The Return of the Evil Genius

A stimulating dialogue between the Evil Genius

and a

Brain in a Vat

on topics in metaphysics, epistemology, and language.



David Cole

University of Minnesota, Duluth



Descartes refuted skepticism in 1641. George Berkeley refuted skepticism in 1710. O.K. Bouwsma refuted skepticism in 1949. Hilary Putnam refuted skepticism in 1981. The locus classicus for the form of skepticism refuted is Descartes' Meditations -- which also goes on to set out a famous realist refutation of skepticism. Indeed, Descartes is the principal inventor of the philosophic enterprise of skepticism refutation so central to Modern philosophy and its epistemic preoccupations. What the cited successors of Descartes and many others have in common is a rejection of both skepticism and Descartes' refutation of it. Berkeley, Bouwsma and Putnam offer a sophisticated, undercutting, refutation which undermines the very coherence of skepticism. And for all, issues of language and meaning rise to the fore, determining the coherence of any skeptical doubts and constraining the possibilities of systematic error and deception (for Berkeley, these appear as issues about the objects of thought, which in turn are the foundation of the meaning of language).

Berkeley was a metaphysician, despite many anti-metaphysical themes in his critique of contemporary philosophy, mathematics and science. For many Twentieth Century anti-metaphysical thinkers, heirs of the linguistic turn, the considerations of language and meaning raised by O.K. Bouwsma in "Descartes' Evil Genius", constitute a clearer refutation of skepticism, free of positive metaphysical axes to grind. Bouwsma presents most of his argument through the device of dialogue between the Evil Genius (the putative possibility of whom motivates the final increment of Descartes' hyperbolic doubt in the first of the Meditations) and his would-be subject of deception. Bouwsma argues that only deception potentially open to detection is possible. This or that can be fake, but not the whole world. The Evil Genius must fail in his attempt at total, transcendent, deception. An interesting feature of Bouwsma's argument is that it does not attack the intelligibility of the creation of experience by an Evil Genius, but denies that this would be deceit, and this for reasons having to do with language and meaning.


More recently, Hilary Putnam has raised similar considerations, although this time again with an explicit metaphysical point, anti-realism. In a sense, this brings us full-circle, back to Berkeley's attack on the scientific realism of the materialists. Putnam holds that there is a logical dependence of metaphysical realism upon the coherence of skepticism: if metaphysical realism were true, there would be a completely mind-independent reality, and the possibility of systematic error would arise. Thus Putnam seeks to refute realism by attacking the coherence of skepticism, and he does the latter by raising considerations of meaning similar to those pointed to by his predecessors Berkeley and Bouwsma. Putnam discusses a popular skeptical possibility updated from that of his predecessors, namely the alleged possibility that we are all brains in a vat, with experience generated by a computer. Like his predecessors Berkeley and Bouwsma, Putnam argues that our circumstances set limits on what we can mean, and that we cannot coherently entertain the possibility that we are brains in a vat. This because if we were brains in a vat, we would speak/think in `vat-English'. This language is syntactically the same as English, but in vat-English the terms don't refer to the English referents. For example, the English word "tree" refers to trees, but the vat-English homonymous "tree" cannot refer to trees, for there is no causal connection between use of the vat-English term and real trees. Putnam summarizes the argument as follows:


...if...we are really the brains in a vat, then what we now mean by `we are brains in a vat' is that we are brains in a vat in the image or something of that kind (if we mean anything at all). But part of the hypothesis that we are brains in a vat is that we aren't brains in a vat in the image (i.e. what we are `hallucinating' isn't that we are brains in a vat). So, if we are brains in a vat, then the sentence `We are brains in a vat' says something false (if it says anything). In short, if we are brains in a vat, then `We are brains in a vat' is false. So it is (necessarily) false. (15)

Like Bouwsma, Putnam does not deny that we might have our experience generated by art -- i.e. in particular, that there could actually be brains in tanks, and that we could be such things -- but like Bouwsma he does deny that one can coherently wonder if he or she is such a subject of transcendent deception.



By continuing the dialogue between the Evil Genius and a new subject in a more mundane and hopefully more comprehensible situation, a brain in a tank, I shall try to show that it is not at all clear that the considerations of meaning raised by Bouwsma and Putnam will vanquish the Evil Genius -- even if bolstered by recent work on reference and necessity. This topic is peculiarly suited to Dialogue form: paramount are issues of meaning and language, and what it is possible to talk about (and whereof one must remain silent). And then there is the intrinsic drama of the powerful Evil Genius sparring with the subject of synthesized experience. In this paper the tables are turned and the brain in the tank raises several objections to the Genius' philosophical protest, along the lines advanced by Bouwsma and Putnam, that he is no deceiver. The objections that are raised are directed against the view that the Genius and his victim speak different languages, the view that "real" is defined as one pole of various contrasts but solely internally to our experience, and also the view that no matter what an experience generator such as the Evil Genius might reveal to us, we could not justifiably conclude that we had been universally deceived regarding the ontological import of our experience.

