Alas, I have little time to read actual philosophy and am now reduced to reading introductory anthologies in considering changing my textbook. Nonetheless, in one of the anthologies I am considering, there is an article by David Chalmers in which Chalmers argues that consciousness (which I will refer to as qualia for short) supervenes on the physical organization of the conscious organism. For example, my visual experience of blueness is determined by my brain's physical structure and organization, say a firing in a certain set of neurons in my visual cortex. (Let's ignore complications involving determination, dependence and supervenience for now.)
I am in complete agreement with this conclusion, but his argument for that claim seems to me deeply flawed. The argument is from a thought experiment involving what he calls 'dancing qualia'. Here's the thought experiment.
Suppose that one constructed a circuit of silicon chips that structurally and functionally mimicked perfectly the neural circuit (forgive the metaphor) underlying a visual experience of blueness. Now, in a reductio ad absurdum argument, we suppose that the experience is not determined by this physical organization but instead differed from the neural circuit in the silicon circuit. He imagines that we suppose the visual experience might be one of redness instead of blueness. I'll assume that there is no visual experience at all fixed by that silicon circuit since that appears to be more like the claims of Searle and other anti-functionalists.
Now, suppose that we could connect a human brain to this silicon circuit or to the section of the visual cortex ordinarily responsible for that experience, and this connection could be mediated by a single switch so that the brain could be switched seamlessly between brain circuitry and silicon circuitry. The result of this switching would be that the person would alternate between having a visual experience of blueness and having no visual experience at all (given our hypothesis). But, also given our hypothesis, the functional organization of the brain/silicon chips is exactly the same, so the firings fed into the rest of the brain are functionally indistinguishable. It follows that the person's inner cognitive state could be switched back and forth between blueness and nothing without anything else in the person's cognitive state changing. The person would continue to believe she had just had a blue experience, would say and believe exactly the same things about her "blue experience" and be emotionally affected in exactly the same way by this "blue experience". Chalmers thinks this result is absurd: it is not possible that one's qualia could dance in and out of one's experience without having any effect on one's other mental states.
The problem with this thought experiment is that we know that people can have various cognitive deficits which, taken together, could conceivably have something like the dancing qualia result that Chalmers thinks absurd without any change in their other mental states or cognitive processes. One relevant syndrome is blindness denial. In this state, people are blind yet unaware of their blindness; they assert and apparently believe that they are fully capable of sight despite their manifest blindness, and they confabulate impressively to account for their difficulties in using their eyesight. Similarly, we might imagine that someone had a kind of unconsciousness denial in which she did not have a conscious experience but still had all the functional equivalents. She would believe she had had conscious visual experiences but would not have had them, and the functional equivalence of her experiences would make the confabulation unnecessary. Conversely, there is the blindsight syndrome in which people are capable of some degree of visual processing without any awareness that they are in fact processing that information. So, when forced to guess whether there is a light in their supposedly blind field, they will do so at a rate much higher than chance, yet they will remain steadfastly certain that they can see nothing in that area. So, it might be possible to have functional equivalence of visual processing without any consciousness of seeing anything at all (although blindsight patients are far from functionally equivalent to fully sighted people). And, of course, the denial that they can see anything would make these patients clearly different from the imagined silicon-equivalents. The point of these examples is not that phenomena exactly like that described by Chalmers are actual but that it is conceptually possible that someone could have the functional equivalent of conscious visual processing without any realization that one's visual processing was not in fact conscious at all. So, in the case in which we switch back and forth between the conscious and the non-conscious visual processing, it is not absurd to think that the person not be aware of this difference, that the person's cognitive and other mental states not be affected at all.
I suppose one might respond that the imagined case is not one in which the person's brain is functioning abnormally or is damaged in any way, so we would have no reason to think this complex syndrome occurs. However, we might easily think that the silicon-chip circuit was in itself a kind of abnormal functioning or brain damage. The damage need not affect the rest of the brain if the circuits are, as supposed, functionally equivalent.
A better response is to suggest that if the two circuits are fully functionally equivalent, then supposing that one such circuit is conscious and the other is not is inherently an unfalsifiable claim. So, since the person believes the state to be conscious, we should suppose that it is conscious. I have some sympathy for this idea, and it is amusingly presented by Raymond Smullyan's "An Unfortunate Dualist", but it's also clearly a different argument from Chalmers's. If we supposed it was possible for the functional organization of two states to be equivalent without the consciousness being the same, then we've already made the assumption that two states can differ without falsifiable differences between them.