62332

Download the full version of the ebook at ebookfinal.
com
Mind and Mechanism 1st Edition Drew V. Mcdermott
https://ebookfinal.com/download/mind-and-mechanism-1st-
edition-drew-v-mcdermott/
OR CLICK BUTTON
DOWNLOAD EBOOK
Download more ebook instantly today at https://ebookfinal.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...
Intelligence Beyond Thought Exploding the Mechanism of

Mind Dada Gavand
https://ebookfinal.com/download/intelligence-beyond-thought-exploding-
the-mechanism-of-mind-dada-gavand/
ebookfinal.com
Washington Merry Go Round The Drew Pearson Diaries 1960

1969 1st Edition Drew Pearson
https://ebookfinal.com/download/washington-merry-go-round-the-drew-
pearson-diaries-1960-1969-1st-edition-drew-pearson/
ebookfinal.com
Endocrine Secrets 5th Edition Michael T. Mcdermott Md
https://ebookfinal.com/download/endocrine-secrets-5th-edition-michael-
t-mcdermott-md/
ebookfinal.com
Motion Simulation and Mechanism Design with COSMOSMotion

2007 Kuang
https://ebookfinal.com/download/motion-simulation-and-mechanism-
design-with-cosmosmotion-2007-kuang/
ebookfinal.com
Naturalistic Realism and the Antirealist Challenge 1st
Edition Drew Khlentzos
https://ebookfinal.com/download/naturalistic-realism-and-the-
antirealist-challenge-1st-edition-drew-khlentzos/
ebookfinal.com
Arsenic Contamination of Groundwater Mechanism Analysis

and Remediation 1st Edition Satinder Ahuja
https://ebookfinal.com/download/arsenic-contamination-of-groundwater-
mechanism-analysis-and-remediation-1st-edition-satinder-ahuja/
ebookfinal.com
Molecular Hematology 5th Edition Drew Provan
https://ebookfinal.com/download/molecular-hematology-5th-edition-drew-
provan/
ebookfinal.com
Hollywood Film 1963 1976 1st Edition Drew Casper
https://ebookfinal.com/download/hollywood-film-1963-1976-1st-edition-
drew-casper/
ebookfinal.com
Atherosclerosis Experimental Methods and Protocols 1st

Edition Angela F. Drew (Auth.)
https://ebookfinal.com/download/atherosclerosis-experimental-methods-
and-protocols-1st-edition-angela-f-drew-auth/
ebookfinal.com
Drew V. McDermott
Mind and Mechanism

Mind and Mechanism
Mind and Mechanism
i
This Page Intentionally Left Blank
ii
Mind and Mechanism
Drew McDermott
A Bradford Book
The MIT Press
Cambridge, Massachusetts
London, England
iii

c 2001 Massachusetts Institute of Technology
All rights reserved. No part of this book may be reproduced in any form by any
electronic or mechanical means (including photocopying, recording, or informa-
tion storage and retrieval) without permission in writing from the publisher.
This book was set in Sabon by Interactive Composition Corporation.
Printed on recycled paper and bound in the United States of America.
Library of Congress Cataloging-in-Publication Data
McDermott, Drew V.
Mind and mechanism / Drew McDermott.
p. cm.
“A Bradford book.”
Includes bibliographical references and index.
ISBN 0-262-13392-X (hc. : alk. paper)
1. Mind and body. 2. Artificial intelligence. 3. Computational
neuroscience. I. Title.
BF161 .M29 2001

153—dc21 00-066986
iv
For Judy
v
vi
There are still harmless self-observers who believe that there are “immediate cer-
tainties”: for example, “I think,” or as the superstition of Schopenhauer put it,
“I will”; as though knowledge here got hold of its object purely and nakedly as
“the thing in itself,” without any falsification on the part either the subject or the
object. . . . The philosopher must say to himself: When I analyze the process that
is expressed in the sentence, “I think,” I find a whole series of daring assertions
that would be difficult, perhaps impossible, to prove; for example, that it is I who
think, that there must necessarily be something that thinks, that thinking is an
activity and operation on the part of a being who is thought as a cause, that there
is an “ego,” and, finally, that it is already determined what is to be designated
by thinking—that I know what thinking is. . . . In short, the assertion “I think”
assumes that I compare my state at the present moment with other states of myself
which I know, in order to determine what it is; on account of this retrospective
connection with further “knowledge,” it has, at any rate, no immediate certainty
for me. In place of the “immediate certainty” in which people may believe in the
case at hand, the philosopher thus finds a series of metaphysical questions pre-
sented to him, truly searching questions of the intellect, to wit: “From where do
I get the concept of thinking? Why do I believe in cause and effect? What gives
me the right to speak of an ego, and even of an ego as cause, and finally of an ego
as the cause of thought?” Whoever ventures to answer these questions at once by
an appeal to a sort of intuitive perception, like the person who says, “I think, and
know that this at least, is true, actual, and certain”—will encounter a smile and
two question marks from a philosopher nowadays. “Sir,” the philosopher will
perhaps give him to understand, “it is improbable that you are not mistaken; but
why insist on the truth?”
—Nietzsche (1886), pp. 213–214
vii
viii
Contents
Preface xi
Acknowledgments xv
1 The Problem of Phenomenal Consciousness 1

2 Artificial Intelligence 29
3 A Computational Theory of Consciousness 93
4 Objections and Replies 137
5 Symbols and Semantics 167
6 Consequences 215
Notes 243
References 249
Index 259
ix
x
Preface
There are many reasons to work in the field of artificial intelligence (AI).
My reason is a desire to solve the “mind-body” problem, to understand
how it is that a purely physical entity, the brain, can have experiences.
In spite of this long-range goal, my research has been concerned with
seemingly much tinier questions, such as, how might a robot know where
it is? How would a computer represent the sort of routine but arbitrary
fact or belief that people seem to keep track of effortlessly? (I’m thinking
of “facts” such as “If you go swimming too soon after eating, you might
get a cramp.”) It may seem misguided to pursue tactical objectives that
are so remote from the strategic objective, but scientists have learned that
in the end finding precise answers to precise questions is a more reliable
means to answering the big questions than simply speculating about what
the big answers might be. Indeed, scientists who venture to attempt such
speculation are often looked at askance, as if they had run out of useful
things to do.
Hence, by writing a book on the mind-body problem from a compu-
tational perspective, I am risking raised eyebrows from my colleagues. I
take that risk because I think the mind-body problem is important, not
just technically, but culturally. There is a large and growing literature on
novel approaches to the problem. Much of it is quite insightful, and some
is totally wrong (in my opinion, of course). Even authors I agree with
often fail to understand the role of computational ideas in explaining the
mind. Claims like these are often made with only the flimsiest arguments:
• An intelligent computer program would treat every reasoning problem
as a deduction.
xi
xii Preface
• There are two computational paradigms to choose from: symbolic com-

puting and neural networks; they are quite different, and have fundamen-
tally different properties.
• People think serially at a conscious level, but are “massively parallel” in-
side; so it’s appropriate to model them with a program only when studying
conscious problem solving.
•Whether something is a computer depends entirely on whether a person
uses it as a computer.
• When a computer program manipulates symbols, the symbols must have
a formal semantics, or the program will do nothing interesting.
•Whether the symbols in a computer mean anything depends entirely on
whether people treat them as meaning something.
•A computer could be made to behave exactly like a person, but without
experiencing anything.
I will show that all these claims are false, meaningless, or at least ques-
tionable.
If these misconceptions mean nothing to you, good; by reading this
book first, you will avoid them. Unfortunately, if you’ve read one or two
books on the subject of computation and the mind, you have probably
absorbed some of the nontruisms on the list without even noticing it.
I often assume that the mind-body problem is interesting to everyone,
but I have discovered, by watching eyes glaze over, that it isn’t. One reason
is that it is surprisingly difficult to convey to people exactly what the
problem is. Each of us has little trouble separating mental events from
physical ones, and so we gravitate to a theory that there are two realms,
the mental and the physical, that are connected somehow. As I explain
in chapter 1, this kind of theory, called dualism, though it seems at first
obviously true, runs into enough difficulties to move it to the “obviously
false” column. This should get anyone interested, because many of us have
religious beliefs—important religious beliefs—that presuppose dualism.
Hence we have a stake in what theory ultimately replaces it.
Some parts of the book are a bit demanding technically. There is a little
mathematics in chapter 2, a survey of the state of the art in artificial intel-
ligence. Chapter 5 addresses the knotty technical issues surrounding the
notion of symbols and semantics. I was tempted to leave all these hard
Preface xiii
bits out, to keep from driving away a large class of intelligent readers who
suffer from “mathematics anxiety” or “philosophy narcolepsy.” I decided
to leave chapter 2 in to counteract the general tendency in surveys of AI to
talk about what’s possible instead of what’s actually been accomplished.
The problem with the former approach is that people have an odd series
of reactions to the idea of artificial intelligence. Often their first reaction
is doubt that such a thing is possible; but then they swing to the opposite
extreme, and start believing that anything they can imagine doing can be
automated. A description of how it might be possible to program a com-
puter to carry on a conversation encourages this gullibility, by painting a
vivid picture of what such a program would be like, without explaining
that we are very far from having one. Hence, throughout the book, I try
to differentiate what we know how to build from what we can imagine
building.
I left chapter 5 in for a different reason. I think the most serious objec-
tions to a computational account of mind rest on the issue of the observer-
relativity of symbols and semantics, the question of whether symbols can
mean anything, or can even be symbols in the first place, unless human
beings impute meanings to them. This may not seem like the most serious
objection for many readers, and they can skip most of chapter 5. Readers
who appreciate the objection will want to know how I answer it.
With these caveats in mind, let me invite you to enjoy the book. The
puzzles that arise in connection with the mind-body problem are often
entertaining, once you’ve wrapped your mind around them. They are
also important. If people really can be explained as machines controlled
by computational brains, what impact does that have on ethics or religion?
Perhaps we can’t answer the question, but we should start asking it soon.
xiv
Acknowledgments
I must thank David Chalmers for writing a great, if totally misguided,

