CHAPTERONE
Frontal Assault
Nobody has the slightest idea how any thing material could be
conscious. Nobody
even know swhat it would be like to have the slightest idea So
much for the philosophy of consciousness. (Fodor1992)
One of my clearest memories of early childhood finds me sitting
alone in my
bedroom at twilight when I was about five, pondering a curious
family of questions. Why does the universe exist? What if it
didn't? What would be left over if
it stopped existing? Wouldn't something still exist? What colour
would it be?
Even now , the questions elicit the same peculiar twisting sensation
from my
stomach. And now, as then, I find the basic mystery of why anything
exists the
most unfathomable of all.
Probably it was fortunate for the sake of a happy childhood that
the gravity
of the second most difficult question didn't impress itself on
me until years later:
how is it that it feel like it does to be me (or anyone else)
if that feeling is done
with nothing more than a stringy ball of nerve fibres, glia,
and other organic
structures made mostly of water? How can it possibly feel like
anything?
It is unusual that I agree significantly with Fodor , especially
in print . But I
think his pessimistic prognosis that an unassailable conceptual
roadblock separates consciousness and the material world is almost
right. After all, conscious
experience seems entirely different from mundane features of
material objects
like mass or colour or fuzziness . Such attributes generate comparatively
few
deep metaphysical quandaries. While the road from fundamental
physics and
chemistry to their macroscopic appearance is, in practice, a
very long and complex one which has yet to be explored in every
detail, in principle nothing stands
in the way of a complete low level account of, say, the particular
bulk, reddishyellow hue, and deceptively appealing fuzziness
of a ripe peach. No one puzzles
over the question 'how can a material object like a peach possibly
be fuzzy ?' .
But on the face of it, at least, the characteristic features
of phenomenal experience stand in nothing at all like the same
sort of relationship to low level physics
and chemistry that renders peach fuzz so metaphysically unintriguing.
While a fuzzy peach looks to be a straightforward example of
material sub stances arranged in the right way, it isn't easy
to see , even in principle, how
Nature could have built something like a conscious mind out of
matter. Taking
as a working hypothesis the notion that fundamental particles
of matter lack
phenomenal experience entirely, it isn't easy to see how any
set of such particles-
however they might be organised, energised, shaken, or stirred
could
somehow come to be conscious. Nevertheless, each day we find
ourselves surrounded
by creatures who seemingly enjoy conscious experience without
incorporating
within themselves any extra 'secret ingredients' beyond myriad
ordinary
material particles. (At least, contemporary science has yet to
uncover any
such extra ingredients.)
After years of flirting with extravagant but, ultimately, explanatorily
bankrupt
nonphysical candidates for a solution to this apparent mystery,
some significant
conclusions have finally begun to emerge which lead me to think
the
instantiation of conscious minds by purely material structures
really may be
fathomable after all. This book explores some of those conclusions
and the lines
of thought behind them, constructing a story within which conscious
experience
and the material world may plausibly be unified.
Taking the position that understanding the puzzles of conscious
experience
cannot proceed without some grasp of who or what the subject
of conscious
experience might be, I ultimately defend the view that the conscious
subject-
the I-is a materially instantiated structure of dynamically changing
information.
On this account, phenomenal experience is 'what it is like to
be' such a changing
structure. While it may be true that it is not identical to any
material thing, I
suggest that a conscious mind may nonetheless be implemented
by matter. For
now, I will say little more about this central unifying idea,
apart from noting that
information is understood here in a wholly naturalistic, precise,
and objective
sense and quantified within a formal framework provided by the
field of algorithmic
information theory. Several chapters of preliminaries-concerning
information
and representation, problems of perspective, functionalism, supervenience,
and mental states-must be completed before much sense can be
read
into this view of the conscious subject, which doesn't receive
full treatment until
the latter half of Chapter 6. Drawing on resources from traditional
philosophy,
computer science, mathematics, physics, and neuroscience, the
job of developing
and exploring this account amounts to a full frontal assault
on the clutch of
confusions which philosophers collectively refer to as 'the mind/body
problem'.
Many themes underlie this campaign (some of which receive more
attention
below in section 2), but of paramount concern is my aim of embracing
the rich
'ineffable feel' of the phenomenal world while still operating
within the constraints
set by the laws of physics. The method I advocate here takes
the laws of
physics as given and then examines what sort of picture of consciousness
and
of cognition might be built up within some framework consistent
with those
laws. In other words, we assume for the sake of enquiry that
the creatures of
interest, such as humans, are instantiated in a purely physical
world and then
attempt to construct a picture of how (or whether) those creatures
could possibly
be conscious in such a world. The main task then is to evaluate
whether such a
picture accounts fully for the way our conscious experience really
does seem to
us in the actual world or whether some significant feature has
been left behind.
