PHILOSOPHY 203 (W)

 Lecture 1: Skepticism and rationality.

Epistemology is literally the theory of knowledge (episteme). But the most interesting questions in the area are not about knowledge at all, but rather about the closely related concept of reasonable belief.

What's the difference between knowledge and reasonable belief? The most obvious difference is that knowledge has to be true. If the earth is not at rest, then obviously you can't know that it is at rest. But it does not follow that you cannot reasonably believe that it is. Indeed, for centuries people believed precisely this on the basis of excellent (though ultimately misleading) evidence. Given the evidence then available, the ancient Babylonians had every right to believe that the earth was at rest. This was the sensible thing to suppose given what they knew about their environment. Their belief was therefore reasonable, and yet it did not amount to knowledge. And this is enough to show that the two notions are distinct. To put the point with the sort of fussy precision that you will come to recognize as characteristic of modern philosophy:

Statements of the form 'S knows that P' imply that P is true. Statements of the form 'S reasonably believes that P' do not imply that P is true.

This is an important difference between the two concepts. But another difference is much more significant for our purposes. The best way to bring it out is to distinguish between two kinds of skeptical argument.

As you will see at length over the course of the term, skeptical arguments are an important device in philosophy. The skeptic - usually a fictional character, but sometimes a real person - offers a sweeping challenge to our most basic received opinions on some important topic. Thus the religious skeptic rejects conventional religious opinion; the skeptic about science challenges so-called "scientific knowledge"; the moral skeptic challenges the prevailing ethical view, and so on. It is characteristic of the skeptic that his challenge is essentially destructive. When a religious innovator like St. Paul or Mohammed challenges the received wisdom of his time and place, he seeks to motivate a new religion or the modification of an existing one, and we would not ordinarily describe this as a skeptical challenge. By contrast, when the skeptic takes on religious doctrine, his aim is to undermine the arguments that are supposed to support it without erecting any alternative doctrine in its place. The skeptic says:  "For all I know, your position may be right, but you are certainly not entitled to hold it on those grounds: Your arguments are lousy; your evidence is weak. So your beliefs are unsupported and you should set them aside."

We can distinguish two species of skeptical challenge. The moderate skeptic targets our claims to knowledge in some central area. He fastens on the fact that knowledge implies certainty. If I claim to know that God exists, or that the earth is more than 5000 years old, or that there are people in this room right now, I am expressing my certainty that these facts obtain. But more than this, I am claiming a right to this certainty. I may be quite convinced that tomorrow's lottery number will be 1988563: my belief may be unshakable; I may be psychologically certain. But if I cannot offer any reason for my belief, then I cannot claim to know what tomorrow's number will be. I cannot be epistemologically certain. By contrast, when I claim to know that my initials are G. R., I am not just reporting my confidence in this fact: I am asserting that I have every right to this confidence.

Statements of the form "S knows that P" imply that S has a right to be certain that P.

The moderate skeptic fastens on this fact because claims to rational certainty are surprisingly easy to dislodge. Suppose you claim to know that your car is parked around the corner. The skeptic may then ask, "How do you know it hasn't been stolen in the past five minutes? Can you rule that possibility out on the basis of evidence and argument?" You probably can't. And when you grant him this much you concede, in effect, that you can't really be certain that your car is still parked where you left it. And once you've gone that far you must presumably admit that — come to think of it — you don't really know where your car is parked after all.

But of course you still have an opinion about where it is. You're still pretty sure that it's just around the corner. Moreover, this opinion is not just a hunch. You remember where you parked; you know that your memory is pretty reliable; you know that car theft is relatively uncommon around here. All of this constitutes excellent rational support for your belief that your car is still parked around the corner. The moderate skeptic's argument may show that your evidence does not warrant perfect confidence in this case. But it does not show that your evidence does not support your commonsensical belief to some very high degree that falls short of perfect certainty.

