Suppose we agree that
- Causal theories of reference have successfully answered the sceptic by ensuring that most of our empirical beliefs refer precisely to the situations that cause the beliefs (cf. Davidson).
Let's think about what a causal theorist should say in Koons' scenario. For simplicity, assume physicalism. I think some of what I say generalizes to the non-physicalist case, but only some. I think we need some more detail about the scenario. Here are three versions:
- The apparently perceptual beliefs occur without any cause.
- Mental perceptual states (states relevantly like our state of being appeared to red-cubely) occur without any cause, and then cause beliefs.
- Sensory nerve stimulations occur without any cause, and then causal mental perceptual states and beliefs.
I will consider version (2), though (3) and (4) are also interesting. As it stands, the causal theorist has to say that my description in (2) is incoherent. Granted, states that are neurally just like our own belief states occur in the victim. But these states, being systematically uncaused, are not about anything, and hence are not belief states. It seems, then, that on version (2) of the story, the causal theorist has escaped scepticism—it is still true that most of the victim's beliefs are true. But while (2) may be incoherent, there is still a sceptical scenario in the vicinity. For wouldn't it count as a sceptic's victory if the sceptic were to make us conclude that, perhaps, most of the empirical-belief-like mental states we have are not in fact beliefs? We can imagine what it is like. Suppose that an epistemic authority told us: "Yesterday, the borogoves were very mimsy." We would then acquire a neural state that would cause us to say: "I believe that yesterday, the borogoves were very mimsy." This neural state would not be a belief, perhaps, because the sentence is in fact nonsense—it does not express anything—but it would masquerade as a belief. And the scenario that most of the things that appear to us as empirical beliefs are like that would, surely, be a sceptical scenario. It seems plausible that any scenario on which it is not the case that most of those states that are not introspectively distinguishable from beliefs are not correct beliefs (i.e., are not correct or are not beliefs) is a sceptical scenario. If so, then a Koonsian scenario where most of our belief-like neural states are uncaused would be a sceptical scenario, and one that the causal theorist does not have an answer to.
I think the causal theorist's best answer compatible with (1) would be to deny that in the case where the empirical-belief-like states are uncaused that these states would be conscious, or to say that if our minds littered with too many empirical-belief-like states that are not empirical beliefs, then we would not be able to form the concept of an empirical belief. Perhaps that answer works. But I am not sure. One might, for instance, have a lot of empirical beliefs early on in one's life, which are in fact veridical, and these early ones could anchor one's concept of "empirical".
If this line of reasoning is right, then (1) requires an acceptance of the Causal Principle.