Only if one or more significant features cannot be accommodated
should we
then consider rejecting both the framework itself and perhaps
even the original
assumption that consciousness exists in a purely physical world.
This 'world first', or 'physics first', approach, which begins
with the
physical and attempts to work toward a phenomenal goal, contrasts
starkly with
methods traditionally dear to the hearts of philosophers. More
often, the preferred
starting place includes some set of features of phenomenal experience
thought to be 'manifestly clear' to a subject's introspection,
and the goal is to
reconcile those features with what we observe in the physical
world.
2
But frequently
these innocuous clear beginnings lead on in very short order
to conclu-sions
clearly conflicting with what a physicalist picture can deliver.
It shouldn't
be too surprising, however, if starting only with what seems
manifestly clear
from heady introspection sometimes leads us to places which are
in fact impossible
to accommodate with actual mere physics.
For instance, it is not unusual to include in an initial flotilla
of assumptions
the notion that human minds can generate arbitrarily many well-formed
English
sentences, or that they can perform addition over the full domain
of natural
numbers, or that their conception of a valid proof can be infinitely
extended.
Likewise, it is not uncommon to suppose that human minds are
capable, in
principle, of perfect rationality-in the sense of being able
to find all deductively
valid consequences of a set of beliefs-or even that they are
capable of determining
whether any particular proposition (say, in first order logic)
follows
from a given set of assumptions. Despite the fact that not one
of the above assumptions
finds support from a single scrap of empirical evidence, even
many
naturalistically minded philosophers happily invite them on board.
A popular
justification for doing so, it seems, boils down to a belief
that these 'in principle'
ideals are what really need explaining, and that actual world
deviations from
the ideal are little more significant than the noisy boundary
conditions which
always blur experimental measurements away from the ideal predictions
of a
good theory. As it happens, however, convincing reasons suggest
all the above
assumptions are false for any cognizer physically instantiated
in the real
world-even though the failure of each is consistent with the
appearance, from
the vantage point of the cognizer concerned, that it is true.
Many of the mind/body confusions which this book attempts to
clear away
find their roots in just such lavish assumptions about what seems
clear from introspection.
For progress to be made on the puzzle Fodor thinks no one should
even hope to understand, it is helpful for philosophers to adopt
a little more
modesty about their own capabilities, acknowledging that-in the
absence of
some argument to the contrary, and mere appearances aside-the
default as4 CHAPTER ONE
sumption should be that we humans are as subject to the constraints
of the laws
of physics as any helium atom in the core of a star or any fuzzy
peach poised to
fall from a tree. This book, I hope, marks significant moves
in that direction.
1. STRUCTURE AND OVERVIEW
In view of its interdisciplinary nature, the text has been organised
in a way
which I hope will accommodate readers with a wide variety of
backgrounds and
personal interests in the material. Below, I outline my attempts
to package farflung
topics in an accessible way, describe the book's anticipated
audiences,
and briefly preview each chapter.
1.1 Pathways Through the Book
Because perhaps only a minority of readers whose interests happen
to coincide
with my own will want to pursue the arguments set out here linearly
from start
to finish, I have structured the book with a view to making other
reading strategies
as painless as possible. Especially technical remarks are typically
flagged in
advance, and in each such case I suggest an alternative route
through the text
which will allow readers with less interest in details to skip
ahead without
missing central points. Around two hundred cross-references to
particular
chapters, sections, or page numbers, together with an extensive
index, should
help those taking a nonlinear path through the book to locate
other points supporting
key lines of thought.
3
With a few exceptions, the discussions require little in the
way of specialist
background knowledge, but by no means do I intend the book as
a comprehensive
introduction to any of the subdisciplines which feature in it,
nor to the
mind/body problem itself. Although in places the text still reveals
roots in a
doctoral dissertation, the overhead dedicated to one dissertation
favourite, reviews
of existing literature, is substantially reduced-particularly
in areas (such
as representation) where fresh perspectives are in the offing,
perspectives which
may fit only awkwardly or not at all into the categories provided
by that literature.
Likewise, I aim mainly to present views positively, largely dispensing
with cautionary lists detailing what they are not and contrasting
them with the
many similar cousins for which they might be mistaken. And while
I hope the
broad collection of about five hundred references provides a
helpful start for
readers following up particular threads, it remains far from
complete. I have
tried to avoid the 'my bibliography is bigger than your bibliography'
syndrome-
the urge to cite the kitchen sink-which seems to be spreading
ram-pantly
through populations of undergraduates and seasoned researchers
alike
(perhaps under the influence of easy to use electronic abstract
databases).
