C. A. Campbell
Charles Arthur Campbell (1897–1974) was a Scottish philosopher who taught at the University of Glasgow and the University of North Wales. He is best known for his work on the free will problem, but he also made important contributions to ethics and the philosophy of religion. In the following selection, he attempts to elaborate and defend a libertarian view of free will.
Campbell’s view is a version of the doctrine of agent causation. The central idea is that a free choice (in his case the choice of whether or not to exert the effort required to overcome the balance of one’s desires and do what one perceives to be the morally right thing) is caused by the agent or self, but not by some specific event (or set of events) occurring within the agent—because otherwise one could ask for the cause of that event, and so on, leading either to causal determinism or to one or more random events. Thus the cause of the choice is a thing or substance, not an event.
In Defence of Free Will
...Let us begin by noting that the problem of free will gets its urgency for the ordinary educated man by reason of its close connection with the conception of moral responsibility. When we regard a man as morally responsible for an act, we regard him as a legitimate object of moral praise or blame in respect of it. But it seems plain that a man cannot be a legitimate object of moral praise or blame for an act unless in willing the act he is in some important sense a ‘free’ agent. Evidently free will in some sense, therefore, is a pre-condition of moral responsibility....
We raise the question at once, therefore, what are the conditions, in respect of freedom, which must attach to an act in order to make it a morally responsible act? It seems to me that the fundamental conditions are two. I shall state them with all possible brevity, for we have a long road to travel.
 For ultimate analysis, the agent’s power of alternative action would seem to be an inexpugnable condition of his liability to moral praise or blame, i.e. of his moral responsibility.This first condition, however, is quite clearly not sufficient. It is possible to conceive an act of which the agent is the sole cause, but which is at the same time an act necessitated by the agent’s nature....In the case of such an act, where the agent could not do otherwise than he did, we must all agree, I think, that it would be inept to say that he ought to have done otherwise and is thus morally blameworthy, or ought not to have done otherwise and is thus morally praiseworthy. It is perfectly true that we do sometimes hold a person morally responsible for an act, even when we believe that he, being what he now is, virtually could not do otherwise. But underlying that judgment is always the assumption that the person has come to be what he now is in virtue of past acts of will in which he was confronted by real alternatives, by genuinely open possibilities: and, strictly speaking, it is in respect of these past acts of his that we praise or blame the agent now.
We may lay down, therefore, that an act is a ‘free’ act in the sense required for moral responsibility only if the agent ( a ) is the sole cause of the act; and ( b ) could exert his causality in alternative ways....
And now, the conditions of free will being defined in these general terms, we have to ask whether human beings are in fact capable of performing free acts; and if so, where precisely such acts are to be found. In order to prepare the way for an answer, it is desirable, I think, that we should get clear at once about the significance of a certain very familiar, but none the less formidable, criticism of free will which... the Libertarian has to meet. This is the criticism which bases itself upon the facts of heredity on the one hand and of environment on the other. I may briefly summarize the criticism as follows.
The externality of these influences is taken for granted in our reflective practical judgments upon persons. On those occasions when we are in real earnest about giving a critical and considered estimate of a man’s moral calibre—as, e.g., in any serious biographical study—we impose upon ourselves as a matter of course the duty of enquiring with scrupulous care into his hereditary propensities and environmental circumstances, with a view to discovering how far his conduct is influenced by these factors. And having traced these influences, we certainly do not regard the result as having no bearing on the question of the man’s moral responsibility for his conduct. On the contrary, the very purpose of the enquiry is to enable us, by due appreciation of the external influences that affect his conduct, to gain as accurate a view as possible of that which can justly be attributed to the man’s own self -determination. The allowances that we all of us do in practice make for hereditary and environmental influences in passing judgment on our fellows would be meaningless if we did not suppose these influences to be in a real sense ‘external’ to the self.
 Nor can there be in the end much doubt, I think, in what function of the self that activity is to be located. There seems to me to be one, and only one, function of the self with respect to which the agent can even pretend to have an assurance of that absolute self-origination which is here at issue. But to render precise the nature of that function is obviously of quite paramount importance: and we can do so, I think, only by way of a somewhat thorough analysis—which I now propose to attempt—of the experiential situation in which it occurs, viz., the situation of ‘moral temptation’.Let us proceed, then, by following up this clue. Let us ask, why do human beings so obstinately persist in believing that there is an indissoluble core of purely self -originated activity which even heredity and environment are powerless to affect? There can be little doubt, I think, of the answer in general terms. They do so, at bottom, because they feel certain of the existence of such activity from their immediate practical experience of themselves.
