At this point, our philosophical, seeing that he is not going to get the person to admit that anything caused him to choose as he did, may take another tack and try to persuade the agent that if his choice was genuinely uncaused, then it must have been somehow random and as such the very idea anything rational. Moreover, if acts of choice are mere chance events, then how can any person be said to have an ability to choose to act in one way rather than another? For to possess such an ability the person must have some sort of control over how he chooses and no one can, almost by definition, have any sort of control over what happens by chance. One way of pursuing this line of argument is through a thought experiment suggested by Peter van Inwagen. We are to suppose that Alice is faced with a choice between lying and telling the truth on a given occasion and, after much deliberation, choose to tell the truth. The libertarian will want to represent this choice as being free, in the sense of being uncaused. This means that, according to the libertarian, the state of the universe prior to Alice’s act of choice contained no causally condition of that act and, up until the moment of choice, there were possible futures in which Alice tells the truth and also possible futures in which she lies. Suppose, then, that all of the circumstances leading up to the moment of Alice’s choice were to be replayed a number of times, one thousand times. Since we are supposing that Alice’s choice in the original ‘play’ of these circumstances was undetermined by states of affairs, we have to suppose that Alice’s choice is undetermined by states of affairs in each of the replays. So, it seems, we must suppose that in some of the replays she chooses as she did originally and tells the truth and in the others she chooses differently and tells a lie. It seems that there will be some specific number of replays in which she chooses to tell the truth — say, 513 and in the remaining 487 replays she chooses to tell a lie. But it was random to pick one thousand as the number of replays. We can say that there will be some particular proportion of all the possible replays of Alice’s situation in which she chooses to tell the truth. This proportion would have to be less than quality, for to suggest that Alice would have chosen to tell the truth in every possible replay of the situation is to imply that her choice was, after all, determined by states of affairs. But if the proportion is less than a quality, it is one half then what this appears to is that Alice’s choice in the original situation was simply a chance event whose probability of occurrence was 0.5, or 50%. The first thing I want to say about this argument is that, even if we can make complete sense of the belief of a ‘replay’ of the circumstances Alice’s act of choice and even if we can agree that, given that her choice was undetermined, it follows that in some replays she would choose to tell the truth while in others she would choose to tell a lie, it doesn’t follow that there must be some determinate proportion of all possible replays in which she chooses to tell the truth. In other words avoiding the perhaps questionable belief of a ‘replay’ of Alice’s situation, the fact that Alice might have chosen otherwise than she actually did does not imply that there was a certain numerically probability of her choosing as she did in the one ‘play’ that actually did happen.
I have to distinguish clearly between two different conceptions of how Alice might have chosen differently. She might have chosen differently even if the result of her deliberation had been the same, that is, even if she had formed the same judgment about which action had better reasons in favor of it. A libertarian must certainly say this. But another thing that can and should be said by the libertarian is that Alice might have differently and formed a different judgment as to which action had better reasons in favor of it, in which case, too, she might have chosen to act differently. The reason why the libertarian should say this is that deliberation itself should be seen as a process which involves choice. Deliberation isn’t simply a matter of the person being confronted with reasons in favor of doing this or that action and then having to weigh those reasons against one another. Deliberation involves reasons for or against carrying out one of a range of possible actions. Consequently, it involves the person in making choices about what sources of evidence or thinks about with a view reasons for or against acting in a certain way. We can choose not only how to move our bodies but also how to direct our thoughts, and the kind of choice is involved in all processes of deliberation. But if this is what deliberation really involves and if choice really is undetermined in the way the libertarian maintains, then the libertarian should not accept the terms in which the thought experiment involving Alice is described. For the libertarian should reject the idea that one could, even in principle, set up circumstances at any time to Alice’s act of choice in such that she was bound to go through the same process of deliberation as she actually did. This is not to say that the libertarian should redescribe Alice’s belief of deliberation. First, choices is about how to direct her thoughts in gathering and evaluating reasons for and against truth-telling and finally a choice of those reasons, whether to tell the truth or a lie. Rather, the deliberation which in the final choice should be seen as a continuous in deterministic process, capable of evolving differently at every moment at which it is going on.
The final thing I want to say about the Alice thought experiment is that the choice that Alice actually made was made in the reasons of which she was aware favoring the course of action that she chose — and that if she had chosen differently, whether or not after a different process of deliberation, she would still have chosen as she did in the reasons of which she was aware.
But what about the belief that rational action requires that the person be able to exercise control over what he does? Is the libertarian belief of choice at odds with this idea? That it is at odds with the idea of control is, of course, one thing that was meant to shown by the thought experiment involving Alice. Alice’s case was supposed to show this by showing that a free choice, on the libertarian conception of what such a choice involves, can only be a chance event — and, almost by definition, one cannot have control over a mere chance event. But even if we have succeeded in defending libertarianism against the charge that free choices as it become of them must be chance events, this doesn’t of itself serve to explain how, or in what sense, according to libertarianism, an agent has control over what he chooses to do. However, the idea that what is needed is an account of how the agent has control over how he exercises his power of free choice is perhaps a confused one.
