Refuting the Identity Theory
Suppose someone were to make an outrageous identity claim—say, that Donald Trump is identical to Barack Obama. It would be easy to refute that by pointing out that the two men are to be found in different places, so cannot be the same man—following the principle (beloved of detectives) that one person can’t be in two places at the same time. It is not so easy to refute it by asserting that Obama and Trump look different, sound different, and have a different educational history, because that leaves it open for the identity claimer to say that one person can disguise himself and appear as if two people (Superman and Clark Kent, for example). One person can have two modes of presentation but not two locations. The same point applies to two events: if someone claims that the explosion was the same event as the fireworks display, you can refute this by pointing out that the events occurred two miles apart—though it isn’t so easy to argue that the appearances were of different events. Such identity claims are vulnerable to what we might call the “location argument” (as opposed to the “appearance argument”). They are thus falsifiable by recourse to location. However, in the case of the mind-brain identity theory this recourse is not available, because the mental event has no identifiable location: you can’t point to it in space and observe that its location differs from that of its correlate in the brain. If someone says C-fibers are identical to D-fibers, you can point to the D-fibers and observe that they are not where the C-fibers are, thus refuting the identity claim. But you can’t do that with pain: it has no identifiable location. So, the primary method for falsifying identity claims about empirical particulars is not available. You are left weakly protesting that pains don’t appear like states of the brain—to which the identity theorist will retort that one thing can have two appearances (though not two locations). This gives the identity theorist an unfair dialectical advantage: no matter how preposterous his claim is he cannot be refuted in the standard fashion. Even if the pain exists in a separate immaterial substance, you cannot demonstrate this by pointing to its different location. Not because it has the same location as the brain state but because it has no location, or none that can be pointed out. It is simply in the nature of the case that it lacks a distinct observable location, not because the identity theory is true. One suspects that identity theorists have got a good deal of mileage from that fact: they can complacently point out that no one has ever refuted their claim in the only way it can be, conclusively anyway. But this is nothing to their credit, since it follows equally from the dualist perspective: the pain could be as ontologically distinct from its neural correlate as you could wish and still not be capable of being located elsewhere. The location argument thus can’t be used against the identity theorist, but not because there is any truth in his position. Once this is recognized we can lean more heavily on such arguments as that pains and brain states seem very different, or that you can know all about the latter and not know about the former, or that in some possible worlds there are pains and no C-fibers. It shouldn’t worry you that you can’t falsify the identity claim in the canonical way. It should worry the identity theorist that he can’t locate pains in space at all, independently of the truth of his identity claim. The case is not like a single mountain seen from different perspectives.
 If it turned out (per impossibile) that your pains exist in the back of your closet, you would have a cast-iron proof that they cannot be brain states (no sign of your brain in there); and then it would be perfectly clear why pains and brain states should have such different appearances—they are totally distinct entities. You might then wonder about the properties of your closet—does it have a peculiar aura of subjectivity about it?