172 J.J. Haldane
For example, in response to my claim that naturalism cannot account
for conceptual thought, Smart seeks to resolve the puzzle of intentionality
by appealing to Quine’s proposal that reports of mental acts be treated as
describing attitudes of thinkers to sentences and predicates. Thus ‘Joe wants
a unicorn’ is to be rendered for philosophical purposes as ‘Joe wants-true of
himself “possesses a unicorn” ’ (chapter 3, p. 157). And in reply to my argu-
ment that the meaning of a general term such as ‘cat’ cannot be identified
with its extension (the set of things, cats, to which it applies) – because, for
example, claims about the prospects for cats in the future may be meaning-
ful though they do not concern actually existing animals – Smart contends
that ‘cat’ refers to the setof cats, past, present and future. To which he adds
‘I believe in future cats (they are up ahead of us in space–time)’ (chapter 3,
p. 157).
Our difference over the nature of intentional states clearly involves com-
peting philosophical theories, and even Smart’s last, seemingly straightfor-
ward, reply has a complex of abstract metaphysical theses lying behind it. The
primary issue between us is not so much about what it is that ‘cat’ refers to, as
about how it is possible for a general term to have ‘sense’ and thereby to refer
at all. On my neo-Aristotelian account someone who is competent in the use
of the term ‘cat’, or has the corresponding concept, has intellectual possession
of an abstract ‘intensional’ entity, a formal structure which is also possessed by
cats (exemplifying this structure naturally or materially is what makes them to
be cats).^1 Informed by the concept one can then raise questions about types of
circumstances which do not obtain but in which cats might exist. (Likewise
someone who has the concept of a unicorn possesses a thought-structuring
principle which gives ‘shape’ to his or her thoughts notwithstanding that
there are no unicorns outside the imagination.) Smart’s rejoinder requires that
reference be accountable for in general without invoking abstract senses, and
that in particular we accept a theory of the material universe as, in effect, a
four-dimensional object some temporal parts of which feature stretches of cat.
Here I am not concerned to dispute these ideas, though I regard them
as untenable. I only wish to alert readers to the fact that they are thoroughly
metaphysical, quite revisionary of ordinary ways of thinking, and far from
being obviously true. The same points hold good for Smart’s Quinean-
inspired discussion of modality, and for his suggestion that the notion of
causation is not metaphysically robust and may be eliminable. Similarly his re-
peated attempts to use Ockham’s razor to excise non-physical entities presume
a background against which non-naturalist, and more specifically theistic
hypotheses stand out as ontologically extravagant. Notice also that contrary
to its being presented in a matter-of-fact, commonsensical fashion, the back-
ground in question is a distinctly theoretical one. Readers may have assumed
that a physicalist would have nothing to do with abstract entities, but Smart’s