book on consciousness that seemed to demand a response. I also thank
him for strenuous conversations about the role of qualia in mental life.
Various drafts of this manuscript were read by my siblings (John, Phil,
Jim, and Marcia), Will Miranker, two anonymous referees, and my wife,
Judy Nugent. Their comments were all very helpful, although there are still
plenty of flaws in the book, for which none of these people is responsible.
xv
xvi
1
The Problem of Phenomenal Consciousness
Science has pushed man farther and farther from the center of the uni-
verse. We once thought our planet occupied that center; it doesn’t. We
once thought that our history was more or less the history of the world;
it isn’t. We once thought that we were created as the crown and guardian
of creation; we weren’t. As far as science is concerned, people are just a
strange kind of animal that arrived fairly late on the scene. When you
look at the details of how they work, you discover that, like other life
forms, people’s bodies are little chemical machines. Enzymes slide over
DNA molecules, proteins are produced, various chemical reactions are
catalyzed. Molecules on the surfaces of membranes react to substances
they come into contact with by fitting to them and changing shape, which
causes chemical signals to alter the usual flow of events, so that the ma-
chine’s behavior can change as circumstances change.
Traditionally there was one big gap in this picture: the human mind.
The mind was supposed to be a nonphysical entity, exempt from the laws
that govern the stars, the earth, and the molecules that compose us. What
if this gap closes? What if it turns out that we’re machines all the way
through?
This possibility may seem too implausible or repugnant to contemplate.
Nonetheless, it looms on the horizon. For some of us, it seems like the
most likely possibility. The purpose of this essay is to increase the plau-
sibility of the hypothesis that we are machines and to elaborate some of
its consequences. It may seem that a wiser or more moral strategy would
be to avoid thinking about such a weird and inhuman hypothesis. I can’t
agree. If we are indeed physical systems, then I get no comfort from the
fact that most people don’t know it and that I can occasionally forget it.
1
2 Chapter 1
Mind as Self-Fulfilling Description
I will be arguing that people have minds because they, or their brains, are
biological computers. The biological variety of computer differs in many
ways from the kinds of computers engineers build, but the differences are
superficial. When evolution created animals that could benefit from per-
forming complex computations, it thereby increased the likelihood that
some way of performing them would be found. The way that emerged used
the materials at hand, the cells of the brain. But the same computations
could have been performed using different materials, including silicon. It
may sound odd to describe what brains do as computation, but, as we
shall see, when one looks at the behavior of neurons in detail, it is hard to
avoid the conclusion that their purpose is to compute things. Of course,
the fact that some neurons appear to compute things does not rule out
that those same neurons might do something else as well, maybe some-
thing more important; and there are many more neurons whose purpose
has not yet been fathomed.
Even if it turns out that the brain is a computer, pure and simple, an
explanation of mind will not follow as some kind of obvious corollary. We
see computers around us all the time, none of which has a mind. Brains
appear to make contact with a different dimension. Even very simple
animals seem to be conscious of their surroundings, at least to the extent
of feeling pleasure and pain, and when we look into the eyes of complex
animals such as our fellow mammals, we see depths of soul. In humans the
mind has reached its earthly apogee, where it can aspire to intelligence,
morality, and creativity.
So if minds are produced by computers, we will have to explain how.
Several different mechanisms have been proposed, not all of them plau-
sible. One is that they might “excrete” mind in some mysterious way, as
the brain is said to do. This is hardly an explanation, but it has the virtue
of putting brains and computers in the same unintelligible boat. A variant
of this idea is that mind is “emergent” from complex systems, in the way
that wetness is “emergent” from the properties of hydrogen and oxygen
atoms when mixed in great numbers to make water.
I think we can be more specific about the way in which computers
can have minds. Computers manipulate information, and some of this
The Problem of Phenomenal Consciousness 3
information has a “causative” rather than a purely “descriptive” char-

acter. That is, some of the information a computer manipulates is about
entities that exist because of the manipulation itself. I have in mind en-
tities such as the windows one sees on the screens of most computers
nowadays. The windows exist because the computer behaves in a way
consistent with their existing. When you click “in” a window, the result-
ing events occur because the computer determines where to display the
mouse-guided cursor and determines which window that screen location
belongs to. It makes these determinations by running algorithms that con-
sult blocks of stored data that describe what the windows are supposed
to look like. These blocks of data, called data structures, describe the
windows in the same way that the data structures at IRS Central describe
you. But there is a difference. You don’t exist because of the IRS’s data
structures, but that’s exactly the situation the window is in. The window
exists because of the behavior of the computer, which is guided by the very
data structures that describe it. The data structures denote something that
exists because of the data structure denoting it: the data structure is a wish
that fulfills itself, or, less poetically, a description of an object that brings
the object into being. Such a novel and strange phenomenon ought to have
interesting consequences. As I shall explain, the mind is one of them.
An intelligent computer, biological or otherwise, must make and use
models of its world. In a way this is the whole purpose of intelligence, to
explain what has happened and to predict what will happen. One of the
entities the system must have models of is itself, simply because the system
is the most ubiquitous feature of its own environment. At what we are
pleased to call “lower” evolutionary levels, the model can consist of simple
properties that the organism assigns to the parts of itself it can sense. The
visual system of a snake must classify the snake’s tail as “not prey.” It
can do this by combining proprioceptive and visual information about
where its tail is and how it’s moving. Different parts of its sensory field
can then be labeled “grass,” “sky,” “possibly prey,” “possible predator,”
and “tail.” The label signals the appropriateness of some behaviors and
the inappropriateness of others. The snake can glide over its tail, but it
mustn’t eat it.
The self-models of humans are much more complex. We have to cope
with many more ways that our behavior can affect what we perceive. In
4 Chapter 1
fact, there are long intervals when everything we perceive involves us. In
social settings, much of what we observe is how other humans react to
what we are doing or saying. Even when one person is alone in a jungle,
she may still find herself explaining the appearance of things partly in
terms of her own observational stance. A person who did not have beliefs
about herself would appear to be autistic or insane. We can confidently
predict that if we meet an intelligent race on another planet they will
have to have complex models of themselves, too, although we can’t say
so easily what those models will look like.
I will make two claims about self-models that may seem unlikely at
first, but become obvious once understood:
1. Everything you think you know about yourself derives from your self-
model.
2. A self-model does not have to be true to be useful.
The first is almost a tautology, although it seems to contradict a traditional
intuition, going back to Descartes, that we know the contents of our
minds “immediately,” without having to infer them from “sense data” as
we do for other objects of perception. There really isn’t a contradiction,
but the idea of the self-model makes the tradition evaporate. When I say
that “I” know the contents of “my” mind, who am I talking about? An
entity about whom I have a large and somewhat coherent set of beliefs,
that is, the entity described by the self-model. So if you believe you have
free will, it’s because the self-model says that. If you believe you have
immediate and indubitable knowledge of all the sensory events your mind
undergoes, that’s owing to the conclusions of the self-model. If your beliefs
include “I am more than just my body,” and even “I don’t have a self-
model,” it’s because it says those things in your self-model. As Thomas
Metzinger (1995b) puts it, “since we are beings who almost constantly
fail to recognize our mental models as models, our phenomenal space is
characterized by an all-embracing naive realism, which we are incapable
of transcending in standard situations.”
You might suppose that a self-model would tend to be accurate, other
things being equal, for the same reason that each of our beliefs is likely
to be true: there’s not much point in having beliefs if they’re false. This
supposition makes sense up to a point, but in the case of the self-model we
run into a peculiar indeterminacy. For most objects of belief, the object
exists and has properties regardless of what anyone believes. We can pic-
ture the beliefs adjusting to fit the object, with the quality of the belief
depending on how good the fit is (Searle 1983). But in the case of the
self, this picture doesn’t necessarily apply. A person without a self-model
would not be a fully functioning person, or, stated otherwise, the self does
not exist prior to being modeled. Under these circumstances, the truth of
a belief about the self is not determined purely by how well it fits the
facts; some of the facts derive from what beliefs there are. Suppose that
members of one species have belief P about themselves, and that this
enables them to survive better than members of another species with be-
lief Q about themselves. Eventually everyone will believe P, regardless of
how true it is. However, beliefs of the self-fulfilling sort alluded to above
will actually become true because everyone believes them. As Nietzsche
observed, “The falseness of a judgment is . . . not necessarily an objection
to a judgment . . . . The question is to what extent it is life-promoting . . . ,
species-preserving . . . ” (Nietzsche 1886, pp. 202–203). For example, a
belief in free will is very close (as close as one can get) to actually having
free will, just as having a description of a window inside a computer is
(almost) all that is required to have a window on the computer’s screen.
I will need to flesh this picture out considerably to make it plausible. I
suspect that many people will find it absurd or even meaningless. For one
thing, it seems to overlook the huge differences between the brain and
a computer. It also requires us to believe that the abilities of the human
mind are ultimately based on the sort of mundane activity that computers
engage in. Drawing windows on a screen is trivial compared to writing
symphonies, or even to carrying on a conversation. It is not likely that
computers will be able to do either in the near future. I will have to argue
that eventually they will be able to do such things.
Dualism and Its Discontents
The issues surrounding the relation between computation and mind are
becoming relevant because of the complete failure of dualism as an ex-
planation of human consciousness. Dualism is the doctrine that people’s
minds are formed of nonphysical substances that are associated with their
6 Chapter 1
bodies and guide their bodies, but that are not part of their bodies and are
not subject to the same physical laws as their bodies. This idea has been
widely accepted since the time of Descartes, and is often credited to him,
but only because he stated it so clearly; I think it is what anyone would
come to believe if they did a few experiments. Suppose I ring a bell in your
presence, and then play a recording of the 1812 Overture for you. You are
supposed to raise your hand when you hear the sound of that bell. How
do you know when you hear that sound? Introspectively, it seems that,
though you don’t actually hear a bell ringing, you can summon a “mental
image” of it that has the same tonal quality as the bell and compare it at
the crucial moment to the sounds of the church bells near the end of the
overture. (You can summon it earlier, too, if not as vividly, and note its
absence from the music.) Now the question is, where do mental sounds
(or visual images, or memories of smells) reside? No one supposes that
there are tiny bell sounds in your head when you remember the sound of
a bell. The sounds are only “in your mind.” Wherever this is, it doesn’t
seem to be in your brain.
Once you get this picture of the relation between mind and brain, it
seems to account for many things. I’ve focused on remembering the sound
of a bell, but it also seems to account for perceiving the sound as a bell
sound in the first place. The bell rings, but I also experience it ringing.
Either event could occur without the other. (The bell could ring when I’m
not present; I could hallucinate the ringing of a bell.) So the experience is
not the same as the ring. In fact, the experience of the ring is really closer
than the physical ringing to what I mean by the word or concept “ring.”
Physics teaches us all sorts of things about metal, air, and vibration, but the
experience of a ringing doesn’t ever seem to emerge from the physics. We
might once have thought that the ringing occurs when the bell is struck,
but we now know that it occurs in our minds after the vibrations from the
bell reach our minds. As philosophers say, vibration is a primary quality
whereas ringing is a secondary quality.
Philosophers use the word quale to describe the “ringyness” of the expe-
rience of a bell, the redness of the experience of red, the embarrassingness
of an experience of embarrassment, and so forth. Qualia are important
for two reasons. First, they seem to be crucially involved in all perceptual
events. We can tell red things from green things because one evokes a red
quale and the other a green one. Without that distinction we assume we
couldn’t tell them apart, and indeed color-blind people don’t distinguish
the quale of red from the quale of green. Second, qualia seem utterly un-
physical. Introspectively they seem to exist on a different plane from the
objects that evoke them, but they also seem to fill a functional role that
physical entities just could not fill. Suppose that perceiving or remem-
bering a bell sound did cause little rings in your head. Wouldn’t that be
pointless? Wouldn’t we still need a further perceptual process to classify
the miniature events in our heads as ringings of bells, scents of ripe apples,
or embarrassing scenes?
So far I have focused on perception, but we get equally strong intuitions
when we look at thought and action. It seems introspectively as if we
act after reasoning, deciding, and willing. These processes differ from
physical processes in crucial respects. Physical processes are governed by
causal laws, whereas minds have reasons for what they do. A causal law
enables one to infer, from the state of a system in one region of space-time,
the states at other regions, or at least a probability distribution over those
states. The “state” of a system is defined as the values of certain numerical
variables, such as position, velocity, mass, charge, heat, pressure, and so
forth—primary qualities. We often focus for philosophical purposes on
the case of knowing a complete description of a system at a particular
time and inferring the states at later times, but this is just one of many
possible inference patterns. All of them, however, involve the inference of
a description of the physical state of the system at one point in space and
time from a description of its state at other points. By contrast, the reason
for the action of a person might be to avoid a certain state. A soldier
might fall to the ground to avoid getting shot. People are not immune to
physical laws; a soldier who gets shot falls for the same reason a rock
does. But people seem to transcend them.
This idea of physical laws is relatively new, dating from the seventeeth
century. Before that, no one would have noticed a rigid distinction between
the way physical systems work and the way minds work because everyone
assumed that the physical world was permeated by mental phenomena.
But as the universe came to seem mechanical, the striking differences
between the way it works and the way our minds work became more
obvious. Descartes was the first to draw a line around the mind and
8 Chapter 1
put all mental phenomena inside that boundary, all physical phenomena
outside it.
Nowhere is the contrast between cause and reason more obvious than
in the phenomenon of free will. When you have to make a decision about
what to do, you take it for granted that you have a real choice to make
among alternative actions. You base your choice on what you expect to
happen given each action. The choice can be difficult if you are not sure
what you want, or if there is a conflict between different choice criteria.
When the conflict is between principle and gain, it can be quite painful. But
you never feel in conflict in the same way with the principle of causality,
and that makes it hard to believe that it is a physical brain making the
decision. Surely if the decision-making process were just another link in a
chain of physical events it would feel different. In that case the outcome
would be entirely governed by physical laws, and it would simply happen.
It is hard to imagine what that would feel like, but two scenarios come
to mind: either you would not feel free at all, or occasionally you would
choose one course of action and then find yourself, coerced by physics,
carrying out a different one. Neither scenario obtains: we often feel free
to choose, and we do choose, and then go on from there.
Arguments like these make dualism look like a very safe bet, and for
hundreds of years it was taken for granted by almost everyone. Even those
who found it doubtful often doubted the materialist side of the inequality,
and conjectured that mind was actually more pervasive than it appears.
It is only in the last century (the twentieth) that evidence has swung the
other way. It now seems that mere matter is more potent than we thought
possible. There are two main strands of inquiry that have brought us to
this point. One is the burgeoning field of neuroscience, which has given us
greater and greater knowledge of what brains actually do. The other is the
field of computer science, which has taught what machines can do. The
two converge in the field of cognitive science, which studies computational
models of brains and minds.
Neither of these new sciences has solved the problems it studies, or
even posed them in a way that everyone agrees with. Nonetheless, they
have progressed to the point of demonstrating that the dualist picture is
seriously flawed. Neuroscience shows that brains apparently don’t con-
nect with minds; computer science has shown that perception and choice
apparently don’t require minds. They also point to a new vision of how
brains work in which the brain is thought of as a kind of computer.
Let’s look at these trends in more detail, starting with the brain. The
brain contains a large number (1011 ) of cells called neurons that
apparently do all its work.1 A neuron, like other cells, maintains different
concentrations of chemicals on each side of the membrane that surrounds
it. Because many of these chemicals are electrically charged ions, the result
is a voltage difference across the membrane. The voltage inside the cell
is about 60 millivolts below the voltage outside. When stimulated in the
right way, the membrane can become depolarized, that is, lose its voltage
difference by opening up pores in the membrane and allowing ions to
flow across. In fact, the voltage briefly swings the opposite direction, so
that the inside voltage become 40 millivolts above the outside voltage.
When that happens, neighboring areas of the membrane depolarize as
well. This causes the next area to depolarize, and so forth, so that a wave
of depolarization passes along the membrane. Parts of the cell are elon-
gated (sometimes for many centimeters), and the wave can travel along
such an elongated branch until it reaches the end. Behind the wave, the cell
expends energy to pump ions back across the membrane and reestablish
the voltage difference.
When the depolarization wave reaches the end of a branch, it can cause
a new wave to be propagated to a neighboring cell. That’s because the
branches of neurons often end by touching the branches of neighboring
neurons. Actually, they don’t quite touch; there is a gap of about one
billionth of a meter (Churchland 1986). The point where two neurons
come into near contact is called a synapse. When a depolarization wave
hits a synapse, it causes chemicals called neurotransmitters to be emitted,
which cross the gap to the next neuron and stimulate its membrane. In
the simplest case one may visualize the gap as a relay station: the signal
jumps the gap and continues down the axon of the next neuron. When
a neuron starts a depolarization wave, it is said to fire. Many neurons
have one long branch called the axon that transmits signals, and several
shorter ones called dendrites that receive them. The axon of one neuron
will make contact at several points on the dendrites of the next neu-
ron. (A neuron may have more than one axon, and an axon may make
contact on the dendrites of more than one neuron.) A depolarization
10 Chapter 1
wave travels at a speed between a few hundred centimeters per second

