I would like to discuss Papineau’s recent reply to the Swampman objection in ‘The Status of Teleosemantics or How to Stop Worrying about Swampman’ (Papineau, 2001, Australasian Journal of Philosophy).
As it is well known, the usual reply to the Swampman objection is simply to bit the bullet and accept that Swampeople do not have mental representations (Dretske, 1995, Millikan, 1993). Indeed, Papineau (1993) also defended this reply, but apparently a graduate student convinced him that he was wrong. As he tells the story, this student pointed out to him that ”if [Swampeople] have not mentality, as teleosemantics implies, then it would seem to follow, absurdly, that it would be right to kill Swampeople and eat them as meat”. According to Papineau, this conclusion is unacceptable. Even if we can accept (with some difficulties) that Swampeople lack representations, it is much harder to hold that we do not have any moral obligations towards them. So Teleosemantics is in trouble.
An obvious reply he considers is that killing Swampmen may be wrong because, even if they lack representational states, they can still be conscious. However, he dismisses this argument on the following grounds:
No doubt cows and pigs have some kind of conscious sentience, but to most people this does not make it wrong to kill them quickly and painlessly. Killing sentient beings is only clearly wrong when they also have complex enough minds to make plans, form relationships, engage in projects, and so on.
I have some worries about this reply. First, I am not sure the fact that most people think killing cows is ok undermines the fact that killing them is wrong. Furthermore, it is not obvious to me that it is only clearly wrong to kill beings with complex minds (see Peter Singer (1975)). On the other hand, Papineau does not seem to consider other ethical views like rule utilitarianism, according to which it might be morally wrong to kill beings that behave like us (that looks to me like a plausible moral rule since beings that behave like us very often are humans).
In any event, Papineau thinks that we should attribute representational states to Swampeople, because that is the only way of explaining our moral judgments concerning them. Of course, that puts some pressure on his view, since he also wants to stick at Teleosemantics. In order to resolve this tension, he tries to show that both claims are compatible by denying that Teleosemantics is an identity theory. According to him, Teleosemantics only tells us what fills in the ‘belief’ and ‘desire’ role in the actual world:
Swampeople only follows if this essential core is conjoined with the claim that “belief” and “desire” are rigid designators of those states. (…) But it is equally consistent with the central core of Teleosemantics to hold that belief and desire are not rigid designators, and that Swampeople do have beliefs and desires, on the grounds that in the context of Swampassumptions these psychological terms do not refer to selectional states after all, but to states that would then be present in Swamppeople (p.13)
So according to Papineau, ‘belief’ and ‘desire’ designate some functional roles that in the actual world are realized by states with certain selectional stories. ‘Beliefs’ and ‘desires’, then, differ from terms like ‘water’ and other rigid designators, in the sense that they may pick up different things in different worlds.
Certainly, this is a non-standard view of Teleosemantics. But what is the problem of adopting Papineau’s view? One thing we might say is that it is not clear what the alleged ‘belief’ role is. If we cannot appeal to selective story, how can we determine this role? Consider my belief Apples are tasty. Certainly, we partially individuate beliefs by content; so, if Papineau is right, there must be a non-selectional story that explains why my belief Apples are tasty has the content it has. However, I take it what we learned from classical functionalism is that this story is not likely to come.
Further, suppose we get a purely functionalist story about the belief role, say, by appealing to functional role semantics. Then, the normal thing to do would be to identify this role with the belief across worlds. Once we had a way of spelling out what belief states are (and how they acquire their content!) that does not appeal to selection, I do not see why we would need Teleosemantics at all.
Well, maybe that is too harsh. Certainly, we attribute mental states to other people, and in so doing we do not seem to be taking into account the selectional story of the species. Maybe what Papineau means by ‘belief role’ is whatever allows us to attribute mental states to other people. Let us grant that for the sake of the argument. A second concern is how to understand the idea that a selectional story can be the realizer of such a role. It seems that if ‘belief’ designates some kind of functional role, the most plausible realizer would just be some mental state (say, C-fibers firing). Once we have the actual mental state that realizes the ‘belief-role’ in current thinkers, I do not see why we need to appeal to selection at all. Again, Teleosemantics seems to play no role in the theory.
Finally, let me consder a third possibilty. Suppose we accept we can identify a ‘belief role’ independently of selection, and that the realizers in the actual world are something like ‘sate that has been selected for by natural selection (or maybe: state produced by a mechanism selected for by natural selection). Still, it seems to me that if the role of Teleosemantics were to explain what fills the belief role in the actual world, it would lose much of its interest. In that case Teleosemantics would not tell us anything about what beliefs essentially are.
In short, I think Papineau’s view of Teleosemantics is insufficiently motivated, since there are other plausible ways of explaining why we feel obliged to behave morally towards Swampeople. Furthermore, I think the resultant interpretation of Teleosemantics would make these theories completely uninteresting from a philosophical point of view.
Naturalizing Stuff 