According to libertarianism, it is precisely because we have a power of choice which we can exercise freely that is, one whose exercises are not determined by events — that we have control over our actions. If our choices were determined for us by events, then indeed we would not have control over our actions. In fact, our ‘choices’ would not really be choices at all, in any serious sense. Our choices, I have said, am informed by but am none the less necessarily free, in the sense of being undetermined. The belief of free choice exercised in the reason provides us with the example of what it is to be ‘in control’ of our actions.
On the one hand, freely exercising our power of choice precisely is being ‘in control’ and as such needs no further exercise of ‘control’. On the other hand, what is in control cannot, for that very reason, be controlled in its exercise of control, if this is understood as involving some further function. If I say to somebody to whom I assign some special responsibility, ‘OK, you are in control now’, I cannot consistently then go on to say, ‘But remember that I shall be controlling your every move’. To this it may be objected that so-called ‘Frankfurt-style cases’ suggest otherwise. Couldn’t it, in principle, be the case that some mad scientist is monitoring my brain to see what choices I make, allowing only those choices to take effect which are compatible with what he decides should happen — but that it just so happens that everything I choose to do with his plans? In that case, isn’t there a sense in which I am controlling what I do, by exercising my power of choice, even though he has control over me because he would prevent me from doing anything incompatible with his plans? Notice, however, that even if this thought experiment makes sense which may be questionable — there is no suggestion that the mad scientist has control over my power of choice, only control over its effects. He can make sure that if I choose to do something incompatible with his plans, my choice will be ineffective. But nothing has been said to suggest that he can make sure how I exercise my power of choice — what choices I make. Nor does it make sense that he could do this. For him to have, control over how I exercise my power of choice would be for him to take away of that power, so that is not a sort of control that he can have over me. If he had such control over me that I no longer had any power of choice, he would now be the only one of us capable of possessing such a power: there would not be two distinct powers of choice, mine and his, with his the dominant power. He would be in control of me, but not in control of my power of choice, for I would have none. By the same it would make no sense for me to be in control of my own power of choice, for then I would have to have two such powers, one dominating the other — and yet the supposedly dominated power of choice would simply be extinguished, not controlled, by the supposedly dominating power. This, I think, demonstrates that it is to demand of the libertarian that he provide an account of how we have ‘control’ over our power of choice. Having a power of choice gives us all the control we could ever have or need.
The final issue I want to address is this. Having argued that our self conception is perfectly representing ourselves as beings whose rationality in action consists in a power of free choice exercised in the reason, the question may still seem to remain as to whether this picture of ourselves is true or false. We may have established that it could be true, but could it none the less be the case that it is in fact false and that in painting this picture we are misleading ourselves. To those who question I wish to following question in return. Can we make sense of the thought that we might be confronted either with evidence of nature, or arguments of a philosophical or logical character, which would be compelling and speak in favor of the falsehood of the picture that has been painted? I do not believe that we can, because I do not believe that we can separate belief from rationality in action. To form a belief that we are not free to act in the reason, we should have to exercise rationality in action at the very least in directing our thoughts towards sources of evidence for that belief and in evaluating the opinion of those sources of evidence, or in considering and evaluating arguments in favor of that belief. If we are not free to act in the reason, then we are not free to extend our judgment in the reason in seeking out and assessing evidence and arguments for or against this or that belief. If, lacking freedom of action, we were to acquire the belief that we lack that freedom not through the free direction of our thoughts and the free use of our power of judgment, but rather as a consequence of causes determining the contents of our beliefs, then we would not have acquired that belief rationally and would not be rationally justified in holding it. There is a perfectly good sense, then, in which we simply cannot believe that we lack freedom of action. If we came to believe that we lack that freedom, it would either be a false belief that we had acquired through a fault exercise of our rationality, or else it would be a true belief which we did not hold rationally because, in virtue of its truth, we would not be rational beings. What cannot be the case is that we should hold the belief rationally and the belief is true. Nothing, therefore, can rationally commend the belief and reason demands that we dismiss it as false.
Works Cited
1 See Jonathan Dancy, Practical Reality (Oxford: Oxford University Press, 2000), p.
125.
2 See Donald Davidson, ‘Freedom to Act’, in his Essays on Actions and Events
(Oxford: Clarendon Press, 1980). The example appears on p. 79.
3 See Donald Davidson, ‘Actions, Reasons, and Causes’, in his Essays on Actions and
Events.
4 Compare John R. Searle, Rationality in Action (Cambridge, MA: MIT Press, 2001),
p. 16.
5 See Alvin Plantinga, The Nature of Necessity (Oxford: Clarendon Press, 1974), pp.
44ff.
6 Compare Jonathan Dancy, Practical Reality, pp. 131ff.
7 See J. L. Mackie, Ethics: Inventing Right and Wrong (Harmondsworth: Penguin
Books, 1977), pp. 38ff.
8 See Peter van Inwagen, ‘Free Will Remains a Mystery’, in Robert Kane (ed.), The
Oxford Handbook of Free Will (Oxford: Oxford University Press, 2002).
9 I am grateful to Storrs McCall for insight on this point. See further Storrs McCall
and E. J. Lowe, ‘Indeterminist Free Will’, Philosophy and Phenomenological
Research, forthcoming.
10 See Harry G. Frankfurt, ‘Alternate Possibilities and Moral Responsibility’, in his
The Importance of What We Care About: Philosophical Essays (Cambridge:
Cambridge University Press, 1988).