Bouwsma considers two cases of attempted deception. In the first, the Evil Genius changes everything into paper. Flowers, people, trees, become paper. This attempt is soon detected. In the second attempt, the Evil Genius destroys all matter, but produces experiences just-as-if there were matter. This attempt at deception also fails, argues Bouwsma, but for quite another reason. The undetectable putative deception cannot be deception. The Evil Genius communicates directly with his would-be subject of deception, Tom. The latter explains to the Genius "You do not mean by illusion what we mean, and neither do you mean by flowers what we mean" (p. 96). Thus there can be no deception about whether what we mean by "flowers", e.g., exist. In the end, the Evil Genius disappears, a great and powerful supernatural Spirit vanquished by philosophic considerations.

But as any viewer of movies with titles that end in Roman numerals knows, things are rarely this simple. And given that we are here dealing not with a mere mutilator in a hockey mask, nor even a Texan with a chain-saw, but the Evil Genius, as powerful as God, in the original story, it would not be surprising if it were somewhat more difficult to dismiss him than Bouwsma thought. I shall attempt to show this, as I have indicated, by changing the context slightly in which these questions are examined from the immaterial Spirit of Descartes and Bouwsma, to the more contemporary scientist stimulating dis-embodied brains. More obscurity surrounds the notion of immateriality crucial in the Cartesian and Bouwsma arguments than is desirable in an attempt to get clear about the epistemic and semantic issues. Therefore, it seems best to consider these questions by expanding the scenario Putnam considers to a dialogue between a brain in a tank and an experimenter, although I will suppose the latter is a reformed and philosophically enlightened Genius. The considerations to be raised also, I believe, bear on several other recent attempts to refute skepticism, especially those which seek to restrict the range of counter-possibilities relevant to the truth of knowledge claims. Examples include Nicholas Rescher (Scepticism, 1980) and Peter Klein (Certainty, 1981).




Let us then suppose that after the events recounted by Bouwsma in "Descartes' Evil Genius", the Genius rode off, a broken Spirit. To fail at deceit, after having destroyed virtually everything in the universe, left him humiliated and alone. But let us suppose that a good consequence also ensued: he became less evil.

The Genius recreated the physical universe he had destroyed in his unsuccessful deception attempt, became material, and moved to San Francisco. Impressed by its success at demonstrating his failure, he took up philosophy. And, just to keep his hand in his ancient craft, but with a concession to the advance of science, he kept six tanks in his basement. In them were brains, afferent and efferent neurons connected via an interface to The Genius' new personal computer. Now you may think that the Genius must not have learned his lesson, and had lapsed back into his evil, if not deceitful, ways. Not so.

Professor Bouwsma had apparently shown that deceit was not possible in such a case, for what the experience-subject will regard as illusion and reality will necessarily be defined internal to his experience. To deceive, "you must learn the language of those you are to deceive" -- and then you must produce what your would-be victim can regard as properly denominated "deceit", "unreal" or "illusion". Those distinctions apply only amongst the objects of experience, and could never encompass the totality of a subject's experience. One can fool some of the people some of the time -- but one can't fool anyone all of the time. And this is an important philosophical discovery, one which eluded Descartes himself.

And so although he was now convinced that he must fail at deceit, the Evil Genius nevertheless prided himself on the high quality of experience which he was able to produce. And to further perfect his art, he kept those six subjects of his synthesized experience in his basement. It was a hobby.

The Genius induced in his subjects the experience they might have had had they lived in a world somewhat like our own, only rather prettier and more pleasant generally. The laws of nature were more lenient. It was Oz-like. The Genius controlled what was apparently happening to the tankbound subjects in his basement, but he did not control their reaction. The subjects were left to form beliefs and think and wonder and speculate as they might. The genius provided the most modern and up-to-date experience possible. Some days there was skateboard-experience, followed perhaps by Introduction-to-Philosophy-1A-experience. There was television-watching-experience, and psychology-and-neurophysiology-class-experience. Such a whimsical genius was he, there was even visit-to-the-Standard-Research-Institute-and-see-a-live-brain-in-a-tank-experience. And so it came to pass one day in the basement of that philosophical house (not long after a Phil 1A episode) that one of the envatted subjects came to wonder if it could know for certain anything about the world. Might I not, it naively wondered, be the immaterial victim of the activity of a malign immaterial spirit, the world of material objects including my body being but an illusion? Worse, might I be but a brain in a tank, with my experience generated by a computer which exists in a whole world of which I am unaware?

