My 2 cents on mind and matter is that it doesn't matter if mind is matter or not with regard to the reliability of perception. Because whatever 'they' are they are effectively integrated. A surgeon with gloves on manages to perform well even if his hand is not in direct contact with the scalpel. There is nothing in principle saying that fundamentally different kinds of being cannot share integration points, especially if these points have been constructed and assigned by (god forbid) a god. I have always found the arguments against Cartesian dualism unconvincing for the above stated reasons. In fact, if a supernatural dualist model is going to work well it is going to have to be so well integrated as to appear indistinguishable. This is why it is (contrary to what many cognitive scientists have claimed) pretty much impossible to disprove a supernatural dualist model using the latest scanning techniques. The fallacy lies in somehow assuming that if the dualist model was true then brain scans would turn up different sorts of results, i.e. that the activities of the mind would not show corresponding integrated registrations in the brain. To the contrary, if the dualist model is going to be a good working model, that is precisely how it ought to work. There ought to be parts of the brain that are specialized for certain activities of the mind, and this to such a degree that these functions of the brain should be able to be artificially stimulated to produce sensations or experiences normally initiated by the mind. I guess what I am saying is that there is nothing about Cartesian dualism that I find modally unsound in principle. That doesn't mean it's true. I'm prepared to accept that it might not be. But not because of any inherent contradiction in the model. I have always been somewhat bemused by the notion that if there is a supernature, it would necessarily be unable to have any conceivable causal relation to nature. I suppose in a sense if what one means by nature is 'everything there is' then of course nothing else but nature exists. But that stills fails to say how it is that separate categories of ontology should necessarily be inherently unable to possess inter-causal relations. The idea of mechanistic causation is a kind of Gollumium 'my precious' of the modern zeitgeist. I think we often use it to think we've disproved other constructs of reality, when, in fact, we really haven't. I mean after all, look at these brain scans.