and a hundred meters per second, depending on exactly how the axon is
configured.
Nowadays we take it for granted that the reason neurons fire is to con-
vey information. That’s because we’re familiar with the transmission of
information in physical forms that are remote from the forms they take
when they are first captured or ultimately used. It doesn’t strike us as
odd that sound waves, disturbances in air pressure, are encoded as little
bumps on CDs or electrical impulses in wires. Two hundreds years ago
this idea would not have been so obvious, and someone looking at the op-
eration of the brain might have been quite puzzled by the depolarizations
traveling through neural membranes. Those who take dualism seriously
might demand proof that the signals were actually conveying informa-
tion. Fortunately, it’s not hard to find proof. First, we need to show that
as the situation is varied in one place in the brain, the behavior of neurons
elsewhere in the brain varies accordingly. This phenomenon has indeed
been demonstrated over and over. Light is received by the retina, and neu-
rons in the visual system become active; sound is received by the ear, and
neurons elsewhere in the brain respond. Of course, it is not enough to
show variation. We must also show that there is a code of some kind, so
that a piece of information is represented by a consistent pattern of neural
behavior that is different from the pattern for other pieces of information.
I’ll talk about that shortly.
If neural impulses were on their way to a nonphysical mind, one would
expect to see neurons transmitting data as faithfully as possible, preserving
the content intact until it reached the place where true perception began,
and where qualia arose. I’ve sketched this possibility in figure 1.1. At the
point where the interface to the mind appears, one might see neurons
with inputs and no outputs. On the “other side” of the mind, one would
see neurons with outputs but no inputs, which react to the decisions of
the mind by sending impulses to the muscles relevant to the action the
mind has decided on. The gap between the last layer of input neurons and
the first layer of output neurons might not be so blatant. There might be
no spatial gap at all, just a “causality gap,” where the behavior of the
output neurons could not be entirely explained by the behavior of the
input neurons, but would also depend in some way on mental events.
The Mind
Relay Stations
Figure 1.1
The naı̈ve dualist picture
Lest figure 1.1 be thought of as a straw man, in figure 1.2 I have re-
produced a figure endorsed by Sir John Eccles, one of the few unabashed
dualists to be found among twentieth-century neurophysiologists (Eccles
1970, figure 36, detail). He divides the world into the material domain
(“World 1”), the mental domain (“World 2”), and the cultural domain
(“World 3”), which I have omitted from the figure. The brain is mostly
in World 1, but it makes contact with World 2 through a part called the
“liaison brain.” The liaison brain is where Eccles supposes the causality
gap lies.
Unfortunately for this dualist model, the behavior of neurons doesn’t
fit it. For one thing, there are few places at which data are simply trans-
mitted. Usually a neuron fires after a complex series of transmissions are
received from the neurons whose axons connect to it. The signals coming
out of a group of neurons are not copies of the signals coming in. They
12 Chapter 1
World 2
Liaison Brain
(World 1)
Memory stores
(World 3b)
Brain
World 1
Body
Aff. World 1 Eff.
path. path.
World 1
Figure 1.2
From Eccles 1970, p. 167, figure 36
are, however, a function of the signals coming in. That is, if there is a
nonphysical “extra” ingredient influencing the output of the neurons, its
effects must be very slight. As far as we can tell, any given input always
results in essentially the same output.
We have to be careful here about exactly how much to claim. Neurons’
behavior changes over time, as they must if they are to be able to learn.
However, such changes are reflected in changes of synapses’ sensitivity to

inputs, so that the output becomes a function of the input plus the state
of the synapses, and their state changes as a function of its input and
previous state. More subtly, we must deal with the fact that no system
ever behaves exactly the same way twice. There is always a tiny variation
in the output. Is this a place where mental effects can slip in?
The answer will be no if the variations are irrelevant to the information
carried by the neuronal signals. The only way to judge what’s relevant is to
understand better how neurons encode information. So far, we understand
it only partially. In many cases, the information encoded is the average rate
at which the neuron fires. For instance, when light strikes a light-sensitive
cell in the retina of the eye, the cell responds by firing several times in
succession.2 It behaves for all the world as if the brightness of the light
is encoded by the number of depolarization waves that are transmitted
per time unit. The size of each depolarization is irrelevant. Indeed, they
are all of about the same size. The membrane either depolarizes or it
doesn’t. Each depolarization, called a spike because of its appearance
when charted against time, is rapid and complete. What changes is the
number of spikes that occur. The greater the stimulation, the faster the
firing rate. Some cells can be negatively stimulated, or inhibited, in which
case they fire more slowly than their normal rate.
Firing rates are not the only coding scheme in the brain. They could not
be, because before a neuron fires its dendrites are combining information
from several axons using as a medium only the smoothly varying voltages
across postsynaptic membranes. In figure 1.3, the axon on the left has fired
and caused the membrane voltage to increase, from its normal value of
−70 mV to more like −40 mV. However, the axon on the right is connected
to this cell by an inhibitory synapse, so that as it fires it “hyperpolarizes”
the membrane, driving it even more negative than usual. The net effect
depends on the geometry, chemistry, and input firing rates, and is not well
understood.
Given that any physical quantity is capable in principle of serving as a
code for a piece of information, it might seem impossible to determine if a
given physical setup is properly described as actually encoding anything.
Fortunately, there is another principle we can appeal to. Suppose someone
proposes a certain code for some module of the brain. Then we can look
14 Chapter 1
Axon 2
Axon 1
Dendrite
Cell Body
⫹40 mV
Position along
membrane
Voltage
⫺70 mV
Figure 1.3
Graded action potentials along dendritic membrane
at the data represented by the inputs to that module under the proposal,
look at the data encoded by the outputs, and ask whether the output is
an interesting function of the input, in the sense that one could see why
an organism would want to compute that function. If the answer is yes,
then that is evidence that the code is real. Such codes are now found
routinely. For example, in the visual system of the brain there are arrays
of cells whose inputs represent brightness values at all points of the visual
field and whose outputs represent the degree to which there is a vertical
brightness edge at each point. Such an edge is defined as a brightness
profile that changes value sharply along a horizontal line, from light to
dark or vice versa. Other arrays of cells are sensitive to edges with other
orientations.
Finding edge detectors like this is exciting because there are independent
theories of how information might be extracted from the visual field that
suggest that finding edges will often be useful. (For example, if the sun
goes behind a cloud, all the brightnesses become smaller, but many of
the edges stay in the same place.) But what I want to call attention to is
that the edges are being found before the signals “reach the mind.” Edges
are not something perceived qualitatively, or at least not exclusively. Here
we find edges being found in a computational sense that is essentially
independent of mind.
At this point we have been led to notice the importance of the second
major intellectual strand in the story, namely, the science of computa-
tion. We usually use the phrase “computer science” to refer to it, but
that doesn’t mean it’s about laptops and mainframes. It’s about physical
embodiments of computational processes, wherever we find them, and it
appears that one place we find them is in groups of neurons.
Let’s turn our attention from neurons for a second and think in terms of
artificial systems. Suppose we build an artificial ear. It takes sound waves,
analyzes them for different frequencies, and prints out the word “bell”
if the frequency analysis matches the profile for a bell, and “not a bell”
otherwise. I don’t mean to suggest that this would be easy to do; in fact,
it’s quite difficult to produce an artificial ear that could discriminate as
finely as a person’s. What I want to call attention to is that in performing
the discrimination the artificial ear would not actually experience ringing
or the absence of ringing; it would not experience anything. If you doubt
that, let’s suppose that I can open it up and show you exactly where the
wires go, and exactly how the software is written. Here a set of tuning
forks vibrate sympathetically to different frequencies; here an analog-to-
digital converter converts the amplitude of each vibration to a numerical
quantity; there a computer program matches the profile of amplitudes to
a stored set of profiles. There’s no experience anywhere, nor would we
expect any.
Hence it should give us pause if the structures in the brain work in
similar ways. And in fact they do. The ear contains a spiraling tube, the
cochlea, different parts of whose membrane vibrate in resonance with
different frequencies. These vibrations cause receptor cells to send trains
of spikes of different frequencies, the frequency of a spike train encoding
the amplitude of a particular sound frequency. The overuse of the word
“frequency” here is confusing, but also illuminating. One physical quan-
tity, the rate at which a neuron fires, is being used to encode a completely
different quantity, the magnitude of a certain frequency in the spectrum of
a sound, just as voltages are used in digital computers to represent entities
16 Chapter 1
that have nothing to do with voltages. The representations feed further

