Not much company, however, as far as I know. Most philosophers seem to think it quite easy to rob the will of some freedom. Thus Elizabeth Anscombe, in an essay called "Soft Determinism," appears to suppose that a man who can't walk because he is chained up has lost some freedom of will. He "has no 'freedom of will' to walk," she says, or, again; no "freedom of the will in respect of walking."⁴ "Everyone will allow," she says, "that 'A can walk, i.e. has freedom of the will in respect of walking' would be gainsaid by A's being chained up."⁵ And again, "External constraint is generally agreed to be incompatible with freedom",⁶ by which she seems to mean: incompatible with perfect freedom of will, because incompatible with freedom of will to do, or freedom of the will in respect of doing, whatever the constraint prevents.

. The horrid tribe of "soft determinists" are supposed by Anscombe to allow, of course, what "everyone will allow," namely that a physical impossibility of, say, walking does restrict one's freedom of will if it comes of some external constraint. What they won't admit is that every such physical impossibility, from whatever cause, external or internal, is equally fatal to some fraction of one's freedom of will. Anscombe, on the other hand, has "never thought that freedom was compatible with physical impossibility,"⁷ That is, with physical impossibility of any derivation, whether from chains or from brains, so to speak. I hope I am not misunderstanding her. At any rate I am not misquoting her. She believes, as she says, that a 'can' of freedom" which holds in face of physical impossibility is pure nonsense.⁸

Well, it's awfully difficult about brains. I won't really get into our brains, this evening. But I do want to dispute, first, what Anscombe thinks "everyone will allow." I don't allow it. I don't see (do you?) that my freedom of will would be reduced at all if you chained me up. You would of course deprive me of considerable freedom of movement if you did that; you would thereby diminish my already unimpressive capacity to do what I will. But I don't see that my will would be any the less free. What about my "freedom of will to walk," you will ask (or perhaps you won't, but there the phrase is, in Anscombe's essay); what about my "freedom of the will in respect of walking"? I reply that I don't understand either of those phrases. They seem to me to mix up incoherently two different things: free will, an obscure idea which is the one I am after, on this expedition, and physical ability to walk, a relatively clear idea which has nothing to do with free will.

If instead of being perfectly free in the matter of whether, all things considered, to walk over there or not, I were absolutely restrained somehow from deciding that option in favor of walking, nothing would follow as to my physical ability or inability to walk. And if, having decided to walk over there, I found that I could not, after all these years of being carried about by my devoted pupils, nothing would follow as to my freedom or unfreedom of will. Not according to my idea of free will, anyway. I wonder if Anscombe is operating with another one, or what? How can it seem so clear to her that every physical impossibility of doing this or that reduces one's little portion of free will, as if only an impossible creature for whom nothing was physically impossible, not even eating Chicago, could satisfy a clear necessary condition of perfect freedom of will? What about God, for example? He can't eat Chicago any more than I can, after all.

But here someone might reply (in effect) that in God's case that limitation ought to be counted as "grammatical" or "conceptual" or "logical," not physical, and therefore as nothing against His perfect freedom of will. And this reply would make for a possibility of reconciling my idea of free will with what Anscombe says, up to a point. Why am I not inclined to think that my will is any less free than it might be, in that I can't if I like eat Chicago? In one view of me, what gets in the way of that meal isn't, of course, "logic"; it's my size, for example. I'm simply not big enough. But perhaps, in thinking as I do "that's no skin off my free will," I am tacitly treating my size as given and treating the impossibility of eating Chicago as geometrical, in my case, rather than physical. If so, and if one were to allow a certain legitimacy to that way of looking at the matter, one might allow that I don't have to think myself deficient in free will because I am too small to wolf down whole cities and too big to live in a shoe. One might similarly make room for my disinclination to count my will as any the less free for my being in no position right now to throw myself into the Seine. That is, one might agree that various trivial consequences of location -- of what H.H. Price once called the misfortune that we are not ubiquitous -can be written out of the question how free the will is as in a (very) broad sense logical embarrassments, not physical impossibilities of the kind that straightforwardly restrict our freedom of will, like the physical impossibility of walking if you are properly chained up.