If that is even a possibility, then I can know nothing. Take any example of something that I think I know: say, that there are rocks. If I can know that there are rocks, then I can know that I am not being deceived, by an evil scientist stimulating my brain, into believing that there are rocks. But I can't know that I am not I am not being deceived into believing that there are rocks by an evil scientist stimulating my brain. So it follows that I can not know that there are rocks -- or anything else about the world.

The Genius, who regularly monitored the neural activity of his little subjects, was kept well apprised of their thoughts. In the face of such extreme provocation, and knowing full well the disastrous sort of philosophizing which began in such putative wonder, the Evil Genius, as before (Bouwsma, op. cit.), succumbed to the temptation to converse with this new subject of his experience generation. He was quite keen on administering a philosophical lesson, as many are eager to pass on hard learned truths. He also resented the suggestion that he was still evil. (And you see how far toward being a Good Genius he had moved: how evil it would have been to remain silent as his subject sank into darkest skepticism!)

The Genius, with the aid of his computer, appeared to his subject as a voice out of a burning bush. "Oh you of little philosophical sophistication! The world is, more or less, just as it appears to be, by definition. Sometimes I think that you don't know your own language! See those Munchkins over there? Those are one of the finest achievements of the experience generators' art. And since those are what you call "Munchkins", of course they exist!"

For several reasons, the subject (call it "Biat") was quite taken aback, and shifted uncomfortably in its tank. Biat looked at the burning bush and his exclamation appeared over the computer driven monitor speaker in the basement. "Who are you? What is going on?" The Genius, his flames reaching high when he became excited, explained the details of the tank and computer arrangements. He concluded, "But, of course, there can be no deceit here. Just innocent fun, like ant-farming."

The subject, who did not know what "ants" were, was confused by other matters as well, and not at all happy. "Do you mean that life is but a dream; my "world" is an illusion?"

"No, no! Don't be unphilosophico-grammatical! Your world is just as you suppose it to be. There are flowers and trees, people, Munchkins, chairs and frippets. For each of these terms (semantic consciousness raising requires semantic ascent) has an extension within your experience. Now you and I speak different languages: what you call "tree" and what I call "tree" are two completely different things (in fact, two completely different sorts of things): your tree is synthesized by a computer; mine is not. To emphasize the difference, we might do well to suffix our terms with an "e" for me and a "v" for you. I see(e) trees(e) whereas you see(v) trees(v). You speak Vat-English, which is syntactically largely identical to English, but semantically is completely different."

Understandably, Biat fell silent for awhile, and the Genius, glowing at first as a vehicle for Reason, became a little worried. "What's the matter? Aren't you reassured?"

"No," came the reply, "but where's the good in talking to you about it? We don't, according to you, even speak the same language."

Evil Genius: Hmmm. That's true. But wait -- that needn't interfere with our conversing. We can readily translate. I know perfectly well what you mean when you say "tree", and you know at least something about what I mean when I say "tree" -- a tree for me is something which appears to me much as your trees appear to you. And so there is really no trouble in talking back and forth.

Biat: Well, it does seem that we are able to talk, although I am not sure about your account of the situation. But that is the least of my worries. It certainly seems to me that my world is not real. It is all sham and data processing tricks! You live in the real world, and I am the victim of your deception.

Evil Genius: Now, now. You still haven't got it. Your chairs(v) are perfectly real(v). You forget that we mean different things by the syntactic entity, the mere string of letters or noise, "real". You mean by it "real(v)" which is used by you to mark out a distinction in your world, in your experience. It contrasts with those things which are "unreal(v)", "fake(v)", "illusory(v)" or "sham(v)" (and there is a great deal of important philosophical digging to be done at those roots!). Whereas, when I make the sound "real", I mean "real(e)", and mark out the distinction in my world. In either case, there can be no deception without detection, or at least the possibility thereof.

Biat: But it seems that we do mean the same thing by "real", at least to the extent that "real" used of objects such as chairs and trees excludes the possibility that a real chair "exists" only in someone's mind. But you've just told me that the chairs I have had to do with exist only in my mind. "Real" cannot apply to such a computer simulation of a chair. Chairs are necessarily public objects.

Evil: Nice point. You are coming along nicely! But the computer simulations are just as public as one might want: just as the mind must couple with the world, via sensory channels, to experience a chair, so must a mind couple with the computer to experience the simulated chair. Neither is a hallucination. There is really no important difference.

