158 J.J.C. Smart
These are such things as typical external causes and effects, waxings and
wanings, and positions in similarity spaces. This depends on the ability to per-
ceive bare similarities and differences between our inner goings on without
our being able to say in what respects these similarities consist.^16 My argu-
ment here does not rest entirely on the plausibility of physicalism. I can draw
on the elusiveness of so called ‘raw feels’ to which B.A. Farrell drew attention
in a fine article ‘Experience’ nearly half a century ago,^17 as well as the work of
the later Wittgenstein, Gilbert Ryle and others, and more recently fine books
by Robert Kirk and Austen Clark.^18
Nevertheless there does seem to be a strong tendency to believe that in
consciousness we are aware of radically ‘psychical’ properties unreconcilable
with materialism. David Armstrong has suggested that this tendency can be
understood by comparison with what happens in the headless woman illu-
sion.^19 A woman is seated on a brightly lit stage with a black background. She
has a black cloth over her head. The audience think that she has no head:
they confuse not seeing her head with seeing that she has no head. Similarly
we may be aware only of the neutral properties, and not being aware of them
as physical we think of them as non-physical.
We are familiar with times at which we go on ‘automatic pilot’. Some-
times, cycling to the university, I have realized that I have steered my cycle
on the bicycle path, crossed a busy road and avoided traffic, gone round
bends, and so on and yet I have no recollection of having done so. In a sense
I am conscious: I am not asleep or anaesthetized. I have reacted correctly
to stimuli. Still, in the full sense I have not been conscious. I have had
experiences, which I hold to be brain processes, but I have not been aware
of them. Armstrong has suggested that consciousness of my experiences
is a sort of direct monitoring by one part of my brain of other brain pro-
cesses that constitute sense experiences and the like. This monitoring would
certainly have survival value. Armstrong holds that it would be analogous
to proprioception. In proprioception we can be directly aware of such things
as the positions of our limbs. In the same way we can be directly aware,
in what I have called a ‘topic neutral’ way, of goings on in our brain that
constitute sensations and imagings. Consciousness comes out as awareness
and monitoring of awareness, and there can presumably be awareness of
awareness of awareness, though without finite minds this will not go very
far up the possible hierarchy. I like this suggestion of Armstrong’s that
consciousness is a sort of proprioception, not requiring neuronal receptors
external to the brain, but of the brain directly by itself.^20 The suggestion
implies that a robot constructed to monitor its own control system would
have consciousness as a sort of proprioception; this may not be wholly satis-
fying, but it is not clear what more is needed or if it is needed, how it should
be described.