But this line of reconciliation could go only so far, because I still don't see that my freedom of will would be affected even by chains. Especially obviously not if I didn't know about the chains: Suppose I am chained up so that I can't walk, but don't yet know it. I deliberate about what to do next and decide on a little tour of my cell. Then I discover that I can't walk. They've chained me up, the swine! But wasn't the part of my "will to walk" in these events antecedent to that discovery? Do I have reason to think not only, "They've chained me up!" but, "Good God, they've been tampering with my will!"? No, I don't. Of course, one wants not just freedom, in a will, but efficacy too, so to speak. One might well find the possibility of embracing one's fate, however authentically, unconsoling, and wish instead for some effective power of resisting it. But all such power is a different thing from free will. I am inclined to defend this intuiti

on as follows: What we propose to do is up to us, if our wills are free. But what the world will make of what we do is of course up to the world. (I think I am echoing Brian O'Shaughnessy here.) It's nothing against my freedom of will if I earnestly move my feet in the right sort of way but it isn't in the world to make walking out of my so moving them because, as it happens, I am upside down with my legs in the air. It's nothing against my freedom if will if I "can't walk" because the floor will collapse, or because it has been arranged for me to explode if I shift my weight. These difficulties in the way of my actually getting any walking accomplished are on the side of the world, not the will; and they don't in themselves interfere with the will's part in walking (that is, in these cases, its part in deciding and trying to walk). They don't affect its freedom, therefore. Where there's a will, there just isn't always a way. After the christening, the ship majestically slides to the bottom of the harbor, and so we haven't managed after all to launch her, though that was our intention and our wills were in perfect working order. We freely did what we could do, and hoped that it would turn out to have been a launching, so to speak; but the world declined to cooperate. It does cooperate quite often enough for unconditional intentions and unconditional decisions to make means "I'll be there." Sometimes, however, the car won't start. That's life. But it's much too remote from the springs of action to pass for a shackle on the will.

It's no accident, I imagine, that in Anscombe's example the man is chained up. If you chain somebody up, you seem to be getting closer to where the will is than if you build a high wall around him at a radius of ten miles from where he sits reading Spinoza. You seem to be working in toward the connection in him between will and world, which might strike you as a good point at which to try disabling his will itself. An even better idea might be a judicious injection of curare, if it works as I've been told it does. There he will be, fully conscious but physically so relaxed that he cannot move a muscle. One might hope, even, to have incapacitated him from trying anything physical, and thereby definitely constrained his will, because one might hope that it was right to suggest that one logically can't try to perform a "basic" physical action, and that trying to perform any other, nonbasic, physical action must consist in actually performing one or more that are basic.

But that suggestion was wrong, it seems to me. I don't see why the prisoner incapacitated by curare shouldn't be trying and failing to move, if he isn't too far gone to try anything. And I therefore don't see that we've managed to subvert his freedom of will. (I'm out of jail and we're working on somebody else, here.) We have radically disabled this man from exerting his will to any physical effect. Indeed, we have gone as far as possible, without killing him, in the direction of as it were dislodging his will from his body, which we don't want to do. We don't want him disconnected, as from an odd piece of furniture: that (if it were possible) would alter his situation "logically," again, which isn't what we're after. We want to get at his will. But this business with the curare hasn't taken us far enough. We have in this (for all we can tell, still defiant) prisoner a limiting case of physical disability, which does involve a sort of disability of the will too. (Why not say so? His will is physically powerless, one might say.) But do we have what (I'm supposing) we wanted: a man not only powerless and flat on his back, but robbed of some freedom of will? I don't see that we do. Perhaps he loves us by now, and can't wait to show it. Or on the other hand, perhaps he has other plans.

I don't mean to be thrilling about man's unconquerable will. Or at any rate, not very thrilling. We can just be knocked out, after all, without bothering about our freedom. And even cold sober I sometimes find that I have done quite complicated things "automatically," so to speak, an automatism that is most vivid when it goes awry: What on earth am I doing in the bathroom? Oh, I meant to fetch a certain book from the (adjacent) bedroom, of course. But I have nevertheless gone straight to the bathroom, not in unfreedom of will but in ridiculous absence of mind. No doubt this sort of thing happens to me more often than it did to Marco Polo, and even I don't find that I have in this curious way gone to San Francisco. But there is a portent in everyday automatism, all the same. Why shouldn't it be brought under control and greatly extended in range, by horrible new techniques? This method of getting people to do what you want would just bypass the will, as far as I can see. Or one can leave the will alone and get excellent results even now, by manipulating belief instead. Convince me that your enemy is the Antichrist and I will no doubt behave satisfactorily, in full freedom of will. How else should one behave toward Antichrist?

So I'm not proposing a round of self-congratulation. But all the same, isn't there a distinction between something obscure that might reasonably be called "freedom of will" and every kind of freedom to work one's will, to do as one will, to have one's will or way, so to speak? The living dead might unfortunately have lots of freedom of action. Or freedom of movement, anyway. And a human being of intact free will might have almost no freedom of action, indeed no freedom of physical action at all. So it seems to me.