Biat: Public is public! While what I experience is not exactly a hallucination, it is not real, anymore than is an animated cartoon or a computer video game.

Evil: What you experience, including that chair, is public -- their are five more brains in tanks here, and all may experience these chairs (except for brain number 3, whose interface is down right now).

Biat: But you can't sit in a chair(v)! It's just an image. So its not real.

Evil: Ah, but you can sit(v) in a chair(v). And that is more than good enough for you, my friend. You just plop down in the nice overstuffed chair(v) over there [a disk drive whirred; a chair(v) appeared] and relax!

Biat: How can you tell me what I mean by "real" and "chair"? I know what I mean, and a chair you simulate isn't real, subscript me as you will. Which just goes to show that there is something wrong with this dual language view of yours. And I'm beginning to be able to articulate my worries. I don't see how, on your dual language theory, that you can account for the fact that I can understand my situation -- how can you ever explain to me what is going on?

Evil: Easy, I just did it a few minutes ago. In fact, I can show you your brain and the tank in which you are kept. Let me just switch you from my "Vat World" program over to the lab monitor television camera. There! As I pan the camera, for the first time you can see how it is down here in the subterranean Vat Cave. And as you compare your current visual experience of this view with your past experience, you can also see how good the quality of your synthesized experience has been -- almost like....

Biat: You were going to say "almost like the real thing"?

Evil: Er, Yes, but I see now that I need not have hesitated. For of course I speak from my standpoint. But you understand how it is.

Biat: Well, that's agreed, but the point is, how can you account for my understanding? You suddenly burst in on my experience and announce that I am nothing more than a brain that you are stimulating. By which I perforce understand, on your own theory, that I am merely a brain(v) which you are stimulating(v). But that is absurd! I can't be a brain(v), something internal to my own experience. I am a full-bodied person(v), five-foot-ten(v). I can only be a stimulated brain(e). So you could not explain to me what is going on if your theory of the dual languages were correct. But you can, Q.E.D.

Evil: Well! And that is exactly the problem which I had with the first would-be subject of deception I had, Tom. I never could explain to him what was going on, although I could see it all well enough. But then that was back in the fifties, and brains seem to have changed so -- quicker on the uptake. And now you seem to be able so quickly to see things objectively. It gives me pause....

Biat: And well it should -- but you think that I give you pause! That's nothing to the brain activity you provoke in this little tank. Please, I have had enough of staring at my own cortex. Switch off the monitor camera and resume the computer generated experience.

Evil: Alright. But I will continue conversing with you in my burning bush persona. Your point about not it being absurd to wonder if you are a brain(v) in a vat(v) is precisely the point made by Hilary Putnam. That was the point I was trying to make, when I first addressed you to lay at rest what you took to be skeptical thoughts. You thought you were wondering that you were a brain(b) in a vat(b) -- but you couldn't wonder about brains(b) and vats(b), because you had never experienced those. You don't know your own thoughts. I thought I would just pop in and let you know what you were thinking.

Biat: Needless to say, I am disturbed and far from convinced that my worries are groundless! ****** Your claim that we speak different languages seems to be based upon an exaggerated view of the importance of reference, and of the particular history of our use of terms.

Evil: The causal history of each of our words is important. Kripke has shown this for proper names, and Putnam has expanded on that view, showing how it applies to our thought about most things, including Gold and water. For example, you can imagine a place called "Twin Earth" which is just exactly like earth, except that instead of water, on Twin Earth there is a substance with the same visible properties but a different atomic composition. Their stuff is not H2O, it is XYZ. When Twin- Earthlings say "water", they are talking about XYZ, not what we call water. Suppose a Twin-Earthling is instantaneously transported to Earth, and upon tasting a dipperful of the contents of the River Charles, said: "Ah, that's delicious water!" Such a being would simply be mistaken -- not only in the gustatory judgement, but in supposing that what it tasted was water. By "water" it refers to a different substance. In this limited respect then, Twin-Earth English contains words that are syntactically the same as Earth English, but that differ in meaning. Your case is just more extensive.

Biat: It is true that our concepts of vat and brain are caused by experience of different things. But we have the same concept. A concept is a capacity to think about a certain kind of thing, and to recognize it and distinguish it from other things. But the capacity is not the same as the ability to exercise it. The exercise of the capacity depends upon contingencies of sensory abilities and connection with the world. For example, I might have color concepts, but have lost the ability to exercise that cognitive capacity because I have gone blind.

