for
every neuron. So all the neurons in your brain are
massively interconnected in huge neural nets and sub nets.
We have a growing body of evidence that our sensory and
mental experience, in activating these synapses, causes the
strength of the connection at the synapse to change. This
variation in connective strength appears to be the basis of
all learning. The altered synaptic connective strength
constitutes, in effect, the memory store.

But much more than this, the network of variable synaptic
connections gives rise to a very remarkable logic with most
amazing properties. Research on the brain is incredibly
difficult. Brilliant scientists are devoting their lives to
understanding the neuron itself. The precise
interconnections of millions of neurons cannot not be
examined directly. However computer simulations of "neural
nets" are revealing the vast potential of this neural logic.

First of all this logic allows information to be distributed
over vast numbers of neurons, which makes it possible to
recover from error and injury. It allows vast amounts of
information to be distributed over the same set of neurons,
and a process involving the summation of these synaptic
strengths seems to form the basis of our "associative
memory", that is our remarkable ability to pull information
out of our memory by reason of its "association" (that is
similarity) with what is presented to our senses. It is
also the basis of our ability to distill useful information,
over a long period of time, from the chaos of sensory data
that is always streaming into our brain. It basically means
that our reason and memory are one.

This also gives us another element in our inventory of
intelligence - the connection. That is what we call the
passing of a signal from one neuron to another at the
synapse. The connection is also the basic operation in the
new neural net computers, which implement some form or other
of neural logic.

Analyzed in terms of connections our memory logic system is
huge. Taking the most common estimate of the number of
neurons as 100 billion and the numbers of synapses per
neuron as ten thousand we come up with a thousand trillion
synapses. Suppose we were to make a drawing of your brain.
If we made it very compact, and placed one synapse every
quarter of an inch, that drawing would cover over 15
thousand square miles, It would take enough paper to cover
Massachusetts, Connecticut, Rhode Island and Delaware. If
we tried to build such a net with vacuum tubes it would
cover the ocean.

We literally have an "ocean of intelligence". But if we do,
a Chimpanzee has a sea of intelligence and your dog or cat
has a large lake full. They have hundreds of millions if
not billions of neurons, and they work just like ours. In
fact, we know how ours work by studying neurons in animals.
But we don't fully understand how they work because they are
too complicated. Most of the attempts to find the physical
basis of memory, for example, have been done with simple
invertebrates, like snails, because even the more primitive
vertebrates are way too complex.

How does your brain change when you learn to play a piano
piece. Most likely it involves a fairly slight perturbation
in the connective strengths of billions of synapses
connecting millions of neurons. Our models of associative
memory indicate that other memories involve perturbations to
the same synapses. How would we measure it? How does the
brain of your dog change when you teach it how to fetch a
stick. It must involve millions of synapses. Aren't we
lucky that when the simplest animals emerged from primitive
multi-cellular blobs almost a billion years ago they hit
upon a feature that would not only allow them to adapt to
the challenges of their primitive environment, but would
also enable Newton and Einstein to unravel some of the most
important secrets of the universe. Aren't neurons
wonderful?

*CONSCIOUSNESS

The brain of a dog is as much like your brain as the heart
of dog is like your heart. Wouldn't you be utterly
astonished, then, if scientists told you that we don't
really know if a dog's heart pumps blood? Should we not
then be utterly astonished when scientists tell us they
don't know if dogs are conscious? Is there anything we do,
that dogs do also, that we could not do if we were not
conscious, no matter how much unconscious intelligence we
had? We will now embark upon a thought experiment which I
believe will give you a satisfactory answer to this
question.

What will you do when you put down these pages? Most likely
you will go someplace. To the next room, out in the back
yard, into the street, to your favorite chair, to the table,
to bed, some place, anyplace. You will be doing something
marvelous, something that requires a tremendous amount of
intelligence, something you can do only because animals have
been developing that capability for hundreds of millions of
years. First you must see. You must take in hundreds of
millions of bits of raw visual information every second.
You must resolve those bits into shapes and features, that
is lines, etc. You must then combine those shapes and
features into objects, like chairs, rocks, or whatever.
Then you must combine those objects into a scene, so that
the whole panorama of vision, all those hundreds of millions
of bits, becomes instantly meaningful to you. You must
orient yourself. That is you must place the scene which you
see at this instant into its proper place in the larger
world that you do not see. That is when you see the room
with your eyes you must also see the house of which is a
part with the eyes of your mind; and also the yard, the
street, the area, the country and the world that contains
the house.