And what goes for physical impossibility goes for physical necessity too, in whatever sense there is any such thing. If I'm forcibly carried to your rotten garden party, or deposited in the middle of it by a defective parachute, I'm there against my will, no doubt; I'm not there of my own free will. But then, I'm not there "of my will" at all, am I? It's your will that has been done by these brutes; or it's nobody's will, in the case of the parachute. It's none of my doing that I have shown up or appeared at, or come to, your party (if those phrases can be Griced, as perhaps they can). I haven't put in an appearance at your party. They can't say I attended it. And so they can't say I attended it of an unfree will, so to speak, or in unfreedom of will, or any such thing. My will was free as ever, as I landed in your garden. Why else was I so angry?

An annoying ambiguity of the noun "will" must be dealt with, however summarily. Consider: "She went to work with a will" (what will? well, a will to please, or to succeed, or the like); or "He has a fierce will to live"; or the cartoon in the New Yorker of a televised message card reading "Please stand by. We have temporarily lost the will to continue." In these contexts "will" is a noun like "wish," though for some reason it has no such common plural as in "They ignored my wishes." (One doesn't have wills. I can't think why not.) But a will, in this sense, is a will that... or a will to.... A will in this sense may, like a wish, be strong or not so strong. Or one may have no particular will in some matter. ("What is your will, majesty?" Answer: "For God's sake leave me alone. Just do anything. This crown is killing me.") Is a will in this sense "free", or "not free"? One might say so: "What is your will, sire?" Answer: "Do as this man with the grenade says. But remember: that's not my free will." This dark utterance suggests the guess that his "free will", and therefore his real will, is something else. Perhaps his eyesight has improved and he can see that it's only a pineapple. What a puzzle! How can he have no "free will" in the matter of this demented peasant? Perhaps he hasn't made up his mind whether to hurt the fellow's feelings or not, and is temporizing. Or of course he may think it is a grenade. But then what does he mean, it isn't his "free will" that we do what the fellow says? Perhaps he means that it wouldn't have been his will in pleasanter circumstances, and we're to remember that. But in the meantime, if it is his will, we'd better do it. He gets so furious when he's crossed. And so forth.

But this talk of a will in some case that's not one's free will is very opaque. I don't know what to make of it, really. The word "will" in the usual sense of the expressions "free will" and "freedom of the will" is different: it's like "intellect," "mind," "heart," "imagination," not like "wish." "We know the will is free" doesn't invite the question "The will to what?" Of course the idea of freedom of the will is a hard idea to get hold of. But at least one has it, and needn't make it up. One isn't supposed to believe in the will any more, perhaps, not even as one believes in the mind or the imagination. But on the other hand, we say "She has a strong will," "They broke his will," "He has no will of his own," "He didn't do it of his own free will," "He has brain power but no will power," and so on. So we do still believe in the will, however vaguely. And why not? Vows of ontological commitment are another and more dubious matter, but free will is what we've got if the will is free, as of course it is. I don't exactly know what any of that means, but I don't know how to doubt that we've got free will, either. The will is free, whatever that means. Indeed, the ulterior motive of this paper was to discover something about what these expressions do mean by discovering why, hopelessly American as I am, I am so strongly inclined to agree with Descartes' prima facie absurd estimate of how free the will is. It seems to me not a grandiosity but a simple truth. Maybe so simple that there's nothing in it, in a sense. But if so, that might be instructive.

Suppose I am right so far. Chains, or the curare of my imagination, can't get at the will directly and can't in themselves affect its freedom. I would say the same, of course, about other sorts of "external" limitation on what one can do. (External to the will, I mean.) Lack of opportunity, for example, or lack of means. I would shoot him, but where is he? There he is, but now where's the gun? Incompetence: here's the gun, and here's the safety catch, but I can't remember how to release it. And so forth. All that kind of thing, not all of it "physical," is nothing to do with free will. Or at any rate, it can't affect my free will directly. Suppose you were to agree. But now, can't that kind of thing constrict the operations of the will indirectly, nevertheless, by way of one's knowledge or conviction that, for whatever reason, one can't do this or that? Never mind the chained man's physical incapacity. What about his recognition of it? Knowing or believing that he can't walk, how can he so much as try? And trying things is a function of the will, surely: one in which it can be partially paralyzed after all, it seems. Indeed, it needn't even be a fact that (say) I can't walk. Convince me that I can't, by suitably arranged illusions and lies, and there go some former freedoms of my will: my freedom to decide to walk, for example, or to choose to walk, or even to try to walk, and therefore my freedom to decide or choose to try to walk. I can, of course, still wish I could walk. But if wishing is an operation of the will -- as perhaps it is, in contrast with perfectly passive desire -it isn't much of one. So: convince me that I can't walk, and sooner or later I do lose some freedom of will. Or do I?