Evil: Well, perhaps once one has acquired a concept, one can continue to refer and think about things falling under that concept, despite loss of the ability to recognize and discriminate. But the point here is that you never were causally connected with things in the world outside the computer. You are in no position to continue thinking about them, for you have never been able to think about them.

Biat: You are surely correct that I must have been able initially to hypothesize the existence of brain and vat to think about them. But of course I am causally connected with my brain -- my real brain, not any brain image. And on your own account -- and demonstration, using the lab camera -- I have been causally connected with a very real vat(e) and a computer(e). The causal connections between my experiences and thoughts, on the one hand, and my brain and the things it causally interacts with form the causal bridge between the Vat World of objects(v) and the world of objects(e) -- a connection sufficient for reference on your own causal account.

Evil: I cannot deny that you have concepts similar to mine, and that that is the basis for our ability to communicate. And I can't deny that your experience is causally connected with what I know to be brains and vats. But somehow it is not the right connection.

Biat: So now that I show you that I can meet conceptual and causal conditions for thinking about brains and vats, you attempt to revise the requirements I must meet. You empiricists are always supposing that it is impossible to think about anything outside your own provincial immediate experience. I think that you have deceived me horribly. And I think that there is but one language at work here. Whether I am connected to cameras and microphones or merely a computerized simulator, I speak and think with a single language and understand my situation all too perfectly. My "world", the world I have known all my life, is but an Evil Genius production, a two-reeler that you can turn on and off at will!

Evil: Yes, I could do that, but don't let it trouble you. Nor should you expect to get any philosophic mileage out of it. Any Spirit of means has such power over what exists -- Descartes thought God sustained the world at each instant and could "off" it with a shrug. But the world would be nonetheless real for that power if unexercised. Such power over existence in the hands of an individual (He has the whole world in those hands) is not a sufficient condition for the unreality of that over which he has the power.

Biat: Perhaps, but it makes it all seem somehow less substantial.

Evil: Yes, well I like to think of myself as a little like Berkeley's God, creating a lovely little world (I'm not saying it's perfect, but the best of all practical worlds, given this old computer, my programming skills, and the high cost of RAM). Its a world that exists in your perception of it. Surely, you don't suppose that if Berkeley had been right (and some Berkeley Berkeley fans think he was), then the world would not have been real, do you? You couldn't accuse Berkeley's benevolent God of deception! [The taller of the two conversing San Francisco residents smiled triumphantly. He then leaned back from his microphone and, pushing a button on his console, delayed the sunset scheduled for Biat, for it would have detracted from the rhetorical force of his flames.]

Biat: With all due respect to the Bishop, I'm not at all sure that his metaphysics doesn't jeopardize the reality of the world. And even Berkeley is sensitive to the fact that objects, to be real, must exist outside at least minds of humans. And further Berkeley is not willing to sleep in the bed he made when it comes to an account of other minds -- he accords them substantial independent existence.

Evil: Yes, and so you qua brain in a tank have an ontological aspect which you, qua five foot ten inhabitant of your phenomenal world, do not.

Biat: But there is more to it than that. Heretofore I have carefully distinguished the tree I see from my seeing of it, and from any visual sensations I might have in seeing it. The tree and my experience of the tree are two different things -- or so I thought. Now you come along and tell me that the trees that I distinguished from my tree-experience don't exist.

Evil: No -- that's not what I tell you, I tell you. It's just that you were wrong to distinguish the tree from your experience of it. My message is closely related to Berkeley's. Indeed, the parallels are striking. For a tree(v), to be is to be perceived. Biat: Well, people who live in glass tanks shouldn't kick stones, so I'll try another style of refutation. The question is exactly whether one can sensibly identify a stone, say, or a tree, with one's experience of it.

Evil: Yes, and I think I can make out a very strong argument for my position. I shall present the argument in the form of a constructive dilemma with the conclusion that a being cannot be deceived by a brain-stimulator, such as myself, into falsely believing, e.g., that there are rocks. This should take care of your skeptical doubts.


1) Either there is a scientist generating one's experience or there is not.

2) If there is no such scientist, then no scientist is deceiving one into believing that there are rocks.

3) Now if there is such a scientist, then rocks will be other than what one now thinks they are: they will be complexes of sensations or sections of computer program or ideas in minds, or whatever -- I do not know what their essence would be. But rocks would exist, whatever their "ultimate nature". Therefore the scientist could not deceive one into falsely believing that rocks exist.

4) Conclusion: One can know that one is not being deceived into falsely believing that there are rocks by a scientist stimulating one's brain.