Now your dog doesn't see the country or the world in his
mind's eye. But you can be quite sure that he has a mind's
eye, that beyond the visible room he sees the house, the
yard, the street, the area with which he is familiar. Why
should we assume this to be so? Because your dog is
perfectly oriented. He moves around the house, the yard,
the area just like you do. He knows where to go, and is
totally unsurprised as the visual scene, all those millions
of bits, changes.

We know that what you do, and what your dog does as you move
around the house and immediate area, takes a tremendous
amount of intelligence. It takes a tremendous amount of
information processing capacity. We know that because it is
so incredibly difficult to get a computer to do anything
like it. Some of the best brains in the world are wrestling
with the problem? We are making progress, but we have along
way to go simply because the problem is so huge. Computers
can now read printed pages, and with difficulty and many
imperfections, they can read handwritten notes. They can be
programmed to recognize specific shapes, and thereby some
machines, airplanes, etc. But a generalized scene
recognition which you find so natural is totally beyond
them. Some day, perhaps, at tremendous expense, with a lot
more progress on every front of computer science - but not
today.

But if we succeed, then won't unconscious machines be doing
the same thing we do? Not quite? No matter how successful
we are in developing computer vision it will never be like
yours. Let us see why.

When something works incredibly well you tend not to be
aware of its existence. You know that it takes intelligence
to solve a problem in calculus because you find it rather
difficult. If you are like most people you are not aware
that it takes intelligence to create the three dimensional
world in which you live because your brain does such a
fabulous job of putting together the billions of "pixels"
streaming from your eyes into your brain every second that
you confuse that accomplishment with the real world.

Now there is something else your brain creates that almost
no one attributes to the brain - something that would
perhaps never occur to you unless you had spent a great deal
of time wrestling both with some of the more neglected
problems of philosophy and with the effort to impart some
kind of intelligence to machines. We refer to something
that brightly illuminates the nature of your consciousness.
That something is your rare, precious and beautiful NOW!
The NOW? you say. Now is an aspect of time. The boundary
between past and future. How can the brain create the now?

Note that we didn't say THE now, we said YOUR now.
Philosophers like William James have called our attention to
what they call the "specious present". How long, actually,
is the present? A second can be divided into a thousand
milliseconds, each of which can be further divided into a
thousand microseconds. Each microsecond contains one
thousand nanoseconds, each of which contains one thousand
picoseconds, of which there are a trillion in every second.
Etc. No matter how small a unit we pick, at any instant of
time, some portions of that unit must be either in the past
of the future or both. We now have computers that can
perform a significant number of operations in a billionth of
a second. In a fraction of billionth of a second each
operation slips from the future into the past.

Yet all of this is no problem for our minds. As we move
continuously though time, and the various events of our
lives slip from the future into the past, our brain keeps
hauling into our awareness just enough of our brand new
past, and just enough of an anticipated future to always
present us with an extended meaningful NOW.

It doesn't have to be this way. What if our brain didn't do
this? What would our experience be like? Perhaps it might
be like a computer, where there is no concern with the
specious present because there is no meaningful NOW. When a
computer needs to make a decision, it performs a compare
operation and turns on or off a couple of bits in a compare
designator to record the result. In a later slice of time
it tests the compare designator. Those few bits are its
only link with the past. The fact that it might perform
several million of these operations every second might lead
an observer to conclude that a very well programmed computer
is pretty smart. But in each slice of time there is no past
or future, just a handful of 1's and 0's.

But in creating your beautiful meaningful now, your brain
not only ties together the remembered past and anticipated
future, but it also unites what is separated in space. We
have shown that when you see you are instantly taking in
millions of bits of information, millions of visual pixels.
You are not at all astonished that you behold all of these
millions of points of light together, at the same time - but
you should be!