I think not, again. In the familiar, obscure sense of "free will" that I am trying to get into clearer view, here, it seems to me no deficiency of free will that one can't just up and go against knowledge and belief (insofar as one can't) because that "can't" is, again, not psychological, or metaphysical either, but "grammatical" or "conceptual" or "logical." (I'll just go on saying "logical," if I may, as if I didn't know any better.) It isn't that the will is hobbled by the prospect of an impossible project. It's that "I know I can't do it, but I've decided to do it anyway," for example, is either a figurative way of speaking or a kind of nonsense. Perhaps it is even a contradiction.

Well, perhaps it is, and if so it might be conceded that what gets in the way of deciding to walk when you know you can't, or think you know you can't, is in the language, not in the will. But there's no such contradiction in trying to do the impossible, knowing one will fail, as far as I can see. So why can't I (as surely I can't) try to jump over this hotel, say? I can't even try to do fifty consecutive pushups, I think. Not really. Why not? Isn't it that my will is in these matters disabled, in a way that has to be seen as robbing it of some freedom? I hope not, but I confess that I have sunk into a little swamp about trying to do the impossible, and don't know how to get out of it gracefully. It isn't an enthralling swamp, so I'll just flounder on, here. Consider the following speech: "Of course you won't be able to do it, we know that, but try. You promised to try. You will certainly be unable to do it, at first, but if you don't try and try again, you will never be able to do it, whereas if you keep on trying you may eventually succeed." That seems all right. So one can knowingly try to do the impossible -- the not yet possible -- under certain conditions. Can one, even, try and try again, although one knows perfectly well that the attempt will never succeed, without even hope against hope that it might? I'm not sure that one can't desperately do that. Indeed, I think one might even do it cynically, to please the authorities. They will be very displeased, say, if I don't at least try to lift a certain sacred boulder, which is obviously much too heavy for me to lift. So I try. That is, I grasp the thing and strain hard, thinking, "When will these savages learn?" Am I only pretending to try? I don't see that one has to say so. Not in every case.

But then how is it that I can't (if, as I think, I can't) try to jump over this building?⁹ If I went sufficiently mad, I could. If I went mad, I could try to fly around it by flapping my arms. Suppose I am pressed to admit that after all trying to fly is a perfectly conceivable little act of the will, which needs no other description. Birds do it. Maybe even bees. So let's do it. But of course we can't. Why not? Because the logic of trying excludes trying to do what you know you can't do? Unfortunately, it doesn't. So what's my problem about trying to fly, or trying to jump over hotels? Isn't it a blot on my freedom of will? Shouldn't I just admit that?

I don't admit it. And I think I know, more or less, how to handle cases like trying to jump over this building, as follows: I could go outside, get down into a crouch, and jump up into the air, thinking (perhaps), "Well, it isn't a necessary truth that I won't somehow be carried on up and over the building." But that wouldn't be trying to jump over the building, as I might try to jump over this lectern. What would? Nothing I can think of, in these present circumstances, including of course the circumstance of my present cognitive "state." It isn't that it's an iron law of the will that someone in that state can't possibly try, no matter how he strains, to leap tall buildings. It's that nothing I can think of to do this evening would be rightly described as trying to jump over this building, in a straightforward sense, unless for example my beliefs were to alter or go very dim. It's the same in principle as if I were invited to try to wiggle my teeth. Given what I vaguely think about human teeth, as contrasted with ears, I am not inclined to count anything I might conceivably do now as "trying to wiggle my teeth." (I might arrive at "willing my teeth to wiggle," as one could say: a would-be magical operation like willing a table to rise, or willing one's arm to rise. But that would be different. And it doesn't work on arms any better than on tables.)

One can, often, try to do what one sees no point in trying. (Or maybe one can't be bothered.) But the sort of case I've just been discussing seems different. It seems over the edge: a logical edge, as I am stupidly putting it. In my present cognitive state, that is, I'm not inclined to count anything I might succeed in doing as "wiggling my teeth," "jumping over this building," "flying around it by flapping my arms," and that's why I'm not inclined to count anything I might do as trying to do any of those things; and that's why my prima facie incapacity to try to do any of them isn't an incapacity, and doesn't show any lack of freedom in my will.

