Académique Documents
Professionnel Documents
Culture Documents
Forthcoming in Philosophy:
The Continuum Companion to Aesthetics, edited by Anna Christina Ribeiro
The Continuum Companion to Berkeley, edited by Bertil Belfrage and Richard
Brook
The Continuum Companion to Epistemology, edited by Andrew Cullison
The Continuum Companion to Ethics, edited by Christian Miller
The Continuum Companion to Existentialism, edited by Jack Reynolds, Felicity
Joseph and Ashley Woodward
The Continuum Companion to Hegel, edited by Allegra de Laurentiis and Jerey
Edwards
The Continuum Companion to Hobbes, edited by S. A. Lloyd
The Continuum Companion to Hume, edited by Alan Bailey and Dan OBrien
The Continuum Companion to Kant, edited by Gary Banham, Nigel Hems and
Dennis Schulting
The Continuum Companion to Leibniz, edited by Brendan Look
The Continuum Companion to Metaphysics, edited by Robert Barnard and Neil
A. Manson
The Continuum Companion to Political Philosophy, edited by Andrew Fiala and
Ma Matravers
The Continuum Companion to Plato, edited by Gerald A. Press
The Continuum Companion to Pragmatism, edited by Sami Pihlstrm
The Continuum Companion to Socrates, edited by John Bussanich and Nicholas
D. Smith
The Continuum Companion to Spinoza, edited by Wiep van Bunge
The Continuum Companion to Philosophy of Language, edited by Manuel
Garcia-Carpintero and Max Kolbel
The Continuum Companion to the Philosophy of Science, edited by Steven French
and Juha Saatsi
The Continuum
Companion to
Philosophy of Mind
Edited by
James Garvey
Continuum International Publishing Group
The Tower Building 80 Maiden Lane
11 York Road Suite 704
London SE1 7NX New York, NY 10038
www.continuumbooks.com
ISBN:HB:0826431887
978-0-8264-3188-2
BD418.3.C6565 2011
128'.2dc22 2010036913
Acknowledgements ix
Contributors xi
How to Use This Book xv
Introduction xix
vii
Contents
Glossary 280
Chronology 319
Research Resources 327
Notes 330
Bibliography 341
Index 375
viii
Acknowledgements
I relied on a large number of people for help in pu ing this volume together.
First and most importantly, I am very grateful to all of the contributors. Some
provided advice and read bits of the manuscript, and their suggestions always
resulted in improvements. I am particularly in the debt of those who saved
me and stepped in to do some last-minute writing. Each delivered good, solid
philosophy in record time. You know who you are. So thanks are owed to:
Fred Adams, Steve Beighley, Margaret A. Boden, Mark Cain, Neil Campbell,
Adam Ferner, Daniel D. Hu o, Dale Jacque e, E. J. Lowe, Tim Mawson,
Barbara Montero, Isabella Muzio, Paul Noordhof, Dan OBrien, Dimitris
Platchias, Ian Ravenscro , Georges Rey, Constantine Sandis, Sarah Sawyer,
Barry C. Smith, and Michael Wheeler.
David Avital, Carly Bareham, Tom Crick and Sarah Douglas at Continuum
are all very nearly equally excellent thank you all for your help and for
pu ing up with a lot. Thanks are also owed to comrades at Crisis, Kim Hastilow,
Ted Honderich, London Street Rescue, Justin Lynas, Anthony OHear, and my
associates at UCLU Jitsu.
I am particularly grateful to Judy Garvey for her unwavering support.
ix
This page intentionally left blank
Contributors
Adam Ferner works for the Royal Institute of Philosophy and is studying for a
Ph.D. in Philosophy at Birkbeck College, University of London. His thesis is on
animalism, animals, and artefacts.
xi
Contributors
(2000), Wi genstein and the End of Philosophy (2006) and Folk Psychological Narra-
tives (2008). He is also the editor of Narrative and Understanding Persons (2007),
Narrative and Folk Psychology (2009) and co-editor of Folk Psychology Re-Assessed
(2007). A special yearbook issue of Consciousness and Emotion, entitled Radical
Enactivism, which focuses on his philosophy of intentionality, phenomenology
and narrative, was published in 2006.
xii
Contributors
Isabella Muzio has a Ph.D. in philosophy from University College London and
teaches philosophy at The Open University. She is writing a book about our
knowledge of our own emotions.
xiii
Contributors
xiv
How to Use This Book
To help you find your way into this book, here is a short overview of each of
the main sections.
Introduction
Here you will find a very short take on the recent history of contemporary
philosophy of mind, its Cartesian roots, and a few words about recent con-
ceptual shi s, as well as a consideration of the general tone of the volume and
the choice of contents. The Introduction also contains a short summary of each
of the books main chapters, and this might serve as a springboard into the
rest of the book.
Current Research
Here you will find eleven original essays wri en by experts in the field. Not
only do they provide overviews of large sub-topics in the philosophy of mind,
the authors take a stand and argue for their own positions. Its this combination
which makes the essays of interest to researchers at dierent levels. Again, there
is an overview of each chapter in the Introduction.
This section, wri en by Paul Noordhof, aims to follow up from the previous
essays with reflection on cu ing-edge thinking in the philosophy of mind. This
xv
How to Use This Book
Glossary
This part of the volume includes short definitions and longer treatments
wri en by experts in various subfields of the philosophy of mind. Each entry
concludes with references to works which might be consulted for further infor-
mation. The aim of course is to provide a good resource for someone who
is reading this companion but might not be familiar with one or more of the
technical terms used by the authors. It also might help someone just ge ing on
with the study of the philosophy of mind who encounters a dicult word or
concept in a book or article some of the terms do not appear elsewhere in
this book. The glossary contains a number of clever distillations of dicult
concepts, so it might also be read independently, just for interest.
Chronology
This timeline is a reference for students and researchers who require speedy
access to a date or title as well as a bit of context for it. If you encounter an
unfamiliar philosopher, book or paper, have a scan of the chronology to get a
feel for where the topic fits in the history of mind. The entries up to 1950 contain
a sentence or two of explanation. Beyond that date we are too close in time to
be sure of the relevance, meaning or impact of a paper or book.
Research Resources
This section includes a list of resources on the Web, associations and research
centres, and periodicals devoted to the philosophy of mind anything of a
practical nature which might help a budding or seasoned researcher.
Bibliography
xvi
How to Use This Book
What Now?
xvii
This page intentionally left blank
Introduction
Renewal
xix
Introduction
xx
Introduction
Reasonable people can argue about what caused the revival in reflection on
mind, but few really doubt that something new is underway. Philosophers have
turned their a ention to a huge number of new or nearly new problems in the
past few decades, and by most lights they are still finding new topics, new
arguments, new theories and distinctions, maybe even new answers. Its no
surprise that there is a large number of books on the philosophy of mind
even many valuable companion volumes but this one really does aim to do
something at least a li le dierent from the rest. For one thing, all of the com-
panions in the Continuum Series have the slightly unusual aim of being of use
to people operating at many levels of enquiry. They arent just for beginning
students or old hands. The hope is that theyll serve as useful desk references
for people at many dierent stages, from classroom work to highly technical
reflection and research. To this end you will find a number of sections designed
to help beginning students hit the ground running as well as sections to assist
advanced researchers move on through further thoughts on various subjects.
For more details, have a look at the section called How to Use This Book.
However, what really marks this volume out against many other compan-
ions or supplements to the philosophy of mind, I think, is that it is a piece of
the philosophy of mind, not just a report on it. Thats been the intention from
the start. The contributors not only scout the relevant territory with a view
to ge ing the reader up to speed with who said what, they also pitch their
essays at fellow researchers. They say what they think and argue for their own
views or against claims made by others. The pleasing result is a companion
which doesnt just nod in agreement or politely show you around. Instead,
youll read philosophers ge ing on with the job, doing philosophy, arguing,
jostling, persuading, objecting, judging and generally trying to get the truth
into clear view. The result might not be consistent or coherent in fact Im
sure theres disagreement somewhere among all the contributors but their
thoughts are interesting and worth reading, and their work will certainly
stimulate further thinking.
xxi
Introduction
Topics
This book doesnt cover everything or even almost everything that ma ers in
the philosophy of mind. I used to cringe a bit when I read similar apologies for
an obvious truth in other books: if only we had more space to cover this and a
bit more time to consider that. No book could possibly include everything
which ma ers, so why make excuses? Never again. Editorial decisions are almost
always painful judgements one thing ruled in and twelve perfectly respect-
able and interesting and important things ruled out. In most cases Ive been
guided by something more than my own feeble thoughts and limited experience:
Ive made repeated use of a number of helpful and patient advisors who
know much more about the mind than I do. The topics covered in this book
from the main essays to the glossary entries all made it in at the expense of
something else based on the closest thing to a consensus I could get from
many people. Thats not to say that I hereby abdicate responsibility if the
mix of topics could have been be er, thats entirely my fault. I also gave the
contributors a free hand to approach topics more or less as they liked, given
the general constraints of the series, and maybe firmer editorial guidance
would have resulted in a more comprehensive or balanced volume. Then again,
giving experts the opportunity to scout the territory as they see it has many
recommendations too.
I also had to consider how the contributions might hang together, given who
agreed to write about what, and that means that some topics could only appear
within others, even though they might deserve star billing alongside a dierent
mix of papers. So, for example, theres nothing specific on qualia or the first
person or intentionality, but each one of these important subjects turns up
again and again in dierent papers. I also thought it might be good to consider
not just large, abstract questions in the philosophy of mind but also some very
specific, narrow problems. The hope is to convey something of the sweep of the
philosophy of mind as it is really practiced. To give you a feel for the topics
under consideration, well now glance at each of the main papers.
There are eleven central essays in this book. Some take up broad topics like the
nature of consciousness and the mark of the mental. Others deal with particular
theories of mind, such as physicalism and dualism. Still others examine specific
sub-topics such as mental causation and personal identity. The central essays
are bookended by two more general pieces. The first, by Ian Ravenscro , sets
the tone with an introductory overview of the philosophy of mind. The last, by
Paul Noordhof, considers some of the possible future directions the philosophy
xxii
Introduction
of mind might take. Well begin with Problems and Questions in the Philo-
sophy of Mind by Ian Ravenscro and briefly summarize the rest.
Ian Ravenscroft
Ravenscro identifies four broad areas of research in the philosophy of mind as
it is practiced today: metaphysical issues concerning the relation between the
mental and the physical, epistemological questions about our knowledge of our
own minds and the minds of others, themes associated with the influence of the
behavioural and cognitive sciences, and methodological issues concerning the
right approach to the study of mind. His main focus is the philosophy of minds
main focus: the mind-body problem. He works through various theories of
mind: substance dualism, reductive and non-reductive physicalism, the super-
venience relation, eliminativism, and instrumentalism. Closely connected to
the question of the relationship between the mind and body is a set of problems
having to do with mental representation. Ravenscro briefly takes up the
representational theory of mind as well as several theories of content. Next, he
considers mental causation and the specific sense in which certain problems
arise for physicalist theories of mind along with some possible solutions to
them. Finally, he considers various answers to what might be contemporary
philosophy of minds understanding of its own central question: how can
phenomenal consciousness exist in a purely physical universe? Jacksons
knowledge argument and several replies to it are considered.
In the end he argues that even if we havent go en past certain apparently
intractable problems, at least we have a more sophisticated set of tools and con-
cepts to help us understand them than at any other time in our history. Maybe we
have a good grip on what it is that we dont know, and thats a kind of progress.
Consciousness
Daniel D. Hutto
Hu o begins by admi ing that there is no clean, clear and neutral account of
what we mean by consciousness. He goes on to do what many others do: pin down
his topic with examples, alongside certain nearby expressions which seem to
strike a chord: what its like, for example. He also lists some of the characteristic
features of conscious experience mentioned in this connection, such as phenom-
enality, intentionality, subjectivity, unity, temporal extension and self-awareness.
He then takes up reductively naturalistic explanatory frameworks which by
turns equate consciousness to something else brain states, functional states,
xxiii
Introduction
and so on. Working through the main arguments for and against non-reductive
naturalism, Hu o se les on the so-called hard problem of consciousness: how
could this functional state or that representational property ever give rise to
conscious experience? The best we can do, it is sometimes concluded, is hope
for a specification of the non-reductive relations which hold between conscious
and other properties. The main replies to this thought are considered and judged.
In the final section, Rethinking Metaphors of Mind, Hu o considers a new,
or anyway dierent sort, of reply to the hard problem and the explanatory gap
between mind and world: the a empt to explain away the dierences which get
in the way of the explanatory identities we posit. He takes up Denne s analysis
of our a empts to understand consciousness, turning eventually to reflection
on representationalism and enactivism. In the end, Hu o concludes that a sat-
isfying naturalistic understanding of consciousness will require a network of
theories operating at dierent levels. More than this, whats needed is an under-
standing of which theories work best at which level no doubt alongside fresh
thinking about consciousness itself.
xxiv
Introduction
Substance Dualism
T. J. Mawson
In this spirited defense of substance dualism, Mawson gets right to it. He
assumes, along with almost everyone else, that there is physical stu which one
might pick out in part with paradigm cases: the stu which makes up tables,
chairs, stars and so on. If you take it that only this kind of stu exists, you are a
physical substance monist. Substance dualism, however, says that there is this
physical stu and another type of stu as well. According to the dualist, the
other type of stu is in essence capable of thought in the broadest sense of
that term, a property most definitely not had by physical stu.
He examines objections to substance dualism. There is of course the point
that any monism has the advantage of simplicity over any dualism, but can we
do be er than Ockhams razor? Mawson considers two sets of problems. First,
there are notorious diculties associated with identifying souls. What makes
one dierent from another, and how can we know anything about souls other
than our own? The second set of problems has to do with perhaps the loudest
objection to dualism: troubles with understanding the alleged interaction
between physical and mental stu. Mawson is unmoved by both sorts of
problems, and he gives replies to each.
He goes on to consider three reasons to believe that substance dualism
is true. Dualism lines up well with certain commonsense intuitions having to
do with personal identity, freedom and the qualitative features of conscious
experience, but physicalism has more than a li le trouble with each one.
He concludes with a Moorean argument which forces a choice between the
simplicity of physicalism on the one hand and the truth of those commonsense
intuitions which favor dualism on the other.
Physicalism
Barbara Montero
Just what is the main thesis of physicalism what does it mean to say that
everything is physical? Montero takes the ma er up in detail, working through
the meanings of the words everything, is, and physical. She begins by
considering dierent approaches to defining the scope of physicalism, conclud-
ing that we should understand everything in the most inclusive way possible:
everything, whatsoever, is physical. Next, she focuses on the relation between
the fundamentally physical properties and higher-level properties such as
mental properties. How should we understand the is relation? Pu ing other
possibilities to one side, her answer is couched in talk of upward determination
with some provisos, worlds which duplicate the fundamental physical properties
xxv
Introduction
and laws of our world also duplicate all the properties of our world. What,
then, is the physical? Again Montero considers a number of answers, plumping
finally for a negative characterization: the physical is the fundamentally
non-mental, non-divine, and non-normative.
Barry C. Smith
We see other people as acting in accordance with beliefs and desires its a
large part of how we understand other people as people, how we see them as
taking a course of action deliberately and with an end in view, and its how we
explain and predict what they do. The sciences of the mind and brain have
revealed a great deal about what makes us tick in other senses. We have a
grip on the cognitive states and mechanisms which also serve to explain our
behavior. Smith considers the relationship between these two views that we
have of ourselves, primarily through the work of Davidson, Denne and Fodor.
He claims that a good account of folk psychology should provide a rational
explanation of action, accommodate the causal ecacy of the mental, and make
room for the dierence between first- and third-person ascriptions of mental
states. Following a careful account of folk psychology and scientific psycho-
logy, as well as a brief wave at eliminative materialism, the work of all three
philosophers is judged on the basis of these criteria.
Smith maintains that even if we can find an account of folk psychology which
satisfies those three requirements, we are still le with a number of questions.
Exactly how do we succeed when we ascribe beliefs and desires to others? What
role does consciousness play? How do the emotions jive with belief-desire psy-
chology? He concludes by pointing towards some potentially fruitful answers.
Sarah Sawyer
There is a large debate between internalists and externalists concerning the very
nature of mental properties is it just whats on the inside that counts? In this
paper Sawyer wades in by first giving an account of both views. Internalism,
roughly, holds that no two individuals could dier psychologically without
diering physically. Externalism, roughly, holds that individuals could be
physically identical but diverge psychologically, given certain external dier-
ences in, for example, their physical or social environments. Well-known
thought experiments owed to Putnam and Burge are considered. Sawyer gives
an account of the many possible forms of each kind of view which seem to
fall out of such reflections.
xxvi
Introduction
Margaret A. Boden
As her title suggests, Boden maintains that theres much more to cognitive
science than merely the science of cognition there are many thinkers working
on various research programs producing a wealth of insights and her paper
captures something of the breadth of the subject. She takes up functionalism
first, as many consider it the core philosophy of the field, and examines several
possible variations in the work of Putnam, Fodor, and Denne . An interest
in neuroscience led many from functionalism to connectionism and parallel
distributed processing. Boden scouts the relevant objections and replies. She
then takes up the contribution cognitive science has made to understanding
the computational processes associated with representation.
Thoughts about representation, and in particular the senses in which inten-
tionality might be understood in terms of situatedness or embodiment, can lead
to reflection on the extended mind. Boden considers Clarks and Chalmers
claim that our minds are somehow partly located in the external world, work-
ing through some diculties for the view along the way. She then considers
the nearby notions of embodiment, enactiveness and phenomenology as they
appear in the continental tradition, as well as the influence this has lately
had on cognitive science. She concludes by examining a large number of con-
tributions cognitive science has made to central features of our understanding
of consciousness, the mind and life. In a concluding note Boden makes a case
for the claim that cognitive science really has provided not just good questions,
but satisfying answers in the philosophy of mind.
Representation
Georges Rey
At least some mental states are representational they stand for, refer to
or are about something else. How might such states fit into our general
understanding of the world? Rey considers the potential of the computational/
xxvii
Introduction
Mental Causation
Neil Campbell
Mental events seem to stand in causal relations to physical events : my hopes
and fears apparently cause my smiles and frowns. Questions about mental
causation have a very long history, but many contemporary philosophers of
mind who are drawn to some version of non-reductive physicalism face a new
version of it. Motivations for non-reductive physicalism seem solid enough.
The anomalousness of the mental leads many to the view, while others are
persuaded by the innocent thought that mental events or properties are
multiply realizable in dierent physical forms. Whatever the motivation,
epiphenomenalism the una ractive possibility that the mental has no eect
on the physical seems only a few steps away.
Some worries about mental causation are raised against Davidsons anoma-
lous monism. Other objections are couched in terms of Kims principle of caus-
al-explanatory exclusion. Campbell considers both objections in close detail.
In his final section he argues that these objections depend on dubious meta-
physical assumptions about the nature of events. The objections, he concludes,
are either misguided or question begging.
Personal Identity
E. J. Lowe
Against the backdrop of some useful clarifications of both the identity relation
and the notion of a criterion of identity, Lowe takes up the question of a criterion
of personal identity. He follows Locke in thinking that before we can establish
a criterion of identity for persons, we have to say what kind of thing persons
xxviii
Introduction
are. We run into a certain sort of trouble here, however, because philosophers
have come up with a very long list of candidates: immaterial substances, material
substances, phases of substances, bundles, transcendental entities, and even mere
fictions. Lowe speculates that our very status as persons and in particular a
certain fact about the first-person pronoun is part of the problem.
He goes on to consider Lockes so-called memory criterion of personal
identity, because it is the first explicitly formulated criterion and, no doubt, the
one which has had the most influence. He considers Reids objection to Lockes
view and some possible modifications which might side-step it, alongside
further objections and replies. He concludes with a consideration of some
alternatives to variations on Lockes criterion, including the possibility that
personal identity is primitive and simple. This last, suggestive possibility goes
some way towards explaining why formulating a criterion of personal identity
is so dicult for us.
Michael Wheeler
Wheeler describes the hypotheses of embodied cognition and the extended
mind as two stopping-o points in the flight from the Cartesian view of intel-
ligent action. In a nutshell the Cartesian view has it that the mind guides action
in a manner largely conceptually independent of the facts of embodiment.
However, recent thoughts about how we actually solve problems in the world
have made the Cartesian account less and less a ractive. Wheelers aim is to
examine the move not just away from the Cartesian picture but from embodied
cognition to cognitive extension.
He first sheds light on the embodied cognition hypothesis by marshalling
examples from recent work in cognitive science. Once a case has been made
for a certain conception of intelligent action, Wheeler argues that we face a
philosophical choice between a radical body-centrism and a new sort of
functionalism. Building up an argument from parity for the extended mind
hypothesis, Wheeler argues that the functionalist option is more a ractive. He
concludes by re-enforcing an aspect of the parity argument.
Paul Noordhof
In this contribution Paul Noordhof brings us right up to date with an account of
how three large topics central to the philosophy of mind have developed over
the last few decades. He takes up recent developments in our understanding of
xxix
Introduction
xxx
1 Problems, Questions and
Concepts in the Philosophy
of Mind
Ian Ravenscroft
Introduction
Philosophy of mind has a long and distinguished history, but it is not the aim of
this chapter to provide an historical overview of philosophical investigations of
the mind. Rather I intend to elaborate on what I take to be the most significant
themes in contemporary philosophy of mind. It is worth noting, right at the
outset, that philosophy of mind has been one of the most dynamic perhaps the
most dynamic areas of English-speaking philosophy over the last half century.
Very roughly, research in this field falls into four broad areas. (Needless to
say, a tangled web of connections exists between these areas, rendering the
boundaries to some extent arbitrary).
1. Metaphysics of Mind: The nature of the mental, and in particular the study
of how the mental relates to the physical, is one of the most enduring
philosophical puzzles and has been at the forefront of contemporary work
in philosophy of mind. This problem is o en simply called the mind-
body problem, and will take centre stage in what is to come. Within the
broad area of the mind-body problem are a host of specialized issues
including mental causation, mental representation and consciousness.
Other specialized metaphysical issues in the philosophy of mind include
perception, memory, action and intention. Very o en these more special-
ized issues are connected with questions in other areas of philosophy.
For example, discussions of moral responsibility o en involve claims
about whether or not an action was intentional.
2. Epistemological Issues: The mind raises special epistemological questions
concerning how, and to what the extent, we have knowledge of our own
mind and the minds of other people. Over the last few decades philo-
sophical work on these questions has been influenced by work in cogni-
tive and developmental psychology and evolutionary biology.
1
The Continuum Companion to Philosophy of Mind
So extensive has been the research over this period, and so enormous
the resulting literature, that any author of an overview like this one is forced
to make hard decisions about what and what not to include. I will focus
on contemporary approaches to what I earlier described as one of the most
enduring philosophical puzzles: the mind-body problem. Even within this
restricted domain I will be forced to make some drastic editorial decisions.
For example, there are important approaches to thinking about the mind-body
problem over which I vault without a glance. (I flinch, but I do not glance.)
In particular, there is an important strand of mid-twentieth century philo-
sophy of mind which includes on the one hand the logical positivists and on
the other the later Wi genstein and Ryle, that I do not discuss at all. This is
not because I believe these thinkers to be unimportant, but because I believe
that I can adequately develop an account of contemporary philosophy of
mind without paying these figures direct a ention. Their influence on contem-
porary philosophy of mind was considerable, but it was channelled through
figures such as Smart and Denne , about whose views I do have something
to say.
In the next section I outline some of the key moves that have been made on
the mind-body problem, beginning with Descartes because, in striking ways,
his problems have turned out to be our problems. Among other ma ers, reduc-
tive and non-reductive physicalism and the idea of supervenience are sketched
in that section. Most contemporary philosophers of mind endorse some version
of the representational theory of mind: they think of the mind as an organ for
2
Problems, Questions and Concepts in the Philosophy of Mind
3
The Continuum Companion to Philosophy of Mind
Dualism
4
Problems, Questions and Concepts in the Philosophy of Mind
5
The Continuum Companion to Philosophy of Mind
6
Problems, Questions and Concepts in the Philosophy of Mind
Reductive Physicalism
I will now leap ahead 300 years to the rise of reductive physicalism (also called
the identity theory) in the 1950s. The theories of mind which populate the
intervening decades are neither unimportant nor uninteresting; however, the
problems on which I will focus in the next three sections mental representa-
tion, mental causation and consciousness take their modern forms in the con-
text of the physicalist theories of the mind which have their origin in reductive
physicalism. To a first approximation, reductive physicalism identifies mental
properties with brain properties. Crucially, the brain properties with which
mental properties are identified are held to be physical properties. To use an old
example from the 1950s an example whose details should not be taken too
seriously the property being in pain is identical to the property having c-fibre
activity.3 It is important to stress that reductive physicalism proposes type
identities. The claim is not merely that every instance of pain is identical to an
instance of a physical property; rather, the claim is that all instances of the
type pain are identical to instances of the type c-fibre firing.
Early proponents of reductive physicalism took the property identities
discovered by science as their model. For example, J. J. C. Smart drew an
analogy between the identity of pain and c-fibre firing (on the one hand) with
the identity of water and H2O (on the other). Crucially, the discovery that water
is H2O was the outcome of a process of scientific investigation; it is not some-
thing that can be discovered by conceptual or linguistic analysis. Similarly,
Smart thought that identities between mental properties and brain properties
would be discovered by scientific investigation. It is no objection to reductive
physicalism that pain and c-fibre activity dont have the same meaning, nor
is it an objection that the proposed identities cannot be discovered a priori. (See
the seminal papers by U. T. Place [1956], H. Feigl [1958] and Smart [1959].)
Non-reductive Physicalism
In the 1960s, two separate developments drove many philosophers to con-
clude that reductive physicalism is mistaken. The first development lead
to functionalism; the second to anomalous monism. In 1967 Hilary Putnam
pointed out that mental properties are, at least in principle, multiply realizable.
7
The Continuum Companion to Philosophy of Mind
8
Problems, Questions and Concepts in the Philosophy of Mind
interact; (2) the principle of the nomological character of causality: causal rela-
tionships always fall under strict laws; (3) the principle of the anomalousness
of the mental: there are no strict laws relating mental and physical events (see
Davidson, 1970, pp. 2234). If, as (1) requires, there are causal relations between
mental and physical events, then by (2), there must be strict laws relating
mental and physical events. And yet by (3) there are no such laws. Davidson
oers an ingenious resolution of this (apparent) inconsistency a resolution
which appeals to non-reductive physicalism. However, before turning to
these ma ers I will briefly explain why Davidson takes the mental to be
anomalous.5
Davidson correctly observes that we do not a ribute mental states one by
one; rather, we a ribute extensive complexes of mental states (Davidson, 1970,
p. 221). For example, if I observe Jones walking towards a mailbox with a le er
in his hand, I will a ribute to him not only the desire to post a le er, but also
the belief that placing a le er in a mailbox is the way to post it; that the red
object nearby is a mailbox; that walking is an eective means of covering
the distance between his present location and the mailbox; etc. But there is a
very large number of sets of mental states that would account for Jones action.
(Jones believes that his le er is about to explode; he desires to prevent the
explosion; he believes that the mailbox is in fact a bomb disposal device placed
there by MI5 . . .) According to Davidson, we select one set of mental states (or
a small number of such sets) from this vast range of possibilities by applying
a principle of charity: we assume that the target is rational. But the notion
of rationality is a normative one that is not found in the physical sciences.
Consequently, we cannot expect to find laws linking the mental realm with
the physical realm.
It is important to note that Davidson is not claiming that the a ribution of
mental states requires that the target be perfectly rational. On the contrary, he
explicitly claims that the a ribution of minor cognitive slips is only possible
because we assume that people are by and large rational: Crediting people
with a large degree of consistency . . . is unavoidable if we are to be in a
position to accuse them meaningfully of error and some degree of irrationality
(Davidson, 1970, p. 221).
Lets now turn to Davidsons resolution of the inconsistency mentioned
above. According to the anomalousness of the mental, there are no strict laws
linking a mental event, described in psychological language, with a physical event,
described in physical language. But if non-reductive physicalism is true, every
mental event token is a physical event token, and therefore has a description in
purely physical terms. Consider an event E which has both a mental descrip-
tion, M, and a physical description, P. By the anomalousness of the mental there
is no strict law linking E, described as M, with some other physical event. But
there maybe a strict law linking E, described as P, with some other physical
9
The Continuum Companion to Philosophy of Mind
Supervenience
10
Problems, Questions and Concepts in the Philosophy of Mind
11
The Continuum Companion to Philosophy of Mind
Here is how it [the intentional stance] works: first you decide to treat the
object whose behavior is to be predicted as a rational agent; then you figure
out what beliefs that agent ought to have, given its place in the world and
its purpose. Then you figure out what desires it ought to have, on the same
considerations, and finally you predict that this rational agent will act to
further its goals in the light of its beliefs. A li le practical reasoning from
the chosen set of beliefs and desires will in many but not all instances
yield a decision about what the agent ought to do; that is what you predict
the agent will do. (Denne , 1987b, p. 17)
12
Problems, Questions and Concepts in the Philosophy of Mind
Mental Representation
It is widely accepted that at least some mental states are about, or represent,
states of aairs. This is sometimes expressed by saying that many mental states
have content. Prominent among the mental states that are widely assumed to
have content are the propositional a itudes, which include beliefs, desires,
fears, hopes and wishes. The name propositional a itude derives from the
13
The Continuum Companion to Philosophy of Mind
idea that such states consist of an a itude (e.g. of belief) towards a proposition
which represents a state of aairs. Thus I can believe that the cat is wearing
a hat and hope that Dr Suess is amusing. Some other kinds of mental states have
contents, for example, perceptions. It is controversial whether the subjective
feels of mental states like perceptions and sensations are representational.
(That is an issue to which we turn briefly in The Knowledge Argument
subsection.)
In this section I sketch some of the key issues surrounding mental repre-
sentation, and some of the key theories of content. I will frequently refer to
the syntactic and semantic properties of mental states. The syntactic proper-
ties of a mental state are those of its narrow properties in virtue of which it
engages in cognitive processes. They are sometimes referred to as a mental
states shape. (Think of how subway tokens engage with the turnstile mecha-
nism in virtue of their narrow properties like shape and mass.) The semantic
properties of a mental state are those properties it has in virtue of its
representational properties. Truth and falsity are semantic properties par
excellence.
14
Problems, Questions and Concepts in the Philosophy of Mind
Theories of content
(1)The approach assumes that humans are largely rational; without such an
assumption there would be li le interest in finding isomorphisms
between the causal structure of our corpus of beliefs and the inferential
structure of a set of propositions. However, there is an extensive body
of work in psychology which suggests that humans are not especially
rational. (For details, and extensive discussion of the philosophical
consequences, see Stich, 1990.)
(2)There exists a very large number of inferential networks isomorphic
to any one causal network; consequently, there will be no unique assign-
ments of contents to beliefs.
(3)The proposal in eect says that the content of a belief is dependent upon
its causal relations, whereas intuitively the causal relations of a belief
are dependent upon its content.
[T]he proposal is that we match the head states that are beliefs with possible
states of the world by the rule that each state of the head gets assigned the
possible state of the world which is such that if it were the way things
actually are, the behaviour that head state causes would realize what the
subject desires. (Braddon-Mitchell et al., 1996, p. 181)
This view requires that the basic unit of semantic interpretation is an agents
entire corpus of beliefs and desires. We assign to the agent those beliefs which,
were they true, would bring about behaviour that satisfied the agents desires;
15
The Continuum Companion to Philosophy of Mind
and we assign to the agent those desires which would bring about behaviour
leading to their satisfaction if the agents beliefs were true.7
This proposal faces very similar diculties to the proposal considered
previously:
Information-Theoretic Approaches
A variety of relations exist between a thought and states outside the body.
According to externalist theories of content, a special subset of those relations
determines the thoughts contents. Which external relations are the content-
conferring relations? Many contemporary philosophers stress information-
bearing relations.8 Back in the 1980s Fred Dretske proposed that a thought, T,
is about a state of aairs, S, in virtue of carrying information about S (e.g.
see Dretske, 1981). That is, T is about S if, and only if, the probability of S given
T is 1. Smoke means fire because the probability of there being fire, given
there is smoke, is 1; similarly, my sheep thought is about sheep because the
probability there is a sheep present, given I have a sheep thought, is 1.
Ingenious though this suggestion is, it is immediately confronted by a pair of
problems. The first is semantic promiscuity: Dretskes view entails that meaning
is superabundant. There are a very large number of states of the world which
carry information about other states of the world. Smoke carries information
about fire; tsunamis carry information about earthquakes; tides carry informa-
tion about the relative positions of the Earth, Sun and Moon. It follows that
meaning is not an especially psychological notion it is not a unique feature
of minds and mind-generated artefacts. The second problem is o en called the
disjunction problem. Under certain conditions I will have a sheep thought in the
presence of a goat. (I might mistake a goat for a sheep on a particularly gloomy
a ernoon.) If that is the case, the probability of there being a sheep present
when I have a sheep thought is less than 1. Indeed, it may be the case that the
probability of there being a sheep or a goat present given I have had a sheep
thought is 1. In that case my sheep thought is about sheep-or-goats rather than
about sheep.
One important line of response to the disjunction problem appeals to
natural selection. Frogs snap their tongues at flies. Say that the frog tokens
the concept FLY when a fly is in its visual field, and (ceteris paribus) tokens of
16
Problems, Questions and Concepts in the Philosophy of Mind
that kind cause tongue snapping. We naturally think that FLY is about flies.
However, small boys sometimes throw BBs (lead pellets) at their pet frogs,
eliciting tongue snappings; that is, frogs misrepresent BBs as flies. Lets assume
that the probability of there being a fly-or-BB, given that the frog tokens FLY
is 1. It follows that the frogs FLY concept is about flies-or-BBs, not flies. In
response, the teleological theory of content introduces the idea of the biological
function of the frogs FLY tokens. It is because the modern frogs ancestors
tokened FLY in responses to flies that they were able to survive and reproduce.
The existence of the modern frog is dependent upon ancestral tokenings of
FLY in response to flies rather than inedible BBs. The theory of natural selection
gives us a principled way of saying what FLY tokenings are for, and that in turn
allows us to distinguish between appropriate tokenings which successfully
represent flies, and inappropriate tokens which misrepresent BBs as flies. (See
especially Millikan, 1984 and Papineau, 1984.)
Jerry Fodor rather doubts that Darwin is going pull Brentanos chestnuts
from the fire (Fodor, 1990d, p. 70). In the frogs ancestral environment, small,
black, fast moving objects in the frogs visual field were almost always flies.
An ancestral frog which responded to small, black, fast moving objects would
have survived and reproduced just as successfully as an ancestral frog which
responded to flies. So the teleological theory of content has no way of deter-
mining whether the modern frogs FLY concept refers to flies or to small, black,
fast moving objects. The disjunction problem has returned. (For countermoves
see Godfrey-Smith, 1994a, pp. 2734.)
Very roughly, Fodors own solution to the disjunction problem turns on the
asymmetric dependence of my goat-caused sheep thoughts on my sheep-caused
sheep thoughts. Goats only get to cause sheep thoughts because sheep cause sheep
thought: no sheep-caused sheep thoughts, no goat-caused sheep thoughts. Sheep
thoughts are about sheep not goats (or sheep-or-goats) in virtue of the depen-
dence of goat-caused sheep thoughts on sheep-caused sheep thoughts (see
Fodor, 1990e). Some philosophers are concerned that Fodors approach still
faces the problem of semantic promiscuity (e.g. see Adams et al., 1994). Fred
Adams (2003, p. 161) oers the following example, which he a ributes to Colin
Allen. Kadu antelopes bite the bark of the acacia tree which in turn emits tannin
as a deterrent. Now while the tannin response evolved as a deterrent to Kadu
rather than to humans, if a human were to damage the bark of the acacia tree
tannins would be emi ed. So the acacias tannin emissions are about Kadu bit-
ings rather than damage by humans (or Kadu-bitings-or-damage-by-humans)
because Kadu bitings cause tannin emissions, and human-caused tannin emis-
sions are dependent on Kadu-caused tannin emissions. (No Kadu-caused tannin
emissions, no human-caused tannin emissions.) So Fodor seems to be commi ed
to the claim that the emi ed tannin molecules are about Kadu bitings.
17
The Continuum Companion to Philosophy of Mind
18
Problems, Questions and Concepts in the Philosophy of Mind
will bring about exactly the same behaviour and will be expressed by u ering
exactly the same sounds or by writing exactly the same marks.
Many philosophers have marked a distinction between wide (or broad)
content and narrow content. Oscars thought that beer is 90 per cent water has
dierent wide content to Twin-Oscars thought that beer is 90 per cent water,
but the same narrow content. As we have seen, wide content supervenes on
the agents wide properties; in contrast narrow content supervenes only on the
agents narrow properties. Some philosophers have argued that psychology
needs only narrow content; in contrast, other philosophers have wondered
whether narrow content is really a kind of content at all. (For a wide ranging
discussion and defence of narrow content, see Segal, 2000.)
Mental Causation
Common sense tells us that physical and mental properties causally interact:
bodily damage causes pain and pain causes wincing. Moreover, our common-
sense notions of agency and responsibility invoke mental to physical causal
relations. We have dierent moral and aective a itudes towards those whose
destructive behaviour is caused by their intentions to behave destructively
than we do towards those whose destructive behaviour is not caused by such
intentions. If it were to turn out that there are no causal relations between
mental and physical properties many of our most cherished views about
human life would have to be reassessed. Fodor makes this point in an especially
dramatic way:
19
The Continuum Companion to Philosophy of Mind
20
Problems, Questions and Concepts in the Philosophy of Mind
calls this problem the exclusion problem because mental properties appear to
be excluded from causal interactions with the physical world (see, especially,
Kim, 1998, Chapters 2 and 3.) In the Some Possible Responses to the Exclusion
Problem subsection below I will briefly canvas some responses to the exclusion
problem.
Sallys desire for tomatoes caused her to purchase tomatoes. If Sally had had
a dierent desire (say for peppers), she would not have purchased tomatoes:
her desires being about tomatoes is what caused the tomato purchase. More
generally, an intentional states representational properties play a role in deter-
mining its causal relations. Now we saw in the Narrow v. Wide Content above
that, according to many philosophers, the representational content of an inten-
tional state depends in part on the agents environment (water refers to H2O
in my mouth but XYZ in my Twins mouth). So if intentional states have
causal powers they have them in part because of their wide properties. How-
ever, the causal powers of an object are entirely determined by its narrow
properties. Fodor makes this point with the following example (Fodor, 1987,
Chapter 2). A quarter activates a vending machine and causes it to emit a
Coke in virtue of its mass, shape and size. Mass, shape and size are all narrow
properties. The wide properties of the coin for example, that it was minted on
a certain date or in a certain place are irrelevant to its causal properties. It is
this fact that makes counterfeit coins possible. If a counterfeiter can succeed in
making a metallic disc with the same narrow properties as a quarter, then he or
she can steal Coke from Coke machines. The upshot of these considerations is
that the representational properties of intentional states are epiphenomenal:
they have no impact on the world.
21
The Continuum Companion to Philosophy of Mind
are types, and so the identities we are considering are type identities.) As we saw
in the subsection on reductive physicalism, reductive physicalists propose that
mental properties are identical to physical properties. So one way to resolve
the exclusion problem is to embrace reductive physicalism. But that involves
rejecting the plausible claim that mental states can be multiply realized. (For an
important assessment of the prospects of rehabilitating reductive physicalism
as a resolution of the exclusion problem, see Kim, 1998, Chapter 4.)
Program Explanation
Frank Jackson and Philip Pe it (1988; 1990) propose that while mental proper-
ties are not causally ecacious that is, mental properties are strictly speaking
epiphenomenal they are nevertheless causally relevant because they pre y
much guarantee that causally ecacious states are present:
The property-instance does not figure in the productive process leading to
the event but it more or less ensures that a property-instance which is required
for that process does figure. A useful metaphor for describing the role of the
property is to say that its realization programs for the appearance of the pro-
ductive property and, under a certain description, for the event produced. The
analogy is with a computer program which ensures that certain things will
happen things satisfying certain descriptions though all the work of producing
those things goes on at a lower, mechanical level. (Jackson et al., 1990, p. 114)
As a consequence of the causal relevance of mental states we can have
powerful explanations of behaviour in which mental states figure, even though
those mental states are not causally ecacious. This strategy concedes some-
thing to the various problems of mental causation without giving up on the
idea that we can predict and explain behaviour by appealing to mental states.
One diculty with this view is that computer programs are causally eca-
cious: the lines of code which constitute a computer program cause (in the right
environment) lines of code in a machine language which in turn cause the
computer to behave in the desired way. So programming cant be taken too
seriously as a metaphor for causal relevance without causal ecaciousness. As
we have seen, it is widely accepted that the mental states supervene on physical
states (in the Supervenience subsection). Perhaps the idea that M programs for
P means only that M supervenes on P (see Kim, 1998, p. 74). But that does not
seem to be very helpful. When M supervenes on P, M nomologically depends on P.
In contrast, the idea of causal relevance requires the dependence of P on M. Its
not clear, then, that the program explanation idea helps us understand how
mental states might be causally relevant without being causally ecacious.
22
Problems, Questions and Concepts in the Philosophy of Mind
mental properties and physical properties to do, so the mental properties are
rejected as causally inecacious. Dual explanandum solutions to the exclusion
problem challenge the idea that there is nothing for the mental properties to do.
Heres an example. Say that Trudys brain instantiates physical property P*, and
that P* is causally sucient to make her arm rise. P* is not, though, sucient to
explain one important aspect of Trudys arms rising: it is not sucient to explain
the fact that Trudys arms rising was the casting of a vote. Trudys arm raising
has two aspects its physical shape and its property of being a vote casting.
The former aspects of the arm raising are fully accounted for by the presence
of P*. However, the presence of P* cannot account for the arm raisings being a
vote casting. The vote casting aspect of the arm raising was caused inter alia
by Trudys intention to vote. We can only make sense of the distinction between
what we might call mere behaviour and intentional action by allowing that
mental states are causally ecacious. Making such an allowance does not
reintroduce the problem of over-determination because P* is not sucient to
explain Trudys action being a vote casting. Solutions along these lines are
advanced in, for example, Yablo 1992 and Thomasson 1998.
One way to understand Princess Elizabeths worry about Cartesian substance
dualism (see the Dualism subsection above) is as a demand for an account
of the mechanisms which link mental and physical properties. That worry
re-emerges here: precisely how is it that Trudys intention to vote caused her
arm raising to be a vote casting? Without such an account it is tempting to
reverse the argument: there are no vote castings because, in order to count as a
vote casting, an arm raising must be caused by an appropriate intention, and
we cannot give an account of how intentions impact on the world.
Consciousness
In the Mind and Body section above, I quoted Colin McGinns question, How
can technicolour phenomenology arise from soggy grey ma er? (McGinn,
1991, p. 1). McGinn is articulating a challenge: explain how consciousness can
exist in a purely physical universe. More precisely, the challenge is to explain
how phenomenal consciousness can exist in a purely physical universe. Phe-
nomenal consciousness is the term used to refer to the subjective properties of
experiences. Thomas Nagel (1974) identified phenomenally conscious experi-
ences as those which it is like something to have. For example, there is something
that it is like to stare at a brightly lit scene, and there is something that it is
like to smell smoke on a damp autumn evening. Phenomenal consciousness is
o en contrasted with access consciousness. A mental state is said to be access
conscious if it is (a) inferentially promiscuous that is, available for use in a
wide range of reasoning tasks, and (b) readily available for the rational control
23
The Continuum Companion to Philosophy of Mind
of action, including speech (see Block, 1994, 1995). Ned Block argues that a
mental state can be access conscious without being phenomenally conscious, or
phenomenally conscious without being access conscious. As an example of the
former he oers the mental states of philosophical zombies: creatures function-
ally identical to humans but lacking phenomenal consciousness. As an example
of the la er he oers, rather controversially, the example of suddenly becoming
aware that the refrigerators compressor, which has been humming for some
time, has stopped. According to Block, we were phenomenally conscious of the
compressors humming all along, but were not access conscious of it; it is only
when the compressor stopped that we could report that it has been humming.
Block stresses that some solutions to the problem of consciousness involve
conflating access and phenomenal consciousness. That is, the author advertises
his or her theory as an account of phenomenal consciousness, but actually
provides a theory of access consciousness.10
For present purposes I will use term the consciousness to refer (exclusively)
to phenomenal consciousness and conscious to mean to phenomenally con-
scious. A further terminological note: I will use the term qualia to refer to
the subjective properties of conscious experiences. Thus the experience of twist-
ing ones ankle has the qualia of hurting; the experience of staring at the sky on
a clear day has the qualia of blueness; and the experience of really wanting a
cigare e has the qualia of craving.11 Occasionally philosophers use qualia
in such a way that, by definition, qualia are non-physical. However, I will
use qualia in a way that leaves open the issue of whether qualia are physical
or not.
I will briefly sketch three kinds of responses to the challenge of locating con-
scious properties in the physical world. David Chalmers (2003a) has oered a
very useful taxonomy of positions in the metaphysics of consciousness. Where
appropriate, I will indicate where positions on my taxonomy map on to his.
Physicalism
24
Problems, Questions and Concepts in the Philosophy of Mind
Optimistic Physicalism
The optimistic physicalist believes that physicalism is true and that, moreover,
human cognitive capacities are up to the task of locating consciousness in the
physical world. Two strands of optimistic physicalism can be identified.
Strongly ptimistic physicalists believe that we have already made significant
progress towards understanding consciousness in physical terms. A very wide
range of theorists fall into this category, including Gilbert Harman (1990),
Daniel Denne (1991a), Fred Dretske (1995) and Frank Jackson (2003).
Weakly optimistic physicalists believe that while we are yet to make significant
progress on the problem, there are no good reasons to believe that we wont
do so in the future. Thomas Nagel (1974), for example, argues that we are cur-
rently unable to reason from a physical description of the brain to a description
of its phenomenal properties because we lack the required concepts. Future
research, though, may one day provide those concepts.
Pessimistic Physicalism
The pessimistic physicalist believes that while consciousness is a physical
phenomenon, humans will never achieve a completely satisfying account of
how consciousness arises in the physical brain. Two strands of pessimistic
physicalism can be identified.
Strongly pessimistic physicalists believe that while consciousness is a physical
phenomenon, developing adequate theories of the emergence of consciousness
from the physical brain will forever transcend human cognitive capacities.
The relevant theories reside, as it were, in a species-wide, cognitive blind spot.
Colin McGinn holds this position (McGinn, 1991, especially Chapter1). While
he calls it transcendental realism, it is sometimes referred to as new mysteri-
anism. (The la er term is Owen Flanagans. See his 1992.)
Weakly pessimistic physicalists believe that, while we may obtain and per-
haps already have obtained an adequate physicalist theory of consciousness,
we are likely to find any such theory unsatisfying. We will grasp the relevant
physical theory, and follow each step of the physicalist explanation, but the
conclusion will not force itself upon us. An analogy: many people understand
the four dimensional theory of time, but cant shrug o the intuition that time
flows. Similarly, we may understand the physicalist theory of consciousness
but be unable to shrug o the intuition that consciousness stands apart from the
physical. Philip Pe it has articulated a view of this sort (Pe it, 2009).
Uncommitted Physicalism
The uncommi ed physicalist accepts that consciousness is a physical phenom-
enon, but expresses no view on whether or not satisfying physicalist theories of
consciousness are available. In the 1990s David Braddon-Mitchell and Frank
25
The Continuum Companion to Philosophy of Mind
Anti-physicalism
Emergentism
According to emergentism, certain complex physical arrangements of ma er
(e.g. human brains) cause new, non-physical mental properties to emerge. The
emergence of non-physical mental properties from complex arrangements of
ma er is a brute fact about the world. In particular, the existence of mental
26
Problems, Questions and Concepts in the Philosophy of Mind
properties is not physically necessitated; that is, the structure of the physical
brain, together with the laws of physics, is insucient to bring about the mental
properties. The classic presentation of this view is Alexander (1920); for a more
recent discussion see McLaughlin (1992).
Phenomenal Fundamentalism
This position, once defended by Bertrand Russell (1927), holds that the intrinsic
properties of the fundamental physical entities are phenomenal (or perhaps
protophenomenal). On this view, the fundamental physical properties such as
mass are in fact relational properties among entities whose intrinsic nature is
phenomenal. (Phenomenal fundamentalism is type-F monism on Chalmers
[2003a] taxonomy.)
27
The Continuum Companion to Philosophy of Mind
(1) If physicalism is true then someone who knows everything about the
physical knows everything simpliciter.
(2) It is not the case that someone who knows everything about the
physical knows everything simpliciter.
Therefore,
(1) seems obviously true. Since physicalism is the doctrine that everything is
physical, the truth of physicalism entails that a person who knows everything
about the physical knows everything simpliciter. Jackson defends (2) by means
of a famous thought experiment. Mary is a brilliant scientist whose colour
visual system is normal but who has been raised from birth in a black and
white environment. She learns everything about the physical aspects of the
human visual system, but has never experienced the colour qualia; in parti-
cular, she has never experienced the qualia of red. Upon her release from the
black and white environment Mary is exposed to a red surface in good light
and exclaims Now I know what red looks like. It is natural to say that Mary
learnt something when she le the black and white environment; that is, it is
natural to say that she gained knowledge of the qualia of red. But if she gained
knowledge then she must have previously lacked knowledge. Thus, even
though (by hypothesis) she knew everything about the physical aspects of
human colour vision, she did not know everything simpliciter. The conclusion,
(3), follows from (1) and (2) by modus tollens.
Upon her release from the black and white environment Mary learnt
something about the qualia of red it was knowledge of the colour qualia that
escaped her when she was in the black and white environment. Since she knew
28
Problems, Questions and Concepts in the Philosophy of Mind
everything about the physical aspects of colour vision when she was in the
black and white environment, the colour qualia must not be physical:
Moreover, Jackson endorses the claim that the world is physically closed:
Therefore,
However,
Therefore,
29
The Continuum Companion to Philosophy of Mind
not have caused her to be in a state of knowledge about the qualia of redness.
(8) expresses the fact that there are only two ways to close o the argument
from the conclusion of the Mary thought experiment to epiphenomenalism by
denying either (4) or (5). Braddon-Mitchell and Jackson assert (9) because they
accept that modern science overwhelmingly supports it. The conclusion, (10),
follows from (8) and (9) by disjunctive syllogism.
Braddon-Mitchell and Jackson call their argument the there has to be a reply
reply. It is a very striking example of what I earlier called uncommi ed phy-
sicalism. Uncommi ed physicalism accepts that consciousness is a physical
phenomenon, but is uncommi ed on the question of whether or not satisfying
physicalist theories of consciousness are available. The there has to be a reply
reply concludes that consciousness is indeed physical, but oers no view on
whether humans will be able to arrive at a good understanding of conscious-
ness as a physical phenomenon.
More commi ed physicalist responses to the knowledge argument can
be grouped into bold and modest versions. Bold physicalist responses to the
knowledge argument insist that Mary learned nothing upon release from the
black and white environment. On this view, the common intuition that Mary
learned something when she finally le the black and white environment is
mistaken. In contrast, modest physicalist responses claim that, although Mary
knew all the physical facts, there were still things she had to learn about red
a er her initial exposure to red surfaces. The a raction of modest physicalist
responses to the knowledge argument is that they preserve the common
intuition just mentioned: that Mary gains knowledge when she leaves the
black and white environment. The diculty for modest physicalist replies
is explaining how Mary gained knowledge even though she already knew all
the physical facts.
A variety of modest physicalist replies to the knowledge argument exist in
the literature. Laurence Nemirow argued that Mary would have all the relevant
propositional knowledge (knowledge that) about colour qualia when she was
in the black-and-white environment, but lacked certain skills (knowledge
how) (Nemirow, 1980; see also Lewis, 1990). The intuition that Mary learns
something when she leaves the black and white environment is explained by
the fact that Mary learns the relevant skills (e.g. she can now imagine a red
surface); however, there were no facts with which she was unfamiliar prior to
her release. Jackson has advanced an ingenious argument against the skills
response. He points out that prior to her release Mary would not know
what other peoples mental lives are like. What, she might wonder, is it like
for ordinary people to look at a ripe tomato in good light? Jackson urges that
it is implausible that Mary is wondering about other peoples skills; it is facts
about other people that she is missing (Jackson, 1986).
30
Problems, Questions and Concepts in the Philosophy of Mind
Another kind of modest physicalist response turns on the idea that the same
propositional knowledge may be stored in dierent representational media.
David Lewis advanced a view of this sort by way of an analogy:
Imagine a smart data bank. It can be told things, it can store the information
it is given, it can reason with it, it can answer questions on the basis of its
stored information. Now imagine a pa ern-recognizing device that works
as follows. When exposed to a pa ern it makes a sort of template, which it
then applies to pa erns presented to it in future. Now imagine one device
with both faculties, rather like a clock radio. There is no reason to think that
any such device must have a third faculty: a faculty of making templates
for pa erns it has never been exposed to, using its stored information about
these pa erns. If it has a full description about a pa ern but no template
for it, it lacks an ability but it doesnt lack information. (Rather, it lacks
information in a useable form.) When it is shown the pa ern it makes a
template and gains abilities, but it gains no information. We might be
rather like that. (Lewis, 1983b, pp. 1312)
Lewis describes a case in which a machine possess all the relevant proposi-
tional knowledge represented in a sentence-like medium, but does not possess
all the relevant knowledge in an alternative, analogical medium. Similarly,
before her release Mary might possess all the relevant propositional knowledge
in a sentence-like medium, but only a er her release does she acquire new
representations of that knowledge in an alternative, phenomenal, medium.
Paul Churchland presses a similar point, and provides reasons for thinking
that the human brain does indeed contain two (or more) distinct representa-
tional media with only a limited ability to translate between them (Churchland,
1989a). However, a nagging doubt remains. There is something that it is like to
have a phenomenal representation in the alternative medium. And knowledge
of what it is like to have a phenomenal representation in the alternative medium
is exactly what Mary seems to be missing.
In recent years Jackson has provided his own modest response to the
knowledge argument (Jackson, 2003). His response turns on two claims
31
The Continuum Companion to Philosophy of Mind
In defence of (12) Jackson asserts that, while we do not yet have agreement
on the correct form a physicalist theory of mental representation should take,
we can nevertheless be confident that such a physical theory is in principle
available. The idea that qualia are essentially representational is not original to
Jackson, having been articulated by, for example, Gilbert Harman (1990) and
Michael Tye (2000). On this view, qualia are distinguished from other forms
of mental representations in virtue of their unique functional roles. Pe it puts
this point succinctly:
A state will count as experiential so far as it functions in a manner typical of
experiences: it generally disposes an agent to come to believe that things are as
they are represented to be; it does not control behaviour except when it leads
to belief; it may remain in place continuing to represent things being thus
and so, even when the subject has come to believe that they are not that way
. . . and so on. (Pe it, 2009, p. 169)
An important feature of many modes of representation is that the represen-
tation need not have the property it represents. Linguistic representations are
like this. The following token
green
represents the property green but is not green. Similarly, my qualia of redness
which, according to representationalism about qualia, represents surfaces as
being red, need not itself be red. Its long been observed that, when I have a
mental image of a ripe tomato, there need be nothing red in my brain (e.g. see
Smart, 1959). It has been suggested, though, that my image has the property of
being phenomenally red of instantiating a special phenomenal property which
is sometimes referred to as phredness. Representationalists about qualia
deny this. My qualia of redness represents red without being either red or
phred. The what it is like of the experience I have when I look at a ripe tomato
in good light is entirely exhausted by the fact that it represents the tomato as
being red. There is nothing else to be explained (see, especially, Jackson, 2003.)
Jackson accepts, however, that Mary still makes an epistemic gain when she
leaves the black and white environment; that is, he articulates what I have
called a modest physicalist reply to the knowledge argument. While Jackson
asserts that Mary knew all the propositional knowledge about qualia prior to
her release, he does not claim that Mary learned nothing about qualia upon
her release. In particular, he endorses Nemirows claim that Mary acquired
certain skills upon her release.
In response to Jacksons representationalist reply to the knowledge
argument, Robert Van Gulick has argued that physicalists can successfully
respond to the knowledge argument without endorsing representationalism.
The knowledge argument contains, he submits, a number of assumptions which
32
Problems, Questions and Concepts in the Philosophy of Mind
the physicalist can challenge. (See Van Gulick, 2009.) Chalmers has raised
objections directly against representationalism (Chalmers, 2003a, p. 111). He
distinguishes between functional and phenomenal representation as follows. A
system has a functional representation of p when it responds to p appropriately.
For example, I have a functional representation of the red trac light when
I respond to it by braking. In contrast, a system has a phenomenal representa-
tion of p when the system is phenomenally conscious of p. For example, I have
a phenomenal representation of the red trac light when there is something
that it is like for me to see the red light. (This distinction is related to Blocks
distinction between access and phenomenal consciousness, introduced at the
beginning of this section.) Chalmers worry is that Mary could have full
knowledge of the functional representational properties of the qualia of red
without having knowledge of the phenomenal representational properties of
red. That is, phenomenal properties cannot be reduced to or even multiply
realized by functional properties.