Look at the face of an acquaintance, either in front of you
or in your mind's eye. Now imagine that your are a
computer, looking at the same face, very successfully let us
assume. We have leaped into the future and given computers
the capacity to recognize human faces. The point is that in
no slice of time will the computer ever see that face.
Rather, in one slice of time it will look at a few bits from
the nose. Ten million time slices later it will be looking
at a few bits from the ear, etc. It will have no meaningful
view because it will never have a meaningful past. It may
experience fifty million time slices a second, but each one
is locked out from all the others. The final result of
recognizing the face will not be the simultaneous perception
of all the millions of visual pixels that you enjoy. Rather
the result will simply be a few bits stored in the computer
to indicate that this is face number xyz23456. That code
will allow it to do what it is supposed to do. But it will
never in any one time slice provide a meaningful view of
that face. Aren't you glad you are you!

Neural net computers will change this somewhat by allowing
the computer to process millions of bits simultaneously, but
still each slice of time will be living in isolation. There
will be no continuous meaningful NOW!

No one knows how your brain creates this meaningful now.
But we can say this! It would be impossible to imagine your
capacity to see and hear without your extended now. It
would be impossible to imagine your extended now without
consciousness.

Now your dog sees and hears. Your cat sees and hears. Not
as richly as you do. But this most challenging of all your
mental faculties is just what the higher animals seem most
fully to share with people. Does your dog or cat see like
you do, or like a computer? Thinking of all the sights and
sounds that enrich your life and come to you in your
extended now, and thinking of the animals that live in your
world, you probably don't find it very difficult to answer
that question.

In ascribing a mental life to animals we are making an
inductive leap. Thus we are dealing with probabilities. I
find it highly probable that the higher animals have an
inner mental life that is qualitatively much like our own in
our less reflective moments. The visual world of dogs is
likely grayer, coarser, less charged with significance, yet
much like ours.

Everything we see, hear, smell, taste or feel is enhanced by
our associated memories, which are ready to pull us off in
any direction on a mental journey composed of recalled
images rearranged in every fanciful way, by those powers of
association that we call imagination. I suspect the same
thing is true of animals. Darwin was correct in ascribing
imagination to animals, but of course the images must be
less vivid and the train of memory less extensive. I think
our ability to reason derives not from a power which we
alone possess, but rather from all of those advantages that
allow us to make the chain of association more persistent
and better controlled - our longer path to maturity, our
greater memory store and related powers of recall, our
powers of communication and record keeping which enable us
to profit from the experiences of so many others removed in
place and time. One reason why I cannot imagine what it is
like to be a dog is because I cannot imagine what it is like
to look at a tree without placing it in a larger universe.

The study of how things might appear to other minds has been
given the name "heterophenomenology", that is, the study of
phenomena from the third person perspective (hetero) as
opposed to the first person perspective (auto). It is thus
the study of the "innenwelt", that is the inner world, of
others. Daisie Radner provides some intriguing perspectives
in her paper, "Heterophenomenology: Learning About the
Birds and the Bees" (Journal of Philosophy, 8 August 1994).
She writes "The question 'What is the animal's experience
like?' has to be answered in the light of the animal's own
physiology and behavior. The key to making progress via
INNENWELT heterophenomenology is to recognize that animal
experiences are not just pale imitations of our own."

Since that seems to be what I am saying let qualify my
position. As regards behavior we have already noted the
extremely versatile navigation abilities of many land
animals seems to bespeak a 3d visual world, so let's look at
physiology. Radner reminds us that we have to consider such
things as color vision. She points out that the visible
spectrum of bees is 300 to 700 nanometers compared to 400 to
800 for humans. Filtered ultraviolet photographs of various
flowers together with training experiments demonstrate that
bees can make discriminations that we cannot. "'How do
flowers look to a bee?' proved to be a very fruitful
question, leading to increased knowledge, no only of honey
bee vision, but also of the co-evolution of plants and their
pollinators."

Of course, no matter how much we learn about the machinery
of perception, imagining the experience of other minds will
always entail an inductive leap. Whenever a signal passes
through a gate it loses its identity and anything can happen
to it. If your red looks like my blue we will never know it
because we will always point to the same red things and the
same blue things. Visual signals could go to the auditory
cortex, and auditory signals could go the visual cortex, and
apparently in rare cases they do. It could be theoretically
possible to intercept the sensory signals of some sentient
creature somewhere between the receptors and the brain and
feed them into a computer, and with massive advances in
computer power and computer smarts, achieve the same
functionality as the brain with no "innenwelt" at all.

There is, however, one "other mind" accessible to you by
logical analysis, and whose perceptual experiences differ
from yours just enough to allow you to extrapolate to other
"other minds". I refer to your earlier self. "Accessible
by logical analysis?" you say. Yes! Most certainly!