Will that do? Perhaps it might for such exotica as these. But I know exactly what I would count as doing fifty consecutive pushups, or reciting the first four pages of The Wings of the Dove. I can even make what would in someone else's case be a beginning on just those projects. One pushup, perhaps. I'll spare you that. The first four words of The Wings of the Dove. I'll spare you them, too. So why can't I so much as try for fifty pushups; for four pages? I want to say: it's not that I can't, it's that those descriptions of me, as I am -- "He's trying for fifty" and so forth -- would be inept whatever I did. And so for the chained man: either he can try to walk, or if he can't, it's not that he can't, any longer. It's that in his present cognitive position and state of mind that description of him would be inept whatever he did. Trying to walk isn't a perfectly imaginable little act of the will, separate and distinct from all belief and contingently blocked in him by belief. Folk psychology is my favorite kind, and I hope it never withers away, but one must not think of such folk-psychological items as trying in that way.

Well, again, suppose you were to agree, though God knows why you should, that there is enough for my purposes in this admittedly messy idea of logical limits on the will that ought not to prejudice our estimate of its freedom even though they are not logical limits of the comforting type that rule out (not only in my case but in practically any case, even on Twin Earth) trying to stand at attention while lying down. You might then agree that chains and so forth can't bind the will either directly or indirectly. Not by way of laying on it powerful spells of knowledge and belief, at least.

Of course, chains and the like can in a certain sense undermine apparent choices. But it isn't the freedom of those choices that chains can abolish, it's the choices, in an objective sense. Determined to remain seated during the national anthem, say, I may think I have done so and then discover that I was firmly stuck to the seat of my chair anyway and am now marooned in the balcony. So I did not, after all, "remain seated by choice," unpatriotic as my intentions were. Of course not, because in the relevant sense I did not, though I thought I did, remain seated. Stuck to my chair as I was, I had no choice of remaining seated or standing, and didn't do either of those things. In this use of "choice," mere physical circumstance can, of course, deprive us of choices, whether we know it or not. But, again, that's nothing to do with the will's freedom. Stuck in the balcony, am I? Well, then, alternatives between which I might have chosen have gone glimmering. Both of them have. They are out of logical range, so to speak again. The thought, as the auditorium empties, "Well, I'll just stay here in the balcony, then, that's what I'll do" isn't even an expression of amor fati. It's just fatuous.

Where am I, now? Some cases have perhaps turned out not to be cases in which there is any lack or loss of free will. But I must of course consider the obvious objection that I have been looking much too far from home for unfreedom of the will. What if it were urged on me that the will's freedom (if it has any) is diminished all the time in everyday life? Every decisive, compelling reason to make one choice rather than another reduces one's freedom of will, someone might object. Luther said, and he meant it, "Ich kann nicht anders." We say all the time, and not always in bad faith, "I'd love to, but I really can't", "I'm afraid we must go", "I couldn't possibly, she's my sister", "I can't help it, I gave him my word," and the like. Some of us do, anyway. Shall I say we never mean it in good faith? Of course we do, often. Shall I say we mean it, all right, often enough, but it's never true? No, again. It's often true. Well, then mustn't I say that when we do mean it, and it is true, we are reporting a constraint, deep inside, on the will itself, which really does deprive it of some little freedom, or even of some big freedom? Why do so many of us in a way envy criminals, especially if we haven't intimately known any? Isn't it their freedom? And not only, not even principally, their freedom of action, such as it is. (Stone walls do, after all, a prison make.) Isn't it their freedom of will that's in a way enviable, or at any rate envied? They may not get far on the wild side; but they walk there, when they can, by choice. Aren't we incapable of that choice? So much stands in the way of it: nervousness, habit, addiction to our comforts, including our spiritual comforts, and to our accustomed pleasures, including such cheap moral thrills as pointing out to the waitress that she has undercharged us. And reason bars the way, too, of course: the awful majesty of duty, or just duty, majestic or not; obligations; other engagements; ordinary prudence; common sense; the usual constraints of cognitive sanity. But in default of reason, wouldn't desire and aversion and so forth do, to keep the will hogtied? We may like to think that if we threw off the restraints of reason we could do appalling things. But could we? Aren't most of us probably too squeamish, if nothing else? And anyway, can one throw off the restraints of reason?

Well, the first thing to be said in the matter of what we rationally speaking "cannot" and "must" do is that these modal laments and excuses and so on are standardly not introspective autobiography. "We have to go" isn't a report to the effect that we are about to be frog-marched out of the house by invisible forces. "We really must" isn't falsified if in the end we don't, any more than "You must believe me" is falsified by your not believing me.¹⁰ These "can't"s and "must"s belong to the system of what's getting considered, not to the considering of it. "I must go" is false if, for example, the pressing engagement I have in mind was actually for yesterday, not today. What's wanted for the discovery that after all I can stay a little longer is (for example) a look at the calendar, not a more searching look within, much less the observation that I do stay a little longer.