FRONTAL ASSAULT 5
As for exceptions to general accessibility, I do assume familiarity
with basic
logical connectives, plus a few mathematical and set theoretic
symbols, and I
assume experience with the philosophical notions of a priori,
a posteriori, intension,
extension, possible worlds, and the like. Prior acquaintance
with Turing
Machines is helpful. The 'worst' exception occurs in Chapter
7; the summary
notes on the quantum formalism beginning on page 145, as well
as some subsequent
sections, will be most useful for those with at least a passing
familiarity
with linear algebra. However, as always, those portions may safely
be skipped
by readers preferring to bypass the particulars, and pointers
on where to pick up
again are of course included.
Overall, I hope the book will be enjoyable for most people interested
in materialist
cognitive science and the challenges of understanding consciousness,
from advanced undergraduates to senior researchers and those
whom publishers'
marketing departments sometimes dub 'the motivated lay reader'.
It may
find a place in graduate seminars or as a supplementary text
in philosophy, cognitive
science, artificial intelligence, or artificial life.
1.2 Chapters Summary
The second chapter is a warm-up exercise. Taking up Dennett's
recent challenge
to defend the philosophical relevance of zombies without begging
important
questions about their capabilities (or lack thereof), the chapter
includes an architecturally
explicit zombie construction and applies it to motivate later
explora-tions
not of how conscious subjects behave but of what internal processes
bring
about that behaviour. Main outcome: nonconscious zombies with
external be-haviour
indistinguishable from that of normal subjects are logically
possible, so
conscious experience requires more than the right external behaviour.
Those later explorations of internal processes depend largely
on the precise
and objective notions of information and of representation which
take centre
stage in Chapter 3. Introducing a purely physical view of information
based on
Gregory Chaitin's version of algorithmic information theory,
the chapter describes
representation in the general case with a formal measure of mutual
information
content between two physical objects. (Note that this appeal
is to a
concept of information which differs from Shannon's, used by
Dretske 1981.)
Along the way, I outline the modern and easily understood information
theoretic
version of incompleteness results, questioning the foundations
of
claims occasionally made about incompleteness and its bearing
on philosophy of
mind. Main outcome: representation is objectively linked to the
physical world.
The next chapter shows why, far from underwriting a convincing
argument
against physicalism, the curiosities of Jackson's example of
Mary, the colour deprived
neuroscientist, arise naturally within a wholly physicalist setting.
Tricky puzzles of perspective, or points of view, first arise
here-to return in
6 CHAPTER ONE
Chapter 6-and the relationship between logic and the physically
instantiated
cognizers using it merits brief remarks. Main outcome: the same
information
may be physically instantiated in many distinct ways and with
disparate ramifications
for conscious experience; Mary's ignorance before seeing red
for herself
is not a matter of information, but of state.
Chapter 5 outlines a new formal framework for understanding functional
systems. Problems of mathematical triviality threaten traditional
approaches to
functionalism based on correlation or correspondence, while teleofunctionalism
requires links to facts historically removed from the system
in question, rendering
it more suitable for questions like 'why is this component here?'
than
those like 'how does this system work?'. This chapter's objective
method of
functional decomposition, using a formal measure of process complexity
called
functional logical depth (inspired by work of Chaitin and Charles
Bennett), circumvents
both difficulties. Main outcome: the revised functionalism is
now ob-jective
and better disposed for explanatory work.
Within a context featuring arguments about supervenience, perspectives,
and mental states, Chapter 6 outlines the first components of
a theory of consciousness
based on the self model, a materially instantiated dynamic data
structure which emerges as a promising candidate for the seat
of phenomenal
experience. On this view, one motivated by the need for a conceptual
link between
consciousness and the material world yet grounded empirically
and thus
falsifiable, conscious experience is 'what it is like to be'
a particular kind of
changing data structure. The self is not identical to any of
those physical com-ponents
underlying it, yet it is implemented by them; I am a self model.
Main
outcome: taking the concept with an appropriate intension, consciousness
supervenes
logically on the physical world, with the self model as link.
Taking a brief side trip to debunk a competing class of theories,
the so-called
'quantum theories of consciousness', the next chapter describes
the work
of Roland and the mechanisms of interactive decoherence which,
auto-matically
and extremely efficiently, eliminate the need for a conscious
observer
in quantum mechanics and guarantee that special quantum effects
are virtually
nonexistent at the levels of description where they are sometimes
imagined to
feature in the human brain. As a bonus, interactive decoherence
accounts for the
automatic emergence of apparently deterministic quasi-classical
reality from a
quantum substrate. Main outcome: quantum physics doesn't need
consciousness,
and consciousness doesn't need quantum physics.