I base this claim on my own personal discovery. As a
student of philosophy in college I began to study of written
Chinese in order to see how a "radically different" language
would impact my thought. This rather casual "thought
experiment" accidentally changed the course of my life and
stayed with me for 45 years. One discovery emerged from
this experiment with ever greater force over the years: we
see with our memory. That is, our experience not only
determines how we evaluate what we perceive, but how our
perceptions will be constructed, how we select from the
millions of pixels streaming into our brain, and how we put
them together. For example, about twenty years ago I
discovered I needed glasses to read Chinese. My brain
apparently could put English together with fewer perceptual
clues. None of us can remember what our native language
sounded like before we learned to speak. We could get an
idea listening to Swahili. Nor can we remember what it
looked like before we learned to read. Nor can we remember
what trees, houses, people, dogs, and cars looked like
before our experience invested them with ever greater
significance and provided the brain with an ever greater
memory store to assist in their construction. But we can
get an idea.

However, we do not have to jump over the decades to see how
experience remolds our perceptions. You may remember the
first time you heard some familiar song. You certainly
cannot remember the way it sounded. It was "half
dimensional". It stood on only one side of the time line in
your extended now. Experiment. Get a tape of some song you
have never heard before. It should be the old fashioned
kind with a real melody. Or maybe there is a Mozart
concerto you are not familiar with. Play it a dozen times
or so. Notice how your memory deepens and enriches your
perception and expands your now as anticipation is added to
recall! Or take a poem. Again the old fashioned kind,
where the second rhymed word will jerk the first one back
into your extended now.

I find it reasonable to assume perceptually well endowed
animals possess the bits and pieces of perception (some
philosophers call them qualia) which are rather like my own,
though certain colors, shades, intensities, tones, etc.
will differ. With these bits and pieces, and with the aid
of experience, and also prewired equivalents (perceptual
instincts), will be constructed experiences which will
differ from mine rather like my earlier experiences differ
from my later ones, though in vastly greater degree.
Language comes into this mix simply as additional
experience. So that is exactly what I mean and do not mean
by inducing that some animal experience is like my own, not
a pale imitation, but a pale analogue.

*WHY SO DEFENSIVE?

Why is it that Darwin was so totally unembarrassed in
talking about animal minds, while so many modern day writers
are so cautious and defensive. Partly the reasons are
historical. Early in this century, psychologists, reacting
to the inconclusive results by obtained by "introspection",
adopted a viewpoint known as behaviorism. Deciding that
mental states could not be known with certainty or measured
objectively, psychology became overwhelming concerned with
what could be measured objectively, namely behavior.

Many writers not only claimed that subjective mental
experiences, especially conscious thinking, should be
ignored because they are private phenomena which cannot be
independently verified. They went so far as to claim that
mental states have no influence on behavior. They are
simply epiphenomena, that is, the incidental by products of
brain function.

Today strict behaviorism has been pretty much abandoned.
Researchers in animal behavior now recognize the importance
of what they call "cognitive states", such as learning,
remembering, problem solving, perception, and recognition."
However, there still remains a widespread view that it is
unscientific to talk about subjective experience or
consciousness.

One of the strangest arguments in this debate lies in the
claim that anyone ascribing mental states to animals is
guilty of "anthropomorphism", that is ascribing strictly
human qualities to non humans. It is not necessarily any
more anthropomorphic to assign thoughts to animals than it
is to assign digestion to animals, unless, of course, the
"Cartesians" have already won the debate. As an argument,
the charge of anthropomorphism is totally circular. But
perhaps it is simply an issue of methodology. Dr. John E.
R. Staddon, a behavioral neuroscientist at Duke University,
points out that a fundamental principle of science is that
one should always seek the simplest explanation for any
observation. Assuming that an animal is aware of its
behavior violates this principle because most behavior does
not require such awareness. (Natalie Angier, "Flouting
Tradition, Scientists Embrace an Ancient Taboo" The New York
Times, August 9, 1994)

There's the rub! We don't have explanations, simple or
otherwise, for most animal behavior, except in the
sketchiest form. We can explain the relatively simple
behavior of computers, only because the hardware and
software designers have produced miles of documentation. We
have pointed out that the complexities of human vision must
be shared to a large degree by higher animals. To
understand vertebrate vision, as we understand a computer
program, we would have to detect and record billions of
discrete signals, at time t, and time t+1, etc. From this
mass of information we would have to go back and figure out
the logic. It is almost impossible to uncover a computer
design from logic and wiring diagrams. We have shown that a
wiring diagram of a moderately complex vertebrate brain
could cover hundreds or even thousands of square miles. If
we made every conceivable breakthrough in reading out memory
storage and signal flow in the brain, the task of
discovering the discrete logic of vision might even turn out
to be fundamentally impossible in a finite universe. What
we can hope for is to gain enough clues to build computer
models that will reproduce the essential functionality of
biological vision.

