This fine summary by Denise of Churchland's book and the mind/body
problem had been posted a little early in the seminar, but now its
time has come, and here are some comments:
> From: "Baden, Denise" <DB193@psy.soton.ac.uk>
> Date: Wed, 15 Nov 1995 11:50:29 GMT
>
> Churchland discusses the nature of conscious intelligence and covers
> both the dualist and materialist points of view. He did not, however,
> answer [Stevan's] question of why consciousness is necessary. A
> question, by the way, which has wound up everyone I have put it to.
For the record, let's repeat the question here: Supposing you are
granted the truth, the completeness, and the necessity/sufficiency of
your favorite theory of brain and behaviour: It either explains exactly
how the brain does everything it can do (including everything WE can
do, and everything we think and feel and know whilst doing it) or it
explains how ANY device, any causal system, can do everything we can do
(etc.), with the brain being only one particular example of such a
device. Either way, the problem is this: What role could CONSCIOUSNESS
be playing in any of this? For if you look at how the brain or the
generalised device works, you will always have a sensible, causal
explanation of its capabilities in terms of its internal structures and
processes. Fine. So those internal structures and processes, whatever
they turn out to be, are doing the work. So what role is left for
consciousness? Why couldn't all that work be done completely
unconsciously, with nobody home, nothing feeling or thinking or knowing
anything?
In trying to answer that question, you will always find yourself up
against the kinds of problems Paul Churchland, amongst others, wrestles
with. They are all variants of one problem: the mind/body problem, which
is: What has the mental stuff got to do with all this brain/device,
structure/process, functional-capability business? What kind of stuff IS
the mental stuff, and what kind of causal role could it possibly be
playing?
Consciousness is undoubtedly real; Descartes saw rightly that it is
self-contradictory to suppose that you are not supposing, to doubt that
you are doubting, to deny that experiences are going on in your head
(though you can doubt you have a head, and you can even doubt that there
is a "you" -- in the sense of some continuously existing entity, for an
immediate experience you have, even if it SEEMS as if it has a long
history, your history, might be illusory, in the same sense that the
outside world might be an illusion: it could all be a big, false, "deja
vu" experience; Descartes could doubt that things were as they seemed,
but he could not doubt that "SEEMING" itself, that is, experiencing,
was going on). So experiences, appearances, are definitely going on,
though all the rest of the details of their contents, the details of
what they are about, are could be incorrect. Hence any theory that would deny
that we have experiences (= consciousness) would be denying the
undeniable, and hence wrong.
But the next step from there is anything but clear! Ok, so Descartes'
Cogito is right. There really ARE mental states, experiences,
consciousness (these are all synonyms): How are they related to the
rest of what there is, which is what physics tells us about? What sorts
of things are mental states in the physical world of matter and energy,
which is the ONLY world physics speaks of.
It is no wonder that people have resorted to "dualism," and even to
belief in the existence of an immaterial (and immortal?) "soul," given
the compellingness of this profound puzzle about what the mind could
possibly be in the world.
My question, however, turns this puzzle rather on its head: The usual
mind/body problem is: Given that there is undeniably a mind, what could
mind possibly be, physically speaking, if we are NOT to be dualists and
say it's some extra kind of "stuff" that defies physics (and would in
fact defeat physics if it were real, since the most fundamental of
physics's laws, the law of the conservation of matter/energy, which is
supported by every experiment ever performed, would be the first to
bite the dust if there were something ELSE besides matter/energy in the
world -- at least if that something else had any causal power at all,
for then things we did with our minds would violate the conservation
laws, coming in out of nowhere and producing physical effects!)?
So the usual ploy is to try to avoid dualism at all costs, by somehow
trying to show how/why mental states are really the same as (certain)
physical states. The trouble is, that when you try to do that -- when
you try to show that, say, pain, is really just the neurochemical
activity of a certain system in the brain -- you invariably have a
causal story in which pain itself -- the FEELING of pain, of course --
seems to play no independent role, and instead has a
take-it-or-leave-it quality: If my neurochemical story about pain is
correct and complete, then there's no harm in saying that that's what
pain is, but then, wouldn't the story run just as well without the
pain? And is it obvious what, other than faith, makes pain the same
thing as whatever that neurochemical system turns out to be?
