Dear Derek,
The Philpapers organisers' aim is for a professional
philosophical forum with emphasis on discussion of current literature. My
transitional status aside I hope we are in line with that and I take your basic
point - that the 'P without Q' zombie argument against physicalism has problems
with premises - as an important philosophical issue worth pursuing. (It would
be good to get input from David Chalmers himself.) Nevertheless, I am worried
that several arguments in above posts go off at a tangent from David's
reasoning and slide back into folk psychological issues irrelevant to the
zombie posit. My understanding is that the point of both philosophy and science
is to identify inconsistencies, contradictions, or other inadequacies in folk
accounts of the world and replace them with workable reasoned alternatives. I
sometimes worry that current professional philosophy seems the opposite - a
last bastion of intuition jealously guarded against reason! But let me get to
the point.
The zombie posit is of an entity based on exactly the same
instantiated laws of physics as a human, that behaves the way that we do in
all respects but has
no conscious experience. By definition we are considering something that
will never fail a Turing test. My point about building robots that can pass
Turing tests is not a point about building zombies but one of several arguments
against the folk idea that human behaviour needs some magic ingredient called
'intention' that requires consciousness in some way. I will return to that but
it is probably best first to address your very first query about what it means
to have no consciousness.
David uses consciousness to mean that X is conscious if 'it
is like something to be X' in Nagel's sense. I agree that although most people
see what is intended, this phrase is pretty oblique. One could say that it is
no help unless we can say what it is
like. 'Dark inside' is also misconstruable. Both phrases are awkward metaphors.
But this may be unavoidable because natural language has not yet moved on from
a folk description format to a rigorous ontological one. We do not have the
language structures we need. However, I think there may be an alternative to
the Nagel phrase that is more robust:
X is conscious if (at least a limited internal part of) the
world is like something to X.
This implies that the world in some way provides data for X
and that these data have distinguishing qualities of appearance to X. X is
acquainted with data in an overt way. Seeing red and blue is all we need to posit.
For me, this meaning of consciousness is fully adequate to start out on the
zombie argument. Moreover, most people get the idea. If you do not you might be
a zombie. You may be a philosopher with your own precepts about consciousness
in to which David Chalmers's conception does not fit (very likely). You might
also be conscious but you may not be able to access the role your consciousness
plays in the way you construct arguments. This is an interesting possibility
partly because it is the opposite of your intuition and partly because it is
probably true for all of us in one sense. One of the most difficult problems
about consciousness in David's sense is how we can possibly talk about our
acquaintance with data rather than just the data themselves. My own view would
take too long to explain here but I conclude that although there is a mechanism
we can never take the accounts of others about their consciousness as evidence
of their consciousness. The same behaviour can always be simulated without.
(And I think there is a real possibility that people differ in their access to
the link between experience and behaviour but again the detail is too long.)
This comes back to your point that we cannot indeed ever
know that a zombie is a zombie. But that is not the point of the zombie posit,
which is a conjecture about ontology that ventures beyond positivist
constraints. In passing, I think the issue about how long a robot (not a
zombie) could keep passing the Turing test is a red herring. Suspension of
disbelief in cartoon films indicates that we never make a final rational
decision on whether or not something is conscious. We might have doubts about
the robot but we would constantly ask 'is it really conscious after all?'. The
reason is that our decision about what is conscious can never be rational
anyway because we have no conception of how 'overt acquaintance with data' in
the way we are familiar with, as opposed to just 'physical signal passing'
could constrain behaviour (and as I say it probably does not). Moreover,
passing Turing tests at a complex level relating to jokes and unrequited love
etc. does not need to be invoked here. All we need talk about is the
consciousness of the black on white of print. Computers decipher print but do
not have our 'awareness of black'.
Having established that David's definition of a zombie is clear, my own
problems relate to the meanings of 'conceivable' and 'physical'. Conceivability
is a dangerous plaything. We are questioning our most cherished beliefs about
ourselves so we need to question the reliability of our conceivings and there
are plenty of reasons to be sceptical. The zombie is conceivable in the very
weak and loose sense that if we do not know the nature of something any
possible nature is conceivable, (we do not know the final rules that link known
physical laws and experience within anybody). But my earlier post made the
point that if we ask what we really mean by physical laws we find a clear
contradiction because we have posited, by default, that the links are the same
as in our conscious selves. (I do not think that inconceivability on the basis
of an intuitive sense of a need for 'intentions' comes in to it.)
The real problem as I see it is the idea that physics is
'closed' or 'complete'. Dynamical physical laws form a closed chain of
intermediary links within any causal description but are so far always
incomplete when it comes to the final cashing out in experience - as Newton
pointed out. There are extra laws of correspondence between dynamics and experience
implicit in physics that we have yet to get a grip on. The solution to
Jackson's Mary story is obvious. There is more to know about what determines
experiences. Maybe I actually agree with David Chalmers in that physicalism is
false but the physicalism falsified is a partial materialist physicalism that
within physics itself is more or less a straw man. We know we need laws beyond
current physical laws and maybe we could call these some other sort of law.
However, since 'physical law' only means a law instances of operation of which
determines experience, I am not sure why we should want to make the
distinction. I suspect that people usually want to make the distinction because
of a residual folk sense of 'physical' as 'stuff'.
Best wishes
Jo E