Conclusion
For most of the history of the philosophy of mind, the mental realm was
taken to be exhausted by the phenomenal realm. But under the impact of
logical positivism and behaviourism in the first half of the twentieth century,
the mental was reconceived so that consciousness was only one, inessential,
aspect of the mental. (Perhaps the last great philosophical work which took
mind and consciousness to be synonymous was C. D. Broads The Mind and
Its Place in Nature [1925]). Once it was recognized that the mental was not
exhausted by the phenomenal, a change of focus took place. Central to that
change of focus was functionalism. By identifying mental states as the occu-
pants of characteristic functional roles, it became possible to identify mental
states with brain states, thus planting the mental firmly in the physical world
and providing new ways to think about the relationship between neuroscience
and psychology.
But two apparently intractable problems remain. Clearly, the problem of
mental causation cannot be functionalized away. Functional roles are causal
roles, and the functional roles characteristic of mental states include psy-
chophysical causal relations. The problem of mental causation is a challenge
to functionalism, not a puzzle that can be resolved by appealing to functional-
ism. The second apparently intractable problem is phenomenal consciousness.
Qualia are resistant to a purely functional approach. Mary knew all the func-
tional roles characteristic of the qualia of redness, and knew what states
occupy those roles. Nevertheless, it seems that there was something about
qualia which she did not know. Similarly, it seems that we can conceive of
33
The Continuum Companion to Philosophy of Mind
34
2 Consciousness
Daniel D. Hutto
What is Consciousness?
There is no u erly clean, clear and neutral account of what exactly is covered
by the concept of consciousness. The situation reflects, and is exacerbated by
the fact that we speak of consciousness in many dierent ways in ordinary
parlance. A consequence of our multifarious uses of the concept is that it has
proved impossible to define its essential characteristics through conceptual
analysis. We have nothing approaching a descriptively adequate philosophical
consensus of what lies at the core of all and every form of consciousness in
terms of necessary and sucient conditions that would be accepted by all
interested parties.
This is not regarded as a cause for despair. The same is true of other philo-
sophically important topics such as knowledge and causation. Despite this,
consciousness remains of pivotal philosophical interest because of its centrality
to our psychological lives and the way that it tantalizingly resists incorporation
into a fully naturalized account of the world.
Recognizing that a empts to provide a philosophically robust definition of
consciousness are likely forlorn, a standard tactic for isolating core features
of consciousness is to provide clear-cut exemplars as specimens. By means of
this strategy we might still, at least, divine philosophically important a ributes
of the quarry. Take your experience of reading these lines. Hopefully their
content is at the focal centre of your a ention, but even if so there will be a
range of other peripheral and background things of which you are consciously
aware: colours, noises, feelings. Some of these may remain present throughout
your intellectual activity while others intrude upon it momentarily, in largely
expected but perhaps occasionally surprising ways, before vanishing from
the stage. Despite such comings and goings you will not feel as if your overall
experience is ruptured or fragmented.
Conscious experience of this sort is u erly mundane and intimately familiar.
It appears to be an all-or-nothing property that pervades the waking lives of
many creatures. Human beings, cats, octopi (apparently), and spiders (perhaps)
are kinds of beings commonly thought capable of possessing consciousness
while inanimate objects, such as chairs, are not. We say of creatures or organisms
35
The Continuum Companion to Philosophy of Mind
that they are conscious if they are awake and sentient. Evidence of this is that
they exhibit a certain degree of sensitivity or coordination with respect to
aspects of their environment. However if the case just described is taken as
our paradigm, then merely exercising capacities for such responding will not
suce for being truly conscious. It is easy to think of examples of complex
intelligent activity, sometimes of a quite sophisticated kind, that are never-
theless apparently habitual, automatic, or unreflective.
Most philosophers insist that, minimally and necessarily, to be conscious it
must also be the case that a being possess or enjoy some degree of occurrent
experiential awareness. In other words, there must be something that it-is-like
for them to be awake, sentient or intelligently controlling its behaviour. A truly
conscious being enjoys experiences that have phenomenal aspects; it feels a
certain way to be such a creature in such and such circumstances.
Experiential awareness can take dierent forms. It may be transitive in the
sense of being awareness of environmental surroundings or aspects thereof.
For example, the subject may be aware of the red speck in the centre of its visual
field. For this reason consciousness is o en regarded as being inherently
intentional, as being directed at certain objects, not others. But it seems possible
to be experientially aware in more intransitive ways too in ways that lack
directedness at specific objects. Diuse and undirected forms of consciousness
are surely possible, as is the case with moods, such as elation, calmness or
depression. Other, even more basic forms of undirected conscious experience
are also imaginable. Either way, to repeat, being conscious appears to require
being in a state of mind with a characteristic feel one in which there is
something-that-it-is-like to be in it. This is seemingly common to all forms
of consciousness; or, more cautiously, at least there are interesting forms of
consciousness that have this feature necessarily.
Conscious beings are essentially experiencers and the particular types of
experiences that they enjoy have distinctive characteristics and notable aspects
(i.e. they have specific phenomenal properties or characters). Experiencing
itchiness, for example, is quite dierent from experiencing anger. Seeing the
peculiar greenness of an aloe vera plant diers from seeing the peculiar green-
ness of a Granny Smith apple. We can specify the dierences by using illocu-
tions such as this or that shade of greenness. But this is to invoke inevitably
crude and (still) relatively abstract categories in order to pick out something
that is much more fine-grained, analog and particular.
Experiencing phenomenal characters, apparently, ma ers. Having experiences
seems to make a dierence to what is done, in line with how such experiences
are evaluated. Encountering the unusual taste and smell of durian, for example,
may evoke reveries or prompt certain other actions, depending on whether one
finds that taste pleasant or unpleasant. In line with this some are inclined to
reserve, more stringently, the accolade of being phenomenally conscious only
36
Consciousness
for those beings that exhibit a certain degree of global control over their actions
or that are capable of reporting, expressing and appraising how things appear
to them. To achieve this, it is argued that conscious beings must not only be
aware of and a end to aspects of their environment but to aspects of experi-
ential mental states themselves. Accordingly this kind of capacity implies at
least some degree of self-awareness or self-consciousness. If one accepts this,
subjects that are truly phenomenally conscious must not only enjoy experiences
with certain phenomenal qualities, they must be aware of the qualities of these
experiences. If so, those states of mind that exhibit phenomenal consciousness
do so at best only partly in virtue of having phenomenal characters.
Still, even if all conscious beings are experiencers of some or other phenom-
enal properties it may be that they experience these in more or less unified
ways. Human experience tends to integrate experienced phenomenal proper-
ties (i.e. those associated with dierent sensory modalities), continuously and
seamlessly over time. Recent empirical studies concerning the phenomenon of
ina entional or change blindness, raise doubts about the extent and degree to
which we actually experience the world in fully detailed and non-gappy ways.
Still, for human experiences at least usually it feels as if the way in which
our experiences inter-relate and change happens in coherent, well-coordinated
and expected manners.
Typical human consciousness, at least, feels as if it were, in important
respects, objective, temporally extended and unified. It involves having a coher-
ent and unified individual perspective on reality. These unique points of view
are internally complex. When we notice and a end to specific worldly features,
such as the greenness of a particular apple, this involves being able to see an
apple as something more than just the sum of its presented features. To see an
apple as something in which greenness, and other properties, might inhere is
to see it as having a continued existence over time. Experiencing a world of
objects and their features always occurs against a larger and more complex
background in which such items are systematically related to other things. To
have experience of the world, as opposed to merely having sentient capacities,
is to experience worldly oerings in a structured way.1
This entails, modestly enough, that dierent sorts of creatures may enjoy
dierent forms of consciousness. What it is like to be a human being may vary
considerably from what it is like to be a dolphin, or more famously still, what
it is like to be a bat. Indeed, even what it is like to be a particular human being
in a particular set of circumstances can dier qualitatively from what it is like
to be a particular human being in another set of circumstances.
Conscious experience is subjective at least in the sense that as Nagel (1974)
proposes it is idiosyncratic. Being phenomenally conscious apparently equates
to having a particular point of view or perspective that involves having a range
of more or less unified experiences with individual phenomenal characters.
37
The Continuum Companion to Philosophy of Mind
Many philosophers hold that since we have no direct access to how things
appear experientially to others, it is enigmatic whether others are conscious or
what the exact character of their conscious experience is like. Thus unless it is
possible to securely infer what it is like for the other from more objective avail-
able facts, then, for all we know, even apparently sophisticated and intelligent
beings may lack conscious experience altogether, or they may enjoy experiences
that possess radically dierent phenomenal or qualitative characters from our
own. Moreover, the way that their experiences are normally integrated with
one another or unified (to the extent that they are integrated or unified at all)
may be quite alien to the way that typical human experience is organized.
Even if a fully transparent conceptual analysis of consciousness is not on the
cards, it seems that there are a number of identifiable or at least apparent
properties that are fundamental to it that make it of real philosophical interest.
Perhaps based on empirical or philosophical reflection it will be decided that
not all of these seeming a ributes are genuine; perhaps they will not all make
the final list of properties that warrant straight explanation. Nevertheless,
phenomenality, intentionality, subjectivity, unity, temporal extension, minimal
self-awareness are prima facie prominent features of consciousness that must
be either explained or explained away.
38
Consciousness
Debates between naturalists of this stripe take the form of in-house assess-
ments of (and sometimes proposed adjustments to) each others headline pro-
posals. Not every framework is regarded as equally promising. For example,
the more extreme versions of behaviourism are almost universally unpopular
today. Their followers hold that any and all genuine mental phenomena need to
be identified with behaviours or dispositions to behave that allow for opera-
tional definitions, however complex, in terms of observable causes and eects.
A major criticism of this framework is that, in relying entirely on dispositions,
however complex, in order to understand the mental it lacks the essential
resources to satisfactorily account for the holistic and complex sorts of inter-
actions that occur between mental states that take place before responses
are produced. In this respect, functionalism, its natural successor, is deemed
superior because it makes space for precisely this sort of complexity.
Reductive functionalists take it that to be a conscious mental state of a certain
type equates to being a functional state of a whole organism: a state that can be
understood in terms of its wider systemic relations or teleological purposes and
that has appropriate causal relations to perceptions, other mental states and
actions. According to the analytic or commonsense version of the doctrine,
mental phenomena, including the experience of sights, sounds, pains and other
conscious mental states, are identified with specifiable higher order causal
roles. And for reductive naturalists these, in turn, are, either directly or indi-
rectly, identified with the physical states that happen to occupy, realize or fill
those roles. Rich mental activity is thus thought to take place between stimuli
and responses. Mental states, activated by environmental triggers, causally
interact in specifiable ways with each other and other bodily states, and only
then produce outward responses.
In large measure functionalisms popularity as a general framework for
thinking about mentality derives from the fact that it gives both philosophers
and psychologists the requisite apparatus and platform for positing inner, caus-
ally ecacious mental states without having to commit, in advance, to specific
details about how such mental states are physically realized or implemented.
This is useful because there appears to be an enormous stock of creatures of the
actual, terrestrial and imaginable, alien varieties that are capable of conscious
experiencing, despite the fact that they lack a physiology similar to our own.
The key functionalist insight is that not every creature which might be capa-
ble of conscious experience need have central nervous systems or brains like
ours. Consequently, we should not expect to uncover any neat, cross-species,
one-to-one correlations holding between particular types of experience and
particular types of neural or physiological states. To assume otherwise is, it is
claimed, to promote an unwarranted species-biased chauvinism.
Moreover, empirical work has revealed that brains, such as ours, are open
to re-wiring; that the neural structures underpinning certain types of mental
39
The Continuum Companion to Philosophy of Mind
activity are highly plastic. For example, patients who have had hemispherecto-
mies in which the cortex of one hemisphere is removed have succeeded in
enlisting other parts of their brains to restore the lost functions, thus managing
to compensate. In the light of such discoveries it appears mistaken to assume
that there will be uniquely dedicated neural configurations supporting very
specific kinds of mentality. It is likely that this is true of conscious experience too.
Considerations of this sort cast doubt on strong, type-type versions of mind-
brain identity theories: those that propose straightforward identifications of
particular kinds of conscious experience with particular types of brain event.
Type-type identity theorists hope to provide class-to-class identifications of
brain events or processes that will be capable of grounding an interestingly
predictive and informative science of the mind. Such theories hope to tell us
that being a middle-A sound is identical with being an oscillation in air
pressure at 440 hertz; being red is identical with having a certain triplet of
electromagnetic reflectance eciencies; being warm is identical with a certain
mean level of microscopically embodied energies, and so forth (Churchland,
1989b, p. 53).
However if some version of functionalism is true then psychological
laws can be cast at a higher order level even though conscious states will be
variously, and thus disjunctively, realized in species specific (and perhaps even
individual-specific and/or circumstance-specific) ways. Consider that the hum-
ble carbure or can be functionally defined in terms of the abstract causal role
it fulfils. Something is a carbure or insofar as it mixes air with liquid fuel.
In theory, such devices could be made of metal, rubber, plastic, possibly even
soul-stu as long as they are capable of discharging the stated function. The
relation between a functional role and what realizes it can be one-to-many as
opposed to one-to-one. Thus a general description of the realizers of any given
type of experience would take the form of a disjunction on the right side of the
relevant equation that includes mention of a, perhaps, indefinitely long chain
of dierent kinds of instantiating states.
The trouble is that, without significant qualification, functionalism can
appear overly inclusive when it comes to saying which sorts of systems ought
to make the list of the conscious. This is illustrated by the fact that it is easy
to imagine beings that produce the appropriate outward behaviour through
functionally identical means but which plausibly lack any kind of experiential
awareness. For example, Ned Block (1978) famously imagined a scenario in
which the behaviours of a complex artificial body are orchestrated by com-
munications between members of the Chinese nation, so as to mimic in all
functionally relevant respects the responses of a human being undergoing
a painful experience. In order to achieve this feat, the Chinese citizens are pro-
vided with rules on how to respond to instructions provided by sky-based
display and are able to communicate with one another by means of two-way
40
Consciousness
radio-links so that, by working together, they are able to remotely generate just
the right kinds of responses in the body.
However outlandish the scenario, it is at least theoretically possible that the
Chinese nation could simulate human pain behaviour in the artificial body
by such means, mirroring at one level of description the ways in which such
behaviour is normally functionally produced in humans. The worry is that if
the two systems are identical in this respect then, assuming functionalism is
true, it appears we have no principled grounds for saying that one is really
undergoing the experience of pain and the other not. The situation is preposter-
ous since, intuitively, we want to ascribe the experience of pain to the individual
human being while withholding that ascription to an artificial body controlled
by the conglomerate of Chinese individuals engaged in this bizarre exercise. Yet
the only path to that verdict seems to require thinking of conscious experience
as something distinct from functional properties per se.
Indeed, this is precisely the moral that many are inclined to draw from this
thought experiment; while functionalism might reveal something important
about the processes and structures associated with experiencing it is incapable of
providing a real insight into the nature of experience itself. On these grounds
functionalists are o en accused of having no reasonable means of accommodat-
ing the phenomenal character of experience. Their proposals apparently miss
out the most important ingredient: how it feels.
Of course, it is possible for the functionalist to bite the bullet and insist, fly-
ing in the face of standard intuitions, that if a systems behaviour is generated
in the right way then it simply is conscious. So if the human is, then so too is
the body governed by the Chinese nation. But this is not the only (nor the
most convincing) line of reply. Arguably, the charge of liberalism might be
answered by adjusting the level of grain of the proposed functional analysis.
This requires a shi of a ention from abstract job descriptions crudely, what
the thing is doing to a greater focus on precise engineering details crudely,
how the thing actually does what it does (Churchland, 1989b, Chapter 2;
Flanagan 1991). Ironically, what is hailed as the chief virtue of functionalism
its capacity to abstract from specific details and its openness to the possibility
of various realizability looks to be a vice when it comes to understanding
consciousness.
Nevertheless, to assume that such objections are fatal is to underestimate
the flexibility of the functionalist approach. It is quite open for defenders of this
framework to insist, more in the spirit of identity theory, that the peculiarities
of how a system is organized and even what materials it is composed of might
ma er to having experience (or having certain kinds of experience). This is
wholly consistent with acknowledging that two systems that dier in their
lower level engineering details might be regarded as functionally equivalent at
some higher level of analysis.
41
The Continuum Companion to Philosophy of Mind
42
Consciousness
43
The Continuum Companion to Philosophy of Mind
44
Consciousness
there is to know under one conceptually based description does not entail
knowing the very same things under another.
What makes this line of reply plausible is that concepts of phenomenal
consciousness are apparently special in key respects. Their special features are
what systematically foster and explain the illusion that experiential properties
can come apart from all other properties, even though this isnt so. Public con-
cepts of experience, such as redness or itchiness are, some hold, recognitional
in nature. They are thought to involve perceptions of worldly properties.
In contrast, phenomenal concepts, such as seems red or feels itchy which
some hold are formed on the basis of re-enacting or having higher order per-
ceivings or believings about first order experiential states are regarded as
purely recognitional or inherently first-personally perspectival (Papineau, 2002;
Tye, 2009).
If phenomenal concepts are formed in special ways and have unique pro-
perties because of this, then it is arguable that it only seems to us that zombies
are possible (i.e. one can conceptually imagine phenomenal properties as being
distinct from all other properties even though they cannot be) and that it only
seems that Mary doesnt know all the facts (in fact she does, but she knows some
of them under a limited description or mode of presentation). If so, those who
are excited by the idea that such thought experiments damage the prospects of
naturalism are subject to a persistent cognitive illusion.2
Those who endorse the phenomenal concepts strategy place dierent bets
on the odds of closing the so-called explanatory gap (Levine, 1983). For even
if naturalists are able to put their metaphysical house in order there remains
a lingering and perturbing question of why a neural state should be the basis
of a certain kind of experience or of any experience instead of none at all. This
question suggests that there is a gap in our understanding that still appears
to remain wide open even if one denies the force of the standard thought
experiments.
One tactic for dealing with it, promoted by McGinn (1991), is to simply con-
cede that our minds are cognitively incapable of forming the relevant concepts
required for closing it (i.e. of providing a constructive, scientific account of con-
sciousness). This is so, he maintains, even though consciousness is a perfectly
legitimate natural and, indeed, wholly physical phenomenon. In principle it is
wholly explicable in physical terms even though we are, forever, cognitively
closed to understanding how this could be so. We are prevented from this
because we lack the appropriate cognitive faculties.
Top-down a empts to understand the psychophysical link between the
experiential and the physical are impeded on one side by the limits of intro-
spection. There is nothing in our experience that provides us with the means of
intelligibly understanding how experience is generated by the processes that
45
The Continuum Companion to Philosophy of Mind
46
Consciousness
2002, p. 3). Accordingly the best policy for dealing with the hard problem is the
same as in dealing with taxes; avoidance is permissible but evasion is illegal.
In this case avoidance looks like the best move since as soon as you suppose
that conscious states are distinct from material states, then some very puzzling
questions become unavoidable (Papineau, 2002, p. 2).
A empts to identify conscious experience with a physical state of some kind
or other would be doomed to fail from the outset if conscious experiences in
fact shared no essential properties in common with such states. But the reply
to this charge is that while this might seem to be the case it simply isnt so. It is
entirely possible that one and the same thing may present itself to us in dier-
ent ways. There are plenty of cases in which a single referent is mistaken for
numerous distinct ones and vice versa because of misleading appearances,
names or descriptions. Noting this is all that defenders of reductive theories
require if they are to establish that their hypotheses about consciousness might
be possibly true in ways that would obviate having to deal with the problems
of consciousness.
To take this line is to hold that there is nothing more inherently absurd in
claiming that conscious experiences might equate to certain kinds of physical
happenings than there is in claiming that the Morning star and the Evening
star are the same planet: Venus. In neither case is the identity immediately
obvious or self-evident. If we allow this then there is no reason to deny, in
advance, that conscious experiences could not be identified with something
physical. To think otherwise, on the basis of appearances of dierence, is to be
under the sway of the stereoscopic or antipathetic fallacy.
While a ractively simple, nevertheless this sort of reply only goes so far. It
does nothing, by itself, to motivate acceptance of any of the proposed identity
claims; at most it makes space for their possible truth; at best it secures the
barest logical possibility of putative identity. And it does not deal with the root
problem that underpins the explanatory gap or the hard problem because it
fails to overcome worries about the intelligibility of making certain identity
claims. The bo om line is that to make their favoured identity claims credible,
it is necessary for reductionists to deal with the appearances of dierence in
some satisfactory way.
47
The Continuum Companion to Philosophy of Mind
that seemingly distinct things can possess all of their apparent properties with-
out tension or contradiction. Thus to make any progress on the problems of
consciousness, to render any given naturalistic equation about consciousness
truly convincing, would involve showing how the properties proposed for the
reduction could be the kinds of things that experiential states or properties
might be.
A nagging concern is that all existing reductionists proposals leave too
many important questions unanswered about the appearances of dierence.
In particular, they give inadequate answers to question, such as: Why do
experiences feel as they do? Who or what does the experiencing? How and
where does this all come together?
A deep-seated problem is that although reductive naturalists outwardly
denounce the picture of mental objects as occupying an inner sanctum of the
mind, they are inclined to take seriously questions and problems that do not
wholly make sense without presupposing this picture in some, perhaps vesti-
gial way. For example, some are tempted to ask: Where is my experience of pain
located? The sense of this question is taken to be straightforwardly akin to
the query; where is my pen located? But this leads directly to the problem of
phenomenal space, which is the problem of finding a place for the world of
experience within the world of physical space. In this context Denne is right
to ask, Now what is phenomenal space? Is it physical space inside the brain?
Is it the on-stage space in the theater of consciousness located in the brain
(Denne , 1991a, p. 130)?
Denne s analysis of the assumptions grounding the enterprise of explain-
ing consciousness is instructive. He believes that most philosophers, and many
lay folk influenced by them, conjure up images of the mind as an inner, mental
theatre complete with a self who examines various on-stage objects in the
spotlight of consciousness (pains, colours, figments of the imagination, etc.).
Those under the sway of this picture think of our verbal reports concerning
consciousness as based directly upon what the self sees on its private, inner
screen. Apparently it introspects mental items in a way similar to that in which
we ordinarily inspect everyday things such as watches or pieces of china.
Denne has done more than most to get us to critically question our thinking
on this score so as to abandon the idea that there is any such place in the physi-
cal world (and, in particular, the brain) where all the events of consciousness
come together. Rather than starting with such dangerous assumptions about
our explanandum he thinks that we have no choice but to begin our investi-
gations into the nature of consciousness by interrogating first-person reports
in a public, intersubjective context. He gives the name heterophenomenology
to this activity. While engaged in it, we, as interpreters, eectively allow the
subjects to verbally describe to us the nature of their experiences. They generate
texts about how things seem to them. They have authority concerning the
48
Consciousness
content of what is described. But what is described are best understood (at
least in the first instance) as notional worlds that are analogous to fictional
worlds, such as Sherlock Holmess London (not the real London). In being of a
like nature to such fictional worlds, The subjects heterophenomenological
world will be a stable intersubjectively confirmable theoretical posit (Denne ,
1991a, p. 81).
It follows that speech acts are the primary interpreted data for the study
of consciousness. These are reports, judgments, and beliefs that are made
concerning purported conscious experiences. The question of whether or
not what is described in these speech acts is real or fictional is le in abeyance.
Ocially, when we start investigating consciousness scientifically using this
method, we are required to begin (but not end) by focusing on the contents of
the speech acts of humans (and other possible speakers), staying studiously
neutral on what if anything lies behind them/explains their etiology.
The ontological moral Denne is inclined to draw is that although we ought
to allow subjects to have the final word in saying how they judge that things
appear to them, this in no way commits investigators to take seriously what
they describe at the level of ontology. He maintains that this is the only sound
way to take the first-person point of view as seriously as it can be taken
(Denne , 2003, p. 19). For him, interrogating such texts are our only means
of neutrally analysing the reports about what is going on in our minds. He
claims that the texts generated in these circumstances, and not something above
and beyond to which they putatively refer, are the raw material for any theory
of consciousness.
In promoting this understanding of where we must start Denne oers a
new metaphor for consciousness; the multiple dra s model. The multiple dra s
model identifies consciousness with our ability to generate a coherent text
concerning our putative mental episodes. James Joyces Ulysses is the model.
But he goes further and advances a positive reductive theory of consciousness
in terms of the ability to generate detailed, coherent serial reports. For Denne
the business of explaining consciousness boils down to explaining how the
brain is able to produce the relevant texts. By his lights we wont have explained
consciousness until we give a naturalistic account of our ability to produce
coherent speech acts through which we describe our experience of what it is
like for us to be conscious. And like all good reductionists he believes that,
Only a theory that explained conscious events in terms of unconscious events
could explain consciousness at all (Denne , 1991a, p. 454). Explaining con-
sciousness, for him, converts to explaining the capacity for a certain kind of
text production.
His task is not to explain the existence of conscious experience as it is usually
imagined to be but rather to explain how our talk about how things seem to us
is produced by underlying sub-systems. Thus he hopes to give an ontogenetic
49
The Continuum Companion to Philosophy of Mind
explanation of how those sub-systems were formed and an account of how they
work. The essence of his proposal is captured in the following remark I am
suggesting conscious human minds are more or less serial virtual machines
implemented ineciently on the parallel hardware evolution has provided
for us (Denne , 1991a, p. 218). In line with his multiple dra s model he calls
the virtual machine that gives rise to consciousness a Joycean machine. And
he is quite aware of the limits of his theory; he notes that If consciousness
is something over and above the Joycean machine, I have not yet provided a
theory of consciousness at all (Denne , 1991a, p. 281).
There are similarities between Denne s approach and ambitious higher
order theories of consciousness those that maintain that being phenomenally
conscious requires a ending to or noticing the aspects of ones mental states
in ways that necessarily involve making reference to those aspects in higher
order acts of perception or thought. Accordingly, phenomenal consciousness
requires the use of higher order perceptions or thoughts (Lycan, 1996;
Rosenthal, 2005; Carruthers, 2000). If such higher order operations are, in fact,
partly constitutive of phenomenal consciousness then the neural basis of
experience must include machinery for inner sensing or for making theory
of mind ascriptions.3
For those who doubt that such mechanisms form part of our basic biological
equipment, Denne s account has an advantage. He does not believe that
the Joycean so ware is built-in; he regards it as the result of cultural design.
He tells us that consciousness is, largely a product of cultural evolution that
gets imparted to brains in early training (Denne , 1991a, p. 219). But critics
regard this as an admission that non-verbals, such as animals and infants, are
incapable of having experiences. Denne s response to this worry is that our
folksy intuitions regarding animal and infant consciousness are not sacrosanct.
Je isoning some of our most deeply held intuitions concerning the nature
of experience may be a price we must pay for adopting a neater criterion of
consciousness.
A deeply objectionable feature of Denne s theory, echoing the problems of
certain versions of behaviourism and functionalism, is that a complex system
would count as conscious if it produced pa erns of behaviour identical to,
say, those of yours or mine when we generate a stream of coherent u erances
that are interpretable as saying how things seem to us. Highlighting this aspect
of Denne s account, many have complained that his theory leaves out what
is critically important for understanding phenomenal consciousness: phenom-
enal qualities themselves. A capacity for experiencing such qualities, it is argued,
is logically independent from (and developmentally prior to) capacities for
propositional believing, reportage and narrative text production.
The problem for Denne s proposal and the oerings of higher order thought
theorists is that, as stand-alone accounts, they allegedly place too much emphasis
50
Consciousness
51
The Continuum Companion to Philosophy of Mind
things might be (given that how the world seems to be and how it actually is
may dier). Consequently, there can be no dierence in phenomenal character
without a corresponding dierence in representational content because
phenomenal character just is a kind of representational content. Weaker, non-
reductive versions of representationalism hold that changes in phenomenal
character lawfully co-vary with changes in content because, although distinct
from representational properties, phenomenal properties perform representa-
tional service.
Both strong and weak versions of representationalism about consciousness
face a number of serious objections (for details see Hu o, 2009). Arguably, a
major problem with all such accounts is that they a empt to understand basic
perceptual activity by illicitly importing features that in fact necessarily depend
on being a participant in sophisticated, linguistically-based practices (e.g.
having mental states with the kind of semantic content that requires assessment
by appeal to public norms and concepts, as in the a ribution of blueness to
aspects of the environment). If so, in imagining basic experiences to have
more properties than is necessary or possible for them to have, such accounts
make the opposite mistake to Denne and his followers.
Plausibly, having a capacity for phenomenal experiencing is more rudimen-
tary and fundamental than the capacity to represent the world as being a
certain truth-evaluable way. Consequently, experiencing aspects of the world
might be thoroughly non-contentful (and not just non-conceptual). Experi-
encing might not be intrinsically content-involving even though there is
something-it-is-like to experience worldly oerings in phenomenologically
salient ways.
This non-representationalist view of experience features as the central
plank of a radically enactivist approach to phenomenality; one that seeks to
understand phenomenal experience by focusing on the ways in which creatures
actively sense, perceive and engage with their environments (see Hu o and
Myin, forthcoming). Enactivists propose that the core features of experiential
properties are best explained by appeal to specific pa erns of sensorimotor
activity, through which complex self-organising systems interact with aspects
of their environment. Their slogan is: Experience isnt something that happens
in us, it is something we do (No, 2004, p. 216; see also Thompson, 2007). They
maintain that Experience is not caused by and realized in the brain, although
it depends causally on the brain. Experience is realized in the active life of
the skilful animal (No, 2004, p. 227). Thus enactivists challenge traditional
internalist thinking about the extent of the supervenience base of conscious-
ness, holding that it constitutively involves not just the brain but also bodily
and environmental features.
In pressing this idea, enactivists are critical of endeavours to understand
the phenomenal character of experience on a purely correlative basis, namely,
52
Consciousness
53
3 The Mark of the Mental
Fred Adams and Steve Beighley
Introduction
Whats a mind? What would it take to build one? There must be a dierence
between having a mind and not having one. What is it? This paper defends the
view that there is a dierence and tries to say what it is.
Why be interested? First, inquiring minds want to know. The question is
intrinsically interesting. It certainly seems that there is a natural divide among
biological1 systems that have minds and those that dont. If this is not an
illusion, we should be able to discover what constitutes that dierence.
Second, we commonly talk about minds and mental states. Is this just a
convenient fiction? Some people think we talk about minds because we dont
yet know the real story about why people (and animals) do the things they do.
Speaking about minds has practical predictive value in itself even if there
arent any minds. Horticulturists say that some plants like light or like dark
places, even though no one literally thinks plants have minds, have likes or
dislikes. Yet, while most people would accept that speaking of likes for house-
plants is a convenient fiction, very few think its fiction when talking about
Grandma or the kids (or even the family pet). Here, we seem more commi ed
to the a ributions of mind and mental states to people and pets being literally
true. We think these uses are literally true and our job is to figure out what
underlies that truth.
Third, science, the law, even other areas of philosophy pin important issues
on mental states. In the law, it is important to know if one who acts is legally
sane. In epistemology, some internalists about justification claim that justifica-
tion supervenes upon mental states. This requires knowing what constitutes a
mental state. Researchers interested in embodied cognition are making some
amazing claims in the scientific literature.2 We are told that the use of com-
puters, cell phones, PDAs, or even pencil and paper while doing a complex
math problem can involve cognition extending into the environment across
the boundaries of body and brain (Clark, 2009). The skin is an arbitrary
boundary, only observed by an outmoded Cartesian view of the mind. Some of
this may be true, but unless we are able to specify what counts as a cognitive
process, it will be impossible to evaluate such claims. For now, researchers are
54
The Mark of the Mental
able to make such bold speculative claims precisely because there is no agreed
upon account of what makes something a cognitive process. If these uses of
tools manipulate representations and information, and if thinking (cognizing)
is a kind of manipulation of representations, then why isnt tool use a kind of
cognizing?
Fourth, suppose we were going to try to build a mind. Many artificial intel-
ligence labs around the world are racing to be the first to build a computer or
a robot that can think. Governments are funding projects to build genuinely
intelligent agents to serve many dierent functions. To win the competition,
these centers must determine what it takes to build a mind, a cognitive agent.
There are, however, more and less radical views about why there might not
be a mark of the mental at all. Eliminativism asserts that there are no minds.
Mysterianism asserts that we can never know how a brain generates a mind.
We will not discuss these views here, but we do so elsewhere.3
A less radical possibility is that researchers cannot agree on what the mark
of the mental is or why people should want one. Susan Hurleys comment
is typical: Criteria of the mental or the cognitive vary widely (if not wildly)
across theorists; it isnt even clear what agreed work such criteria should
do (Hurley, forthcoming, 5). Kim too paints a dismal prospect for a unified
conception of mind, saying:
The diversity and possible lack of unity in our conception of the mental that
the class of things and their states that we classify as mental is also likely to
be a varied and heterogeneous lot . . . A question to which we do not yet have
an answer is this: In virtue of what common property are both sensory states
and intentional states mental. What do our pains and beliefs have in common
in virtue of which they fall under the single rubric of mental phenomena?
They of course satisfy the disjunctive property qualitative or intentional,
but that would be like trying to find a commonality between red and round
by saying that both red things and round things satisfy red or round. To
the extent that we lack a satisfying answer to the question we fail to have
a unitary conception of what mentality consists in. (Kim, 2006, pp. 267)
Hurley makes her point in the context of the extended mind debate, chiding
Adams and Aizawa (2008) for requiring a mark of the mental in order to decide
whether minds extend. She rightly points out that one can study the mind
without having an agreed upon mark of the mental. We know that vision,
memory, and decision making are mental, but if one wants to claim that mental
events extend beyond body and brain, one must have a notion of what makes
something or some process mental.
Kim (2006) argues that there is a dichotomy between sensory states
and intentional states. The former seem to take instances of properties as their
55
The Continuum Companion to Philosophy of Mind
contents (sweet tastes, round feels, loud sounds), while the la er seem to take
propositions as their contents (believes that the Democrats will win; hopes that
we will get out of Iraq). For Kim, a unified account of the mental would require
somethings being the same across these types of states. Kim is pessimistic about
this possibility.4
Incorrigibility
The first single property view that we will consider focuses on the property
of incorrigibility. Rorty (1970a, 1970b, 1972) and possibly Denne (1996) want
incorrigibility to serve as the mark of the mental. Rorty is careful to distinguish
occurrent states (my foot hurts now; Im now thinking it is time for lunch) from
standing states (my background desire for self-preservation; my background
belief that global warming is a bad thing). Given this distinction, Rorty limits
incorrigibility to only first person, reportable, occurrent sensations and thoughts,
admi ing that there is no single mark of the mental for all entities customarily
called mental (Rorty, 1970a, p. 409). According to Rorty mental does not apply
to standing states.5
On this view, incorrigible states are first-person self-reports about thoughts
or sensations, for example, Im tired now, Im angry now, and so on. Rorty says:
56
The Mark of the Mental
57
The Continuum Companion to Philosophy of Mind
We will present three simple objections to Rorty and then argue against
his whole approach. The first objection is the infants-and-animals objection.
On Rortys view non-lingual infants and animals dont have minds, but surely
thats not true. Of course, there have been those who have denied that animals
have minds (Descartes) or that they have conscious minds (Carruthers), but
most philosophers and scientists accept that both pre-lingual infants and non-
lingual animals have minds. Rortys view simply cannot accommodate this.
Infants and animals make noises, but they dont give first-person self-reports,
and so they are denied the incorrigible reports that are his hallmark of the
mental. They have plenty of occurrent mental events that are simply not
captured by Rortys criterion.
The second objection is what well call the Cog-objection. Denne (1996)
sides sympathetically with Rorty as he tells of the project to build an intelligent
agent named Cog at Rodney Brooks MIT lab. Cog is fi ed with cameras
for eyes, microphones for ears, and microprocessors linked in parallel for a
brain. The goal is to get Cog to think and to teach it language. Cog is a self-
reprogramming system, and at some point the reports that come out of Cog
may not match the interpretations placed on its internal states by its program-
mers. Denne agrees with Rorty: for Cog to think would be for Cog to make
incorrigible reports on its internal states.
We think at some point Cog may babble like a baby. At a later point Cog may
come out with It is hot in here. Suppose Cog has an internal mechanism that
is not unlike a quantum mechanism in this respect. To examine its internal
states is to change them. So, in eect, Cogs reports are incorrigible. Does this
mean that Cog satisfies the criteria for the mental? It would seem so, but
this seems like a limitation on us, on what we know or can know, not a break-
through in cognitive science.
Of course, Rortys ocial pronouncement had two conditions:
We suspect that Rorty had to mean that one knew (1) was satisfied in virtue of
a self-report. We know that Cog believes it is hot in here because he u ers
It is hot in here. If there is some other behavioural test for occurrent belief,
then Rortys theory collapses, assuming occurrent beliefs are mental states.
Condition (2) is satisfied by the nature of the internal mechanism that changes
58
The Mark of the Mental
the system upon any a empt to verify its current states. Of course, one could
object that unless we have another way to validate Cogs beliefs, we dont
really know that he made a report or satisfies (1). This takes us to our third
simple objection.
The third objection is that Rorty is not entitled to help himself to the idea
that any verbal u erance is indeed a report. Some Japanese cars (e.g. the 1986
300ZX) have a system that says door not closed, to warn the driver not to
drive until closing the door. Did the car make an internal report? We would
claim not. A sensor detected that the door was ajar and sent a signal to a voice
simulator that emi ed sounds interpretable by English speakers. To be a first
person report telling us the internal state of the car, the u erance must be made
with intention and purpose. These may have existed in the minds of the Nissan
engineers who designed the car but not in the car itself.
The same is true of Cog. If It is hot in here comes out of Cogs audio port,
is that a first-person self-report? We suspect that it is not any more than it is a
report coming out of the 300ZX. Reports are linguistic u erances.8 They are
intentional. They have meaning. They are for the purpose of communicating or
conveying information. This is the type of thing minds do. In essence, genuine
reports have to come from minds.
The more general objection is that Rorty has things the wrong way around.
If genuine reports come only from minds, then it cant be the logical features
of the reports, such as incorrigibility, that make them mental. There must be
something else underlying the ability to make the report that accounts for the
system having a mind, being a mental system.
Furthermore, we deny Rortys worry9 that the distinction between the men-
tal and the physical would collapse if we reject Rortys approach. True, if minds
turn out to be physical things, then the mental will be a subset of the physical,
but this doesnt mean that there is no dierence between a minded thing and a
non-minded thing. Minded things can be physical things arranged in dierent
ways from non-minded things, with dierent functions, dierent causal histo-
ries, and dierent internal and external behavioural capacities. Just because
minds may be physical does not mean there is no dierence between minds
and non-minds. The category of the mental, and therefore the meaning of the
mental, can be a subset of the physical. Gold is a kind, and it is physical. Water
is a kind, and it is physical. They are both physical, but that does not mean that
there is no significant categorical dierence between them. We see no reason to
worry that there will be no interesting conceptual dierence between the mental
and the physical if they both turn out to be physical, as Rorty seems to fear.
Once one takes seriously the notion that possibly everything is physical, even
minds, then Rortys fears vanish. If one takes seriously the thought that minds
are natural kinds, then it is up to science to discover what minds really are.
59
The Continuum Companion to Philosophy of Mind
Intentionality
60
The Mark of the Mental
But he seems to think that he gets o the hook by appealing to the phenomeno-
logical side of conscious mental states. We dont think he makes it o the hook.
If his concept of mind is the phenomenologically conscious (what is given),
why doesnt he make that the mark of the mental? If it is not, then how does
he rule out those information processing states in the brain as being mental?
For instance, some parts of the brain detect low blood sugar or cold extremities
and reduce insulin production or constrict the capillaries to hold blood from
the extremities. Why arent these mental activities if they are intentional states
of the brain? What about unconscious desires or beliefs? Surely these are
mental states even if they lack a phenomenology. We are sympathetic with the
idea that sensory states are intentional, but probably not because they are given
phenomenologically, rather because they are the right kinds of representations.
So we are on board with Crane (1989) until the very end where he acknow-
ledges intentional non-mental states. He needs a principled way to distinguish
the mental intentional states from the non-mental ones. Crane wants to do it
via phenomenology. We think there must be a be er way.
Furthermore, since Crane defends only a weak view that sensory states
are intentional, but their qualia may not be, he acknowledges that this leaves
him open to the question of what makes qualia mental. Qualitative states
certainly seem to be mental states in good standing, and they seem to be on a
par with phenomenological states generally. So it would be strange indeed if
a mark of the mental le them out. Realizing this, Crane acknowledges that
more needs to be said (Crane, 1998, p. 251, n. 26).
We believe accounts like those of Tye (1995) and Dretske (1995) have the
advantage of making the qualitative characteristics of conscious states them-
selves representational and hence intentional, thereby leaving no dangling
qualia unaccounted for by the mark of the mental. While there are some
important dierences between their views, for our purposes here we empha-
size their similarities.11 Consider the sweet taste of sugar in ones mouth. For
Dretske (1995), the qualia of experience arise due to an indicator function in
the sensory system. Although Dretske uses the term function in his theory, we
dont believe too much weight should be placed upon that term (Adams,
2003). What is essential is that there is a type of sustained causing (Adams,
1991). That is, the structure in the brain S that indicates the presence of sugar
will cause some other brain activity or bodily movement M (say, swallowing).
When the S causes M (rather than some contrasting N spi ing) because of
the indication of sugar by S, then and only then does S acquire the function of
indicating sugar. What is important is the sustained, contrastive causing. The
structure S must be sustained in its causing some relevant eect by the fact
61
The Continuum Companion to Philosophy of Mind
that it indicates the presence of sugar. If so, then S comes to represent the pres-
ence of sugar and makes the person in whom S does this conscious of sugar by
virtue of the qualitative experience of sweetness.
Tye (1995) defends a representational theory of the phenomenal mind on the
basis of a co-variation view of representation. The view states that something
S represents that P = df. If optimal conditions obtain, S is tokened in X if and
only if P and because P. In this definition, X is a place holder for a person and
P is a place holder for a proposition, due to the that on the le hand side of
the identity sign. The la er is a particularly bad choice because, first, proposi-
tions dont cause things. Events or instantiations of states of aairs may cause
things and may have propositional content, and may cause things because
they have propositional content, but propositions themselves dont cause
things. Second, Tye himself says the structure of phenomenal representation
is topographical (Tye, 1995, p. 120). To us this suggests that it would be far
be er to interpret P in the definition as a property instance. If so, then the
definition says that S (sensation of sweetness) represents P (sugar in the mouth)
if, under optimal conditions, S is tokened in person X if and only if there is
sugar in the mouth and because there is sugar in the mouth. Hence, the sweet
sensation represents sugar (in the mouth) because it is tokened when there is
and caused by sugar (in the mouth).
On the accounts of both Dretske and Tye, therefore, the qualia themselves
arise out of the representational role of the sensory states. It is because the states
are representational (intentional) that the qualia are as they are. Ones qualita-
tive experience of sweetness is itself an intentional state a representation of
sugar in the mouth under normal conditions.12
A further issue raised by Crane (1998) and Enc (1982) is whether intentional-
ity is a sucient condition for the mental. As weve seen, Crane embraces the
existence of two kinds of intentionality because he is willing to say that infor-
mation processing in the brain is a primitive form of intentionality but is itself
not mental. If this primitive intentionality is aboutness of the type generated
by any informational connection in the world, then it exists everywhere, not
only in the brain. Litmus papers turning pink is about a liquids being an acid.
A thermometers rising is about an increase in temperature. The falling barom-
eter is about a decrease in atmospheric pressure. This kind of informational
aboutness is everywhere in the world, and barring panpsychism it is not su-
cient for the existence of minds. So if this primitive intentionality13 is indeed
intentionality, it is not the right kind to qualify something as mental. Any
single property view of the mental that claims intentionality is the mark of
the mental is going to have to clarify what is the right kind of intentionality
to serve as the mark of the mental and explain why it is the right kind.
Fodor (1986b) introduces the notion of the detection of non-nomic properties
for precisely this type of reason. Lower organisms, such as paramecia, exhibit
62
The Mark of the Mental
63
The Continuum Companion to Philosophy of Mind
not explain this term, we think he means something like the specialization of a
property detector or concept.
Fitch also stresses that the nano-intentionality of cells or systems responds to
the novelty of their circumstances and records, in a non-mental sense, internal
changes based upon the organisms history of interactions. He maintains that if
the instructions for responding are already encoded in the organisms DNA,
then its responding to circumstances accordingly is not nano-intentionality.
He is interested in a biological kind of learning, or so it seems to us,
that depends on processes not simply inherited or primarily fixed by genetic
inheritance.
Fitch thinks genuinely mental representations are internal mental models
or possible worlds instantiated in neuronal firings (Fitch, 2007, p. 20). The
system has to be sensitive to the model, as well as, via the model, sensitive
to the world. These models also have to direct the systems behaviour that
is contingent upon tracking both the world and the internal models. Once
models are in place, then representation and misrepresentation are possible.
Hence, Fitchs view is that these models or complex representations are the
hallmark of the mental and only occur at higher levels of organization in the
nervous systems of animals.
Consciousness
We turn now to Searles view, which is actually not a single property view but a
system view, where the system itself must possess consciousness to be mental.
It is a system view because he appeals to the thesis of the background:
64
The Mark of the Mental
65
The Continuum Companion to Philosophy of Mind
of neurons (Searle, 1990a, p. 588). This may be true at the level of single neurons.
However, in macaque monkeys, at the level of single cell recordings, there
are cells that are sensitive to only goal-directed movements of other monkeys.
This seems aspectual. More importantly, Searle gives no reason why there
cannot be aspectual shape at the level of collections of neurons. A er all, he
insists that the only things that cause consciousness are the neurophysiological
properties of the brain. If these cause aspectual shape in conscious processes,
then there must be something specific in the neural structure that accounts for
the specific aspectual shape of ones conscious experience when accessed. Of
course, Searle may mean that the aspectual shape of conscious states only
exist as an emergent property in the conscious access of the unconscious state.
But since he thinks all conscious states are neural states, then at some level,
perhaps the conscious level of organization in the brain, clusters of neurons
must possess aspectual shape. They must be aspectually organized in their
firings, so to speak. To deny this would make Searle an emergent dualist,
something he would vehemently deny.
In addition, we worry that there may be mental states in good standing that
cannot be brought to consciousness. One type of case is blindsight, where an
individual who lacks a conscious presentation of a visual scene nonetheless
responds purposively by hand orientation in relation to a slit or negotiating
objects in a room (Weiskrantz, 1997). Another type of case is vision for action
(Milner and Goodale, 1995). Some actions guided by the dorsal stream in the
visual system are able to correct for visual illusions, such as Titchener Circles.
An object being reached for might appear larger than it is because of an illusion,
but thanks to dorsal stream processing ones reach is for the actual size of the
circle.17 Milner and Goodale, as well as Weiskrantz, have given examples of
states that are surely mental. They are guiding high-level purposive activity.
Yet they are states that seem not consciously accessible to the blindsight
subjects who cannot describe the scenes before them and whose dorsal stream
processing is guiding their action.
Our view
Our view is a systems view.18 We dont think that every state in a mental system
must possess a single type of property shared by all other core mental states.
Some states contribute to a systems having the property or properties that
make a system mental without themselves possessing the core property or
properties. We think that mental systems share a cluster of properties that we
will articulate below.
On our view one cannot properly say what minds are without saying what
minds are for. Minds allow organisms to track changes in their environment
66
The Mark of the Mental
67
The Continuum Companion to Philosophy of Mind
68
The Mark of the Mental
phylogenetically ancient part of the brains motor system) and serve as input to
thinking, concept application, and reasoning (Evans, 1982, p. 158). Evans says
this just a er discussing a case of blindsight. In this case a subject dierentially
responds to an external visual stimulus while reporting no conscious visual
field. This leads Evans to say if the sensory system only provided information
to the motor system, there would be li le interest in producing qualitative
states blindsight seems to have no need of them. In blindsight visual informa-
tion is processed, but there is no conscious perceptual experience or qualia.
This leads Evans (and perhaps Tye and Dretske) to maintain that it is only when
properly connected to a conceptual system, perhaps conjoined with a motor
system, that qualitative experiences arise. Only then will there be qualitative
visual fields, auditory experiences, and so on.
Why dont amoeba and paramecia have qualitative experiences? Fitch
would say that although they have nano-intentionality, they dont have internal
mental models. Evans, Dretske and Tye would say that they dont have sensory
systems which feed into conceptual systems. Even though all would agree
that information is processed at the first level of intentionality. We think that
there is a perfectly good sense in which these organisms dont need qualitative
experiences. The dierential responses that they make to environmental
changes do amount to a low-level processing of information, Fitchs nano-
intentionality, but the processes involved can all be explained at the level of
chemistry or photo-chemistry alone, plus a bit of history of the organism in
its local environment. There are no dedicated processors no biologically
selected structures that are recruited for the purpose of tracking information
about the environment or the internal states of the organism as they respond to
environmental changes. Thus, there are no more or less permanent structures
that have the biological function of indicating to the organism changes of envi-
ronment and self.20 Thus there are no internal structures that have the function
of both informing a cognitive system and driving a motor system that serves
the needs and desires of the organism. Having such internal, dedicated infor-
mation-processing structures requires explanation that rises above the level of
local chemical reactions. The organisms use of the information generates a
qualitative sensory experience.
It may seem like magic21 that a biological structure whose function is to
deliver information from a sensory transducer to a conceptual, cognitive
system should generate a qualitative visual, tactile, auditory, or gustatory
experience. This leads someone like Searle or Block to say that qualia must
arise from the brains neurochemistry. To us thats no less magical. How do the
chemicals do it?
Along with Dretske and Tye, we maintain that the qualitative nature of the
sensory experiences is explanatory. It is because of the way things look, taste, or
feel, that we do what we do when we experience them. In blindsight cases an
69
The Continuum Companion to Philosophy of Mind
individual may orient her hand vertically because a slit in front of her is
oriented vertically, not because of how it looks phenomenally. Whereas, when
we do the same thing with a full visual field, we do orient our hand because
of how the slot ahead looks. So the qualitative representational content of the
sensory experiences plays an explanatory role in our purposive behaviour.
Of course, how the light of qualitative states comes on is no less a mystery, even
if this is how or why it comes on in conscious sensory systems.
Since we maintain a systems view, and since we think there arent actually
any purely sensory mental systems, all actual minds must have concepts.
Concepts have a semantic cognitive content or meaning, and this content rises
to the level of non-derived semantic content. By non-derived we mean that
the content is not given by or dependent upon the mind of another. So, for
example, the first minds that arose did not derive their mental contents from
the mind of another. Internal structures had to acquire a content that was
meaningful to the organism itself. This means that the system interprets the
world on its own. This dierentiates minds from contemporary computers that
have only derived content, content supplied by the minds of the engineers who
build them, content meaningful only to the engineers, but not to the computers
themselves.22 This also means that the concepts in minds have to rise to the
second level of intentionality.
The best explanation currently available of how this goes is due to Dretske
(Dretske, 1988). Dretske explains how a structure or concept rises to the level
of non-derived meaning in the context of solving the dreaded disjunction
problem the problem of how C could come to mean just F when either Fs
or Gs cause tokenings of C. Dretskes solution to the disjunction problem
has at least two components. The first component is ge ing to the first level
of intentionality, being able to indicate that something is F without also
indicating that it is G. The symbol C must start out with the ability to naturally
mean Fs and not Gs. If all of its natural meaning is disjunctive, if it only
indicates Fs or Gs, then a disjunctive content is the only semantic content it
could acquire.
The second component is the jump to semantic content. Even if Cs indicate
Fs only, to acquire semantic content, a symbol must lose its guarantee of pos-
sessing what Grice called natural meaning. Smoke naturally indicates fire
because in the wild the two always co-occur. Hence smoke has the ability to
indicate the presence of fire. A symbol that has smoke as its semantic content
needs to become locked to smoke and permit robust and even false tokening
that means smoke without infecting its semantic content. It has to lose its
ability to always naturally indicate fire. Dretske appeals to the explanatory
relevance of the natural meaning. For Dretske, it is not just what causes Cs, but
what Cs in turn cause, and why they cause what they cause that is important
in locking Cs to their content (F).
70
The Mark of the Mental
71
The Continuum Companion to Philosophy of Mind
72
4 Substance Dualism
T. J. Mawson
Introduction
Substance dualism could not have a more venerable lineage, being traceable
back through Descartes at least as far as Plato and Socrates. However, the
respect with which people treat the view has declined to such an extent in the
last few hundred years that it has recently been described as, not so much
a position to be argued against, as a cli over which to push ones opponent.
Certainly within contemporary educated circles, were one to venture the
opinion that we have souls, one should expect to find oneself held to have
propounded an extravagance only slightly less great than had one ventured
the opinion that visiting extra-terrestrial life was in part responsible for the
construction of the pyramids or that Elvis may be seen working in the local chip
shop. The most favourable response one could realistically hope for would be
the concession that perhaps, before the development of such things as com-
puter science and neuroscience, such a whimsy might have been excusable, but
even so, now souls must surely go the way of phlogiston and light-carrying
ether: onto the intellectual scrap heap.1
Here I shall advance the claim that, despite the near universality of the
assumption that the theory may be easily cast aside, within the structure of a
hylomorphic substance/property metaphysic, the only reason to suppose that
we do not have souls is that provided by Ockhams razor and even that reason
is conditional upon an assumption, albeit an assumption that it is no more my
intention to cast doubt upon here than it is my intention to cast doubt upon
the substance/property metaphysical structure within which I shall be framing
this debate.2 The assumption is that there is physical stu.
Given that there is physical stu, it would indeed be simplest to suppose
that the mind is ontologically reducible to that or to processes going on in that.
But given that, as I shall also argue, there are some reasons to suppose that
we do have souls that is, that such a reduction cannot be accomplished, so one
might find oneself, probably idly, reflecting on the fact that idealist substance
monism would oer one all the advantages of simplicity oered by physicalist
substance monism while in addition accommodating these reasons for sup-
posing we have souls. This reflection would probably be idle as there is li le
73
The Continuum Companion to Philosophy of Mind
We have to start somewhere and time is pressing, so let us put onto the table
without oering argument in its favour a certain commonsense realism about
the physical world and our knowledge of it as gained through the natural
sciences. First then, let us assume that there is physical stu. This may be
characterized as stu of the sort that we suppose ourselves to encounter with
our five senses in everyday life; that our folk science describes more or less
adequately for our everyday purposes; and that our natural sciences describe
with increasing accuracy as they develop. We may define the sort of stu we
have in mind by paradigm examples of things which are made of it: this desk,
here; that star, there; and so on. In a previous century, we might have called this
physical stu simply ma er, but now we know that ma er may be converted
into energy and vice versa and we hear scientists speculate concerning quarks,
hyper-dimensional strings, and so forth as making up the more commonplace
objects that we encounter in everyday life. These are things which, while strik-
ing us as no doubt physical, do not strike us as in any obvious way material,
so, instead of ma er, we call this stu physical stu . We shall call the view
that this physical stu is all the stu that there is physicalist substance monism
or physicalism for short. Obviously one might hold that in addition to this
sort of stu, there is another type of stu as well. We shall call this second
view substance dualism.
Substance dualism is commi ed then to there being a type of stu that
resists full integration into the natural sciences. What we might call partial
74
Substance Dualism
75
The Continuum Companion to Philosophy of Mind
76
Substance Dualism
77
The Continuum Companion to Philosophy of Mind
makes one fundamental unit of substance dierent from another and how we
know of such units that they are the same over time), these problems cannot be
reasons to favour physicalism over substance dualism and so are not properly
construed as objections to it, rather than perhaps as objections to the wider sub-
stance/property metaphysic within which this debate is taking place. Secondly,
insofar as substance dualism faces problems not faced by physicalism (as it
does in addressing the issue of how we can ever know that units of substance
other than our own exist at all), the fact that it commits one to a certain sort of
scepticism here is a reason to suppose it true, not false.
The second area of concern centres around supposed problems in explaining
the causal interaction between the two sorts of substance the substance dualist
posits. How can mind and body act on one another? Does not any answer to
this question run into insuperable problems from what we already know of
physics, for example concerning the causal closure of the physical world and
the conservation of energy? In short, my analysis here will be that the inter-
actionist substance dualist is not beholden to answer the question of how mind
and body act on one another, rather than merely assert that they do, as it is
not a commitment of interactionist substance dualism that this question will
be answerable by us. Positing that there is an interaction of this kind does not
in fact require one to contradict things which we already know of physics,
although there is potential for physics (were it to move back into a deterministic
mode) to put pressure on the claim that there is in fact interaction of the sort
posited. At the moment then, there is no reason from science to suppose
interactionist substance dualism false.
Let us go into these objections in more detail.
Problems of identication
We may sensibly ask the substance dualist what it is that makes one soul
distinct from another and predict that he or she will have li le informative to
say by way of reply. Obviously, he or she may maintain that it is extremely
unlikely that two souls will have all the same properties as one another, so he
or she may point out any two souls will in fact dier in this fashion. One
will be thinking about strawberries, another, about cream, and so on. But exact
qualitative identity between two souls is not a metaphysical impossibility gen-
erated by the nature of souls per se and even if it were somehow impossible for
two souls to have exactly the same properties, this impossibility would not
ground the numerical dierence between two souls, but rather presuppose it.