Also we have pointed out that removing conscious awareness
from our process of visual discrimination would introduce
all the complexities of understanding a visual continuum
piece by piece in isolated time slices. But even if we are
successful in explaining animal behavior without conscious
awareness, we have introduced a new complication. For now
we have two sets of brains, human and animal, very similar
in construction but vastly different in their ultimate
functionality. Here is a kind of neuro dualism that makes
Descartes a little more consistent than some of his
accidental twentieth century soul mates. Mickey Mouse is
anthropomorphic. THE DESCENT OF MAN is not.

There is also what we might call the "Clever Hans" syndrome.
Scientists are afraid of getting stung. Hans was a trained
horse which appeared capable of doing arithmetic. For
Example, when shown a written display, 3 X 9, Hans would tap
his foot twenty-seven times. He could even divide. Careful
observation and analysis finally revealed, however, that
Hans was not watching the displays but the tester. When
Hans got to the correct answer people were anxious to see if
he would stop. They would make small unintentional motions
which they were not aware of, but Hans had learned to detect
them. If Hans couldn't see the tester he tapped at random.

Yet did this test really reveal the limitations of animal
intelligence? You can buy a calculator that can do what
Hans was supposed to do for about five dollars. What would
it take to build a machine capable of doing what Hans
actually did? Five billion? We wouldn't come close.
Imagine what we could do with a machine that could not only
recognize human beings, but also visually detect
psychological states from extremely subtle body clues.

*THE CHALLENGE OF ANIMAL MINDS

I don't think we will ever teach animals to do mathematics,
that is prove theorems, derive formulas, etc. I do think it
would be possible to teach them to do simple arithmetic with
a limited understanding of what they are doing. That is
"fetch 5 + 7 sticks", where the task is not to fetch 5
sticks and then fetch 7 but rather to fetch 12 sticks." I
make this prediction first by combining an analysis of the
amount of information processing involved in arithmetic with
an extrapolation from available facts about animal
capabilities. I make it secondly by noting that we haven't
come close to proving that they can't be taught arithmetic.
Let me explain.

What is usually left out of discussions of learning, human
or animal, is what I call the culture of learning. We have
made the point that people have millions of pieces of
knowledge that they are largely unaware of. Where did we
get that knowledge? I believe it is obtained from a
learning environment, that is culture.

A culture, or subculture, is an environment in which the
individual is continuously exposed to the elements of that
culture. It is thus an environment of continuous, largely
unattended learning. The first product of culture is
ordinary language, which of course, is our greatest
continuous largely unattended learning experience. But we
are also we are learning about chairs and trees, grass and
tin cans, plates and sidewalks, etc. We slowly learn to
operate in our environment through thousands of hours of
meaningful exposure to things that are a part of that
culture. The expectations we place on our schools shows how
we tend to discount the thousands of hours of learning that
precede any schooling.

In support of his view that the human brain is prewired for
language, Steven Pinker writes that the average six year old
commands about 13,000 words. He adds that preliterate
children "must be lexical vacuum cleaners, inhaling a new
word every two waking hours, day in, day out." Yet children
are exposed to language for many hours a day. People
normally speak at a rate of 250 words a minute, which if
sustained, would be 15,000 words an hour. Even at a
fraction of this rate a child would hear thousands of words
for every one learned.

It is this continuous unattended learning provided by the
child's culture that is so difficult to provide an animal.
Darwin speaks of human cultures where the capacity to count
above four doesn't exist. In our culture children rather
easily learn to count, but it often takes years of
intermittent exposure to numbers before they grasp the
concept that the number they stop on is the number of
objects. Consider then the extraordinary effort that it
would take to provide a dog, a horse or a chimpanzee the
kind of numerical culture that would make learning
arithmetic possible. Consider the endless frustration
during all those hundreds or thousands of hours when it
would appear that the animal was learning nothing. The
question may well not be "Can we teach animals arithmetic?"
but "Is it worth it?" And yet when we consider the endless
mysteries forged in the struggle for survival over so many
eons, which now lie hidden in the endlessly complex animal
brain, we might ask ourselves "Why not?"