computations in each case, leading to the extraction of richer lodes of in-
formation, such as the phonemes and words hidden in a stream of sounds.
What they apparently don’t lead to is any experience of anything.
The problem is that every time we find information being extracted
computationally, that’s one less job for the mind to do. If we find that
color discriminations can be done by nonminds, that is, by sensors and
computers, and we find the brain doing it the way a nonmind would do
it, where does the mind come in? The traditional dualist presupposition
was that the brain does almost nothing except prepare raw sensory data
to be passed to the mind. That’s not the case, so we have to find some
other way for the mind to be present.
One possibility is that the mind is still lurking there, it just requires a bit
more preprocessing than we used to think. In other words, it receives not
raw brightness data, but data already helpfully analyzed into colors, edges,
texture analysis, matchings of the images from the two eyes, and so forth.
What the mind does is experience all these things. Another possibility is
that experience inheres in the brain as a whole; in addition to performing
computation, the collection of molecules has a quite different function,
namely, to provide qualia for some of the discriminations made by the
computations.
The question that now arises is what role these experiences play in
determining behavior. Dualism maintains that they are crucial. If we trace
out the events happening in neurons, we will (at least in principle) find
places where what happens cannot be explained purely in terms of physics
and chemistry. That is, we would find a gap where a physical event P1
caused a nonphysical experience N, which then caused another physical
event P2 . For example, consider the behavior of a wine taster. She sips a
bit of 1968 Chateau Lafitte Rothschild, rolls it around her tongue, and
pronounces “magnifique.” Dualism predicts that if we were to open up
her brain and take a peek, we will see neurons producing spike trains
that represent various features of the wine. These spike trains apparently
would go nowhere, because their sole function is to interface with the
intangible dualist mind. Other neurons would produce impulses sent to
the mouth and vocal cords to utter the word “magnifique,” but these
impulses would apparently not be a function of anything. As discussed
above, the gap might be much smaller than depicted in figures 1.1 and 1.2,
but it would have to be there, and the link across it would be nonphysical.
It is possible that we will encounter such a linkage in the brain, but
almost no one expects to. Eccles (1973, p. 216) proposes that his “liaison
brain” is located in the left hemisphere, because the speech center of most
people is located in the left hemisphere. Needless to say, despite intense
research activity, no such linkage has appeared. Of course, if it existed it
would be very hard to find. The network of neurons in the brain is an
intricate tangle, and there are large sections as yet unexplored. Almost
all experimentation on brains is done on nonhuman animals. Probing a
person’s brain is allowed only if the probing has no bad or permanent
effects. The failure to find a liaison brain in a nonhuman brain might
simply indicate that such brains are not conscious. On the other hand, if
an animal is conscious, it might be considered just as unethical to exper-
iment on its brain as on one of ours. So we may be eternally barred from
the decisive experiment. For all these reasons it will probably always be
an option to believe that dualistic gaps exist somewhere, but cannot be
observed. However, failing to find the causal chain from brain to mind
and back is not what the dualist must worry about. The real threat is
that neuroscientists will find another, physical chain of events between
the tasting of the wine and the uttering of the words.
Suppose we open up the brain of a wine taster and trace out exactly
all the neural pathways involved in recognizing and appreciating a sip
of wine. Let’s suppose that we have a complete neuroscientific, compu-
tational explanation of why she utters “magnifique!” instead of “sacre
bleu!” at a key moment. Even if there is an undetected dualist gap, there
is nothing for it to do. Nothing that occurs in the gap could affect what
she says, because we have (we are imagining) a complete explanation of
what she says.
One further option for the dualist is to say that experience is a non-
physical event-series that accompanies and mirrors the physical one. It
plays no causal role, it just happens. This position is called epiphenom-
enalism. This is a possibility, but an unappealing one. For one thing, it
seems like an odd way for the universe to be. Why is it that the sort
of arrangement of molecules that we find in the brain is accompanied
by this extra set of events? Why should the experiences mirror what the
18 Chapter 1
brain does so closely? Keep in mind that the dualist holds that there is
no physical explanation of this link, so it is difficult to point to physical
properties of the molecules in the brain that would have anything to do
with it. The interior of the sun consists of molecules that move around
in complex ways. Are these motions accompanied by experiences? When
an ice cube melts, its molecules speed up in reaction to the heat of the
environment. Does the ice cube feel heat? It can’t tell us that it does, but
then you can’t tell us either. You may believe that your utterance of the
words “It’s hot in here” has something to do with your experience of
heat, but it actually depends on various neurophysiological events that
are no different in principle from what happens to the ice cube. The ex-
perience of heat is something else that happens, on the side. Difficulties
such as these make epiphenomenalism almost useless as a framework for
understanding consciousness, and I will have little to say about it.
Most scientists and philosophers find the problems with dualism in-
surmountable. The question is what to replace it with. A solution to the
problem of consciousness, or the mind-body problem, would be a purely
physical mechanism in the brain whose behavior we could identify with
having experience. This is a tall order. Suppose we filled it, by finding
that mechanism. As we peered at it, we could say with certainty that a
certain set of events was the experience of red, that another set was a kind
of pleasure, another an excruciating pain. And yet the events would not
be fundamentally different from the events we have already observed in
brains.
Furthermore, if the brain really is just an organic information-
processing system, then the fact that the events occur in neurons would
just be a detail. We would expect that if we replaced some or all of the neu-
rons with equivalent artificial systems, the experiences wouldn’t change.
This seems implausible at first. We think of living tissue as being intrin-
sically sensitive in ways that silicon and wire could never be. But that
intuition is entirely dualistic; we picture living tissue as “exuding” ex-
perience in some gaseous manner that we now see isn’t right. At a fine
enough resolution, the events going on in cells are perfectly mechanical.
The wetness disappears, as it were, and we see little molecular machines
pulling themselves along molecular ropes, powered by molecular springs.
At the level of neurons, these little machines process information, and
they could process the information just as well if they were built using
different molecules.
This idea seems so unpalatable to some that they refuse to speculate any
further. Consciousness is evidently so mysterious that we will never un-
derstand it. We can’t imagine any way to build a bridge between physics
and mental experience, so we might as well give up. (Colin McGinn is
the philosopher most closely associated with this position; see McGinn
1991.) This is an odd position to take when things are just starting
to get interesting. In addition, one can argue that it is irresponsible to
leave key questions about the human mind dangling when we might clear
them up.
Modern culture is in an awkward spot when it comes to the mind-body
problem. Scientists and artists are well aware that dualism has failed,
but they have no idea what to replace it with. Meanwhile, almost ev-
eryone else, including most political and religious leaders, take dualism
for granted. The result is that the intellectual elite can take comfort in
their superiority over ordinary people and their beliefs, but not in much
else. Is this a state of affairs that can last indefinitely without harmful
consequences?
I realize that there is a cynical view that people have always accepted
delusions and always will. If most people believe in astrology, there can’t
be any additional harm in their having incoherent beliefs about what goes
on inside their heads. I suppose that if you the view the main purpose of the
human race as the consumption of products advertised on TV, then their
delusions are not relevant. I prefer to think that, at the very least, humans
ought to have a chance at the dignity that comes from understanding and
accepting their world. Our civilization ought to be able to arrive at a
framework in which we appreciate human value without delusions.
Nondualist Explanations of Consciousness
The field of consciousness studies has been quite busy lately. There seem
to be two major camps on the mind-body problem: those who believe that
we already have the tools we need to explain the mind, and those who
believe that we don’t and perhaps never will. McGinn is in the pessimistic
camp, as is Nagel (1975) and others. I’m an optimist.
20 Chapter 1
Some theorists (notably Bernard Baars 1988, 1996) focus on explaining

the function of consciousness and how this function might be realized by
the structures of the brain. Although this is an important area, it won’t be
my main focus. I agree with Chalmers’s assessment (Chalmers 1996) that
the “hard problem” of consciousness is to explain how a physical object
can have experiences. This is the problem of phenomenal consciousness.
It is a hard problem for all theories, but especially for computational ones.
Like everyone else, I can’t define “phenomenal consciousness.” It’s the
ability to have experiences. I assume that anyone who can read this knows
what it means to experience something (from their own experience!); and
that everyone knows that thermostats don’t have experiences, even though
they can react to temperature differences.
The difficulty of defining consciousness has led some to propose that
there is no thing as consciousness. The standard citations are to
Churchland (1990) and Rorty (1965). This position is called elimina-
tivism. The idea is to replace the concept of consciousness with more
refined (more scientific?) concepts, much as happened with concepts like
“energy” and “mass” in past scientific revolutions. It seems plain that a
full understanding of the mind will involve shifts of this kind. If we ever
do achieve fuller understanding (which the pessimists doubt), any book
written before the resulting shift, including this one, will no doubt seem
laughably quaint. However, we can’t simply wait around for this to hap-
pen. We have to work on the problems we see now, using the tools at hand.
There is clearly a problem of how a thing, a brain or computer, can have
experiences, or appear to have what people are strongly tempted to call
experiences. To explain how this is conceivable at all must be our goal.
O’Brien and Opie (1999) make a useful distinction between vehicle and
process theories of phenomenal consciousness:
Either consciousness is to be explained in terms of the nature of the representa-
tional vehicles the brain deploys, or it is to be explained in terms of the com-
putational processes defined over these vehicles. We call versions of these two
approaches VEHICLE and PROCESS theories of consciousness, respectively.
A process theory is one that explains experience in terms of things the

brain does, especially things it computes. A vehicle theory explains expe-
rience as a property of the entity doing the computing. O’Brien and Opie
themselves propose a vehicle theory in which experience is identified with
(or correlated with?) stable activation patterns in networks of neurons.