In any case, it looks as if the substance dualist should agree that there is nothing
in the nature of souls per se that prevents there being two qualitatively identical
yet numerically distinct souls, for it seems that theres nothing in the nature of
78
Substance Dualism
souls per se that prevents there being an exact duplicate of this universe. In that
universe there would consequently be a person thinking qualitatively identical
thoughts to those that you are currently thinking. That person would, never-
theless, not be you; it would be your duplicate. So, the substance dualist should
say that it is not fundamentally in virtue of their dierent properties that
dierent souls are dierent. Rather, he or she should admit that souls might in
principle dier solo numero. (They have what is sometimes called thisness.)
Need he or she be embarrassed that he or she can say no more than this? I do
not think so.
Presumably the person who believes in units of physical substance will
wish to maintain that at least with regards to some of these there is nothing in
their nature that prevents their diering solo numero too. The classic thought
experiment on this topic involves imagining a universe composed simply of
two chemically pure iron spheres, each of the same diameter, hanging in other-
wise unoccupied space a certain distance away from one another; these spheres
would be qualitatively identical to one another, yet they would be numerically
distinct. Can the physicalist substance monist say more about how these two
spheres manage to retain ontological individuality than that they do, that they
dier solo numero or have thisness? No. So the substance dualist need not feel
embarrassed about being able to say no more than this about how two souls
might retain their ontological distinctness even were they to have qualitative
identity.
As this discussion might have already indicated, this type of issue and
in fact the one we are about to go on to discuss is an artefact of believing
in substance as such, (i.e. of believing in things to which the principle of the
identity of indiscernibles does not apply of necessity). As such, this type of
issue and the one we are about to go on to discuss cannot be a reason to prefer
any theory that claims that substances exist over any other that claims that they
do. Thus it cannot be a reason to prefer substance dualism over physicalism.
Belief in substance raises certain problems at the epistemic level. Of sub-
stance dualism, it is sometimes said, souls might be swapping bodies every
few minutes but each inheriting the psychological properties of the soul that
had just vacated the body into which the new one was now moving. Were
this to be the case, no one would be able to detect these changes, yet people
(Descartes) or significant parts of people (Aquinas) would constantly be swap-
ping bodies. Furthermore, we seem to face on substance dualism a peculiarly
intractable variant of the problem of other minds: how do you know, as you
encounter another person through the medium of the physical world, that he
or she is a person at all, that he or she has a soul in the right sort of causal
connection with the body which you observe directly?
Again we may observe that the first problem aects those who believe in
substance per se and thus in substance of the physical sort; thus, whatever it
79
The Continuum Companion to Philosophy of Mind
80
Substance Dualism
the happenings which are mind-happenings, interior to the skull, then, unless
we meet people who are themselves in fMRI scanners of sucient sophistica-
tion to reveal to us these happenings, we never ourselves see the happenings
that are, on the physicalist account, being angry, or what have you. On physical-
ism no less than substance dualism, we never observe the having of minds
other than our own. How then do we know, if physicalism is true, that others
have minds?
To cut a long story short, the answer to this question is that they tell us that
they do, and we ordinarily have no reason to doubt them. Someone says that he
or she is suering, let us say, from anger. If physicalism is true, they will be
speaking truly if a certain happening is occurring in their brain; but we do not
see this happening and indeed at the current stage of science might not know
that it was their feeling of anger, even if we did see it. But, unless we have
reason to doubt them (e.g. they are performing in a play or some such), we
are surely rational, whatever the theory of mind to which we subscribe, in
believing that they are angry simply on the basis of their saying that they are.
Without taking this sort of epistemic route into knowledge of others minds, it
would be impossible for the physicalist substance monist to construct the
theories by which he identifies to his satisfaction the having of anger with the
brain happening that he could then, in principle, find to be universally corre-
lated with the tendency to report it. (This is sometimes called the privilege
that must be given to first-person reports of the mental.) But if that is so, then
this same route is open to the substance dualist.
It is true that on the substance dualist view, the actual feeling of anger is
something happening in a substance even more recondite than the inner parts
of the brain. It is happening in a soul and thus in something that could never
be revealed by investigation into the physical world however advanced fMRI
scanners became. But the same route which the physicalist substance monist
takes in everyday life, before hand-held fMRI scanners and the like become
commonplace (and which he or she will have to hold as epistemically authori-
tative even were they to do so, to accommodate the issue of privilege), is open
to the substance dualist. This is how the problem of other minds is to be over-
come whatever ones theory of mind: by taking claims to have minds as a prima
facie reason to believe minds are had.
However, moving on to the second point, it seems as if the physicalist
substance monist may argue that whatever problems he or she faces in
coming to knowledge of others minds, and however these are to be overcome,
the substance dualist must face an additional problem unless he or she posits
some direct and very reliable telepathic contact between minds as an alterna-
tive source of knowledge, which positing would itself be most implausible.
This is a true point. But does it speak against or really in favour of substance
dualism?
81
The Continuum Companion to Philosophy of Mind
Were physicalism true, then, a er science has been completed and presum-
ing it has allowed for hand-held fMRI scanners or some such of sucient
accuracy let us call them brainoscopes one could perhaps confidently
bypass first-person reports as a source to knowledge of others minds; one
could, instead of speaking to a person, directly apply ones brainoscope to
someones skull and, on the basis of its findings, confidently report things like,
No need to speak; I see from my brainoscope that you are angry at my having
applied it to your head without first asking your permission. These reports
could be unfailingly accurate. (Note: not all physicalists believe that this will
prove possible, but we are considering the views of one who does in order to
point out the contrast with substance dualism and the extra problem of other
minds that it faces.) Let us consider a physicalist substance monist who con-
tends that, a er science has been completed, one will be in a position to know
that a certain brain state or some such may be identified with anger being felt at
having had a brainoscope applied without having been asked for permission
and, with the technology of the brainoscope properly applied, one will know
that this brain state is being had, so, one will know that the person is angry in
this way. For such a physicalist substance monist, there will then be no gap
into which a sceptical doubt may creep. It might appear that nothing similar
could happen on substance dualism. But, in fact, the substance dualist may
hold that it could. If substance dualism is right, then in a completed science this
technology might well be possible. The substance dualist of course would not
make the extra step of identifying the brain state or what have you that is
revealed by the brainoscope with the mental state, but he or she can acknow-
ledge that there might well turn out to be a perfect correlation of the sort the
physicalist we are considering anticipates our finding, and thus the substance
dualist might admit that the sort of brainoscope that is capable of bypassing
first-person reports in the manner described could well turn out to be possible.
But there is, nevertheless, it must be conceded, a gap for the substance dualist
here relative to his physicalist substance monist counterpart, a gap generating
an extra problem of other minds.
The extra problem for the substance dualist is generated because it will
always remain possible that the brainoscope is in error, even once the science
is completed and the brainoscope working (for all we know) properly, for,
according to the substance dualist, the brain state or what have you that the
completed science finds universally to be conjoined with a thought of a certain
kind (and we are supposing that this is what it will find) and that the braino-
scope correctly reports to be present in this case is not to be identified with
the thought of that kind. According to the substance dualist, one could know
everything about the physical world, yet not know without the possibility of error
what mental state a person was in (or indeed even if they were a person at all)
for there is according to substance dualism an ontological gap between the
82
Substance Dualism
physical world and the mental, a gap which may be bridged by causation, but
causation not being a conceptual relationship any particular bridge across
which may or may not hold and thus any particular judgements using which
may be in error. But now this extra problem of other minds for the substance
dualist looks more like an asset than a liability, for, as we shall see when looking
at Mary-type arguments for substance dualism, it is apparently possible that
someone might know everything about the physical world yet not know some-
thing about the mental, which appearance has to be ruled out as deceptive by
the physicalist substance monist we are considering.
Problems of interaction
The version of substance dualism on which we are focusing suggests that there
is two-way causal exchange between physical substance and soul substance.
This is o en held to generate problems for the view. First, it is suggested that it
runs contrary to a finding of physics. In particular, it looks as if the principle
that ma er/energy is conserved across a closed system such as the physical
universe must be violated if substance dualism of the interactionist sort is true.
Second, it is suggested that there is something problematic in general in any
case regardless of whatever physics might be telling us about non-physical
substances causing changes in physical ones and vice versa. We know, a priori,
that such is an impossibility.4 I do not find either of these two lines of thought
tempting.
Let us suppose for a moment, what we shall later see is in any case false,
that the interactionist substance dualist is commi ed to laws of physics being
violated. It does not seem that an objection arising from this commitment would
be any more than a restatement of the objection from the relative complexity of
substance dualism over physicalism. Obviously it would be simpler were the
universe closed and the laws of physics not violated, and that is indeed, we
have already conceded, a reason to suppose that it is so. We should not double
count this objection to substance dualism. In fact though, the interactionist
substance dualist is not commi ed to his or her souls violating natural laws.
With the advance of physics beyond determinism, another possibility arises.
The substance dualist may maintain that happenings in the brain which are
caused directly by the soul are caused in ways compatible with the preceding
brain state and the laws of nature, but these two not being such as to necessi-
tate what state emerges from them they are caused to be the particular way
that they are by the soul. That the brain be in state q, rather than state r, a er it
has previously been in state p is something which was always allowed for by
the preceding physical states (given indeterminacy), but, in fact, the substance
dualist may maintain, that it ended up in state q was caused by the relevant
83
The Continuum Companion to Philosophy of Mind
persons soul. It is no bar to this theory to point out, if such a fact can be pointed
out (and it is doubtful that it can be), that, of any individual sub-microscopic
event where such quantum indeterminacy plausibly reigns, it seems incapable
of producing cascade eects up to the macroscopic level which results in arms
being moved and so forth. For presumably some brain state leads to macro-
scopic happenings such as arms being moved, and this is made up at the sub-
microscopic level of many such quantum happenings. So, the substance dualist
may maintain that the souls influence on the brain, in causing it (e.g. to raise
ones arm, occurs in a number of disparate tiny locations, any one of which is
perhaps not sucient, or perhaps even necessary, for the event to occur, but
which then jointly cause ones arm to rise). Those quantum happenings in the
brain which are similar in the properties they reveal to the natural sciences as
those happening in an inanimate object where they are indeed uncaused are
in fact, when they happen in the animate object that is the brain, caused by
the soul of the relevant person. The universe is not indeed causally closed, but
no laws of nature need be violated.
So, in short, even were fundamental physics to return to a deterministic
mode, the interactionist substance dualist could maintain that souls are able to
influence physical stu (and vice versa) although by doing so he or she would
be positing that the laws of physics are violated li le bits of energy come into
and go out of existence. However, within the current indeterministic paradigm,
no such violations are required as a part of the substance dualists account of
this interaction. The substance dualist may maintain that the soul operates
in the causal gaps, otherwise filled by randomness, that indeterminism opens
up. And of course even were the dominant paradigm of interpretation of the
laws of nature within the community of physicists to revert to determinism,
it would still be just a paradigm of interpretation; there would be no necessity
that the substance dualist follow it.
Of course, such suggestions on the part of the substance dualist presuppose
that in general a spiritual substance may cause a change in a physical substance
and vice-versa, and someone might hold as a ma er of principle that the only
possible relata of causation are physical events, so such a suggestion may be
ruled out in advance. But why adopt such a principle? It may be rejected by the
substance dualist as mere prejudice if argued for a priori (although of course if
argued for validly a priori, the substance dualist will need to find one or more
premises to which to object) and the substance dualist will insist that such a
principle cannot be discovered a posteriori, for the actual universe is one which
has souls operative in it, so does not follow it. Descartes himself said all that,
it strikes me, needs to be said on this issue in a le er to one of his objectors.
84
Substance Dualism
But I will say, for your benefit at least, that the whole problem contained in
such questions arises simply from a supposition that is false and cannot in
any way be proved, namely that, if the soul and the body are two substances
whose nature is dierent, this prevents them from being able to act on each
other. (Descartes, in Co ingham, vol II, 1994, p. 275).5
Various arguments in favour of substance dualism have been put forward over
the last two and a half thousand years, and it would be impossible to provide
an adequate treatment of all of them in anything smaller than a substantive
book. That being so, in the space that remains for me, I wish to focus on just
three areas where, it strikes me, the substance dualist can plausibly contend
that substance dualism does be er than physicalism in accommodating various
commonsense intuitions we have about ourselves. Of course commonsense
intuitions are hardly the basis for conclusive arguments in favour of substance
dualism. A er all, if our commonsense intuitions about such issues were not
sometimes wrong, there would hardly be any point in the discipline of meta-
physics. I conclude then by discussing what weight we may in general give
to this type of argument relative to the weight we may give to the virtue of
simplicity which, it has been conceded, physicalism has over substance dual-
ism. The three areas are personal identity, freedom, and consciousness. I shall
consider them in order.
Personal identity
What is it that makes a person at a later time, t+1, the same person as existed at
an earlier time, t? Substance dualism has a simple answer: it is fundamentally
the continuity of the same soul (or, for Aquinas perhaps, the same soul and the
same body), and souls themselves do not continue in virtue of anything more
basic continuing (bodies presumably do). For the physicalist substance monist,
the issue is more complicated: there are three options. The person may be
identified with a certain set of properties (usually psychological properties
are chosen); with a part of the physical substance which makes up his or her
85
The Continuum Companion to Philosophy of Mind
body (usually the brain is chosen); or with a combination of these (e.g. psycho-
logical properties p going on in brain b). However, none of these options seems
to oer a satisfactory theory of personal identity. There are problems peculiar
to each, but a general defect may be observed in play in their dealing with
almost all the thought experiments that are used, it is supposed, to illuminate
this issue.
So, for example, one is asked to imagine a brain bisection, a er which the
two resultant hemispheres are transplanted into separate clones of the original
body where they take up more or less functional residence. To add weight to
the situation, perhaps one of the resultant people is then tortured to death
over the next five minutes while the other is given a gin sling to enjoy. Which
of these two resultant people, if either, is the person who originally underwent
the brain bisection? one is asked. Then the details of the experiment are altered;
perhaps one of the two resultant people gets more psychological continuity and
the other more of the physical substance of the original brain. What then do
we say? For some proportions of psychological continuity and continuity of
physical substance, the physicalist must say that it either becomes ontologically
indeterminate whether a resultant person is the same as the original, or it
remains determinate, yet he or she does not know whether he or she is the
same or instead a new person inheriting some of the originals psychology
and/or brain ma er. But our commonsense intuitions about personal identity
do not allow for indeterminacy, as shown most markedly when one thinks of
these possibilities from the first-person perspective of someone about to
undergo the relevant experiment: Either I will survive or I wont; it cannot be
ontologically indeterminate in a few minutes time whether Im there or not.
But nor is there anything unknown le for the physicalist to hang a determinate
fact of personal identity from, something which again we might perhaps see
most sharply by imagining the first-person perspective: If I can know where all
the properties are going and know where all the physical substance is going,
yet still not know where I am going, then I cannot be identified with any
combination of properties or the physical substance; I must be something else,
and the only something else le (once weve swept properties and physical
substance o the table) is soul substance. This is not conclusive of course, for
one could be unbeknownst to oneself identical to some indivisible property
or indivisible bit of physical stu and thus even if one knew in advance of this
property/bit of stu where it was going to go, one would not know that in
knowing this one was knowing where one was oneself going.
However, each of these claims that one is to be identified with an indi-
visible property or an indivisible bit of physical stu would itself be most
implausible. Properties and sets of properties (whether properties of physical
substance or soul substance) are capable of multiple instantiation, and the
sorts of sets with which people might most plausibly be identified (that go into
86
Substance Dualism
Freedom
87
The Continuum Companion to Philosophy of Mind
range is greater than one we do genuinely have options and, when we have
options and end up realizing one rather than another as a result of the right
sort of conscious choice on our part, we suppose that in that way the causal
and moral buck stops with us. We are in this way free agents, responsible to a
greater or lesser extent for the choices we make and thus for the shape of our
lives and the lives of those we aect.
Substance dualism of the interactionist sort gives a straightforward
and simple account of how all of this gets to be so. (That is the long-promised
reason why interactionist substance dualism, rather than, for example, psycho-
physical parallelist or epiphenomenalist substance dualism is the most plausi-
ble.) According to interactionist substance dualism, the soul, while of course
being aected by things going on in the physical world (e.g. in coming to the
beliefs that it has about that world), is not always necessitated to do what it
does by those eects; sometimes it initiates causal chains, which then impinge
upon the physical world when it could yet have initiated dierent causal chains
and thus impinged dierently, had it chosen to do so. When my soul does
so, that is me (Descartes) or a part of me (Aquinas) making a choice. The com-
monsense view of ourselves as articulated in the previous paragraph finds its
metaphysical grounding.7
Physicalism cannot ground this commonsense view. On physicalism, either
what I ended up doing was entirely causally necessitated by preceding states
extending back though time to the big bang or there was a certain amount of
randomness (uncaused-ness) involved in the causal chain that ended up with
my doing whatever it was I did. In neither case would the causal and, one
might hence think, moral buck stop with me; either the happening was
caused by factors beyond my control (for they go back to the big bang, which
is certainly beyond my control); or it was random; or it was some mixture.
Various accounts of how the moral buck might stop earlier than the causal
buck and in the right spot me have of course been advanced by physicalist
substance monists keen to accommodate moral responsibility to their world-
view. So, for example, one might say that if my body does what I want it to
do as a result of me wanting it to do that thing, then thats me being morally
responsible for the doing of that thing, and the fact that my wanting it to do that
thing rather than something else was itself caused by factors beyond my control
does not detract from that. This account is open to easy counter-arguments, but
there are of course much more sophisticated accounts. However, they all suer
from the common feature that whatever psychological states are posited as
sucient to lead to the agent being morally responsible, it seems possible to
imagine a skilled enough hypnotist inducing those states in a person and yet
we not hold such a victim of such hypnosis to any extent responsible for the
actions that then flowed from these states. In cases where we can identify causal
responsibility, moral responsibility, we think, falls straight through to it; we are
88
Substance Dualism
strongly commi ed to the causal and moral buck stopping in the same place.
Substance dualism of the interactionist sort is the view that accommodates
this strong commitment in oering a third way to causal necessitation of the
physical sort stretching back beyond our births and randomness: my actions
are caused by me (Descartes) or my mental part (Aquinas). Substance dualism,
then, gives the best theory of freedom by reference to our commonsense intu-
itions about ourselves as being the initiators of and thus morally responsible
for our actions.8
Consciousness
The classic thought experiment here concerns someone called Mary, who,
we are asked to imagine, has been brought up in an entirely black and white
room. In this room she has access to black and white science textbooks and
science is now completed. She thus learns everything there is to know about
the physical properties of colour and indeed, let us say, about the physical
properties of brains too. She then leaves the room and goes into the outside
world. For the first time, she herself sees a red apple. Is it not plausible to sup-
pose of her that she thereby learns something new: what red looks like? We
may call this new fact a fact about red qualia: what it is like to see red. From
the fact that Mary ex hypothesi knew everything about the physical qualities
of the colour red and the brain prior to leaving her room yet did not know
about this qualiatative, as we may call it, property, so we can conclude that
this qualiatative property is not a physical property of red or the brain; of what
is it a property then? The substance dualist has a ready answer: of red as it is
experienced by the soul.
There have been various physicalist responses to Mary-type thought
experiments; they tend to deny the fact that Mary comes to know about a
qualiatative property; rather, they tend to assert, she comes to have an ability
which she did not previously have, the ability to recognize red objects in a
new way.9 This however seems wrong-headed to me, for Mary plausibly will
not gain the ability to recognize red objects simply by ge ing out of the room
and seeing a red apple for the first time. She will only gain that ability once
someone provides her with information in the following manner: That apple
youre looking at, Mary, its red. In hearing someone say that, she will plausibly
gain a new ability to recognize red objects therea er, but she had already come
to know what red objects looked like prior to hearing someone say that, just by
looking at the red apple. She wouldnt say back to the person whod just said
this to her, Now, for the first time, I know what red is like; shed say something
like this: Now youve told me that that apple is red, I realize that I already
knew just by looking at it what it was that red was like, rather than what it
89
The Continuum Companion to Philosophy of Mind
was that blue was like, and so on. But although I didnt know that it was red,
the qualiatative nature of which I knew about by looking at the apple prior to
your telling me, it was red that I had discovered something new about simply
by looking at the apple. 10
Of course the analysis presented here has been, perforce, terribly brief
(cannot property dualism deal with the issue of consciousness to which we
have recently adverted?), but, even so, it appears that the facts of personal
identity, freedom of choice and consciousness, as they present themselves to
commonsense are, when taken together, easily accommodated by substance
dualism and fail to be accommodated by physicalism. Either these facts are
not facts at all commonsense is wrong or physicalism is wrong.
Conclusion
90
Substance Dualism
our findings thus: theres one argument against substance dualism (its more
complex) and three in favour (it be er explains personal identity, freedom of
choice, and consciousness). If that is so, neither substance dualism nor physical-
ist substance monism will give us everything we want, and we shall naturally
turn to considering how we should weigh simplicity against these other consid-
erations when deciding what we have, on balance, most reason to believe.
Moore has taught us that we may take any valid argument in either of two
directions, as articulating a reason to suppose its conclusion true or a reason to
suppose one or more of its premises false, and that the direction in which it
is most reasonable to take a given argument will depend on whether the
premises are jointly more obviously true than the conclusion is obviously false.
So we may give the considerations presented here some direction by finally
asking ourselves this question: Knowing now that you can only believe one,
which of the following seems more obviously right to you?
z We are persons in more or less the same way that commonsense suggests;
we have freedom of the sort supposed in everyday life; and colours
and indeed mental happenings in general have qualiatative properties.
z The world is as simple a place as physicalism suggests.11
91
5 Physicalism
Barbara Montero
Physicalism, as some see it, takes the fun out of life. In their eyes, if physicalism
is true, the pleasure of a great bo le of wine, the euphoria of that first kiss,
the thrill of a hole in one and so much more are nothing but the workings of
the brain. At the same time, physicalism is probably the most widely held
general philosophical theory of the nature of the world, and many of those
philosophers who think that physicalism takes the fun out of life still defend
it tooth and nail. But what exactly is the theory of physicalism? Here I hope
to make some headway towards understanding physicalism, the theory that
many philosophers both love and hate. In particular, I aim to arrive at an under-
standing of the thesis of physicalism that captures its essence and at the same
time can be used to ground the contemporary debate over whether it is true.
Physicalism is a view about the ultimate nature of the world along the lines
of Thaless view that all is water or Democritus view that all is atoms in the
void. But rather than pronouncing all is water or all is atoms in a void, physical-
ism pronounces that all is physical, or as it is usually phrased, everything
is physical. Of course, this isnt very informative unless you know what it
is to be physical. Indeed, each term everything, is, and physical is
open to various interpretations. In what follows, I examine each of these
components in turn.
92
Physicalism
mental is physical. Indeed, some may even simply refer to the theory that the
mental is physical as physicalism. It may be that this is simply intended as
shorthand for the view that everything (or some significant subset of every-
thing) is physical. Yet this shorthand can be confusing when a more encom-
passing type of physicalism is evoked to justify physicalism with respect to
the mental, such as when physicalists argue that the mental is very likely to
be physical because everything else is physical. Obviously, here the scope of
everything else is not just the mental. So what, then, is supposed to count as
everything else?
Some understand physicalism in the broadest sense possible. It is theory
about everything whatsoever, a theory that says that all reality is physical. On
this inclusive conception, physicalism implies not only that people, animals,
rocks, trees, and all other concrete objects are physical, but also that abstract
objects which on some accounts include numbers, properties, classes, rela-
tions, and propositions are all physical. Even God, if she exists, would need
to be deemed physical given the truth of this conception of physicalism.
Others think that physicalism ought to have a more restricted scope. For
example, some understand it as a theory about only the concrete world, that
is, roughly about phenomena in space or time. Physicalism, then, is true if and
only if all phenomena in space or time are physical. This understanding of
physicalism ensures that the status of the mental is relevant to the truth of
physicalism, since, whatever else they are, mental processes do seem to occur
over time. However, the existence of abstract numbers (regardless of what they
are like in other respects) would not refute such a physicalism. Jerey Poland
can be seen as defending this conception of the scope of physicalism (if we
assume, as many do, that the abstract world has no causal influence on us)
when he claims that physicalists are (or should be) concerned with what exists
in nature that is, with what can be spatially and temporally related to us, with
that with which we can interact and by which we can be influenced, and with
that of which we and the things around us are made (Poland, 2001, p. 228).
A related approach to defining the scope of physicalism is to think of physi-
calism as a theory about the empirical world, that is, about the phenomena that
we come to know via our senses, or to put it more carefully, about phenomena
that are such that our knowledge of them must be justified via our sense experi-
ence. If, as is o en thought, our senses do not justify knowledge of abstracta,
this restriction allows for the existence of non-physical abstract entities to be
consistent with physicalism. However, if abstracta are known via our senses,
then, the truth of physicalism, on this interpretation, implies that abstracta are
physical.
A more encompassing view, such as Andrew Melnyks, takes physicalism
to be a theory about the contingent and/or causal world (Melnyk, 2003). If
abstracta are not causal or if they exist necessarily, this restriction comes close
93
The Continuum Companion to Philosophy of Mind
94
Physicalism
When the physicalist claims everything is physical, what is being said about
everything? Typically physicalists deem something physical if its existence
depends in the right way on basic or fundamental physical properties.1 And
typically the fundamental physical properties they have in mind are the micro-
physical properties countenanced by physics, such as the property of having a
charge, of being a quark, and so forth. In the third section I shall question this
conception of the fundamental physical properties. Here, however, I want to
ask, what exactly is the relation between the fundamental physical properties
and higher-level properties, such as mental properties, which is thought to
make the higher-level properties count as physical? In other words, when
physicalists say that everything is physical, just what is meant by is?
Some hold that the relation between higher-level physical properties and
fundamental physical properties is that of explanation (Jackson, 2006; Witmer,
2006). On this view it is thought that physicalism is true if and only if every-
thing is either a fundamental physical property or law, or can be explained
in terms of such properties and laws. As such, physicalism is an epistemic
thesis about what we can explain. It may have ontological implications since
typically we think that a good indication of whether the fundamental nature
of r is p is the fact that we can explain r in terms of p. Nonetheless, such a view
is primarily an epistemic thesis.
Many philosophers, however, see physicalism as an ontological thesis, a
thesis that tells us about what the world is like, whether or not we can under-
stand how it could be like this. Physicalism, many think, could still be true
even if we never arrive at a physical explanation of, say, pain, as long as pain
is an entirely physical phenomenon. As Joseph Levine puts it, I am prepared to
maintain that materialism must be true, though for the life of me I dont see
how (Levine, 1998, p. 475). And some philosophers such as Brain Loar (1990)
and Colin McGinn (1989) have proposed theories about why we cannot
understand physicalism could be true of the mind, even though they think that
physicalism might very well be true.
To make sense of positions such as these, physicalistic dependence relations
cannot be formulated in terms of explanation. Of course, most advocates of
thinking about physicalism in terms of explanation do not mean that we can
95
The Continuum Companion to Philosophy of Mind
96
Physicalism
the world was to create the fundamental properties of physics. A er this she
could rest, as everything else came along for free.
How close does upward determination take us to physicalism? Upward
determination states that any world that duplicates all the fundamental
physical properties and laws of our world also duplicates all properties of
our world. But now imagine a necessarily existing God. A world that dupli-
cates all the fundamental physical properties of our world would also dupli-
cate such a God. Yet, intuitively, the existence of God refutes physicalism. If
this is correct, then upward determination is not a sucient condition for
physicalism.
To be sure, if this necessary God interferes freely with the workings of the
world, a fundamental physical duplicate of our world might not duplicate all
aspects of our world, for God might arrange things so that in the duplicate
world, although all the fundamental feature of the world are the same, I prefer
coee to tea. As such, upward determination would fail. However, if the role of
God were merely to set the fundamental nature of the world, merely to be the
hand behind the big bang, as it were, then a necessarily existing God would
be consistent with upward determination.
If you accept Humes view that there are no necessary connections between
distinct entities, then such a God cannot exist.3 Such a God is distinct from the
rest of the world, yet her existence is necessary, given the world. Alternatively,
one could restrict the scope of physicalism so that such a God would be consis-
tent with the truth of physicalism. But if you reject Humes view and also think
that the existence of God is incompatible with physicalism, you are led to reject
upward determination as perhaps a necessary condition for physicalism, but
not a sucient one.
The desire to find both a necessary and sucient condition for physicalism
has led some philosophers to hold that explanation plays a role in our under-
standing of physicalism a er all.4 Physicalism, as they see it, is not just the
view that everything is determined by fundamental physical properties, but
that everything is determined and ultimately explained by the fundamental
physical properties. Such a view presumably rules out a necessarily existing
God from counting as physical. And if it doesnt, such a God would seem to be
physical.
But many are content with a mere necessary condition since much of the
action in the literature on physicalism involves various arguments against
physicalism, all of which purport to show that upward determination, which
is taken to be a necessary condition for physicalism, fails to hold. For example,
the zombie argument against physicalism is intended to show that the pos-
sibility of zombies not the lumbering Hollywood variety, but creatures that
duplicate our microphysical structure yet lack consciousness implies that
consciousness is not physical.
97
The Continuum Companion to Philosophy of Mind
The Physical
Now we must address the question, what is the physical? When we say, for
example, that everything is determined by fundamental physical phenomena,
what are these fundamental physical phenomena? Most define the fundamen-
tal physical properties in terms of the entities and properties and perhaps
laws posited by microphysics: the fundamental physical phenomena are those
entities and properties mentioned in the theories of microphysics. But what
is meant by microphysics? Is it current microphysics? This would provide a
relatively clear position: physicalism would then be the view that all of the
fundamental properties are properties of microphysics. Unfortunately, this is a
theory that is rather dicult to accept since we know that current microphysics
is most likely neither entirely true nor complete and thus we now know that
it is most likely not true that all higher level properties are determined by the
properties of microphysics.
A more common understanding of what counts as the fundamental physical
properties in the thesis of physicalism is that they are the properties posited
by an ideal physics, a true and complete physics, or a physics in the end.
Can we formulate physicalism in terms of a true and complete physics?
Of course, we do not currently know what future physics will be like, and
therefore we cannot now determine whether physicalism is true. But perhaps
physicalism can be seen as a hypothesis that awaits scientific confirmation (or,
for that ma er, refutation). Physicalists, on this understanding, are be ing that
it is correct, but do not claim to be able to now determine that it is correct.5
I see no problem with making physicalism a thesis that awaits empirical
support. However, it seems that far from turning physicalism into a thesis
whose truth awaits empirical support, defining the physical in terms of a true
and complete physics actually seems to turn physicalism into a trivial truth.
For what is a true and complete physics, save for one that accounts for the
fundamental nature of everything? If free floating souls exist in our world, a
completed physics will, by definition, account for the most fundamental nature
98
Physicalism
of these souls. Yet neither physicalists nor their foes think that at this time in the
debate physicalism is true merely as a ma er of definition. Physicalists think
the thesis needs to be argued for and, as many hold, will ultimately depend on
what scientific investigation reveals. And their foes clearly do not think that
they are denying what amounts to, more or less, an analytic truth. It seems,
then, that physicalists who define physicalism over a true and complete physics
cannot simply mean by this a theory of everything since then their claim that
the mind is physical is trivially true. Yet, there is also reason to think that they
do not simply intend to refer to the temporal end of physics. For this physics
might still be inaccurate and incomplete; even worse, for all we know, physics
might regress. We need, then, another route to defining the physical.6
Some argue that there are phenomena that physics and perhaps scientific
investigation in its entirety does not aim to cover. Rather, physicists, they argue,
in their role as physicists, are only concerned to account for a certain class of
phenomena and souls and spirits are not in this class. As such, the truth of
physicalism becomes open to debate. The question, then, is: Are there no other
fundamental properties than those that are under the hegemony of a true and
complete physics, where what counts as being an object of study for physics
is restricted in certain ways?
This makes physicalism admirably more risky, but should we assert that
physics has identifiable limits (besides, of course, that which is by definition
unknowable)? As I see it, it makes good methodological sense to hold that
scientific inquiry should not accept a priori barriers. Certainly, it would be
reasonable to say that as things stand, government grant money ought not to
fund physics research into the properties of souls. This research would seem
to be currently hopeless. However, the claim that physics should never investi-
gate the nature of souls even if in some currently unfathomable way a physics
lab reveals signs of souls is a much stronger claim. And, indeed, it seems
that such barriers could hinder progress. In other words, it seems that a good
approach to scientific investigation is that when you discover territory that
does not conform to your map, change the map, not the territory. Such changes
might involve not only expanding our scientific ontology, but changing our
scientific method as well. For example, if standard controlled experiments fail
to reveal phenomena that we nonetheless think exist as some have claimed
could be the case with parapsychological phenomena we should try to find
a way to change the control. If we were somehow convinced that there
was a spiritual realm that was causally isolated from our world, let us try to
understand it.
Where does this leave us? I think that it indicates that, despite the consonance
of the two terms, the physical should actually not be defined over physics.
Physics is the study of the fundamental nature of the world, whatever that
nature may be. But physicalism is more discriminating about what is to count
99
The Continuum Companion to Philosophy of Mind
100
Physicalism
101
6 Folk Psychology and
Scientic Psychology
Barry C. Smith
102
Folk Psychology and Scientic Psychology
having a desire to escape the party unseen would not lead her do anything
unless she also believed she could leave the party unseen by exiting swi ly from
the room.) Similarly, we could have cited a belief, She tiptoed past the drawing
room because she thought no one would hear her leave. Again, the explanation
is partial, and we are assuming that she had a desire to leave without anyone
hearing her. We may mention a belief or a desire even though it is beliefs and
desires together that constitute someones reasons for acting, and in citing these
reasons we are oering rational explanations of an agents behaviour.
The beliefs and desires posited to explain behaviour, are hypotheses that can
be re-worked in the light of further evidence. The beliefs ascribed must make
sense in the light of other beliefs, and similarly, desires must make sense in the
light of further desires. Thus it is part of our competence in giving such expla-
nations that we stand ready to adjust our a ributions of belief and desire if they
are not consistent with what else it makes sense to ascribe to an individual on
the basis of what they say and do elsewhere and at other times. The overall
picture must make sense of the individual as by and large a rational thinker,
and these relations between beliefs and beliefs, between desires and desires,
and between beliefs, desires and actions, are logical or rational relations.
We call beliefs and desires propositional a itudes because an agent can take
dierent a itudes such as believing, desiring, hoping, or fearing towards
the same proposition, for example, that the war on terror will continue for a
long time. Alternatively, one can take the same a itude towards dierent prop-
ositions. It is the logical relations between propositions believed that ensures
the rational connections between propositional a itudes. For example, if we
believe that the US president J. F. Kennedy was assassinated we must also
believe that he is dead, as well as believing that there was someone called
J. F. Kennedy who was president of the United States, etc. If Oswald intended
to kill Kennedy then he must have wanted Kennedy dead and believed that
by firing the gun he could bring about that outcome (i.e. he must believe the
death of Kennedy would satisfy that desire). If he didnt want Kennedy dead
(but wanted merely to frighten him) then we can suppose his intentional action
is incorrectly described as his killing Kennedy even if the shot he fired acciden-
tally resulted in Kennedys death. Perhaps he was genuinely surprised that
Kennedy died a er all, it was a very improbable shot. His subsequent surprise
would make no sense if he all along desired to kill Kennedy and believed that
pulling the trigger would result in Kennedys death. Even if Oswalds observ-
able behaviour, his li ing the rifle, aiming and pulling the trigger was the same
in both scenarios, his surprise would be a reason for supposing that he didnt
103
The Continuum Companion to Philosophy of Mind
intend to kill Kennedy, and for denying that we should call his action a killing.
By contrast, his lack of surprise and satisfaction would make rational sense of
supposing he did intend to kill Kennedy, and that killing is the correct descrip-
tion of his action. The correct description the description under which an
action is intentional depends logically on the intentions of the agent: the par-
ticular beliefs and desires he or she has. These propositional a itudes rational-
ize the agents behaviour, enabling us to see it as an intentional action. Thus if
the intentional description of the action were dierent, the a itudes we ascribe
to the agent would have to be dierent too. In courts of law whether someones
action should be described as murder or manslaughter depends on whether
they had a premeditated intention to kill before they acted as they did. The
behavioural event is the same but whether it counts can be intentionally
described as an act of murder or an act of manslaughter depends on which
mental states it is correct to a ribute to the agent (i.e. what his reasons were in
behaving as he did).
We make sense of peoples beliefs and desires in the light of further beliefs and
further desires we have grounds to a ribute to them, based on what they do
and say elsewhere and at other times. And we aim for the most consistent over-
all interpretation of someones actions and u erances by constructing a net-
work of intentional a itudes and actions that makes best sense possible of their
overall behaviour. We constantly rework and revise our portrait of someones
mental life in the light of further evidence. We find ourselves equipped to do
so without any explicit training. We operate quite instinctively in forming views
about other peoples states of mind. Premack and Woodru (1978) coined the
term theory of mind for this set of abilities. It is not an explicit theory, of course,
but it amounts to a tendency or ability on the part of normal human thinkers
one we develop fully by about the age of four to make hypotheses about the
beliefs, desires, hopes and fears of fellow humans and to explain and, to some
extent predict, their actions. It is the nature and status of this everyday folk
psychology that we will now examine.
Folk Theories
Folk psychology plays much the same role in dealing with others, as folk
(or naive) physics plays in our understanding of physical objects and forces.
We know that folk physics is not literally true. We, the folk, say that the sun
goes down behind the hill even though it is the earth that goes round the sun.
104
Folk Psychology and Scientic Psychology
We say that cold water cools the hot water in our bath, even though what
actually happens is that the hot water heats up the cold and thus loses its kinetic
energy. In these cases, the proper scientific explanation replaces the rough and
ready generalizations of folk theorizing about how the world works.
Is the same true of the generalizations of folk psychology? Will they eventu-
ally be replaced or revised when we learn more from the science of the mind?
Will neuroscience gradually replace the false but appealing assumptions of
the folk? If that is not the model of the relation between folk psychology and
scientific psychology, will belief-desire psychology be vindicated or refined
by scientific psychology? The worry is that if folk psychology is not reducible
to scientific psychology we seem to have competition between two explana-
tions of the same purposeful behaviour and perhaps only one of them can be
genuinely explanatory.
To ask whether the particular claims and generalizations of folk psychology
are true we need to know what they commit us to, and to know what kind of
explanation folk psychology provides. As we have already seen, folk psycho-
logical explanations of behaviour are rational explanations: the giving of reasons as
to why people do what they do. But such explanations are also causal explanations.
The reasons we cite the beliefs and desires that make sense of people acting as
they do are also the causes of their behaviour. When we say Charlo e le the
party because she wanted to catch her train, the because is used in a causal
sense. (Contrast this with, She broke the law because she parked on a double
yellow line.) But if folk psychological explanations are intended as causal expla-
nations explanations of what brought about certain events the key question
is whether folk psychological explanations are literally true of human agents?
Are our descriptions of our own and others behaviour as rational actions
really true: is behaviour really the upshot of beliefs and desires at work in us?
Or are we actually caused to behave as we do by something entirely dierent:
configurations of neural firings in the brain that have nothing like the neat
structure of belief and desire? What aspects of reality are we picking out when
we make ascriptions to people of beliefs and desires, citing these states as the
causes of their actions?
Let us look at the eliminativist challenge. What if states like belief, desire, hope
and fear, cited in our common sense psychological framework, turn out at
no level of organisation to be among the causes of human behaviour? Our
everyday psychological scheme would misrepresent our internal states and
activities, just as conceptions of the world that mentioned witches, ether,
or phlogiston, in earlier thinking, misrepresented the nature of reality. Paul
105
The Continuum Companion to Philosophy of Mind
Churchland has suggested that there may be some reason to accept the legiti-
macy of this challenge to common sense psychology because of the failure of its
theoretical concepts to line up with the categories of neuroscience. The former
may have to be eliminated by, rather than reduced to the categories of a fully
mature neuroscience of cognition and action (see Churchland, 1978).
This is the eliminative materialists option. And even if we reject it, the threat
it poses is real enough. Consider the following passage from Brian Loars Mind
and Meaning (1981):
If it were to turn out that the physical mechanisms that completely explain
human behaviour at no level exhibited the structure of beliefs and desires,
then something we had all along believed, viz. that beliefs and desires
were among the causes of behaviour, would turn out to be false. Naturally,
we would continue to use the belief-desire framework to systematize
behaviour, but that should then at the theoretical level have the air of
fictionalising and contrivance. (Loar, 1981, pp. 145)
106
Folk Psychology and Scientic Psychology
Scientic Psychology
107
The Continuum Companion to Philosophy of Mind
key requirements (1) and (3) on a satisfactory folk psychology. The various
accounts it gives of the cognitive states and processes that subserve specific
abilities for vision, language, audition, and motor-control are states not known
to agents first-personally. Nor are they ordinarily known third-personally.
Instead, we should think of them as states of sub-personal mechanisms not
states we a ribute to persons posited by a theory of the internal cognitive
mechanisms that subserve particular capacities. Such content-bearing cognitive
states do not provide rational explanations of the behaviour in which we display
our abilities and capacities We see distance and depth in the visual field, but
the psychological states and processes responsible for this aspect of the visual
scene do not give the sighted person reasons to see things this way: they causally
explain why creatures with stereoscopic vision do see things this way in virtue
of the content of those states. The common element that such underlying states
of our cognitive systems share with the psychological states posited by folk
psychology is that both are content-bearing or representational states; but the
cognitive states invoked by scientific psychology do not have propositional
contents that sustain logical and hence rational relations to one another.
Scientific psychology, in particular cognitive psychology, extends the domain
of content-bearing psychological states without extending the domain of
rationally governed states, thus marking a division within the mind.
Given this division within the mind between dierent kinds of mental states
and the dierent kinds of psychological explanations in which they feature,
how can a psychology that rationalizes behaviour in terms of beliefs, desires and
intentions accommodate an underlying sub-personal psychology that explains
capacities exhibited in the same behaviour? Does the science underpin or
undermine the folk psychological we give of ourselves and others?
We shall look first at a purely a priori defence of the legitimacy of folk
psychology. This view rejects the idea that common sense psychology could
be answerable to scientific psychology by denying that there could be a science
of the mind.
This is the interpretationist view of the mental set out by Donald Davidson.
Is the commonsense (or folk) psychology we use to make sense of one anothers
verbal and non-verbal behaviour vulnerable to scientific challenge? Could
findings in scientific psychology or neuroscience show that a scheme we use
for explaining human action is fundamentally flawed and mistaken?
An a priori defence of commonsense psychology, if successful, would render
it immune to scientific challenge. This is the view of the mind proposed by
Davidson, who sees mental life as constituted and exhausted by the application
108
Folk Psychology and Scientic Psychology
Levels of description
109
The Continuum Companion to Philosophy of Mind
gives the agents reasons for behaving as she did. The reasons cited are given
in terms of what people believe, desire and intend.
Notice that the target of rational explanation is not behaviour itself, where
behaviour is construed as bodily movement, but action: intentionally described
episodes of behaviour or bodily movements that amount to an agents perform-
ing an action for a reason. It is these intentional actions as observable aspects
of mind for which we give rational explanations. Equally, actions as part of
the mind of a creature only come into view when we interpret that creature
as a rational agent acting on beliefs and desires. Thus rational interpretation
constitutes its own explananda by coming to see certain bodily movements
as actions. When a bodily movement is interpreted as an intentional action
part of someones mind it is the very same event that can be described in
physical and mental terms: all we have are two descriptions of the same thing.
The mental is an intentional level of description of otherwise physical events.
As Davidson puts it: events are mental only as described (Davidson, 1980,
p. 215). Events are particular, datable unrepeatable occurrences, and all events
are physical (i.e. physically describable). Some of these events are also mental,
that is, correctly describable in mental terms. When we re-describe a episode
in a persons physical history a bodily movement in intentional terms,
describing it as the action of an agent intentional under a certain description,
we see it as part of her mental life. And in treating certain behaviours as part of
someones mental life as actions undertaken we thereby introduce beliefs
and desires as the agents reasons for performing those actions, and at that
stage we retrospectively identify those mental states with the physical or
neurological states that are the causes of the bodily movement in question.
Keeping the mental and physical levels of description separate, Davidson
argues for a non-reductive physicalism he terms anomalous monism:
The mental events that cause bodily movements are identical with physical
events: an event described in mental terms can also be described in physical
terms, and it is under its physical description that it can instantiate a strict
causal law couched in purely physical vocabulary. The physically described
event, which is also described in mental terms, enters into causal relation
with another event described in physical terms. Events are related as cause
and eect when they have descriptions that instantiate causal physical laws.
So even though there are no strict laws for predicting or explaining psycho-
logical phenomena, mental events can still cause physical events. The mental
110
Folk Psychology and Scientic Psychology
event that is the cause of some piece of behaviour has a description under
which it instantiates a causal law that links physical events of that kind to
physical events of some other kind. The singular causal statement linking
As reasons on a particular occasion to what he did will be true only if it is
backed by a general causal law that relates events of that kind when physically
described.
It may seem that it is only at the physical level that causality occurs, and so the
mental may seem causally irrelevant or epiphenomenal. Physically described
events would continue to have the eects they do whether or not they were
describable in mental terms. But care is needed here if we are not to misunder-
stand Davidsons position. Davidson believes in the supervenience of the
mental on the physical: there can be no mental dierence without a physical
dierence, and any two events alike in all physical respects will be alike in all
mental respects. So when one event, physically described, causes another,
physically described, and the first is also describable in intentional terms as a
mental event and the cause of the resulting behaviour, it could not have been
that very event if it did not have that intentional description; for if it had been
dierent in some mental respect (i.e. not having any mental description), it
would have had to be dierent in some physical respect on pain of violating
supervenience, and then it would have to be a dierent event all together.
Whenever physically described events are also describable in mental terms,
those events count as mental events and cannot be otherwise unless things
had been physically dierent. So the challenge to causal ecacy that suggests
things could have taken place physically whether the mental was present
or not fails, and Davidson can still claim that the mental events can causes
physical events.1
Beliefs and desires issue in behaviour only as modified by other beliefs and
desires, a itudes and a endings, without limit. Clearly this holism of the
111
The Continuum Companion to Philosophy of Mind
The principles that ensure the coherence of a persons a itudes and actions,
from the point of view of an interpreter, are the principles of rationality and
charity. Beliefs should make rational sense in the light of other beliefs to
which they are logically related by the contents of those beliefs; and the actions
we see people as performing should make sense in the light of the beliefs and
desires we are prepared to ascribe to them and which give them reason to do
and say what they do and say. In addition to rationality, the other constraint on
correct interpretation is the principle of charity, which ensures that a persons
beliefs about their surroundings should be, by and large, correct by the inter-
preters lights. To be interpretable, most of an agents ordinary beliefs about the
world around her should be true. And the correct interpretation of a persons
behaviour will be the one that ensures the best fit with these two principles.
The principles of rationality and charity must guide us in the course of
building up a portrait of someones overall propositional a itudes and actions.
They do so when we ascribe particular beliefs, desires and meanings that make
112
Folk Psychology and Scientic Psychology
sense of someones actions and u erances, in the light of further beliefs, desires
and other meanings that comport with the persons overall behaviour and their
surroundings.
Rationality requires that if there are two interpretations, each consistent
with everything the person does and says, we should favour the one that makes
their network of beliefs, intentions and actions, more and not less rational. The
principles of practical reason suggest that people will by and large do what
they believe will secure their fondest wish at that moment, assuming no coun-
tervailing beliefs and no countermanding desires. For you to do otherwise
would make no sense to us as interpreters. Charity requires interpreters to
make sense of people by a ributing to them beliefs it would make sense for
them to have given their current surroundings. Generalizing, we should inter-
pret people charitably by ascribing to them beliefs that are largely true (by our
lights). That is, we should not gratuitously ascribe to someone a bizarre belief
about what is going on around them, but assume, charitably, that they are like
us in having reliable beliefs about their current surroundings. Notice that the
principles of rationality and charity interact: to ascribe to someone an outland-
ish belief may result, through the holistic connection between a itudes, in a
scheme of interpretation for that person that makes him less rather than more
rational. On the other hand, if ascribing to someone veridical beliefs about his
current surroundings would make his behaviour less rationally explicable than
ascribing a false belief to him, we must forgo charitable interpretation on this
occasion. By and large, Davidson thinks the conditions for interpretation show
that there is a large degree of truth and consistency in the thought and speech
of an agent.2 The coherence of belief is guaranteed by the joint application of
charity and rationality. If there is not enough coherence in someones beliefs we
would not be justified in regarding them as rational agents at all. Even irratio-
nality assumes that a prior rational standard is operative and that a person
is going against it on this occasion. To assume that no rational standard is
operating is to see a creature as non-rational.
As interpreters we must always a empt to see peoples a itudes and actions
as making rational sense by our lights. This is not a subjective view of rationality.
What enables me to see pa erns of rationality in the a itudes and actions of
others is the rationality at work in my own thought and talk. The standards of
rationality that enable me to make sense of others are the standards that enable
me to make sense simpliciter; this is what enables other to make intelligible
sense of me. Thus, there is just one standard of rationality for all or at any rate
for those who are mutually intelligible.
Rationality and charity are relative a priori principles: they are not principles
designed to get at the (independently constituted) facts of someones mental
life; rather it is only when the intentionally described states of a creature con-
form to them that creatures count as having mental states. What it is correct to
113
The Continuum Companion to Philosophy of Mind
When interpreting other people, much of our information about what they
think comes from what people say. But to find out what they are saying we have
to know what their words mean. Even in everyday speech we can still wonder
whether other people are using words in the same way we are. So to confirm
that others mean the same thing as us by their use of words, or to adjust for the
dierence, we have to resort to interpretation. Interpreting speech is part of
interpreting a persons overall behaviour. Thus, in addition to ascribing beliefs,
desires and intentions to an individual we also have to assign meanings to her
words, and ensure the best holistic fit with her linguistic and non-linguistic
behaviour. In this way the interpretative principles are also at work in justifying
our ascription of meanings to peoples words. But if we have to know what
someone means to know what they believe and cannot know what they mean
without first interpreting them ascribing them beliefs and desires we go
round in a circle. We must either solve for two unknowns simultaneously or
find a way to break into the circle. Davidson proposes to do the la er by focus-
ing on the case of a speaker A holding a sentence S true under certain circum-
stances. We can know the speaker holds S true without knowing what S means,
and we can see the holding true as depending on what the speaker means by S
and what she believes to be the case. Now if we charitably interpret a speaker
114
Folk Psychology and Scientic Psychology
115
The Continuum Companion to Philosophy of Mind
correctly called states of mind they are not beliefs, desires, wishes or
intentions. (Davidson, 1986b, p. 160)
Other minds, on Davidsons view, are what we get when we interpret the
behaviour of others. Bodies are what we have before we interpret their
behaviour. (Root, 1986, p. 294)
116
Folk Psychology and Scientic Psychology
117
The Continuum Companion to Philosophy of Mind
For Denne , beliefs and desires go together with the idea of an intentional
system a system whose behaviour can be predicated by ascribing it beliefs
and desires and assuming it will act rationally. To treat a creature as if it were
a rational agent is to adopt an intentional strategy, to take up the intentional
stance towards it. According to Denne , this means we ascribe it the beliefs
it ought to have given its place in the world and its progress through it.
Likewise, we ascribe it the desires it ought to have given its place in the world
and its purposes. We then predict what it will do given the a ributions
we make to it and the assumption that it is rational. (When we dont succeed
we may have to modify particular a ributions, or give up the assumption of
rationality.)
Taking up the intentional stance is a strategy to predict how the creature
(or system) will behave. But note that to predict is not necessarily to explain.
Denne s analogy is the chess-playing computer. We predict its moves by
assuming it has certain beliefs and goals it wants to get its queen out early;
it thinks Im weak on the le flank and treating it as a rational opponent. But
in some sense, it doesnt really have beliefs and desires. This is just a heuristic,
or useful assumption that helps us to predict the computers moves so we can
compete against it. If we want to know what actually accounts for and explains
its behaviour, we have to drop down to what Denne calls the design level by
adopting the design stance.
At this level predictions are in line with the functioning components of the parts
of the system and what they are designed to do. For example, the chess-playing
computers workings are predictable by programmer on the assumption that
everything is functioning properly and working as it should do at the physical
level. The program may not be all we need to know in order to know the beha-
viour of the machine. It may not be working as it was designed to do because
physical components may have failed. This requires us to turn to the next
level down by adopting the physical stance towards the system.
118
Folk Psychology and Scientic Psychology
To know about the workings of the system at the physical level is to adopt
the physical stance towards its workings. At this level we expect to achieve
physical law-like predications. For Denne , the physical and the design stance
(descriptions of the functional organisation and mechanical working of the
system) oer not only predictions but also genuine explanations of the beha-
viour of the system. However, when we adopt the intentional stance, we are
only involved in predication, not explanation.
Just as its not literally true to say the chess playing computer has beliefs and
desires that cause it to act, so its not literally true of us either. We simply ascribe
beliefs and desires to one another and assume people will act rationally. In
this way, we succeed in making fairly reliable predictions about one anothers
behaviour. But at the design level, the level of our inner organisation studied
by cognitive psychologists, there may be nothing like beliefs and desires
causing us to act. Belief-desire talk does not describe our innards. Denne is
a realist about our internal functional organization but a non-factualist, or
instrumentalist about beliefs and desires. Beliefs and desires are a ributed to
us by others to make sense of our behaviour, but they dont feature in our
inner workings, and they dont causally explain our behaviour. To do that
we need a scientific psychology that addresses the functional architecture of
the mind, and at that sub-personal level (another of Denne s distinctions about
systems smaller than persons which lack the properties assigned to persons),
we will have no use for belief-desire talk.
The trouble with this view, as Brian Loar put it in the quote above, is that
our continuing to use the belief-desire framework to systematize behaviour
. . . [would then] have the air of fictionalizing and contrivance. But Denne
wants to reject this portrayal of his view. He wants to maintain that he does, in
some sense, still believe in the reality of beliefs and desires. But in what sense?
Like Davidson, Denne is an a ributionist about the mind; they both hold
an essentially third-personal view of belief-desire psychology. But their views
dier. Davidson is a realist about beliefs and desires. He thinks that belief-
desire psychology gives us genuine explanations but not predictions of human
agency. Denne , on the other hand thinks that belief-desire psychology gives us
119
The Continuum Companion to Philosophy of Mind
predication but not explanation. The dierence between these views hinges on
the rival conceptions of rationality. According to Davidson, rationality is uncod-
ifiable: there are no psychological or psycho-physical laws. Rational explana-
tion is always post hoc and partial and what counts as rational is just what
makes sense to us as rational creatures. By contrast, Denne thinks rationality
is law-like although rationality is at best an approximation of our behaviour.
The real explanations are found at the levels below.
For Denne the laws of folk psychology are at best rough and ready approxi-
mations (or idealizations) that help us to predict and interact with others. Hence:
Folk psychology is best seen not as a sketch of internal processes, but as an idea-
lized, abstract, instrumentalistic calculus of prediction. (Denne , 1987a, p. 48)
120
Folk Psychology and Scientic Psychology
details from the intentional level of folk psychology onto the levels of organisa-
tion below. But we are still allowed to ask the question:
Exactly what feature must we share for [a given belief ascription] to be true
of us? More generally . . . what must be in common between things truly
[italics mine] ascribed an intentional predicate such as wants to visit China
or expects noodles for supper? (Denne , 1987a, p. 43)
Denne s answer is
a shared property that is visible, as it were, from one very limited point of
view: the point of view of folk psychology. Ordinary folk psychologists have
no diculty imputing such useful but elusive commonalities to people. If
they then insist that in doing so they are postulating a similarly structured
object in the head, this is a gratuitous bit of misplaced concreteness, a
regre able lapse in ideology. (Denne , 1987a, p. 55)
121
The Continuum Companion to Philosophy of Mind
Taking up the challenge where Denne s view gives out, Jerry Fodor asks
what all thinkers who share a given thought have in common if its not their
neurophysiology:
How much (and what kinds of similarity between thinkers does the
intentional identity of their thoughts require? This is, notice, a question one
had be er be able to answer if there is going to be a scientifically interesting
propositional a itude psychology. (Fodor, 1986a, p. 9)
Fodor advocates the view that scientific psychology will vindicate common-
sense psychology. The problem is to understand how a science of the mental
can explain what we do without replacing or reducing our common sense
concepts.
Fodor aims to steer between these two dangers reduction (not likely) or
replacement (not palatable) by insisting that the causal laws of belief-desire
psychology that capture intentional generalizations about individuals have
to be explained by the computational and syntactic laws that implement them.
