Of course, no one supposes that zombies exist. But the actual existence of zombies has nothing to do with it. This is a point about a theory about reality and its adequacy to account for that reality, not reality itself.selfAdjoint said:Logical consistency is a weak criterion. Many things which are logically consistent are not true of the universe we experience. Taxicab topology for example. The fact that the zombie world is consistent with physicalism is of no force whatever against physicalism unless you can show that zombies actually exist.
In the case of topology, we have a pretty good model of what physical conditions make for what type of topologies. We can say, "because conditions such-and-such obtain under the influence of known physical laws, the topology of universe X is so-and-so." We have a well-behaved function that maps each set of unique physical conditions onto its own topology. You put in conditions A, and topology B pops out like clockwork; if you get topology C, it's because you've done the math incorrectly. The fact that the physical laws allow topologies not observed in our universe is not problematic, because we can explain
1) why these topologies are not observed. They are not observed because they follow directly from conditions that do not obtain in our universe.
2) why the topology that we do observe is observed. It is observed because it follows directly from conditions that obtain in our universe.
In the case of consciousness, we do not have a similar scenario. If you start from just a physical description of the brain, you have no a priori reason to believe that anything like subjective experience should directly follow. You input the physical description, and what you get back is a zombie. Sure, you can ad hoc it and say "whenever these conditions obtain, subjective experience results." But this is hardly a satisfactory account.
If we were working with topologies, this would be like having the equations tell us that conditions A should lead to topology B, but oops, actually it's empirically known that they create topology C-- worse, we have have no idea why this is the case, and worse still, no rearranging of the equations consistent with the current theory can get us from A to C. But rather than try to radically revise our theories to account for this mystery satisfactorily, we content ourselves to maintain our current set of equations, with a discontinuous, ugly, and unexplained exception for the anomalous set of conditions. Worse, we insist that our set of equations is entirely correct, even though it produces the wrong answer for the anomalous case without the ad hoc patch up.