This "peekaboo" relation between physical/functional explanations and the
mental is the USUAL mind/body problem: "How can the mental be just that
stuff, and why should I believe it?" My version (the one Denise alludes
to above) is: Given all that neurochemical/functional
stuff, and the remarkable things it can do (the pain system detects
injuries, favors injured parts, learns to avoid such injuries in the
future, etc.), what possible ADVANTAGE, what possible functional role,
could the mental be contributing to it? If it is an independent causal
role, then we are up against the laws of physics. But even if it just
dangles, without causing anything, as an EFFECT of the neurochemical
system, without any further independent effects of its own, what's the
POINT of it? What's the FEELING of pain for? And if for nothing, then
why is it there at all? For no matter what physical/functional system
you describe, no matter what its capabilities, you will always have a
choice between two ways of generating those capabilities: Dualistically,
by claiming there are mental properties with independent causal powers
-- but then your system, being at odds with physics, will not be very
believable -- or by generating all of its capabilities physically, and
merely stipulating that the very same system also happens to generate
the mental ones. But then one must ask: Why would it bother to do the
latter? since the job can clearly be done without them?
So your assignment, if you wish to resolve this puzzle (I don't think
it can be resolved), is to describe a possible physical/functional
system of which it CANNOT be said that it would have EXACTLY the same
physical/functional capabilities whether or not it had a mind. (And you
must do this without going dualistic, because otherwise you have to
answer to the conservation legal authorities of physics!)
> 1. Dualism: Claims that the nature of conscious intelligence is
> non-physical, associated with religious ideas and Descartes. Problem of
> how a non physical-substance can interact with matter.
The religion is a bit of a red herring: Yes, the special, puzzling
nature of consciousness is what gives rise to the mind/body problem, and
many have resolved it by becoming florid, unrepentant dualists, positing
not just the power of the soul and its free will in the world, but even
another world (or many), an afterlife (or afterlives), and a Creator (or
many). But the right way to put it is that the problem of consciousness
caused, among other things, religious ideas, not the other way round:
Religion did not cause the puzzle of consciousness; and the puzzle is as
real for an atheist as for a believer.
> Popular Dualism: A person is a 'ghost in a machine' - mind is in
> contact with brain and can exchange energy. But is no evidence for a
> thinking substance.
As I've said, immaterial stuff interacting in any causal way with
material stuff (matter/energy) is completely incompatible with
everything we know about physics. And we know a lot about physics, and
there's an awful lot of physics in the universe. In contrast, there is
very little mind in the universe, it's mostly possessed by a minute
biomass on our own minute planet, and the little we know about THAT
(through biology, biophysics and biochemistry) unequivocally suggests
that it's made of the same stuff, and obeys the same laws, as all other
physical things; it does have a few special properties peculiar to it,
but even those are just physical properties and give no indication of
the kinds of radical properties that the CONTENTS of our experiences
suggest we have: there seems to be a basis neither for free will nor for
any independent causal role, nor even, for that matter, for the EXISTENCE
of anything other than matter/energy in the universe, including its
biological bits -- apart from the undeniable Cartesian reality of the
existence of our experience (and what its contents seem to tell us)!
In a word, some kind of cavalier hand waving about "interaction"
between brain and mind won't do...
> Property Dualism: There is no substance, but brain has a special set
> of properties associated with it which are non-physical eg feeling a
> pain, sensation of red etc. These cannot be reduced or explained by the
> physical sciences as they are purely mental.
True, perhaps, but what is this saying, really, other than restating the
puzzle that, of all the things in the universe that are present, all but
one kind are accounted for...?
> Epiphenomenalism: Mental phenomena are side effects of brain
> complexity. They are caused by brain activity, but have no causal
> effects on brain.
The mind/body problem sooner or later forces you to have to swallow
something unpalatable; or to bite a bullet, if you prefer that image. I
happen to find biting the bullet of epiphenomenalism less unpalatable
than any of the alternatives, most of which amount to saying either:
(1) "Consciousness just IS this [or that, substitute the neural or
functional or computational candidate of your choice here]; trust me!" or
(2) "Consciousness plays the following necessary functional role [trust
me that it doesn't tamper with physics, nor is it merely the same as
(1), which would of course mean it was not necessary after all...].
In this sad set of choices, epiphenomenalism -- which simple says "Seeming
is indeed there, as it undeniably is, but it does not play the causal
role it seems to play, it just IS" -- seems to me the softest bullet to
bite upon...