Since the dilemma is a proof that there is no brain stimulating scientist deceiving you into falsely believing that there are rocks, even if a brain stimulating scientist such as myself exists and generates your experience, you can know that you are not the victim of such deception. Furthermore, I can devise a like argument for any hypothetical transcendental deceiver and any belief about the world (that there are rocks was just a convenient example of a physical object). Thus your skeptical doubts must be based on fallacious reasoning.

The supposition that rocks could be the products of stimulation is of a type with Berkeley's claim that rocks not only could be, but were, ideas, mental entities. Berkeley actually embraces the possibility that worries the skeptic and shows that the conclusion the skeptic draws doesn't follow (a "skeptical solution" to skeptical doubts).

Biat: Excuse me, but isn't your argument a kind of "paradigm case argument"? Aren't those a bit out of favor now?

Evil: I suppose it is related to paradigm case arguments. But the considerations are bolstered by decidedly contemporary considerations from the causal theory of reference. I can make this clear by putting the argument informally as follows. Rocks are those things right there (and I can point at some). Those are what we call "rocks". Now just what their "ultimate nature" may be, we may not know (and it may be a senseless question). But that is neither here nor there. Even though we might have reason to change our minds about this or that property of rocks, perhaps even to the point of deciding that rocks were not material substances, but that they were bundles of sensations produced by a brain stimulating scientist, still we know that there are rocks. We cannot be deceived into falsely believing that there are rocks; we know that there are rocks.

As presented, this argument is indifferent to whether we hold that transcendental metaphysical theories are meaningful or not; the argument concludes that in any case we cannot be now being deceived about the existence of rocks. We all can know how the term "rock" is used; from the fact that it has an extension it follows that there are rocks.

Now it certainly is not the case that you supposed that rocks are bundles of sensations. But then it is certainly not the case that men always thought that the sun was a ball of hydrogen undergoing fusion, nor that rocks were bundles of moving charged particles. These were surprises! But it was the sun, that thing over men's heads during the day which they called "the sun", which was found to be other than as it had been thought. There never was any question of finding out that the sun did not exist after all just because it wasn't a ball of fire (I am supposing that fusion is not fire; the sun is too hot to support combustion, and fire is a sort of combustion).

And as far as transcendent, metaphysical accounts of ordinary things are concerned, these have often changed without existential consequence for the description of daily affairs. Even if Berkeley's account were the correct metaphysics, there would be rocks, as he stoutly asserted -- it is just that they would not be what materialists and (perhaps contrary to the Bishop's own view of the learned) the educated take them to be. And so you see that the account I have given you of your existence is really not a great deal different from Berkeley's account in respects relevant to the skeptical argument: rather than being ideas in God's mind, rocks are ideas in your mind with causes in my equipment.

So the argument for the dilemma goes. Sorry to be so long-winded. The upshot of it is that I cannot deceive you about such a matter as the existence of rocks, so the skeptical argument fails.

Biat: Well, it does seem to me that the dilemma which you advance is certainly a valid argument. And clearly its success in refuting skepticism depends upon the truth of the claim that even if one's putative experience of rocks is generated by a scientist, he cannot deceive us as to the existence of rocks. It is clear that things are not as I had supposed them to be. But it is also clear that the difference, transcendental as it is, will make no difference whatever to my experience. While I continue to believe, we will suppose, that there is a real transcendent world, in which world I and you the scientist and perhaps other tankbound persons like myself exist, my experience is related to that world in a way quite other than that in which I thought it was.

Evil: Nevertheless, tomorrow you will face the same general run of experience and face your apparent companions with whom you must continue to communicate if you are not to be subjected to the unpleasantries befalling the anti-social. But behind the scenes my computer will run as usual, following the general patterns imposed by its program, a pattern which you have come to know and which you call "laws of nature" and use to anticipate your future experience.

Biat: So you assure me. But given all this, are there rocks or not? I do not think that it is clear at this point, despite your argument, just what one should say. It has been revealed to me that all of my sensory experience is generated by someone, you, stimulating my brain. And our question is whether I ought say that there were such things as rocks without considering what there is out there in your world (where there may be rocks). In particular, ought I say that rocks just turned out to be rock-sensations, which certainly do exist?

I think the answer to this question must be negative. The main reason is the presence of the extreme ontological contrast between my experience and, on the other hand, the existence of you the scientist and my brain. You and all the objects in your world, including my brain, I suppose to exist, although I do not know what many of those objects are. But if they exist, if that is what it is to exist, and if they are distinct from my sensations, then rocks can be distinguished from rock-sensations. I know what it would be for rocks to really exist, that is, for it to be the case that there exists you, Evil, my brain, and rocks (perhaps with my brain resting in its tank on a large rock). The rocks which I thought I saw and threw and sat on do not exist, are not really objects at all, although real rocks may exist along with my brain and you in San Francisco.

