I take it that by "sane" one has in mind some notion of normal mental functioning. If so, then the question points to an ambiguity in the notion of normality. One such notion is a statistical one: What is "normal" is simply what is common (or average, or what have you) in a given population. Another notion is normative: What is normal is what is proper, in accord with the rules, and the like. The contrast emerges in different ways in different cases, but its presence can usually be uncovered by considering modal statements: Would what is now normal still be normal if things were very different from how they now are? For example, it is normal for human beings to have five fingers on each hand. Suppose genetically modified corn carried some kind of virus so that, if a pregnant woman were to eat it, her children would have six fingers on each hand. Suppose further that such corn becomes very common, so that babies everywhere start to be born with twelve fingers. No-one knows what the cause is, so this...

One can slightly simplify Mark's case as follows. Suppose one believes that p and also believes that q . One therefore believes that p and q , but also that p or q . The disjunctive belief surely must be "based upon" one's beliefs in the disjuncts, but neither of them is essential: The belief in the disjunction would survive failure to believe either disjunct. Should we say that your knowledge is "based upon" a false belief? It seems to me that, in this case, one has two independent reasons for the disjunctive belief, namely, the beliefs in its disjuncts. The reasons are independent in several senses: (i) one could have each reason even if one did not have the other; (ii) as I'm imagining the case, anyway, one's justification for the beliefs in the disjuncts is independent of one's justification for the othe belief; (iii) each reason is sufficient, on its own, to underwrite knowledge in the disjunction. As it happens, however, only one of the reasons actually underwrites...

Some people think they do know what exists after death. As Alex notes, "Nothing" is one option, and some people believe they have strong enough evidence for this view to make it a reasonable belief, and perhaps even to count as knowledge. On the other side, there are people who would claim to know, on the basis of divine revelation, that they will survive the deaths of their bodies. Presumably, not both camps can be right, and both camps might be wrong about their claims to know . But I don't see any general reason, absent consideration of the details, for supposing neither of them could know. Perhaps one wants to say that neither camp can really know, on the ground that truly conclusive evidence isn't available. But truly conclusive evidence is rarely available for anything, and yet we claim to know lots of things. That is to say: The claim that it is not possible to know what exists after death seems to be based upon a general sort of skeptical challenge that, allowed to run free, would...

Truth is a property that some propositions have and some do not. Itcan be hard to tell which a proposition is. But this much we can say.The proposition that Wittgenstein was Jewish is true if, and only if,Wittgenstein was Jewish. The proposition that Frege was Catholic istrue if, and only if, Frege was Catholic. And so forth. Somephilosophers (Paul Horwich, Scott Soames) think that's about all thereis to be said about truth. I'd disagree, but I hope we can all agreethat, even if that's not all there is to be said, it is something that there is to be said. (Note that there is some kind of sense herealready in which what is true depends upon how things are. For example,whether the proposition that Russell was German is true depends upon,well, whether Russell was German, and that's a question of how things are "out there".) So, that said, how can we know that truth is not an illusion?The obvious way to interpret the question is: How can we know, e.g.,that it isn't an illusion that it is true...

There have been many instances over the centuries in which well-confirmed scientific theories were later shown to be wrong. Usually, they weren't simply wrong. There was something they had right, but then it turned out that there were various sorts of problems, and a very different theory had to be introduced, often with a wholly different background metaphysics. Natural selection, as well confirmed as it now seems, could turn out to be wrong. (And there are very different takes on natural selection itself, anyway.) So yes, our understanding of the world is no more guaranteed to be correct than was that of our forebearers. (I'm not happy with the term "relative", which tends to be used in a different way.) But that doesn't mean that we don't have good reason to believe what we believe. We do. If we find we have better reason to believe something else, then we'll believe that. But until we do find better reason to believe something else, we should, well, believe what we have best reason to believe...

There seems to be a pattern of argument here that needs to be questioned. It is: (i) Method M for reaching judgements isn't completely reliable; therefore, (ii) method M can't be trusted. The conclusion simply doesn't follow. Method M might be very reliable, in which case it can be trusted to a high degree. I doubt "group opinion" is the only method available to us for distinguishing hallucination and illusion from perception. (For example, in the land of the blind, there might be ways of correlating your apparent perceptions with facts on which you can all agree, e.g., that there is a boulder in such-and-such a location that wasn't there yesterday.) But it might be quite a good way of drawing the distinction even if it isn't a perfect way.