teleo-semanticist would be conceding that determinate functions – and so
determinate contents – only appear as one moves up the phylogenetic
scale, to include organisms capable of increasingly complex and sophis-
ticated inferences. This seems intuitively quite plausible.
3.4 The Swampman objectionPerhaps the strongest objection to teleo-semantics is that it entails that
creatures (and/or parts of creatures) which havenotevolved (and so whose
properties and parts fail to haveproper functions) must lack any intentional
states. Davidson’s (1987) example ofSwampman, who is accidentally
conWgured out of an old tree stump in a swamp by a bolt of lightning, in
such a way as to be molecule-for-molecule identical to an actual living
person, is often used to make the point. Swampman acts and responds to
questions just as the normal person does, since his accidentally conWgured
brain has provided him with a complete set of ‘beliefs’, ‘goals’, and
‘memories’. But according to the teleo-semanticist, Swampman in fact
lacks any beliefs, or any other states with intentional content, because,
having come to exist by accident, he lacks states with proper functions.
This is highly counterintuitive. Would we not be strongly inclined to say
that Swampmanbelieves(wrongly) that he is 44 years old, and that he
desires(embarrassingly) to move in with someone else’s wife, for example?
But according to the teleo-semanticist we cannot say these things. Rather,
it is onlyas ifSwampman has such beliefs and desires. In fact he has none,
because none of his states have proper functions.
Papineau (1987) faces this objection, and replies that on this matter our
intuitions need to be reformed. For he points out that he, and teleo-
semanticists generally, are not in the business of oVering aconceptual
analysisof the notion of content. Rather, what is being oVered is a
naturalistic theory of thenatureof content. And it is perfectly possible that
our concepts may be poor reXections of, or have gotten out of line with,
reality. So ourconceptof content would lead us to say that Swampman has
thoughts with content, but thenatureof content is such that, in fact, he
does not. But this reply, although acceptable so far as it goes, does not
address the real problem. For the diYculty for teleo-semantics is not just
that we feel pre-theoreticallyinclinedto say that Swampman has beliefs
and desires. Rather, it is that Swampman’s behaviour is subsumed by just
the same psychological (content-involving) laws as may be used to explain
and predict the behaviour of normal people. Not onlycanwe explain
Swampman’s behaviour by attributing to him beliefs and desires, but so we
should, it seems to us; for it is beliefs and desires which do cause his
behaviour, just as they do other people’s. In which case beliefs and desires
Teleo-semantics 173