The dierence between the intentional and computational laws is not a ma er
of levels of description but levels of organisation within a creature. The compu-
tational laws govern the mechanisms that mediate the connections between
intentional causes and their behavioural eects. They are many and varied, so
no strict reduction of the intentional laws is possible. However, the lack of
reduction doesnt loosen all connection between the levels, for although there
are intentional causal laws that explain why we think and act as we do, we
are always entitled to ask, for any such law, how do those causes bring about
those eects in that individual.
122
Folk Psychology and Scientic Psychology
(1)The laws of folk psychology are intentional laws that cite contents.
(2)The content (semantics) of intentional states is informational and external.
123
The Continuum Companion to Philosophy of Mind
124
Folk Psychology and Scientic Psychology
125
The Continuum Companion to Philosophy of Mind
(e.g. Elm/Beech, or H20 and XYZ). Fodors answer is that both these cases mostly
dont occur, and when they do we can explain what has gone wrong. In Frege
cases, the dierent syntactic forms of the representations explains how they
aect computational mechanisms dierently: semantics and syntax come apart.
Semantics and syntax march in step in twin cases, but because of oddities of
the environment, distinct representations with dierent contents trigger the
same computational mechanisms, which is extremely rare.
There will be minor revisions in our everyday intentional idiom but not
massive revision, or else the laws of intentional psychology would not work
as well as they do, nor would the computational mechanisms operating in us
really be implementations of those laws and thus sustain intentional regulari-
ties. In other words, Fodor is holding out for the claim a hostage to empirical
fortune that there is a class of computational syntactic processes that makes up a
natural domain for psychological explanation. This will be the domain for creatures
like us who satisfy the same intentional generalizations we do, and the relation
between the computational underpinnings and the level of intentional states,
though contingent, will be reliable and explicable. Certainly, all believers and
desirers have this much in common: they must all have mental representations
which are syntactic vehicles for the contents of those states. This is the indepen-
dently motivated claim for the language of thought which does generalize across
thinkers however dierent their beliefs and desires. So regardless of their
specific mental states, and the particular psychological explanations we give
of their behaviour, they will all satisfy the same general explanatory pa erns
of acting in such a way as to do what they believe will secure for them their
fondest wish. And the claim that the internal mediating mechanisms of thinkers
can dier from individual to individual will still have to leave room for the
claim that they have enough overlap to ensure that the syntax of their mental
representations, whatever they are, mediate the same broad behavioural similarities
among thinkers:
The syntax of the mental representations which have the facts that P in their
causal histories [and so are about P] tend to overlap in ways that support
robust behavioural similarities among P-believers. (Fodor, 1994, p. 53, italics
and brackets mine.)
The trick is to find in each of us mentalese sentences, which are the causes
of the sorts of behavioural proclivities that the laws of psychology say that
P-believers share (Fodor, 1994, p. 54). For if no such overlapping set of causes
exists it is hard to see what relation of constraint between the levels Fodors
view imposes. And if it turned out to be none at all, we should have le Fodor
no be er o for a story about the relationship between folk psychology and
scientific psychology than Denne .
126
Folk Psychology and Scientic Psychology
The issue in (a) concerns not what makes the ascription of a belief to an indi-
vidual true, but rather how we go about ascribing beliefs and other a itudes in
the first place. What equips us to make such ascriptions? What provides us with
the means to a ribute mental states to other creatures? We shall briefly look
at two empirically dierent accounts of how we a ribute mental states to one
another, namely the theory-theory and the simulation accounts of our theory of mind.
So far as (b) is concerned, consciousness has had li le if anything to do with
the issues we have been discussing so far. It is a central aspect of the human
mind and a full and final theory of human psychology will have to account
for it, but it plays a much less prominent role in our folk psychology. We know
very li le about the conscious character or experiences of other people and yet
we are still adept at figuring out their reasons for action.
Finally, under (c), we surely need to accommodate the emotions as part of
folk psychology. A er all, in our everyday dealings with one another, we use a
wide repertoire of emotional terms such as anger, fear, humiliation, jealousy,
joy, happiness, envy, longing, and loneliness to help explain and predict one
anothers behaviour. And yet, when we turn to philosophical descriptions of com-
mon sense psychology, we have, until very recently, seen li le or no acknow-
ledgement of the role emotions play in our mental lives. Can a psychology that
rationalizes behaviour in terms of beliefs, desires and intentions leave room for
emotions to play more than a merely disruptive role as sand in the mecha-
nism? Let us end by looking at each of these issues in turn.
127
The Continuum Companion to Philosophy of Mind
Some have objected that we do not need to theorize about others to know what
they are up to and to know their mental states. The mirror neuron system in
humans and monkeys activates the same populations of neurons in the pre-motor
128
Folk Psychology and Scientic Psychology
Consciousness
When we talk about folk psychological concepts like belief and desire
concepts with both first and third person applications all mention of con-
sciousness is missing. Davidson makes no use of it, nor of perceptual experience.
Experience does not give grounds for believing what we do. Instead, he argues
that the only thing that can justify a belief is another belief. And the only things
that can bring about beliefs in non-rational ways are the causal impacts the envi-
ronment has on us: sensation plays a crucial role in the causal process that
connects beliefs and the world (Davidson, The Myth of the Subjective in his
2001, p. 46). When we think about what we see we form beliefs about what
we are looking at. So Davidson pays no a ention to consciousness. No doubt
we have it but it is not playing an important epistemological or metaphysical
role in characterizing mental states. We know a lot about other peoples minds
while knowing li le or nothing about their conscious experiences.
129
The Continuum Companion to Philosophy of Mind
Emotions
Option (a) has it that emotions play no real role in explaining others. The
virtue of this strategy is that it would cause minimal mutilation to existing
philosophical accounts of everyday or folk psychology. But it neglects the role
130
Folk Psychology and Scientic Psychology
131
The Continuum Companion to Philosophy of Mind
132
7 Internalism and Externalism
in Mind
Sarah Sawyer
133
The Continuum Companion to Philosophy of Mind
and been thought to tell against externalism. In the fi h section I deal with epis-
temological considerations concerning the direct, non-empirical, authoritative
nature of self-knowledge that have been thought to tell against externalism.
I then conclude briefly in the sixth section.
Kinds of Externalism
Predicative externalism
134
Internalism and Externalism in Mind
S would have been related to twilver in just the same way as she is actually
related to silver, and hence it is plausibly the concept twilver that S would have
acquired. Consequently, where S thinks that silver jewellery is cheaper than
gold jewellery, counterfactual S thinks instead that twilver jewellery is cheaper
than gold jewellery. The dierence in representational content between the
belief S has and the belief S would have lies in the dierence between the objec-
tive properties to which she is related (silver) and would be related (twilver)
respectively. If these considerations are persuasive, then what determines the
representational content of a subjects beliefs goes beyond her intrinsic physical
make-up and her discriminative capacities (which are hypothesized to be
identical in the actual and the counterfactual scenarios alike) and depends in
addition on the objective properties to which she is related.4
This kind of thought experiment is taken by many to establish externalism
specifically with respect to natural kind concepts: concepts that carve nature at
its joints and feature in the true final set of scientific theories: concepts such as
(perhaps) quark, electron, hydrogen, water, heart, tiger, planet.5 However, reflection
on two further kinds of counterfactual scenario favours a more general exter-
nalism. The first draws upon the possibility of incomplete linguistic under-
standing6; the second draws upon the possibility of non-standard theory7.
I outline each in turn.
First suppose that a subject S has a wide range of ordinary beliefs a ribut-
able by means of the term game: she believes that some games are more fun
than others, that chess is a game, that children like party games, and so on.
However, she believes in addition (and mistakenly) that games must involve
at least two people, a point she would readily accept correction on if her
mistake were pointed out to her. Next consider a counterfactual scenario
in which her intrinsic physical make-up is hypothesized to remain constant
while her linguistic community is hypothesized to dier. In the counterfactual
scenario the term game is defined and standardly used to apply to games
that involve at least two people. Since game and game involving at least
two people mean dierent things, the word-form game in the counterfactual
scenario has a dierent meaning than it does in the actual situation. In like
fashion, the concept expressed by the word-form diers in the actual and the
counterfactual situations. In the actual situation the word-form game expresses
the concept game and includes in its extension games such as solitaire and
patience. In the counterfactual situation, in contrast, the word-form game
expresses a dierent concept that does not include in its extension either soli-
taire or patience. Consequently, S may in fact believe that pass the parcel is a
game, but had she been a member of the counterfactual linguistic community
she would have possessed a distinct concept, believing instead that pass the
parcel is a shgame, say. Once again, Ss intrinsic physical make-up and classifi-
catory capacities are identical in the actual and the counterfactual situations,
135
The Continuum Companion to Philosophy of Mind
and the dierence in representational content between the belief she has and
the belief she would have lies beyond her intrinsic physical properties, this time
anchored by the classificatory practices of the wider linguistic community of
which she is and takes herself to be a part.
Behind this counterfactual scenario lies a certain understanding of linguistic
meaning according to which the conventional linguistic meaning of a term
(roughly its dictionary definition) is a complex abstraction from communal
rather than individual use. Linguistic meaning is determined by actual and
possible agreement among the most competent users, where the most compe-
tent users are those to whom others do and would defer if a question about an
individuals use were to arise. On this view, understanding the meaning of a
word is not an all-or-nothing thing, but rather comes in degrees. And it is the
possibility of understanding a word incompletely that allows for the dierence
in linguistic meaning in the actual and the counterfactual situations to be con-
sistent with there being no dierence in intrinsic physical make-up between
actual and counterfactual S. The dierence in linguistic meaning is then taken
to imply a dierence in concept expressed.
The final consideration that favours a general externalism trades on the fact
that even a subject with a full understanding of the linguistic meaning of a term
can doubt whether the dictionary definition that reflects that meaning correctly
characterizes the things referred to by that term. Thus suppose a subject S has a
full understanding of the term sofa and yet comes to wonder whether sofas
are really religious artefacts and not pieces of furniture made for si ing. Her
proposed theory about sofas is false, but this need not compromise either her
full understanding of the term sofa or her ability to think with the concept
sofa; rather, it reflects a strange view about the nature of sofas thought of as
such. Now hypothesize a counterfactual situation in which Ss false theory
is standard and true of a dierent yet superficially indistinguishable class of
entities (call them safos).8 The linguistic meaning of the term sofa in the
actual situation diers from the linguistic meaning of the term sofa in the
counterfactual situation even though the entities referred to are superficially
indistinguishable. This is because the actual linguistic community and the
counterfactual linguistic community have agreed upon dierent characteriza-
tions of the relevant entities. Moreover, the concept expressed by the term
diers in the two situations because the entities referred to dier: in the actual
situation they are sofas (pieces of furniture made for si ing), whereas in the
counterfactual situation they are safos (religious artefacts). Consequently, while
actual S believes that sofas are religious artefacts, counterfactual S believes that
safos are religious artefacts.
Behind this counterfactual scenario is a certain understanding of the dier-
ence between the linguistic meaning of a term and the concept expressed by
that term. The linguistic meaning of a term goes beyond individual use and is
136
Internalism and Externalism in Mind
Singular externalism
137
The Continuum Companion to Philosophy of Mind
138
Internalism and Externalism in Mind
Kinds of Internalism
139
The Continuum Companion to Philosophy of Mind
140
Internalism and Externalism in Mind
as illustrated by the use of new terms in the examples just given. And on both
views, S possesses the same concepts as her counterfactual self in virtue of
having the same discriminative capacities and epistemic outlook on the world,
but she has dierent concepts from those in her linguistic community. This
stands in marked contrast to predicative externalism, according to which S has
dierent concepts from her counterfactual self but shares many concepts with
those in her linguistic community despite varying degrees of understanding
and competence which result in a wide variety of discriminative capacities and
epistemic outlooks across individuals within that community.
The third and fourth kinds of predicative internalism are more complicated.
They acknowledge that the counterfactual scenarios outlined establish that
S and counterfactual S have dierent thoughts in some sense, but aim none-
theless to retain a sense of content which is preserved across intrinsic physical
duplicates, in order to respect the internalist conception of sameness of epistemic
outlook. Both therefore maintain that a thought has a narrow and a broad
content and are thus kinds of two-factor theory.
According to the first of these, the internal component of a thought (its
narrow content) is a function that determines its external component (its broad,
truth-conditional content) given a context (an environment).16 Thus when S
and counterfactual S u er the sentence Silver jewellery is cheaper than gold
jewellery, the narrow content of their thoughts is the same, but the broad
content of their thoughts diers simply in virtue of their location in dierent
environments: Ss thought concerns silver, and is true if and only if silver
jewellery is cheaper than gold jewellery; whereas counterfactual Ss thought
concerns twilver, and is true if and only if twilver jewellery is cheaper than
gold jewellery.
There are similarities between this two-factor theory of predicative thought
and the two-factor theory of singular thought discussed in the Singular Exter-
nalism section above. On both views the only form of truth-conditional content
is broad. And yet on both views the thoughts of intrinsic physical duplicates
share a kind of content even though they have dierent truth conditions.
However, the similarity does not extend beyond the superficial level, and the
dierences are important. According to the two-factor theory of singular
thought, singular thoughts have contents that are intrinsically representational
independent of context, and can be applied to (or thought of) dierent indi-
viduals in dierent circumstances. The two factors involved in a singular
thought are first, a content (is nutritious, say), and second, (potentially) an
individual of whom the content is thought (a particular apple, for instance).
141
The Continuum Companion to Philosophy of Mind
The content of a singular thought is not itself divided into a narrow and a broad
component. According to the two-factor theory of predicative thought now
under consideration, in contrast, the content of a predicative thought is itself
divided into a narrow component and a broad component. Crucially, the
narrow component is not representational: only the broad component is. The
narrow component is a function and can be understood only in terms of its
inputs and outputs: that is, only in terms of the broad, truth-conditional content
it produces once the individual is situated in a particular environment. This
puts pressure on the idea that the narrow component of a predicative thought
is properly conceived as a form of content at all.
The second kind of two-factor theory of predicative thought is also a racted
both by the externalist interpretation of the counterfactual scenarios and by the
internalist conception of sameness of epistemic outlook. However, it aims to
draw a distinction between broad and narrow content consistent with all con-
tent being representational in some sense. On this view, the broad content of a
subjects thought is determined, in line with externalist considerations, in part
by relations she bears to objective properties in her environment. The narrow
content of a thought, on the other hand, is individuated by the epistemic pos-
sibilities it allows and excludes.17 The underlying thought here is that intrinsic
physical duplicates are in the same epistemic position in the sense that they
cannot distinguish between the relevant actual and counterfactual situations
and that the narrow content of a thought encapsulates this fact. For example,
suppose that S and counterfactual S both u er the sentence Silver jewellery
is cheaper than gold jewellery. The thoughts they thereby express have dier-
ent broad, truth-conditional contents: one concerns silver whereas the other
concerns twilver. However, the thoughts are taken to have the same narrow
content because a purely qualitative description of a situation in which silver
jewellery is cheaper than gold jewellery is identical to a purely qualitative
description of a situation in which twilver jewellery is cheaper than gold
jewellery. Given the epistemic position of S and counterfactual S, both situa-
tions verify their thoughts and hence they share a narrow content. The success
of the position clearly depends on the possibility of describing situations in
purely qualitative terms terms not subject to externalist considerations. As
such, the position depends upon the truth of a restricted rather than a general
form of externalism. If all terms were subject to externalist considerations then
there would be no terms available to feature in the qualitative descriptions
required to ground this notion of narrow content. Moreover, there is a question
about whether it makes sense to think of a thought as having two forms of
representational content where only one of these is truth conditional.
I have discussed and argued against all four forms of internalism elsewhere
and will not repeat the arguments here.18 Instead I now turn to some of the
142
Internalism and Externalism in Mind
Metaphysical Considerations
143
The Continuum Companion to Philosophy of Mind
144
Internalism and Externalism in Mind
Epistemological Considerations
145
The Continuum Companion to Philosophy of Mind
146
Internalism and Externalism in Mind
false, since it does not capture the content of her thought at t1. Consequently, S
does not know at t2 what she was thinking at t1. This is taken to undermine
self-knowledge of externally individuated past thoughts. Moreover, it is argued,
self-knowledge of externally individuated current thoughts is also undermined.
A er all, if S does not know at t2 what she was thinking at t1, and there is no
reason to think she has forgo en anything in the interim, there is reason to
think she never knew at t1 what she was then thinking.
As with the initial argument, there are various ways the argument might be
taken depending on whether one thinks mere possibility, close possibility or
actuality the relevant epistemic factor. But here two dierent responses have
emerged. According to the first, the argument shows that externalism does
undermine the direct, non-empirical, authoritative nature of ones knowledge
of ones past thoughts, but it does not show that externalism undermines the
direct, non-empirical, authoritative nature of ones knowledge of ones current
thoughts.31 This can be made plausible, for instance, by acknowledging a new
way in which one might be said to forget something (namely, by being switched
between subjectively indistinguishable environments), or alternatively, by
maintaining that forge ing is not the only way in which one might fail to
know at t2 what one knew at t1 (since one might instead be switched between
subjectively indistinguishable environments). The general moral here is that
although one may need to rely on empirical considerations to the eect that the
environment has remained broadly stable in order to know what one thought
in the past, the non-empirical warrant for knowledge of ones current thoughts
is not thereby undermined. The disruption, as it were, is confined to knowledge
of past thoughts.
According to the second line of response, the argument does not show that
externalism undermines the direct, non-empirical, authoritative nature of ones
knowledge either of ones current or of ones past thoughts.32 This second line
of response is bolder and can be made plausible, for instance, by showing
how the content of past thoughts can be preserved in memory even across
undetectable switches between diering environments. This has been advo-
cated by Burge, who distinguishes substantive event memory, which refers
back to earlier events, from preservative memory, the function of which is to
hold the contents of thought in place so the subject can determine logical and
epistemic relations between them. Preservative memory does not refer back to
earlier thinkings, but rather holds contents in place for the purposes of, for
instance, critical reasoning.33 While substantive event memory might be under-
mined by externalism, preservative memory will not be, precisely because
preservative memories do not refer back to independent events.34
It is important to note that externalists have not oered a theory of self-
knowledge in response to the achievement problem in either of its guises.
Rather, they have tried to show how the arguments are misguided. Two things
147
The Continuum Companion to Philosophy of Mind
148
Internalism and Externalism in Mind
Conclusion
149
The Continuum Companion to Philosophy of Mind
150
8 The Philosophies of
Cognitive Science1
Margaret A. Boden
Theres no such thing as the philosophy of cognitive science. Rather, there are
competing philosophies of and within the field. Thats partly because the
concepts and techniques of artificial intelligence and artificial life have been
changed and enriched since the 1940s. For although psychology is the thematic
heart of cognitive science, its intellectual heart is AI/A-Life or AI, for short.
Psychology (both animal and human) is the thematic heart because cognitive
science studies all aspects of the mind or mind/brain, or, if you prefer, embodied
experience and behaviour. It ranges from low-level vision to enculturated
thought, from infantile development to adult personality, and from individual
behaviour to social phenomena. It investigates not only cognition, but emotion
and motivation too. So the field is badly named: outsiders are o en misled,
assuming that it deals only with cognition.
Cognitive science diers from other forms of psychology in using computa-
tional concepts of various kinds. Very broadly speaking, these fall into two
main types: formalist/symbolic and connectionist/dynamical. Much of the
philosophical interest lies in the dierences between these approaches.
Some computational concepts, in the broad sense intended here, denote
formal computations on symbolic representations. These typify classical AI, or
GOFAI, Good Old-Fashioned AI (Haugeland, 1985, p. 112). Others draw on
cybernetic ideas about embodied and self-organizing systems. These include
situated robotics, wherein the robots rely on direct reflex responses to environ-
mental cues; dynamical systems understood in terms of physical laws; and
self-equilibrating neural networks. And all approaches sometimes include
the sort of computation (mutation and natural selection) thats eected by
evolution.
Cognitive scientists o en express their theories as computer models, because
this is the best way of testing their coherence and implications. (Testing for their
truth, of course, involves comparisons with the actual phenomena.) The com-
putational concepts implemented in such models are substantive theoretical
151
The Continuum Companion to Philosophy of Mind
terms. That is, the mind and the brain are theorized as mechanisms that actu-
ally carry out computations of one kind and/or another. Put another way, the
mind is conceptualized as what computer scientists call a virtual machine,
defined in abstract (computational) terms but implemented in the brain.
The varieties of computation mentioned above are best suited for dierent
tasks. All of them (and more) will be needed for an adequate account of the
complex virtual machine which is the human mind (Sloman, 2000; Minsky,
1985, 2006). Some interesting hybrid systems have already been implemented,
in which psychological phenomena are modelled by a combination of GOFAI
and connectionist methods (Norman and Shallice, 1986; Cooper et al., 1995).
However, the various techniques are o en wrongly assumed to be mutually
exclusive (e.g. Dreyfus and Dreyfus, 1988). Thats part of the reason for the
multiplication of philosophies of cognitive science.
Another reason for this pluralism is the fundamental divide between
analytic and continental philosophy. Almost all the early philosophers of
cognitive science came from the analytic community, and were commi ed to
functionalism. Philosophical traditions such as phenomenology were scarcely
mentioned, although Hubert Dreyfus (1967, 1972) was an important exception.
Recently, there have been a empts to combine these two traditions, or even
to forsake one for the other; see the Embodiment, Enactiveness, and Phenome-
nology section below.
152
The Philosophies of Cognitive Science
153
The Continuum Companion to Philosophy of Mind
154
The Philosophies of Cognitive Science
GOFAI did. Broadly inspired by the brain, it was less biologically implausible
too. Even so, there are many dierences between PDP systems and real neural
networks; multiple realizability had been diluted only a li le.
In GOFAI, in the psychological research that inspired it (Bruner et al., 1956;
Hunt 1962), and in philosophical writings based on it (Fodor, 1975), concepts
were defined in terms of necessary and sucient conditions. This contributed
to the notorious bri leness of GOFAI systems: in the absence of explicit excep-
tions, just one missing criterion would render a concept inapplicable. Later
psychological work suggested that concepts were less neat and tidy (Rosch
and Mervis, 1975). And some philosophers, of course, had already said so
(Wi genstein, 1953).
In PDP systems, concepts are implemented not by all-or-none activations of
neatly listed defining criteria, but by equilibrium states of the whole network,
involving mini-representations of many dierent, and even partially conflict-
ing, facets. Since each facet is continuously weighted, and the weights can
change according to the contextual evidence, concepts arent accepted/rejected
outright but are given varying degrees of confidence.
Many philosophers were enthusiastic (Clark, 1989, 1990, 1993, 1996; P. M.
Churchland, 1986, 1989b; P. S. Churchland, 1986; Churchland and Sejnowski,
1992; Thagard, 1988, 1989, 1990). They used connectionist ideas not just as a
way of glossing concepts, but as a way of thinking about mind in general
including epistemology, philosophy of science, and ethics.
One major implication of PDP appeared to be that the GOFAI/Chomskian
emphases on explicit rules and on innateness were each mistaken. One PDP
network not only learnt from input examples to form the past tenses of regular
and irregular verbs, but the changes in its performance over time seemed to
match aspects of infants behaviour, which Chomsky had claimed were incon-
trovertible evidence for innate explicit rules (Rumelhart and McClelland,
1986).
The Chomskians fought back. While some focused on criticizing the details
of that specific network (Pinker and Prince, 1988), others argued that PDP
models in general, because of their holistic nature, couldnt satisfy the general-
ity constraint (Evans, 1982), nor capture the compositionality, productivity,
and systematicity of language (Fodor and Pylyshyn, 1988). For Fodor (and for
Newell too: 1980, 1990), connectionism was not a theory of cognition as such,
but of its implementation.
Fodor still insists that a GOFAI-based psychology is the only [theory of
cognition] weve got thats worth the bother of a serious discussion (Fodor,
2000, p. 1). Worth the bother or not, this dispute has triggered a huge literature.
Some philosophers have claimed that PDP helps us to understand the
nature of non-conceptual content, and how it can lead to genuinely conceptual
meaning (Cussins, 1990). Paul Smolensky (1987, 1988) rebu ed Fodors critique
155
The Continuum Companion to Philosophy of Mind
of PDP (for Fodors response, see Fodor and McLaughlin, 1990), saying that
GOFAI-explanations of thought and language are approximations to finely
detailed sub-symbolic accounts. Smolenskys fellow connectionist Clark (1989,
Chapter 8; 1991) disagreed. In arguing that that PDP could deliver systematic-
ity, he admi ed that productivity was problematic for PDP systems (more so
than sequential order). But he said that compositionality neednt be built into
the basic architecture, as it is for GOFAI systems; instead, it could emerge from
it. Given the public availability of language, our brain can model individual
words and thereby create a virtual machine with formalist properties. So
Smolensky was wrong: GOFAI accounts of thinking arent mere approxima-
tions of PDP accounts, but are true (or false) descriptions of processes in the
relevant virtual machine.
The PDP group themselves had suggested such a machine: they believed
that, to be able to do logic/mathematics or engage in hierarchical thinking, the
brain must somehow emulate a von Neumann computer (Norman, 1986). But
despite various a empts (e.g. Touretsky and Hinton, 1985, 1988; Hinton, 1990;
Elman, 1990, 1993), no one has yet shown how this is possible. PDP models
still cant match the formalist strengths of GOFAI.
The eorts of many philosophers to justify either GOFAI or connectionism as
the key to the philosophy (and psychology) of mind are ill judged. The mind is
a complex virtual machine that includes both these types of computation and
probably many others (see the What is Computation? section below). More-
over, the fact that the brain consists of interconnected neurones doesnt
justify dismissing connectionism as mere implementation, or GOFAI as
non-biological. For a virtual machine is defined in terms of its computational
functions, not its physical implementation.
It follows that brain anatomy alone cant justify eliminative materialism.
It may be that the concepts of folk psychology arent useful in a scientific
psychology. But thats to say that they may not figure crucially in the brains
virtual machine. Their relevance/irrelevance cant be decided purely on the
grounds that the brain is a mass of interconnected neurones.
Varieties of Representation
Cognitive science typically posits cerebral models (Craik, 1943) that is,
representations in the mind/brain. And in positing such models/representa-
tions, cognitive scientists also oer hypotheses about the computational pro-
cesses that manipulate them (build, compare, transform, combine them).
In GOFAI, in psychological theories inspired by GOFAI, and also in
Putnams functionalism, these are thought of as symbolic representations with
compositional semantics. Another way of pu ing this is to say that the virtual
156
The Philosophies of Cognitive Science
157
The Continuum Companion to Philosophy of Mind
(Papineau, 1984, 1987; Millikan, 1984), causal regularities (Dretske, 1984, 1995a),
or biological autopoiesis (Jonas, 1966; Di Paolo, 2009). Since the concept of
representation is o en used in defining cognitive science, disagreements about
what representation is can be reflected in conflicting judgments about the
fields scope and success. Given Fodors formalist sense of representation, con-
nectionism is either a refutation of cognitive science (as Dreyfus claims) or a
mere implementational adjunct to it (as Fodor believes). Given a more catholic
definition of representation, connectionism is an interesting example of cogni-
tive science, and further examples have been mentioned above. Similarly,
dynamical and autopoietic theories and situated robotics (all of which deny
representations) must be excluded from the field unless one can show, as some
critics have argued, that they too involve (non-formalist) representations.
Because of these problems, and despite the huge influence in philosophical
circles of Fodors view, its best not to use the term representation in defining
cognitive science.
Some experimental psychologists outside cognitive science explicitly deny
the existence of mental representations. For instance, James Gibsons (1950,
1966) ecological approach emphasizes direct (computation-free) responses
to environmental cues. There have been many debates between Gibsonians
and the followers of David Marr, who analyzed low-level vision and object-
recognition in terms of a hierarchy of representations (Marr, 1982; Ullman, 1980;
Hinton, 1980; Sloman, 1989; Norman, 2002). Marrians typically complain that
although Gibsonians have reported important empirical data about responses
to perceptual cues, they fail to ask how those responses are possible.
Gibsons theory includes the notion of aordances (Gibson, 1977). He
claimed that perception doesnt only, or even primarily, provide knowledge
about the things in the external world, but rather provides (direct) knowledge
of the possibilities for action that are aorded by those things. A gap, for example,
is perceived as something that can be moved through, and a smile (or other
bodily expression of positive emotion) as aording approach without fear.
Not all aordances have to be learnt: the visual cli experiment suggests
that even very young babies can see (sic) that a sudden drop in the floor is
dangerous, and refuse to crawl over it (Gibson and Walk, 1960).
Fodor has complained that Gibson gave no principled way of deciding
what counts as an aordance (Fodor and Pylyshyn, 1981). For him, some
specific story about computations over representations is needed to explain
perception of any kind. Today, that view is less prominent.
The (non-computational) representational theory of perception has been a
philosophical embarrassment ever since Descartes. Accordingly, ideas about
direct, representation-free perception have also been developed within philo-
sophy by the continental phenomenologists. Some of them, such as Maurice
Merleau-Ponty (1962), borrowed heavily from the Gestalt psychologists; and
158
The Philosophies of Cognitive Science
all would have been more sympathetic to the Gibsonians than to the Marrians.
But none admi ed the possibility of a scientific (naturalistic) explanation of
intentionality, not even in neuroscience, never mind AI. According to them,
what counts as a representation cannot be understood in terms of information-
processing mechanisms of any kind, nor in terms of biological evolution either.
Rather, intentionality is seen as a basic philosophical notion, to be understood
(though not explained) in terms of situatedness, embodiment, in-dwelling, and/
or Dasein (see the Embodiment, Enactiveness, and Phenomenology section).
In the early days of cognitive science, the psychologist Jerome Bruner wrote
about how cognitive technologies (motor action, imagery/drawing, and
language) enable us to think (Greenfield and Bruner, 1969; Cole and Bruner,
1971). Similarly, Newell and Simon (1972) stressed the use of external memory
such as wri en sums and memos in problem-solving. And the cognitive anthro-
pologist Anthony Wallace (1965) showed how drivers regulate their journeys
by monitoring the input/feedback from roads, trac lights, landmarks, and road
signs, and from clutch, indicators, and gear shi . These cognitive scientists, and
others a er them, have provided a wealth of evidence that cultural artefacts
in general (and above all, language) enable us to think thoughts, and to raise
and answer questions that would otherwise have been beyond our powers.
However, they didnt go so far as to claim that our minds are partly
constituted by these artefacts. Nor did those early visionaries who predicted
that the availability of certain types of information technology then, only
just conceivable, but now common would deeply change the way we think
(Bush, 1945; Engelbart, 1962). Scientists and engineers in general dont normally
ask constitutive questions.
Philosophers, however, o en do. So the anthropologist Cliord Geertz (1973),
speaking in philosophical mode when recommending a non-psychological
view of culture and anthropology, said that the mind is located outside the
head. And the philosophers Clark and David Chalmers went even further,
locating the mind and the self largely in the external (physical/cultural) world
(Clark and Chalmers, 1998; Clark, 2003a, 2008b).
Clark and Chalmers argued that while there is a clear distinction between
brain (or body: Clark, 1997) and artefacts, considered as material things, there
is no principled distinction between the (abstractly defined) processes of control
that link them in behaviour. If an amnesiac constantly consults a personal
notebook to decide what to do, and even how to do it, the complex feedback
loops and information sources involved cant be neatly assigned to either brain
or world, but only to a closely coupled merger of the two. If, very broadly
159
The Continuum Companion to Philosophy of Mind
speaking, the mind is what the brain does (the credo of functionalism), then the
mind itself is partly constituted by things in the world.
Fodor (2009) isnt convinced. Computation and control, he insists, are
grounded in the brain. He insists, also, that the defining feature of mind is con-
tent, or aboutness, and that only brain processes can have content in a non-
derivative sense. If these processes continually consult, aect, and are aected
by aspects of the outside world, it doesnt follow that those aspects notebooks,
i-phones, whatever are actually parts of the mind. One doesnt have to
share Fodors views on content to agree; its not clear that theres any great
advantage, either philosophical or scientific, in making the strongly counterin-
tuitive constitutive claim. Nevertheless, some philosophers are sympathetic:
Clark and Chalmerss article won a Philosophers Annual prize as one of the
years ten best papers in philosophy. And some cognitive scientists, likewise,
were persuaded. In certain circles, extendedness and externalism have
become buzz-words.
This philosophical approach is especially well suited to cognitive scientists
talk of distributed and situated cognition. Much as AI-work on PDP refers
to distributed processing and distributed representations, so AI-work on
autonomous agents refers to distributed cognition (Bond and Gasser, 1988).
The cognition (problem-solving, as well as knowledge) is thought of as being
distributed over a number of interacting agents. And those agents are usually
described as situated, meaning that they arent controlled by a top-down
problem-solving GOFAI program but by bo om-up cues from the environ-
ment, to which they respond directly, in a near-reflex fashion (Agre and
Rosenschein, 1996; cf. Brooks, 1991a, 1991b).
Both these ideas are highlighted by the cognitive anthropologist Edwin
Hutchins (a follower of Wallace) in a study of ship navigation (Hutchins, 1995).
He shows that this skill isnt located inside the head of a single person, or
even several. It emerges from a complex coupling of individual personalities,
social roles and conventions, maps and instruments, ship-design, mariners
knowledge, problem-solving (o en spread across several crew members), and
a variety of bodily skills.
The thesis of the extended mind need not be combined with the concept of
embodiment (although the notion of situatedness is very close to it). But even
before the prize-winning article, Clark had argued for a philosophy of mind
grounded in embodiment.
160
The Philosophies of Cognitive Science
161
The Continuum Companion to Philosophy of Mind
162
The Philosophies of Cognitive Science
Consciousness
163
The Continuum Companion to Philosophy of Mind
and interprets the persons behaviour (Boden, 1972, pp. 23660, 32733;
Denne , 1991a, Chapter 13).
All those puzzling phenomena were addressed at length by Denne (1991a),
who analysed them in broadly computational terms. Other cognitive scientists
have oered general theories of consciousness somewhat similar to his, such
as Bernard Baars (1988) integrative idea of the global work space, Thomas
Metzingers (2003) self-model theory, and Slomans architectural account
described below (see also Crick and Koch, 1990; Frith et al., 1999). But Denne s
is among the most philosophically sophisticated.
Of Denne s many provocative claims, the most controversial is his position
on what Chalmers (1996) calls the hard problem namely, how to explain
(or even admit) the existence of subjective experiences, or qualia. Early critics of
functionalism declared that it couldnt account for qualia, because zombies,
behaving exactly like humans, but with no conscious experiences, are in
principle possible (Block, 1978; Zi, 1959). Denne disagrees: the concept
of zombies, he says, is incoherent, and belief in their possibility ridiculous
(Denne , 1991a, Chapter 10, p. 4; 1995). Sloman (1996c, 1999) argues, similarly,
that nothing could have the same computational architecture as us (necessary
for it to behave exactly like us), yet lack sensation.
On one key point, however, these two computationalist zombie-deniers
disagree with each other. For Denne , qualia are mere fictions (Denne ,
1988; 1991a, Chapter 12). Once everything has been said about the behavioural
aspects (the many subtle discriminative and associative activities) of seeing
blue, or tasting chablis, nothing more remains. For Sloman, by contrast, qualia
do exist. They arent visible as overt behaviour, nor verbally describable to
others, but nor are they events in some essentially mysterious immaterial
world. They are internal (but self-accessible) computational states, whose
generation and functions are possible only in complex computational
systems virtual machines of a particular architectural type (Sloman and
Chrisley, 2003).
Slomans philosophy of mind takes the design stance seriously. Drawing
on a wealth of experience in AI, and knowledge of a wide range of animal
behaviour, Sloman makes relatively specific suggestions about which sorts of
computational structures and processes could and which could not generate
particular types of cognition and control (Sloman, 1978, 1993, 2000). For
example, various types of anxiety, and the complex emotion of grief, are made
possible by distinct types of computational mechanism, some of which can
already be modelled, up to a point, in computers (Wright et al., 1996).
Sloman regards most philosophical discussion of consciousness as fixated
on highly confused concepts (Sloman and Chrisley, 2003; Sloman, 2010).
These include not only long-familiar notions (such as qualia), but also suppos-
edly more precise and more recent terminology (such as phenomenal and access
164
The Philosophies of Cognitive Science
consciousness: Block, 1995). This conceptual confusion partly accounts for the
rise of new mysterian accounts of consciousness (Flanagan, 1992, pp. 8f.),
according to which it is either u erly unintelligible to human minds (McGinn,
1989, 1991) or intelligible only in terms of arcane, even undiscovered, aspects
of quantum mechanics (Penrose, 1994; Hodgson, 1991) and/or of an informa-
tion-imbued universe (Chalmers, 1996).
Searle (1980, 1992) isnt one of the new mysterians, but he might be termed
an old mysterian. In rejecting functionalist/computational accounts of con-
sciousness (and intentionality), he insists that we know that it is a biological
phenomenon, just as digestion and photosynthesis are. This knowledge,
however, is merely the fact that we now have even more evidence than
Descartes did to believe that the brain causes/generates consciousness. What
we want to know is How? At the level of material stu (compare: lactose,
chlorophyll), the emergence of consciousness is intuitively unintelligible. Searle
oers no new ideas, mysterian or not: no micro-tubules, no dual-aspect
information, no special types of computation or computational architecture.
He simply says that the problem of consciousness is a scientific problem, and
leaves the scientists to get on with it.
There are now several conferences, and an international journal, dedicated
to machine consciousness. Sometimes, this phrase is glossed by references to
mechanisms explaining human or animal consciousness. O en, however, it
is used to suggest that computers of a certain sort are, or anyway would be,
genuinely conscious. Some such claims are philosophically (and computation-
ally) thin (e.g. Aleksander and Dunmall, 2003; Aleksander, 2005). A few are
more substantive. For instance, Sloman has outlined a specification for a
machine whose normal functioning could lead it to discover within itself
something akin to qualia, as a result of developing an ontology for describing
its sensory contents (Sloman and Chrisley, 2003; cf. Sloman, 2010, Section 13).
Many philosophers assume that life is necessary for mind (e.g. Scriven, 1953,
p. 233), and that, because computers arent alive, strong AI is impossible
(Geach, 1980, p. 81). However, the necessity of the life/mind linkage is more
o en taken for granted than explicitly justified and when arguments are given,
they are usually weak.
Cyberneticists in general assume that the same principles of control govern
both life and mind that is, they are strong continuity theorists with respect to
the ontological similarity of life and mind (Godfrey-Smith, 1994b). However,
they pay scant a ention to issues such as self-reflection and reasoning.
Even language, though sometimes mentioned, isnt considered in any detail.
165
The Continuum Companion to Philosophy of Mind
What is Computation?
166
The Philosophies of Cognitive Science
167
The Continuum Companion to Philosophy of Mind
about what computers actually do (Smith, 1985). It has developed into a philo-
sophy of presence that is a new metaphysics: an account of the emergence of
objects, individuation, particularity, subjectivity, and meaning. Mind, or inten-
tionality, is a form of active registration that requires a relatively high degree of
disconnectedness, or autonomy, as well as connectedness. This gives rise to sub-
jectivity and objectivity alike. Smith sees no fundamental distinction between
intentionality in people and computers: computation is inherently participatory,
and computers have intentional capacities ultimately grounded in practice,
analogous to the human practices stressed by Heidegger and Dreyfus (Smith, 1996,
pp. 305, 149). Both physical objects and intentional subjects, he argues, arise
from the participatory engagement of distinguishable regions of the meta-
physically basic dynamic flux. This flux is described by field-theoretic physics,
and having no objects involves neither individuality nor particularity. Objects
emerge, or are constructed, as a result of dynamic participatory relations. So where
others were broadening Dasein from humans to animals (see the Embodiment,
Enactiveness, and Phenomenology section above), Smith broadened construc-
tive dynamical interaction from animals to rocks, and even to atoms.
Smith claims to have retained the major insights of both continental and
empirical-analytic traditions, without any of their problematic ontological
assumptions. He also claims that his metaphysics gives us norms as well as
facts: a way of living right as well as a way of speaking truthfully (Smith, 1996,
p. 108). Some philosophers are deeply impressed: John Haugelands jacket-
blurb says Smith recreates our understanding of objects essentially from scratch
and changes, I think, everything. Others are only partly persuaded, allowing
that the physical implementation of computation is important although reject-
ing Smiths account of it (Searle, 1990b, 1992, p. 209). Yet others are highly scep-
tical, and also repelled by his vividly purple prose. (My own view is that he has
helped himself to the dynamic flux, his version of Kants noumenal world,
without proper licence, despite his claim, in the final 60 pages, to have pulled
this concept up by its own bootstraps.)
As weve seen, Smith isnt the only one to regard the concept of computation
(understood intuitively as what computers do) as problematic. So the core
thesis of cognitive science, like that of physicalism, should be interpreted
transparently, not opaquely (Chrisley, 2000). The claim isnt that mind can be
explained by our current ideas about computation, but that its explicable by
whatever theory turns out to be the best account of what computers do.
Conclusion
Thirty years ago, Sloman predicted that philosophers today would be pro-
fessionally incompetent if they werent well informed about developments in
168
The Philosophies of Cognitive Science
AI (Sloman, 1978, p. xiii). He was right. This chapter has described a host of
examples where the findings of cognitive science, and especially the concepts
developed within AI/A-life, have provided intriguing questions, and some-
times plausible, or even satisfying, answers, for the philosophy of mind. These
questions and answers cannot properly be ignored. Even those who reject the
fundamental assumption that there can be a naturalistic psychology should
(like Morris) at least engage lightly with some of the data/theories of cognitive
science.
Moreover, cognitive science can contribute to longstanding problems
ranging from the philosophy of science (Sloman, 1978; Thagard, 1988, 1989;
Churchland, 1989b; Whitby, 1996) and of religion (Arbib and Hesse, 1986; Boyer,
1994), through metaphysics (Smith, 1996; Sloman, 1996a), to ethics (May et al.,
1996), aesthetics (Turner, 1991; Boden, 2000b, 2007; Boden and Edmonds, 2009),
and philosophical logic (McCarthy and Hayes, 1969; Pearl, 2000). Thats hardly
surprising. Insofar as philosophy is the study of systems of thought, and ways
of knowing, it is concerned with the mind. So a science of the mind is likely to
have connections to, and implications for, most areas of philosophy.
Arguably, the study of metaphysics is dierent. But even there, the concern
is with how we should identify and think about the most basic categories of
being. AI models using natural language, for example, must o en employ
notions of space, time, and cause. The AI definitions of these categories leave
much to be desired. But AI researchers with philosophical expertise have con-
tributed not only to AI modelling but also to philosophical understanding of,
for example, causation (Sloman, 1996a; Pearl, 2000).
As that last example suggests, the potential for dialectical enrichment goes
both ways. When philosophers engage with the field in a serious spirit, they
may come up with views that influence the science itself. They may help it to
advance not only by clarifying current scientific concepts but also by oering
new insights for empirical study.
Examples of this salutary eect include Denne s work on the intentional
stance, which has been used by cognitive ethologists to guide experimentation
and theorizing on the minds of many species (Grin, 1978, 1984). (Phenome-
nologists who refuse intentionality to animals will describe these data in dier-
ent terms; but they must admit that interesting new data have been discovered
by cognitive ethologists, and that the writings of philosophers have played
some part in furthering this.) Another piece by Denne which influenced
empirical research was his (Denne , 1978) snippet on modelling the whole
animal, although some cognitive scientists, to be sure, had already worked out
that message for themselves (Arbib et al., 1974; Arbib, 1982).
Two more examples, concerning the acquisition of conceptual representations,
are due to Clark, writing in cooperation with a developmental psychologist
(Clark and Karmilo-Smith, 1993) and a connectionist AI scientist (Clark and
169
The Continuum Companion to Philosophy of Mind
Thornton, 1997). My own analysis of the concept of creativity, and of the com-
putational mechanisms underlying it, has prompted research in both AI and
psychology (Boden, 2004). Yet more instances are due to Sloman, including
his work on temporary representations (which influenced the recent dual
pathways theory of vision: Goodale and Milner, 1992), on virtual machines
and computation, on emotions and mental architecture, and on the space of
possible minds.
So the appellation cognitive scientist covers many people who are primarily
regarded as philosophers. Theres no fundamental distinction between those
who actually do science and those who restrict themselves to commenting on
it. Thats not to say that science can, of itself, answer philosophical questions.
But the days when respectable philosophy, and especially the philosophy of
mind, was thought to exclude any reference to science should by now be well
and truly over.
170
9 Representation
Georges Rey
Introduction
171
The Continuum Companion to Philosophy of Mind
CRTT
The program
Since at least the demise of behaviourism, it has been taken for granted that
people and many animals exhibit all sorts of pa erns of behaviour that seem
explicable only on the assumption that they are capable of perceiving, remem-
bering, reasoning, planning, decision making, and, for some, systematically
expressing their thoughts in natural languages. There are two general proper-
ties of such thinking things: the transitions between their states are o en
rational, and they occur o en by virtue of the intentional content of the states.
Thus, people are sensitive to deductive, inductive and abductive relations
among their thoughts, and to decision theoretic pa erns in much of their
planning and behaviour, most of which can only be expressed by reference
to the truth-valuable contents of the constituent states. Its standardly truth-
valuable contents that are at least one of the relata of imply, infer, confirm
or rationalize, as when, for example, the hypothesis that it has rained is con-
firmed by the evidence of the wet streets, and the act of taking an umbrella
is rationalized by a desire that one not get wet.2
Dualists such as Descartes and Brentano have argued that rationality and
intentionality arent assimilable to a general physical theory of the world. How-
ever, many people have thought that recent advances in various formalizations
of reasoning and computation suggest otherwise, and so have pursued a CRTT.
Building on the proposals of Alan Turing regarding the nature of computation,
CRTT postulates that there exists a medium of formal representation (a lan-
guage of thought) and a set of computational processes defined over it that
could account for rational (and many irrational) phenomena.3 Since these
computations and representations could be physically realized in the brain,
it promises an answer at least to Descartes challenge that reason cant be
extracted from the power of ma er.
There is still the intentionality of the representations to be accounted for,
and it might be feared that CRTT depends too heavily on our understanding of
artifactual computers, whose intentionality, like that of a book, is derived from
the original intentionality of its programmers. To be sure, the plausibility of
CRTT is based partly from such an understanding, and frequently specific
CRTT proposals are tested by running them on artifactual computers. How-
ever, this reliance on artifactual machines is inessential. Indeed, CRTT departs
from current research in artificial intelligence (AI) in two important respects,
(a) it is not, as AI routinely is, concerned merely with understanding how to
get a computer to solve some problem in some way or other; rather, it seeks
to understand the actual inner workings (at a computational level) of humans
and animals, and (b) it decidedly does not suppose that human or animal
172
Representation
(Non-)physical (non-)locality
Its important to appreciate both a serious problem that CRTT aims to solve and
a constraint on its solution. An organisms sensitivities begin to present dicul-
ties for any purely physicalistic proposal once one considers the detection of
non-local, non-physical properties by local physical agents. It is a fundamental
fact about the kinds of entities that include people and animals that their causal
interaction with their environment is entirely local and physical. In the case
of human beings this interaction is confined to six or seven sense modalities,
or transducer systems, that convert various physical phenomena into electro-
chemical impulses and motor systems that convert them back to bodily beha-
viour. As animals become more intelligent, the relations between these inputs
and outputs become increasingly complex, and it becomes increasingly hard
to explain how an animal or other device is capable of realizing them.
For example, successful models of animal navigation need to assume that
the animal can represent some of the spatio-temporal structure of that experi-
ence (e.g. how long it has been since it last found food at a given place). Indeed
as Gallistel notes:
173
The Continuum Companion to Philosophy of Mind
But how does an animal keep track of temporal structure in a way that
enables it to perform these computations and so sensitively modify its beha-
viour? What must its brain be like to do this? For starters, it had be er
somehow represent events as having specific temporal properties, and then
somehow store those representations for further use in combination with other
such representations in ways available for executing the vector algebra of dead
reckoning (Gallistel, 1990, 2008). Understanding mechanisms that could be
sensitive in these and other ways is one of the fundamental problems for any
psychology.
Moreover, human sensitivities to stimuli are not confined to non-local
properties. As contemporary psycholinguistic research shows, even very young
children seem to be sensitive to such categories as noun or verb that are in no
plausible sense physical properties of the stimuli (the properties, at any rate, do
not appear in physics or physiology). What kind of mechanism could possibly
produce such sensitivities? Noun phrases patently do not share any transduc-
ible property. It would seem impossible to rig a device to be sensitive to all and
only noun phrases in any way other than by in one way or another building into
the device the principles of grammar, and, with them, at least representations
of the relevant grammatical categories over which those principles are defined.
Similarly for the plethora of other non-physical properties to which people
are obviously sensitive (e.g. timidity, audacity, pomposity; being composed by
Beethoven, a fall in corporate profits, a proposal of marriage, or a declaration
of war). The only plausible way something could be sensitive to these things
is by having a mind, and the hope is that CRTT will explain how this helps
by showing how it emerges from computations over representations.
174
Representation
Referential Opacity
175
The Continuum Companion to Philosophy of Mind
(a)the terms involve a relation between real things: if x kills or kisses y, then
both x and y had be er exist: you cant kill/kiss something that doesnt
exist; this is contrast with a mental verb like think, since you can think
that x is angry without there being anything of which you think it: the
Greeks thought Zeus was o en angry, but there was no Zeus; and
(b) whatever objects are related by a transparent verb are so related no
ma er how the objects are named or described, so long as the names
of descriptions of them truly apply. Thus, if Oedipus killed Laius, then,
since Laius is, in fact, his father, it follows that Oedipus killed his father.
Or, to put it another way, if the expressions Laius and his father both
refer to the same thing, then they can be substituted for each other in a
normal, referentially transparent verbal contexts such as x killed y.
Again, this contrasts with a mental verb such as think: Oedipus can
think he killed Laius without thinking he killed his father; its the fact
that the one claim doesnt imply the other that gets him into trouble. By
contrast with kills, most propositional a itude terms such as x thinks
that p, are opaque, so to say; the light of reference doesnt shine
directly onto the referent, if any, of expressions in the direct object clause
of the term, but somehow bounces o the expressions themselves.
176
Representation
would then be false, for lack of Zeus: you cant bear a real relation to some-
thing that doesnt exist. But theres surely a reading of (b) that makes it true,
since, again, Zeus is not meaningless. Its merely empty (which Ill confine to
meaningful expressions).
So what does an empty term like Zeus represent? Well, its an interesting
fact that an almost universal response is that of Quines (1953a) fictitious
philosopher, McX: it represents an idea in your head, as many people
might be inclined to put it. But this is absurd, since (a) whatever else might
be in your head, there are certainly no bearded gods there; and, in any
case, (b) if Zeus is an actual idea in your head, then Zeus would turn out to
exist aer all!
There have been a wide variety of replies to this puzzle.6 Again, to maintain
neutrality between disputes, we should simply note that the word represent
(and, for that ma er, virtually any intentional idiom) seems to suer from a
systematic ambiguity along the following lines:
The first usage might be called the existential, the second the (purely)
intentional usage of represent (and other intentional idioms).7 Ive expressed
the second, intentional use with deliberate vagueness. It would be tempting to
say so a purely intentional use of representation of y, for lack of any y, is
really about an intentional content. But this wouldnt be correct, since someone
thinking about Zeus and his philandering ways isnt thinking about the phi-
landering ways of an intentional content. Speaking more carefully, we should
say something such as: when x represents a y that doesnt exist, a person is
standing in the thinking relation to the content [y]; but this doesnt entail she
is thinking about [y]. Even in a purely intentional usage, thinking about y is
one thing; thinking about [y] quite another.
Arguably, psychology is concerned with pure intentional content, even in
cases that might be described existentially. Although its convenient to describe
navigating birds and bees as representing, for example, the azimuth of the sun,
(e.g. see Gallistel, 1990), this seems to carry the misleading suggestion that the
birds and the bees actually represent the sun as the sun, which is doubtful;
presumably they dont really have the concept [sun], and would react just as
well to lights in a planetarium.8
177
The Continuum Companion to Philosophy of Mind
Internalist Strategies
One of the most natural ideas about meaning is that it is some sort of intro-
spectible idea inside ones mind and/or head, say, an image or a inclination
to make one inference rather than other.
The idea that mental representations are images, or perhaps maps by which
we steer (Lewis, 1994) can be traced back to Aristotle (see Cummins, 1989).
There seems to be a presumption that imagistic representation is somehow
unproblematic: an image, such as a green triangular one, represents what it
resembles. As appealing as this idea has been, it wont get us very far. In the first
place, one doesnt find entities that are actually triangular and green in the
brains of people who think about green triangles. Perhaps talk of resemblance
is just a way of talking about a correspondence between features of neural
events and real world properties, but then the naturalization problem is the
problem of specifying why one correspondence rather than another provides
the correct interpretation: as Wi genstein (1953) noted, there are infinite
numbers of projections from a triangle to arbitrary objects in the world.
Moreover, even if some mental representations can be usefully thought of
as images, it is extremely implausible that all, or even a significant proportion
of representations can be. Images simply dont combine to produce logically
complex images. What image, for example, could represent a negative fact, as
in the thought that there are no green triangles? A green triangle with a black
X superimposed upon it? How is that to be distinguished from a unnegated
image of a triangle with a black X superimposed on it? And the problem, of
course, only gets worse when one considers conditionals, quantified sentences,
modal claims, and all the indefinitely complex thoughts that people are
manifestly capable of thinking (e.g. try forming a distinctive image for If not
every green triangle is inside a square, then either all the figures are illusory
or Im blind). And over all this there hovers the problem of abstract ideas, such
as those of uncle, number and justice; what are the distinctive images for them?
While imagistic representations might well play some role in some cognitive
processes, they simply seem inadequate for any serious logical thought.
Moreover, as we already noted, in a CRTT, computations are defined over local
178
Representation
Conceptual roles
One kind of proposal that has been immensely popular in the twentieth century
is that the meaning of a representation is determined by its epistemic (eviden-
tiary and logical), conceptual role in reasoning. Logical operators provide the
most plausible examples; thus, a certain expression # might plausibly mean
[and] in virtue of the fact that thinking P#Q tends to cause thinking P and
thinking Q, which in turn tends to cause thinking P#Q (cf. Peacocke, 1992).
However, extending the account past [and] merely to other logical particles
is dicult. There are substantial disputes about or and the law of excluded
middle, as well as about how to understand conditionals. Are we to take these
dierences to be dierences in the meaning of the logical words, or simply in the
theories that people have about them (cf. Williamson, 2006)? Moving away
from purely logical cases, the problem seems even more daunting. With a li le
imagination, it seems always possible to construct a story whereby someone
(particularly a philosopher) could reasonably deny a standard role for a con-
cept while still seeming to possess it, simply by having a suciently bizarre
theory about the world. Thus, some creationists seem to be denying that humans
are animals; some nominalists, that numbers are abstract; and some idealists
that tables are material objects. There would seem to be no epistemic connec-
tion so secure that, with a li le ancillary theory, someone couldnt break it and
yet still be competent with the relevant concept. So how could some epistemic
connection provide the meaning of a concept?
A further problem, stressed by Fodor and LePore (1992), is that conceptual
roles dont seem compositional in the way that content ought to be. The content
[pet fish] ought, a er all, to be some kind of compositional function of the con-
tent [pet] and the content [fish]. However, its not at all clear that the conceptual
role of [pet fish] is a compositional function of the role of [pet] and the role of
[fish]; there may well be inferences that are peculiar to pet fish that are not
shared by pets or fish alone (e.g. only pet fish live in bowls).
It is tempting to try to limit the relevant roles to certain constitutive infer-
ences. The most famous eort to do this was the verifiability theory of meaning,
which was an a empt to spell out the meaning of a concept in terms of the
sensory evidence that would confirm or disconfirm its application. It is widely
thought that such a theory encountered insuperable problems. For purposes
here the most serious were Quines (1953b) observations about confirmation
179
The Continuum Companion to Philosophy of Mind
holism: claims are (dis)confirmed not individually, but only in conjunction with
a great many other claims, making it dicult to see how to isolate certain
epistemic connections as constitutive of the meaning of a specific claim. This
was the heart of Quines famous challenge to any theory of meaning, and to
eorts to distinguish analytic claims, such as that bachelors are unmarried,
whose truth seemed to be due to meaning alone, from claims that were simply
tenaciously believed.9
These observations about belief variability and confirmation holism have
led many advocates of a conceptual role semantics to the desperate measure
of semantic holism: all of a terms epistemic connections are constitutive of its
meaning. But this seems to make a havoc of psychology. It would entail that
any change in any of a persons beliefs would ipso facto be a change in the
content of all of their thoughts. Since by accretion, reasoning or forgetfulness,
our beliefs are constantly changing, no one would ever have the same thoughts
twice; indeed, one wouldnt remember anything with the same content. It would
be a cosmic coincidence if two people ever shared a belief, since to agree on
anything theyd have to agree on absolutely everything. There could be no
serious generalizations about mental states, not even of the well-confirmed
kind regarding visual illusions. This was perhaps no problem for Quine who
was a behaviourist and sceptical of intentional psychology. However, it is
a serious problem for post-behaviourist cognitive approaches such as that
of CRTT.