Where Chapter 6 focused mainly on the conceptual territory between
the self
model and supervening consciousness, Chapter 8 sets out in the
opposite direction,
from the self model down toward the sorts of lower level materials-neural
tissue, in the case of humans and other terrestrial life forms
which may implement
such data structures. After tidying the information theoretic
description
FRONTAL ASSAULT 7
of self models given in Chapter 6 and introducing basic tools
from neuroscience,
the chapter shows how one particular research programme, Stephen
Grossberg's adaptive resonance theory, bears especially on the
task of implementing
self models in real neural systems. Main outcome: the right neural
systems
can do the representational work which self models require, and
real organisms
may plausibly have evolved so as to implement them.
Turning from cognitive models to the mathematical presuppositions
underlying
them, Chapter 9 examines the relationship between models and
properties
of the physical systems being modelled. In particular, the chapter
examines re-cent
work by Hava Siegelmann and Eduardo Sontag suggesting that analogue
models (based on the real numbers) may display computational
capabilities
which exceed those of digital models (based on the rational numbers).
To the
extent that the two sorts of models differ in their capabilities,
an interesting
question then arises as to whether one or the other makes for
a better match with
reality. Main outcome: contrary to popular dogma, the choice
of number sys-tems
does make a difference for models of cognitive systems, but the
significance
of that difference for the real world has yet to be established.
Finally, Chapter 10 recaps some of the text's central themes
and reflects on
the broader significance of the mind/body problem for understanding
who we
are, as individuals and as a civilisation. It finishes with some
remaining open
questions and possible directions for future research.
Readers who find my attempts in each of these chapters to situate
the discussion
within an overall picture a little too obscure might also want
to skim
through that last chapter's section 2, starting on page 243,
which, unlike the
above summary, surveys the principal developments of each discussion
on the
assumption that readers will already have acquired some familiarity
with them.
2. UNDERLYING THEMES AND METHODS
Although the book treats many distinct topics, all are united
both in the sense
that they relate to some aspect of the mind/body problem and
in terms of under-lying
themes. Below I begin with a theme I specifically attempt to
sidestep and
move on to two more positive ones. Readers less interested in
comparatively
dull methodological issues should skip ahead to section 3 below.
2.1 Dynamics and Computation
One fracas I hope to avoid is the war raging between advocates
of dynamical
and computational methods in cognitive science. At a recent workshop
marking
the opening of Sussex University's Centre for Computational Neuroscience
and
Robotics, for example, I was astonished at the number of participants
who ap-peared
quite abruptly to have rediscovered dynamical systems and were
vigor8 CHAPTER ONE
ously promoting a dynamical approach to artificial life and related
fields as the
greatest new advance-perhaps 'revelation' would be more fitting-in
decades.
Worse than the hype
4
was the impression that one should be exclusively either
dynamicist or computationalist and that for the sake of scientific
purity the two
must not be mixed-a notion asserted with characteristic vehemence
by Tim
Smithers, the well known 'non-representationalist' roboticist
from the University
of the Basque Country. I hardly dared raise a voice for a hybrid
view for
fear of witnessing my own public stoning!
Typically, allegiances to a particular approach run deep, often
it seems to
such an extent that what appears at a glance a straightforward
observation supporting
one or the other position becomes utterly invisible to those
favouring the
opposing viewpoint; all too frequently, arguments are simply
ignored or contradicted
rather than rebutted. I hope in this book to challenge both sides
while incensing
neither. For my part, while generally inclined more in the direction
of a
dynamical approach, I believe tools of both types play a valuable
in cognitive
science: both dynamics and a computationally defined variety
of representation
feature crucially in my own view. In Chapter 3, I outline a method
of quantifying
information content (and representation) which, while itself
wholeheartedly
computational, applies equally well to all physical systems,
whether interpreted
as computational or dynamical entities. Later, in Chapter 6,
I argue
against the typical computationalism-friendly notion of an instantaneous
mental
state in favour of an inherently temporal replacement, advancing
a view of consciousness
itself which, while described in 'computational' information
theoretic
terms, subsequently receives a dynamical account of neural instantiation
in
Chapter 8. (For readers harbouring strong feelings on the debate
and an irrepressible
urge to categorise this book in the language of it, the approach
I adopt
here might be verbosely labelled, with tongue in cheek, as 'computationally
described,
dynamical semi-representationalism'.)
Contra Smithers, I emphatically do not believe that one should
decide in advance
to adopt one framework or the other, on pain of tainted science.