To be sure, a man who keeps saying "I really must go" and nevertheless stays on until he is physically incapable of going may be suspected of insincerity; but what he kept saying wasn't therefore false, as may be seen by recasting it in the second person and putting it in the mouth of his host: "You really must go!" There is a kind of contradiction in acting against decisive reason to act otherwise; and if you do it often enough you're impossible. But impossible people, given for example to sarcasms like, "Oh can't I? Just watch me," aren't impossible to find. We, of course, aren't like that, except in some of our worst (and, conceivably, some of our best) moments. But unfreely not like that? As if it weren't up to us to act, or not act, as there is compelling reason to do? As if so acting were otiose, and we might instead just wait for the compelling reason to work? I am inclined to say (but I know I have barely begun to argue) that even the most irresistible case for doing a thing isn't irresistible like that. How could it be? It's a mere case. Can't one always not do what one must? I am on duty; I can't go on fooling around in the linen closet like this; I must make my rounds. I absolutely must. And yet of course from another point of view I needn't at all. I could just stay here, couldn't I, and my station and its duties be damned. And if in a sense I couldn't, because I would become hysterical, and hysteria isn't what you had in mind when you whispered "Stay!", that's irrelevant, isn't it? What I would need in order to do what you want isn't more freedom of will, it's stronger nerves.

The point is: I have to make my rounds, that's true, and it follows, trivially, that I can't stay. But I don't have to make my rounds in any sense that binds my will. Even Luther's reasons were only reasons. Even the reasons that favor obliging a man with a gun are only reasons. Naturally, one gives them a lot of weight, insofar as one acts for reasons at all in such a case. One does. Or one doesn't. "It's entirely up to you," the man with the gun says, smiling ironically. He's right. It is up to you, even if his smile is right too. Everyone will gravely agree, later -- or perhaps they won't, depending on the case -- that you had no choice but to do as he said. There was no alternative. What else could you do? But these remarks are not even psychological, much less absolute. They are alike "You had no alternative; there was nothing else to be done" in reviewing a game of chess, which doesn't imply that you couldn't have made some quite other, idiotic move instead -- or stormed out of the room, for that matter. They are not, in that sense, remarks about you. You occur vacuously in them, so to speak, and they have no bearing on your freedom of will in making the move you made.

Well, I shouldn't quite say that. They have this sort of bearing: You have been ordered by God himself to shoot me, and do. Later on, you insist that you had no free will in the matter, no choice. It isn't as if you shot me of your own volition. You hated doing it. Indeed, you shot me again to put me out of my misery. In all this, you were only doing what you had to do. All right. In religious obedience (if we believe you) you had no freedom of choice in the matter of whether to shoot me or not. But why did you have no choice between religious obedience and disobedience? The same answer won't do, and eventually, as these questions proliferate, we will arrive at a formulation of what you did under which you won't be able to say, in this way of saying it, that you had no free will in the matter of whether to do the thing or not. You will run out of constraining reasons.

Or is there some system of reasons in which you can hole up from which there is no exit except into nonsense? Well, maybe. But even if there is, and yes, one quite sees that every alternative to shooting me was ruled out in that system, from which you could not intelligibly have prescinded even in the particular case, still -what of it? You had to do it. You absolutely had to do it. That was the objective position, or anyway you thought so, in your craziness. But having to do a thing does not settle magically the question whether to do it or not. Reasons, of whatever species, logically can't close that question. It's a question of a different genre, and is not relative to any system of reasons. It isn't for reasons, in the end, that we act for reasons. "I'm sorry, I have no choice," "You leave me no alternative," and so forth, are objective. They are misused or used in bad faith (there is bad faith in English as well as in French) if they are put on offer as reporting the sad results of scrupulous introspection. "I must shoot you. I can't help it." Well, yes. That's the intrasystematic position. I see what you mean. Now give me that gun!

That's all I'm going to say about reasons, because I must say at least something about desire and aversion, and all that lot. That is, I am determined to. But it won't be nearly enough. First, insofar as the role of these items in conduct is that of giving us reason to behave well or badly, this species of reason is in itself as powerless as any other. It has to be treated with the respect that it may or may not deserve. It's no good your just running it up the flagpole to see if you salute. But of course there is another role of desire and the rest in action. Acting in view of one's desire is one thing, like acting in view of someone else's desire. Acting from desire is quite another and commoner thing. Indeed, pure cases of acting in consideration of a (felt) desire but not at all from it must be rare at best. (Your new policy is to satisfy your desires. You notice that, by George, you feel a faint desire for chalk, and rush out into the night, hoping the desire won't vanish before you can satisfy it.) Well, is the will perfectly free in action from desire, out of fear, on impulse, and so on? It certainly can seem not. And a parallel question arises for action from such motives as malice and greed. And for compulsive or obsessional conduct, and so on. But I don't see unfreedom of will even in this most promising part of the forest. Of course if it's really like chains, or like being violently thrown into bed, then it is, and there's no unfreedom of will in it. You haven't in the relevant sense done anything. You froze. Or it was a seizure. Why shouldn't there be seizures of whatever you like: curtseying, bowing, waltzing? There are seizures of obscenity, one reads. They must be embarrassing, unless one is obscene anyway.