Another example, much vaguer, is that of John Searle (1992), who appeals
to unknown “causal powers” of brain tissue to explain experience.
Then there is the theory of Stuart Hammeroff (1994; Penrose 1994),
which explains consciousness in terms of the quantum-mechanical be-
havior of brain cells, specifically their microtubules. I don’t know how
this approach fits into O’Brien and Opie’s dichotomy, because the details
are so fuzzy. But if the theory is a process theory, it’s a theory of an un-
usual kind, because the processes in question cannot, by hypothesis, be
explained mechanistically. Instead, the proposal seems to be that the abil-
ity of the mind to arrive at sudden insights is explained in the same way as
the sudden collapse of wave functions in quantum mechanics. Grant that,
and perhaps you may be willing to grant that phenomenal consciousness
gets explained somehow, too.
The problem with all these theories besides their vagueness is that they
are vulnerable to subversion by competing process theories (McDermott
1999). This is a flaw they share with dualism. As I explained above, any
dualistic or “vehicular” explanation of the mind will have to accommo-
date all the facts that mundane process models explain, a set of facts that
one can expect to grow rapidly. Every time we explain a mental abil-
ity using ordinary computational processes, we will have to redraw the
boundary between what the vehicle theory accounts for and what the
process theory accounts for. There are only two ways for this process to
be arrested or reversed: either the vehicle theorists must explain more, or
the process theorists must fail to explain very much. In the first case, the
vehicle theory must compete with process theory on the process theory’s
home field, explaining how particular behaviors and competences can
result from implementing a computation in one medium rather than an-
other. That’s hard to picture. In the second case, the vehicle theory will get
a tie by default; no one will have explained consciousness, so the vehicle
theory will have explained it as well as anyone.
If process theory doesn’t fail, however, vehicle theory will be left in an
odd position indeed. Suppose we have two entities E1 and E2 that be-
have intelligently, converse on many topics, and can tell you about their
experiences. Their similarity is explained by the fact that they implement
the same computational processes. However, the one implemented using
22 Chapter 1
vehicle Vgenuine is, according to the theory, conscious. The other, imple-
mented using vehicle Vbogus , is only apparently conscious. When it talks of
its experiences, it’s actually making meaningless sounds, which fool only
those unfamiliar with the theory. Unfortunately, no matter how elegant the
theory is, it won’t supply any actual evidence favoring one vehicle over
the other. By hypothesis, everything observable about the two systems
is explained by the computational process they both instantiate. For in-
stance, if you wonder why E1 is sometimes unconscious of its surround-
ings when deeply involved in composing a tune, whatever explanation you
arrive at will work just fine for E2 , because they implement exactly the
same computational processes, except that in the case of E2 you’ll have
to say, “It’s ‘apparently conscious’ of its surroundings most of the time,
except when its working on a new tune, when it’s not even apparently
conscious.”
Vehicle theories are thus likely to be a dead end. This is not to say that
the explanation of consciousness may not require new mechanisms. The
point is, though, that if they are required they will still be mechanisms.
That is, they will explain observable events. Phenomenal consciousness is
not a secret mystery that is forever behind a veil. When I taste something
sour, I purse my lips and complain. A theory must explain why things
have tastes, but it must also explain why my lips move in those ways, and
the two explanations had better be linked.
Many critics of computational theories reject the idea that phenomenal
consciousness can be explained by explaining certain behaviors, such as
lip pursing, or utterances such as “Whew, that’s sour.” But even those who
believe that explaining such behavior is insufficient must surely grant that
it is necessary. We can call this the Principle of the Necessity of Behavioral
Explanation: No theory is satisfactory unless it can explain why someone
having certain experiences behaves in certain characteristic ways. Natu-
rally, process theories tend to explain behavior well and experience not
so well, whereas vehicle theories tend to have the opposite problem.
Process theories tend to fall into two groups, called first-order and
higher-order theories. The former are those in which in some contexts
the processing of sensory information is “experience-like” in a way that
allows us to say that in those contexts the processing is experience. Higher-
order theories, on the other hand, say that to be an experience a piece of
sensory information must itself be the object of perception. First-order

theories include those of Kirk (1994) and Tye (1995).
Shapiro (2000) sketches what might be considered the “primordial”
first-order theory:
Why not think of our perceptual experiences as sometimes entering a channel
that makes us phenomenally aware of what they represent and other times by-
passing this channel? When I’m driving around town thinking about a lecture I
need to prepare, my perceptions of the scenery bypass the phenomenal awareness
channel. When, on the other hand, I need to attend more closely to the world, to
heighten my awareness of the world, I elect (perhaps unconsciously) to make my
perceptions of the world conscious. Accordingly, my perceptions of the world get
funneled through the phenomenal awareness channel. Phenomenal awareness, on
this picture, requires no higher-order representational capacities.
He says this model is pure speculation, so I don’t suppose he would de-

fend it if pressed, but it makes a convenient target in that its obvious
flaw is shared by all first-order theories. The flaw is that nothing is said
about what makes events in one channel conscious and those in another
channel unconscious. In Kirk’s version, sensory data that are presented to
the “main assessment processes” (1994, p. 146) are experienced, whereas
other sensory data are not. Why not, exactly? Just as for dualist theories
and vehicle theories, first-order computational theories suffer from a dis-
connection between the hypothesized process and its visible effects. The
theories tend to treat “conscious” as a label stuck onto some signals in
the brain, without any explanation of how this labeling causes people to
be able to report on those signals (while being unable to report on those
without the label).
Second-order theories do not have this flaw, because they say exactly
what the labels consist in: computational events whose topic is the sensory
processes being labeled. Here the labels are as concrete as data structures
in a computer, so there is no difficulty following the chain of causality.
The difficulty is convincing anyone that the resulting system has real ex-
periences.
I will be pursuing a second-order theory in the rest of the book. Chap-
ter 2 is a survey of the present state of research in artificial intelligence.
Chapter 3 is a detailed explanation of my theory of consciousness.
Chapter 4 deals with various objections, including those alluded to in
the previous paragraph. The most serious objections are based on the
24 Chapter 1
observation that a computational theory of mind can’t be correct because

concepts such as “computer” and “symbol” are ill defined. Chapter 5
deals with this issue. Chapter 6 deals with various consequences of the
theory, including the impact on religion and ethics.
One feature that will strike many readers is how little I appeal to
neuroscience, unlike a great many recent theories of consciousness
(Flanagen 1992; Crick and Koch 1998; Churchland 1986, 1995). One
reason neuroscience is so popular is that it has produced some detailed,
interesting proposals for how the brain might work. Many people won’t
be satisfied with an explanation of thinking and consciousness that doesn’t
ultimately appeal to the facts of neurophysiology.
But another reason is a prejudice that the basic case of phenomenal
consciousness is the quivering bit of protoplasm in contact with the cruel
world. The brain feels, in the end, because it is made of living, feeling
parts. I have the opposite intuition: that feeling has nothing to do with
being alive. The great majority of living things never feel anything. When
evolution invented feeling, it stumbled onto a phenomenon that can be
elicited from a living system, but not just from living systems. A theory
of phenomenal consciousness must reflect this neutrality.
Another way to put it is this: Different organisms sense vastly differ-
ent things. We rely primarily on visual inputs and so we receive com-
pletely different data about the world than a bat would at the same point
in space and time. But gathering data and having experiences are two
different things, and it may be that adding phenomenal consciousness to
data processing is always a matter of adding the same simple twist to the
system. If we ever find life on other planets, we will probably find that
the data they gather from their environment, and the anatomical struc-
tures they use to process them, are specific to that environment and their
needs; but if they’re conscious it will be because of the same trick that
our brains use. What they’re conscious of will be different, but the way
they’re conscious may be the same. The same conclusions would apply to
conscious robots.
This book is mainly about philosophical questions, but I confess that
I do not always feel comfortable using the usual philosophical tools to
approach them. Usually this is owing to differences in training and dis-
position. I hope those who would feel more at home in a discussion
conducted in the pure philosophical style will nevertheless bear with me,
in spite of my neglect of some of the problems and issues that philosophers
focus on.
For example, philosophers spend a lot of time arguing about functional-
ism. This term has several meanings. Some people treat it as synonymous
with computationalism, the doctrine that the mind can be explained en-
tirely in terms of computation. Since I’m defending a version of computa-
tionalism, to that extent I’m defending functionalism, too. However, there
are also other meanings assigned to the term, which reduces its utility. One
version may be summarized thus: what mental terms and predicates refer
to is whatever fills certain causal roles in a functional description of the
organism involved. For example, a statement such as “Fred is in pain” is
supposed to mean, “Fred is in state X, where X is a state with the prop-
erties pain is supposed to have in a worked-out functional description of
Fred, e.g., the property of causing Fred to avoid what he believes causes
X.” Actually, stating the full meaning requires replacing “believe” with a
functionally described Y, and so forth.
The purpose of this version of functionalism is to show that, in principle,
mental terms can be defined so that they can be applied to systems without
making any assumptions about what those systems are made of. If pain
can be defined “functionally,” then we won’t be tempted to define it in
terms of particular physical, chemical, or neurological states. So when we
find an alien staggering from its crashed spaceship and hypothesize that
it is in pain, the claim won’t be refutable by observing that it is composed
of silicon instead of carbon.
I am obviously in sympathy with the motivation behind this project.
I agree with its proponents that the being staggering from the space-
ship might be in pain in spite of being totally unlike earthling animals.
The question is whether we gain anything by clarifying the definitions
of terms. We have plenty of clearcut mental states to study, and can
save the borderline cases for later. Suppose one had demanded of Van
Loewenhook and his contemporaries that they provide a similar sort of
definition for the concept of life and its subconcepts, such as respiration
and reproduction. It would have been a complete waste of time, because
what Van Loewenhook wanted to know, and what we are now figuring
out, is how life works. We know there are borderline cases, such as viruses,
26 Chapter 1
but we don’t care exactly where the border lies, because our understand-
ing encompasses both sides. The only progress we have made in defining
“life” is to realize that it doesn’t need to be defined. Similarly, what we
want to know about minds is how they work. My guess is that we will
figure that out, and realize that mental terms are useful and meaningful,
but impossible to define precisely.
In practice people adopt what Dennett (1978a) calls the “intentional
stance” toward creatures that seem to think and feel. That is, they sim-
ply assume that cats, dogs, and babies have beliefs, desires, and feelings
roughly similar to theirs as long as the assumption accounts for their be-
havior better than any other hypothesis can. If there ever are intelligent
robots, people will no doubt adopt the intentional stance toward them,
too, regardless of what philosophers or computer scientists say about the
robots’ true mental states. Unlike Dennett, I don’t think the success of the
intentional stance settles the matter. If a system seems to act intentionally,
we have to explain why it seems that way using evidence besides the fact
that a majority of people agree that it does. People are right when they
suppose babies have mental states and are wrong when they suppose the
stars do.
So I apologize for not spending more time on issues such as the structure
of reductionism, the difference between epistemological and metaphysi-
cal necessity, and the varieties of supervenience. I am sure that much
of what I say could be said (and has been said) more elegantly using
those terms, but I lack the requisite skill and patience. My main use of
the philosophical literature is the various ingenious thought experiments
(“intuition pumps,” in Dennett’s phrase) that philosophers have used in
arguments. These thought experiments tend to have vivid, intuitively com-
pelling consequences; that’s their whole purpose. In addition, the appar-
ent consequences are often completely wrong. I believe in those cases it is
easy to show that they are wrong without appeal to subtle distinctions;
if those familiar with the philosophical intricacies are not satisfied, there
are plenty of other sources where they can find the arguments refuted
in the correct style. In particular, Daniel Dennett (1991), David Rosen-
thal (1986, 1993), Thomas Metzinger (1995a), and William Lycan (1987,
1996) defend positions close to mine in philosophers’ terms, though they
each disagree with me on several points.
Several other nonphilosophers have proposed second-order theories of