Whether
a good explanation of observed phenomena should be dynamical,
computational,
or hybrid in nature is a question properly evaluated in light
of empirical
evidence and the resulting matches between candidate theories
and reality; it is
not, to my mind, a matter of evaluating empirical evidence in
light of a preexisting
conviction that explanation is to be found in one and only one
particular
form. Likewise, contra van Gelder (see note 4 above), I find
very little ontological
mileage in the methodological distinction between dynamical and
computational
approaches to cognitive science. As far as I can see, whether
we
speak in computational language, dynamical language, or little
green men language,
we still talk about the same actual world. ('World first', not
'words
first'!) In the terminology of Chapter 6, the 'things' talked
about supervene
FRONTAL ASSAULT 9
logically on the physical world. Depending on the needs and aims
of any given
situation, we might find it useful sometimes to describe cognitive
processes
computationally, while at other times dynamical or hybrid descriptions
will
prove most helpful. There may be a de facto rough cultural division
between
cognitive scientists who generally prefer computational tools
and those who
generally prefer dynamical tools, just as there is a de facto
rough division between
those who prefer reverse Polish notation calculators and those
who prefer
algebraic calculators, but such a division needn't necessarily
reflect ontologically
significant facts about the real world of cognizers which they
are studying.
2.2 Priorities and the Naturalistic Urge
Closely related to the business of choosing appropriate languages
of explanation
are questions about the goals of research itself and the choices
of tools which
those goals motivate. One opinion popular in the United Kingdom,
at least,
counts the task of furthering the cause of a particular discipline
or even department
as, if not quite an end in itself, at least a highly significant
component of
academic research. On this view, a central aim of the philosophy
researcher
should be to advance the understanding of philosophy for philosophy's
sake;
the general field of philosophy is primary, while the particular
issues to be explored
take a back seat. A diametrically opposed tack shapes my own
research
in general and this book in particular.
My main interest lies not with furthering a particular discipline
(or, even
less, a department), but with examining fascinating questions,
whatever disciplines
may count as their own those questions or their solutions. I
make no
apologies for importing methods or conclusions from outwith traditional
philosophy,
particularly from scientific fields, in the course of exploring
the
mind/body problem.
5
Although many philosophical colleagues in the United
Kingdom are quick to dismiss my own work as "science, not
philosophy!"-as
though the two were mutually exclusive-I am only too happy to
recruit for
'philosophical' purposes the tools of other fields. (Were I to
encounter tomorrow
some good reasons for believing the art of pottery held the key
to understanding
problems of consciousness, I would learn to make pots!) In the
spirit
of thinkers like Kitcher (1992) and those featured in Kornblith
(1994), I take
philosophy and the natural sciences as forming a continuum. The
eloquent
words of E.W. Hobson, Sadleirian Professor of Pure Mathematics
and Fellow
of Christ's College in Cambridge, speaking at the conclusion
of his 1921-22
Gifford Lectures at the University of Aberdeen, express the idea
succinctly:
If we were in possession of, and able to grasp, a unified view
of the Universe, in
which all the elements of existence and valuation were completely
synthesizedwe
should not require to mark out frontiers between Science and
Philosophy or Theology
The untrammelled freedom which must be allowed to workers in
all depart10 CHAPTER ONE
ments of the great cultural work of humanityshould notinvolve
the erection of
rigid impassable barriers which shall mark off domains which
hold no communication
with one another. On the contrary, workers in one department
will often receive
the most valuable enlightenment, and most important suggestions,
from quarters outside
their own special line. (Hobson 1923, p. 501)
Borrowing the contemporary words of Sussex researcher Inman Harvey
(personal communication), I also agree wholeheartedly with the
notion of
"doing philosophy of mind with a screwdriver"-testing
philosophical ideas
with robotics and artificial life in real world laboratories.
In an article entitled
'Artificial Life as Philosophy', Dennett (1994) advocates a similar
view.
But despite a distinguished tradition of positive and mutually
beneficial interaction
between philosophy and the sciences, philosophy of mind is, perhaps
more than at any time in its history, feeling the heat from scientific
areas as
those fields 'encroach upon' the study of questions still considered
by many to
be philosophers' territory. Quite justifiably, methods and strategies
from areas
like quantum physics, chaos theory, neuroscience, and artificial
life have been
stirring up the field, eliciting responses from the philosophers'
camp ranging
from the sort of dismissiveness recalled above to enthusiastic
endorsements of
method x as that which will at last naturalise the study of mind.
Although my
own tendencies draw me firmly in the naturalistic direction,
explicitly arguing
the case for the relevance of scientific fields to 'philosophical'
questions really is
not my battle. I hope others will simply evaluate my judgement
in choosing the
tools I have in light of the lines of argument they support,
rather than under the
shadow of some pre-existing prejudice as to how best to secure
the borders of
philosophy against infiltration.