But addiction, alcoholism, child molesting and so forth -- to begin with a very complicated kind of case -- aren't patterns of bizarre physical seizure. There is automatism in them, in the sense of finding a cigarette in your hand. And again, that's not unfreedom of will, according to me. But there is a lot else in them. The individual drink, say, is typically not quite automatic, or even not automatic at all. It may be coldly deliberate, in fact, and not preceded by any notable craving. (There is nihilistic drinking that makes one think of possession rather than unfreedom of will.) It's simply not true, locally, that the alcoholic can't help himself. I mean: he could perfectly well, and sometimes does, empty the bottle into the sink, or the like. But usually he doesn't. He drinks it, in (as far as I can see) full freedom of will. And goes on like that, year after year. How very odd that the obvious reasons to stop behaving in this way don't weigh with him as decisively as one might expect! Well, yes, it is odd, but there it is: they don't, and he doesn't stop, though other drinkers have just stopped, no doubt. After all, people have on purpose stopped eating and died.

Is there reason to believe about any ordinary alcoholic that he literally couldn't stop, even at the cost of devoting his life to not drinking and letting everything else go? I wonder. But in any case, he probably isn't going to stop, except by taking extraordinary steps which would be inconvenient and humiliating. (Or worse, he may feel. Who could ask Odysseus to spend a lifetime lashed to the mast?) And since in the usual case he knows this as well as we do, or better, he is in no position to decide to stop drinking as one might decide to stop shaving. He is past that point. One can't simply decide to stop doing what one knows one probably won't in that way stop doing. But that's "logic" again, not unfreedom. And it doesn't get in the way of a good deal of pretending to decide to stop drinking, of course. Indeed, even really deciding to stop isn't altogether ruled out by terrible experience of oneself. This time, the alcoholic thinks, he'll really stop. Then he doesn't, again.

Is it possible that he can't? Of course he could by extraordinary means, but setting such means aside, as he is likely to do, is he perhaps actually incapable of stopping? Perhaps. Perhaps he just hasn't the strength of will to hold out, as one might be unable to withstand torture. (That sort of thing isn't in one's repertoire. One just will, after a time, decide to tell them what they want to know while one still has one's faculties.) All right. But strength of will is one thing and freedom of will is another. Isn't it? Or do you think not? I think it is.

Besides, the alcoholic may love drinking. Moreover, a few drinks ("just one more time") and he alters. He's a changed man. This changed man thinks whatever he thinks, pours the stuff down, gets belligerent with the bartender, imagines that no one can tell, and so forth. That's what he's like. The alcoholic in propria persona doesn't quite know about this alteration, of course. How could he? Well, videotapes might help. But one despairs. What's the matter with him? In a word: alcoholism, and I see no unfreedom of will anywhere in it. Unfreedom of will would be a marvelous excuse for the complex misbehavior of such disorders. But there are other causes for pity in them. There's no need to drag that one in.

Or to drag it into simpler cases either: flying off the handle, say, and its relatives. Again, if it's like literally exploding, it is, and the will doesn't come into it. But perhaps it isn't quite like that. Rather: you didn't control yourself. Perhaps you actually couldn't have controlled yourself much longer, anxiously knew that, anxiously didn't know what form the coming disintegration would take, and (in the nick of time) preempted it. One does that.

But controlling yourself is a project, possible or impossible, like controlling a dog. If you try hard and can't do it, better luck next time. You gave it your best shot. Or you didn't, in perfect freedom of will. And, what about your shouting and stomping and smashing the china while we say sayonara to your better self, as if to an incompetent spirit control? We're all dismally used to Mr. Hyde, here, yelling "I am the famous Dr. Jekyll, by God, and you will keep a civil tongue in your head if I have to sew it in myself!" Has he no free will? Why not? He's having a marvelous time. Look at him! Thinks he's a regular force of nature, he does!

