Introduction

1. Radical scepticism

Much of current epistemology (the theory of knowledge) in philosophy is best seen as a response to the thesis that we never have the slightest reason to believe anything. This radical skepticism must be distinguished from the more commonsensical idea that nothing can be known for certain and that we can never be absolutely sure of anything. Common sense assumes that practical certainty is possible even if absolute certainty is not. Radical skepticism departs from common sense and denies that even practical certainty is ever attainable. Indeed, it denies that anything is ever even the slightest bit more likely to be true than anything else.

The problem is not that there are radical skeptics who need to be convinced that they are wrong. The problem is that an extremely natural line of argument seems to lead inevitably to radical skepticism. Common sense keeps us from accepting such a conclusion; but that leaves the philosophical problem of saying what goes wrong with the reasoning that seems to lead there. To repeat, the problem has not been to find an argument against skepticism, it has been to find out what is wrong with an argument for skepticism.

One such argument begins by asking how you know that the color red looks to someone else as it looks to you. Perhaps things that look red to you look green to him, things that look blue to you look orange to him, and similarly for other colors. If his spectrum is inverted compared with yours, there may be no way to discover it, since what is for you the experience of red he will call the experience of green. Even if he describes colors exactly as you do, there could still be a systematic difference between his and your experience of color. This suggests that you can have no reason to suppose that others see the world as you do rather than as does the man with an inverted spectrum.

Further reflection suggests that there is no reason to suppose that visual perception gives other people experiences that are anything like your visual experiences. Perhaps someone else has what would be for you auditory experiences. When he looks at the blue sky it is like what hearing middle C on the piano is for you. There seems to be no way to tell, since he would have been brought up to call that sort of experience the experience of blue. Indeed it is not clear that you have the slightest reason to suppose that others have anything you could recognize as experience. When others see things, their visual experience may be something you could not even imagine.

It might even be suggested that there is no reason to suppose that others have any experience at all. The suggestion is that, even if you could know that the people around you were made of flesh and blood, born of women, and nourished by food, they might, for all you know, be automatons in the sense that behind their elaborate reactions to the environment there is no experience. But the suggestion is not merely that you do not know whether other people have any experience but also that you haven't the slightest reason to suppose they do.

Similarly, it might be suggested that you have not the slightest reason to believe that you are in the surroundings you suppose you are in, reading a book called Thought. It may look and feel to you as it would look and feel if you were in those surroundings, reading such a book. But various hypotheses could explain how things look and feel. You might be sound asleep and dreaming — or a playful brain surgeon might be giving you these experiences by stimulating your cortex in a special way. You might really be stretched out on a table in his laboratory with wires running into your head from a large computer. Perhaps you have always been on that table. Perhaps you are quite a different person from what you seem: you are a volunteer for a psychology experiment that involves having the experiences of someone of the opposite sex reading a book written in English, a language which in real life you do not understand. Or perhaps you do not even have a body. Maybe you were in an accident and all that could be saved was your brain, which is kept alive in the laboratory. For your amusement you are being fed a tape from the twentieth century. Of course, that is to assume that you have to have a brain in order to have experience; and that might be just part of the myth you are being given.

You might suppose that there are actually little differences among these hypotheses. After all, you have the same experiences on all of them. So you may feel that, no matter what is actually the case, everything will work out fairly well if you continue to act on the assumption that things are roughly what they seem. But that seems to go beyond your evidence. It may be that up until now everything has worked out well on the assumption that things are roughly what they seem to be. But what reason is there to suppose that things will continue to work out well on that assumption? An inductive inference is needed here: "Things have worked out well in the past; so they can be expected to work out well in the future." But how could you justify this use of induction? You might argue that, since inductions have generally worked out well in the past, you are entitled to expect them to work out well in the future. But that would be circular. You would be giving an inductive argument in order to defend induction.

Furthermore, do you really have any reason to suppose that things have worked out well in the past? You may seem to remember that they have. But how can you justify relying on your memory? Surely not on the grounds that your memory has been reliable in the past, for what reason do you have for thinking that it has been? The fact that you seem to remember that memory has been reliable in the past is irrelevant. You cannot legitimately appeal to memory to justify memory. You seem to have no reason to suppose that things have worked out well if and when in the past you have acted as if things are much as they seem.

Such reflections lead to the philosophical problems of other minds, the external world, induction, and memory. It is important that these problems are not simply whether we could ever know for sure or be absolutely certain of various matters. They are more radical: How is it possible that we should have the slightest reason to continue to believe as we do? How can there be more reason to believe what we do about the world than to believe anything else, no matter how incredible it may seem?

Recall that our interest in these problems is not that we want to be able to refute the radical skeptic. It is rather that we suppose that we often do have reasons to believe one thing rather than another and can sometimes even be practically certain, even if not theoretically certain, that we are right. Common sense tells us that something must be wrong with the arguments for radical skepticism. The...