Luka Yovetich didn’t e-mail this question in as an official “Question for NoodleFood,” but rather just posted it in the comments. However, since it got lost in the shuffle, I thought I ought to drag it out for more prominent display. He wrote:

It’s actually not so clear to me anymore that the skeptic’s (about knowledge) view is self-refuting. Why can’t he say that we have justified beliefs but not knowledge? When he says that we can have beliefs that are justified, the non-skeptic says “AHA! BUT do you KNOW that we can have justified beliefs?! If so, then you’re position is self-refuting. If not, then you have not right to make that claim.”

But why can’t the skeptic just say that he doesn’t know that people can have justified beliefs? Why can’t he say that he has a right to make his claim because he has a justified belief that people can have justified beliefs? The non-skeptic might say that that leads to an infinite regress and is, therefore, unacceptable. Yet, it seems, that the non-skeptic can fall into an analogous regress when making claims about knowledge. They claim to know that p. Then the skeptic says: “But do you know that you know that p?” And the non-skeptic might say that he does. And then the skeptic asks if the non-skeptic knows that he knows that he knows that p. And so on…

The non-skeptic might say that he doesn’t have to know that he knows that p in order to know that p. That’s a reasonable sounding move. But it sounds no more reasonable than an analogous move by the skeptic, when talking about justified belief. He might say that one doesn’t have to have a justified belief that he has a justified belief that p in order to have a justified belief that p. Right?

I mean, can somebody help me see the importance difference between the two cases? What makes it so this isn’t a relevant analogy? Or if it is a relevant analogy, what other considerations should we take into account that make the skeptic’s view untenable?

A few points:

The regress of “P. I know that P. I know that I know that P. I know that I know that I know that P. I know that I know that I know that I know that P. I know…” is nothing but crow-busting nonsense. Sure, we can utter the words, continuing to add “I know that…” every time, but the individual claims quickly lose all meaning. “I know that P” refers to the fact that I have knowledge, where that knowledge could be either implicit or explicit. “I know that I know that P” means that my knowledge is explicit, in the sense that I’ve self-consciously reflected upon it. Beyond that, adding extra “I know that…” adds nothing and refers to nothing, since we can’t be double-self-consciously aware of some fact.

So following that line of regress beyond the stage of explicit knowledge isn’t at all helpful. The question for the skeptic must be: “How are your beliefs justified?” So the skeptic has some possible answers. He can say, “I have no idea” or “I just feel it” or “My soul is one with the world.” That’s an unlikely response, since we can then dismiss his claim of justified belief. If he instead says, “My justification consists of X, Y, and Z,” we can then ask for the justification of X, Y, and Z. The skeptic cannot claim some self-evident truths at the bottom of that chain of justification, since then he wouldn’t be a skeptic. He cannot appeal to perceptual self-evidencies, as the Objectivist does to ground his claims to knowledge. So what if the skeptic claims that it’s justification all the way down? Well, then he does have an infinite regress, meaning that the justification for the conclusion is diminishing with every step backwards.

Let me quantify this point to make it more clear, particularly since that will reveal the skeptic to be in an even worse position that already noted. Let’s say that each justified belief has a probability of .800 of corresponding to the facts. So if P is justified by X, Y, and Z, and X is justified by A and B, Y is justified by C, D, and E, and Z is justified by F and G. (We’ll stop at this point, just so that the example doesn’t become too complicated to grasp. Also, for the sake of simplicity, all of the justifying beliefs are independent and necessary.) So if A to G are justified beliefs, then each will have a probability of .800. However, that will not lead to X, Y, and Z being justified beliefs with a probability of .800. Rather, X and Z will have a probability of .640, while Y will have a probability of .512. That means that the conclusion P will have a probability of .210. And that probability of the final conclusion P will continue to drop as we factor more links in the chain of justification.

However, as I already indicated, the skeptic is in a worse position than that analysis indicates, since he cannot know the exact probability of his supposedly justified beliefs, since that would constitute knowledge. Worse still, he cannot factor in the possibility of inaccuracy in those probabilities, since that would require knowledge that, for example, his .800 probabilities are .963 probable. And maybe some beliefs should be given more weight than others, e.g. that A counts for 83% of the justification, while B only adds 17%. And maybe some probabilities are dependent upon others, such that if A is right, then the probability of C decreases by .027. And maybe if A is wrong, then M and N could serve as substitutes. However, the skeptic couldn’t even know how to calculate such probabilities, nor could he be certain that he performed the arithmetic correctly.

So a skeptic about knowledge cannot claim that he has justified beliefs, since the possession of justified beliefs presupposes knowledge. To put the point another way, if nothing can be certain, then nothing can be probable or possible either.

I’m not sure if that’s the best way of addressing Luka’s question — and perhaps Luka has some other approach in mind for the skeptic. Either way, let’s have it out in the comments!

(Just so that folks know, I have four questions still in the pipeline. I’ll be posting them at least a few days apart, since I want to give the comments on each time to develop.)