180
Representation
be their use of Aristotle: where Sophies use refers to the Ancient Greek
philosopher, Twin Sophies refers to Twin Aristotle!). If meaning is what
determines extension, then, as Putnam pithily put it, meanings just aint in
the head (Putnam, 1975, p. 227).
Whatever one thinks of Putnams examples, still a further problem can be
raised in terms of a general question for any theory that takes thought to
be formal symbol manipulation. Imagine a computer that used exactly the
same program one day to play chess, the next to fight a war; there would seem
to be no purely internal facts that would distinguish what it was representing
on the one occasion vs. the other. It would seem that something external to the
computer must determine the meaning of at least some of its representations.
Externalist Theories
181
The Continuum Companion to Philosophy of Mind
sucient for a naturalistic theory, since at every stage of such causal chains
there are events (such as ostensions, dubbings, communications, understandings)
that require intentional characterization. For example, they would seem to
require that the original dubber had one thing rather than another in mind on
the occasion of the dubbing. But what determines that it was the infant,
Aristotle, and not, for example, his nose, or the kind, Greek, human being, or
animal, that Aristotles parents had in mind? All these things, a er all, are
equally in the causal path described by the dubbers ostending finger. (This is
the qua problem, discussed at length in Devi and Sterlney, 1987.)
A natural answer to the question of what a dubber dubbed might be: whatever
kind of thing she would discriminate as that thing with the term; that is, what-
ever she would apply the term to, as opposed to everything she wouldnt.
Along these lines, a number of philosophers have considered ways in which
states involving token expressions might have, in addition to actual causal
histories, certain counterfactual dispositional properties to co-vary with certain
phenomena in the world. Intentional meaning is treated as a species of so-called
natural meaning, the kind of meaning that is said to obtain between dark clouds
and rain, red spots and measles, expansions of mercury in a thermometer and
ambient temperature. One event naturally means another if there is a causal
law connecting them, or, as Dretske (1981) put it, the one event carries informa-
tion about the other. A sentence, on this view, means what it carries information
about: the sentence its raining means that its raining, since it carries the
information (causally co-varies) with the fact that its raining.10
Nevertheless, so stated, the view is open to several immediate objections.
As stated, almost everything would mean something, since almost everything
is reliably caused by (and carries information about) something. So there must
be some further condition on genuine mental meaning. Here, most tokenings
of representations are produced in the absence of the conditions that they never-
theless mean: Thats a horse can be thought on a dark night in the presence of
a cow, or just idly in the presence of anything. In his influential discussions
Fodor (1987, 1991b) calls these la er usages wild; the property whereby
tokens of symbols can mean things that arent on occasion their actual cause
he calls robustness.
The problem for any co-variational theory is to account for robustness. In
doing so it needs to solve what has come to be called the disjunction problem:
given that among the causes of a symbols tokenings there both are meaning-
forming and wild causes, what distinguishes them? In particular, what makes
182
Representation
it true that some symbol F means [horse] and not [horse or cow on a dark
night], or [horse or cow on a dark night or w2 or w3 or . . .] (where each wi is
one of the purportedly wild causes)?11
Several proposals have been advanced for handling the disjunction problem.
They have in common trying to constrain the occasions on which the nomic
connection is meaning-forming.
Ideal co-variation
183
The Continuum Companion to Philosophy of Mind
of certain intentional states, such as a ending to, thinking of, wanting to get
things right).
In order to avoid these problems Fodor (1987, 1990c) went on to propose
another kind of co-variation, what has come to be known as the asymmetric
dependency relation. Although it makes no explicit appeal to ideal epistemic
conditions, much of its motivation can be appreciated by thinking of the ideal
co-variational theory in the background.
184
Representation
a ention here to problems that arise for any externalist theories but only a er
se ing out another important class of them.
Teleofunctional theories
Millikan (1984), Papineau (1987), Dretske (1988) and Neander (2004) have been
working on a general account of meaning based upon the role mental states
play in a biological account of the evolution and life of an organism. On this
view, intentional states possess certain information-carrying, evolutionarily
determined functions, even if there are no conditions under which they pres-
ently execute them. These functions fix the states intentional content. Thus, a
frogs tokening F whenever a black speck crosses its retinal field might be
interpreted as meaning [fly], since its flies that are responsible for the survival
of frogs. Such teleological ideas can be combined with co-variational approaches.
Thus, Dretske (1988) proposes treating meaning as the recruitment of a co-
variation between an internal state and a worldly phenomenon on behalf of
some adaptive response. The fact of the co-variation itself (say, between a state
of the frog and the motion of a fly) is an explanatorily significant cause of the
frogs shooting out its tongue at such nutritious prey.
All such teleofunctional approaches face some general problems. A worry
that has been much discussed about selectionist theories in general is that they
are Panglossian in assuming that all useful traits were evolutionarily selected.
Critics of teleosemantics, like Fodor (1987, 1991b), argue that we have no reason
to think that large parts of thought and language, and their meaning, may not
be unselected eects of whatever was selected, and so of no help in determining
semantic content.
Indeed, certain properties of thought and language do seem to outstrip any
pressures from natural selection. Insofar, for example, as many of our concepts
involve commitments to full potentially infinite universal quantifications, it is
dicult to see how they, as opposed to more modest finite cousins, could have
been selected by a finite history (see Peacocke, 1992, p. 131).
But perhaps the most serious worry about such teleofunctional approaches
to content is that, given the vicissitudes of natural selection, the wrong contents
might get assigned to psychological states. Pietroski (1992) imagines a case
in which an animal gets a racted by red flowers on high hills and increases
its selectional advantage by avoiding predators in the low lying valleys. The
teleofunctionalist would appear to be commi ed to claiming that the state of
responding to the red flowers actually has something like the content [avoid
predator], even if the animal couldnt actually recognize a predator when face
to face with it. Standard selectionist explanations of intentional states tend to
presuppose the contents of such states and so would be unable to explain them.
185
The Continuum Companion to Philosophy of Mind
A general diculty that arises for any pure externalist account is the dis-
tinctions of the mind seem to outrun the distinctions that the external world
independently provides. There are not only distinctions among concepts (such
as [renate] and [cordate]) that happen in the real world to be co-extensive, but
there are distinctions among necessarily co-instantiated concepts (i.e. concepts
that are instantiated in all the same possible worlds and/or counterfactual
situations).
There are two kinds of such co-instantiated concepts: those expressed by
necessarily co-extensive terms, such as triangle and trilateral, eucalyptus
and gum, circle and point equidistant locus of co-planar points (whatever
satisfies one in each of these pairs necessarily satisfies the other); and what
might be called the necessarily co-divided ones, such as rabbit, vs. unde-
tached rabbit parts, vs. temporal stage of a rabbit. Dierent things satisfy
each of these three expressions, but whenever an agent is presented with some-
thing that satisfies one of them shes presented with something that satisfies
the others (cf. Quine, 1960, Chapter 2; Gates, 1996). Or consider simply the
phenomenon of subception, whereby many animals are able to recognize
groups of things of certain modest cardinality (Gallistel, 1990, Chapter 10).
Do such animals plausibly have the same concept [three] that people have?
Theres this reason to think not: unlike other animals, most people have a
concept controlled by general principles, for example, Peanos axioms for
arithmetic (e.g. zero is a number; every number has a successor), whereby we
can be led by reasonings into understanding a potential infinity of complex
arithmetic truths.
Recalling our early discussion of empty terms and purely intentional uses
of represent, a particularly crucial set of cases of necessarily co-instantiated
concepts are the necessarily uninstantiated ones (e.g. [largest prime], [round
square]). One way to think of dealing with such cases is to try dealing with
them as logically complex, so [largest prime] would actually involve a logically
complex construction about of symbols meaning [large] and [prime] (see Fodor,
1990c). However, there seem to be plenty of non-complex cases. As Plato pointed
out, no one ever perceives a genuine circle: all the figures we could possibly
encounter are only very crude approximations, at best ragged complex poly-
gons.14 Moreover, although laws may arguably relate properties that happen
to be non-instantiated (e.g. some specific mass that nothing ever happens to
possess), if one limits oneself to properties that genuinely figure in scientific
laws, it would be dicult to see how there could be laws about, for example,
unicorns, angels or ghosts. Its not at all clear that these things are even meta-
physically possible, much less suciently possible for there to be scientific laws
relating them to entokenings of representations.15
186
Representation
Ecumenical Approaches
There are two ways to combine the two strategies: two factor theories and
appeals to basicality.
Two factor theories posit both an internal and external factor to meaning.
One suggestion might be to allow the representational vehicle itself to be a
component of content. This suggestion seems a natural way of distinguishing
necessarily co-instantiated concepts that have dierent constituent structure.
Thus, the thought that water is wet is distinct from the thought that H2O is wet
by virtue of the fact that the content of the one thought involves a syntactically
complex expression, H2O, and the other doesnt.
However, such a suggestion by itself is unlikely to suce for all the cases;
a er all, there can be logically simple expressions names, simple predicates
that are necessarily co-extensive (Mark Twain/Sam Clemens, Zeus/Jupiter,
eucalyptus/gum), and although they might be associated with dierent con-
tents within the mind of a single person, it doesnt seem as though they should
always express dierent contents across people. Indeed, as the above examples
187
The Continuum Companion to Philosophy of Mind
show, dierent people could have the same thought about Sam Clemens, Zeus
or gum trees without the vehicles of thought being actually spelt the same!
Intuitively, all that would seem to ma er is that the role of the vehicles in their
thought be the same.16 Consequently, some two-factor theorists (e.g. Loar, 1981;
Block, 1986a) suggest that the narrow component be identified with a terms
conceptual role, along the lines but also subject to the diculties sketched
above.
Another suggestion is to identify the narrow content with a rule in the
head that along the lines of Grice (1961/65) and Putnam (1975) may leave
some sort of blank space to be filled in by the specialist, a kind of indexical
element that permits a full semantic content to be determined by the context
with which the agent interacts, much as the semantics of indexical terms
such as I,now, this and that do (cf. Kaplan, 1989). White (1982) and Fodor
(1987) develop this strategy, generally identifying the narrow content of a
LOT expression with a function (in the set theoretic sense) that maps a context
onto a broad content. For example, the narrow content of Sophie and Twin
Sophies water is the function that maps Sophies context onto H2O and her
twins context onto XYZ. When Sophie u ers Water is wet, she thereby
expresses the content [H2O is wet], while when Twin Sophie u ers it she
expresses the content [XYZ is wet]. Two symbols have the same narrow con-
tent just in case they serve to compute the same such function: it is this that
is shared by Sophie and her twin. (See Chalmers (forthcoming) for further
development of this strategy.)
Of course, this strategy will be subject to the same Quinean worries we
raised earlier with regard to conceptual role theories: how do we distinguish
those roles that are essential and constitutive of meaning from those that are
mere ma ers of belief. A tentatively promising approach to those worries has
recently emerged in the work of Paul Horwich and Michael Devi .
Basicality?
188
Representation
Although one might quarrel with the examples and worry about the defla-
tionary context in which Horwich proposes his view, there seems to me some-
thing right about it which could be applied to CRTT, at a first pass, along the
following lines:
Note that the explanatorily basic property need not be a purely internal one
but might well involve relations to external phenomena. That is, (BAS) has
Fodors proposals potentially as special cases, cases in which the basic proper-
ties are ones about actual language use, or about how symbols manage to be
locked onto actual phenomena in the world. (BAS) is simply not limited to
such cases.17
Although (BAS) is by no means a reduction of intentionality (intentional
notions are still mentioned in it), it still does some important work. Insofar as
basic properties are suciently local, it permits conceptual stability despite
wide epistemic and other sorts of surface variation in how people use words
and concepts. The issue is not whether people agree in their surface behaviour
but whether their responses are controlled by the same basic properties, an
issue not so easily addressed. And (BAS) allows for empty concepts such as
[unicorn] and [circle], and response-dependent ones such as [funny] and [good],
where the basic properties seem to be mostly in our internal responses, not
in the variable things to which we are responding.
(BAS) also concedes to Quine that there may well be no adequate way to
define theoretical terms such as electron or species, since the basic facts in
these cases may be precisely as theoretically diuse as Quines holistic view of
theoretical confirmation emphasizes. But on the other it may allow for some
local basic facts of the sort that seem to explain the intuitions about meaning
that people have about trivial cases such as bachelor. Thus, it seems to capture
not only internalist and externalist intuitions, but also what was always reason-
ably driving a Quinean scepticism about intentional content. If this is correct,
then it may well be at least the most ecumenical strategy to pursue in trying
to provide an adequate theory of the content of mental representations.
189
10 Mental Causation
Neil Campbell
190
Mental Causation
In his paper Mental Events Davidson (1970) sought to reconcile three claims
that appear to be true yet seem to be mutually inconsistent:
The apparent inconsistency is that the truth of (1) and (2) entails the falsity
of (3). If a mental event such as my deciding to close the door causes the door
to close then (2) seems to imply that there ought to be a law connecting my
deciding to close the door and its closing, but this is just what the third claim
denies. Davidsons method of reconciliation involves a particular understand-
ing of the second claim. According to Davidson, when events stand in a causal
relation they have true descriptions that instantiate a strict law; not every true
description of the events is amenable to the formulation of such laws. In fact,
given the holism of the mental and the rational principles that guide mental
ascription Davidson argues that mental vocabulary is unsuitable for the for-
mulation of strict laws. Since only physical predicates are appropriate for the
formulation of strict laws and mental events enter causal relations with phys-
ical events, it follows from (2) that mental events have physical descriptions
and hence, are themselves also physical events. Since there are no strict psy-
chophysical laws mental concepts cannot be reduced to physical concepts,
so we have an ontological reduction of mental to physical events without a
conceptual reduction of mental to physical properties. Since it is individual
events and not properties that are identified, Davidsons anomalous monism is
a token rather than type identity theory.
Davidsons brand of non-reductive physicalism would seem to provide a
simple and elegant account of mental causation. Mental events such as deci-
sions or choices cause other events, including physical events, because they are
themselves physical events. A number of critics (Honderich, 1982, 1983, 1984;
Hess, 1981; Stoutland, 1976, 1980, 1985; Kim, 1989a, 1993a; Antony, 1989), how-
ever, have argued that anomalous monism entails the inecacy of mental prop-
erties and consequently fails to provide an adequate account of mental causation.
Although this argument takes many forms, the basic reasoning is roughly as
follows.1 Davidson faces a dilemma when it comes to the issue of mental causa-
tion. At the heart of the dilemma is the observation that we ordinarily distin-
guish between the properties of an event or object that are causally responsible
for the production of a given eect and those that are irrelevant. For example,
if I throw my glass on the floor and the impact of the glass against the concrete
191
The Continuum Companion to Philosophy of Mind
causes the glass to sha er, some of the properties of the cause seem to ma er
and some dont. The fact the glass was blue and that it contained water seem
peripheral to the sha ering, whereas the velocity at which the glass was travelling
when it struck the floor, the angle of impact and the structure of the glass seem
more important. When it comes to identifying the law that connects events like
the first with events like the second it seems only natural to suppose that the la er
rather than the former properties will be implicated. That is, there is more likely
a law connecting the structure, velocity, and angle of impact of the glass with
its breaking than one couched in terms of the colour and contents of the glass.
Since mental events are, according to Davidson, identical to physical events,
it seems that mental events have both mental and physical properties. Given the
example of the sha ering glass it is reasonable to suppose that when there
is causal interaction between a mental event and a physical event we should
be able to identify which properties of the mental event enabled it to play the
causal role it did. This is o en expressed in the form of the question, Is it
the mental event as mental (i.e. in virtue of its mental properties) that causes
behaviour, or is it the mental event as physical that has causal ecacy?
Davidsons claim that the only strict laws there can be are physical laws sug-
gests that it is in virtue of the events physical properties that it caused what
it did. That is, if it is in virtue of the law-engaging physical properties of the
impact of the glass against the floor that caused the glass to sha er, by analogy
it seems reasonable to suppose that it is in virtue of the law-engaging physical
properties of a mental event that the event caused what it did. While this is
consistent with Davidsons three claims his critics think this falls too far short
of a robust account of mental causation. For what this first option means is that
mental events cause behaviour solely in virtue of their physical (i.e. neuro-
biological) properties. That is, when I decide to get up for a drink and then rise
from my chair my rising is caused by the event that was my deciding, but my
behaviour is not caused in virtue of the fact that the cause was a deciding, was
a desire for a drink, or in virtue of any of its mental properties. These are all
as irrelevant to the production of the eect as the colour and contents of the
glass were irrelevant to its sha ering. This hardly seems like mental causation
anymore. As Jerry Fodor once famously said,
192
Mental Causation
certain physical properties of the event are causally responsible for his reach-
ing. For many of Davidsons critics this is not good enough.
Since the first option is not very appealing what if one argued for the other,
according to which mental events cause in virtue of their mental properties?
This would certainly address Fodors concern, for then his wanting would
literally be causally responsible for his reaching because it would be in virtue
of the fact that his wanting has the mental properties it does that he in fact
reaches. However, this option fares no be er for the Davidsonian because it
entails psychophysical laws. This is because, as we saw earlier, it seems that
events cause in virtue of their law-engaging properties. If a mental event
causes in virtue of its mental properties, then this reintroduces psychophysical
laws, contradicting Davidsons third claim. Worse still, the reintroduction of
psychophysical laws revives the possibility of psychophysical reduction. So
although claiming that mental events cause in virtue of their mental properties
might provide a robust account of mental causation, it does so at the cost of
mental anomalism and, at least potentially, of non-reductive physicalism itself.
Neither of the two options then, is a ractive to the non-reductive physicalist,
in which case it looks like Davidsons brand of non-reductive physicalism
stumbles on the question of mental causation.
193
The Continuum Companion to Philosophy of Mind
corpus but in events and relations in the world. For this reason he adopts what
he calls explanatory realism, which claims that a proposition C is an explanans
for E in virtue of there being some determinate relation R holding between
events c and e. Kim, then, takes R to be the explanatory relation that grounds
the explanans relation between propositions C and E.
On the realist view, our explanations are correct or true if they depict these
relations correctly, just as our propositions or beliefs are true if they correctly
depict objective facts; and explanations could be more or less accurate
according to how accurately they depict these relations. Thus, that c is related
by explanatory relation R to e is the content of the explanation consisting of
C and E; it is what the explanation says (Kim, 1988, p. 226).
194
Mental Causation
another in terms of c2. If we explore the various ways in which c1 and c2 might
be related, it turns out that the explanations fail to be complete or independent.
Kim identifies six possibilities: (1) c1 is identical to c2, (2) c1 is distinct from c2 but
is reducible to or supervenient on it, (3) c1 and c2 are both partial causes of e,
(4) c1 is a proper part of c2, (5) c1 and c2 are dierent links in the same causal
chain leading to e, and finally, (6) e is causally overdetermined by c1 and c2.
There is no need to discuss all of these options. It is clear that if c1 and c2 are both
partial causes, then neither event is sucient on its own for the eect, and so
according to Kim an explanation that appeals to either cause alone will be
incomplete because it leaves out a central causal factor. Similarly, if c1 and c2 are
sequential links in a causal chain, then the explanation in terms of c2 fails to
be independent of the explanation in terms of c1 in virtue of the dependence of
c2 on c1. Hence, Kim plausibly assumes that if the events referred to in two
explanations are not independent of one another, then the explanations them-
selves also fail to be independent. The only time there can be two complete and
independent explanations of the same event is option (6), when the event is
causally over-determined (i.e. when there are two independent causes, each
of which is sucient for the eect). Kim admits this possibility but claims that
genuine cases of over-determination are suciently rare, in which case the
principle of explanatory exclusion is a plausible general principle.
Kim has used the exclusion principle to place considerable strain on the
concept of mental causation. Suppose that George rises from the couch. On the
one hand we have an explanation for his rising that appeals to the instantiation
of a mental property, such as his desire for a beer; on the other hand, since his
rising is a physical event it seems to have a purely physical explanation in terms
of a neurobiological property. This la er claim implicitly appeals to the causal
closure of the physical domain which states roughly that for the occurrence
of any physical event there is a physical cause which is sucient for it.3
Kim claims we should find something puzzling about having both of these
explanations for why George rises from the couch:
When these two claims are viewed together, we should find the situation
perplexing and somewhat unse ling . . . We want to ask: Which really did
it? Whats the real story? The premises of the two causal explanations are
mutually consistent; however, there is something perplexing and perhaps
even incoherent about accepting both as telling us what caused Georges
behaviour, without an account of how the two accounts are related to each
other. Each explanation specifies a cause of Georges behaviour. But how
are the two supposed causes related to each other? (Kim, 1990a, p. 125)
Kim then surveys the six possibilities mentioned earlier and shows that none of
these is very promising. He claims the possibility that the mental and physical
195
The Continuum Companion to Philosophy of Mind
causes of Georges behaviour are each partial causes, distinct links in the same
causal chain, or proper parts of the same cause are highly implausible, and I
agree. Over-determination is not an option either because this would require
an unexplained coincidence of causes that is systematic. Occasional cases of
causal over-determination can be tolerated (e.g. the smouldering cigar and the
lightning strike simultaneously ignited the haystack) because circumstances, as
unusual as they may be, can lead to a coincidental convergence of sucient
causes. Appealing to over-determination to explain mental causation requires
that all intentional actions are systematically over-determined by mental and
physical causes. While this is not impossible, it is not a very a ractive option
since there is something troubling about the idea of systematic coincidences.
The only live options for Kim are identity and supervenience.
Although Kim was at one time optimistic about using supervenience to
account for the relation between intentional and neurobiological explanations
(see Kim, 1984) he points out that such an approach faces some serious
problems. The trouble is that in order to use supervenience to show that one
explanation depends on the other one must oer a characterization of the
supervenience relation between mental and physical properties that captures a
suciently robust notion of dependence, and this has been lacking in standard
accounts. Weak and global supervenience are, according to Kim, too weak,4
and strong supervenience arguably implies reduction, which is precluded by
the non-reductive physicalist. Hence, the only real hope for a solution lies in
the identification of mental with physical properties. Indeed, this is precisely
what Kim argues in his most recent work (Kim, 2005), but this is once again to
give up on non-reductive physicalism and espouse a version of reductionism.
These considerations show that the non-reductive physicalist lacks an appro-
priate account of how the two explanations of Georges behaviour are related,
in which case the neurophysiological explanation excludes the intentional
explanation. This puts the legitimacy of all psychological explanations in jeop-
ardy, for the above line of reasoning generalizes to every case where human
beings seem to act for reasons.
196
Mental Causation
2003; Gulick, 1992; Bontly, 2002; Gille , 2001; Burge, 1993b; Menzies, 2003) have
argued that Kims reasoning can be generalized to all irreducible supervenient
properties in which case there is no causation anywhere but at the fundamental
physical level, which is absurd. Another reply is to acknowledge the possibility
that mental properties over-determine their eects, though few seem to take
this idea very seriously (Bontly, 2005; Ezquerro and Vicente, 2000; Vicente, 1999;
Kallestrup, 2006; Sparber, 2005; Raymont, 2003). Finally, some authors (LePore
and Loewer, 1987) place their hopes in the idea that mental properties can be
shown to be causally relevant to physical causation and that this relevance,
though not the same as ecacy, is robust enough to rescue mental causation.
In the remainder of this discussion I would like to explore an alternative
approach. I will suggest that both formulations of the objection share meta-
physical assumptions about the nature of events that render the objections
either incoherent or question-begging, in which case non-reductive physicalism
is not as implausible as the arguments make it seem. Since the form of this
response is slightly dierent depending on the target argument, I will discuss
each in turn, but since both responses rely on the same general idea, the second
will build on the first.
The reply to the argument against anomalous monism was actually provided
by Davidson himself (Davidson, 1993), though he did not make it as perspicu-
ous as one might like. Davidsons main line of response was to claim that given
his view of events and causation the objection raised against him makes no
literal sense. Featuring heavily in this reply is his claim that causation is an
extensional relation between events.
197
The Continuum Companion to Philosophy of Mind
have the power to change things, not our various ways of describing them
(Davidson, 1993, p. 12). This means there is no room for the idea of an event
causing as mental or as physical or in virtue of its mental or physical proper-
ties; events, not their properties, cause, and to be a mental or physical event
just is to be described using mental or physical vocabulary. Hence, Davidsons
view of events as concrete particulars and his nominalism about properties
prevent the epiphenomenalist objection from ge ing o the ground. The entire
objection depends on the seemingly innocuous assumption that events have
properties and cause what they do in virtue of some subset of those properties.
But Davidsons metaphysics is incompatible with this assumption because
events are simple entities that dont have properties as constituents; properties
are instead simply ways of describing events, and if events dont care how
we describe them, as his extensionalist thesis claims, it is hard to see how one
could claim that events cause in virtue of either their mental or their physical
properties. Neither option is possible, yet it was by forcing one option or the
other that the epiphenomenalist objection got going in the first place.
So Davidsons response, with which I am sympathetic, is that the objection
is formulated on the basis of a certain view of the metaphysics of events and of
causation that is foreign to his philosophy.6 The alternative view treats events
as property exemplifications, according to which an event is a structured com-
plex of which some property is a part. This is not to say that this alternative
metaphysics is false or should be rejected; indeed, there is something very
a ractive about the account of events and causation assumed by Davidsons
critics.7 The point is that it is illegitimate to import these assumptions into
the argument against Davidsons position, unless, that is, one can show that
Davidson himself accepts these assumptions. Since it is quite clear that he does
not, the objection takes aim at a position that bears only a faint resemblance
to Davidsons own views, and hence misses the mark. That this is the case is,
I think, made quite plain in the following rejoinder from Kim:
The issue has always been the causal ecacy of properties of eventsno
ma er how they, the events or the properties, are described. What the critics have
argued is perfectly consistent with causation itself being a two-termed
extensional relation over concrete events; their point is that such a relation
isnt enough: we also need a way of talking about the causal role of properties,
the role of properties of events in generating, or grounding, these two-termed
causal relations between concrete events. (Kim, 1993a, p. 21)
While it might be true that what many of Davidsons critics want is a way of
talking about the causal role of properties, Kim continues to assume that this
way of speaking about causation and about properties makes sense within a
198
Mental Causation
199
The Continuum Companion to Philosophy of Mind
same object, but F G, then the tokening of F and the tokening of G are distinct
events. This is made quite explicit in Kims Identity Condition for events:
200
Mental Causation
a beer has a distinct cause from his bodily movement. Kims identity condition
renders the dual explananda strategy impossible, for any a empt to ground
multiple explanations of a single event in distinct properties, the event tokens
will run afoul of Kims identity condition for events. Thus the a empt to show
that a single event can have more than one explanation by fragmenting it into
multiple explananda (according to which property is tokened) has no prospect
for success within Kims metaphysics. Such an approach could only succeed on
a Davidsonian course-grained account of event identity.8 But this means there
is an important sense in which Kim has begged the question in his use of the
principle in debates about mental causation.9 The principle obviously holds
for someone who, like Kim, accepts a fine-grained theory of events, but there
are many who prefer a course-grained Davidsonian approach, and there is cer-
tainly much room for doubt about whether or not the exclusion principle holds
under the conditions of this alternative metaphysics.10 Thus, by making it seem
as though the principle of explanatory exclusion holds regardless of ones
metaphysical theory of events, Kim does the issue a serious disservice.
If I am correct that Kim and other critics of non-reductive physicalism have
assumed something like Kims property exemplification view of events, this
goes a long way to discrediting the two arguments surveyed in my discussion.
However, there is an even more serious concern here, at least about Kims use
of exclusion to argue against non-reductive physicalism. Since the argument
relies on Kims theory of events it is entirely question-begging. This is because
Kims version of the property exemplification theory of events already assumes
the falsity of non-reductive physicalism.
As we saw, a central claim of Kims theory of events is the identity condition,
which he uses to individuate events. The identity condition states that two
events are identical if and only if their constitutive elements are identical. That
is, for event [x, P, t] to be identical to event [y, Q, t], x must be identical to y,
so we have here the same constitutive object, t must be identical to t, so the
events occur at the same time and have the same duration, and property P must
be identical to property Q. So if P is a physical property and Q is a mental prop-
erty, on Kims schema the mental event [y, Q, t] can be identical to the physical
event [x, P, t] only if we can identify the mental property Q with the physical
property P. This precludes the very possibility that defines most forms of non-
reductive physicalism, namely, that mental events are physical events but that
mental properties cannot be identified with physical properties. This means
that if an argument against non-reductive physicalism assumes Kims theory of
events the argument begs the question, for the property exemplification theory
already assumes non-reductive physicalism is false. To the extent that Kims
version of the argument from exclusion depends on the property exemplifi-
cation theory then, the exclusion argument begs the question.11
201
The Continuum Companion to Philosophy of Mind
The lesson the above observations hold for those concerned about the
problem of mental causation seems to be that the problem cannot be isolated
from metaphysical questions about the nature of events and the role of proper-
ties in their individuation. Since ones assumptions about such ma ers can
have a profound eect on ones treatment of the problem of mental causation,
it seems only prudent to clear such ma ers up first, or at least to be forthright
about them from the start.
202
11 Personal Identity
E. J. Lowe
203
The Continuum Companion to Philosophy of Mind
204
Personal Identity
is identical with itself and with no other thing, one might wonder how facts of
identity can fail to be u erly trivial and uninteresting.
Part of the solution to this conundrum is provided by distinguishing, as we
must do anyway, between identity and identification. Identification is a cognitive
act and a far from trivial or easy one. One and the same object may o en be
identified in dierent ways, even by the same thinker, and it may not be evident
to such a thinker that, indeed, he or she has identified the same object in two
such ways. To be able to identify an object is, typically, to be in possession of
some descriptive information which applies uniquely to that object. But, as
Frege (1960 [1892]) pointed out, a thinker can be in possession of two such
pieces of information without necessarily thereby knowing that they apply to
the same object. To use his famous example, it was an astronomical discovery of
considerable magnitude that the Evening Star (Hesperus) is the Morning Star
(Phosphorus). Similarly, it would be a stunning discovery to find out that the
victor of Austerlitz (Napoleon) is the author of this chapter (Jonathan Lowe).
As we shall see, the role of identity criteria is to impose certain constraints
on what can count as an acceptable answer to such a question concerning iden-
tification. But in order to understand that role, we first need to say a li le bit
more about identity as such.
Identity, as has been remarked, is a reflexive relation a relation which,
of necessity, holds between everything and itself. We can formalize this as
follows:
(x)(x = x)
As was also remarked earlier, identity is subject to Leibnizs law, which for
our purposes may be formalized in this way:
Here F stands for any condition that may hold true of an object, so that the
above formula eectively arms that, for any things x and y, if x is identical
with y, then anything true of x is also true of y, and vice versa. From the fore-
going two principles, it is easy to derive two other logical properties of the
identity relation: its symmetry and its transitivity, expressible by the following
two formulas:
(x)(y)(x = y y = x)
(x)(y)(z)((x = y & y = z) x = z)
Together, these four formulas exhaust the properties of the identity relation
from a purely logical point of view. They pin that relation down uniquely, as
205
The Continuum Companion to Philosophy of Mind
Criteria of Identity
Here, RK denotes what we may call the criterial relation for objects of kind
K. And note that such a relation must, of course, be an equivalence relation
reflexive, symmetrical and transitive because identity itself is an equivalence
relation and RK has to hold between Ks just in case they are identical. The
best-known example of such a one-level identity criterion is the axiom of exten-
sionality of set theory, which tells us that if x and y are sets, then x is identical
with y if and only if x and y have the same members, so that in this case having
the same members is the relevant criterial relation. However, Frege, who founded
206
Personal Identity
the formal theory of identity criteria, favoured two-level identity criteria, which
may be wri en in the form:
Here, fK denotes what could aptly be called the K-function. The best way to
illustrate this is by means of Freges own famous example of such an identity
criterion, his criterion of identity for directions (see Frege, 1953 [1884], p. 74).
A direction (in the geometrical sense of the word) is always a direction of
something, namely, a line. And Freges criterion of identity for directions is
just this: the direction of line x is identical with the direction of line y if and only
if x and y are parallel. So, in this case, the K-function is the direction of function
and the criterial relation for directions is parallelism between lines. Observe that
both the relation of having the same members and the relation of parallelism
between lines are, as required, equivalence relations.
It should be easy to see why the two dierent forms of identity criteria
receive their respective names. A two-level criterion specifies the identity-
conditions of things of a kind K in terms of an equivalence relation between
things of another kind; thus, in the case of Freges criterion, it specifies the iden-
tity-conditions of directions in terms of an equivalence relation between lines.
In contrast, a one-level criterion specifies the identity-conditions of things of a
kind K in terms of an equivalence relation between those very things; thus, in
the case of the axiom of extensionality, it specifies the identity-conditions of
sets in terms of an equivalence relation between those sets. We shall see that
this dierence between the two forms of identity criteria is significant in the
context of a search for an adequate criterion of personal identity. For a two-level
criterion of personal identity will be appropriate only if we can think of persons
as being objects of a functional kind, in the sense that directions are.
Something more should now be said about the requirement that a criterion
of identity be non-trivial and, more particularly, non-circular. Clearly, it would
be blatantly circular to allow the criterial relation in a one-level criterion of
identity for Ks simply to be the relation of identity itself. It is true, but just
trivially so, that if x and y are Ks, then x is identical with y if and only if x and
y are identical. But sometimes a putative identity criterion can be circular in a
less obvious way: for example, the putative identity criterion for sets which
states that if x and y are sets, then x is identical with y if and only if x and y
include exactly the same sets. It is indeed logically necessary and sucient for
the identity of sets x and y that x and y include exactly the same sets (bearing
in mind that every set includes itself), but since what we are seeking is an
informative way of specifying the identity-conditions of sets, it is clearly
unsatisfactory to do so by appealing to a criterial relation in this case, the
relation of including the same sets which is itself defined at least partly in terms
207
The Continuum Companion to Philosophy of Mind
What is a Person?
Locke, as we noted earlier, very wisely observed that This being premised to
find wherein personal identity consists, we must consider what Person stands
for (Locke, 1975 [1690], II, XXVII, p. 9). We cannot hope to formulate an
adequate criterion of identity for objects of a kind K unless we have a pre y
good idea as to what Ks are. But what exactly are we asking when we ask a
question of the form What are Ks? The short, but, I think, correct answer is
that we are inquiring into the nature or essence of Ks. As for what the word
essence means in this context, we again do well to quote Locke who said that
in the proper original signification of the word it denotes the very being of
anything, whereby it is, what it is (Locke, 1975 [1690]: III, III, p. 15). From this
we may glean that, at the very least, we do not know what a K is unless
we know to what ontological category Ks belong. Unfortunately, in the case
of persons this immediately gives rise to a problem, namely, that dierent
208
Personal Identity
philosophers over the ages and across cultures have had very dierent views as
to what, in this sense, persons are. Some have held that persons are essentially
immaterial substances (spirits or souls), some that they are combinations
of such a substance with a material one (a body), some that they are purely
material substances (such as living animals), some that they are phases of such
substances (rather as caterpillars and bu erflies are dierent phases of the
same kind of insect), some that they are non-substances (such as bundles of
experiences, or functional roles that substances can occupy), some that they
are not even individual entities of any kind but rather universals of a certain
type, some that they are transcendental entities which cannot be identified
with items of any kind that are located in the world of space and time, some
that they are literally non-entities having a purely fictional status.
What is the source of this remarkably wide dierence of opinion concerning
the nature or essence of persons? Perhaps this: the key ingredient in anyones
conception of a person seems to be the conviction that, at least, he or she herself
is a person. Thus, possession of the first-person perspective is at the heart of any-
ones conception of a person, whatever else may also be part of it. A person
is, first and foremost, something that conceives of itself as thinking, feeling or
doing various things (see Lowe, 1996, Chapter 1). Such a conception is one that
requires the deployment of the first-person pronoun, I, or some expression
equivalent to that, for its articulation. But the peculiar feature of this pronoun,
from a semantic point of view, is that its competent use apparently does not
require of the user any very specific conception of what kind of thing it desig-
nates. This is why Descartes (1984 [1641], II) could famously claim to be certain
of the truth of the cogito I think and thereby certain of his own existence, while
still professing uncertainty as to what he was. In the end, of course, he concludes
that he is essentially a thinking thing, a substance whose essence is thinking
(in the broadest sense of that term) and which excludes any other property
(at least, any material property, such as shape or mass).
Locke, as we have already seen, is less prescriptive concerning the nature
or essence of persons, saying only that a person is a thinking intelligent Being,
that has reason and reflection, and can consider it self as it self, the same
thinking thing in dierent times and places (Locke, 1975 [1690], II, XXVII, p. 9).
This definition of personhood certainly builds in the notion that a person is a
self-aware subject of thought and experience, but it is far from clear that we
should take Locke to be implying, by his use of the capitalized word Being,
that persons are substances, much less that they are essentially immaterial
substances (or, indeed, that they are essentially material ones either). In fact,
it would appear that Locke held human persons to be, strictly speaking, non-
substances, with their ontological status being that of modes, or bundles of
modes (mode being Lockes preferred term for an individualized property, or
209
The Continuum Companion to Philosophy of Mind
210
Personal Identity
moral concern for your or my future as such, given that the animals that you
and I supposedly are have identity-conditions which dont entail that those
futures are ones in which you or I exist as persons at all?
Reflections such as these suggest that it is strongly built into the common-
sense conception of a person that all persons are essentially persons, so that
my ceasing to be a person would entail my ceasing to exist altogether. Lockes
definition of personhood, whatever its defects, is clearly intended by him to
have this consequence and to that extent seems to be more in tune with com-
mon sense than a view like animalism is. This, in any case, is a good point
at which to look more closely at Lockes own proposed criterion of personal
identity, not only because it is interesting in its own right but also because it
is, in eect, the first explicitly formulated criterion of personal identity to be
found and has remained highly influential. This is not to deny that preceding
philosophers were implicitly commi ed to various criteria of personal identity
which can be deduced from their writings. The point is just that Locke has
the distinction of being the first philosopher who explicitly acknowledged
the notion of a criterion of identity although he did not use that term for it
and applied it to the case of persons.
According to Locke,
[S]ince consciousness always accompanies thinking, and tis that, that makes
every one to be, what he calls self; and thereby distinguishes himself from all
other thinking things, in this alone consists personal Identity, i.e. the sameness
of a rational Being. And as far as this consciousness can be extended
backwards to any past Action or Thought, so far reaches the Identity of that
Person; it is the same self now as it was then; and tis by the same self with
the present one that now reflects on it, that that Action was done. (Locke,
1975 [1690], II, XXVII, p. 9)
211
The Continuum Companion to Philosophy of Mind
Here is one way in which one might a empt to frame Lockes proposed
criterion in the form of a one-level identity criterion, as such criteria were
formulated earlier:
where t1 and t2 are any two times at which both x and y exist (with t1 being
earlier than t2) and e is a variable ranging over individual conscious experi-
ences, such as a conscious experience of having a particular thought or under-
taking a particular action. What the foregoing formula says, in plain English, is
just this: if x and y are persons, then they are the same person if and only if any
conscious experience had by x at any earlier time is remembered by y at any
later time, and vice versa (restricting ourselves here to times at which both x
and y exist, of course, since no person can experience or remember anything at
a time at which he or she doesnt exist). This criterion entails, obviously, that
a person must always remember every conscious experience that he or she
ever formerly had. That, however, is extremely implausible. Indeed, its implau-
sibility was fairly soon exploited by Thomas Reid (1975 [1785]) to construct
a refutation of Lockes proposed criterion by means of his well-known brave
ocer example, as follows.
We can readily imagine there being an elderly general who remembers
saving the regiments standard when in ba le as a young ocer and who, as
a young ocer, remembered stealing apples as a boy. But it also seems quite
conceivable that the elderly general has entirely forgo en the boyhood episode.
Suppose, indeed, as seems prima facie conceivable, that the elderly general remem-
bers every experience of the young ocer and the young ocer remembers
every experience of the boy, but that the elderly general remembers only some
of the experiences of the boy. Then it seems to follow that, by Lockes criterion
(as we have stated it), the elderly general is the same person as the young ocer
and the young ocer is the same person as the boy, but the elderly general is not
the same person as the boy. This, however, blatantly conflicts with the transiti-
vity of identity and implies that Lockes proposed criterial relation for personal
identity remembrance of past experience, as we may call it is not, as required,
an equivalence relation. We might seek to remedy ma ers by relaxing Lockes
criterion so as to require only that a person remember some of the experiences
that he had at any earlier time in his life. (Very possibly, indeed, this is all that
Locke himself really meant to imply.) But then the counterexample can be
modified by having the elderly general remember only some of the young ocers
experiences, who in turn remembers only some of the boys, while the general
remembers none of the boys which again seems perfectly conceivable.
212
Personal Identity
213
The Continuum Companion to Philosophy of Mind
reason. It seems at least prima facie conceivable that two distinct persons, A
and B, existing at a time t2 should both stand in this relation to a single person,
C, existing at an earlier time t1. But if the relation is both transitive and sym-
metrical, this implies that A and B stand in the relation to each other. Why?
Call the relation R for short. Then, we are given that (1) A is R to C and (2) B is
R to C. But if R is symmetrical, then it follows from (2) that (3) C is R to B.
And given that R is also transitive, it follows from (1) and (3) that (4) A is R to B.
Here it may be suggested that an advocate of the modified Lockean criterion
should just bite the bullet and accept that in such circumstances A and B are
not, a er all, two distinct persons. But this simply isnt sustainable, even by the
standards of the modified Lockean criterion. For the kind of circumstances
that we are now envisaging are ones in which a single person, C, supposedly
undergoes a process of fission, spli ing into two distinct persons who go
on to build up, therea er, quite dierent and unconnected stores of autobio-
graphical memory. This, supposedly, might occur as a result of the bisection of
Cs brain into its two hemispheres, each of which is then transplanted into the
head of a dierent human body (see Nagel, 1979 [1971]). At the later time t2, it
simply will not be true to say that A stands in the R relation to B, because there
will be experiences that B had a er the fission event which are not connected
to any memory that A has at t2. Suppose, for example, that at some time a er the
fission event, B experiences a toothache. Will it be the case that at t2 A remem-
bers the past experience of someone who remembers the past experience of
someone . . . who remembers Bs toothache experience? Surely not: for the
memory-chain in question will take us back to C at the moment of fission, but
not forward from there to Bs toothache experience.
Another thing that we should bear in mind in assessing the merits of the
modified Lockean criterion is this. While it is necessary that a criterial relation
should be an equivalence relation, this is not sucient, since such a relation
is required to hold between objects of a kind K just in case they are identical.
Consequently, it cannot hold between distinct Ks. However, it may readily be
argued that the relation of connectedness of remembered past experience doesnt
meet this demand, if we are prepared to countenance, in addition to cases of
personal fission, cases of personal fusion (for instance, as a result of a reversal
of the kind of brain-bisection and double transplant operation described
earlier). For if a single person, C, existing at a time t2, stands in this relation to
both of two distinct persons, A and B, existing at an earlier time t1, then it
follows since at most one of A and B can identical with C that C stands in this
relation to at least one person who is not identical with C.
Of course, it may be objected that these imagined cases of personal fission
and fusion are purely imaginary and not really possible. But that is much too
big a debate to be entered into here. Suce it to say that such cases present
a prima facie problem for the modified Lockean criterion. A rather dierent
214
Personal Identity
215
The Continuum Companion to Philosophy of Mind
called the only x and y principle: this is the principle that, in order to se le a
question about whether an object x is identical with an object y, only facts
about x and y should be deemed relevant, not facts concerning other objects
(see Wiggins, 2001, p. 96). The proposed condition on non-branching violates
this principle, because it amounts to the requirement that person xs identity
with person y is conditional upon there existing no other person, z, in addition to
x and y as a result of some fission or fusion process involving them. However,
it may be questioned why we should regard the only x and y principle as
sacrosanct. Why shouldnt we just concede that identity can sometimes
be extrinsically determined, by being dependent on the existence or non-
existence of other objects in addition to those whose identity is at issue?
Another objection is that the new version of the Lockean criterion is at odds
with our moral convictions concerning the importance of personal survival.
For, if my survival requires the future existence of someone who is identical
with me, then it seems that, by the neo-Lockean criterion, my surviving or
not surviving can turn upon the seemingly irrelevant ma er of whether or
not I will at some point undergo fission or fusion (however far-fetched such
scenarios may seem). To this, however, it may be replied that the real lesson
of this is that my survival, in the sense in which it is or should be something
of importance to me, should be defined not in terms of my identity with some
future person but rather in terms of there being at least one such person who
is linked to me by a connected chain of quasi-memories; if there is more than
one, as in a fission case, then so much the be er, on this view (see Parfit, 1984,
Chapter 12 and Chapter 13).
Does this mean, then, that the final, non-branching version of the neo-Lockean
criterion is finally acceptable? Does it satisfy all the requirements of an adequate
criterion of personal identity? Very arguably, it does not, for it still seems vul-
nerable to a charge of implicit circularity, although one of a dierent sort from
Butlers. Recall again that the neo-Lockean criterion appeals to the notion of a
person, P, remembering or, rather, quasi-remembering some past experi-
ence, e. But now we need ask ourselves this: how are memories and experiences
themselves individuated? Such items are mental states or events. But what are
their identity-conditions? We can already rule out the Davidsonian criterion of
identity for events as a way of se ling this question because we found it to be
implicitly circular. It was so because it sought to identify events on the basis of
the sameness of their causes and eects, while also taking these causes and
eects to be events themselves. So it defined sameness of events in terms
appealing to sameness among events, a blatantly circular procedure, leaving us no
216
Personal Identity
clearer as to what the identity of events consists in. But the neo-Lockean cri-
terion likewise appears to be implicitly circular, albeit in a rather more round-
about way. For it is strongly arguable that the only adequate criterion of identity
for mental states and events will be one which makes reference to their subjects,
which, in the case of personal memories and experiences, will be the persons
who have those memories and experiences (see Strawson, 1959, Chapter 3).
Let us focus on the case of experiences, although the same reasoning
will apply equally to memories. On the view now being recommended, part of
what makes an experience of mine numerically distinct from a qualitatively
indistinguishable experience of yours is the very fact that it is mine as opposed
to yours. The only other possible distinguishing feature seems to be the time
at which an experience occurs. In short, the following seems to be a very plau-
sible criterion of identity for personal experiences:
where P1 and P2 are variables ranging over persons and t1 and t2 are vari-
ables ranging over times. In plain English, what this formula says is just this: if
x and y are personal experiences, then they are the same personal experience if
and only if x and y are qualitatively indistinguishable experiences had by the
same person at the same time. It is quite clear that the criterial relation invoked
by this criterion is, as required, an equivalence relation. But, equally, it is obvi-
ous that it appeals to the notion of sameness between persons and hence presup-
poses that notion. Accordingly if, as I strongly suspect is the case, this is the
only adequate criterion of identity for personal experiences, then the neo-
Lockean criterion of personal identity is implicitly circular inasmuch as it
will need to rely on the foregoing criterion for a specification of the identity-
conditions of the experiences to which it appeals for the purposes of identifying
persons. Clearly, at any rate, we cannot both individuate persons in terms of
their experiences (as the neo-Lockean criterion a empts to do) and individuate
personal experiences in terms of the persons having them (as the foregoing
criterion does). And to the extent that the foregoing criterion of identity for
personal experiences looks to be in good order, it is the neo-Lockean criterion
that must be rejected as inadequate (see Lowe, 2009, Chapter 7).
217
The Continuum Companion to Philosophy of Mind
it is being endorsed or being a acked. But something should be said now about
some alternative approaches. First of all, so far we have considered only the
prospects for a one-level criterion of personal identity. But on some views of
what persons are, a two-level criterion might seem more appropriate for
instance, if persons are taken to be functional states or roles that objects of
appropriate kinds can occupy. Thus, one such view would be that a persons
body, or a special part of that body, such as its brain, is the object that occupies
the functional role in question. Suppose that being a person is a functional
role of a brain (e.g. it might be taken to be the role of being a producer of first-
person thoughts). Then a criterion of personal identity could be expected to take
something like the following two-level form:
218
Personal Identity
former view. I feel myself to be some thing, with distinctive properties such as
thought and feeling, rather than my being merely some property or feature of
some other thing, such as my brain. But it must be confessed that a satisfactory
criterion of personal identity that supports this conviction is still very elusive.
On the other hand, we should be open to the possibility that personal identity
is so basic in our ontological scheme that we should not really expect to be able
to formulate such a criterion. For, as we have seen, criteria of identity for objects
of a kind K always appeal to objects of other kinds in specifying a criterial rela-
tion for K-identity. If persons are really fundamental in our ontological scheme,
we should not expect to be able to appeal to such other kinds of objects in their
case. That being so, we should probably conclude that personal identity is
primitive and simple, in the sense that nothing more informative can be said
about the identity of persons than that in some cases it just obtains and in
others not (see Lowe, 2009, Chapter 7).
219
12 Embodied Cognition and
the Extended Mind
Michael Wheeler
220
Embodied Cognition and the Extended Mind
221
The Continuum Companion to Philosophy of Mind
According to this hypothesis, there are actual (in this world) cases of intelligent
action in which thinking and thoughts (more precisely, the material vehicles
that realize thinking and thoughts) are spatially distributed over brain, body and
world, in such a way that the external (beyond-the-skin) factors concerned are
rightly accorded cognitive status. In other words, actions and loops through non-
biological structure [sometimes count] as genuine aspects of extended cognitive
processes (Clark, 2008b, p. 85). So, if the extended mind hypothesis is true, it is
not merely the case that thinking is sometimes (and perhaps sometimes essen-
tially) causally dependent in complex and intricate ways on the bodily exploita-
tion of external props or scaolds. Indeed, bare causal dependence of thought
on external factors is not sucient for genuine cognitive extension (a point
rightly emphasized by Adams and Aizawa, 2008). Rather, if the extended mind
hypothesis is true, thought must sometimes exhibit a constitutive dependence
on external factors. This is the sort of dependence indicated by talk of beyond-
the-skin factors rightly being accorded cognitive status. Stretching our thesbian
metaphor beyond reasonable limits, this is the twist in the tale of intelligent
action where the scenery and the props get a mention in the cast list.
In one short chapter, I cannot hope to give a comprehensive field guide to
embodied cognition and the extended mind. So my goal will be more modest.
I shall endeavour to cast light on a specific issue which lies at the very heart
of the contemporary debate, namely the character of, and the argument for,
the transition from embodied cognition to cognitive extension (see also, Clark
2008a, 2008b; Wheeler forthcoming a and c; Rowlands, forthcoming). Here,
then, is where I am going. In the second section, I shall present some empirical
research from cognitive science which illuminates the embodied cognition
hypothesis, henceforth EmbC. In the third section, I shall suggest that once one
has accepted the resulting picture of intelligent action, there remains a philo-
sophical choice to be made over how to conceptualize the role of the body in
the action-generation process, a choice between what Clark (2008a) identifies as
a radical body-centrism and a newly interpreted functionalism. In the fourth
section, I shall explore the connection between the second of these options and
the extended mind hypothesis, henceforth ExM. My suggestion will be that
the basic character of one of the central philosophical arguments for ExM, the
argument from parity, makes that functionalist option more a ractive. In the
fi h section, I shall seek to strengthen the emerging picture by showing how
a key element of the argument from parity may be secured.
Body Matters
As I shall use the term, orthodox cognitive science encompasses the bulk of
research in both classical cognitive science (according to which, roughly, the
222
Embodied Cognition and the Extended Mind
223
The Continuum Companion to Philosophy of Mind
224
Embodied Cognition and the Extended Mind
225
The Continuum Companion to Philosophy of Mind
Clark (2008a) observes that there are two dierent, although o en tangled,
strands of thinking at work within contemporary accounts that stress embodi-
ment. In the following passage, he unravels those strands for us.
One . . . depicts the body as intrinsically special, and the details of a creatures
embodiment as a major and abiding constraint on the nature of its mind:
a kind of new-wave body-centrism. The other depicts the body as just one
element in a kind of equal-partners dance between brain, body and world,
with the nature of the mind fixed by the overall balance thus achieved: a
kind of extended functionalism (now with an even broader canvas for
multiple realizability than ever before). (Clark, 2008a, pp. 567)
In order to see this division of ideas in its proper light, one needs to say
what is meant by functionalism, as that thesis figures in the debate with which
we are concerned here. The final emphasis is important, because although Clark
does not address the issue, the kind of functionalism plausibly at work in
the transition from EmbC to ExM is not the kind most usually discussed by
philosophers, although I think it is the kind most usually assumed in cognitive
psychology. To bring our target version of functionalism into view, we can
exploit McDowells (1994) distinction between personal-level explanations, which
are those concerned with the identification and clarification of the constitutive
character of agency (roughly, what it is to competently inhabit a world), and
226
Embodied Cognition and the Extended Mind
sub-personal explanations, which are those concerned with mapping out the
states and mechanisms (the parts of agents, as it were) that causally enable
personal-level phenomena. Functionalism, as I shall understand it here, is a
sub-personal causal-enabling theory. It is not, as it is in its more common
philosophical form, a way of specifying constitutive criteria for what it is to
undergo types of personal-level mental states. Depending on ones account
of the relationship between personal and sub-personalsub-personal levels of
explanation, one might be a sub-personal functionalist while rejecting function-
alism at the personal level. In this paper I shall say nothing more about personal-
level functionalism. My concern is with the sub-personal version of the view,
i.e., with the claim that what ma ers when one is endeavouring to identify
the specific contribution of a sub-personal state or process qua cognitive is not
the material constitution of that state or process, but rather the functional role
which it plays in the generation of personal-level cognitive phenomena by
intervening between systemic inputs, systemic outputs and other functionally
identified, intra-systemic, sub-personal states and processes.
With that clarification in place, lets return to the division of ideas recom-
mended by Clark. In the present context, it will prove useful to re-draw that
division in terms of a closely related distinction between two kinds of material-
ity, namely vital materiality and implementational materiality (Wheeler, forth-
coming c). The claim that the materiality of the body is vital is tantamount to
the first strand of embodied thought identified by Clark, (i.e. that the body
makes a special, non-substitutable contribution to cognition, generating what,
elsewhere, Clark [2008a, p. 50] calls total implementation sensitivity). On
the other hand, if the materiality of the body is merely implementational in
character, then the physical body is relevant only as an explanation of how
mental states and processes are instantiated in the material world. The link
between implementational materiality and functionalism becomes clear when
one notes that, on any form of functionalism, including the sub-personal one
presently on the table, multiple realizability will be at least an in-principle prop-
erty of the target states and processes. Because a function is something that
enjoys a particular kind of independence from its implementing material
substrate, a function must, in principle, be multiply realizable, even if, in this
world, only one kind of material realization happens to exist for that function.
And since the multiple realizability of the mental requires that a single type of
mental state or process may enjoy a range of dierent material instantiations,
the specific material embodiment of a particular instantiation cannot be a major
and abiding constraint on the nature of mind. Put another way, the implemen-
tational materiality of the mental (or something akin to it) is plausibly necessary
for mental states and processes to be multiply realizable. And this remains
true when ones functionalism and thus the level at which the behaviour-
generating causal states and processes qua cognitive are specified is pitched at
227
The Continuum Companion to Philosophy of Mind
228
Embodied Cognition and the Extended Mind
According to ExM, there are actual (in this world) cases of intelligent action in
which thinking and thoughts (more precisely, the material vehicles that realize
thinking and thoughts) are spatially distributed over brain, body and world, in
such a way that the external (beyond-the-skin) factors concerned are rightly
accorded cognitive status. To see how one might argue philosophically for this
229
The Continuum Companion to Philosophy of Mind
view, we need to make contact with what, in the ExM literature, is called the
parity principle. Here is how that principle is formulated by Clark (drawing
on Clark and Chalmers, 1998, p. 8):
If, as we confront some task, a part of the world functions as a process which,
were it to go on in the head, we would have no hesitation in accepting as part
of the cognitive process, then that part of the world is (for that time) part of
the cognitive process. (Clark, 2008b, p. 77)
The general idea here seems clear enough: if there is functional equality with
respect to governing intelligent behaviour (e.g. in the way stored information
is poised to guide such behaviour), between the causal contribution of certain
internal elements and the causal contribution of certain external elements, and
if the internal elements concerned qualify as the proper parts of a cognitive
system (state, process, mechanism, architecture . . .), then there is no good
reason to deny equivalent status to the relevant external elements. Parity of
causal contribution mandates parity of status with respect to the cognitive. But
if the general idea of the parity principle is clear enough, the details of how
to apply it are not, so we need to pause here to get clear about those details
(for a similar analysis, see Wheeler, forthcoming c).
One interpretation of the parity principle is suggested by the way in which
it is applied by Clark and Chalmers themselves to the near-legendary (in ExM
circles) case of Inga and O o (Clark and Chalmers, 1998). In this imaginary
scenario, Inga is a psychologically normal individual who has commi ed to
her purely organic (neural) memory the address of the New York Museum of
Modern Art (MOMA). If someone asks her the location of MOMA, she deploys
that memory to retrieve the information that the building is on 53rd Street. O o,
on the other hand, suers from a mild form of Alzheimers, but compensates for
this by recording salient facts in a notebook that he carries with him constantly.
