A Defense of Skepticism
Peter Unger


There are certain arguments for skepticism which conform to a familiar . . . pattern or form. These arguments rely, at least for their psychological power, on vivid descriptions of exotic contrast cases. The following is one such rough argument, this one in support of skepticism regarding any alleged knowledge of an external world." The exotic contrast case here concerns an evil scientist, and is described to be in line with the most up-to-date developments of science, or science fiction. We begin by arbitrarily choosing something concerning an external world which might conceivably, we suppose, be known, in one way or another, e.g., that there are rocks or, as we will understand it, that there is at least one rock.

[Argument] Now, first, if someone, anyone knows that there are rocks, then the person can know the following quite exotic thing: There is no evil scientist deceiving him into falsely believing that there are rocks. This scientist uses electrodes to induce experiences and thus carries out his deceptions, concerning the existence of rocks or anything else. He first drills holes painlessly in the variously colored skulls, or shells, of his subjects and then implants his electrodes into the appropriate parts of their brains, or protoplasm, or systems. He sends patterns of electrical impulses into them through the electrodes, which are themselves connected by wires to a laboratory console on which he plays, punching various keys and buttons in accordance with his ideas of how the whole thing works and with his deceptive designs. The scientist's delight is intense, and it is caused not so much by his exercising his scientific and intellectual gifts as by the thought that he is deceiving various subjects about all sorts of things. Part of that delight is caused, on this supposition, by his thought that he is deceiving a certain person, perhaps yourself, into falsely believing that there are rocks. He is, then, an evil scientist, and he lives in a world which is entirely bereft of rocks.

[Argument continued] Now, as we have agreed,

1. if you know that there are rocks, then you can know that there is no such scientist doing this to you, [i.e., deceiving you to falsely believe that there are rocks.] But

2. no one can ever know that this exotic situation does not obtain*; no one can ever know that there is no evil scientist who is, by means of electrodes, deceiving him into falsely believing there to be rocks. That is our second premise, and it is also very difficult to deny. So, thirdly, as a consequence of these two premises, we have our skeptical conclusion:

3. You never know that there are rocks. But of course we have chosen our person, and the matter of there being rocks, quite arbitrarily, and this argument, it surely seems, may be generalized to cover any external matter at all. From this, we may conclude, finally, that

4. nobody ever knows anything about the external world.

[Comments] This argument is the same in form as the "evil demon" argument in Descartes' Meditations*; it is but a more modern, scientific counterpart, with its domain of application confined to matters concerning the external world.' Taking the Meditations as our source of the most compelling skeptical argument the philosophical literature has to offer, we may call any argument of this form the classical argument for skepticism ....

These arguments are exceedingly compelling. They tend to make skeptics of us all if only for a brief while. Anyone who would try to further skepticism, as I will try to do, will do well to link his own ideas to these arguments. For then, the very notable feelings and intuitions which they arouse may serve as support for the theses he would advance ....



Our skeptical conclusion would not be welcome to many philosophers. Indeed, most philosophers would be inclined to try to reverse the argument, perhaps in the manner made popular by G. E. Moore They would not, I think, wish to deny the first Premise which in any case seems quite unobjectionable, at least in essential thrust. But even in its early formulation, they would be most happy to deny the second Premise which is the more substantive one."

[Reverse Argument] The Moorean attempt to reverse our argument will proceed like this: [l.] According to your argument, nobody ever knows that there are rocks. But I do know that there are rocks. This is something concerning the external world, and I do know it. Hence, [3.] somebody does know something about the external world. Mindful of our first premiss, the reversal continues: I can reason at least moderately well and thereby come to know things which I see to be entailed by things I already know. Before reflecting on classical arguments such as this, I may have never realized or even had the idea that from there being rocks it follows that there is no evil scientist who is deceiving me into falsely believing there to be rocks. But, having been presented with such arguments, I of course now know that this last follows from what I know. And so, while I might not have known before that there is no such scientist, at least I now do know that there is no evil scientist who is deceiving me into falsely believing that there are rocks. So far has the skeptical argument failed to challenge my knowledge successfully that it seems actually to have occasioned an increase in what I know about things.

