– Ned Block, in ‘Troubles with Functionalism’ ties his absent qualia argument against functionalism, more or less, to Kripke-esque identity theory. The basic outline is this: if S is a functional state, and Q is a mental state, then functionalism holds that Q=S. Block argues that it is nomologically possible that such a system could in fact be in S without Q – or, at least, its possession of Q can be doubted. Thus, on the Kripkean scheme, if Q=S, then it is necessary that Q=S. However, Block argues that the identity relation can be doubted, since it is nomologically possible that a system can be in S without having Q. We can see, then, the trouble for the functionalist claim if there are absent qualia.
– Chalmers, in ‘Absent Qualia, Fading Qualia, Dancing Qualia’ argues for the impossibility of absent qualia by showing that if absent qualia are possible, then fading qualia are possible – but we have good reason to suppose that fading qualia are impossible. To make a long argument short, Chalmers ends up with the idea that if fading qualia are possible, then a subject with fading qualia whose rational processes are functioning and is fully conscious would be completely wrong about his conscious expereince – and this, Chalmers says, is just implausible, given his own presupposition of ‘functional invariance’.
– John Searle, in ‘The Mystery of Consciousness’, notes the following about the fading qualia argument (which, to my mind, seems a rather sound criticism):
‘The basic idea of the argument is to show that there could not be a mismatch between functional organization and consciousness, because if there were it would be possible to imagine a system’s conscious states fading out (“fading qualia”) even though its functional orginization and hence its behaviour remained constant. And it would also be possible to imagine a system’s conscious states changing in a way that was not systematically related to its behavior (“dancing qualia”). But these, he (Chalmers) says, are impossible because any change in mental content must be “mirrored in functional organization” and therefore be in behavior. But this argument just begs the question by repeating the point at issue and does not establish it.’ (p 151)