If someone asks him the way to MOMA, he automatically and unhesitatingly
pulls out the notebook and looks up the relevant fact, viz. that the museum is
on 53rd Street. Clark and Chalmers claim that there is a functional equivalence
between (a) the behaviour-governing causal role played by O os notebook,
and (b) the behaviour-governing causal role played by the part of Ingas brain
that stores the same item of information as part of her purely organic memory.
By the parity principle, then, O os memory turns out to be extended into the
environment. Moreover, argue Clark and Chalmers, just as, prior to recalling
the information in question, Inga has the non-occurrent dispositional belief
that MOMA is on 53rd Street, so too does O o, although while Ingas belief is
realized in her head, O os is realized in the extended, notebook-including
system.
230
Embodied Cognition and the Extended Mind
Other critics of the parity principle have appealed to the psychological data
on various extant inner cognitive capacities, as delivered by cognitive science,
in order to construct similar failure-of-parity arguments (e.g. see Adams and
Aizawa, 2008 on primacy and recency eects in organic memory; for discus-
sion, see Wheeler forthcoming a and c). The general version of the worry, how-
ever, is this: if (a) the relatively fine-grained functional profiles of extant inner
cognitive systems set the benchmark for parity, then (b) any distributed (over
brain, body and world) systems that we might consider as candidates for
extended counterparts of those cognitive systems will standardly fail to exhibit
full functional equivalence, so (c) parity will routinely fail, taking with it the
parity argument for cognitive extension.
Right now things might look a li le bleak for a parity-driven ExM, but
perhaps we have been moving too quickly. Indeed, it seems to me that the kind
of anti-parity argument that we have been considering trades on what is in fact
a misunderstanding of the parity principle. To see this, one needs to think more
carefully about precisely what the parity principle, as stated above, asks us to
do. It encourages us to ask ourselves whether a part of the world is functioning
as a process which, were it to go on in the head, we would have no hesitation
in accepting as part of the cognitive process. So we are encouraged to imagine
that exactly the same functional states and processes which are realized in the
actual world by certain externally located physical elements are in fact realized
231
The Continuum Companion to Philosophy of Mind
232
Embodied Cognition and the Extended Mind
of that state or process that ma ers, but rather the functional role which it
plays in the generation of personal-level cognitive phenomena by intervening
between systemic inputs, systemic outputs and other functionally identified,
intrasystemic, sub-personal states and processes. There is nothing in this schema
that requires multiple realizability to be a between-the-ears phenomenon. So
functionalism allows, in principle, for the existence of cognitive systems whose
boundaries are located partly outside the skin. It is in this way that we arrive at
the position that, following Clark, I shall call extended functionalism (Clark,
2008a, 2008b; Wheeler forthcoming a and c).
We have seen already that there will be functional dierences between
extended cognitive systems (if such things exist) and purely inner cognitive
systems. So, if extended functionalism and the parity principle are to fly
together, what seems to be needed is some kind of theory that tells us which
functional dierences are relevant to judgements of parity and which arent.
To that end, here is a schema for a theory-loaded benchmark by which parity
of causal contribution may be judged (Wheeler forthcoming a, b and c). First
we give a scientifically informed account of what it is to be a proper part of a
cognitive system that is fundamentally independent of where any candidate
element happens to be spatially located. Then we look to see where cognition
falls: in the brain, in the non-neural body, in the environment, or, as ExM pre-
dicts will sometimes be the case, in a system that extends across all of these
aspects of the world. On this account, parity is conceived not as parity with the
inner simpliciter, but rather as parity with the inner with respect to a scientifically
informed, theory-loaded, locationally uncommied account of the cognitive. So the
parity principle now emerges not as the engine room of the extended mind, but
as an heuristic mechanism that helps to ensure equal treatment for dierent
spatially located systems judged against an unbiased and theoretically moti-
vated standard of what counts as cognitive. It is a bulwark against what Clark
(2008b, p. 77) calls biochauvinistic prejudice.
This idea of a scientifically informed, theory-loaded, locationally uncom-
mi ed account of the cognitive is tantamount to what Adams and Aizawa
(e.g. 2008) call a mark of the cognitive. In the interests of expository elegance,
I shall default to Adams and Aizawas term. The most obvious next step in
this dialectic would be for me to specify the or, given the possibility that
the phenomena in question will reward a disjunctive account, a mark of the
cognitive. In the next section I shall make a tentative proposal.4
Newell and Simon, two of the early architects of artificial intelligence, famously
claimed that a suitably organized physical symbol system has the necessary
233
The Continuum Companion to Philosophy of Mind
and sucient means for general intelligent action (Newell and Simon, 1976,
p. 116). As anyone familiar with cognitive science will tell you, a physical sym-
bol system is (roughly) a classical computational system instantiated in the
physical world, where a classical computational system is (roughly) a system
in which atomic symbols are combined and manipulated by structure sensitive
processes in accordance with a language-like combinatorial syntax and seman-
tics. I shall take it that the phrase means for general intelligent action points to
a kind of cognitive processing. More specifically it signals the sort of cognitive
processing that underlies the same scope of intelligence as we see in human
action . . . in any real situation behavior appropriate to the ends of the system
and adaptive to the demands of the environment can occur, within some limits
of speed and complexity (Newell and Simon, 1976, p. 116). What we are con-
cerned with, then, is a human-scope cognitive system. Notice that the concept
of a human-scope cognitive system is not a species-chauvinistic notion. What
ma ers is that the system exhibit roughly the same degree of adaptive flexibil-
ity we see in humans, not that it have our particular biological make-up, species
ancestry or developmental enculturation.
Against this background, Newell and Simons physical symbol systems
hypothesis may be unpacked as the dual claims that (a) any human-scope cog-
nitive system will be a physical symbol system, and (b) any physical symbol
system of sucient complexity may be organized so as to be a human-scope
cognitive system. In eect, then, the hypothesis is equivalent to the claim that
being a suitably organized physical symbol system is the mark of the (human-
scope) cognitive. To unpack that claim, the physical symbol systems hypothesis
advances a scientifically informed, theory-loaded account of the (human-scope)
cognitive, one that supports a computational form of functionalist theorizing.
But can it tick all our boxes by being a locationally independent account too?
The answer, it seems, is yes. For while classical cognitive scientists in general
thought of the symbol systems in question as being realized inside the head,
there is nothing in the basic concept of a physical symbol system that rules
out the possibility of extended material implementations. Indeed, as I shall
now argue, the idea of an extended physical symbol system has much to recom-
mend it.
In a series of compelling treatments that combine philosophical reflection
with empirical modelling studies, Bechtel (1994, 1996; see also Bechtel and
Abrahamsen, 1991) develops and defends the view that certain human-scope
cognitive achievements, such as mathematical reasoning, natural language
processing and natural deduction, are the result of sensorimotor-mediated
interactions between internal connectionist networks and external symbol
systems, where the la er feature various forms of combinatorial syntax and
semantics. It is useful to approach Bechtels suggestion (as he does himself)
234
Embodied Cognition and the Extended Mind
235
The Continuum Companion to Philosophy of Mind
236
Embodied Cognition and the Extended Mind
237
The Continuum Companion to Philosophy of Mind
238
13 Current Issues in the
Philosophy of Mind
Paul Noordhof
In the broadest terms, the issues which lie at the heart of discussions in
philosophy of mind have not changed and are unlikely to change in the near
future, or even, I hazard a guess, the quite distant future. People will still seek
to understand the nature of consciousness in its various forms; to understand
the nature of intentionality or, indeed, other ways in which our mental life
may concern the world around us; to describe, and account for, the special
access each of us has to our own mental lives; to scrutinize the basis of self-
consciousness; to worry about whether mental phenomena have an appropri-
ate causal explanatory impact, and so on. Nor do new philosophical theories
in these areas arrive thick and fast. Instead, at dierent times, dierent theories
and their motivations receive particular development and emphasis. These
facts about philosophical discussion give rise to unfortunate moments when
placed on the spot by inquiring vice chancellors and sceptical governments
or other, prospective, suppliers of research funding. Nevertheless, since we are
among friends here, we can acknowledge it without embarrassment.
Other awkward moments o en occur when some of the folk just mentioned,
or the others professing to have an interest in the role of philosophy in intel-
lectual life at large, remark that while the themes upon which it focuses are
big (with the slight suggestion, about some of them, that philosophers are on
a fools errand), the current contributions of philosophers are technical and
specialized, as if, given that we can all recognize the themes, we should all be
able to, without too much study, appreciate what progress has been made.
However, precisely because philosophical progress is one of developing our
understanding of particular types of theories by se ing them out in more detail
and/or developing the motivation for them, it is entirely unsurprising that the
statement of these developments will seem more specialized and technical
than is desirable. Philosophers should make every eort to help with this but
not to apologize for it.
Once you get down to the requisite level of detail for the progress and details
to show up, identification of the main themes currently in play and suggestions
about how these will develop, will be controversial (I guess it is no news that
239
The Continuum Companion to Philosophy of Mind
philosophers love an argument) and more than likely receive dierent spins
by dierent philosophers working in the field. That might explain why my
overriding sense is one of trepidation. Nevertheless, I shall suppress this in
what follows and state as clearly as I can what seem to me the main points of
emphasis in a selection of the fields detailed above.
Specifically, I shall discuss how development in our understanding of phys-
icalism has more or less stabilized around an approach which reveals some-
thing rather interesting: first, some of the reason for believing it is undermined
while, second, it is more entrenched as a starting point because working within
this framework brings useful explanatory rigor. I shall discuss how the debate
about mental causation is slowly turning into one which focuses on the onto-
logical commitments of thought and talk about the mental, which purports
to be about physical properties which are not to be identified with those of
physics. Then I turn to a empts to dismiss dualist intuitions regarding the
explanatory gap between neural properties of the brain and phenomenal con-
sciousness independent of an admission of ignorance, or development of a
theory of phenomenal consciousness. I explain why it is increasingly recog-
nized that such a empts fail. Recent theories of phenomenal consciousness
have sought to explain it by appeal to representational properties and a separate
theory of subjective awareness (roughly, a theory of how we are conscious of
the phenomenal content determined by representational properties). I explain
how the appeal to representational properties has deepened our understanding
of the motivation for invoking qualia and issues in the philosophy of perception
and intentionality. I then outline how apparently distinct theories of subjective
awareness have converged. In the final section, I examine relatively new
approaches to the understanding of intentionality which a empt to come to
terms with the diculties that proponents of reductive theories of intention-
ality faced. In this context, I briefly touch on the putative normativity of the
mental.
I mourn not being able to discuss the many interesting developments in
our understanding of self-consciousness, self-deception, mental illness and
introspection as well as some of the fascinating texture of more specific mental
states such as that of imagination, auditory perception and so on. It proved
impossible to do so in a piece of acceptable length and some kind of structure.
Physicalism
At the risk of losing a few readers at the beginning, it seems to me that there is
more general agreement now as to how we should go about characterizing
240
Current Issues in the Philosophy of Mind
241
The Continuum Companion to Philosophy of Mind
Just as broadly physical objects are composed from, and thereby nothing
over and above, arrangements of narrowly physical objects, so broadly physical
properties are those which are constituted from, and, thereby, nothing over and
above arrangements of, narrowly physical properties. The idea is relatively
unproblematic in the case of objects. Integrated spatial arrangements of small
scale physical objects seem naturally to result in large scale objects. In many
cases, physics will concern these small scale objects and other sciences focus
on the larger scale. Opponents of physicalism o en remark that physics con-
cerns the large scale too (Crane and Mellor, 1990). Wise physicalists will not
deny this. They will just insist that any object not recognized by physics will
be composed from objects which are so recognized.
Unfortunately, talk of constitution is not so easily taken across to properties.
If we take properties to be universals so that, for example, there is only one
property of being a hydrogen atom then it is hard to see how the property of
being CH4 methane can be composed of the properties of being a carbon
atom and being a hydrogen atom (Lewis, 1986). There are just two of these
la er properties. So it cannot be the case that the property of being methane
is made up of, at least, five elements: one property of being a carbon atom,
and four properties of being a methane atom.
If we take the relevant notion of property to be property instances, then this
case becomes less problematic. There is no diculty in thinking of the property
instance of being methane as composed, in part, from one property instance
of being a carbon atom, and four property instances of being a hydrogen
atom. Instead, the diculty is that the connection between narrowly physical
property instances and broadly physical ones doesnt seem invariably to fit
this model. In the case of CH4, the arrangement of the properties of atoms
identified is the way of being methane. There are many dierent ways in which
arrangements of instances of narrowly physical properties may be arranged to
make up an instance of a property of being a mountain, or a cell, or an ant-eater.
What is the connection between a particular way in which some narrowly
physical properties are arranged and the broadly physical property which is
said to result in these cases? This is the well known phenomenon of variable
realization applied to non-mental cases.
One kind of minimal answer appeals to supervenience. Supervenience
has been characterized in a number of dierent ways and ge ing clear on the
precise connections between all the various dierent alternatives is, perhaps,
not the most engrossing way of spending a Sunday a ernoon. A currently
popular one holds that physicalism is true of our world if a minimal physical
duplicate of it, as far as arrangements of narrowly physical properties are
concerned, is a duplicate simpliciter (Jackson, 1998, p. 12). For some word, w,
physical properties broadly conceived are just any which meet the following
condition. Given P is instantiated in w, then it is instantiated in all minimal
242
Current Issues in the Philosophy of Mind
243
The Continuum Companion to Philosophy of Mind
nor broadly physical) that breaks the connection between a particular arrange-
ment of narrowly physical properties and the mental property for which they
are putatively sucient: a blocker. Then, intuitively, the connection between
the arrangement of narrowly physical properties and the mental property is
not tight enough. The possibility of the connection being blocked seems to
imply that the connection is merely nomological and, hence, no dierent from
that which holds between arrangements of narrowly physical properties
and the non-physical mental properties characteristic of emergent dualism.
Nevertheless, appeal to the idea of minimal physical duplicates sets aside such
worlds because the relationship is broken by non-physical properties and,
hence, the possibility does not reveal that physicalism is false (Hawthorne,
2002, pp. 1046).
Some argue that, as a result, physicalism should be understood in terms
of suciency in the absence of blockers and, hence, reject the intuitive verdict
that physicalism is false when there is the possibility of a non-physical blocker
(e.g. Leuenberger, 2008, pp. 14860). Prima facie, this is unacceptable for the
reason identified above. The possibility of blocking suggests that the connec-
tion is no be er than emergent dualists endorse. Moreover, if the connection
between arrangements of narrowly physical properties and mental properties
is loose enough to be blocked by non-physical properties, then the connection
between the arrangements of narrowly physical properties and mental proper-
ties holds of nomological necessity, in which case, there is no reason the con-
nection could not also be blocked, if the physico-psychological laws were
dierent, by the presence of some of the same narrowly physical properties
which, in our world (lets say), were responsible for the instantiation of mental
properties. So there is no worry about excluding the worlds with non-physical
blockers by talking of minimal physical duplicates. The looser connection
would show up in the minimal physical duplicate worlds too. (I discuss one
line of objection to this point under the second diculty that this formulation
of physicalism faces below.)
Blockers dont have to be causal blockers, although talk of stu which one
shouldnt add when preparing food, or algoplasm, which makes phenomenal
properties disappear, suggests that this is standardly how they are conceived
(Hawthorne, 2002; Leuenberger, 2008; also Montero in this volume). Instead,
blockers may disrupt a relationship in which at least one of the relata is, partly,
relational. For example, I might have the property of being alone in a room
which I (or my counterpart) do not possess in a minimal narrowly physical
duplicate of my world because, in that world, there is a non-physical poltergeist
in it with me. In these circumstances, there is no reason to expect that, in a mini-
mal physical duplicate world with dierent laws, the property of loneliness
may fail to be instantiated because one of the physical properties has become a
blocker. However, all this shows is that we should refine our understanding of
244
Current Issues in the Philosophy of Mind
245
The Continuum Companion to Philosophy of Mind
246
Current Issues in the Philosophy of Mind
realization with no common property. What Moore is not taking as the basis
of his ethical non-naturalism, is that there is an intrinsic action-guiding or
ought property. Yet, it is the la er which would threaten the characterization
of physicalism. In those circumstances, we would have Moore claiming that,
since there are intrinsic ought properties, non-naturalism is true while, at the
same time, we would be asserting that the fact that other properties stand in
the very same way to arrangements of narrowly physical properties shows
that these other properties are natural.
Roll forward to the most prominent current defender of ethical non-
naturalism: Russ Shafer-Landau. He describes his position as taking the same
approach to moral properties as non-reductive physicalists do to the mental
(Shafer-Landau, 2003, pp. 728). He accepts that moral facts are intrinsically
normative but seems to take this as no more problematic than the properties
of phenomenal consciousness. It is not that he supposes that this intrinsic
normativity is what makes moral facts non-natural and, hence, a problem for
those who seek to characterize physicalism by appeal to the same kind of
relationship of supervenience. So the proposal does not need defence against
those who appeal to the very same relationship to characterize ethical non-
naturalism.
Even if we dont need to appeal to some kind of explanatory relationship, it
is reasonable to ask whether there is some kind of explanation of the relation-
ship. I do not want to rule this out, indeed, far from it. The point of defending
the supervenience-only approach is simply to keep our explanatory options
open. For example, consider currently the most popular explanation on oer:
the subset view. According to it, a property F is realized by another property G
if and only if the causal powers of an instance of the property F are a subset of
the causal powers of an instance of the property G (This is a rough preliminary
formulation, see Shoemaker, 2007, pp. 1213, 2231 for details). It is question-
able whether this is, in fact, true of any variably realized property (e.g. see
Noordhof, 1997, p. 246; Noordhof, 1999b, pp. 11314). However, for the sake
of argument, suppose that it is. Taking it as a general requirement places restric-
tions on the kind of explanation we should look for in every case in which we
suppose that the mental supervenes upon the narrowly physical.
Endorsement of it is behind Jaegwons Kims conclusion that there is a
non-physical residue to the mental: qualia (a particular kind of phenomenal
property, see below). He takes it that reduction of the mental to the physical
requires that it is functionalizable and accepts that, in the case of qualia, exhaus-
tively specifying their nature in functional terms is not going to be plausible
because of the possibility of spectrum inversion (Kim, 2005, pp. 16573). The
conclusion, of course, does not follow if other kinds of explanation are allowed.
If physicalism only requires that mental properties are metaphysically necessi-
tated by physical properties, then there are various ways in which this might
247
The Continuum Companion to Philosophy of Mind
248
Current Issues in the Philosophy of Mind
249
The Continuum Companion to Philosophy of Mind
Either of (C1) or (C2) is will be compatible with the evidence that the
physical world is causally closed. However, only the first would establish that
non-physical mental properties, if they exist, are epiphenomenal.
A lot of energy has been focused, though, on whether a particular form
of physicalism, non-reductive physicalism, is compatible with the ecacy of
mental properties. If it is not, then non-reductive physicalism has no advantage
over property dualism, given the success of the over-determination argument.
Suppose that mental properties supervene upon physical properties in the
way identified above. Suppose that, in particular, an instance of a physical
property P1, p1, necessitates an instance of a mental property M1, m1 and that,
for any candidate eect of m1 we accept (from considerations of the causal
closure of the physical) that it is caused by p1. (As others do, I simplify. p1
would obviously be an arrangement of narrowly physical properties). To fix
ideas, consider m1 to be an extended period of pain, which eventually becomes
unbearable, from leaving ones hand on a hot plate, p1, a firing of neurones
in the brain specifically related to pain and the target eect: the behaviour of
withdrawing ones hand from the hot plate.
250
Current Issues in the Philosophy of Mind
The challenge can come in two forms. The first concerns whether the instances
of mental properties can be ecacious if the instances of physical properties
are. As the question is o en put, is there any further work for the mental prop-
erties to do? (For details of this kind of argument, see Ravenscro , in this
volume). The second, and more pressing, challenge is whether, even if these
instances of mental properties are ecacious in virtue of their relationship
with instances of narrowly physical properties, they are ecacious in virtue of
being mental properties.
Responses to the first challenge emphasize that the instances of mental
properties are either identical to instances of narrowly physical properties
or realized by arrangements of them (identity: Robb, 1997; Macdonald and
Macdonald, 1986, 1995; Macdonald, 1992; realized: Noordhof, 1999b, 2010;
Shoemaker, 2007). Appeal to identity is a ractive because it reproduces the
solution to how mental events and states can be ecacious by being token
identical to physical events and states for property instances. They are men-
tioned in a footnote by Campbell but assimilated to the event response he
favours on behalf of Davidson (Campbell, in this volume). However, the dis-
tinct virtue of the property instances approach is that it allows for finer grained
identifications of relevance than allowed by Campbell, given his rejection of
the property exemplification view of events. The main problem is to justify
the claim that mental property instances are identical to physical property
instances, given that they are distinct properties.
Appeal to realization avoids this diculty, and is independently a ractive
because it seems implausible to say that instances of mental properties are
identical to narrowly physical properties as opposed to realized by them. Such
approaches, though, are said to come unstuck when they try to a ribute to
the mental properties ecacy that they concede belongs to arrangements of
narrowly physical properties in virtue of which they are realized.
There are various a empts to deal with this relying, for example, on the
claim that the causal profile of mental properties is a subset of the causal profile
of its realizer or that they share ecacious property instances (Shoemaker,
2007; Paul, 2007). Opponents of the first have asserted that instances of mental
properties are redundant because they recapitulate a subset of the causal role
that the physical property instance already brings (e.g. OConnor and Churchill,
2010, pp. 513). This seems unfair. If mental properties have a causal profile
which is a subset of the causal profile of the narrowly physical properties, then
it is reasonable to hold that anything which is a manifestation of this profile
reveals the ecacy of mental property instances. It is in virtue of this part of
the profile that a narrowly physical property is ecacious. Opponents of the
second have questioned whether instances of mental properties do anything if
their ecacy is the result of, for example, sharing an instance of mass with an
251
The Continuum Companion to Philosophy of Mind
252
Current Issues in the Philosophy of Mind
253
The Continuum Companion to Philosophy of Mind
Phenomenal consciousness
It has become standard to distinguish between what Ned Block dubbed access
consciousness and phenomenal consciousness. A state is access conscious if and
only if it is poised to play a certain role in our practical deliberations or cogni-
tions (e.g. see Block, 1995, p. 231). Obviously quite a lot is packed into talk of
poised here. Intuitively, some state may be poised to play a role, although
we are not conscious of its presence, because the state has an impact on the
workings of our minds at the relevant points. So less illuminatingly we may say
that access consciousness just involves a being aware that we have a certain
state, or being aware of it. I dont want to presume here that access conscious-
ness is awareness of facts rather than objects. Nothing more is added to this
awareness. By contrast, a state is phenomenally conscious if, and only if, there
is something it is like to be in that state.
Many reductive theories of phenomenal consciousness break down phe-
nomenal consciousness into two components. The first component concerns
the content of phenomenal consciousness, what we may dub the phenomenal
content (some call it phenomenal character). The phenomenal content is just a
characterization of anything it is like for a subject to undergo the mental life
for example, his or her pain, the taste of banana cake, the sound of the waves
breaking on the shore, perceiving that there is a desk lamp lit on the desk, the
sense of yearning to go outside. All of these count. To introduce a stipulation
now, which I will discuss in more detail later, let phenomenal properties
be those properties which determine the phenomenal content of our mental
states. Thus, there is a phenomenal property of my experience of the waves
breaking on the shore which determines that the phenomenal content of my
experience is the waves breaking on the shore. Let manifest objects and proper-
ties be those things which our experience concerns, for example, the waves
(object) breaking (property).
The second component derives from the fact that it seems at least possible,
according to some theories of phenomenal content, that states should have
it and yet not be conscious because the subject is not aware of the state instan-
tiating the relevant content. So there is a story to tell about what makes the
subject aware of the phenomenal content of a particular state he or she is in.
This second component is o en called subjective awareness. I will only discuss
access consciousness in the context of this la er notion since, arguably, a theory
of access consciousness is a theory of subjective awareness.
Discussion of phenomenal consciousness in recent years has centred on
two dierent issues. The first concerns the proper explanation of why it
seems that phenomenal consciousness resists explanation in a way which is
acceptable to physicalists. The second concerns specific theories of phenomenal
254
Current Issues in the Philosophy of Mind
255
The Continuum Companion to Philosophy of Mind
Therefore,
(3)M = A(P).
Here M stands for some phenomenal property, and A(P) for an arrange-
ment of narrowly physical properties which are supposed to realize the
phenomenal property. Because there are no descriptions we can know a priori
are true of M, we cannot know a priori that M has a causal role, R. Therefore,
256
Current Issues in the Philosophy of Mind
257
The Continuum Companion to Philosophy of Mind
258
Current Issues in the Philosophy of Mind
The basic form of the answer seems obvious and remarking upon it is
quite familiar. It is that, obviously, the reason why these things seem to be very
dierent is that when we experience the arrangements of narrowly physical
properties that, allegedly, realize an experience, we have a dierent object of
experience to what that experience concerns, for example, the waves breaking
upon the shore. Yet it is the la er that characterizes what it is like to have the
experience.
By itself, this comment goes very li le way to resolving the situation. Sup-
pose the current object of my experience is a certain arrangement of physical
properties which, let us presume for the sake of argument, realizes our experi-
ence of a red wall. Then what we want to know is why our experience of
this arrangement of physical properties in no way reveals the fact that the
arrangement is an experience of a red wall, or if it does, then in what way it
does and how can this be squared with what it is like to be in that experience.
We cannot just trade upon the fact that there are dierent objects of experience,
we want to understand how one object of experience should manifest itself
in the other object of experience.
The point can be driven home if we consider the special case in which we
are experiencing an arrangement of physical properties which either is exactly
like the arrangement of physical properties which is a realization of this kind of
experience in another or one which is very like the arrangement if, for example,
we are having an introspective experience of one of our own experiences. In
the former case, we want to know why our experience of the arrangement of
physical properties in another does not reveal that it is an experience of the
same type as we are currently undergoing. In the la er case, we would be having
an introspective experience of e* (an experience very like our introspective
experience of it) and asking why it in no way reveals that it is very like the
arrangement of physical properties which realize the introspective experience.
Looked at this way, an a empt to bridge the explanatory gap has two
components. First, we need a theory of what makes a particular experience
of a property P. Second, we need an account of what makes P show up with
the character it has in our experience. In the next section, I will be considering
various accounts of the first, including the currently fashionable one which
I favour: representationalism. The second component will not receive a ention
though it is important. As many have noted, if we a empt to bridge the
explanatory gap by focusing on what our experiences concern, then the
explanatory gap is reintroduced regarding the properties the experiences
concern. For example, if colour is a profile of surface reflectance properties,
then why does it look the way it does in experience? However, the room for
manoeuvre here is far greater. The world is not blurred, but my blurred per-
ception due to myopia has a perfectly satisfactory explanation to bridge the
non-blurred-blurred gap.
259
The Continuum Companion to Philosophy of Mind
Phenomenal content
260
Current Issues in the Philosophy of Mind
(e.g. that I have a fig now) and, outside the mental arena, sentences have truth
conditions, and suppose that they will also be the basis for an account of the
phenomenal content of an experience. For example, some of the representa-
tional properties of a particular perception may determine that the content of
the experience is that there is a grey elephant at such and such a location.
Although, in one sense of determine, what determines the phenomenal content
of the experience is that the elephant and the property of being grey are con-
stituents. In another sense, the one to which the representationalist appeals,
the representational properties determine the phenomenal content by certain
objects and other properties being the manifest objects and properties of the
experience. Representationalists o en, and increasingly controversially,
motivate their account by appealing to the fact that, in characterizing what
an experience is like, we dont notice properties of the experience so much as
go right through to what the experience concerns the manifest objects and
properties suggesting that representation is at work to determine that content.
It is reasonably clear how the representationalist approach may seek to
make a contribution to the question of the explanatory gap mentioned at the
end of the last section. If we had an account of the representational properties
of an experience, we could see what elements of an arrangement of physical
properties helped to explain how that experience was an experience of a red
wall. To illustrate, rather than because the account is independently plausible,
if what made an experience of a red wall was the fact that it was causally cor-
related with red walls in optimal circumstances, then experiencing these causal
facts would be to experience what made that experience have the character
it does.
A acks to the representationalist picture come from, broadly, two angles.
On the one hand, critics seek to identify experiences which have phenomenal
dierences but where it is plausible that their representational content is the
same (Peacocke , 1983; Block, 1990, 1996; Shoemaker, 1990, 1991). On the other
hand, critics emphasize the obvious dierences between two very dierent
kinds of mental states: beliefs and perceptions (e.g. Martin, 2002).
The general form of the debate under the first heading has been that, for
every phenomenal dierence identified by the critics of representationalism
which dierence is supposed to encourage the introduction of qualia as addi-
tional phenomenal properties the representationlist identifies a representational
dierence a er all. Thus qualia are usually thought to make a dierence to
what an experience is like by being possessed by the experience without
being representational properties and, for some, without being the objects of
awareness. Others are inclined to accept that they are objects of awareness and,
hence, in one respect, the way in which they determine the phenomenal content
of our mental lives is by being manifest properties se led by the third account
of phenomenal content appealing to brute awareness.
261
The Continuum Companion to Philosophy of Mind
262
Current Issues in the Philosophy of Mind
263
The Continuum Companion to Philosophy of Mind
the state in question in fact, partly, presents that character. Thus, alleged non-
presentational states in fact involve higher order states which are presentations
of some of these lower order states properties (for a suggestion of this kind of
response, see Byrne, 2001; Noordhof, 2003b).
A number of issues are raised by this possibility. The first is that there are
cases of phenomenal dierence where it is part of our understanding of this
phenomenal dierence that it does not involve a presentation of this dierence.
The classic example is blurred vision. When we blurriedly experience some-
thing, our experience precisely doesnt seem to be an experience of something
which is blurred (Crane, 2006, pp. 1301). A second is that if we do take puta-
tive non-presentational states as involving higher order presentations, are
we, in eect, undermining the a raction of representationalism by, as it might
seem, allowing that what is represented in this case are the very phenomenal
properties that representationalists were seeking to do away with, viz. qualia?
I raise the second issue only to set it aside. I guess your view about that will
depend upon what you pack into the idea of qualia. In principle, representa-
tionalists should not be concerned about what counts as the manifest properties
of experience so long as it is representational properties which determine that
they are the manifest properties.
The idea that there might be phenomenal dierences which correspond to
representational dierences and yet which dont involve presentations of
these properties the blurred vision case may, fruitfully, be related to the
idea that there might be dierent degrees of phenomenal presence. Here are
two examples. Suppose that you are currently experiencing a red tomato.
Then it seems plausible to say that our experience is of something which has a
backside. Indeed, one might think, the phenomenal content of our experience
would be rather dierent if we experienced the tomato as not having a backside
rather than as experiencing it as having a backside. How should we capture this
dierence? Second, many feel happy about saying that our visual experiences
present colours and shape to us but they balk at saying it presents water as
opposed to something which flows, is transparent, etc. But just as with the
case of the backside of objects, it seems undeniable that there is a phenomenal
dierence at some level between experiencing water and experiencing this
more cautiously described kind of thing.
It is tempting for the representationalist to deal with these cases by
suggesting that the phenomenal content is the result of the judgements we are
inclined to make on the basis of the experience. Thus, it is because we dont
take our blurred experience of the world to be an experience of a blurred world
that the blurriness is not presented as a property of the world. But the imple-
mentation of this strategy is not straightforward. For example, suppose that
you are convinced that the forest you are currently seeing is an elaborate holo-
gram. Then you wont be inclined to take the experience to be of trees and so
264
Current Issues in the Philosophy of Mind
265
The Continuum Companion to Philosophy of Mind
266
Current Issues in the Philosophy of Mind
similarity. Thus it is said that when we hold that perception and hallucination
are phenomenally similar, that just means that they are indiscriminable to
introspection alone. The account of why they are indiscriminable may vary.
All that is denied is that phenomenal similarity should be understood simply
in terms of similarity in phenomenal content (Martin, 2006, p. 369). In the case
of perceptual experience, the determinant of its phenomenal content is that it
involves a certain relation to an object and its properties. In the case of halluci-
nation, no account is given of its phenomenology. Some hold that that is because it
has no phenomenology (Fish, 2009). Others may allow that it has a phenomenal
content but just not like that of perceptual experience.
Giving an account of the various ways in which hallucination may be
indiscriminable from the corresponding perceptual experience and resolving
the question of whether hallucination does have a phenomenal content are
pressing issues for proponents of the relational view. I think it is fair to say that
it is upon their treatments of these issues that the success of their approach
turns. Nevertheless, as we remarked, representationalists have a diculty of
their own: the apparent substantial phenomenal dissimilarity between percep-
tual experience and belief or judgement.
The standard strategies are to hold that perceptual experience involves a
richer, perhaps non-conceptual, content. Thus the phenomenal dierence is to
be accounted for by supposing that there are more representational properties
at work in the case of perception and that what is represented is organized non-
propositionally (e.g. spatially and/or without requiring that the subject who
has the states possesses the relevant concepts; see, for example, Tye, 1995).
While these are plausible dierences from belief, it is questionable whether
they will do. Representionalists hold that representational properties have
phenomenal significance. Thus even in the case of belief, one would expect its
representational properties to have some phenomenal significance. The belief
that there is a dog barking should have a whi of a dog about it. The belief
that it is brown should give us a flash of brown. But none of this is the case.
For this reason, representationalists are going to have to appeal to an analy-
sis of the second component of phenomenal consciousness, namely subjective
awareness. They need to do this anyway to explain why there is no phenome-
nology in the case of unconscious mental states which, we may presume, have
representational properties. But perhaps, in addition, such an appeal will help
dierentiate between conscious judgement or belief and conscious experience.
Subjective awareness
Theories of subjective awareness divide broadly into those which take the
consciousness of a certain state to be an intrinsic fact and those who take it to
267
The Continuum Companion to Philosophy of Mind
This theory raises in particularly stark form many of the problems which have
plagued all extrinsic accounts of consciousness, so I will rehearse them here
first. Possibly the most familiar objection is that it makes our experience of
our own conscious states highly fallible. The internal scanning mechanism
may misfire so that either we introspect that we are in a dierent state to the
one we are, in fact, in, or we think we are having a certain mental state when
we are not. You could, for example, perceive that you are in serious pain when,
in fact, thats just a false case of introspection.
In the case of pain, it is possible to construct an evolutionary argument
in favour of radical fallibility being extremely unlikely. Our pain system is
important for us to recognize when our bodies are being damaged and crea-
tures which fail to appreciate when this is happening are unlikely to survive
long enough to pass on their genes. On the assumption that the capacity to
introspect is coded into our genes as one might expect, given it is a sensory
mechanism we can expect considerable accuracy (Lycan, 1996, p. 18). This
is much less obvious in the case of other mental states. It would be productive
to consider this argument on a case by case basis and consider whether our
relative confidence concerning our judgements about whether we are in those
states correspond to what we might predict by the evolutionary story.
268
Current Issues in the Philosophy of Mind
Nevertheless, rather than focus on this, I want to discuss more immediate and
challenging diculties that follow on from this initial observation.
If what makes a mental state conscious is that we perceive it, then what
about the objects in the world? They are the objects of perception. Do we
thereby make them conscious? For example, gaze at a rock. Does your gazing
make this a conscious rock? You might reply that, yes, the rock is an object of
consciousness but it is not, itself, conscious. However, in making this response,
you are presuming that the conscious state is the product of the perceptual/
introspective mechanism. That is not the position of those who adopt this
theory of consciousness. Their claim is that the conscious state is the one
which is the input into the process the object of the perceptual-introspective
mechanism.
So, instead, the reply that is that rocks arent conscious because we do not
call them mental states (Lycan, 1996, pp. 234). This reply is either weak or
under-described. If the denial reflects a stipulation regarding our use of the
term conscious, then the reply seems weak. It concedes that, in terms of the
reality, there is considerable similarity between the state the rock is in, as an
object of perception, and the state one of our mental states is in, as the object
of mental perception. It just denies that we apply conscious to the former. The
situation would be no dierent to remarking that sparkling white wine pro-
duced by the champagne method could not be champagne if it did not come
from the champagne region. This might be an important point regarding mar-
keting but, unless there is something else to be said on the taste front, entirely
uninteresting to the drinker. On the other hand, if it is suggested that we limit
applying conscious to mental states because, in addition, they are states of a
cognitive/aective system and this aects the nature of consciousness, then the
reply is certainly under-described. It is also questionable whether the result is,
strictly speaking, a perceptual model of consciousness since now, the a ribu-
tion of consciousness to something is not just a ma er of it being perceived but,
in addition, these further facts Ive mentioned. This is probably why, as I can
tell, the foremost proponent of the perceptual model William Lycan seems
to have in mind a response of the former sort, with its a endant weakness.
The third challenge that the perceptual model faces concerns what we should
say if the perceptual-introspective mechanism misfires and proclaims that
we are undergoing a certain state when we are not. What are mental halluci-
nations like? If you claim that, when a subject hallucinates that she, say, is
imagining a green frog, then it is exactly for her as if she were imagining this.
Then it seems that we have to allow the possibility that there are some intrinsi-
cally conscious states, our hallucinations. If we allow intrinsic consciousness
here, then why not everywhere? (cf. Byrne, 1997, pp. 1212, with regard to
higher order thought theory).
269
The Continuum Companion to Philosophy of Mind
270
Current Issues in the Philosophy of Mind
conceptual capacity the capacities needed to have thoughts and also relates
it to self-consciousness because the higher-order thought is characterized as
that I am in such and such a mental state. Those who wish to a ribute phenom-
enal consciousness to other animals and children at an early stage of develop-
ment, o en balk at this commitment. The only way this can be resolved,
I suspect, is if we have a much clearer idea of what is involved in the basic
conceptual capacities at work. Naturally, higher-order thought theories down-
play their sophistication, but the question is whether they can do so while, at
the same time, making it legitimate to think that the states they postulate
are genuine thoughts rather than non-sensory higher-order representations
of some non-thought-like kind (e.g. Rosenthal, 1986, p. 344; Rosenthal, 1991a,
pp. 323; Rosenthal, 1993).
It is standard to appeal to the presence of higher-order perception or thought,
rather than the disposition to have these things to account for consciousness.
The reason oered is that being conscious of something is an occurrent state
rather than a dispositional one (Rosenthal, 1993, pp. 2089). However, this
seems to be a confusion. Once it is recognized that the conscious state is the
state which the disposition to have an introspection or thought concerns, then
we have an explanation of in what sense being conscious of something is
occurrent namely, that there is an occurrent state which is the object of a
disposition.
One thing that either type of theory emphasizes is that the higher-order
perception or thought itself is not conscious. What would make it conscious,
in turn, is an even further higher-order thought or perception about it. This
is usually thought to be an unsatisfactory feature by those who propose a
reflexive intrinsic theory of consciousness (e.g. Kriegel, 2009, Chapters 1 and 4).
Such theorists begin by pu ing forward something like the following.
271
The Continuum Companion to Philosophy of Mind
accept this description of their position. So the interesting question is how they
avoid it.
One suggestion has been that the conscious state has the causal roles both
for the experience that there is a desk lamp there and the awareness of this
experience. Co-instantiation of the causal roles explains the reflexivity of the
state (Shoemaker, 1994, pp. 2425). Obviously co-instantiation of roles by
itself wont do. There are any number of roles which may be coin-instantiated.
What is of particular significance is that one of the roles is that of an awareness
of the experience. At the minimum, this means that one of the roles must be of
a state which represents the experience in question. This makes the proposal
a version of representationalism in which the correct account of the repre-
sentational properties of awareness of the experience is to be given in terms of
causal role.
Some have argued that appeal to coincidence of roles a dispositional
ma er is once more ill-suited to explain the occurrent nature of consciousness
(Kriegel, 2005, pp. 378). If what I have argued before is correct, then this is
a mistake. The occupant of these coinciding roles is occurrent. In any event, it
has influenced the currently most prominent exponent of an intrinsic theory of
consciousness to provide a weakened version of this approach to take this
into account (Kriegel, 2005, pp. 4451). According to Uriah Kriegel,
M is conscious if there are M* and M** such that both M* and M** are proper
parts of M, M is a complex (not merely a sum but an integrated whole
like a molecule) of M* and M** and M* represents M by (indirectly)
representing M**. (Kriegel, 2009, pp. 2268)
The idea is that M contains proper parts, M* and M** such that M** causes and
hence is represented by M* and M* indirectly represents M by representing
M** in roughly the way that a painting can represent a house by representing a
portion of it, the rest being obscured by trees etc.
Obviously a lot of work is being done by talk of an integrated whole and the
idea of indirect representation. Since there seems nothing to rule out integrated
wholes in which representation is only of a part, talk of representation of a part
of the overall state, which thereby represents a whole, requires a particular
kind of, as yet unspecified, integration. To remark it is unspecified is not to say
that, over time, such an account cannot be provided. The question is whether
indirect representation is what we need. In the pictorial case, although the
whole house is represented, it is plausible that only part of the house the part
not occluded is phenomenally present in the strongest sense. The remainder
may have a weaker grade of phenomenal presence, depending upon the out-
come of the discussion mentioned earlier. However, it is questionable whether
this is what we want in the case of our consciousness of a mental state. Is there
272
Current Issues in the Philosophy of Mind
some part of our conscious state of which we are not conscious when we
are conscious of it? A natural candidate would be the reflexive element.
However, proponents of the reflexive account of consciousness usually empha-
size that the reflexive character is part of what we are aware of (e.g. Kriegel,
2009, p. 117).
Nevertheless, perhaps this can be turned into a virtue. It has o en been
remarked that we are not conscious of the nature of our mental states but just
conscious of what they are of in being conscious of them. It might be argued
that a lower grade of phenomenal presence implied by indirect representation
nicely explains this.
One of the motivations for the moving to an intrinsic, reflexive account of
consciousness was the conviction that unconscious states were not the kind
of states to make the states they concern conscious and yet, if it was insisted
that the higher order states were conscious, a vicious regress would ensue.
Obviously, one might ask the same question of M*. Are we conscious of it and,
if so, what makes us so. Kriegels answer is that it is part of a globally conscious
state. However, it seems we may have replaced an infinite regress with an
explanatory circle. We are conscious of M (and M**) by having M* representing
M** and conscious of M* by having M* as part of M. If that move is allowed to
the proponent of reflexive consciousness, why isnt an equivalent available to
the proponent of an extrinsic account of consciousness.
Nor does it seem that this approach to consciousness avoids the possibility
that we could have M** without M* and hence the same questions arise as to
whether we would, in those circumstances, have a hallucinatory mental state
or no phenomenal state at all. Reflexive accounts of consciousness are likely to
have to draw upon the same materials as extrinsic accounts of consciousness
here.
Although we began this section with two theories in play, it appears that
they have converged upon something remarkably similar. A conscious state
involves, at least, two parts. The first is the state of which we are conscious; the
second is the element which makes us conscious of it. The second element,
therefore, cannot be present and provide us with consciousness independent of
the first element. There is nothing it is like just to be in a state with the second
element. The dierences lie in the account of indirect representation and the
claim that the second element is part of the state of which we are conscious.
The plausibility of these claims relate back, I have suggested, to the diering
degrees of phenomenal presence we may recognize in our conscious states.
If this convergence is to be evaluated further, work is required on the content
of the second element. I have suggested that an appeal to a causal relationship
between it and the first isnt mandatory to explain its content. Perhaps a suc-
cessful explanation of the content may restore some of the dierence between
the two approaches once more. It also seemed that the second element couldnt
273
The Continuum Companion to Philosophy of Mind
274
Current Issues in the Philosophy of Mind
objects and properties which do not exist. Allowing that this is so is compatible
with also allowing, as externalists insist (see Sawyer in this volume for a good
discussion of externalism), that what contents a subject entertains may be set-
tled by the environment. Even if you allow that a subject thinks about water
rather than twater because they have water in their environment, it does not
follow from that that there can be no circumstances in which the subject thinks
about water without there being any in the environment. However, when we
turn our a ention to the case of perceptual experience, it has seemed to a grow-
ing number that it does not have these features. We cannot perceive a dog
unless the dog is present and we stand in a relation to it.
This has given rise to an interesting range of research questions. Can there be
non-representational intentionality as well as representational intentionality?
Are there both propositional and non-propositional forms of intentionality?
Can intentionality come in both conceptual and non-conceptual forms? What
are the relationships between the answers to these questions?
Such questions have given impetus to a field which appeared to have been
going through a bit of an arid spell. If we compare the flurry of activity in the
late 1970s and 1980s with regard to providing a reductive analysis of inten-
tionality (i.e. in terms compatible with the truth of physicalism) with the cur-
rent state of aairs, it seems clear that the emphasis has changed (Field, 1978;
Dretske, 1981, 1988); Fodor, 1984, 1987); Millikan, 1984, 1986, 1989); Papineau,
1984, 1987, 1993a); Stalnaker, 1984; Block, 1986c; Cummins, 1989; Whyte, 1990,
1991). To some extent, the reductive programme has stalled. As far as causal-
informational accounts have concerned, there has been li le further develop-
ment since Fodors A Theory of Content back in 1990. The most significant
competitor Millikan and Papineaus appeal to biological function still faces
controversy over whether it successfully has dealt with the criticisms concern-
ing indeterminacy and error which it was introduced to supply (Millikan, 1984;
Papineau, 1987, 1998b; see Ravenscro and Rey in this volume, for a description
of the disjunction problem which lies at the heart of this debate). A disappoint-
ing feature of the current lack of further developments has been the tendency
to use toy reductive accounts of intentionality in the development of represen-
tationalist accounts of consciousness and the like both to resolve (in the case
of bodily sensations) and present diculties (in the case of inverted earth)
as if the toy accounts in question were close to being right as opposed to
having been shown to be, more than likely, mistaken.
One consequence of this situation is that there has been renewed interest
in the nature of conscious intentionality with the suggestion that the solution
to the diculties which bedevil earlier reductive accounts will be found here.
For example, Galen Strawson takes experience (conscious experience) to be the
primary form of intentionality because he believes that only cognitive experien-
tial qualities will bring sucient determinateness to our intentional states to
275
The Continuum Companion to Philosophy of Mind
276
Current Issues in the Philosophy of Mind
the required materials will turn on a detailed examination of the grounds for
a ributing these functions, and a resolution of the issue of indeterminacy.
At the moment it looks like dierent a ributions of content may be a ributed
to an organism compatible with oering an explanation of their cognitive and
conative processes in terms of biological functions. However, perhaps this is a
failure to consider these ma ers at the right level of detail.
Further consideration of biological function also has utility in discussing the
second strain of recent investigation into the nature of intentionality to which
I wish to draw a ention. Problems with developing a naturalistic account
of intentionality have led some philosophers to consider the extent to which
intentionality, and the states which possess it, have normative properties. There
seems to be a sense in which it is legitimate to hold that if you perceive that
a white rabbit is nibbling grass to the le of the burrow, and you consider the
ma er, you ought to have the corresponding belief. If you grasp the concepts of
2 + 2 and 4, then you ought to arm that 2 + 2 = 4 (if you consider the ma er).
If you possess the concept dog, and are in the business of making dog judge-
ments, then you ought to apply the concept to dogs you come across and so on.
These oughts appear to be normative oughts and are related to the nature of
concepts and the meaning of words.
Apart from the intrinsic interest in understanding these normative claims,
the focus on normativity has promised to provide traction on the issue of
coming to a final evaluation of the possibility of a reductive account of inten-
tionality. Insisting upon the normative nature of content provides one way of
articulating why reductive accounts of intentionality arent possible. Everybody
knows you cant naturalize morality; now, the claim continues, we can see why
we cant naturalize intentionality either. Taming this normativity, on the other
hand, provides a way of keeping the option open.
A distinction is o en drawn between two ways in which these oughts can be
taken: as related to a norm or standard or as action-guiding (e.g. Ha iangadi,
2007, pp. 378). It is pre y much uncontroversial that they can be understood
in the first sense and, so understood, present no more diculty for naturalism
than the diculty it already faces with regard to intentionality in general. The
question is whether they should be understood in the second. Do they provide
guidance to action independent of our desires to follow these standards?
The action-guiding sense of ought in this context is compared with the case of
morality in which it is supposed that, when it is said we ought to behave
or ought not to behave in a certain way, this is independent of our desires to
behave, or not behave, that way.
It seems unlikely that the oughts relating to content have any action-guiding
role independent of the aims of the mental states in which they figure. Never-
theless, it is less clear whether we may take these aims to be rooted in our
interests. For example, is the aim of belief to be accounted for in terms of our
277
The Continuum Companion to Philosophy of Mind
desire to act on the basis of what we take to be true (see Noordhof, 2001, for a
suggested positive answer)? Or is believing the truth desirable in a moral sense
(see Horwich, 2006, for a suggestion along these lines)? If the aims of mental
states are not rooted in our interests, then there may be scope to recognize a
class of categorical content-related oughts in the context of the states through
which these contents are possessed. However, I am not convinced that a posi-
tive answer to this last question threatens the reductive programme any more
than if they are conditional upon our interests. Categorial oughts may be more
controversial, in that it is questioned whether we really ought to do such and
such if we are uninterested in doing so, but the threat to the reductive pro-
gramme if any does not stem from where the normative properties are a rib-
uted but that they are at all. It is just as bad to suppose that if you have certain
desires, you ought to do such and such. What exactly does it mean to a ribute
such a property to the world (if thats what you are doing)?
Biological function provides a potential source of these aims which would
enable an understanding of them in physicalistically acceptable terms. How-
ever, the application of such an appeal to the understanding of the aims of
mental states is not straightforward. Consider the case of the aim of belief. It is
o en remarked that the aim of belief to be true explains why we cannot
consciously believe at will. How should we interpret the cannot? On one
interpretation, the cannot is as strong as metaphysical necessity. It is literally
not possible for a creature consciously to be able to do this. Appeal to biological
function has no capacity to explain this unless it is not possible for there to be
non-evolved creatures with beliefs.
Perhaps mental states can only have aims if the creature whose states they
are is biologically evolved. Then we would have to conclude that, either no
non-evolved organisms can have beliefs or that beliefs should be understood
by their causal profile independent of the aim. For ease of discussion, I shall
adopt the la er option (which is independently defensible (see Noordhof, 2001;
for a view to the contrary, see Velleman, 2000). It does not ma er for the issue
raised below since it can be simply re-cast for a state with the causal profile
alone. In any event, it is plausible that there may be creatures with states
with that causal profile which are non-evolved. So there is a limitation to the
biological explanation of the impossibility of consciously believing at will
right here.
However, things are worse than that. Appeal to biological function seems
in no be er shape to explain why creatures cant consciously believe at will
even if they are biologically evolved and the cannot is simply nomological. It is
a familiar fact that things with a certain biological function can malfunction
(e.g. the human heart can fail to pump blood). While we might accept that crea-
tures that can consciously will that they are in a state with the causal profile of
belief may die out quickly, there should be anomalies just as there are in all
278
Current Issues in the Philosophy of Mind
other cases. Creatures are born with hearts that dont work; why arent they
born with consciously willable beliefs? In my view, the proper answer to this
question reveals something about the nature of consciousness, viz. that it
makes manifest the a ractiveness of being disposed to act on what you take
to be true (Noordhof, 2001).
Working out the connection between functions derived from our biological
heritage, features of consciousness, cognitive aims and so forth provides a
rich domain of study. It enables us to think about the role of various kinds of
normativity in the proper understanding of our mental life.
Concluding Remarks
279
Glossary
action. Actions have been central to modern philosophy of mind ever since
Descartess contemporaries first criticized the account of mental causation sug-
gested by his substance dualism. Since behaviour involves physical movements
it was deemed that its causes must also be physical on pain of having to explain
how, when and where dualistic causation takes place. Descartes located it
in the pineal gland, a view rejected by modern substance dualists. Other posi-
tions in the philosophy of mind, such as those of property dualism, anomalous
monism, behaviourism, identity theory and eliminativism have all been moti-
vated by concerns related to the causation of behaviour.
While some philosophers use the terms action and behaviour interchange-
ably, others reserve the former for behaviour that is intentional and/or volun-
tary (at least under some description). Voluntary mental acts, such as the act of
calculating inside ones head, pose a problem for any species of behaviourism
which treats all behaviour as (necessarily) being publicly observable.
In the philosophy of mind actions are typically identified with events and/or
processes, however there is much dispute over which events actions are to be
identified with. Some insist that actions are identical to bodily movements. For
example, Davidson (who maintained that all actions are intentional under some
description) identified actions with movements of our bodies. Yet as Hornsby
has rightly cautioned, we must not conflate my moving my body with the
(mere) bodily movement I bring about when I move my body (for the term
bodily movement may be used in both a transitive and an intransitive sense).
This motivates the competing view that actions are the causes of bodily move-
ments (according to Hornsby, for example, all actions are tryings which may or
may not cause our bodies to move). We may wish to further follow Hornsby in
also distinguishing between the thing one did and (the event of) ones doing it
(thus mirroring the Fregean distinction between the thing one believes and
ones believing it). A third view, put forth by von Wright, identifies action with
the causing of an event. While this does not immediately rule out the possibility
that actions are events, it is perhaps misguided to always seek a precise location
of Xs causing of Y, as evidenced in the literature by various counterintuitive
claims concerning the spatio-temporal location of killings involving slow deaths.
Such issues regarding the ontology and individuation of actions are of
particular relevance to the question of whether reasons are causes which in
turn relates to questions concerning causation, agency, control and (ultimately)
free will.
280
Glossary
C. S.
Danto, A. (1973), Analytical Philosophy of Action, Cambridge: Cambridge University
Press.
Hornsby, J. (1980), Actions, London: Routledge.
Moya, C. (1990), The Philosophy of Action, Cambridge: Polity Press.
Stout, R. (2005), Action, Teddington: Acumen.
animal minds. There is much debate over the extent to which some or all
non-human animals can be said to have minds. The arguments typically
revolve around what it is to have a mind, and in particular what it is to have a
so-called mental state such as a belief or desire. For example, some philo-
sophers maintain, contra behaviourism, that a dog cannot believe that the cat
is at the top of the oak tree unless it has the concept of an oak tree, where a
concept is a linguistic representation of some kind. The question of animal
minds is thus also closely related to questions about language and/or concept
acquisition. A dog may now desire to go for a walk immediately, but it cannot
now desire to go for a walk next Tuesday. Frankish here contrasts a behaviour-
based concept of mind with the language-involving concept of supermind.
C. S.
Beko, M., and D. Jamieson (eds) (1996), Readings in Animal Psychology, Cambridge:
MIT Press.
Glock, H. J. (2000), Animals, Thoughts and Concepts, Synthese, 123, 3564.
Searle, J. (1994), Animal Minds, Midwest Studies in Philosophy, 19, 20619.
281
Glossary
capable of predicting and explaining the true causes of any given happening
only trades in strict causal laws. These assumptions frame his famous argument
that if mental events do in fact cause physical events then they must be logically
identical to some physical event or other, since only an explanation couched at
the level of an ideal physics can possibly get at the true cause of any physical
happening.
D. H.
Davidson, D. (1980), Mental Events, in Essays on Actions and Events, Oxford:
Clarendon Press.
(1987), Problems in the Explanation of Action, in Metaphysics and Morality,
Oxford: Blackwell.
Macdonald, C. (1989), Mind-Body Identity Theories, London: Routledge.
M. C.
Denne , D. (1998), Brainchildren: Essays on Designing Minds, Cambridge, MA: MIT
Press.
Haugeland, J. (ed.), Mind Design II, Cambridge, MA: MIT Press.
282
Glossary
behaviourism has also been associated with Ryle and Wi genstein, who both
emphasized the conceptual relations between behaviour and mental ascrip-
tions, though they arguably fell short of confirming hypotheses about psycho-
logical events in terms of behavioural criteria (to use Sellars characterization
of what makes someone a behaviourist). A recent counterexample to some
(though by no means all) strands of behaviourism is that of Galen Strawsons
Weather Watchers, viz. beings who are hypothesized to have a mental life
despite being constitutionally incapable of any sort of behaviour, as this is
ordinarily understood.
C. S.
Smith, L. (1986), Behaviorism and Logical Positivism, Stanford: Stanford University
Press.
Stout, R. (2006), The Inner Life of a Rational Agent, Edinburgh: Edinburgh University
Press.
belief. While Russell and McTaggart both wrote about states of mind as early
as 1921, and Turing talks of the internal states of machines in his influential
1950 Mind article Computing Machinery and Intelligence, it was not until the
rise of functionalism in the 1960s (and in particular Hilary Putnams 1964
paper Minds and Machines) that philosophers began to refer to beliefs as
mental states, alongside desires and other mental phenomena. Yet belief does
not generally appear to be a state of mind in the ordinary (emotional) sense of
the term, besides which one can only be in so many states of mind at any given
moment. Some theorists hold that there is an episodic occurrent form of belief
in which thoughts are actively brought to mind. It is true that we could not
have many of these at any given time, but this is not because they are states we
might be in. One may be in a state of nervousness or anxiety but it ordinarily
makes no sense to say that a subject, mind, or brain is in a state of belief (though
one may find oneself in a state of disbelief).