[Comments] While the robust character of this reply has a definite appeal, it also seems quite daring. Indeed, the more one thinks on it, the more it seems to be somewhat foolhardy and even dogmatic. One cannot help but think that for all this philosopher really can know, he might have all his experience artificially induced by electrodes, these being operated by a terribly evil scientist who, having an idea of what his "protege" is saying to himself, chuckles accordingly. One thinks as well that for all one can know oneself, there really is no Moore or any other thinker with whose works one has actually had any contact. The belief that one has may, for all one really can know, be due to experiences induced by just such a chuckling operator. For all one can know, then, there may not really be any rocks. Positive assertions to the contrary, even on one's own part, seem quite out of place and even dogmatic.

[Counter Argument) Suppose that you yourself have just positively made an attempt to reverse; you try to be a Moore (and claim to know that there is no scientist who implanted electrodes and is deceiving you.) Now, [case i ] we may suppose that electrodes are removed, that your experiences are now brought about through your perception of actual surroundings, and you are, so to speak, forced to encounter your deceptive tormentor. Wouldn't you be made to feel quite foolish, even embarrassed, by your claims to know? Indeed, you would seem to be exposed quite clearly as having been, not only wrong, but rather irrational and even dogmatic. And [case 2] if there aren't ever any experiences of electrodes and so on, that happy fact can't mean that you are any less irrational and dogmatic in saying or thinking that you know. In thinking that you know, you will be equally and notably irrational and dogmatic. And, for at least that reason, in thinking yourself to know there is no such scientist, you will be wrong in either case. So it appears that one doesn't ever really know that there is no such scientist doing this thing.

[Extension and Qualification] Now, if you think or say to yourself that you are certain or sure that there is no scientist doing this, you may be doubly right, but even that does not seem to make matters much better for you. You may be right on one count because you may, I will suppose, be certain that there is no such scientist, and so be right in what you think. And in the second place, there may be no evil scientist deceiving you, so that you may be right in that of which you are certain. But, even if doubly right here, it seems just as dogmatic and irrational for you ever sincerely to profess this certainty. Thus it seems that, even if you are certain of the thing, and even if there is no scientist, you shouldn't be certain of it. It seems that you are wrong, then, and not right on a third count, namely, in being certain of the thing. It seems much better, perhaps perfectly all right, if you are instead only confident that there is no such scientist. It seems perfectly all right for you to believe there to be no evil scientist doing this. If you say not only that you believe it, but that you have some reason to believe this thing, what you say may seem somewhat suspect; at least on reasoned reflection, but it doesn't have any obvious tint of dogmatism or irrationality to it. Finally, you may simply assert, perhaps to yourself, that there is no evil scientist who is deceiving me into falsely believing that there are rocks. Perhaps strangely, this seems at least pretty nearly as foolhardy and dogmatic as asserting, or as thinking, that you know the thing.

[Comments] This idea, that claims to know about external things are at least somewhat foolhardy and dogmatic, applies in all possible situations, even the most exotic cases. Suppose, for example, that you actually do have a sequence of experience which seems to indicate that an evil scientist was deceiving you into falsely believing that there are rocks, You seem to be confronting an exotic scientist who shows you electrodes, points out places of insertion on your skull or shell, and explains in detail how the whole thing works. And you seem to see no rocks outside the window of this scientist's laboratory. The scientist assures you that there really are no such things as rocks, that he only created an impression of such things by stimulating certain groups of cells in your brain. After enough of this sort of thing dominates your experiences, you might suppose that you know that there is an evil scientist who deceived you in the past, but he now does not. And you may also tome to suppose that you know that there were nearer any rocks at all. But should you think you know? These latter experiences might themselves find no basis in reality, for all you really might know. For all you can know, it may be that all the time your experiences are induced by electrodes which are operated by no scientist, and it may be that there are no scientists at all, and plenty of rocks. Whether or not this is the case, you may always have new experience to the effect that it is. Is the new experience part of an encounter with reality, or is it too only part of an induced stream, or perhaps even a random sequence of experience? No matter how involved the going gets, it may always get still more involved. And each new turn may make any previously developed claim to know seem quite irrational and dogmatic, if not downright embarrassing. No matter what turns one's experience takes, the statement that one knows there to be no scientist may be wrong for the reason that there is a scientist. But it will always be wrong, it seems, for the reason of dogmatism and irrationality, however this last is to be explained