So I claim that the third premise of the constructive dilemma is false; rocks cannot be discovered to be identical with sensations generated by you or anyone (nor can they be identified with just spontaneously occurring sensations).

The third premise of your dilemma involves the claim that a stimulating scientist could not deceive the beings whose brains he stimulates with regard to the existence of rocks, for rocks would exist as a result of his attempted deception -- rocks would be the product of his stimulation. (I believe this is the gist of your and Bouwsma's argument.) But this is not plausible. The discovery of you, a brain stimulating scientist, compels me to consider another more inclusive viewpoint on the matter. From your standpoint, my experience is not of any existent objects (although my experience is generated by existent laboratory equipment). There are no objects corresponding to my rock-experience (although there may be objects corresponding to my concept of rock in the world in which you and my brain exist -- but such objects are not appropriately causally connected to my experience). I can appreciate this fact even though I am the one having the rock-experience, just as a victim of hallucinations may appreciate the fact that he is hallucinating without thereby diminishing the vividness of the hallucinations. Finally, as I have mentioned, once aware of what is going on, I must leave open the possibility that there really are rocks somewhere, and that in fact the tank in which my brain rests itself is supported on a large and handsome rock. But leaving that possibility open is clearly inconsistent with identifying rocks with the sensations I have in the tank. Tanks cannot rest upon rock-sensations, only upon rocks. Since I must leave open the possibility that rocks exist apart from my rock-experience, I cannot identify any part of my subjective experience with rocks. Real rocks cannot be sensations; what I experience in the tank are not really rocks.

Evil: I'm not so sure about that. Mooreover, it seems that an objector might still argue as follows. He says, holding up his hands, "these are hands". Could it be that he is mistaken about this? He might continue: "The word `hand' was defined ostensibly by these things which I am holding up, hands. I therefore know that these are hands."

Biat: Well, even if we suppose that he is in fact holding up his hands, it seems clear to me that neither we nor he know that this is the case. We could be, for all we know, now being deceived into falsely believing that he is holding up his hands (by powerful deceivers such as yourself). Similarly, he cannot know that that which he is holding up, if he is holding anything up, is what was ostensibly defined as a hand.

Suppose that your objector then once more holds up his hand and says this: "I hereby define this (he looks at and shakes his hand) as a mitt. Now I know that there are mitts."

With regard to this claim, a skeptical objection can still be made. Our man cannot know whether mitts are real objects or not. We might compare his situation to that of two linguistic philosophers on a working holiday in the desert who both see a mirage, not knowing that it is a mirage. The one says, "I hereby define that (he indicates the apparent position of the mirage) as an Asis".

Now a mirage is not an object at all: we say that one is "seeing a mirage" when there is no object there to be seen; it is a trick of light. Hence if our man does not know whether he is seeing a mirage or not, then he does not know that an Asis really exists or not, even though he has in some sense ostensibly defined "Asis". Similarly, our hand waving man has in some sense ostensibly defined the term "mitt", but it does not follow from that fact that mitts actually exist nor that he can know that they exist.

So even if terms are ostensibly defined, we cannot know that they have an extension, if we limit the extension of terms to things that really exist. It seems to me that your constructive dilemma counterargument to my common-sense realism fails.

The Evil Genius grew angry and glared at Biat's busy hemispheres. "Perhaps a little sensory deprivation will help you to "see" things my way," he threatened.

Biat: I certainly have no wish to make someone with your abilities (and particular power over myself) angry! And I have grown attached to my synthesized experience, enjoying much of it -- although I can't imagine exactly how I can go on living in the future as I have lived in the past. Can I possibly take a computer synthesized romance seriously? As Robert Nozick points out, dear Evil, there is something basically deficient in being confined to synthesized experience (although he speaks, I think wrongly, of a "man-made reality" and a "deeper" reality -- reality concerns what is really going on, not what a machine makes appear is going on). But that just shows that it is certainly possible for us to understand what it is to have one's experience synthesized.

Be that as it may, I had merely hoped that the reasons I advanced for my position -- merely explaining my dismay at the discovery of my true situation -- would convince you! Don't you see:

First, the dual language theory will not work because were it true you could never have explained to me that my experience did not come about as I thought it did. Yet you in fact had no trouble explaining to me my situation, and almost everyone can understand what it is to be a brain in a tank -- including someone, such as myself, who has been given reason to seriously entertain the possibility that he is in fact a brain in a tank. If your theory of language meaning were correct, I could not, as I do, understand what it is for things to exist which transcend, or exist beyond and apart from, all my experience. And, further, I do understand what it would have been if you had never made yourself and my real situation known to me -- I still would have been, unbeknownst to me, nothing more than a brain in a tank.