You see how this is going to go. "Daddy is not himself." And that's right, he isn't. He'll be back, lachrymose as always. But he's out of the picture at the moment, again. Or is this he, after all, raging in monstrous freedom? Either way, how absurd to say to him later, "You had no free will, there. You suffered another of your sad lapses of free will, that's all. Don't take on so." I bet I can handle any case you like, though I know I have only begun to deal with even these few. I hope I haven't cheated. I doubt that I am going to see unfreedom of will anywhere in our lives. Suppose that indeed I am not, not because, miraculously, it isn't there, but because the idea of it is incomprehensible, a picture without application! But I've asked you to do a lot of supposing. You may well be sick of supposing. What about determinism, you may wonder. Fortunately, Elizabeth Anscombe has taught me, by her essay "Causality and Determination," that I needn't go in for Lapacean fantasies, and I gather that John Earman is intent on conveying the same reassurance. That's fine. But one wouldn't care to think one's freedom of will secured by the physical possibility in pure theory that one will stay in bed for the rest of one's life, with Russian explanations ready in case anyone asks, much less by the theoretical possibility that instead of doing one's duty one will suddenly deliquesce into a nasty liquid all over the rug. Are we or aren't we as approximately deterministic as alarm clocks, say? That seems an awful question.

Unlike Anscombe, Wittgenstein thought it didn't matter, I find. At any rate there is a set of rather bad notes on a lecture about freedom of the will given by Wittgenstein in perhaps 1945 or 1946, from which it appears that he saw no contradiction (or other incoherence) in the idea that a free decision might nevertheless be "determined" by natural laws. He seems to have been quite clear in his mind about that. I, on the other hand, am not. I am foggy about it, possibly because I am bogged down in superstitions about natural law and the causal nexus. Still, I do sometimes think I see that logical inconsistency isn't the real trouble between determinism and free will: that they aren't in strict truth inconsistent. It then seems to me that their conflict comes rather of our being lured by the hypothesis of determinism into an alarming view of ourselves as a species of objects, which as such can't be thought to do anything (in a sense in which we had supposed that we did things), however busy and smiling they are. Automata are just automata, after all, whether physical through and through or on the contrary psychophysical. And, on second thought, whether deterministic or indeterministic. If one had hoped for a friend, or at least a pet, even a rather unpredictable object is cold comfort.

I think one might be able to hang onto one's friends, and enemies, in spite of these objects. One might begin by questioning the success of any attempt to introduce them into the conversation, by ostension or by description. It really is not clear that if I pat myself all over, saying, "You know: this thing here," you do know what thing I mean, or that I can explain what thing I mean, either. So Douglas Long argued explicitly years ago, and others have held too.11 Anscombe seems to be among them, actually, and perhaps Wittgenstein was as well. 21 They have not been widely believed. Nevertheless, they may be absolutely right.13 That would help. A new inquiry into the very idea of physical necessity would help.14 And even what I have said tonight might help. But I can't go on about any of that, I really can't. It's out of my hands, I absolutely must stop. So, I will. Thank you.

1. What follows is for the most part the text read on the occasion, with minor defensive alterations. One passage, about alcoholism, has been rewritten and is now perhaps intelligible, for better or worse.
2. Passions of the Soul, I, Article XLI. I got this reference from Paul Hoffman.
3. Conversation with Burman, tr. by John Cottingham (Clarendon Press, Oxford, 1976), sec. 31, p. 21.
4. G.E.M. Anscombe, Collected Philosophical Papers, vol. II: Metaphysics and the Philosophy of Mind (University of Minnesota Press, 1981), pp. 166, 167.
5. Ibid., p. 167.
6. Ibid., p. 170.
7. Ibid., p. 172.
8. Ibid., p. 172.
9. Read "try to jump over this building" opaquely. That I might mistake a building in the distance for a miniature building at my feet, and try to jump over it, is irrelevant, I hope, even if true. (This counterexample, or one like it, was put to me by Palle Yourgrau, as I recall.)
10. I think it was John Koethe who suggested this analogy.
11. See Long's articles "The Philosophical Concept of a Human Body," Philosophical Review, LXXIII (July, 1964) and "The Bodies of Persons," Journal of Philosophy LXXXI, 10 (May 30, 1974), and now also Jay Rosenberg, Thinking Clearly about Death (Prentice-Hall, 1983). Strawson's Individuals is not a case in point, as far as I can tell, since the "material bodies we possess" of that book seem to be just the bodies in question.
12. See Anscombe's "The First Person," loc. cit., p. 33, 3rd full paragraphy, and on Wittgenstein, John Cook's "Human Beings," in Studies in the Philosophy of Wittgenstein ed. by P. Winch, (New York Humanities Press, 1969), an interpretation that Cook would now repudiate, if I am not mistaken.
13. Though not necessarily for the reasons they give, where they give any. 14. And another into my relation to the necessities, if any, of my nervous system's operations and their bodily consequences. Am I under all these necessities? Do they all face me with impossibilities, in the sense of Anscombe's rejection of any 'can' of freedom that holds "in face of physical impossibility"?