Largely because it is so exotic and bizarre, the case of a deceiving scientist lets one feel acutely the apparent irrationality in thinking oneself to know. But the exotic cases have no monopoly on generating feelings of irrationality.

[Ordinary Cases]
1. If you are planning a philosophical book and trying to estimate the energy you will spend on each of the several chapters, you might think that you know that it will not take much to write the third chapter. For the argument there may seem already so clearly outlined in your head. But experience may later seem to show that this argument is far from dear. And much time and effort may become absorbed with no clear fruits to show for it. In that case, you will, I suggest, feel somewhat embarrassed and foolish, even if there is no other person to whom your idea that you
knew was ever communicated. If you just believed, or even if you were quite confident that this chapter would not take much effort to write, then, I suggest, you would not feel nearly so foolish or embarrassed, oftentimes not at all.

2. Again, you may think you know that a certain city is the capital of a certain state, and you may feel quite content in this thought while watching another looking the matter up in the library. You will feel quite foolish, however; if the person announces the result to be another city, and if subsequent experience seems to show that announcement to be right. This will occur, I suggest, even if you are just an anonymous, disinterested bystander who happens to hear the question posed and the answer later announced. This is true even if the reference was a newspaper, The Times, and the capital was changed only yesterday. But these feelings will be very much less apparent, or will not occur at all, if you only feel very confident, at the outset, that the city is thus-and-such, which later is not announced. You might of course feel that you shouldn't be quite so confident of such things, or that you should watch out in the future. But you probably wouldn't feel; I suggest, that you were irrational to be confident of that thing at that time. Much less would you feel that you were dogmatic in so being.

3. Finally, if you positively asserted something to another in a conversation, as though reporting a known fact, later contrary experiences might well cause you to feel that you had overstepped the bounds of good sense and rationality. The feeling is that you have manifested a trait of a dogmatic personality. If you happen to be right, your extremely positive approach is not likely to be questioned. In case subsequent events seem to indicate you are wrong about the matter, then you come in for a severe judgement, whether or not this judgement is ever made out loud. This is a rather familiar social experience. (As I say this, even in trying to make my style a little less cautious, to be readable, I leave myself open to just such a judgement by putting the matter in such a positive, unqualified way.) I suggest that such feelings ought to be far more familiar, occurring even where you are right about the matter. They should not just occur where you are in fact wrong about things. Accordingly, we should avoid making these claims in any case, whether we be right or whether wrong in the matter, e.g., of which city is the capital of that state.

It is hard for us to think that there is any important similarity between such common cases as these and the case of someone thinking himself to know that there are rocks. Exotic contrast cases, like the case of the evil scientist, help one to appreciate that these cases are really essentially the same. By means of contrast cases, we encourage thinking of all sorts of new sequences of experience, sequences which people would never begin to imagine in the normal course of affairs. How would you react to such developments as these, no matter how exotic or unlikely? It appears that the proper reaction is to feel as irrational about claiming knowledge of rocks as you felt before, where, e.g., one was apparently caught in thought by the library reference to the state's capital. Who would have thought so, before thinking of contrast cases? Those cases help you see, I suggest, that in either case, no matter whether you are in fact right in the matter or whether wrong, thinking that you know manifests an attitude of dogmatism. Bizarre experiential sequences help show that there is no essential difference between any two external matters; the apparently most certain ones, like that of rocks, and the ones where thinking about knowing appears, even without the most exotic skeptical aids, not the way to think.