> So it is an illusion that our beliefs & desires
> determine our actions. (I think this argument, radical though it may
> be, is supported by a study by Nisbett & Wilson, which showed that the
> explanations offered for one's own behaviour often have little or no
> origin in introspection, despite sincere beliefs to that effect, but
> are instead spontaneously confabulated on the spot as explanatory
> hypotheses to fit the behaviour & circumstances observed. These are
> often wrong, since the 'introspective' reports given prove to be a
> function of wholly external features of the experimental situation
> which were under the control of the experimenters.)
Yes, in fact most of cognitive psychology (and the difficulty we have in
coming up with a successful causal explanation of how we can do most of
what we do) is testimony to the fact that the true causes of our
thoughts and actions are not available to our introspections; but the
same would be true of even the elementary (noncognitive? or minimally
cognitive) act of deciding to press this button NOW rather than NOW:
"I" seem to be the "originator" of that decision, but Libet's evidence
from the readiness potential suggests we're wrong about that too!
Libet, B. (1985) Unconscious cerebral initiative and the role of
conscious will in voluntary action. Behavioral and Brain Sciences
8: 529-566.
> Main arguments for dualism:
>
> 1. Religious belief, or belief in a soul
Belief is not an argument. Better to see dualism, or rather, the
mind/body problem as an argument for religion!
> 2. Introspection - we feel that we have mental states
The Cartesian evidence is indisputable (on pain of self-contradiction: "I
doubt that I am doubting"...).
> 3. Argument from irreducibility- can we reduce fragrance of a rose to
> the physical?
Weak stuff: The problem is not the variegated fragrances of a rose; the
problem would be just as real and just as big if there were just ONE
experience in the world, say, a toothache that measures 7 on the "Ouch"
scale. Just one simple experience, relative to as much physical
complexity as you like, and it's still "irreducible"...
> 4. Parapsychological phenomena - but not much evidence
Right; the "evidence" for parapsychology is pathetic. It's mostly errors
and deceptions; when not that, it's coincidences. If the biomass is tiny
relative to the universe, the evidence for psi is submicroscopic...
On the other hand, you don't need mind over matter in the form of psi.
The very act of pressing the button NOW rather that NOW would be just a
radical if I were right that "I" decided it, through the mental force of
my will, rather than my brain "deciding" FOR me, through the force f
ordinary physical/chemical interactions. For if it really was my mind,
independently of my brain, than button-pressing with one's own hands
would be as remarkable, and as psychic, as spoon-bending at a
distance...
> Main arguments against dualism:
>
> 1. Religious arguments lack empirical evidence
Of course not, but it's also a bit of a trick to try to claim that the
problem of dualism is a problem about religious beliefs, because the
mind/body problem is just as much of a problem if one completely rejects
the supernatural.
> 2. False to believe that introspection reveals things as they really are
But not false to believe that introspection reveals that mental states
really exist. (There would be no other way to know it! And that's
another dimension of the mind/body problem, known as the other-minds
problem: The only mental states you can know about are your own; those
of others must be assumed to exist on faith...)
> 3. Many abilities thought to be mental are duplicated by computers
Many abilities thought to require mental states can be duplicated by
systems that have no mental states. But the real mystery is not:
How could you possibly do that without consciousness? but rather "What
could consciousness possibly add to the physical/computational basis
for the capacity to do that that it does not already have?
> 4. The effect of chemicals, drugs, lesions etc on the mind suggests
> that consciousness is very much affected by physical states.
Since most dualists are interactionists, this is irrelevant (except, of
course, that interactionism itself leads to the problems with
conservation laws I already mentioned).
> 5. Evolutionary argument that the human species & all of it's
> features are the physical outcome of a physical process.
Yes, yes, but none of this helps: The problem is untouched. You can have
all your smart computers, and the fact that all organisms were shaped by
evolution, and that drugs affect the brain, and that introspection
doesn't reveal how the brain works, and that there is no supernatural --
and you're still face to face with the mind/body problem, with SOME
form of dualism the only thing that makes sense (with epiphenomenalism
its mildest form).
> 6. No need to postulate 2 substances - physical and mental/spiritual,
> when only physical properties are necessary to explain phenomena.
Here's the sleight of hand: Physical properties, until further notice,
are only explaining physical phenomena. The fact that, say, drugs
(or, for that matter, all stimuli) AFFECT mental states does not show
they ARE mental states...
> 2. Philosophical behaviourism: claims that talk of ghostly inner states
> like beliefs and desires is a shorthand way of talking about actual &
> potential patterns of behaviour. But sensations do have an intrinsic
> qualititive nature.