Beliefs so construed are held to have representational contents which relate
to (possible or actual) sates of aairs, much as the line on a gramophone record
relates to the notes played by the recorded orchestra. Representationalist
thought dates at least as far back as Hume, if not Plato. Wi genstein a acked
this characterization in his Philosophical Investigations which, among other
things, sought to repudiate his own, earlier, picture theory of the mind. To believe
that p, he came to think, is not to form any kind of representation of p, but sim-
ply to take p to be the case, thus explaining the impossibility of meaningfully
stating p, but I dont believe p. Moores paradox was founded on the worry
that the statement in question could nonetheless be true (since p, but Moore
doesnt believe that p is unproblematic). But there is no real paradox here, for
on the case-stating view (but not on the state-reporting one) to say that one
283
Glossary
believes that p is not to explicitly say anything about oneself, though one may
be disclosing much through conversational implicature.
Frankish helpfully distinguishes between two strands of belief, associated
with two distinct kinds of mental processing and, more generally, two concep-
tions of mind. The first (basic belief) is typically non-conscious, passive, non-
occurrent and a ributable on purely behavioural grounds (see animal minds).
The second (superbelief) may be held consciously, typically requires linguistic
conceptualization, and is frequently occurrent. Frankish argues that basic
beliefs and superbeliefs may conflict. For example, I superbelieve that the
indicator in the new car is on the le , yet on each turn I move my hand to the
right, as in the old car, suggesting that my basic belief (which I may become
aware of) contradicts it. This oers an alternative approach to Moores paradox,
on which the asymmetry between first and third person disappears. If p and
I dont believe p refer to a itudes of dierent types, then they may both be
assertable, even if both (or neither) of them are self-descriptions.
The ontology of (both kinds of) belief also crosses paths with theories of
truth. Inspired by Freges distinction between believing and the thing believed,
White has suggested that the term belief is simply ambiguous: it can either
refer to a proposition (which may be true or false) or to the believing of such a
proposition. However we do not believe our beliefs any more than we desire or
desires or fear our fears, and it is at best awkward to talk of believing, desiring,
or fearing propositions. In contrast to what we believe, a belief may be imagina-
tive, and this need not coincide with ones believing being imaginative (for I may
unimaginatively just latch onto your imaginative belief). Following Gilbert
Ryle, it is arguably a category mistake to think that we have beliefs in the same
sense in which we can be said to have pencils, mortgages, or family in Tanzania.
To claim, instead, that we paradigmatically ascribe beliefs to beings when
they behave (or when it is assumed they would behave) as if something were
the case is not to deny we may legitimately talk of beliefs that no one has
ever had; on the contrary there are numerous things which one could take
to be the case even if nobody has ever actually done so. Indeed, two people
may have exactly the same belief, for the beliefs we have are not particular
tokens of universal types, though ones having a belief is an instance of a
general case.
C. S.
Collins, A. (1987), The Nature of Mental Things, Notre Dame: Notre Dame Press.
Frankish, K. (2004), Mind and Supermind, Cambridge: Cambridge University
Press.
Steward, H. (1997), The Ontology of Mind, Oxford: Oxford University Press.
White, A. (1972),What We Believe, in N. Rescher (ed.), Studies in the Philosophy
of Mind, APQ monograph series no. 6, Oxford: Blackwell.
284
Glossary
causal closure. Causal closure is the claim that all physical events have a phys-
ical cause. If one accepts causal closure one denies the causal ecacy of the
non-physical. Ghosts cannot drag iron chains, and God cannot intervene in
the physical world. Non-physical minds cannot interact with physical bodies
and thus causal closure is thus at odds with mind-body dualism.
D. O.
Papineau, D. (2002), Thinking about Consciousness, Oxford: Oxford University
Press.
M. C.
Preston, J., and M. Bishop (2002), Views into the Chinese Room: New Essays on Searle
and Artificial Intelligence, Oxford: Oxford University Press.
Searle, J. (1980), Mind, Brains and Programs, Behavioural and Brain Sciences, 3,
41724.
285
Glossary
information when it computes. Such a view of the mind is central to what has
become known as classical cognitive science and is associated with Putnam,
Fodor and others. Interpreted more broadly, neural networks of the kind postu-
lated by connectionists are computers. So, to some, connectionism is a form of
computationalism.
M. C.
Fodor, J. (2008), LOT 2: The Language of Thought Revisited, Oxford: Oxford University
Press.
Haugeland, J. (ed.) (1997), Mind Design II, Cambridge, MA: MIT Press.
D. O.
Margolis, E. (ed.) (1999), Concepts: Core Reading, Cambridge: MIT Press.
M. C.
Bechtel, W., and A. Abrahamsen (2002), Connectionism and the Mind, second edition,
Oxford: Blackwell.
286
Glossary
Haugeland, J. (ed.) (1997), Mind Design II, Cambridge, MA: MIT Press.
D. P.
Armstrong, D. M. (1981), What is Consciousness?, in D. Rosenthal (ed.), The Nature
of Mind, Ithaca, NY: Cornell University Press, pp. 5567.
Block, N. (1995), On a Confusion about a Function of Consciousness, Behavioral
and Brain Sciences, 18, 22747.
Rosenthal, D. (2002), How Many Kinds of Consciousness?, Consciousness and
Cognition, 11 (4), 65365.
287
Glossary
D. H.
Evans, G. (1982), The Varieties of Reference, Oxford: Oxford University Press.
I. M.
Hume, D. (1978), Treatise of Human Nature, ed. L. A. Selby-Bigge, Oxford: Clarendon
Press.
Sobel, D., and D. Copp (2001), Against Direction of Fit Accounts of Belief and
Desire, Analysis 61 (1), 4453.
C. S.
Mumford, S. (1998), Dispositions, Oxford: Oxford University Press.
dualism. Any view which posits the existence of just two kinds of something,
as opposed to just one kind of something (monism) is dualistic in nature. In the
288
Glossary
R. D.
Chalmers, D. (1996), The Conscious Mind, Oxford: Oxford University Press.
Descartes, R. (1642/1984), Meditations on First Philosophy, in J. Co ingham,
R. Stootho and D. Murdoch (trans. and eds), The Philosophical Writings of
Descartes, vol. 2, Cambridge: Cambridge University Press.
D. O.
Churchland, P. (1979), Ma er and Consciousness, Cambridge: MIT Press.
emotions. Mental states such as fear, happiness, anger and (on some views)
desire, commonly associated with distinctive pa erns of sensation, thought
and behaviour. What constitutes the underlying nature of such states is a ma er
of much controversy as is the issue of whether they form a unified ontological
category at all.
289
Glossary
The etymology of the term emotion points to the most obvious theory of
emotions as visceral movements. Theories which emphasize this visceral or
sensation aspect of emotions have been widely criticized however for leaving
out the whole intentional dimension of emotions and for being unable to
account for either the rationality of emotions or for their complex role in ratio-
nalizing other states and actions. Such concerns underlie a number of more
recent theories of emotions cognitivist and perceptual theories in particular
according to which emotions are just special kinds of judgements or experi-
ences of the world as having emotion-specific evaluative properties such as
of being frightening, dangerous, worth avoiding or worth obtaining. These
theories retain however, in common with sensation theories, the view that
emotions are essentially conscious episodes, thus leaving them open to the
charge that they cannot make room for the existence of unconscious or non-
conscious emotions (i.e. ones not currently occupying ones a ention). As against
such theories, it is thus sometimes argued that emotions are best thought of as
mental dispositions dispositions to behave in various ways, to have ones
experiences of the world transformed or coloured in distinctive ways, and to
have a wide range of other mental states, both episodic and dispositional,
including thoughts, beliefs, desires and even other emotions.
How each of these approaches might be refined to accommodate the varied
aspects and wide ranging roles of emotions in our mental lives is the focus of
much current research. Questions being actively addressed include ones about
the rationality of emotions, about their role in motivating action, about the
nature of emotional experience and emotional expression as well as of course
fundamental questions about the ontological nature of emotions and about
our ability to know them.
I. M.
James, W. (1884), What is an Emotion? Mind, 9, 188205.
Sartre, J. (1962), Sketch for a Theory of the Emotions, London: Methuen.
Wollheim, R. (1999), On the Emotions, New Haven: Yale University Press.
290
Glossary
D. O.
Jackson, F. (1982), Epiphenomenal Qualia, The Philosophical Quarterly, 32, 12736.
D. P.
Chalmers, D., and F. Jackson (2001), Conceptual Analysis and Reductive Explana-
tion, Philosophical Review, 110, 31561.
Levine, J. (1983), Materialism and Qualia: The Explanatory Gap, Pacific Philosophical
Quarterly, 64, 35461.
McGinn, C. (1989), Can We Solve the Mind-Body Problem? Mind, 98, 34966.
extended mind. The extended mind thesis holds that at least some cognitive
processes essential for enabling the completion of specific acts of cognition are
not wholly within the boundaries of the skin or the skull. Focusing on cases of
belief formation, Clark and Chalmers argue that sometimes successful cogni-
tion unavoidably depends on the use of environmental resources (e.g. appeal to
291
Glossary
D. H.
Clark, A. (2009), Supersizing the Mind: Embodiment, Action and Cognitive Extension,
Oxford: Oxford University Press.
Clark, A., and D. Chalmers (1998), The Extended Mind, Analysis, 58, 719.
292
Glossary
I. M.
Moran, R. (2001), Authority and Estrangement: An Essay on Self-Knowledge, Princeton,
NJ: Princeton University Press.
D. O.
Alston, P. (1971), Varieties of Privileged Access, American Philosophical Quarterly,
8, 22341.
Nagel, T. (1979), What is It Like to Be a Bat? in Mortal Questions, Cambridge:
Cambridge University Press.
293
Glossary
actions (i.e. our own and those of others) in terms of reasons. Engaging success-
fully in this practice requires being able to answer a particular sort of why
question by competently deploying the idiom of mental predicates (beliefs,
desires, hopes, fears, etc.) and a ributing these mental state terms appropri-
ately. Sometimes folk psychology is used to refer, more restrictively, to the
complete set of propositions and generalizations (or at least a perspicuous
presentation of the core body of these) that its practitioners are implicitly
commi ed to when using mental state terms in order to make sense of actions.
Used in this way, folk psychology is o en imagined to denote the theory that
would be obtained by systematically describing all of the relevant folk com-
mitments in a systematic way. There are several accounts of the nature of folk
psychology and how best to explain the relevant abilities associated with it
for example, it has been variously characterized as being, in essence: a kind
of theory; a practice involving modelling or simulation of mental states; and a
narrative practice. Among some of those who regard it as a theory its elimi-
nation has been called for by stressing its folk status and highlighting its lack
of fit with growing modern science.
D. H.
Carruthers, P., and P. Smith (eds) (1996), Theories of Theories of Mind, Cambridge:
Cambridge University Press.
Goldman, A. (2006), Simulating Minds: The Philosophy, Psychology and Neuroscience
of Mindreading, Oxford: Oxford University Press.
Hu o, D. D. (2008), Folk Psychological Narratives: The Socio-Cultural Basis of
Understanding Reasons, Cambridge, MA: MIT Press.
294
Glossary
M.C.
Denne , D. (1998), Brainchildren: Essays on Designing Minds, Cambridge, MA: MIT Press.
Fodor, J. A. (2000), The Mind Doesnt Work That Way, Cambridge, MA: MIT Press.
functionalism. Tables are not defined in terms of their physical structure since
they can be made out of all kinds of stu, including metal, wood, and plastic.
They are defined according to their function: a table is (roughly) something
which we use to put things on. Functionalists have a related view of the mind.
I can feel pain and so perhaps can a tuna fish, yet our brains are structured
dierently. It is also conceivable that creatures on other planets can feel pain,
and perhaps future robots, but such beings will have very dierent brains to us
and to tuna. Mental states are not therefore defined in terms of their physical
structure; they are, rather, defined by their causal relations. Pains are the kind
of state that are caused by bodily damage and that lead to avoidance behaviour
and depression. If an alien creature is in the kind of state that bears those
relations with its behaviour and other mental states, then it is in pain. As Hilary
Putnam (1975, p. 291) claims: we could all be made of Swiss cheese and it
wouldnt ma er.
A key problem for functionalism lies in accounting for the subjective feel
of mental states, or what is called their phenomenology or qualitative nature.
Pain may have the causal relations that functionalists say it does, but pain
also feels a certain way it hurts and it is not clear how functionalists can
account for this fact.
D. O.
Putnam, H. (1975), Philosophy and Our Mental Life, in Philosophical Papers, vol. 2,
Cambridge: Cambridge University Press.
295
Glossary
D. O.
Putnam, H. (1975), The Meaning of Meaning, in Philosophical Papers, vol. 2,
Cambridge: Cambridge University Press.
I. M.
Rosenthal, D. (1991), Two Concepts of Consciousness, in D. Rosenthal (ed.), The
Nature of Mind, Oxford: Oxford University Press.
296
Glossary
exist in but transcend the acts of perception, wanting, hoping, etc. Thus, at least
some intended objects are intended as existing in the external world outside of
thought if they do not simply belong to the thinkers imagination.
D. H.
Fodor, J. A. (1994), The Elm and the Expert, Cambridge, MA: MIT Press.
Putnam, H. (1975), The Meaning of Meaning, in Mind, Language and Reality:
Philosophical Papers, vol. 2, Cambridge: Cambridge University Press.
(1988), Representation and Reality, Cambridge, MA: MIT Press.
introspection. The special way each person has of knowing the contents of his
or her own mind, although what this way is, whether it is in fact special and
whether it is a way of knowing at all, are all ma ers of some controversy. The
etymology of the term introspection suggests that introspection is essentially
a form of inner perception and that to introspect is in some sense to look
inside. But, one might ask, in what sense? The verbs to perceive and to see
are pre-theoretically used in a variety of ways, in particular to mean quite gen-
erally to know or to understand as when speaking of seeing what someone
means. Similarly, to look is o en used to mean to think about or to investi-
gate as when promising to look into some ma er. In its modern theoretical
sense however perception refers fundamentally to external sense perception.
Speaking of introspection in the theoretical context of philosophy can thus be
seen to establish an analogy between so called inner perception and external
297
Glossary
sense perception, or between our way of knowing our own mind on the one
hand and our way of knowing that which lies outside it, through sense percep-
tion, on the other. Few current philosophers however believe this analogy to be
legitimate.
Introspection is argued instead to be either a (particularly quick and well
informed) process of inference from observation of our behaviour (and so not
a special way of knowing, distinct from our way of knowing the mental states
of others), or, some essentially non-epistemic process of avowal of our current
mental states (such as that of mere expression of our beliefs and desires and
so not a way of knowing at all). Beyond standard epistemic and non-epistemic
accounts of introspection, the recent literature has seen a number of further,
o en more subtle, views of introspection being espoused on both sides. The
central task for any such theory remains that of providing an account of at least
the appearance of our having a special way of knowing our own minds which
is unlike our way of knowing the minds of others and which displays certain
key characteristics (e.g. of immediacy, first-person authority and immunity to
certain types of error).
I. M.
Armstrong, D. M. (1968), A Materialist Theory of Mind, London: Routledge and
Kegan Paul.
Heal, J. (1994), Moores paradox: a Wi gensteinian approach, Mind (January).
Shoemaker, S. (1996), Self-Knowledge and Inner Sense, in S. Shoemaker, The
First Person Perspective and Other Essays, Cambridge: Cambridge University Press.
M. C.
Cain, M. J. (2002), Fodor: Language, Mind and Philosophy, Cambridge: Polity.
298
Glossary
D. H.
Davidson, D. (1980), Actions, Reasons and Causes, in Essays on Actions and Events,
Oxford: Clarendon Press.
Heil, J., and A. Mele (eds) (1993), Mental Causation, Oxford: Oxford University Press.
Kim, J. (2008), Mind in a Physical World: An Essay on the Mind-Body Problem and
Mental Causation, Cambridge, MA: MIT Press.
mind-body problem. The mind-body problem is the central and most dicult
problem in the philosophy of mind. It is the challenge of explaining the meta-
physics of the relation between the physical body, especially the brain and
nervous system, and the events of consciousness experienced by a thinking
psychological subject.
299
Glossary
Eorts to solve the mind-body problem can be divided into two main cate-
gories, those that (1) try to reduce or eliminate all meaningful reference to the
mind to or in favor of purely physical material entities, properties and events,
and (2) those who argue that no such reductions or eliminations could possibly
be adequate to the relevant data we find not only in reflecting on the content of
our own subjective psychological lives and those we a ribute to others, that
we learn about from their expressions of thought, but also in the objective
external behavior of other psychological subjects. The main theories in category
(1) include eliminative or reductive (a) behaviorism, (b) materialism, and (c)
functionalism, the la er subsuming (d) computationalism as a special type
that is otherwise simply identified with functionalism. These theories maintain
either that there are no such things as thoughts, states of consciousness, or the
mind, or that whatever can truly be said of mental entities and events can be
interpreted in a vocabulary consisting entirely of terms for purely physical
material substances, entities, properties and events. The main theories in
category (2) include: (a) what is alternatively called, a er Descartes, Cartesian,
substance, or ontic dualism, and (b) property dualism. Property dualism in
category (2b) is in turn sub-divisible into (i) intentionalist- and (ii) qualia-based
philosophies of mind that emphasize either intentionality or the existence and
nature of qualia as explanatorily ineliminable and physically or materially
irreducible. There is also no reason why a (2b) property dualism could not
accept both (i) and (ii) in opposition to mind-body eliminativism or reductivism
in trying to solve or at least clarify the mind-body problem, and then alterna-
tively ordering (i) and (ii) in terms of explanatory or other priority, with either
(i) taking precedence over (ii) or the reverse, or of treating (i) and (ii) as distinct
but explanatorily equally significant and important grounds for denying the
truth of any mind-body solution in category (1). Since it is absurd to suppose as
eliminativism does that, despite appearances, thoughts and the mind do not
exist, appearances themselves being states of mind, and since Cartesian, sub-
stance or ontic dualism is widely believed to be indefensible, it is possible
to speak in practical terms of the mind-body problem as a contest between
some form of physical reductivism, behavioral, material, or functional (compu-
tational) on the one hand, and, on the other, some form of property dualism
that emphasizes intentionality over qualia or qualia over intentionality, or
gives equal importance to both intentionality and qualia in understanding the
nature of mind.
D. J.
Descartes, R. (1641), The Meditations on First Philosophy, in which the Existence of God
and the Distinction of Mind and Body are Demonstrated, trans. and ed. J. Co ingham,
Cambridge: Cambridge University Press.
Jacque e, D. (2009), Philosophy of Mind: The Metaphysics of Consciousness, London:
Continuum Books.
300
Glossary
McGinn, C. (2000), The Mysterious Flame: Conscious Minds in a Material World, New
York: Basic Books.
moral psychology. The branch of ethics that deals with human and animal
psychology, particularly as it relates to moral judgement and motivation. Its
primary aim is to investigate the relation between vice and virtue and our
(general and individual) cognitive and conation abilities including those of
belief, desire, impulse, intention and volition. Various views within moral psy-
chology debate the extent to which ethical norms and reasons are relative to
agential character traits or dispositions (e.g. (a) whether or not moral judge-
ments necessarily motivate and (b) whether one can have a normative reason
for doing something even if there was nothing in their motivational set that
could ever (either directly or indirectly) move them to do it).
C. S.
Blackburn, S. (1998), Ruling Passions, Oxford: Oxford University Press.
Co ingham, J. (1998), Philosophy and the Good Life, Cambridge: Cambridge University
Press.
Smith, M. (1994), The Moral Problem, Oxford: Blackwell.
D. O.
Kim, J. (1992), Multiple Realizability and the Metaphysics of Reduction, Philosophy
and Phenomenological Research, 52, 126.
301
Glossary
The connections have weights so that they can amplify or dampen the
strength of an impulse they carry and can be either excitatory or inhibitory.
When units in the input layer are stimulated by the outside world, impulses
pass along connections to the intermediate layer so stimulating activity there.
Impulses are then passed to the output layer resulting in pa erns of activity
at that level. Consequently, the system transforms pa erns of activation at
the input layer into pa erns of activation at the output layer. The systems
input-output behaviour is determined by the nature and weight of the con-
nections and the threshold values of the units. Adjusting the connection
weights will alter the systems input-output behaviour. As the pa erns of
activation can have semantic significance, the network can serve as an infor-
mation processor.
M. C.
Bechtel, W., and A. Abrahamsen (2002) Connectionism and the Mind, second edition,
Oxford: Blackwell.
Haugeland, J. (ed.) (1997), Mind Design II, Cambridge, MA: MIT Press.
other minds. I know that I have a mind, but how can I be sure that others do?
I see people writing shopping lists, running for the bus and talking to each
other, but it is possible that all these people are just mindless automata, their
actions akin to the behaviour of non-sentient robots. I might be the only mind
in existence! This is called the problem of other minds.
Various solutions to this problem have been oered. I could come to know
that others have a mind by analogy. I know that my behaviour is caused by my
mental states, and since others behave in a similar way to me, I can infer that
their behaviour is caused by their mental states. In contrast behaviourists claim
that behaviour is not merely the surface eect of underlying mental causes.
As Ryle puts it in The Concept of Mind, Overt intelligent performances are not
clues to the workings of minds; they are those workings. Boswell described
Johnsons mind when he described how he wrote, talked, ate, fidgeted and
fumed. For the behaviourist, then, it is my perceptual experience of the beha-
viour of others that justifies my belief in other minds. Last, and most popular,
is a theoretical account of the mind. I am justified in believing that others have
minds in the same way that I am justified in believing that stars are giant nuclear
reactions. The physics of nuclear reactions can be used to predict and explain
the behaviour of stars, and folk psychological categories can be used to predict
and explain the actions of people. The reasoning applied here is inference to
the best explanation. If there is a theory that explains the occurrence of certain
phenomena be er than any alternative theory, then we are justified in believing
that theory. I am therefore justified in believing in folk psychology and the
existence of other minds.
302
Glossary
D.O.
Avramides, A. (2001), Other Minds, London: Routledge.
D. P.
Maund, B. (2003), Perception, Chesham: Acumen.
Robinson, H. (1994), Perception, New York: Routledge.
perceptual content. Perceptual content refers to the things and their properties
that feature in ones perceptual experience, or in other words, it is what is con-
veyed to one by ones perceptual experience via the five sense-modalities and
proprioception. According to some philosophers, perceptual content is repre-
sentational since it seems that our perceptual experiences are by their nature
such that they present the world as being a certain way. Perceptual experiences
seem to have accuracy conditions; they are accurate in certain circumstances
and inaccurate in others and therefore are assessable for accuracy.
If one thinks that experiences have representational content then one thinks
of them as belief-like in some respects: Believing for instance, that there is a cup
on the table is being in a state with representational content. But if one claims
that an experience has representational content, that does not commit one to
identifying experiences with beliefs, since (1) experience may not be the same
a itude as belief, and (2) if (1) is false, they may be both a itudes to a dierent
kind of content. Regarding the la er, some philosophers believe that perceptual
states can represent the world without the subject of those states possessing the
concepts required to specify their content. On this view, ones experience of the
world is not constrained by ones conceptual capacities.
We can further ask, in virtue of what do perceptual experiences have the
content they have and represent the state of aairs they represent? Externalism
about perceptual content (also called phenomenal externalism) holds that the
contents of experience are not determined by the internal states of the brain but
303
Glossary
D. P.
Crane, T. (2003), The Intentional Structure of Consciousness, in Q. Smith and
A. Jokic (eds), Consciousness: New Philosophical Perspectives, Oxford: Oxford
University Press.
Gunther, Y. (2003), Essays on Non-Conceptual Content, Cambridge, MA: MIT Press.
Lycan, W. (2001), The Case for Phenomenal Externalism, Philosophical Perspectives,
15, 1735.
D. P.
Chalmers, D. (2007), Phenomenal Concepts and the Explanatory Gap, in T. Alter
and S. Walter (eds), Phenomenal Concepts and Phenomenal Knowledge: New Essays on
Consciousness and Physicalism, Oxford: Oxford University Press, pp. 16795.
Stoljar, D. (2005), Physicalism and Phenomenal Concepts, Mind and Language, 20,
46994.
304
Glossary
D. P.
Brentano, F. (1995/1874), Psychology from an Empirical Standpoint, ed. L. McAlister,
trans. A. Rancurello, D. Terrell and L. McAlister, London and New York:
Routledge.
Carruthers, P. (2006), Conscious Experience versus Conscious Thought, in
U. Kriegel and K. Williford (eds), Consciousness and Self-Reference, Cambridge,
MA: MIT Press.
Smith, D., and A. Thomasson (eds) (2005), Phenomenology and Philosophy of Mind,
Oxford and New York: Oxford University Press.
Strawson, G. (1994), Mental Reality, Cambridge, MA: MIT Press.
305
Glossary
and physical properties, and the former are distinct from and irreducible to
the la er.
D. P.
Chalmers, D. (2002), Consciousness and Its Place in Nature, in id. (ed.), Philosophy
of Mind: Classical and Contemporary Readings, New York: Oxford University
Press, pp. 24772.
Kim, J. (2005), Physicalism, or Something Near Enough, Princeton: Princeton University
Press.
privileged access. The special epistemic position each person stands in with
respect to the contents of his or her own mind. It is o en noted in the literature
on self-knowledge that we seem to be be er placed than anyone else to say
what mental states we are in. Whether we are be er placed by virtue of having
a special way of accessing our own mental states (e.g. some form of inner sense)
or merely by virtue of tending to have more evidence available (due to our
greater proximity with ourselves) is a ma er of some controversy. Equally con-
troversial is the wider issue of whether the authoritative status of our judge-
ments about our own mind is truly the result of a cognitive achievement on
our part (and so the result of a form of access), or merely the consequence of
some special feature of our self-ascriptive judgements of the form I believe
that p or I am happy or indeed I am in pain.
I. M.
Alston, W. (1971), Varieties of Privileged Access, American Philosophical Quarterly, 8.
D. H.
Russell, B. (1918), The Philosophy of Logical Atomism, The Monist; reprinted in R. C.
Marsh (ed.), Logic and Knowledge: Essays 19011950, London: Unwin Hyman, 1956.
306
Glossary
C. S.
Freud, S. (1964), New Introductory Lectures on Psychoanalysis, trans. J. Strachey,
London: Hogarth Press.
psychology. The study of how human and animal minds function, with the
primary aim of providing explanatory theories of how our knowledge of the
mind helps to explain or manipulate behaviour. Much psychological theory
thus o en falls between philosophy and science (in particular psychiatry);
indeed the godfather of modern psychology is o en said to be Nietzsche who
put forth his concept of a will to power as an explanation of the behavioural
drives of all living things. Not unlike the philosophy of mind, psychology
divides itself into numerous specialities such as cognitive, motivational, social,
neural, educational, perceptual, or cultural psychology. Such theories may then
be variously applied in fields as diverse as advertising, military strategy and
emotional counselling.
C. S.
Smith, E. et al. (eds) (2002), Atkinson and Hilgards Introduction to Psychology,
fourteenth edition, Belmont, CA: Wadsworth Publishing.
qualia. The word qualia is used in stronger and weaker ways in the philo-
sophical literature. Sometimes it is used to designate the distinctive quality,
307
Glossary
D. H.
Chalmers, D. (1996), The Conscious Mind, Oxford: Oxford University Press.
Denne , D. C. (1991), Consciousness Explained, New York: Penguin Books.
Flanagan, O. (1993), Consciousness Reconsidered, Cambridge, MA: MIT Press.
reasons. Agents are said to act for or in the light of reasons when they act with
some aim or purpose in mind. Many philosophers hold that all intentional
action is performed for a reason while almost all hold the converse view that all
actions performed for reasons are intentional. There is much ontological debate
over whether reasons should be conceived of as facts, states of aairs, proposi-
tions, mental states, or some disjunctive combination. Reasons why we do things
need not be reasons for which we do them, nor are the la er always normative
reasons for acting (though it had be er be possible to act for a good reason).
C. S.
Dancy, J. (2000), Practical Reality, Oxford: Oxford University Press.
Parfit, D. (1984), Reasons and Persons, Oxford: Oxford University Press.
Sandis, C. (ed.) (2009), New Essays on the Explanation of Action, London: Palgrave
Macmillan.
308
Glossary
comparing explanations for relative diculty and counting the terms needed
in the vocabularies of target theories and their putative reductions. Where
philosophy of mind is concerned, mind-body reduction in particular is the
reduction of the mind or mental properties to exclusively non-mental, purely
physical behavioural, material, or functional (computational) properties.
Whether such a reduction can possibly succeed in the case of all truths about
consciousness, mental or psychological phenomena, the mind and its thoughts,
is one way of formulating the mind-body problem.
D. J.
Horst, S. (2007), Beyond Reduction: Philosophy of Mind and Post-Reductionist Philosophy
of Science, Oxford: Oxford University Press.
Searle, J. (1992), The Rediscovery of the Mind, Cambridge, MA: MIT Press.
Stich, S. (1983), From Folk Psychology to Cognitive Science: The Case Against Belief,
Cambridge, MA: MIT Press.
309
Glossary
correct just in case its content is correct, and a proposition which gives that
content is correct just in case it is true.
D. P.
Crane, T. (1992), The Non-Conceptual Content of Experience, in id. (ed.) The
Contents of Experience, Cambridge: Cambridge University Press, pp. 13657.
Jackson, F. (1977), Perception: A Representative Theory, Cambridge: Cambridge
University Press.
Macpherson, F., and D. Platchias (forthcoming), Representationalism, Cambridge,
MA: MIT Press.
Tye, M. (1995), Ten Problems of Consciousness, Cambridge, MA: MIT Press.
self. The I who experiences, thinks, believes and desires. This conception of
the self as the subject of psychological states has long been argued to be respon-
sible for creating the Cartesian fiction of the self as a purely mental entity
which persists unchanged through time and in which psychological states
inhere. Hume famously rejected this view, arguing that we have no impression
of such a self but only of a sequence of constantly changing perceptions
(i.e. mental states) and hence no idea of the self except as a bundle of such
perceptions. Most current philosophers also reject, though o en for indepen-
dent reasons, the Cartesian view of the self as a mental entity, yet o en retain
Descartess assumption that the I of psychological self-a ribution refers to an
entity, namely to the human being which thinks, believes, desires and which
is located in space at the point of origin of our spatial experiences or at the
point of reference of our spatial thoughts of the form I am here or that is
over there.
I. M.
Descartes, R. (1912 [1637]), A Discourse on Method, Meditations and Principles, Toronto:
Dent.
Hume, D. (1978 [1740]), Treatise of Human Nature, ed. L. A. Selby-Bigge, Oxford:
Clarendon Press.
Valberg, J. (2007), Dream, Death and the Self, Princeton: Princeton University Press.
310
Glossary
I. M.
Bermudez, J. (1998), The Paradox of Self-Consciousness, Cambridge, MA: MIT Press.
Kant, I. (1929), Critique of Pure Reason, London: Macmillan Press.
Sartre, J. (1969), Being and Nothingness, London: Routledge.
I. M.
Boghossian, P. (1998), Content and Self-Knowledge, Philosophical Topics, 17, 526.
Moran, R. (2001), Authority and Estrangement: An Essay on Self-Knowledge Princeton,
NJ: Princeton University Press.
311
Glossary
M. C.
Fodor, J. (1990), A Theory of Content and Essays, Cambridge, MA: MIT Press.
Putnam, H. (1975), The Meaning of Meaning, in id., Mind, Language and Reality:
Philosophical Papers, vol. 2, Cambridge: Cambridge University Press.
312
Glossary
D. J.
Kim, J. (1993), Supervenience and Mind: Selected Philosophical Essays, Cambridge:
Cambridge University Press.
Rowlands, M. (1995), Supervenience and Materialism, Avebury: Ashgate Publishing.
Tooley, M. (1999), Laws of Nature, Causation, and Supervenience, London: Routledge.
313
Glossary
D. H.
Dretske, F. (1988), Explaining Behaviour: Reasons in a World of Causes, Cambridge,
MA: MIT Press.
Millikan, R. (1984), Language, Thought and Other Biological Categories, Cambridge,
MA: MIT Press.
I. M.
Evans, G. (1982), The Varieties of Reference, Oxford: Clarendon Press.
Martin, M. (2002), The Transparency of Experience, Mind and Language, 17, 376425.
314
Glossary
states. How the machine responds to what it scans will depend upon its state.
Its response will have several elements involving: (1) either leaving the square
unchanged or writing a new symbol on it (i.e. a 1, 0 or a blank); (2) moving
one square to the le or one square to the right or halting; and (3) moving into
some other state or remaining in the same state. The machines response
to any possible symbol for each of the states that it can be in is specified by a
machine table.
Turing proved that for any computable mathematical function there is a
Turing machine that can compute it. A universal Turing machine can be pro-
grammed (by means of strings of symbols printed on its tape) to mimic any
possible Turing machine and so is capable of computing any computable math-
ematical function. Putnam developed an early version of functionalism that
compares mental states with Turing Machine states as the la er are defined by
the machine table in terms of their relations to inputs, outputs and other states
rather than their material constitution.
M. C.
Hodges, A. (1983), Turing: The Enigma, New York: Simon and Schuster.
Putnam, H. (1975), Mind, Language and Reality: Philosophical Papers, vol. 2, Cambridge:
Cambridge University Press.
Turing test. A test proposed by Turing as a precise alternative for the meaning-
less question of whether a machine is capable of thought. The test is based upon
the imitation game where the tester presents both a machine and a human
with a series of questions via a teletypewriter. On the basis of the answers
received, the tester tries to work out which respondent is the human and which
is the machine. The machine passes the test if the tester fails to identity it as the
machine.
M. C.
Millikan, P., and A. Clark (1996), Machines and Thought: The Legacy of Alan Turing,
vol. 1, Oxford: Clarendon Press.
Turing, A. (1950), Computing Machinery and Intelligence, Mind, 59, 43360.
315
Glossary
the sky and flow down rivers into the sea. They are, though, dierent liquids
since they are made of dierent stu. Philosophers of mind discussing this
thought experiment o en refer to the twin on Earth as Oscar, and the one on
Twin Earth as Toscar, and the liquids on their planets as water and twater
respectively.
Putnam took his thought experiment to show that the meanings of natural
kind terms such as water could not be wholly determined by items that are
literally inside the head of a thinker, items such as mental images, or brain or
computational states. Everything inside the heads of Oscar and Toscar is the
same when they are talking about what they both call water, yet Oscars word
refers to water and Toscars to twater. Their words therefore have dierent
meanings even though everything in their heads, and their behaviour, is the
same. As Putnam puts it, Meaning just aint in the head.
Others, including John McDowell and Greg McCulloch, take Putnams
thought experiment to entail a stronger conclusion, one concerning not just
the meanings of words, but the contents of thoughts. Oscar has thoughts with
the content water; Toscar has thoughts with the content twater. Even though
everything in their heads is the same, their thoughts are dierent. The content
of thought is partly determined by our relation to the world: the mind aint in
the head (McCulloch, The Life of the Mind, p. 41). This is cognitive externalism.
Putnam himself has now adopted this position.
D. O.
Pessin, A., and S. Goldberg (1996), The Twin Earth Chronicles: Twenty Years of
Reflection on Hilary Putnams The Meaning of Meaning, New York: Armonk.
what its like. Nagel famously wrote that the fact that an organism has con-
scious experience at all means, basically, that there is something it is like to be
that organism . . . fundamentally an organism has conscious states if and only
if there is something it is like to be that organism something it is like for the
organism the facts of experience [are] facts about what it is like for the experi-
encing organism. (Nagel, 1974, pp. 435, 439, emphasis in the original). This
quote points directly at what-it-is-likeness, the salient but dicult to describe
feature of a conscious state. If theres something it is like for one to be in a men-
tal state then the state is experiential. If there is nothing its like for one to be
in that state, its not. However, to say that there is something its like for one to
be in an experiential state is not merely to mean that there is something that
an experience is like. That there is something that an experience is like is a
mere truism in that it is plain that there is nothing such that it is not like
something. We can say for instance, that there is something that a rock or a
table is like. What-it-is-likeness in the Nagelian sense concerns the individual.
If there is something it is like for the individual to be in a particular mental state
316
Glossary
D. P.
Jackson, F. (1982), Epiphenomenal qualia, Philosophical Quarterly, 32, 12736.
Nagel, T. (1974), What is It Like to be a Bat?, Philosophical Review, 83 (4), 43550.
Tye, M. (1995), What What Its Like is Really Like, in Ten Problems of Consciousness,
Cambridge, MA: MIT Press, 13355.
will. The faculty of the will grounds our abilities to do things intentionally,
voluntarily, and at will, where this typically (but perhaps not necessarily)
involves acting for reasons. Contrary to an influential British tradition dating at
least as far back as Hobbes, to act voluntarily is not to perform an act of volition
but rather to act (intentionally or otherwise) according to ones own will (or
desire) and not under coercion or duress. This need not involve performing an
act of will, the la er requiring a high degree of motivational strength, courage
and/or eort. A persons will is said to be weak if they are too easily prone
to change their mind under the influence of others. In philosophy, however,
weakness of will has become a technical term for the phenomenon of acting
against ones be er judgement, commonly also referred to in the literature as
akrasia or (even more misleadingly) incontinence.
C. S.
Hacker, P. M. S. (2000), Willing and the Nature of Voluntary Action, in id.,
Wi genstein Mind and Will Part I: Essays, Oxford: Blackwell.
O Shaughnessy, B. (1980), The Will A Dual Aspect Theory, vols. 1 and 2, Oxford:
Oxford University Press.
317
Glossary
(Chalmers, 1996; Jackson, 2001), from this epistemic gap one can infer an onto-
logical gap: if we cannot deduce Q from P then we cannot explain phenomenal
consciousness in terms of physical processes, and if we cannot explain it in
terms of physical processes then phenomenal consciousness is not a physical
process.
D. P.
Brueckner, A. (2001), Chalmers Conceivability Argument for Dualism, Analysis,
61, 18793.
Chalmers, D. (1996), Can Consciousness Be Reductively Explained? in The
Conscious Mind, Oxford: Oxford University Press.
Nagel, T. (1998), Conceiving the Impossible and the Mind-Body Problem,
Philosophy, 73, 33752.
Shoemaker, S. (1999), On David Chalmerss The Conscious Mind, Philosophy and
Phenomenological Research, 59 (2), 43944.
318
Chronology
This is necessarily idiosyncratic, but the hope is to provide a useful desk refer-
ence for philosophers of mind who require speedy access to a date or title as
well as a bit of context for it. There are a few lines of explanation for almost
every entry. I tried to encapsulate in a single sentence as much of what ma ers
as possible. Obviously this almost never works, but a narrative structure really
does help one get a grip on the major events in the history of our thinking about
the mind. The narrative ends in 1949 with Ryles Concept of Mind. My excuse for
stopping there is that its dicult to say what the impact or meaning of more
recent books and papers might be, because we are too close in time to know.
Forgive omissions; it is very hard to tell which books and events should be
included early on or during unfamiliar centuries, and it gets extremely dicult
as we approach the present.
800 BCE
Homeric poems taking shape between the late ninth and early eighth century; they
characterize the soul thinly, as something lost at death, something which then
howls o to Hades.
600 BCE
Thales (fl. 600) might view psyche as a mover, force, or impetus, something which
initiates the movement of moving things, from animals and people to magnets.
Anaximenes (c. 585c. 528) possibly believes that psyche holds a living thing
together and rules or controls it.
Pythagoras (fl. 530) accepts metempsychosis; possibly first to locate the soul in
the head.
500 BCE
Anaxagoras (c. 500c. 428) seems to argue for a materialist world actuated by a
cosmic intelligence, mind or nous.
Heraclitus (fl. 500) might believe that psyche is fire, somehow responsible for the
changes a ending waking, sleeping and death.
Parmenides (early to mid-fi h century) distinguishes between false appearances
and reality as revealed by reason; might flirt with idealism.
400 BCE
Empedocles (c. 495c. 435) probably formulates the first theory of perception; his
talk of the cosmic psychological principles, love and strife, suggest panpsychism
to some.
319
Chronology
Socrates (c. 469c. 399), the man not the mouthpiece; might conceive of soul as the
bearer of moral qualities.
Democritus (c. 460c. 370) elaborates the atomism of Leucippus, including
materialist conceptions of perception and the soul; might be first to tie soul to
intelligence.
Plato (c. 427c. 347) distinguishes soul from body, argues for immortality of the soul,
ties soul to reason, Phaedo; divides the soul into three parts: reason, spirit and
appetite, Republic).
Aristotle (c. 384c. 322) oers an extended, systematic discussion of psychological
phenomena, De Anima and Parva Naturalia, (c. 350); soul characterized as the
form of a living thing.
Epicurus (c. 341c. 271) argues for a radical materialism and for the impossibility
of the soul surviving death.
300 BCE
Zeno of Citium (c. 335c. 263) founds Stoic School, active until c. 520, which
perpetuates a variety of materialist notions of soul, typically conceived as a
breath-like substance diused throughout the body.
200 BCE
The Septuagint produced between the third and first centuries BCE; conceptions
of the soul and mental phenomena as depicted in the Hebrew Bible translated
into Greek.
100 BCE
Lucretius (c. 98c. 51) propounds and expands the philosophy of Epicurus, pro-
ducing the first philosophical treatment of mind in Latin, De Rerum Natura.
BCE/CE
Philo (c. 20 BCEc. 50 CE) blends Greek philosophy and Hebrew thought about
the soul.
100
The Church Fathers (end of the first century to as late as 749 CE) subordinate
philosophical accounts of mind to scriptural ones, raise religious questions, and
shape the intellectual agenda accordingly.
Tertullian (c. 160c. 225) advocates traducianism; argues that soul must be somehow
corporeal if it can be tormented in hell, On the Soul.
200
Origen (c. 185c. 254) holds that souls were created by God for contemplation
but, falling away in distraction, became enveloped in bodies, On First Principles.
Plotinus (204/5270/1) founds Neoplatonism, articulates a conception of soul as part
divine and part entwined with body, as well as an intricate theory of perception,
The Six Enneads.
320
Chronology
300
Augustine (354430) oers a detailed description of and reflection on introspected
mental life, Confessions; has thoughts on action theory, On Free Will; might argue
by analogy for other minds, anticipate the cogito, and influence the Cartesian
conception of mind, On the Trinity and City of God.
400
Boethius (c. 480c. 524) translates Aristotle and Plato into Latin; emphasizes rational
nature of the soul, Contra Eutychen.
900
Avicenna (c. 9801037) integrates Islamic philosophy and Greek thought about
mind and soul, formulates floating man thought experiment, On the Soul.
1100
Averroes (11261198) Latin translations of his commentaries on Aristotle bring
Greek views on mind, through Islamic lenses, back to the West; also develops
his own complex psychology and metaphysics of the soul, Long Commentary on De
Anima.
Vespasian Homilies (c. 1150) contain possibly the first use in English of a variation
on the word soul (sawle), meaning life or life force.
1200
William of Moerbeke (c. 12151286) undertakes a complete translation of Aristotle
into Latin (c. 1250).
Aquinas (c. 12241274) reinterprets Aristotle in the light of Christian teaching,
articulates full-blooded conceptions of mind, soul, intellect, memory, appetite,
self-knowledge, imagination, perception, etc., Summa Theologiae.
1300
William of Shoreham (fl. 1330) writes religious poems containing a forerunner of
mind (mende), which might be the first use in English tied to cognition.
1400
Marsilio Ficiono (14331499) is the first to translate all of Plato into Latin, and the
Platonic conceptions of soul and mind are rekindled, Theologia Platonica de
Immortalitate Animae, (1474).
Pomponazzi (14621525) might anticipate property dualism, On the Immortality
of the Soul.
1500
Shakespeare (15641616) writes Hamlet (c. 1600); some detect Cartesian pre-
suppositions in certain soliloquies; others hear Hamlets repressed desires.
321
Chronology
1600
Hobbes (15881679) defends a causal, empiricist, mechanistic and materialist
conception of mental phenomena, Leviathan (1651).
Descartes (15961650) articulates Cartesian dualism; disentangles new thoughts on
mind from Aristotelian, Platonic and Scholastic thinking; thereby ushers in modern
philosophical reflection on the mental, Meditations on First Philosophy (1641).
Geulincx (16241669) (and Graud de Cordemoy [162684]) follows Descartes;
argues for pre-established harmony before Leibniz, Opera Philosophica (c. 1668).
Spinoza (16321677) rejects Cartesian dualism in favour of dual-aspect monism:
there is one substance, God, Ethics (1677).
Locke (16321704) formulates modern conception of self; raises questions about
personal identity; claims that experience is the source of ideas; sets out limits to
understanding, Essay Concerning Human Understanding (1690).
Malebranche (16381715) largely follows Descartes, but argues for occasionalism,
Dialogues on Metaphysics and on Religion (1688).
1700
Leibniz (16461716) argues for pre-established harmony, Discourse on Metaphysics
(1686).
Berkeley (16851753) writes an account of perception, Essay Towards A New Theory
of Vision (1709) and argues that to be is to be perceived, Principles of Human
Knowledge (1710).
Hartley (17051747) founds associationist school of psychology, Observations on
Man, His Frame, His Duty, and His Expectations (1749).
Reid (17101796) brings common sense to an account of sensation, conception, and
perception; uses memory to inform a notion of self, An Inquiry into the Human
Mind on the Principles of Common Sense (1764), Essays on the Intellectual Powers
of Man (1785).
Hume (17111776) brings the experimental method to bear on mind, follows the
sceptical implications of empiricism through, propounds the bundle theory of self,
Treatise of Human Nature (1739), Enquiry Concerning Human Understanding (1748).
Adam Smith (17231790) considers the nature of sympathy, The Theory of Moral
Sentiments (1759).
Kant (17241804) argues that the structuring activity of the mind makes possible
a world of experience; gives an account of reason, perception, judgement, the
understanding, imagination, etc., a Copernican Revolution in the conception
of mind, Critique of Pure Reason (1781), Critique of Practical Reason (1788), Critique
of Judgement (1790).
Bentham (17481832) articulates modern psychological hedonism, Introduction to
the Principles of Morals and Legislation (1789).
1800
Hegel (17701831) gives an account of the evolution of consciousness as it plays
out in human history, The Phenomenology of Spirit (1807).
Schopenhauer (17881860) sees blind craving, will, at the depressing centre of
human action; our inner experience of it points to the hidden nature of all things,
The World as Will and Representation (1819).
322
Chronology
J. S. Mill (18061873) elaborates on the connection between right and wrong and
pleasure and pain; connects social and political reform to psychology, A System
of Logic (1843).
Kierkegaard (18131855) claims that subjectivity is truth, Concluding Unscientific
Postscript to Philosophical Fragments (1846).
T. H. Huxley (18251895) memorably couches a version of epiphenomenalism
in terms of whistles and steam engines, On the hypothesis that animals are
automata, and its history (1874).
Wundt (18321920) investigates the self-examination of experience, Principles
of Physiological Psychology (1873/4), establishes a laboratory of experimental
psychology in 1879.
Brentano (18381917) reintroduces the Scholastic conception of intentionality as the
mark of the mental, and his elevation of introspection paves the way for the
phenomenological movement, Psychology from an Empirical Standpoint (1874).
Peirce (18391914) raises objections to Cartesian methods and suggests panpsy-
chism, along with further thoughts on signs and representation, The Fixation of
Belief (1877), The Monist series (18911893).
James (18421910) largely sets the agenda for both the philosophy of mind and
psychology by advancing influential accounts of the brain, the mind-body
relation, the stream of consciousness, memory, sensation, imagination, will, and
emotions all peppered with compelling introspective reports, The Principles of
Psychology (1890).
Nietzsche (18441900) calls the subject a grammatical fiction, On the Genealogy
of Morals (1887).
Bradley (18461924) leads the turn towards idealism in the English-speaking
world, rejects empiricist psychology, The Principles of Logic (1883), Appearance and
Reality (1893).
Husserl (18591938) rejects psychologism and formulates the phenomenological
method, Logical Investigations (1900/1); the method of epoch and transcendental
phenomenology itself appear, Ideas (1913).
Bergson (18591941) oers an alternative to phenomenology, finds multiplicity
in consciousness, regards intuition as method, Time and Free Will (1889), Maer
and Memory (1896).
1900
Freud (18561939), father of psychoanalysis, formulates such concepts as
repression, psychosexual motivation, unconscious desire, as well as the id, ego
and super ego, The Interpretation of Dreams (1900), The Ego and the Id (1923),
Introductory Lectures on Psychoanalysis.
Dewey (18591952) brings pragmatism to bear on mind, rejects dualisms in favour
of naturalism and evolution; mind emerges socially; founds the functional
approach to psychology, The Reflex Arc Concept in Psychology (1896), Experi-
ence and Nature (1925).
Whitehead (18611947) rejects materialism for the view that nature is a structure of
evolving processes, Process and Reality (1929).
Russell (18721970) champions analytic method, moves from reflection on sense
data to neutral monism, rejects idealism and psychologism, Knowledge by
323
Chronology
324
Chronology
Anscombe (19192001) Intention (1957), The First Person (1975), Metaphysics and
the Philosophy of Mind (1981).
Strawson (19192006) Individuals (1959), Freedom and Resentment and Other Essays
(1974) Scepticism and Naturalism (1985).
Smart (1920) Sensations and Brain Processes (1959), Philosophy and Scientific Realism
(1963).
Place (19242000) Is Consciousness a Brian Process? (1956).
OShaughnessy (19252010) The Will (1980), Consciousness and the World (2000).
Armstrong (1926) Perception and the Physical World (1961), Bodily Sensations (1962),
A Materialist Theory of the Mind (1968), The Nature of Mind and Other Essays (1980),
Consciousness and Causality (1984), The Mind-Body Problem (1999).
Putnam (1926) The Nature of Mental States (1967), Mind, Language and Reality
(1975) The Meaning of Meaning (1975).
Chomsky (1928) Syntactic Structures (1957), Aspects of the Theory of Syntax
(1965).
Williams (19292003) Problems of the Self (1973).
Rorty (19312007) Philosophy and the Mirror of Nature, (1979).
Shoemaker (1931) Self-Knowledge and Self-Identity (1963), Identity, Cause and Mind:
Philosophical Essays (1984), The First-Person Perspective, and other Essays (1996).
Dretske (1932) Seeing and Knowing (1969), Knowledge and the Flow of Information
(1981), Naturalising the Mind (1995), Perception, Knowledge and Belief (2000).
Searle (1932) Minds, Brains and Programs (1980), Intentionality (1983), Minds, Brains
and Science (1984), The Rediscovery of the Mind (1992).
Fodor (1935) The Language of Thought (1975), Propositional Aitudes (1978), Representa-
tions (1979), The Modularity of Mind (1983), Psychosemantics (1987), A Theory of
Content (1990).
Nagel (1937) What is it Like to be a Bat? (1974), Mortal Questions (1979), View from
Nowhere (1986).
Honderich (1933) A Theory of Determinism (1988), On Consciousness (2004).
Millikan (1933) Language, Thought, and Other Biological Categories (1984), White Queen
Psychology and Other Essays for Alice (1993).
Kim (1934) Supervenience and Mind (1993).
Rosenthal (1939) Two Concepts of Consciousness (1986), Consciousness and Mind
(2005), Consciousness and Its Function (2008).
Kripke (1940) Naming and Necessity (1972), Wigenstein on Rules and Private Language
(1982).
Lewis (19412001) An Argument for Identity Theory (1966), Psychophysical
and Theoretical Identifications (1972), Mad Pain and Martian Pain (1980),
Philosophical Papers, Volume II (1986).
McDowell (1941) Mind and World (1994), Mind, Value and Reality (1998).
Jackson (1943) Perception (1977), Epiphenomenal Qualia (1982), What Mary
Didnt Know (1986).
Block (1942) Psychologism and Behaviorism (1981), On a Confusion about the
Function of Consciousness (1995), Consciousness, Function and Representation (2007).
Denne (1942) Brainstorms (1981), Content and Consciousness (1986), The Intentional
Stance (1989), Consciousness Explained (1992), Kinds of Minds (1996), Brainchildren
(1998), Sweet Dreams (2005), Neuroscience and Philosophy (2007).
325
Chronology
326
Research Resources
Journals
AI & Society
AI Communications
Annals of Mathematics and Artificial Intelligence
Artificial Intelligence
Artificial Intelligence and Law
Artificial Intelligence Review
Behavior and Philosophy
Behavioural and Brain Sciences
Brain and Cognition
Brain and Language
Brain and Mind
Cognition and Emotion
Cognitive Linguistics
Cognitive Psychology
Computational Intelligence
Consciousness and Cognition
Consciousness and Emotion
Cybernetics and Human Knowing
International Journal of Approximate Reasoning
Journal of Artificial Intelligence Research
Journal of Cognitive Systems Research
Journal of Consciousness Studies
Journal of Culture and Cognition
Journal of Experimental and Theoretical Artificial Intelligence
Journal of Intelligent Systems
327
Research Resources
Websites
Cogprints
hp://cogprints.org/view/subjects/phil-mind.html
Consciousness and the Brain: Annotated Bibliography
www.consciousness-brain.org/
Dictionary of the Philosophy of Mind
hp://philosophy.uwaterloo.ca/MindDict/
Episteme Links
www.epistemelinks.com/
A Field Guide to the Philosophy of Mind
hp://host.uniroma3.it/progei/kant/field/
Internet Encyclopedia of Philosophy
www.iep.utm.edu/
KLI Theory Lab
www.kli.ac.at/theorylab/index.html
Mind Papers
hp://consc.net/mindpapers
Stanford Encyclopedia of Philosophy
hp://plato.stanford.edu/
The Turing Archive for the History of Computing
www.alanturing.net/
328
Research Resources
329
Notes
Chapter 1
1 One problem with Armstrongs definition of substance is that objects like the Sun
cannot exist alone for they are essentially spatio-temporal objects and so depend for
their existence on the existence of space and time. I will not a empt to resolve this
issue here.
2 In order to avoid concerns raised by quantum indeterminacy, physical closure
is sometimes expressed by saying that every physical event has a physical cause
sucient to determine its objective probability. See Yablo (1992); for a helpful discussion
see Section 2.3 of Robb et al. (2008).
3 The example is not to be taken too seriously because the neurobiology of pain is
very complex, and cannot be captured by the slogan pain = c-fibre firing. I will use
c-fibre firing as a convenient label for the complex neurobiological process which
are involved in the human pain response.
4 We can distinguish between the property of having a property which occupies the
causal role characteristic of mental state M, and the property which occupies
the causal role characteristic of mental state M. Functionalists dier as to which of
these properties is to be identified with M.
5 Strictly speaking, Davidson only denied the possibility of strict laws linking pro-
positional a itudes with physical events. I introduce propositional a itudes in the
Mental Representation section.
6 For alternative eliminativist strategies, see Stich (1983).
7 Notice the similarity between this proposal and Denne s intentional stance (see
the Eliminativism, Instrumentalism and the Intentional Stance subsection).
8 Traditionally, the key external relations were held to be those of similarity or resem-
blance, but this idea is fraught with diculties (for a quick overview see Ravenscro ,
2005, pp. 1267).
9 Putnam originally used his example to develop a point about the reference of natural
kind terms in ordinary human languages. However, his example has been widely
used to discuss parallel considerations in the philosophy of mind.
10 What Chalmers calls the hard problem is the problem of accounting for phenomenal
consciousness, and what he calls the easy problem includes the problem of accounting
for access consciousness (see Chalmers, 2003a, p. 103).
11 Strictly speaking, qualia is the plural; the singular is quale. However, in line with
most contemporary usage, I will use qualia as both the plural and the singular.
Chapter 2
330
Notes
2 Several prominent critics question the phenomenal concepts strategy on the grounds
that it is questionable that concepts of the requisite first-personal sort might exist
(see Prinz 2007; Tye 2009). Nevertheless, it is arguable that ordinary, public concepts
of experience could do the same work that phenomenal concepts are meant to do
equally well (or badly).
3 For a catalogue of other problems and worries concerning higher order theories
and their relatives see Block (forthcoming).
Chapter 3
331
Notes
are non-intentional. Crane gives good reasons to think that this is not true, but we
cannot go into the details here.
11 See Adams and Dietrich (2004) for an account that emphasizes the dierences.
12 Of course there can be hallucinations or artificially caused experiences, but the quali-
tative character of the experience would derive from past representational episodes.
13 All of these are examples of natural signs. Natural signs (such as smoke being a sign
of fire; footprints being a sign of a passerby) have a kind of informational aboutness.
These seem to be the wrong kind of intentionality, at least in part because they need
not produce a phenomenology, cannot be falsely tokened, and have not risen to the
level of semantic meaning. Smoke naturally indicates fire, but smoke means smoke
(and its tokening need not indicate fire).
14 Famously, Descartes believed that non-human animals not only could not think
(were not intentional systems), but could not feel because they were not intentional
systems. Descartes believed that to feel pain, for instance, one must be able to think
the thought Im in pain. So, only intentional systems were able truly to have
phenomenological states. When Fodor says paramecia are only sensory systems, he
does not say whether they may have a phenomenology. He also does not say whether
a purely sensory system may be credited with mental states.
