set of conditionals about what the subjectwouldthink or do in the presence
of (some but not all) other – hypothetical – beliefs and desires. (This point
goes a considerable distance towards answering Fodor’s charge ofidiosyn-
crasyof content; for the same conditionals can be true of diVerent thinkers,
despite wide variations in their actual beliefs. See Carruthers, 1996c, ch.4.)
A second argument for functional-role semantics is a kind of ‘what else?’
argument from those working on the semantics of beliefs and statements.
It is generally maintained that we cannot individuate contents (explana-
tory as opposed to semantic ones, that is – see chapter 6) purely in terms of
reference, since this will slice themtoo thick. That is, it will fail to distin-
guish from one another thought-contents which are, intuitively, distinct.
For example, Oedipus’ belief, ‘Jocasta is more than forty years old’, diVers
in content from his belief, ‘Mother is more than forty years old’, even
though Jocasta is, in fact, his mother. For, not knowing that Jocasta is his
mother, it would be possible for him to have the one belief without the
other. Yet if the reference is the same, whatelsecan distinguish these
beliefsexcepttheir diVerences of causal role? For example, Oedipus’ belief
that he is thirty years old will tend to cause him to have the second of the
above beliefs in the presence only of the belief that a girl of ten cannot have
a child, whereas theWrst of the above beliefs will only be caused if healso
believes that Jocasta is his mother.
It is worth also considering a slightly more elaborate example, lest
someone should respond (as Fodor does, 1994, ch.2) by appeal to com-
positionality, pointing out that Jocastaandmy mothercan be distin-
guished by virtue of the fact that the latter contains the conceptmother,
whereas the former does not. The example is as follows. Imagine that I
have learned to use colour-terms normally, by sight, but have also learned
to use a hand-held machine whichin fact(unknown to me) responds to
colour – perhaps vibrating in the hand with an intensity proportional to
the dominant wavelengths it receives. And suppose that I have been
trained, ostensively, to use a set of terms, ‘der’, ‘wolley’, ‘neerg’, and so on;
where these terms arein factco-extensive with the colour-terms, ‘red’,
‘yellow’, ‘green’, and so on. Then, plainly, there are circumstances in which
I might believe that the tomato is red without also believing that it is der, or
vice versa. And this despite the fact that it is the very same worldly
property, surely, which my thought concerns. Here, as before, the ar-
gument is: whatelsecan distinguish the contents of the beliefs, ‘The
tomato is red’ and ‘The tomato is der’, except their functional roles? For
there is, by hypothesis, no overt compositional or syntactic diVerence
between them. What makes the two beliefs diVerent, in fact, are such
things as that, for example, (unless Ialsobelieve that anything der is red) it
is only the former which will cause me to point to the tomato if asked to
178 Content naturalised