2.3 Description Complementarity and Puzzles of Perspective
After the central goal mentioned at the start of the chapter,
that of embracing the
'ineffable feel' of phenomenal experience while simultaneously
heeding the
constraints of physical law, the next most significant thread
running through
this book concerns the relationships between different levels
of description
6
of
the same cognitive system and between first and third person
perspectives on
cognitive systems. I believe both relationships bear crucially
on the project of
understanding how (or whether) matter can implement a mind.
With respect to the first, a good many problems in the history
of philosophy
originate, in my opinion, with a failure to appreciate the complementarity
between
alternative descriptions of the very same thing-a failure which,
at times,
borders on stubborn-minded silliness. Throughout the explorations
of the topic
which appear here, I have in mind a background context of working
hypotheses
inspired by the likes of Davidson (1973) and Hellman and Thompson
(1975); I
opt for what the latter call 'ontological determination'-the
physical is all there
FRONTAL ASSAULT 11
is, and everything that happens physically is governed entirely
by the low level
laws of physics-coupled with 'explanatory anti-reductionism'.
In other words,
nothing ever happens which is not, at the lowest level, entirely
a result of the
laws of physics;
7
yet, in giving intelligible explanations of processes, we may
well have to rely on entities constructed at a higher level of
description commensurate
with that at which we describe the processes themselves.
For instance, it would be no explanation of how a clock keeps
time with its
hour and minute hands to describe the interactions, in accordance
with the laws
of physics, of every single subatomic particle within it. A good
explanation
would describe instead the interactions of cogs and pendulums
or transistors
and quartz crystals (depending on the clock) and the relationship
of those interactions
to the movements of the hands. Yet nothing ever happens to (or
be-tween)
cogs, pendulums, transistors, quartz crystals, or arms of a clock
which
doesn't follow straightforwardly from the lowest level laws of
physics. The approach
to explanation favoured in this book embraces both observations
to-gether.
The view contrasts starkly with those of Durkheim (1938) or Radcliffe-Brown
(1952), for instance, who maintain that good explanations require
not
only an appropriate level of description, but a new independent
level of things
whose causal properties, significantly, do not follow from those
of their constituent
parts.
8
In the guise of a discussion of evolution and Conway's Game of
Life (Poundstone 1985), Dennett (1995c, pp. 166-75) offers the
tidiest look at
levels of description I have encountered to date.
Interest in the relationships between alternative descriptions
of the same
thing doesn't end with different levels; a special case of complementary
descriptions
exists in the distinction between first and third person perspectives
on
cognitive systems. Just as I believe there is a common failure
to appreciate the
relationships between levels of description, so, too, is there
a common failure to
appreciate the factors underlying the truism that reasoning about
a system as a
third person differs greatly from being that system in the first
person. If, for
instance, I cannot come to know 'what it is like' for me to be
in a particular state
until I have actually been in that state or one relevantly similar
to it, why should
it be at all puzzling that I cannot come to know what it is like
for someone else
to be in such a state? And is there any reason to think I should
be able to grasp
what it is like for me to be in a particular state before having
actually been in
such a state? Such perspectival issues figure in a wide range
of questions about
the relationship between mind and the physical world. My direct
quarrel with a
standard view that problems of perspective support a case against
physicalism
begins in Chapter 4 and returns in Chapter 6, but a reluctance
to accept the standard
view underlies much of the rest of the book.
12 CHAPTER ONE
3. COGNITIVE DISSONANCE
It was in one of the first philosophy texts I encountered as
an undergraduate,
Richard Taylor's (1963) Metaphysics, that I vividly recall reading
the tale of a
poor fellow called Osmo, who perishes while trying to escape
a future foretold
for him in a special book. With perfect accuracy, the mysterious
volume describes
every event in Osmo's entire life, from his birth through to
the present
and on into the future and his eventual death. Like the protagonist
in a Greek
tragedy, Osmo cannot escape his Fate: eerily, whatever the book
foretells always
comes to pass, and try as he might, his every failed struggle
to forge a
new and different future for himself only underscores the book's
apparent infallibility.