I specifically propose the further exploration of one major
area of animal intelligence. I have indicated my belief
that a great deal of our intelligence derives from our
capacity for meaningful perception. If this is the case
then one way to probe the magnitude of animal intelligence
would be to probe the limits of their ability see. How much
of our own visual continuum can animals discriminate? How
well can they remember such visual discriminations? To what
extent can they learn to associate what they see with
something else? What is the full range of conceptualization
that inheres in animal perception? Stanley Coren uses his
command "find glove" as an obedience exercise. If we
generalize this finding exercise, and let the animal find
things in plain view we could probe the limits of visual
discrimination, sound discrimination, thing memory and word
memory and powers of association in one extended experiment.
If continued over the years with much repetition (as we have
absorbed our millions of knowledge items) I suspect the
number of "findees" would grow continuously.

The essence of higher intelligence is generality. We might
be very much impressed with a certain feat of perception,
but if it turned out to have the specificity of a smoke
detector, we would know the intelligence involved was
minimal. On the other hand a feat of visual discrimination
on the part of a dog might be extremely mundane. But if the
dog has the ability to make thousands of such
discriminations then substantial intelligence must be
involved.

The whole of human history has been massively impacted by
the fact that intelligence is largely invisible. Imagine
for a moment that you could not see or feel your legs.
Imagine also that you could run at seventy miles and hour.
You would just think about going somewhere and off you would
go, effortlessly, powerfully, speedily. Think of all the
strange explanations we would make about this strange power.

Yet for thousands of years this has been exactly the case
with our marvelous power to think and feel. Little or
nothing was known of its origin in the body. When the
Egyptians prepared their mummies, that useless slush in the
skull was flushed away. "In my heart" was once take
literally. The Chinese word for heart can also mean mind.
The ancients "knew" the brain could not be the seat of the
mind because it could not feel pain. Almost every culture
has tended to find members of other cultures slow to learn
simply because of the massive amount of invisible learning
that prepares us for formal instruction in the highest
treasures of the culture.

The invisible world of animal thoughts cannot be seen with
our eyes. But we can explore it with our minds. And we
need not feel embarrassed if we find it to be both very rich
and very real.

____This the end of "Of Course Your Dog Can Think"____


AN INVITATION TO EXPERIMENT

In my exploration of animal intelligence, OF COURSE YOUR DOG
CAN THINK, I have attempted to justify my belief that the
capacity for meaningful perception is central to the
intelligence of humans and all higher animals. The eyes do
the "easy" part. The difficult part is a done by the brain.

The brain must first of all make selections from the
billions of bits of information streaming into the brain
from millions of sensors in the eyes. It must extract all
kinds of individual shapes and features, such as lines,
circles and rectangles. It must integrate these features
into several levels of more complex entities, such as skin,
nose, face, human. etc. It must recognize all these
entities by associating the visual panorama with a vast
memory store. It must place all this information in a
larger world, which the brain itself constructs, by
combining current perceptual information with huge amounts
of remembered information, in accordance with whatever
structural rules are already built into the brain.

Inherent in all of this is an awesome capacity for
conceptualization. That is, in recognizing complex objects
and placing them in a larger world the brain must not only
recognize entities, but categories, as Darwin pointed out.
It must also deal with number. It must perform positional
analysis. It must recreate many of objects from partial
information. It must continually fill in all kinds of
missing pieces. It must sense motion and direction from
continually changing visual scenes.

Almost all the logical capacities that we associate with
abstract, disciplined, analytical thinking seem to be
present in some degree in the numerous complex processes
that provide us with our wonderfully mysterious capacity to
see.

Now the higher animals seem to share our capacity for
logical analysis only in the most rudimentary degree. But
they seem to have an extraordinary capacity to see. We
cannot find out what animals are thinking by talking to them
as we do with people. But perhaps we can probe the
boundaries of animal intelligence by probing the limits of
their ability to see. We can do this first an foremost by
attempting to determine how much they can see. How many
things can they recognize? How fine is their ability to
discriminate? The more things an animal can learn to
recognize, the greater is the generality of their visual
capacity, and hence the greater must be the inherent logical
capacities of the animal brain.