The moderate skeptic thus challenges a claim to knowledge without denying that the underlying belief is well supported by the evidence. These arguments are important, and much can be learned from studying them. But they are of limited interest, in part because we can usually just concede the skeptic's point without much fuss. Presented with a moderate skeptical challenge we can usually say: "Sure, now that you mention it, I agree that I don't really know very much, and I'm being sloppy when I say that I know this or that. But that doesn't mean that my ordinary opinions are unwarranted or that I should agonize here and now about how to revise them. Thanks very much for calling this fact to my attention: I hereby resolve be more cautious in my claims to Knowledge from now on. But I see no reason to change my mind about anything important."

This response to the moderate skeptic exploits the following crucial fact:

Statements of the form "It is reasonable for S to believe that P" do not imply that S is justified in being certain that P.

A skeptical demonstration that you can’t be sure that the sun will rise tomorrow is therefore fully compatible with the commonsensical view that one has excellent reason to believe that the sun will rise tomorrow. That's why moderate skepticism – however interesting – poses no significant threat to our ordinary view of things.

On the other hand, a strong skeptical argument is much harder to dismiss. The strong skeptic maintains, not simply that that we do not know what we think we know, but that our ordinary assumptions in the target area are altogether unsupported by the available evidence. Suppose you believe that God exists on the basis of some sort of evidence. Maybe you take some religious text as an authority. Maybe you believe that the only way to explain the apparent order in the universe is to posit the existence of an intelligent creator. Now suppose I show you that these arguments are no good: that your supposedly authoritative scripture is unreliable on other matters, or that there are better explanations for the apparent cosmic order. Then I have destroyed your justification. I have shown that your reasons for believing in God come to nothing. Not only is your belief uncertain: it is totally unsupported by evidence or argument.

Now you can't just shrug this off. On the face of it, the only rational response to a strong skeptical argument is to give up the belief in question, at least until you find some better reason for accepting it. You might give it up by adopting some alternative position. (In this case that would amount to a conversion to atheism.) Typically, however, what the skeptic wants us to do is not to convert, but rather to suspend judgment: to frame no opinion at all until some new bit of evidence or argument comes along.

The view that this is the rational response depends on the following principle, which I shall call the requirement of strong rationality:

In the conduct of our intellectual lives we are obliged to believe only what we have reason to believe: to believe only what can be supported on the basis of evidence and cogent argument. Where evidence and argument do not settle the matter, we are obliged to suspend judgment.

In some applications, this principle is obviously correct. We all agree, for example, that when we discover that one of our opinions is based on a bad argument or on a prejudicial assessment of the evidence or on the testimony of an unreliable informant, we are under some obligation to revise it if it cannot be supported in other ways. In other areas, however, the principle will be more controversial. It implies, for instance, that someone who believes in God simply on faith - i.e., for no reason at all - is doing something wrong. And some of us will certainly want to resist this claim.

The first half of this course is devoted to an exploration of the requirement of strong rationality. Suppose you accept it. What may you believe? We're going to consider a series of strong skeptical arguments, the cumulative effect of which is to suggest that someone who accepts the requirement must suspend judgment about practically everything. We begin next time with the topic of religious belief. But for now you should ask yourself whether you are inclined to accept the requirement of strong rationality, and if not, how you would qualify it. For those of you who reject it, some qualified version of the principle is surely necessary. For the alternative is the view that there are no rational constraints on belief whatsoever. This latitudinarian position has some attractions. Criticism, after all, is very often unpleasant. And a general principle according to which each of us may believe whatever he likes would put an end to rational criticism once and for all. But the costs of adopting this sort of policy would vastly outweigh the benefits. Do you really want to say that scientists, engineers, judges, jurors, policy makers and professors may believe whatever they like, without regard for evidence? Do you really want to say that when it comes to building bridges or prescribing drugs or grading papers, one opinion is always as good as another, "Who am I to criticize"? Surely not. Even if you do not accept the principle in full generality, you must therefore accept some restricted or qualified version of it. So we must ask: How might this restriction be formulated? When is it permissible to believe without reasons, and when is this impermissible? How might an answer to this question be justified?