Taylor's point in articulating the story-apart, perhaps, from
terrorising impressionable
young philosophers-is to suggest that we all should view the
fu-ture
with the eyes of a fatalist. Fatalism should appeal, Taylor argues,
for purely
logical reasons: since some complete description of all my life's
future events is
true right now, and since that description could already have
been written down
in a book (say, by an omniscient and prolific author), I lack
any sort of freedom
to change it, and I ought just to stop worrying about it! Despite
Taylor's protests
to the contrary, the argument for fatalism seems a textbook example
of
modal fallacy. But nevertheless, I think I can almost grasp what
might have led
him to suggest the line of thought, despite his fluency with
the logic of modal
operators. Although most philosophers reject Taylor's reasoning,
the years have
not dulled my impression that something remains deeply disturbing
about the
fantasy scenario he describes; a book accurately foretelling
my every experience-
perhaps even my every thought-for the rest of my life truly would
be a
frightening prospect.
That peculiar discomfort, I believe, bears a possibly illuminating
likeness to
the distaste many express for the idea that human cognition and
consciousness
might one day be explained within a purely materialist framework.
I am not
suggesting that those unhappy about such a possibility are confused
about modal
logic! Rather, although the analogy is imperfect with respect
to logical
structure, there seems to me an appealing kind of symmetry between
the two
cases. On one hand, it is disturbing to think that some book
could in principle
expose, for all to see, every single event in my entire life-even
though I know
perfectly well that the logical possibility of such a volume,
in and of itself, in no
way diminishes my freedom in bringing about the events described
within it.
9
There might be other reasons why I lack freedom, but the possibility
of such a
book is certainly not one of them. And on the other hand, some
may likewise
find it disturbing that a complete materialist theory of consciousness
could in
principle expose, for all to see, the underlying physical foundation
of every sinFRONTAL ASSAULT 13
gle pain or taste or visual impression (or hope or lust or intuition)
throughout a
subject's entire life. This might seem disturbing even though
such a prospect
would in no way diminish the painfulness of the pain or the lustfulness
of the
lust; that rich phenomenal experience might have a physical explanation
would
render it no less rich phenomenal experience.
A usually unarticulated further worry might grow from the notion
that any-thing
which admits of a physical explanation is no different in principle
from
any other physical thing. The possibility of a book explaining
my entire conscious
existence in the language of physics threatens to reduce that
experience
itself to nothing more than mere physics, siphoning off its value
down to the
level of some least common denominator appropriate for other
physical things
like bricks and globules of sludge. Probably it is only natural
to feel some discomfort
at the idea that the very sciences which our own ingenuity created
could
drain away our fundamental value in this way. On this view, perhaps
our value
can only be preserved by finding it a nonphysical refuge categorically
separate
from (and impossible to unify with) the merely mechanical transactions
of
bricks and balls of sludge.
Needless to say, these are worries I do not share. On the contrary,
I feel that
such a naturalistic physical explanation could only add to the
sheer marvellousness
of phenomenal experience. There can be little doubt that it would
be amazing
if it turned out instead that the subject of my phenomenal experience
was
really an independent, immaterial, ethereal spirit of some sort-that
I was such a
spirit and my consciousness one of that spirit's properties.
But how much more
truly amazing it would be to discover that the conscious 'I'
is instantiated purely
physically-that somehow, despite having nothing but that stringy
ball of nerve
fibres and other organic matter with which to do it, I still
manage to enjoy my
full remarkable repertoire of rich conscious experience! That
matter simply organised
in the right pattern and changing in the right ways could actually
instan-tiate
me, with all my vivid phenomenal experience intact: that is a
marvel worthy
of the name. And, indeed, if what really counts is its in enabling
our conscious
lives, why should a pattern be of any less value than an immaterial
spirit?
In the next nine chapters, I set out what I believe are some
useful steps toward
understanding the sorts of links between mind and matter which
might
make such a view attractive in its own right and which may allay
Osmo-style
discomfort. Do I believe I have given a rigorous, complete, and
definitive solution
to the mind/body problem(s)? Of course not! But I do think a
'solution'-
or, rather, a set of solutions to a cluster of related problems-will
eventually be
found to share the general form, and perhaps some of the details,
of what I outline
here. Many times in these pages I deliberately reach far out
on a limb while
constructing some view or other. In so doing, I aim to lay out
a sort of tree of
possibilities; not all of its branches will ultimately bear fruit,
I am sure. But I'm
14 CHAPTER ONE
convinced that while some limbs will eventually need cutting
out and others will
benefit from considerable reshaping, the central thematic trunk
is healthy and
planted just about where it ought to be. I hope that laying out
the tree as I have
will provide new opportunities for progress through the process
of evaluating
and snipping back those bits which don't belong, strengthening
those which do,
and shaping this nascent theory into something robust and comprehensive.
NOTES
1 Nowhere do I argue positively for any particular rendition
of material monism; I find it hard
to grasp what a good argument for material monism would even
look like.