No quarrel here. But the hope of the behaviourists to solve the
mind/body problem by persuading us that mental states are really just
potential behavioural patterns -- a hope that never had much persuasive
power for me, since it's clear that, whatever it is, my toothache is NOT
just a "potential behavioural pattern" -- is nevertheless not unlike
Paul Churchland's hope of persuading us that mental states are really
just brain patterns! Neither form of "shorthand" quite does the
trick...
> 3. Reductive Materialism/Identity Theory: Each type of mental state is
> identical with some physical state or process within the brain.
But, despite his disavowal, Churchland's own views are a form of
reductive + eliminativist theory. He somehow tried to finesse the
reduction/elimination by calling it "co-evolution": You start out with
some naive idea about what the mental is, perhaps even dualism, but you
do experiments and you follow theory, and you learn more and more about
the brain and behaviour, and it's not that you reduce what you thought
the mental was to what neurobehavioural science tells you it is, nor do
you eliminate what you had thought it was, to replace it by what
science says, but the two views -- your original amateur one, and
science's -- "co-evolve" and eventually converge on something that no
longer has a mind/body problem and no longer requires or risks dualism.
I think this is just hand-waving. Of course people can be talked into or
out of anything, including the mind/body problem. But we must
distinguish what politicians and lawyers can do to us from what
scientists do to us -- and even if you find all three of them pretty
much of a muchness, one can still ask: What sort of thing, after
all, is TRUE about the mind and the brain? Never mind how my notions
might "co-evolve" under the persuasive power of the latest ooh-aah facts
about the brain, facts that, being about MY brain, I am bound to be
over-impressed by, and perhaps even see as explaining what they don't
really explain, or solving problems they don't really solve. How does
one -- whether by reduction, elimination or co-evolution -- get a handle on
the relation between the stuff the world's made of and the stuff our
thoughts and feelings are made of?
See: ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad95.churchland.bookrev
> 4. Functionalism: Thinks it unlikely that a 1:1 match will be found
> between our current psychological terminology and the physical
> processes revealed by neuroscience. Believes the defining feature of a
> mental state is the set of causal relations it bears to 1. environmental
> effects on the body, 2. other types of mental states, 3. bodily
> behaviour.
This is a more complicated issue. Functionalism's candidate solution to
the mind/body problem is that we were looking in the wrong place: It's
no wonder we could not equate the mental with the physical, because the
physical details don't matter, and could have been very different, yet
still have been the physical embodiment of the same mental state. What
matters is the FUNCTIONAL relations, not the physical specifics.
Functionalism comes in several flavours, but the only really explicit
one is computationalism, according to which mental states are just
physical implementations of the right computational states. To put it
more simply: Find the right computer programme, the one that "captures"
a mental state, and run it on ANY computer, no matter what the computer
is like physically, and it will be the embodiment of that same mental
state; it will HAVE that mental state.
Computationalism is, alas, wrong, and Searle's Chinese Argument shows
how (and the Symbol Grounding Problem shows why); See:
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad89.searle
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad90.sgproblem
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad93.symb.anal.net.searle
Churchland is not a computationalist, in fact he rejects all forms of
functionalism, even the robotic functionalism that I would say has the
best chance of at least explaining what we can DO (if not what we can
feel). He rejects it in favour of the kind of hybrid neuro-computational
theory (neural nets, really) that he thinks will make it no longer
necessary to worry about the mind/body problem...
> 5. Eliminative materialism: Believe our current psychological framework
> is a misleading conception of the causes of human behaviour and will be
> overthrown when a more accurate framework emerges from the
> neurosciences (I can see, for example, 'black box' terminology being
> overthrown by connectionist, or neural network models). This view
> challenges entrenched assumptions about our view of mental states such
> as desire, fear, purpose etc.
As I said, Churchland himself is probably closest to being an
eliminativist himself, though he is as adept at not quite saying so as
he is in persuading us that the co-evolutionary approach will somehow
settle the mind/body problem -- without quite saying how...
> That covers the main approaches and theories. The rest of the book
> discusses the problem of other minds, i.e. how can we know we
> experience the same things. Churchland also talks about self
> consciousness, and the likely contributions of future research in
> artificial intelligence and neuroscience to the debate.
The other-minds problem is worth considering, as it is the flip side of
the mind/body problem. See:
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad91.otherminds
For my reaction to Churchland's latest views on consciousness, AI and
neuroscience, see:
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad95.churchland.bookrev
and
ftp://cogsci.soton.ac.uk/pub/harnad/Harnad/harnad93.symb.anal.net
Chrs, Stevan
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:57 GMT