Every pet owner is already engaged in this great experiment
to some degree. If you ask your dog to fetch something you
are asking your dog to demonstrate a capacity to recognize
something. Dogs seem especially well suited to the kind of
experiments I am proposing because they have a high capacity
to respond of verbal commands.

We are interested in the magnitude of an animals visual
capacity. We are also interested an animals capacity to
conceptualize as demonstrated by an ability to discriminate
and also an ability to generalize. As Darwin pointed out,
dogs have the capacity to recognize individual dogs, but
also dogs in general. They certainly seem to have the same
capacity to recognize humans. But how far does this
capacity extend? How much information does a dog need to
recognize an individual human? Can a dog recognize people
from photographs? Can a dog recognize categories of people?
Men, women, some races, well dressed, etc. How many
different artifacts around the home could a dog learn to
discriminate? How many categories of such articles could it
learn to recognize.

Donald Griffin, in his book "Animal Minds" describes
experiments with "concept learning" in pigeons, where food
is obtain by pecking at photographs projected on small
screens. That is reinforcement came not from pecking at
specific pictures but rather at pictures which belonged to
various categories. Positive pictures signaled the
availability of food. Negative ones did not. The pigeons
learned to distinguish photos showing wide varieties of
people from those containing no people. They also learned
to distinguish oak leaves from other leaves, scenes with and
without trees, scenes with and without bodies of water, etc.
Typically they would peck at 70 to 80 percent of the
positive pictures and 20 to 30 percent of the negative ones.
Pigeons were also trained to follow "an abstract relational
rule" by pecking at patterns in which one object was placed
inside rather than outside of a closed figure.

With a dog the reward would be human approval and
companionship. Verbal commands would initiate the trial.
The process would involve learning words as well as things.
Could a dog learn relation rules? Actually relational rules
are inherent in the discrimination of any complex object or
scene (nose between the eyes and the mouth, for example).
Dogs have much bigger brains than pigeons.

I am not proposing any laborious research project. What I
am proposing is a rather casual experiment carried on over
an extensive period of time. I am proposing that the human
animal companionship be molded into a kind of culture of
learning, deriving from the fact that probing the limits of
an animal's ability to generalize and discriminate could be
a rather interesting way to play with a pet. Something
significant could be learned in weeks, more could be learned
in months. What would really be exciting would be to
discover if a dog's capacity to add to its knowledge store
can continue over the years. I think it could.

Stanley Coren uses his command "find glove" as an obedience
exercise. If we generalize this finding exercise, and let
the animal find things in plain view we could probe the
limits of visual discrimination, sound discrimination, thing
memory and word memory and powers of association in one
extended experiment. If continued over the years with much
repetition (as we have absorbed our millions of knowledge
items) I suspect the number of "findees" would grow
continuously, frequently into the hundreds and in some cases
into the thousands.

I base this in part on research which indicates that animals
have the capacity to remember the location of thousands of
specific items. I base it also on the simple fact that
hunting, such an important part of the ancestral past of
most pets, requires a powerful ability to see, and also to
record geographical information. I base it most of all on
my own discovery that so much of our own cognitive capacity
derives from the hidden learning of literally millions of
items, absorbed from our cultural and physical environment
over many years. The thousands of hours we have spent
absorbing our native language is the prime example. Thus no
test of an animal's ability to learn which does not continue
over the years can remotely compare with enduring learning
experiences that have provided people with most of their
cognitive abilities. Thus no test has shown that animals
cannot learn to remember and distinguish hundreds or
thousands of items.

I would be happy to learn of the results of any experiments
along these lines. I am interested first of all in adding
the results to my own fund of knowledge. I would endeavor
to make any significant discoveries known to the larger
scientific community. I would also make an effort to
summarize the results and share them with all
correspondents. Initial discoveries may indicate further
fruitful directions of research. No doubt most pet owners
have already made many discoveries along these lines, but
not aware of the fact that complex vision requires vast
cognitive resources, they many not have accorded their
casual discoveries the significance which they may deserve.

Harold Wefald
ALWAYS LEARNING
PO Box 2267
Gaithersburg, MD 20886-2267