consciousness. Marvin Minsky is especially explicit about the role of self-
models in consciousness (Minsky 1968). Douglas Hofstadter’s proposals
(Hofstadter 1979; Hofstadter and Dennett 1981) are less detailed, but in
many ways more vivid and convincing. Michael Gazzaniga (1998) bases
his ideas on neuroscience and psychology.
Adding another voice to this chorus may just confuse matters; but per-
haps it may persuade a few more people, and perhaps even clarify the
issues a bit.
28
2
Artificial Intelligence
Cognitive science is based on the idea that computation is the main thing
going on in the brain. Cognitive scientists create computational models
of the sorts of tasks that brains do, and then test them to see if they
work. This characterization is somewhat vague because there is a wide
range of models and testing strategies. Some cognitive scientists are inter-
ested in discovering exactly which computational mechanisms are used by
human brains. Others are interested primarily in what mechanisms could
carry out a particular task, and only secondarily in whether animal brains
actually use those mechanisms.
This discipline has been around for about half a century. Before that,
psychologists, linguists, neuroscientists, and philosophers asked questions
about the mind, but in different ways. It was the invention of the digi-
tal computer that first opened up the possibility of using computational
models to explain almost everything.
It is possible to approach cognitive science from various points of view,
starting from psychology, neuroscience, linguistics, or philosophy, as pre-
vious authors have done (Jackendoff 1987; Dennett 1991; Churchland
and Sejnowski 1992). My starting point is going to be computer science.
The application of computer science to cognitive science is called artificial
intelligence, or AI. AI has been used as a platform for a more general dis-
cussion before (Minsky 1986; Hofstadter and Dennett 1981), but rarely
in a way that takes philosophical questions seriously.1
One misapprehension is that artificial intelligence has to do with intelli-
gence. When the field started, it tended to focus on “intellectual” activities
such as playing chess or proving theorems. It was assumed that algorithms
29
30 Chapter 2
for hard problems like these would automatically apply to other areas re-
quiring deep thought, while “lower level” activities like walking or seeing
were assumed to be relatively straightforward applications of control the-
ory. Both of these assumptions have been abandoned. There turns out to
be no “General Problem Solver”2 that can solve a wide range of intel-
lectual problems; and there turns out to be no precise boundary line be-
tween lower-level activities and high-level “thinking.” There are plenty of
boundaries between modules, but no difference in computational mech-
anism between peripheral and central modules, and no clear support for
the notion that as you penetrate deeper into the brain you finally reach
an area in which you find “pure intelligence.”
It would be better, therefore, if AI had a different name, such as cog-
nitive informatics. Since we can’t wave a magic wand and cause such a
terminological revolution, we will keep the old label for the field, but let
me repeat: AI will have a lot to say about what role computation plays in
thinking, but almost nothing to say about intelligence.
People tend to underestimate the difficulty of achieving true machine in-
telligence. Once they’ve seen a few examples, they start drawing schematic
diagrams of the mind, with boxes labeled “perception” and “central ex-
ecutive.” They then imagine how they would carry out these subtasks,
and their introspections tell them it wouldn’t be that hard. It doesn’t take
long for them to convince themselves that intelligence is a simple thing,
just one step beyond Windows 98. The truth is that boxes are much easier
to label than to fill in. After a while, you begin to suspect the boundaries
between the boxes were misdrawn. Fifty years into the AI project, we’ve
become much more humble. No one would presume any longer to draw a
schematic for the mind. On the other hand, we have more concrete ideas
for solving particular tasks.
Some of the topics in this chapter may be familiar to some readers. I
urge them not to skim the chapter too hastily, though, because when I
allude to computational mechanisms later in the book, I’m thinking of
mechanisms such as the ones I describe here. That’s not to say that there
won’t be plenty of new algorithms and data structures discovered in the
future. I’m just not assuming the need for, or predicting the discovery of,
any powerful new ideas that will revolutionize the way we think about
computation in the brain. There is nothing up my sleeve.3
Artificial Intelligence 31
Computer Chess
To get a feel for what artificial intelligence is trying to do, let’s look at
a particular case history, the development of computer programs to play
games such as chess. Playing chess was one of the first tasks to be con-
sidered by researchers, because the chessboard is easy to represent inside
a computer, but winning the game is difficult. The game has the reputa-
tion of requiring brains to play, and very few people ever get good at it.
There is no obvious algorithm for playing chess well, so it appears to be a
good domain for studying more general sorts of reasoning, or it appeared
that way at first. In the 1950s, Allen Newell, Clifford Shaw, and Herbert
Simon wrote some papers (Newell et al. 1958) about a program they de-
signed and partially implemented, which used general-purpose symbolic
structures for representing aspects of chess positions. However, over the
years chess programs have become more and more specialized, so that
now there is no pretense that what they do resembles human thinking, at
least not in any direct way.
Almost all chess programs work along the lines suggested in early
papers by Claude Shannon and Alan Turing (Shannon 1950a, 1950b;
Turing 1953), which build on work in game theory by von Neumann and
Morgenstern (1944). A key feature of chess is that both players know ev-
erything about the current state of play except each other’s plans. In card
games a player is usually ignorant of exactly which cards the opponent
has, which adds a dimension we don’t need to worry about in chess, where
both players can see every piece. We can use this feature to construct a
simple representation of every possible continuation from a chess posi-
tion. Imagine drawing a picture of the current position at the top margin
of an enormous blackboard. Assuming it is your turn to move, you can
then draw all the possible positions that could result from that move, and
join them to the original position by lines. Now below each such position
draw a picture of every position that can be reached by the opponent’s
next move. Continue with your move, and keep going until every possible
position reachable from the original position has been drawn. (This had
better be a really big blackboard.) The process will go on for a long time,
but not forever, because at any position there are only a finite number of
available moves and every chess game must come to an end. The resulting
32 Chapter 2
1
Player
1 0 0 1 1
Opponent
P
1 1 1 0 1 0 1 1 1 1 1 1
Player
1 1 1 1 0 1
Figure 2.1
Game tree
drawing is called a game tree (figure 2.1). Each position is indicated by a

square. Pretend the square contains a chess position reached at a certain
point in a game; I’ll explain the numbers shortly. The root of the tree is
at the top and the leaves at the bottom. It looks more like a tree if you
turn the picture upside down, but it’s called a tree even though it’s usually
drawn with the root at the top. The root is the original position, and the
leaves are the final positions, where the game has ended and no further
moves are possible.
We now label the leaves with 0, 1, or −1, depending on whether the
game is a draw, a win for you, or a win for your opponent, respectively.
To describe what happens next I need to introduce another definition. Let
the children of a position B refer to the positions reachable by making
one move in B. (I apologize for the mixing of a family-tree metaphor with
the botanical-tree metaphor.) Now pick a position P that has only leaves
as its children, that is, a position in which no matter what move is made
the game is over. Suppose that at that position it is your turn to move.
If there is a child position labeled 1 (meaning “you win”), then you can
win in position P by making the corresponding move. Hence P can be
labeled 1 as well. If no child is winning, then if there is a child labeled 0
(draw), you can at least draw, so P should be labeled 0. Otherwise, no
matter what you do you will lose, so P is lost, and should be labeled −1.
Actually, this is impossible in chess; the last person to move can’t lose. We
Artificial Intelligence 33
include this possibility for completeness, because we’re going to see the
same pattern elsewhere in the tree: at any position where it’s your turn
to move, the position should be labeled 1 if any child is labeled 1; 0 if
no child is labeled 1 but some child is labeled 0; and −1 if every child
is labeled −1. In other words, at every position where it’s your turn to
move, you should label it with the maximum of the labels attached to its
children. At positions where it’s the opponent’s turn to move, you should,
by a similar argument, place a label equal to the minimum of the labels
attached to the children.
If we continue in this way we will eventually label every position. The
original position will be labeled 1 if you can force a win, −1 if your
opponent can, and 0 if neither of you can. This claim may not seem
obvious, so let’s consider in more detail the strategy you can follow if the
label is 1. There must be some child position labeled 1, so pick one and
make the move that gets you there. At this position it is the opponent’s turn
to move, so if it’s labeled 1 then every child position must be labeled 1. So
no matter what the opponent does, you will be in a situation like the one
you started with: you will have a position with at least one child position
labeled 1. You will be able to make the corresponding move, and then
wait for the opponent to pick among unappetizing choices. No matter
what happens, you’ll always have a position labeled 1 and can therefore
always force a situation where the opponent must move from a position
labeled 1. Sooner or later you’ll get to a position one step from a leaf,
where you have a move that wins.
This idea was first made explicit by von Neumann and Morgenstern.
It sounds like a surefire recipe for winning at chess (and any other board
game with complete information about the state of play and no random
element such as dice or roulette wheels). Of course, no person can play
the game this way, because the equipment (an enormous blackboard) is
too unwieldy, and it would take way too long to work out the whole game
tree. But computers can get around these problems, right? They can use
some kind of database to represent the tree of positions, and use their
awesome speed to construct the tree and label all the positions.
No, they can’t. There is no problem representing the board positions.
We can use a byte to represent the state of a square, and 64 bytes to
represent the board.4 It’s not hard to compute the set of all possible moves
Other documents randomly have
different content
“It will be, but I don’t want to be idle.”
“Perhaps you are right. I will be on the look-out for you, and if I
find something more congenial I will inform you at once.”
Ben did find it slow work following his old business. He missed the
nightly applause, and the pleasant consciousness that he was
earning three times his necessary expenses.
But it was agreeable to think that he had some money in the

savings bank to fall back upon. Mr. Snodgrass urged him to use a
part of it, and even hinted that he should be glad to borrow ten
dollars, but Ben knew the novelist too well to feel that it would be a
safe investment.
It was about this time that a young man of twenty took an

unoccupied room at Mrs. Robinson’s house. He professed to be
earning twelve dollars a week in a counting house on Pearl Street as
assistant bookkeeper.
He was dressed in quite a pretentious style, and had a large stock

of flashy neckties. He had seen Ben on the stage at the People’s
Theater, and this led him to cultivate his acquaintance.
“You must have saved up a lot of money while you were acting,”
he said one day.
“A little, Mr. Grayson,” Ben admitted. “I have sixty dollars in the

Union Dime Savings Bank.”
“Humph! I don’t think much of savings banks.”
“What do you consider better?”
“I’ve got a friend doing business in Wall Street. Give it to me and

I’ll get him to buy a few shares of stock for you on a margin.”
“I think I would prefer to leave the money where it is.”
“All you will get there is a paltry four per cent.”
“The interest doesn’t amount to much, to be sure, but the money

is safe.”
George Grayson did not press the matter, but invited Ben out to
play pool at a place on Sixth Avenue.
“I never played the game,” said Ben.
“No matter; you’ll pick it up directly.”
“But I can’t afford to play it.”
“It only costs five cents.”
Knowing nothing of the game, Ben accepted this as true, and

curiosity led him to accompany his new acquaintance.
“I’ll coach you,” said Grayson.
They made choice of tables and commenced playing. Two other

young men, friends of Grayson, joined them.
The game occupied only about ten minutes. Ben succeeded in

pocketing one ball, and naturally stood last.
“Well, Ben, you’re beaten!” said Grayson. “The rule is to pay at the
end of each game.”
Ben took a nickel from his pocket and handed it to the attendant.
“What’s this for?” he asked.
“My friend told me that the game cost five cents.”

“Yes, five cents a cue.”
“Well, I only used one cue.”
“Come, young feller, no fooling! There were four played, and as

you were beaten you pay for the whole. Fifteen cents more.”
“That’s straight, Ben,” said Grayson.
“But you told me it would be only five cents.”
“Don’t argue the matter or all the boys will be laughing at you.”
Ben saw that he had been deceived, but took the advice of his
tricky companion.
“Now for another game!” said Grayson.
“You can count me out,” said Ben.
“What! Does it worry you so much to get beaten?” sneered his

companion.
“No, but I can’t afford to play.”
“You say that with sixty dollars in the bank!”
“I shouldn’t have it there long, if I played pool every evening.”
Grayson whispered some words in the ear of the next player and
he laughed rather derisively. Ben thought he caught the word
“miser.” At any rate he had had enough of pool playing, and soon
after left the hall.
He did not feel very cordial towards Grayson, but the latter made
friendly advances, and as he said no more about pool Ben gradually
admitted him to companionship.
Two or three times he asked Grayson the street and number of
the business firm which employed him, but only received an evasive
answer.
There came a dull time, so far as news was concerned, and Ben
found that the sale of papers fell off, so that he was no longer able
to earn seventy-five cents a day. This was the very smallest sum on
which he could live even with the strictest economy, and, reluctant
as he was to do it, he found that he must draw some money from
the savings bank.
During Ben’s career as an actor he had increased his stock of

underclothing, and, having only a gripsack, had invested in a small
sized trunk, which he found much more convenient.
In the tray of this trunk he had placed his savings bank book. He
opened the trunk and looked confidently for the book. But to his
surprise it was not to be found.
“Perhaps I put it in the lower part of the trunk,” he said to himself,

though he felt sure it had been in the tray. He continued his search,
but it proved to be vain.
Ben sat down before the open trunk and tried to recall all the
incidents connected with the last time of opening it. But the more he
thought the more puzzled he became.
Then it flashed upon him that the book might have been stolen.
He went at once to the room of his literary friend, Sylvanus
Snodgrass, and told him of his discovery.
“It has been stolen!” said Sylvanus instantly. “I introduced an

