demonstrated (for instance, by galvanic skin responses) to be taking place
at some level.
But no attempt is made at explainingwhya box located as the CAS is
located in cognition should contain states which are phenomenally con-
scious. Why could there not be a system whose function was to make its
contents available for executive decision and for reporting in speech, but
whose contents lackedfeel? In fact, this is one of those places where
Block’s (1995) distinction between phenomenal consciousness andaccess-
consciousness starts to bite. That there is some system which makes its
contentsaccessiblein various ways, does not in itself explain why those
contents should belike anythingfor their subjects to undergo. In fact, all
the proposed theories which we shall be concerned with attempt to explain
the subjective aspect of phenomenally conscious states in terms of the
distinctive sort ofcontentpossessed by such states – that is, all these
theories opt for the vertical branch at choice-point (2).
We shall now work through the four remaining choice-points inWgure
9.2, contrasting:Wrst-order with higher-order theories (3); higher-order
experience (or ‘inner sense’) theories with higher-order thought theories
(4); actualist as against dispositionalist forms of higher-order thought
theory (5); and forms of the latter which do or do not implicate natural
language (6). Our goal will be to convince you, by the end of the chapter, of
the merits of dispositionalist, non-language-involving, higher-order
thought theory.
3.1 FOR theoriesDretske (1995) and Tye (1995) independently develop very similarWrst-
order representationalist (FOR) theories of phenomenal consciousness. In
both cases the goal is to characterise all of the phenomenal – ‘felt’ –
properties of experience in terms of the analog (non-conceptual) represen-
tationalcontentsof experience. So the diVerence between an experience of
green and an experience of red will be explained as a diVerence in the
properties represented – reXective properties of surfaces, say – in each case.
And the diVerence between a pain and a tickle is similarly explained in
representational terms – the diVerence is said to reside in the diVerent
properties (diVerent kinds of disturbance) represented as located in par-
ticular regions of the subject’s own body. The main argument in support of
Wrst-order theories is that they can explain thetransparencyof conscious
experience, discussed in section 2.6 above. If some such theory is correct,
then it is obvious why, in trying to concentrate on my consciousex-
perience, I end up concentrating on the states which my experiencerep-
resents– it is because there is nothingmoreto a phenomenally conscious
Cognitivist theories 249