2 Philosophers, understandably concerned to be clear about what
it is they're trying to explain
before setting out to explain it, frequently adopt the convenient
assumption that our first hand,
direct experience of consciousness suffices to fix the concept
appropriately well. But to paraphrase
an example due to Aaron Sloman, the assumption is as unjustified
as the claim that
even before Einstein's analysis of the concept of simultaneity,
we really all knew what it was
anyway just through our first hand experience. As for the case
of simultaneity, Sloman suggests,
it may be that only after constructing some good theories, capable
of supporting coherent
concepts, will we be able to grasp properly what it was we were
trying to explain in the
first place!
3 Since endnotes are easily found at the end of each chapter,
when cross-referencing them I usually
mention the page number for the discussion which is endnoted
rather than giving the page
number for the endnote text itself.
4 Not all the excitement is hype. While I am dubious about his
tendency to overstate the ontological
significance of what amount to different dynamical (or non-dynamical)
descriptions of
physical entities-see section 3 of Chapter 9-Tim van Gelder (1995a,
b) offers a sober but
provocative analysis of each approach. In a forthcoming target
article for Behavioral and Brain
Sciences, he succinctly maps out distinctions between the computational
hypothesis-
roughly, the view that cognizers are digital computers-and the
dynamical hypothesis-
roughly, the view that cognizers are dynamical systems.
5 Very often it is expedient to import science into philosophical
discourse-nothing clinches
an existence proof like empirical evidence, for instance!-but
some of the really difficult problems
of philosophy might well be those which either cannot be naturalised
or which would be
very difficult to naturalise. (Ethics comes to mind as a possible
example.) I have the utmost
respect for these sorts of areas, but my temperament generally
tempts me more in the direction
of those where some 'easier' progress can be made.
6 Although I usually refer to 'levels', I mean also to include
parallel or otherwise complementary
descriptions which may not exist in an hierarchical relationship
with one another.
7 In the language of Chapter 6, everything supervenes on microphysics;
see section 1 of that
chapter in particular.
8 On the last especially, also compare the emergentists such
as Alexander (1920), Broad
(1925), or Pepper (1926); see McLaughlin (1992) for recent discussion.
9 Specifically, the purported argument to the contrary requires
shifting the scope of a necessity
operator from a whole conditional to just the consequent of the
conditional. But although it is
necessarily true that if such a book correctly proclaims that
I will perform action x, then I will
perform action x, from this it simply does not follow that such
a correct proclamation entails
that I necessarily will perform action x.
CHAPTER TWO
Zombies and Their
Look-Alikes
It is an embarrassment to our discipline that what is widely
regarded among philosophers
as a major theoretical controversy should come down to whether
or
notphilosophers' zombiesare possible/conceivable. (Dennett 1995b,
p. 325)
Zombies of analytic philosophy, unlike the voodoo victims of
Haitian folklore,
are hypothetical creatures entirely bereft of conscious experience
who nonethe-less
behave indistinguishably from the rest of us. Philosophers' zombies
walk
and talk as if they're conscious, they appear to wake up in the
morning, and
over breakfast they even speculate on the meaning of dreams they
claim to have
had. They don't realise they're zombies, of course-no feeling
of peculiarity
spoils the pristine emptiness of their barren phenomenological
landscapes-and
an enterprising philosopher engaging one in conversation about
the topic might
well hear all manner of insightful commentary about what the
notion of zombies
reveals about philosophy of mind.
Daniel Dennett says zombies don't exist. Most people, no doubt,
would
agree. But more importantly, he suggests they are logically impossible:
misguided
philosophers who claim they are conceivable merely fail to imagine
the
full repertoire of zombie behaviours. All too often, philosophers
define zombies
as above and then proceed to argue for some behavioural clue
or other which
would give them away. But of course there can't be any such clue.
"The philosophical
tradition of zombies would die overnight", Dennett says
(1995b, p.
325; Dennett 1995a is similar), "if philosophers ceased
to mis-imagine them". It
is unusual that I disagree significantly with Dennett, especially
in print
(Mulhauser 1997a). But I think his pessimistic prognosis can't
be quite right.
Dennett's challenge offers a tailor-made warm-up exercise for
the rest of this
book:
Show me, please, why the zombie hypothesis deserves to be taken
seriously, and I
will apologize handsomely for having ridiculed those who think
so If the philosophical
concept of zombies is so important, so useful, some philosopher
ought to
be able to say why in non-question begging terms. I'll be curious
to see if anybody
can mount such a defence, but I won't be holding my breath. (Dennett
1995b, p. 326)
Science
& Mathematics
The Uncle
Taz Library
Uncle
Taz Home Page 2
Site search Web search
powered by FreeFind