15 Actually, Fitch says conflicting things. At times he seems to say that the reasons
computers cant think is that they dont have nano-intentionality. At others times he
seems to say that hes only talking about vertebrates.
16 We dont know whether Searle would think it is possible to have a purely sensory
conscious being, but we dont find anything he has said that rules it out.
17 See discussion in Ma hen, 2005, Chapter 12.
18 See Adams and Aizawa, 2008.
19 See Dretske, 1981, for the interpretation of the mathematical model for use by
cognitive science that we are following.
20 We accept that even biological structures that are selected to be dedicated informa-
tion processors have a physical-chemical make-up, but physics and chemistry alone
wont explain why these structures are where they are and are doing what they are
doing. For that you need to appeal to selectional history (Dretske, 1995).
21 There is a whole literature on this feeling of magic. See Levine, 2001, and this may
be part of why McGinn adopts the mysterian view.
22 We do not say only biological systems can have minds (as Fitch comes close to
saying and Searle is taken to say). To think, computers would need concepts the
non-derived meanings of which were meaningful to them, not only to us or to their
designers.
23 Acknowledgements: James Garvey, Ken Aizawa, John Barker, Fred Dretske, William
Tecumseh Fitch, Annie Steadman, and to University of Delaware Oce of Under-
graduate Research.
Chapter 4
332
Notes
University Press, 1982) and Dualism, in S. Stich and T. Warfield (eds), The Blackwell
Guide to Philosophy of Mind, (Oxford: Blackwell, 2003).
2 Substance dualism is perhaps not commi ed to hylomorphism; see Swinburne,
The Evolution of the Soul, pp. 3302. For general discussion of the position see
H. Robinson, Aristotelian dualism, Oxford Studies in Ancient Philosophy 1, (Oxford:
Oxford University Press, 1983), pp. 12344, and D. Oderberg, Hylemorphic
dualism, Social Philosophy and Policy, 22, (2005), 7099.
3 For articulation of the nature of the classic Cartesian view see M. Rozemond,
Descartess Dualism (Cambridge, MA: Harvard University Press, 2002) and
J. Hawthorne, Cartesian dualism, in P. van Inwagen and D. Zimmerman (eds),
Persons Human and Divine (Oxford: Oxford University Press, 2007). For discussion
of Aquinass view, see A. Kenny, Aquinas on Mind (London: Routledge, 1994).
4 See J. Kim, Lonely Souls: Causality and Substance Dualism, in T. OConnor and
D. Robb (eds), Philosophy of Mind:Contemporary Readings (London: Routledge, 2003).
See also R. Larmer, Mind-body interactionism and the conservation of energy,
International Philosophical Quarterly, 26, (1986), 27785; E. Mills, Interaction and over-
determination, American Philosophical Quarterly, 33, (1996), 10515, and Inter-
actionism and physicality, Ratio, 10, (1997) 16983; and E. J. Lowe, The problem of
psychophysical causation, Australasian Journal of Philosophy, 70, (1992), 26376. Lowe
himself endorses a form of dualism, albeit not the one argued for here in E. J. Lowe,
The causal autonomy of the mental Mind, 102, (1993), 62944; Subjects of Experience
(Cambridge: Cambridge University Press, 1996), and Non-Cartesian substance
dualism and the problem of mental causation, Erkenntnis, 65,1, (2006), 523.
5 Rene Descartes, in J. Co ingham, R. Stootho and D. Murdoch (trans. and eds), The
Philosophical Writings of Descartes (Cambridge: Cambridge University Press, 1994),
vol. II, p. 275.
6 For a longer discussion of this argument, see my Belief in God (Oxford: Oxford
University Press, 2005), pp. 8799.
7 For articulation of this view, see J. Eccles and K. Popper, The Self and its Brain (New
York: Springer, 1977).
8 I qualify this conclusion somewhat in my forthcoming book Free Will (Continuum).
9 The original may be found in F. Jackson,Epiphenomenal qualia, Philosophical
Quarterly, (1982), 12736 and, for a good contemporary discussion, see the papers
collected by P. Ludlow, Y. Nagasawa and D. Stoljar (eds), Theres Something About
Mary: Essays on Phenomenal Consciousness and Frank Jacksons Knowledge Argument
(Cambridge, MA: MIT Press, 2004).
10 On the issue of the problem of consciousness, Denne , for example, arguably
re-conceptualizes it until it becomes tractable to the natural sciences in his Conscious-
ness Explained (Boston: Li le Brown, 1991), but perhaps thereby merely fails to
address the real issue; see David Chalmers, for example, in The Conscious Mind (New
York: Oxford University Press, 1996 ). See also D. Ross, Denne s Conceptual Reform
in Behaviour and Philosophy, 22 (1994), 4152 and N. Latham, Chalmers on the addition
of consciousness to the physical world, Philosophical Studies, 98, (2000), 6793.
11 I am grateful for the comments of Richard Swinburne on an early dra of this
chapter.
Chapter 5
333
Notes
other properties, as how the properties of physics determine (many think) the
chemical properties (i.e. once the properties of physics are in place, the chemical
properties are also in place), and (3) are not determined by anything that they do not
themselves determine. See Montero (2006) for a suggestion on how to formulate
physicalism if there is no fundamental level.
2 Indeed, a world that is entirely understandable to a human mind seems, if anything,
to point a non-physicalistic view, as it would seem to hint at a creator that made the
world intelligible to humans. (This isnt to say that any anti-physicalistic view implies
the existence of a creator, but just that the existence of a creator would seem to imply
an anti-physicalistic position.)
3 See Wilson (2010).
4 Terence E. Horgan (1993).
5 For an explanation of this stance see McLaughlin (2001).
6 See Melynk (2003). I argue against his view in Montero (1999).
Chapter 6
1 Jaegwon Kim has pressed this objection repeatedly. He oers the analogy of killing
someone by firing a gun. Had the gun had a silencer, the shot would still have killed
the victim. Thus, the noise of the shot was causally irrelevant. However, by analogy
had the shot been fired with by a gun with a silencer, the shooting would have been
a dierent event, and so would the death. Kim also presses the epiphenomenal charge
by claiming that it is not the mental property but only the physical property of an
event that is causally ecacious. However, for Davidson, causality is a relation
between events not properties, nor does it hold in virtue of properties of events. And
besides, Davidson doesnt believe in properties, and so Kims objections are beside
the point (see Davidson, 1993).
2 Davidson (1986a) A Coherence Theory of Truth and Knowledge, p. 150.
3 Davison rejects the locution, knowing what I mean, preferring, meaning what
I say, and he points out that if we didnt mean what we say we would be
uninterpretable.
4 Note that Denne notes that this objection is based on too limited a view of causation.
See Denne , 1991b. However, he doesnt elaborate.
5 See Fodors 1974 paper Special Sciences: The disunity of science as working
hypothesis.
Chapter 7
334
Notes
See Burge (1979). Putnam himself accepts the adaptation as can be seen by remarks
in his (1996).
4 Putnams original example was of water on Earth and twater on Twin Earth, where
twater is a substance superficially indistinguishable from water but with a dierent
chemical composition. I have not used his example here because it has an irrelevant
complication, namely that S on Earth could not be an intrinsic physical duplicate of
S* on Twin Earth if S were partly composed of water and S* were partly composed of
twater.
5 The view of natural kinds that is taken to underwrite this form of externalism can be
found in Kripke (1972) and Putnam (1975). However, the view remains controversial.
For a thorough overview of recent positions on natural kinds and natural kind terms
see Wilkerson (1998).
6 See Burge (1982). The original example involves an incomplete understanding of
the term arthritis.
7 See Burge (1986a) and (1986b).
8 The example is from Burge (1986b).
9 For reasons of space I have not been able to discuss Davidsons views, which in some
sense straddle the internalism/externalism divide. According to Davidson, roughly
speaking, the content of a thought is determined by the way in which a subject is
best interpreted in the context of a shared world. Thus although the meaning of a
subjects words and the contents of her thoughts are and must be grounded in the
discriminative capacities and epistemic outlook of the individual, they are externally
individuated nonetheless because in order to interpret an individual one must
make essential reference to objective properties in the environment to which she
(and you) are jointly related. See Davidson (1984b) especially Chapters 9 and 10, and
(2001), especially Chapters 1, 2 and 9.
10 See Evans (1982) and McDowell (1977, 1984). See also McDowell (1986).
11 See, for example, Salmon (1986) and Soames (2002).
12 See Burge (1977) for more detail. See also Segal (1989). It should be noted that the
view is consistent with both internalism and externalism about the representational
content of singular thoughts thus conceived. It counts as a form of singular inter-
nalism simply because the representational content of a singular thought is not
dependent for its individuation on the object the thought concerns. For all this,
the predicative element of the thought may depend for its individuation on the
properties to which the thinker is related, for reasons akin to those given in the
Predicative Externalism section above.
13 This view of proper names is advocated by Burge in his (1973). For recent defences
of the view see Elugardo (2002) and Sawyer (2009).
14 Epistemic outlook can itself be understood individualistically or anti-individual-
istically. This makes it dicult to oer a neutral characterization of the internalist
position conceived generally.
15 See Segal (2000) for a thorough defence of this kind of view.
16 See Fodor (1980), McGinn (1982b) and Stich (1983).
17 For this view see Chalmers (2003c).
18 See Sawyer (2007).
19 See Burge (1979), Section IV.
20 Both type- and token-physicalism are commi ed to this claim, although type-
physicalism is commi ed to the stronger claim about property identity as well.
21 That psychological states are the causes of actions (re)gained prominence following
Davidsons (1980), Chapter 1. For an alternative view see Morris (1992).
22 For arguments to this eect see Fodor (1987) and (1991a).
23 See Burge (1989) and (1993b) and Wilson (1995).
335
Notes
24 For recent discussion see Noordhof (2006a) and Kezer and Schouten (2007).
25 In my discussion of both problems I focus on predicative externalism, but the
problems and responses in each case are analogous for singular externalism.
26 For arguments concerning the strength of the argument see Ludlow (1995) and (1997)
and Warfield (1997). See also McLaughlin and Tye (1998) and Brown (2004).
27 See, for example, Burge (1988), Heil (1988) and Peacocke (1999), Chapter 5. Burge
has gone further in identifying a set of what he calls cogito-like judgements that
provide a limiting case of direct, non-empirical, authoritative self-knowledge that
withstand any amount of disruption to presupposed background conditions for
self-knowledge. Cogito-like judgements are self-referential and self-verifying in
virtue of being so. Consequently, they cannot but be true even if their contents are
externally individuated. Examples include: I am now thinking that writing requires
concentration, and I hereby judge that examples need elaboration. See Sawyer (2002) for
a defence of the view.
28 See, in particular, Burge (1988).
29 Davidson oers a dierent response that follows from his particular form of externalism
referred to in Note 9 above. For his response see Davidson (2001), Chapters 1 and 2.
30 See Boghossian (1989) for the original presentation of the argument.
31 See Ludlow (1998).
32 See Burge (1998) and (1993a).
33 In addition, and following discussion of cogito-like judgements about current
thoughts in Note 26, Burge extends the realm of cogito-like judgements to include
thoughts about past thoughts. Thus the was and thereby in judgements such as I was
thereby thinking that p relate to elements in the original thought preservatively rather
than referentially. See Burge (1998).
34 There are related issues concerning the implications of externalism for reasoning.
In addition to Burge (1993a) and (1998), see Boghossian (1992) and Schier (1992).
See Goldberg (2007) for an account of the implications of externalism for content
preservation and discursive justification.
35 McKinsey sparked the controversy with his (1991), although he rejects the inter-
pretation of his argument that led to the controversy. See also Brown (1995).
36 See Brueckner (1992) and Goldberg (2003).
37 For this strategy, see Davies (2000), for example, and Wright (2004), for example.
There are dierences between the positions of Davies and Wright but the general
strategy is the same. The strategy is criticized in Sawyer (2006).
38 See Sawyer (1998) and (2006).
39 Externalism has further epistemological implications that I have not had the space to
discuss here. See, for example, Majors and Sawyer (2005) in which it is argued that
externalism (and only externalism) grounds a reliabilist theory of justification. See
also Majors and Sawyer (2007) for the ramifications of externalism in epistemology
and meta-ethics.
40 I have not had the space to explain this fully here. See Majors and Sawyer (2005).
Chapter 8
1 This chapter is based on the much fuller discussions in my book Mind as Machine:
A History of Cognitive Science (Boden, 2006). No specific references are given below.
But the most directly relevant parts are Chapters 4 and 16 entire, and Sections 1.iii
and iii.bd; 6.iii.c and iv.c; 7.i.eh, iii, and vi.dh; 9.vii and x; 11.iiiii; 12.x; 13.vii;
14.ii and viiixi; and 15.i.
336
Notes
Chapter 9
337
Notes
11 This problem is of a piece with the problem Kripke (1982) a ributes to Wi genstein
(1953) about how to distinguish someone who has added 57 and 63 and obtained 5 as
an error from someone who is computing a dierent function, quaddition, which
is identical to addition except for the case of 57 and 63.
12 Its important to note that Fodor doesnt intend his proposal as a sucient condition
on intentionality tout court, but only as a sucient condition for meeting disjunction
objections; see his 1990c, pp. 12731.
13 In Rey (forthcoming) I argue that the representations of geometrical figures (geons)
in early vision, as well of the standard entities of linguistics (words, phonemes) are
empty along these lines.
14 One strategy is to appeal to real, but simply uninstantiated properties, such as unicorn-
hood, which (arguably) can exist even without unicorns (see Fodor, 1987, 1991b).
Another is to claim that at least syntactically simple expressions are not genuine
representations (see Millikan, 2000). See Rey (forthcoming) for discussion.
15 Note that the Zeus/Jupiter example shows one cant rely here on co-reference alone.
16 This is a condensation of a longer discussion in my (2009) in which I try to distill what
seems to me common and correct in Fodors and Horwichs proposals, while rejecting
what seems to me mistaken in each (viz., Fodors externalism and Horwichs defla-
tionism). Devi s (1996) proposal also looks to explanatory roles, but is not so focused
on the asymmetric basicality condition, which seems to me crucial to replying to the
Quinean challenge. Note that (BAS) is deliberately neutral between Horwichs
(explanatory) and Fodors (asymmetric) ways of expressing what seems to me the
common important idea.
Chapter 10
338
Notes
7 Indeed, the consensus in the literature seems to be that the Davidsonian account of
events should be avoided because it renders causation an u erly mysterious relation.
The virtue of a property exemplification view of events is that it permits an account
of why a cause produces its eect in terms of an appeal to its causally relevant
properties (Gibb, 2006).
8 For further discussion see (Campbell and Moore, 2009).
9 Kim thinks that he has not begged any questions since he claims that if it is
aspects of events, rather than events simpliciter, that are explained, then explanatory
exclusion would apply to these event aspects (Kim, 1989a, p. 96) and elsewhere, We
have so far spoken indierently of both events simpliciter and events being a certain
kind or having certain properties as causes and eects. This makes no dierence:
however we individuate causes and eects, we face the same problem [about
explanatory exclusion] (Kim, 1990a, pp. 401). I think we have seen ample reason
to disagree with these claims. Interestingly, Marras and Yli-Vakkuri (2008) have
recently observed that Kims improved version of the exclusion argument (the
supervenience argument) also presupposes a fine-grained account of event identity
and also begs the question against non-reductive physicalism.
10 Indeed, Marras (1998) ably shows it does not.
11 This observation has also been made in a slightly dierent context by Marras
(Marras, 2007; Marras and Yli-Vakkuri, 2008). It is overly simplistic simply to suggest
that the choice we face is between Davidsons and Kims conception of events, though
that is o en the sense one gains from the literature. While theirs are the most promi-
nent views there are others worth considering, such as Chisholms (1970, 1976) and
Lombards (1986), and of course there is Horgans (1978) claim that we can and
should make do without introducing events into our ontology at all. Macdonald and
Macdonald (2006) adopt a modified version of Lombards account and argue that it
allows for the causal ecacy of mental property instances. However, their approach
strikes me as a return to the Davidsonian position since events are identified with
property instances. It would be interesting to explore whether or not the exclusion
argument can be maintained on an alternative model of events. If, like Lombard,
one adopts the position that events can have more than one constitutive property
it seems an exclusion argument formulated in terms of such events would be
even more vulnerable to the dual explanandum reply. Unfortunately, I do not have
sucient space to explore this question here.
Chapter 12
1 More precisely, its an example of the version of embodied cognition with which
we shall be concerned here. There are other versions of the view. For instance,
some embodied cognition theorists concern themselves with the way in which
embodiment has an impact on our understanding of perceptual experience (e.g.
ORegan and No, 2001, No, 2004). Others argue that our embodiment structures
our concepts (Lako and Johnson, 1980, 1999). This sample is not exhaustive.
2 The classic presentation of what I am calling the extended mind hypothesis is by
Clark and Chalmers (1998). See Menary (forthcoming) for a recent collection of
papers. Rather confusingly, the view has always traded under a number of dierent
names, including close variants of the original moniker, such as the hypothesis of
extended cognition (Rupert, 2004) and the extended cognition hypothesis (Wheeler, forth-
coming a), but also active externalism (Clark and Chalmers, 1998), vehicle externalism
(Hurley, 1998; Rowlands, 2003), environmentalism (Rowlands, 1999), and locational
externalism (Wilson, 2004).
339
Notes
340
Bibliography
Adams, F. (1991), Causal Contents, in B. McLaughlin (ed.), Dretske and His critics
(Oxford: Blackwell), 13156.
(2003), Thoughts and Their Contents: Naturalized Semantics, in S. Stich and
T. Warfield (eds), The Blackwell Guide to the Philosophy of Mind (Oxford: Blackwell),
14371.
Adams, F., and Aizawa, K. (2008), The Bounds of Cognition (Oxford: Blackwell).
(1994), Fodorian Semantics, in Stich and Warfield (1994), 22342.
Adams, F., and Dietrich, L. (2004), Swampmans Revenge: Squabbles among the
Representationalists, Philosophical Psychology, 17: 32340.
Agre, P. E., and Rosenschein, S. J. (eds) (1996), Computational Theories of Interaction
and Agency (Cambridge, MA: MIT Press).
Aikns, K. (1993), A Bat without Qualities, in M. Davies and G. Humphreys (eds),
Consciousness (Oxford: Blackwell).
(1996), Lost the Plot? Reconstructing Daniel Denne s Multiple Dra s Theory
of Consciousness, Mind and Language, 2: 143.
Aizawa, K. (2003), The Systematicity Arguments (Dordrecht: Kluwer).
Aleksander, I. (2005), The World in My Mind, My Mind in the World: Key
Mechanisms of Consciousness in Humans, Animals, and Machines (Exeter: Imprint
Academic).
Aleksander, I., and Dunmall, B. (2003), Axioms and Tests for the Presence of
Minimal Consciousness in Agents, in O. Holland (ed.), Machine Consciousness
(Exeter: Imprint Academic), 718. Special issue of the Journal of Consciousness
Studies, 10 (45).
Alexander, S. (1920), Space, Time, and Deity, 2 vols. (London: Macmillan).
Anscombe, G., and Geach, P. (1954), Descartes: Philosophical Writings (Indianapolis:
Bobbs-Merrill).
Anscombe, G. E. M. (1957), Intention (Oxford: Blackwell).
(1965), The Intentionality of Sensation: A Grammatical Feature, in A. No and
E. Thompson (eds) (2002), Vision and Mind, Selected Readings in the Philosophy of
Perception (Cambridge, MA: MIT Press).
Antony, L. (1989), Anomalous Monism and the Problem of Explanatory Force,
Philosophical Review, 98: 15387.
(1997), What Its Like to Smell a Gardenia, Times Literary Supplement 4897
(7 February).
Arbib, M. A. (1982), Modelling Neural Mechanisms of Visuomotor Coordination
in Frog and Toad, in S. Amari and M. A. Arbib (eds), Competition and Cooperation
in Neural Nets, Lecture Notes in Biomathematics, 45 (Berlin: Springer-Verlag),
34270.
Arbib, M. A., and Hesse, M. B. (1986), The Construction of Reality (Cambridge:
Cambridge University Press).
341
Bibliography
Arbib, M. A., Boylls, C. C., and Dev, P. (1974), Neural Models of Spatial Perception
and the Control of Movement, in W. D. Keidel, W. Handler and M. Spreng (eds),
Cybernetics and Bionics (Munich: Oldenbourg), 21631.
Armstrong, D. (1968), A Materialist Theory of Mind (London: Routledge & Kegan
Paul).
(1981), What is Consciousness?, in The Nature of Mind (Ithaca, NY: Cornell
University Press).
Ayers, M. (1990), Locke, vol. 2 (London: Routledge).
Baker, L. R. (2000), Persons and Bodies: A Constitution View (Cambridge: Cambridge
University Press).
Balog, K. (1999), Conceivability, Possibility and the Mind-Body Problem, The
Philosophical Review, 108: 497528.
Bechtel, W. (1994), Natural Deduction in Connectionist Systems, Synthese, 101:
43363.
(1996), What Knowledge Must Be in the Head in Order to Acquire Language,
in B. Velichkovsky and D. M. Rumbaugh (eds), Communicating Meaning: The
Evolution and Development of Language (Hillsdale, NJ: Lawrence Erlbaum
Associates).
Bechtel, W., and Abrahamsen, A. (1991), Connectionism and the Mind: An Introduction
to Parallel Processing in Networks (Oxford: Basil Blackwell).
Bedau, M. (1999), Supple Laws in Psychology and Biology, in V. Hardcastle (ed.),
Where Biology Meets Psychology (Cambridge, MA: Bradford Books, MIT Press),
25171.
Beer, R. D. (1996), Toward the Evolution of Dynamical Neural Networks for
Minimally Cognitive Behavior, in P. Maes, M. Mataric, J. Meyer, J. Pollack and
S. Wilson (eds), From Animals to Animats 4: Proceedings of the Fourth International
Conference on Simulation of Adaptive Behavior (Cambridge, MA: MIT Press), 4219.
(2003), The Dynamics of Active Categorical Perception in an Evolved Model
Agent, Adaptive Behavior, 11 (4): 20943.
Behan, D. (1979), Locke on Persons and Personal Identity, Canadian Journal of
Philosophy, 9: 5375.
Block, N. (1978), Troubles with Functionalism, in C. W. Savage (ed.), Perception
and Cognition: Issues in the Foundations of Psychology, Minnesota Studies in the
Philosophy of Science, vol. 9 (Minneapolis: University of Minnesota Press),
261325.
(1986a), Advertisement for a Conceptual Role Semantics, in P. French, T. Uehling
and H. We stein (eds), Studies in the Philosophy of Mind (Minneapolis: University
of Minnesota Press).
(1986b), Advertisement for a Semantics for Psychology, in P. French, T. Uehling
and H. We stein (eds), Midwest Studies in Philosophy, vol. 10 (Minneapolis:
University of Minnesota Press), 61578; reprinted in Stich and Warfield (2003),
81135.
(1986c), Functional Role and Truth Conditions, Proceedings of the Aristotelian
Society, Supplementary Volume 60: 157181.
(1990), Inverted Earth, in J. E. Tomberlin (ed.), Philosophical Perspectives, vol. 4
(Atascadero, CA: Ridgeview), 5379.
(1994), Consciousness, in Gu enplan (1994), 21019.
342
Bibliography
343
Bibliography
Braddon-Mitchell, D., and Jackson, F. (1996), The Philosophy of Mind and Cognition
(Oxford: Blackwell).
Brentano, F. (1874), Psychology from an Empirical Standpoint (London: Routledge).
Bringsford, S. (1994), Computation, Among Other Things, is Beneath Us, Minds
and Mahines, 4 (4): 46988.
Broad, C. D. (1925), The Mind and Its Place in Nature (London: Routledge & Kegan
Paul).
Broadbent, D. E. (1958), Perception and Communication (Oxford: Pergamon Press).
Brooks, R. A. (1991a), Intelligence without Reason, in Proceedings of 12th
International Joint Conference on Artificial Intelligence (San Mateo, CA: Morgan
Kauman), 56995; reprinted in Brooks, Cambrian Intelligence: The Early History
of the New AI (Cambridge, MA: MIT Press), 13386.
(1991b), Intelligence without Representation, Artificial Intelligence, 47: 13959.
Brown, J. (1995), The Incompatibility of Anti-Individualism and Privileged
Access, Analysis, 53: 14956.
(2004), Anti-Individualism and Knowledge (Cambridge, MA: MIT Press).
Brueckner, A. (1992), What an Anti-Individualist Knows A Priori, Analysis, 52: 11118.
Bruner, J. S., Goodnow, J., and Austin, G. (1956), A Study of Thinking (New York:
Wiley).
Burge, T. (1973), Reference and Proper Names, Journal of Philosophy, 70: 42539.
(1977), Belief De Re, Journal of Philosophy, 74: 33862.
(1979), Individualism and the Mental, in P. French, T. Uehling, Jr., and
H. We stein (eds), Midwest Studies in Philosophy, vol. 4 (Minnesota: University
of Minnesota Press), 73121.
(1982), Other Bodies, in A. Woodfield (ed.), Thought and Object: Essays on
Intentionality (Oxford: Clarendon Press).
(1986a), Individualism and Psychology, Philosophical Review, 95 (1): 345.
(1986b), Intellectual Norms and Foundations of Mind, Journal of Philosophy,
83 (12): 697720.
(1988), Individualism and Self-Knowledge, Journal of Philosophy, 85: 64963.
(1989), Individuation and Causation in Psychology, Pacific Philosophical
Quarterly, 70: 30322.
(1993a), Content Preservation, Philosophical Review, 102: 45788.
(1993b), Mind-Body Causation and Explanatory Practice, in J. Heil and A. Mele
(eds), Mental Causation (Oxford: Oxford University Press).
(1998), Memory and Self-Knowledge, in P. Ludlow and N. Martin (eds),
Externalism and Self-Knowledge (Stanford: CSLI).
Bush, V. (1945), As We May Think, Atlantic Monthly, 176 (July): 1018; reprinted
in R. Packer and K. Jordan (eds), Multimedia: From Wagner to Virtual Reality
(London: W. W. Norton, 2001), 13553.
Butler, J. (1975) [1736], Of Personal Identity, in J. Perry (ed.), Personal Identity
(Berkeley and Los Angeles: University of California Press).
Byrne, A. (1997), Some Like It Hot: Consciousness and Higher-Order Thoughts,
Philosophical Studies, 86: 10329.
(2001), Intentionalism Defended, The Philosophical Review, 110: 199240.
Calude, C., Casti, J. L., and Dinneen, M. (eds) (1998), Unconventional Models of
Computation (London: Springer).
344
Bibliography
Campbell, J. (1994), Past, Space, and Self (Cambridge, MA: MIT Press).
Campbell, N. (1997), The Standard Objection to Anomalous Monism, Australasian
Journal of Philosophy, 73 (3): 37382.
(2000), Supervenience and Psycho-Physical Dependence, Dialogue: Canadian
Philosophical Review, 39: 30316.
(2003), Causes and Causal Explanations: Davidson and His Critics, Philosophia:
Philosophical Quarterly of Israel, 31 (12): 14957.
(2007), Explanatory Pluralism, International Journal of the Humanities, 5 (3):
2529.
(2008a), Explanatory Exclusion and the Individuation of Explanations, Facta
Philosophica, 10.
(2008b), Mental Causation: A Nonreductive Approach (New York: Peter Lang).
Campbell, N., and Moore, D. (2009), On Kims Exclusion Principle, Synthese, 169
(1): 7590.
Cariani, P. (1992), Emergence and Artificial Life, in C. G. Langton, C. Taylor,
J. D. Farmer and S. Rasmussen (eds), Artificial Life II (Redwood City, CA:
Addison-Wesley), 77597.
Carruthers, P. (1996), Language, Thought and Consciousness (Cambridge: Cambridge
University Press).
(2000), Phenomenal Consciousness: A Naturalistic Theory (Cambridge: Cambridge
University Press).
Cartwright, R. T. (1960/87), Negative Existentials, in his Philosophical Essays
(Cambridge, MA: MIT Press).
Chalmers, D. (1995), Facing Up to the Problem of Consciousness, Journal of
Consciousness Studies, 2: 20019.
(1996), The Conscious Mind: In Search of a Fundamental Theory (Oxford: Oxford
University Press).
(1999a), First-Person Methods in the Science of Consciousness, Bulletin from
the Center for Consciousness Studies.
(1999b), Materialism and the Metaphysics of Modality, Philosophy and Phenome-
nological Research, 59: 47393.
(2002a), Consciousness and Its Place in Nature, in Chalmers (2002b), 24772.
(2002b), Philosophy of Mind: Classical and Contemporary and Readings (New York:
Oxford University Press).
(2003a), Consciousness and Its Place in Nature, in Stich and Warfield (2003),
10242.
(2003b), The Matrix as Metaphysics, March 20th. Available at h p://
whatisthematrix.warnerbros.com.
(2003c), The Nature of Narrow Content, Philosophical Issues, 13: 4666.
(2008), Foreword to Andy Clarks Supersizing the Mind, in Clark (2008b), ixxvi.
(forthcoming), Two Dimensional Semantics, in E. Lepore and B. Smith (eds),
Oxford Handbook to the Philosophy of Language (Oxford: Oxford University Press).
Chisholm, R. M. (1956), Perceiving: A Philosophical Study, in D. Rosenthal (ed.),
The Nature of Mind (Oxford: Oxford University Press, 1990), chapter 11.
(1970), Events and Propositions, Nous, 4 (1): 1524.
(1976), Person and Object: A Metaphysical Study (London: G. Allen & Unwin and
La Salle, IL: Open Court).
345
Bibliography
346
Bibliography
347
Bibliography
348
Bibliography
349
Bibliography
(1985b), Machines and the Mental, Proceedings and Addresses of the American
Philosophical Association, 59: 2333.
(1986), Misrepresentation, in R. Bogdan (ed.), Belief (Oxford: Oxford University
Press).
(1988), Explaining Behavior: Reasons in a World of Causes (Cambridge, MA: MIT Press).
(1993), Conscious Experience, Mind, 102: 26383.
(1994), Dierences That Make No Dierence, Philosophical Topics, 22 (12): 4158.
(1995), Naturalizing the Mind (Cambridge: MA: MIT Press).
Dreyfus, H. L. (1967), Why Computers Must Have Bodies in Order to be
Intelligent, Review of Metaphysics, 21: 1332.
(1972), What Computers Cant Do: A Critique of Artificial Reason (New York:
Harper & Row).
(2003), Existentialist Phenomenology and the Brave New World of The
Matrix, The Harvard Review of Philosophy, 11 (Fall): 1831. Amended version of
Dreyfus and Dreyfus (2002).
Dreyfus, H. L., and Dreyfus, Stephen (2002), The Brave New World of the
Matrix, 20 November. Available at h p://whatisthematrix.warnerbros.com. For
an amended version see Dreyfus (2003).
Dreyfus, H. L., and Dreyfus, Stuart E. (1988), Making a Mind Versus Modelling
the Brain: Artificial Intelligence Back at a Branch Point, in S. Graubard (ed.),
The Artificial Intelligence Debate: False Starts, Real Foundations (Cambridge, MA:
MIT Press), 1543.
Eccles, J. (1987): Brain and Mind: Two or One?, in C. Blakemore and S. Greenfields
(eds), Mindwaves (Oxford: Blackwell).
Eccles, J., and Popper, K. (1977), The Self and Its Brain (New York: Springer).
Efron, A. (1992): Residual Assymetric Dualism: A Theory of Mind-Body Relations,
Journal of Mind and Behaviour, 13: 11336.
Elman, J. L. (1990), Finding Structure in Time, Cognitive Science, 14: 179212.
(1993), Learning and Development in Neural Networks: The Importance of
Starting Small, Cognition, 48: 7199.
Elster, J. (1999), Alchemies of the Mind: Rationality and the Emotions (Cambridge:
Cambridge University Press).
Elugardo, R. (2002), The Predicate View of Proper Names, in G. Preyer and
G. Peter (eds), Logical Form and Language (Oxford: Clarendon Press).
Enc, B. (1982), Intentional States of Mechanical Devices, Mind, 91: 16182.
Engelbart, D. C. (1962), Augmenting Human Intellect: A Conceptual Framework. Report
No. AFOSR 3233, prepared for the Air Force Oce of Scientific Research (Menlo
Park, CA: Stanford Research Institute); reprinted, abridged, in R. Packer and
K. Jordan (eds), Multimedia: From Wagner to Virtual Reality (London: W. W. Norton,
2001), 6490.
Evans, G. (1982), The Varieties of Reference, J. McDowell (ed.) (Oxford: Oxford
University Press).
Evere , A., and Hofweber, T., Empty Names, Fiction and the Puzzles of Non-existence
(Stanford: CSLI).
Ezquerro, J., and Vicente, A. (2000) Explanatory Exclusion, Over-Determination,
and the Mind-Body Problem. Paper read at the Twentieth World Congress of
Philosophy at Boston.
350
Bibliography
Feigl, H. (1958), The Mental and the Physical, in H. Feigl, G. Maxwell and
M. Scriven (eds), Concepts, Theories and the Mind-Body Problem: Minnesota Studies
in the Philosophy of Science, vol. 2 (Minneapolis: University of Minnesota Press),
370497; republished in 1967 with a new postscript, preface to the postscript and
additional bibliography.
(1978), Mental Representation, Erkenntnis, 13; reprinted with postscripts in
N. Block (ed.), Readings in Philosophy of Psychology, vol. 2 (London: Methuen).
(1992), Physicalism, in J. Earman (ed.), Inference, Explanation and Other
Frustrations (Berkeley: University of California Press), 27192.
Fish, W. (2009), Perception, Hallucination, and Illusion (Oxford: Oxford University
Press).
Fitch, T. (2007), Nano-Intentionality, Biology & Philosophy, 23: 15777.
Flanagan, O. (1991), The Science of the Mind (Cambridge, MA: MIT Press).
(1992), Consciousness Reconsidered (Cambridge MA: MIT Press).
Fodor, J. A. (1968), Psychological Explanation: An Introduction to the Philosophy of
Psychology (New York: Random House).
(1975), The Language of Thought (New York: Thomas Crowell).
(1980), Methodological Solipsism Considered as a Research Strategy in Cognitive
Psychology, Behavioral and Brain Sciences, 3: 6373.
(1983), The Modularity of Mind: An Essay in Faculty Psychology (Cambridge, MA:
MIT Press).
(1984), Semantics, Wisconsin Style, Synthese, 59: 231350; and in his A Theory
of Content (Cambridge, MA: MIT Press, 1990), 3149.
(1986a), Banish Dis-Content, in J. Bu erfield (ed.), Language, Mind and Logic
(Cambridge: Cambridge University Press).
(1986b), Why Paramecia Dont Have Mental Representations, in P. French,
T. Uehling, Jr., and H. We stein (eds), Midwest Studies in Philosophy, vol. 10
(Minneapolis: University of Minnesota Press), 323.
(1987), Psychosemantics: The Problem of Meaning in the Philosophy of Mind
(Cambridge, MA: MIT Press).
(1988), Connectionism and Cognitive Architecture: A Critical Analysis,
Cognition, 28: 371.
(1990a), Fodors Guide to Mental Representation, in Fodor (1990c), 329.
(1990b), Making Mind Ma er More, in Fodor (1990c), 13759.
(1990c), A Theory of Content and Other Essays (Cambridge, MA: MIT Press).
(1990d), A Theory of Content, I: The Problem, in Fodor (1990c), 5187.
(1990e), A Theory of Content, II: The Theory, in Fodor (1990c), 89136.
(1991a), A Modal Argument for Narrow Content, Journal of Philosophy, 88: 526.
(1991b), Replies, in B. Loewer and G. Rey (eds), Meaning in Mind: Fodor and His
Critics (Oxford: Blackwell), 255319.
(1998), Concepts: Where Cognitive Science Went Wrong (Cambridge, MA: MIT
Press).
(2000), The Mind Doesnt Work That Way: The Scope and Limits of Computational
Psychology (Cambridge, MA: MIT Press).
(2009), Where is My Mind?, London Review of Books, 31 (3), 12 February.
Fodor, J. A., and Lepore, E. (1992), Holism: A Shoppers Guide (Cambridge, MA: MIT
Press).
351
Bibliography
352
Bibliography
(1966), The Senses Considered as Perceptual Systems (Westport, CT: Greenwood Press).
(1977), The Theory of Aordances, in R. Shaw and J. Bransford (eds), Perceiving,
Acting, and Knowing: Toward an Ecological Psychology (Hillsdale, NJ: Lawrence
Erlbaum), 6782.
Gille , C. (2001), Does the Argument from Realization Generalize? Responses
to Kim, Southern Journal of Philosophy, 39 (1): 7998.
Gillet, C., and Loewer, B. (2001), Physicalism and Its Discontents (Cambridge:
Cambridge University Press).
Gille , C., and Rives, B. (2005), The Non-Existence of Determinables: Or a World
of Absolute Determinates as a Default Hypothesis, Nos, 39: 483504.
Glenberg, A., and Adams, F. (1978), Type I Rehearsal and Recognition, Journal
of Verbal Learning and Verbal Behavior, 17: 45563.
Godfrey-Smith, P. (1994a), A Continuum of Semantic Optimism, in Stich and
Warfield (1994), 25977.
(1994b), Spencer and Dewey on Life and Mind, in R. A. Brooks and P. Maes
(eds), Artificial Life IV (Cambridge, MA: MIT Press), 809.
Gois, I. (2007), On a Misconception about Consciousness, unpublished Ph.D.
thesis, University of London.
Goldberg, S. (2003), On Our Alleged A Priori Knowledge That Water Exists,
Analysis, 63: 3841.
(2007), Anti-Individualism, Content Preservation, and Discursive Justification,
Nos, 41: 178203.
Goodale, M. A., and Milner, A. D. (1992), Separate Visual Pathways for Perception
and Action, Trends in Neuroscience, 13: 2023.
Grau, C. (ed.) (2005), Philosophical Essays on the Matrix (New York: Oxford University
Press).
Green, C. (2003), The Lost Cause (Oxford: Forum).
Greenfield, P. M., and Bruner, J. S. (1969), Culture and Cognitive Growth, in
D. A. Goslin (ed.), Handbook of Socialization Theory and Research (Chicago: Rand
McNally), 63354.
Grice, H. P. (1961/1965), The Causal Theory of Perception, in R. Swartz (ed.),
Perceiving, Sensing, and Knowing (New York: Doubleday).
Grin, D. R. (1978), Prospects for a Cognitive Ethology, Behavioral and Brain
Sciences, 4: 52738.
(1984), Animal Thinking (Cambridge, MA: Harvard University Press).
Griths, P. (1997), What Emotions Really are: The Problem of Psychological Categories
(Chicago: Chicago University Press).
Grush, R. (2004), The Emulation Theory of Representation: Motor Control, Imagery,
and Perception, Behavioral and Brain Sciences, 27: 377442.
Gulick, R. V. (1992), Three Bad Arguments for Intentional Property Epipheno-
menalism, Erkenntnis 36 (3): 31132.
Gu enplan, S. (1994), A Companion to the Philosophy of Mind (Oxford: Blackwell).
Harman, G. (1990), The Intrinsic Quality of Experience, in J. E. Tomberlin
(ed.), Philosophical Perspectives, 4: 3152; reprinted in N. Block, O. Flanagan and
G. Gzeldere (eds) (1997), The Nature of Consciousness, 66375.
Harnish, M. (2002), Minds, Brains, Computers: An Historical Introduction to the
Foundations of Cognitive Science (Oxford: Blackwell).
353
Bibliography
354
Bibliography
Horgan, T. (1978), The Case Against Events, The Philosophical Review, 87 (1): 2847.
(1983), Supervenience and Microphysics, Pacific Philosophical Quarterly, 63:
2943.
(1993), From Supervenience to Superdupervenience: Meeting the Demands
of a Material World, Mind, 102 (408): 55586.
Horgan, T., and Woodward, J. (1985), Folk Psychology is Here to Stay, Philosophical
Review, 94: 197225.
Hornsby, J. (1980), Actions (London: Routledge).
Horwich, P. (1998), Meaning (Oxford: Oxford University Press).
(2005), Reflections on Meaning (Oxford: Oxford University Press).
(2006), The Value of Truth, Nos, 40: 34760.
Hudson, H. (2001), A Materialist Metaphysics of the Human Person (Ithaca, NY:
Cornell University Press).
(2007), I Am Not an Animal!, in P. van Inwagen and D. Zimmerman (eds),
Persons: Human and Divine (Oxford: Clarendon Press).
Humphreys, N. (1982), Consciousness Regained (Oxford: Oxford University Press).
Hunt, E. B. (1962), Concept Learning: An Information Processing Problem (New York:
Wiley).
Hurley, S. L. (1998), Consciousness in Action (Cambridge, MA: Harvard University
Press).
(forthcoming), Varieties of Externalism, in R. Menary (ed.), The Extended Mind
(Aldershot: Ashgate).
Husbands, P., Smith, T., Jakobi, N., and OShea, M. (1998), Be er Living through
Chemistry: Evolving Gas Nets for Robot Control, Connection Science, 10:
185210.
Hutchins, E. (1995), Cognition in the Wild (Cambridge, MA: MIT Press).
Hu o D. D. (2009), Mental Representation and Consciousness, in W. Banks
(ed.), Encyclopedia of Consciousness, vol. 2 (London: Elsevier), 1932.
Hu o, D. D., and Myin, E. (forthcoming), Radicalizing Enactivism (Cambridge, MA:
MIT Press).
Irwin, W. (ed.) (2002), The Matrix and Philosophy: Welcome to the Desert of the Real
(Chicago: Open Court).
Jackendo, R. (1987), Consciousness and the Computational Mind (Cambridge, MA:
MIT Press).
Jackson, F. (1982), Epiphenomenal Qualia, Philosophical Quarterly, 32: 12736.
(1986), What Mary Didnt Know, Journal of Philosophy, 83: 2915; reprinted in
his Mind, Method and Conditionals (London and New York: Routledge, 1998),
705; and in Rosenthal (1991), 3924.
(1993), Armchair Metaphysics, in J. Hawthorne and M. Michael (eds), Philosophy
in Mind (Amsterdam: Kluwer).
(1998), From Metaphysics to Ethics (Oxford: Oxford University Press).
(2003), Mind and Illusion, in A. OHear (ed.), Minds and Persons (Royal Institute
of Philosophy Supplement 53), 5171.
(2006), On Ensuring That Physicalism is Not a Dual A ribute Theory in Sheeps
Clothing, Philosophical Studies, 131 (1): 22749.
Jackson, F., and Pe it, P. (1990), Program Explanation: A General Perspective,
Analysis, 50: 10717.
355
Bibliography
356
Bibliography
357
Bibliography
358
Bibliography
359
Bibliography
360
Bibliography
Maturana, H. R., and Varela, F. J. (1980), Autopoiesis and Cognition: The Realization
of the Living (Boston: Reidel).
Mawson, T. (2005), Belief in God (Oxford: Oxford University Press).
May, L., Friedman, M., and Clark, A. J. (eds) (1996), Minds and Morals: Essays on
Cognitive Science and Ethics (Cambridge, MA: MIT Press).
Mellor, D. H. (1995), The Facts of Causation (London and New York: Routledge).
Melnyk A. (1997), How To Keep The Physical in Physicalism, Journal of
Philosophy, 94: 62237.
(2003), A Physicalist Manifesto: Thoroughly Modern Materialism (Cambridge:
Cambridge University Press).
Menary, R. (2007), Cognitive Integration: Mind and Cognition Unbounded (Basingstoke:
Palgrave Macmillan).
(ed.) (forthcoming), The Extended Mind (Cambridge, MA: MIT Press).
Menzies, P. (2003), The Causal Ecacy of Mental States, in S. Walter and H.-D.
Heckmann (eds), Physicalism and Mental Causation (Exeter: Imprint Academic).
Merleau-Ponty, M. (1962), Phenomenology of Perception (London: Routledge & Kegan
Paul).
Metzinger, T. (2003) Being No-One: The Self-Model Theory of Subjectivity (Cambridge,
MA: MIT Press).
(ed.) (1995), Conscious Experience (Paderborn: Ferdinand Schningh).
Millikan, R. (1984), Language, Thought and Other Biological Categories: New Foundations
for Realism (Cambridge MA: MIT Press).
(1986), Thoughts without Laws, The Philosophical Review, 95: 4780; and in her
White Queen Psychology and Other Essays for Alice (Cambridge, MA: MIT Press,
1993), 5182.
(1989), Biosemantics, The Journal of Philosophy, 86: 28197; and in her White
Queen Psychology and Other Essays for Alice (Cambridge, MA: MIT Press, 1993),
83101.
(1990), Truth Rules, Hoverflies, and the Kripke-Wi genstein Paradox, The
Philosophical Review, 99: 32353; and in her White Queen Psychology and Other
Essays for Alice (Cambridge, MA: MIT Press, 1993), 21139.
(2000), On Clear and Confused Ideas (Cambridge University Press).
Mills, E. (1996), Interaction and Overdetermination, American Philosophical
Quarterly, 33: 10515.
(1997), Interactionism and Physicality, Ratio, 10: 16983.
Milner, A., and Goodale, M. (1995), The Visual Brain in Action (Oxford: Oxford
University Press).
Minsky, M. L. (1965), Ma er, Mind, and Models, Proceedings of the International
Federation of Information Processing Congress, 1: 459 (Washington, DC: Spartan).
(1985), The Society of Mind (New York: Simon & Schuster).
(2006), The Emotion Machine (New York: Pantheon).
Montero, B. (1999), The Body Problem, Nos, 33 (2): 183200.
(2006), Physicalism in an Infinitely Decomposable World, Erkentnis, 64 (2): 177191.
Montero, B., and Papineau, D. (2005), A Defense of the Via Negativa Argument
for Physicalism, Analysis, 65 (3): 2337.
Moore, G. E. (1942), A Reply to My Critics, in P. A. Schilpp (ed.), The Philosophy
of G. E. Moore (La Salle, IL: Open Court), 535677.
361
Bibliography
Morris, M. (1991), Why There are No Mental Representations, Minds and Machines,
1: 130.
(1992), The Good and the True (Oxford: Clarendon Press).
Moya, C. (1990), The Philosophy of Action (Cambridge: Polity Press).
Nagel, T. (1974), What is It Like to Be a Bat?, Philosophical Review, 83 (4): 43550;
reprinted in Rosenthal (1991), 4228.
(1979) [1971], Brain Bisection and the Unity of Consciousness, in his Mortal
Questions (Cambridge: Cambridge University Press).
Neander, K. (1998), The Division of Phenomenal Labour: A Problem for Represen-
tational Theories of Consciousness, in J. E. Tomberlin (ed.), Philosophical Perspec-
tives, 12 (Boston, MA, and Oxford: Blackwell), 41134.
(2004), Teleological Theories of Mental Content, in Stanford Encyclopedia of
Philosophy. Available at h p://plato.stanford.edu/.
Nemirow, L. (1980), Review of Nagels Mortal Questions, Philosophical Review,
89: 4756.
(1990), Physicalism and the Cognitive Role of Acquaintance, in W. Lycan
(ed.), Mind and Cognition (Oxford: Blackwell), 4909.
Neurath, O. (1931), Physicalism: The Philosophy of the Vienna Circle, in
R. S. Cohen and M. Neurath (eds), Philosophical Papers 19131946 (Dordrecht:
Reidel, 1983), 4851.
Newell, A. (1980), Physical Symbol Systems, Cognitive Science, 4: 13583.
(1990), Unified Theories of Cognition. (Cambridge, MA: Harvard University Press).
Newell, A., and Simon, H. A. (1972), Human Problem Solving (Englewood Clis, NJ:
Prentice-Hall).
(1976), Computer Science as Empirical Enquiry: Symbols and Search, Com-
munications of the Association for Computing Machinery, 19 (3): 11326; reprinted
in M. A. Boden (ed.), The Philosophy of Artificial Intelligence (Oxford: Oxford
University Press, 1990), 10532.
Newell, A., Shaw, J. C., and Simon, H. A. (1958), Elements of a Theory of Human
Problem-Solving, Psychological Review, 65: 15166.
Ney, A. (2007), Can an Appeal to Constitution Save the Exclusion Problem?,
Pacific Philosophical Quarterly, 88 (4): 486506.
No A. (2003), Causation and Perception: The Puzzle Unravelled, Analysis, 63:
93100.
(2004), Action in Perception (Cambridge, MA: MIT Press).
(2006), Experience without the Head, in T. Szabo Gendler and J. Hawthorne
(eds), Perceptual Experience (Oxford: Oxford University Press), 41133.
Noonan, H. (1998), Animalism versus Lockeanism: A Current Controversy,
Philosophical Quarterly, 48: 30218.
(2003), Personal Identity, 2nd edn (London: Routledge).
Noordhof, P. (1997), Making the Change: The Functionalists Way, British Journal
for the Philosophy of Science, 48: 23350.
(1999a), Causation by Content? Mind and Language, 14: 291320.
(1999b), Micro-Based Properties and the Supervenience Argument: A Response
to Kim, Proceedings of the Aristotelian Society, 99 (1): 109114.
(2001), Believe What You Want, Proceedings of the Aristotelian Society, 2001: 101,
24765.
362
Bibliography
(2002), Imagining Objects and Imagining Experiences, Mind and Language, 17:
42655.
(2003a), Not Old . . . But Not That New Either: Explicability, Emergence and
the Characterisation of Materialism, in S. Walter and H.-D. Heckmann (eds),
Physicalism and Mental Causation (Exeter: Imprint Academic), 85108.
(2003b), Review of Consciousness, Color and Content by Michael Tye, Mind
and Language, 18: 53845.
(2003c), Something Like Ability, Australasian Journal of Philosophy, 81: 2140.
(2006a), Environment-Dependent Content and the Virtues of Causal Explanation,
Synthese, 149: 55175.
(2006b), The Success of Consciousness, in A. Freeman (ed.), Radical Externalism
(Exeter: Imprint Academic), 10927.
(2010), Emergent Causation and Property Causation, in C. Macdonald and
G. Macdonald (eds), Emergence in Mind (Oxford: Oxford University Press), 6999.
Norman, D. A. (1986), Reflections on Cognition and Parallel Distributed Processing,
in J. L. McClelland, D. E. Rumelhart and the PDP Research Group, Parallel
Distributed Processing: Explorations in the Microstructure of Cognition, vol. 1,
Foundations (Cambridge, MA: MIT Press), 53146.
Norman, D. A., and Shallice, T. (1986), A ention to Action: Willed and Automatic
Control of Behavior, in R. Davidson, G. Schwartz and D. Shapiro (eds), Con-
sciousness and Self Regulation: Advances in Research and Theory, vol. 4 (New York:
Plenum), 118.
Norman, J. (2002), Two Visual Systems and Two Theories of Perception: An A empt
to Reconcile the Constructivist and Ecological Approaches, Behavioral and Brain
Sciences, 25: 73144.
Nussbaum, M. C. (1984): Aristotelian Dualism, Oxford Studies in Ancient Philosophy,
2: 197207.
OConnor, T. (1994), Emergent Properties, American Philosophical Quarterly, 31:
91104.
OConnor, T., and Churchill, J. R. (2010), Is Non-Reductive Physicalism Viable
within a Causal Powers Metaphysic?, in C. Macdonald and G. Macdonald (eds),
Emergence in Mind (Oxford: Oxford University Press), 4360.
OLeary-Hawthorne, J., and McDonough, J. K. (1998), Numbers, Minds and Bodies:
A Fresh Look at Mind-Body Dualism, Philosophical Perspectives, 12: 34971.
ORegan, J. K., and No, A. (2001), A Sensorimotor Approach to Vision and
Visual Consciousness, Behavioral and Brain Sciences, 24 (5): 883975, 93973.
Oderberg, D. (2005), Hylomorphic Dualism, Social Philosophy and Policy, 22: 7099.
Olson, E. (1997), The Human Animal: Personal Identity without Psychology (New York:
Oxford University Press).
(2007), What are We? A Study in Personal Ontology (Oxford: Oxford University
Press).
Osherson, D. (1995/98), An Invitation to Cognitive Science, 4 vols. (Cambridge, MA:
MIT Press).
Papineau, D. (1984), Representation and Explanation, Philosophy of Science, 51:
5573.
(1986), Semantic Reductionism and Reference, in J. Bu erfield (ed.), Language,
Mind and Logic (Cambridge: Cambridge University Press).
363
Bibliography
364
Bibliography
365
Bibliography
366
Bibliography
367
Bibliography
368
Bibliography
369
Bibliography
(1992), The Emperors Real Mind, Review of Roger Penroses The Emperors New
Mind: Concerning Computers Minds and the Laws of Physics, Artificial Intelligence,
56, 35596.
(1999), Review of R. Picards Aective Computing, AI Magazine, 20: 1 (March),
12733.
(2000), Architectural Requirements for Human-Like Agents Both Natural
and Artificial. (What Sorts of Machines Can Love?), in K. Dautenhahn (ed.),
Human Cognition and Social Agent Technology: Advances in Consciousness Research
(Amsterdam: John Benjamins), 16395.
(2002), The Irrelevance of Turing Machines to Artificial Intelligence, in M. Scheutz
(ed.), Computationalism: New Directions (Cambridge, MA: MIT Press), 87127.
(2010), An Alternative to Working on Machine Consciousness, International
Journal of Machine Consciousness, 2 (1): 118.
Sloman, A., and Chrisley, R. L. (2003), Virtual Machines and Consciousness, in
O. Holland (ed.), Machine Consciousness (Exeter: Imprint Academic), 13372.
Special issue of the Journal of Consciousness Studies, 10 (45).
Smart, J. J. C. (1959), Sensations and Brain Processes, Philosophical Review, 68:
14156; reprinted in Rosenthal (1991), 16976.
(1978), The Content of Physicalism, Philosophical Quarterly, 28: 23941.
Smith, A. D. (2002), The Problem of Perception (Cambridge, MA: Harvard University
Press).
Smith, B. C. (1985), Prologue to Reflection and Semantics in a Procedural Language,
in R. J. Brachman and H. J. Levesque (eds), Readings in Knowledge Representation
(Los Altos, CA: Morgan Kauman), 3140.
(1996), On the Origin of Objects (Cambridge, MA: MIT Press).
(1998), On Knowing Ones Own Language, in C. Wright, B. C. Smith and
C. Macdonald (eds), Knowing Our Own Minds (Oxford: Oxford University Press).
(2002a), The Foundations of Computing, in M. Scheutz (ed.), Computationalism:
New Directions (Cambridge, MA: MIT Press), 2358.
(2002b), Keeping Emotions in Mind, in P. Goldie (ed.), Understanding Emotions:
Mind and Morals. Ashgate Epistemology and Mind series (London: Ashgate).
Smith, T., Husbands, P., and OShea, M. (2002), Neuronal Plasticity and Temporal
Adaptivity: GasNet Robot Control Networks, Adaptive Behavior, 10: 16183.
Smolensky, P. (1987), The Constituent Structure of Mental States: A Reply to
Fodor and Pylyshyn, Southern Journal of Philosophy, 26: 13760.
(1988), On the Proper Treatment of Connectionism, Behavioral and Brain Sciences,
11: 174.
Smythies, J. R., and Belo, J. (eds) (1989), The Case for Dualism (Charlo esville:
University of Virginia Press).
Snowdon, P. (1990), Persons, Animals, and Ourselves, in C. Gill (ed.), The Person
and the Human Mind (Oxford: Clarendon Press).
(1996), Persons and Personal Identity, in S. Lovibond and S. G. Williams (eds),
Essays for David Wiggins: Identity, Truth and Value (Oxford: Blackwell).
Soames, S. (2002), Beyond Rigidity: The Unfinished Semantic Agenda of Naming and
Necessity (Oxford: Oxford University Press).
Sober, E. (1992), Learning From Functionalism: Prospects for Strong Artificial
Life, in C. G. Langton, C. Taylor, J. D. Farmer and S. Rasmussen (eds), Artificial
Life II (Redwood City, CA: Addison-Wesley), 74966.
370
Bibliography
371
Bibliography
372
Bibliography
373
Bibliography
374
Index
375
Index
376
Index
377
Index
378
Index
379
Index
380
Index
381
Index
382
Index
383
Index
384
Index
385
Index
what-it-is-likeness31, 32, 378, 49, 51, Wi genstein, L.155, 174, 178, 314, 338n11
8990, 255, 2589, 268, 308, 31617 Philosophical Investigations283
Wheeler, M.157, 162, 220, 222, 223, Wollheim, R.131
232, 233, 276, 339n2 Woodru, G.104
Whitby, B.169 Wright, C.336n37
White, S.188
Whyte, J. T.275 XYZ see H2O XYZ
Wiggins, D.215, 216
Wigner, E.100 Yablo, S.23, 252, 330n2
Wilkerson, T. E.335n5 Yli-Vakkuri, J.339nn9, 11
will, significance of317
Williamson, T.179 Zahavi, D.330n1 (Ch 2)
Wilson, J.245, 334n3, 335n23, 339n2 Zi, P.164
Witmer, G.95 zombies24, 34, 434, 97, 164, 31718
386