Mind-body identity is really a very big deal in contemporary philosophyof mind. Should it be? Do materialists and naturalists need to commit to the identity of mind and body? I don't think so. In the context of mind-body debates, the debate about identity turns out to be about possibility, and this should not be much of a concern for naturalists.
First, a word about evidence.
Suppose that I have evidence that strongly supports a proposition, p. Suppose further that q implies p, and that the only evidence that I have in support of q is that which supports p. Should I accept q?
- In certain cases, obviously no. For example, suppose that I have strong evidence that Peter (who happens to be a banker) is a thief. Suppose, further, that my only evidence that lots of bankers (including Peter) are thieves is my evidence concerning Peter. Here, I would clearly be wrong to believe the stronger proposition. I have no reason to think that Peter is a bellwether for the banking industry.
- In certain other cases, yes. Suppose that I have strong evidence that a certain food gave Mary a lot of pleasure. Suppose, further, that I have no evidence in favour of this food except what it did for Mary. Even so, I would be right to believe that this food would give lots of people (including Mary) lots of pleasure. I have no reason to think that Mary is unique.
The general principle goes something like this, put in terms of possible worlds:
If q implies p, q is true in a subset of p-worlds. If I have no reason to believe that q is more probable given p, than it is unconditionally, I cannot treat the p-favouring evidence as favouring q.
In light of this, consider the proposition that an empirical propostion p is metaphysically necessary. The evidence in favour of this modal proposition is evidence of p, and of what causes p. For example, consider the proposition that:
Necessarily, over-ripe soft fruit are mushy.
The evidence in favour of this proposition could only be that in the actual world over-ripeness causes mushiness in soft fruit. Such evidence supports counterfactuals like "If this fruit were over-ripe, it would be mushy." But should one, on this basis, accept that the connection is necessary? Presumably not. For the actual and nearby worlds are not indicative of all possible worlds.
Let’s call such a statement a strong modal. The only possible evidence for a strong modal is evidence regarding a small set of worlds, where the evidence regarding the subset is not projectible to the larger set in which the modal is assessed.
Identity statements in the philosophy of mind are strong modals. Consider whether pain is identical with C-fibre stimulations (CFSs). David Chalmers argues (strengthening an argument given by Kripke) that there could be creatures (“zombies”) that undergo CFSs but don’t experience pain (or anything else). This shows that CFSs are possibly not the same as pain. But if x is possibly not identical to y, then x is not identical with y. So pain is not identical with CFSs (or any other physical state, by parity of reasoning). (Chalmers' argument bears on the Hard Problem of Consciousness, and this is a separate issue.)
So here’s the situation: our evidence in favour of pain being identical with CFSs is restricted to the actual and nearby worlds. The evidence supports counterfactuals such as "If he weren't in pain, he wouldn't have a CFS," and these have implications concerning nearby possible worlds. But the truth conditions of the identity claim concern all possible worlds. If pain is dissociable from CFS in any possible world, no matter how distant, pain is not identical with CFS. Identity is a strong modal.
Proponents of naturalism should suspend belief about strong modals.
In light of the above, I am inclined to think that it is unwarranted to think that pain is identical with any material state. As a committed materialist, am I in trouble? I don’t think so. After all, I don’t, qua committed materialist, want to take a position on any strong modal truth. Materialism is surely empirical and scientific. It is about the actual world; it has nothing to do with other possible worlds, except nearby ones. Materialists ought therefore couch their doctrine in non-modal terms.
Put this in another way: what Chalmers and Kripke have shown us materialists is that we didn’t properly understand the status of identity. We thought we were making an empirical claim when we said that the mental is identical with the physical. But we were asserting something that went beyond the empirical. We have (let us say) strong evidence that in this world mental causation is wholly mediated by the physical, that in this world there is no non-physical causation. It is over-reaching to say that this entitles us to a conclusion about identity.
What then is the relationship between mental and physical states? I don't want to get into heavy-duty ontology here. So let me just offer a simple (even trite) observation. Psychology (including folk psychology) is an autonomous area of investigation. That mental states are investigated and known independently of neurophysiological states means that the conceptual relations between the two domains are indeterminate. This (plus some empirical propositions) is the reason why the psychological is thought to be multirealizable. Multirealizability, in turn, leads to the possiblility of non-identity, and thus to non-identity. We cannot both treat of psychology autonomously and have mind-body identity. Notice how this ties the identity question to conceptual autonomy . . . not to materialism.