And so it seems you were deceived in supposing that you had to learn my language, as Professors Bouwsma would have it, in order to deceive me -- or even that I had a language different from your own for you to learn. From sometime before the Trojan Horse, and perhaps before the existence of language, there has been deceit -- from the first feigned love or feigned innocence concealing aggressive intentions, to take just two examples.

Second, while it is certainly true that I made out the real/unreal distinction internal to my experience prior to your making yourself known, it does not follow that I am committed to continuing to hold that the "objects" of my synthesized experience which I formerly thought real are real. Thus while it is true, as J. L. Austin notes, that "real" is "firmly established in, and very frequently used in, the ordinary language" I use every day, and thus has "a fixed meaning", it does not follow that its extension is fixed at all. I am not contrasting the real with some other sort of thing, (the artificial, synthetic, specious, make-believe); I am contrasting flowers with the appearance of flowers. I am contrasting the case where there are flowers with the case where there are no flowers but, because of the activity of a transcendental agent, there nevertheless appear to be flowers. You have revealed to me that all my experience is of the same type -- the part I thought real and the part I thought unreal. I used to think that most of my experience corresponded to and represented the welter of objects which I encountered in meanderings about the world -- but today I find out that I have traveled nowhere and encountered no objects -- except you, Genius. You are the first real object of experience for me. And so I see that I have not experienced real objects in the past. It is similar to the distinction between witch and non-witch, which was revised as to extension with the ebbing of paranoia. Just because a distinction is made and both poles are thought to have extension does not mean that they do.

Third, the reason I have come to hold that all my previous experience was not of real objects is this: fundamental to my understanding of the world is the distinction between objects and my experience of and thoughts about those objects. So basic is this distinction to my understanding that it seems that I cannot identify objects with my experience of them, as you and Berkeley would have me do. The same Professor Putnam that you have relied upon in arguing that we speak different has suggested in other writings that the reason we cannot be mistaken about our pains and other sensations is that sensations have "appearance logic". However that may be, it seems clear that objects have the complementary logic -- they cannot have "appearance logic", they must exist apart from our experience of them. The contrast between the objective and the subjective is basic to my understanding of the world. Real objects are more than just "permanent possibilities of sensation", as John Stuart Mill thought. A recurring nightmare fits that description. Objects are those things which can exist apart from my perception of them, and indeed apart from anyone's perception. There is no logical dependence of objects upon sensation. If a person cannot understand how it is logically possible for objects to have existed before there was sentient life on earth, that person does not have the concept of object which I have. Objects are necessarily not identical with experience.

Our experience of objects must arise in certain ways, as in perception, and it seems it must, as a matter of the essence of objects, be the case that we may always be mistaken about the exact nature of the objects we experience. Perhaps it is the possibility of error which is a necessary condition for meaningful understanding of the world, and not, as positivists have held, the possibility of knowledge.

Once I understood that I was a brain in a tank, I saw that the "brains" I had "perceived" heretofore were but computer simulations, not objects at all, and hence not real brains. And the same goes for all the other "objects" of my experience -- they were not really objects at all.

Fourth, it seems that if the constructive dilemma you presented would be too sweeping in its ontological consequences -- for if that argument were sound, what should not exist? If rocks could exist, although but bundles of sensations, then why wouldn't Santa Claus exist, although but words in Santa-Claus-stories? And even the round-square has a name (or description) to be identified with. But surely even your sense of reality is too robust for that; things cannot be identified with stories of them, computer programs simulating them, their names or description, nor sensations or ideas of them.

Fifth, I have shown that I am causally connected with objects in the real world, and that I have the conceptual capacity to refer to them.

Sixth, I ....

But the reasons adduced so far proved conclusive -- for poor Biat. The Evil Genius threw his well worn copy of the Philosophical Investigations at the tank and Biat fell to the floor and silent. The Evil Genius proceeded on a rampage, using his great powers to cause such a rash of hallucinations as no one had seen since the golden days of Haight Ashbury. How it will all end, no one knows. People stand in street gutters waiting for a BART subway train. At Berkeley, a distinguished visiting British philosophy professor lectured stark naked to an amazed audience while insisting that he was fully clothed and knew himself to be so! Nearby, Vat-English has been proposed as an alternative to Ebonics for most classroom instruction. Last, a word of warning: many have been seen staring into space for hours, nodding and eyes scanning, apparently all the while supposing themselves to be reading.