Global skepticism is often refuted by using pragmatic arguments such as "nobody is honestly a skeptic" etc, because it is thought that skepticism is theoretically irrefutable - after all, one can doubt beyond any proofs, even doubt logic etc, at least in principle. But this refutation is not perhaps very convincing, because it seems to more like ignore the problem of skepticism than really fighting it - something that does not look right from the point of view of honest and thorough philosophy. If we cannot show that we can know things, how should philosophy justify its existence?

I think I can do that. So, let's see whether I am totally in error, or just a bit off course ;).

The claim of a skeptic can be characterized like the following sentence:
(S1) You cannot prove that. (Where "that" is pretty much anything).

Let's analyse a bit. What is a proof? If we are talking about logical proofs, which show things beyond question, we see that every logical proof is something that has premises as well as a conclusion. Logic derives true conclusions from true premises. If A is a set of premises, then the statement p is proved iff A |-- p, that is, p can be logically derived from A. Consider the following example:

This is a proof. What the skeptic would do is that he would say: yes, but you do not know P1 or P2. Then he would demand a proof for those. But as the proof of P1, for instance, is based on other set of premises A2, whose premises are again based on another set of premises A3, and so on and so on, it is clear that we cannot prove this beyond doubt. Right?

Well, not exactly. What the skeptic is asking is a proof - so what he is asking in S1 is something that has presupposed premises and a conclusion. He gets exactly that, but is not satisfied. Why? Because he is not really asking for a proof at all. Every proof is of the form A |-- p, but the skeptic is not satisfied with that. The problem is the skeptic who misuses the word "proof". What he is actually looking for is this:

Allright. Now we must first understand that by proof it is meant something that uses logic. The use of logic cannot ensure truth (note that this disarms the skeptic's arguments about letting go of logic because logic cannot be proved - he does not understand that he must step outside logic anyway), so we should not use logic to find truth. This is simple: if logic cannot ensure truth, then even if something was proved through logic beyond any doubt, it still would not do - so logic is out of the game.

Does the skeptic have a meaningful truth? I guess so, because he speaks of truth, so it must mean something. Unless, of course, the skeptic wants to be a mere pain in the ass and spew out nonsense. So if there is truth to be found, but we cannot prove that we have found it, then we must look somewhere else for it. Taking logic out of the picture leaves us with not very much room to manouver: it must be in the senses.

Okay, so we take a simple proposition: "I feel". The skeptic could agree that yes, it seems like we feel, but... and I will say: halt. He looks at me, eyes wide open, and blinks twice. I say: what is that "but" there? Is it not a sign of you using logic or reason on the mere perception or assumed one? But you already proved beyond any doubt that logic cannot guarantee truth, which lead to the idea that logic must be eliminated from the consideration. If the only thing you have is your perceptions, where does that logic step in? After all, proving that it is logically possible that you do not feel does not have anything to do with the truth of whether you feel or not, right?

Does the skeptic have to answer: "guess so"?

I think he does. The skeptic blows his cover when he takes logic out. If he says anything meaningful at all, he will end up in the situation where we perceive, but cannot doubt our perceptions using logic. Does he doubt that he feels? I do not think anyone would honestly do that - this leads to the pragmatic refutation of skepticism, but with the theoretical structure behind it to force the issue. In fact, I believe Husserl was on the ball when he spoke of apodictic truths as something that belong to the sensibility, or faculty of perception or whatever. The certainty the skeptic looks for is in the perceptions he gets. Unless, of course, he would be willing to take logic back in, in which case he would end up being torn to pieces with logic.

This does seem to show that the skeptic cannot simply throw in "I do not believe you" like he wishes, because then he is not making any sense. There is, after all, a difference between rational doubt and mere doubt. Doubting is an intensional state that has no relation to the world as such - doubting something shows nothing of the something. That can be easily shown by assuming that there is a proof that shows conclusively that we know something: the skeptic can, nonetheless, just say "I don't believe you".