incident like this into my last serial story for the Bugle.”
“But who could have stolen it?” asked Ben, perplexed. “The
servant wouldn’t do it I am sure.”
“No, she is an honest Swedish girl. She wouldn’t be capable of it.”
“I agree with you, but some one must have taken it from the
trunk.”
“Of course! Let me think,” and the novelist leaned his head on his
hand and wrinkled up his forehead in the throes of mental
speculation.
“I have it!” he exclaimed suddenly.
“What! the bank book!”
“No; I begin to understand the mystery.”
Ben regarded him patiently. He knew that Sylvanus would soon

impart to him his suspicions.
“Last evening I took a walk with Grayson,” said the novelist. “I

noticed a new and handsome ring upon his finger. I asked him
where it came from. He said, ‘It was given me by a friend,’ but he
spoke hesitatingly. ‘It must have cost as much as ten dollars,’ I said.
‘Fifteen!’ he answered. ‘That is, I saw a ring like it in a shop window
for fifteen dollars.’
“Depend upon it, Ben, that ring was bought with your money, and
George Grayson opened your trunk and stole your bank book.”
“I don’t like to think so,” said Ben, troubled.
“I feel sure of it.”
“What would you advise me to do?”
“Go to the bank, give notice of your loss, and find out whether
any money has been drawn from the bank on your account.”
This seemed to be sensible advice, and Ben acted upon it the next
morning. Mr. Snodgrass accompanied him to the banking house at
the junction of Broadway and Sixth Avenue at Thirty-second Street.
Ben went up to one of the windows—the one where the paying

teller pays over the money—and gave notice of the loss of his book
—giving the number.
“When did you see the book last?” asked the official.
“Wednesday.”
“And to-day is Friday.”
“I should like to know if any money has been drawn on it?” asked
Ben.
The books were referred to, and the answer came, “Forty dollars
were drawn day before yesterday. Didn’t you sign the order?”
“No.”
The receipt was looked up, and the signature examined.
“Isn’t that your signature?”
“No, sir.”
“Then it must have been imitated. The resemblance is very close.”
Ben was forced to admit that it was.
At this moment Sylvanus, who had been looking out of the front
window, came up and said hurriedly, “Grayson is coming, and he has
a bank book in his hand.”
CHAPTER XXII.
GEORGE GRAYSON COMES TO GRIEF.
Ben quickly informed the paying teller of the new arrival, and he
and Snodgrass took a position on the left hand side of the main
entrance, where there was a chance of their escaping observation.
Grayson entered the bank with a jaunty step and walked up to the
window of the paying teller. He did not stop to write a check for the
sum he wished to withdraw, the check being already drawn and
inclosed in the book.
According to custom he passed in the book and waited for the

money.
The teller eyed him attentively, but did not do so in a manner to

excite suspicion.
Opening the book he said, “You drew forty dollars yesterday.”
“Yes,” answered Grayson composedly, “I thought that would be all

I should need, but I am making a little investment, and have drawn
fifteen dollars more.”
“Very well.”
The paying teller took the book and went to the ledger, ostensibly
to compare the signature with that on the check. At the same time
he whispered to a young employee, who immediately left the bank
to summon a policeman.
George Grayson kept his place at the window, looking more cool
and unconcerned than he would had he known what was going on.
Somehow there seemed to be a good deal of delay in getting the
money. The paying teller occupied a considerable time in turning
over the pages of the ledger.
Apparently he had selected the wrong book, for he then went to

another and began to examine that. Now and then he turned his
eyes to the front entrance.
Grayson suspected nothing at first, but after a while it occurred to

him to wonder why he had to wait so long, especially as two other
persons had come into the bank and were standing behind him
waiting for their turn.
Thus far he had not discovered Ben and his friend the novelist, but
chancing to turn his head after a time he caught sight of the two.
Then he understood.
“I must bolt,” he said to himself, and leaving his place he hurried

to the door. But he met the boy coming up the steps with a
policeman.
The boy spoke a word to the officer, who sprang forward and
grasped Grayson by the arm.
“What do you mean?” demanded Grayson haughtily, assuming a

look of virtuous innocence.
“Come back into the bank with me,” said the policeman, “and you
will learn.”
“I am in great haste,” replied Grayson, trying to shake off the

officer’s hand.
“Not so fast, my friend,” said the officer.

“This is an outrage,” blustered Grayson. “I have committed no
wrong.”
“In that case you won’t be detained long. Come in.”
Grayson, much against his will, had to obey.
By this time the bank official had come out in front of the
partition.
“This man has forged a draft on the account of another person,”

he said.
“Is the owner of the book here?”
The teller indicated Ben.
“This is a conspiracy,” blustered Grayson, but he was slow in

meeting Ben’s eye.
“Is your name Ben Bruce?” demanded the teller.
“Ye-es,” answered Grayson in a tone of hesitation.
“That’s a lie,” broke in Sylvanus. “He has always represented

himself as George Grayson.”
“I will take him to the station house,” said the officer, “and depend
on you to appear as prosecutor.”
These words were addressed to Ben.
Grayson’s face changed. He felt that he was in a tight place.
“Look here, Bruce,” he said insinuatingly, “can’t we fix this thing?

I’ve got a ring here that I paid twelve dollars for, and I have a few
dollars in my pocket. I’ll give you them, and agree to pay the
balance as soon as possible if you’ll let me go.”
“Shall I be allowed to do this?” asked Ben, who felt disposed to be
lenient.
“It is too late,” said the officer. “I will trouble you to come to the
station-house with me to make known the charge.”
Ben did so, and matters took their course. After some delay he
received back the savings-bank book with the ring and about ten
dollars. George Grayson was sentenced to a term of imprisonment.
Ben pitied him and would gladly have spared him this, but the law
was inexorable.
Grayson tried to shake off the officer’s hand.
“Not so fast, my friend,” said the officer.—
Page 175.
Ben Bruce.
CHAPTER XXIII.
A STRANGE ADVENTURE.
The summer passed slowly. Business was unusually dull even for
this time of the year, and Ben’s earnings were proportionately small.
Week by week he was obliged to draw from his fund in the savings
bank until he had less than five dollars to his credit there.
He had not written to his mother or to Albert Graham for a

considerable time, not having any good news to communicate.
How was he coming out? That was the question which he

anxiously asked himself without obtaining any satisfactory answer.
He began to think that he might feel compelled to pawn his watch
once more, with a very remote chance of redeeming it.
It was about this time that he had a surprising adventure. He was

selling papers at ten o’clock in the morning when suddenly a lady,
handsomely dressed, stopped opposite him and regarded him
attentively.
“Will you have a morning paper, ma’am?” asked Ben.
“Yes, I will buy all you have,” was the unexpected answer.
“There are twenty-five cents’ worth,” said Ben, counting them

over. It occurred to him that the lady was a philanthropist, who took
this way of helping him.
“Here is a dollar. Never mind the change.”
“Thank you. You are very kind. Will you take the papers, or shall I
carry them for you?”
“Never mind! Leave them in that doorway, or give them to some
other newsboy. I want to employ you for a time.”
Tom Hooper happened to be passing, and Ben, considerably to

Tom’s surprise, went up to him and handed him his papers.
“You can have these papers, Tom. They are a present from this
lady.”
Tom accepted them with pleasure, for he felt sure of disposing of

at least a part of them.
“Now,” said Ben. “I am at your service, madam.”
“Please call a cab.”
Ben complied with the lady’s request.
“Help me in,” she went on, “and get in yourself.”
As the coachman closed the door she said, “Drive to the Fifth
Avenue Hotel.”
The mysterious lady sat on the back seat and signed to Ben to
place himself opposite to her.
It began to look queer to Ben. If the lady intended to employ him,

it seemed odd that she should treat him on such equal terms.
However, Ben was discreet, and feeling that he would know in time
forebore to ask questions.
The cab stopped at the side door, or lady’s entrance.
“You may follow me,” said the lady as she paid and dismissed the
cab driver.
Ben followed the lady up-stairs to a room on the second floor.

The lady opened the door and entered.
“Now sit down,” she said, “and we will have a little conversation.”
Ben seated himself in a large arm-chair and waited for

developments. The lady sat down opposite him.
“Are you a good actor?” she asked.
“I acted a few weeks at the People’s Theater on the Bowery,”

answered Ben.
It was the lady’s turn to look surprised.
“Is it possible?” she exclaimed. “You—a newsboy now—have been

an actor?”
“Yes, madam.”
“I am glad of it. But how do you happen, after such an

engagement, to be reduced to selling papers in the street?”
“The play had to give place to another, and I lost my engagement.

I had to live and took up selling papers for want of something
better.”
“I want you to play a part in a drama of real life.”
“At what theater, madam?”
“At no theater. You are to personate my son. You are to call me

mother, and your name will be Edwin Harcourt.”
“But, madam, will any harm come of it?”
“None whatever. You will be aiding the cause of justice.”
“Then I am willing.”
“I have taken the adjoining bedroom for you: go in and put on the
suit of clothes you will find on the bed. Brush your hair carefully, and
try to do me credit.”
Ben smiled.
“I will try to do so,” he said.
“Of course I shall see that you are well paid.”
“I have no doubt on that point. But——”
“Ask no more questions now. Dress yourself quickly, as we have a

call to make.”
CHAPTER XXIV.
BEN PLAYS A PART.
The suit which Ben had put on was of fine imported cloth, and
evidently expensive.
It fitted marvelously well as Ben could see for himself. It was

better than the suit he had purchased in Boston, and which was now
half worn.
When he was dressed he stepped into the adjoining room.
Mrs. Harcourt regarded him with evident satisfaction.
“The suit fits you admirably,” she said. “It is very becoming.”
“That is what I don’t understand,” said Ben. “How could you select
a suit for me before you knew me?”
The lady smiled.
“Suppose I say that I looked for a boy to match the suit? It shows
that I have a correct eye, does it not?”
“Yes, madam.”
Ben had still to submit to a critical inspection.
“Your shoes need polishing,” the lady said. “Go down below and
get a shine. You will find a bootblack in the lower part of the hotel.
Have you change?”
“Yes, madam.”
“Say ‘yes, mother.’ It is as well that you should get used to the
name.”
“But I have a mother. Won’t it do as well to call you aunt?”
“No; bear in mind that you are acting. On the stage people are
husbands and wives, mothers and sons, for the occasion only.”
“All right. I will look upon you as a stage mother then.”
“Yes, but the illusion must be kept—during our engagement.”
“I will remember.”
“Now go down-stairs and come back with better looking shoes.”
Ben went below and had his shoes blacked. When the operation
was ended he went up-stairs.
He found Mrs. Harcourt dressed for the street.
“Ring the bell, Edwin,” she said, “or rather go down yourself and
order a cab.”
Ben started a little at the unfamiliar name. Then he smiled as he

reflected that he was playing a part.
“All right, mother,” he said.
“Good, Edwin. I see you are working into your part.”
In five minutes they were rattling up Fifth Avenue in a cab. The

driver, who had his instructions, turned into East Fifty-seventh
Street, and paused in front of a handsome brown stone house.
“Is Mr. Anderson in?” asked the lady.

“Yes, ma’am, but he isn’t feeling well. I don’t know if he can see
you.”
“Nonsense!” exclaimed the lady sharply. “Tell him his niece, Maria
Harcourt, has just arrived from Europe and wishes to see him.”
“Very well, ma’am,” said the girl, overawed, “I’ll tell him.”
She went up-stairs and quickly returned, saying, “He will see you.”
“Of course he will. Edwin, you may stay here until I return, unless
you are sent for.”
“All right, mother.”
Ben was about to omit the designation “Mother,” but a quick

glance from Mrs. Harcourt showed that she expected him to use it.
We will follow Mrs. Harcourt up-stairs.
In a room fitted up as a library, sat, or rather reclined, in an easy-

chair, an old man evidently quite feeble. He essayed to rise, but Mrs.
Harcourt moving forward rapidly prevented him.
“No, Uncle Henry,” she said, “don’t get up.”
She bent forward and just touched his chin with her lips.
“I am glad to see you, Mamie,” he said. “Have you just returned

from Europe?”
“Yes, uncle.”
“Have you brought the boy with you?”
“Yes, uncle; he is down-stairs.”

“Didn’t I hear that he was sick with typhoid fever somewhere in—
in——”
“Geneva. Yes, uncle, my poor Edwin was very sick, but fortunately
he recovered and is now the picture of health.”
“Basil was under the impression that he was dead.”
“It was for the interest of Basil to report so, Uncle Henry.”
“I don’t think he had any reason to misrepresent, Maria.”
“If Edwin should die, Basil’s income would be increased by five

thousand dollars, and the Mordaunts would profit also.”
“True, but——”
“Well, we won’t discuss the matter. I will try to think as well of him
as I can. The fact is, however, that Edwin is alive and well. If you will
give me an order on your bankers for the last six months’ income I
shall be glad.”
“Can I not see the boy?”
“Certainly, Uncle Henry, but promise me not to keep him long, as I

have to take him to get some clothes.”
“Very well, Maria. I only wish to see him. I don’t feel well enough
for a prolonged interview.”
“First, then, Uncle Henry, write me a letter to your bankers, asking

them to pay the boy’s income now due, and you may as well tell
them to remit regularly without further instructions, as I don’t want
to trouble you every time.”
“Very well, Maria.”

When this business was over, Mrs. Harcourt went down-stairs,
where she found Ben waiting patiently for her return.
“Are you tired of waiting, Edwin?” she said playfully.
“Oh no.”
“No, what?”
“Mother,” said Ben a little awkwardly. He had not yet accustomed

himself to his new part.
“Now, Edwin, listen attentively to what I say. I am going to take

you up-stairs to see an old gentleman, an uncle of mine, in fact, who
is, between ourselves, rather feeble in intellect. Whatever he asks
you answer in such a way as to humor him, otherwise he will
become violent. For instance he may ask you about traveling in
Europe, perhaps about being sick. Fall into his humor, and don’t let
him suspect that you think him queer.”
“All right—mother.”
“Remember, I trust to your discretion.”
“I will do as well as I can. What is the name of the gentleman?”
“Mr. Anderson. I call him my uncle Henry. Now follow me.”
Ben followed Mrs. Harcourt up the broad staircase, and into the
presence of the frail old gentleman. Mr. Anderson looked up as they
entered the room and signed for Ben to approach.
“Come here, my boy,” he said. “I have but little eyesight left. I

need to have you near me.”
Ben approached and stood beside the easy-chair.

“Why, you are looking fine,” said the old man in some surprise.
“You don’t look as if you had been sick.”
“No, sir.”
“You feel perfectly well, then, in spite of your recent sickness?”
“Yes, sir.”
“I am very glad. And you enjoyed traveling?”
“Yes, sir.”
“You are very well grown. I did not expect to find you so large.”
“He has grown rapidly, Uncle Henry,” said Mrs. Harcourt.
“Basil would be glad to see you. He thought you were dead!”
“He looks very much alive, doesn’t he, Uncle Henry?”
“Yes, yes. And so you enjoyed Europe, did you, Edwin?”
“Yes, sir.”
Ben felt a little awkward as he said this, but he remembered that

the old gentleman was feeble-minded and felt that he was justified
in humoring the delusion.
“Won’t you stay to lunch, Maria?” asked Mr. Anderson.
“I am sorry we can’t do so, uncle, but Edwin and I have some

calls to make.”
“Where are you staying?”
“At the Fifth Avenue Hotel.”

“I should be glad to have you stay here. The house is large
enough.”
“I wouldn’t for the world interfere with your quiet ways, uncle.
Remember that you are an invalid, and need to have things quiet
around you. Edwin is a boy of a lively temperament, and he will feel
more comfortable at the hotel.”
“No doubt you are right, Maria. Shall you stay long in the city?”
“My plans are not formed yet, Uncle Henry, but I will apprise you
of them when I have made up my mind. And now I must really say
good morning.”
“Good morning, Maria. Good morning, Edwin.”
Ben shook the old man’s hand, and followed Mrs. Harcourt out of
the room.
“Well?” said the lady interrogatively. “What do you think of him?”
“He didn’t seem to me feeble-minded.”
“Probably not. He was unduly quiet. He has strange delusions,

however. Last night he fancied himself to be Christopher Columbus. I
don’t know if he has got over it yet.”
“He seems to be a very pleasant old man.”
“Yes, he was in a pleasant mood. Perhaps when you next see him
it may be different. Now let us go to the carriage. I am going to Wall
Street.”
CHAPTER XXV.
THE MYSTERY DEEPENS.
The cab stopped in front of a handsome office building on Wall

Street.
Mrs. Harcourt dismissed it.
“I shall have some other calls to make, Edwin,” she said, “and
won’t take a carriage till I am through. Now let us go up-stairs.
“Remember,” she said, as they were ascending the stairway, “we

are for the present mother and son.”
“I’ll remember.”
“Should anything be said to you answer as briefly as possible.”
“Very well.”
Ben felt puzzled. He did not at all comprehend what was going on,
but concluded that it was all “in the play.”
Mrs. Harcourt opened the door of a large office and entered.

Several clerks were working behind a counter or partitioned wall,
which separated the inner from the outer office.
A young man came forward and said politely, “What can I do for
you, madam?”
“Is Mr. Stormleigh in?”
“Yes, madam.”
“Give him my card.”
“Certainly.”
He returned presently with an invitation to Mrs. Harcourt to follow

him.
“Stay here, Edwin, till I return or send for you,” she said, and Ben
seated himself in a chair near the window.
In the inner office sat a pleasant-looking man of fifty.
“I am glad to see you, Mrs. Harcourt,” he said rising. “Let me see,

how long is it since we met?”
“Five years.”
“Indeed. You look as young as ever.”
“I am afraid you are a flatterer, Mr. Stormleigh.”
“Your son——” began Mr. Stormleigh in a tone of hesitation.
“My son is in the outer office.”
“What? I heard a rumor that he was dead.”
“And that was probably the reason you did not send me the last
quarterly income due to me as his guardian?”
“Yes. Of course, if he were dead, it would no longer be due.”
“Thank Providence, the dear boy is in the best of health.”
“I am heartily glad of it. And you brought him with you?”
“Yes.”
“May I see him?”
“I will call him.”
Ben was summoned, and Mr. Stormleigh regarded him with

evident approval.
“Really, Mrs. Harcourt, you have reason to feel proud of such a

fine-looking boy.”
“Have I not? Edwin, shake hands with Mr. Stormleigh. He is an old

friend of mine, besides being your trustee.”
“Well, my boy, how old are you?”
“Sixteen.”
Mrs. Harcourt looked relieved. The age tallied exactly.
“And now, Edwin,” said the lady, “I won’t detain you. You may go
down at once to the Fifth Avenue Hotel and await me there. Or, if
you want two hours for yourself, meet me at the end of that time at
my room. I am not sure whether you have any money. Here is a ten-
dollar bill.”
“Thank you—mother.”
Mrs. Harcourt remained fifteen minutes longer, receiving a large

check from Mr. Stormleigh, which she deposited to her credit in the
Park National Bank.
“What are your plans, my dear Mrs. Harcourt?” asked the banker.
“Shall you remain in America?”
“I am not sure. I may go back to Europe, taking Edwin with me.”
“Is he at school?”
“I shall probably place him at school, but my plans are not fully
formed.”
“He does not appear to have any resemblance to the late Mr.
Harcourt.”
“Boys often change in looks as they get older.”
“True.”
“And you have not seen Edwin for several years.”
“And then I only had a glimpse of him.”
“Well, I must really go. I have no doubt you have important

business, so that you will be glad to get rid of me.”
“I confess that I am quite busy this morning. Call again, however,

when you have an opportunity.”
Meanwhile Ben went down-stairs, more and more mystified. He

thought Mrs. Harcourt a very mysterious character.
She had treated him handsomely, however. He had on an elegant

suit and a ten-dollar bill in his pocket. His life seemed to be entirely
changed.
In the morning he had been a Bowery newsboy; now he was

boarding at the Fifth Avenue Hotel. That reminded him that he must
give notice to his landlady that he would not sleep in his room at
present.
“But how long will this last?” he asked himself.
If only a week he might as well keep the room, as the price was
so small, and he was in funds. Having no urgent business, he
decided to walk up Broadway.

62332

Uploaded by

Copyright:

Available Formats

62332

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

62332

Uploaded by

Copyright:

Available Formats

Download the full version of the ebook at ebookfinal.

Mind and Mechanism 1st Edition Drew V. Mcdermott

Download more ebook instantly today at https://ebookfinal.com

Intelligence Beyond Thought Exploding the Mechanism of

Washington Merry Go Round The Drew Pearson Diaries 1960

Endocrine Secrets 5th Edition Michael T. Mcdermott Md

Motion Simulation and Mechanism Design with COSMOSMotion

Arsenic Contamination of Groundwater Mechanism Analysis

Molecular Hematology 5th Edition Drew Provan

Hollywood Film 1963 1976 1st Edition Drew Casper

Atherosclerosis Experimental Methods and Protocols 1st

Mind and Mechanism

This book was set in Sabon by Interactive Composition Corporation.

Printed on recycled paper and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data

BF161 .M29 2001

1 The Problem of Phenomenal Consciousness 1

• There are two computational paradigms to choose from: symbolic com-

I must thank David Chalmers for writing a great, if totally misguided,

Mind as Self-Fulfilling Description

information has a “causative” rather than a purely “descriptive” char-

Dualism and Its Discontents

wave travels at a speed between a few hundred centimeters per second

However, such changes are reflected in changes of synapses’ sensitivity to

that have nothing to do with voltages. The representations feed further

Nondualist Explanations of Consciousness

Some theorists (notably Bernard Baars 1988, 1996) focus on explaining

A process theory is one that explains experience in terms of things the

(or correlated with?) stable activation patterns in networks of neurons.

sensory information must itself be the object of perception. First-order

He says this model is pure speculation, so I don’t suppose he would de-

observation that a computational theory of mind can’t be correct because

Several other nonphilosophers have proposed second-order theories of

drawing is called a game tree (figure 2.1). Each position is indicated by a

But it was agreeable to think that he had some money in the

It was about this time that a young man of twenty took an

He was dressed in quite a pretentious style, and had a large stock

“A little, Mr. Grayson,” Ben admitted. “I have sixty dollars in the

“Humph! I don’t think much of savings banks.”

“What do you consider better?”

“I’ve got a friend doing business in Wall Street. Give it to me and

“All you will get there is a paltry four per cent.”

“The interest doesn’t amount to much, to be sure, but the money

“I never played the game,” said Ben.

“No matter; you’ll pick it up directly.”

“But I can’t afford to play it.”

“It only costs five cents.”

Knowing nothing of the game, Ben accepted this as true, and

“I’ll coach you,” said Grayson.

They made choice of tables and commenced playing. Two other

The game occupied only about ten minutes. Ben succeeded in

“What’s this for?” he asked.

“My friend told me that the game cost five cents.”

“Well, I only used one cue.”

“Come, young feller, no fooling! There were four played, and as

“That’s straight, Ben,” said Grayson.

“But you told me it would be only five cents.”

“Now for another game!” said Grayson.

“You can count me out,” said Ben.

“What! Does it worry you so much to get beaten?” sneered his