Now the objective validity or otherwise of this belief is not at the moment in question. I am here merely pointing to its existence as a psychological fact. No amount of introspective analysis, so far as I can see, even tends to disprove that we do as a matter of fact believe, in situations of moral temptation, that it rests with our self absolutely to decide whether we exert the effort of will which will enable us to rise to duty, or whether we shall allow our desiring nature to take its course.
I have now to point out, further, how this act of moral decision, at least in the significance which it has for the agent himself, fulfils in full the two conditions which we found it necessary to lay down at the beginning for the kind of ‘free’ act which moral responsibility presupposes.
For obviously it is, in the first place, an act which the agent believes he could perform in alternative ways. He believes that it is genuinely open to him to put forth effort—in varying degrees, if the situation admits of that—or withhold it altogether. And when he has decided—in whatever way—he remains convinced that these alternative courses were really open to him.
Here then, if my analysis is correct, in the function of moral decision in situations of moral temptation, we have an act of the self which at least appears to the agent to satisfy both of the conditions of freedom which we laid down at the beginning. The vital question now is, is this ‘appearance’ true or false? Is the act of decision really what it appears to the agent to be, determined solely by the self, and capable of alternative forms of expression? If it is, then we have here a free act which serves as an adequate basis for moral responsibility. We shall be entitled to regard the agent as morally praiseworthy or morally blameworthy according as he decides to put forth effort or to let his desiring nature have its way. We shall be entitled, in short, to judge the agent as he most certainly judges himself in the situation of moral temptation. If, on the other hand, there is good reason to believe that the agent is the victim of illusion in supposing his act of decision to bear this character, then in my opinion the whole conception of moral responsibility must be jettisoned altogether. For it seems to me certain that there is no other function of the self that even looks as though it satisfied the required conditions of the free act.
 Nevertheless I agree that we shall have to weigh carefully several criticisms of high authority before we can feel justified in asserting free will as an ultimate and unqualified truth.Now in considering the claim to truth of this belief of our practical consciousness, we should begin by noting that the onus of proof rests upon the critic who rejects this belief. Until cogent evidence to the contrary is adduced, we are entitled to put our trust in a belief which is so deeply embedded in our experience as practical beings as to be, I venture to say, ineradicable from it. Anyone who doubts whether it is ineradicable may be invited to think himself imaginatively into a situation of moral temptation as we have above described it, and then to ask himself whether in that situation he finds it possible to disbelieve that his act of decision has the characteristics in question. I have no misgivings about the answer. It is possible to disbelieve only when we are thinking abstractly about the situation; not when we are living through it, either actually or in imagination. This fact certainly establishes a strong prima facie presumption in favour of the Libertarian position.
Fortunately for our purpose, however, there are some lines of criticism which, although extremely influential in the recent past, may at the present time be legitimately ignored....
...Libertarianism is certainly inconsistent with a rigidly determinist theory of the physical world. It is idle to pretend that there can be open possibilities for psychical decision, while at the same time holding that the physical events in which such decisions manifest themselves are determined in accordance with irrevocable law. But whereas until a few years ago the weight of scientific authority was thrown overwhelmingly on the side of a universal determinism of physical phenomena, the situation has, as everybody knows, profoundly altered during the present century more especially since the advent of Planck’s Quantum Theory and Heisenberg’s Principle of Uncertainty. Very few scientists to-day would seek to impugn free will on the ground of any supposed implications of the aims or achievements of physical science....
I may turn at once, therefore, to lines of argument which do still enjoy a wide currency among anti-Libertarians. And I shall begin with one which, though it is a simple matter to show its irrelevance to the Libertarian doctrine as I have stated it, is so extremely popular that it cannot safely be ignored.
My answer is that the Libertarian view is perfectly compatible with prediction within certain limits, and that there is no empirical evidence at all that prediction is in fact possible beyond these limits. The following considerations will, I think, make the point abundantly clear.
There is no question, on our view, of a free will that can will just anything at all. The range of possible choices is limited by the agent’s character in every case; for nothing can be an object of possible choice which is not suggested by either the agent’s desires or his moral ideals, and these depend on ‘character’ for us just as much as for our opponents. We have, indeed explicitly recognized at an earlier stage that character determines the situation within which the act of moral decision takes place, although not the act of moral decision itself. This consideration obviously furnishes a broad basis for at least approximate predictions.(1)
There is one experiential situation, and one only, on our view, in which there is any possibility of the act of will not being in accordance with character; viz. the situation in which the course which formed character prescribes is a course in conflict with the agent’s moral ideal: in other words, the situation of moral temptation. Now this is a situation of comparative rarity. Yet with respect to all other situations in life we are in full agreement with those who hold that conduct is the response of the agent’s formed character to the given situation. Why should it not be so? There could be no reason, on our view any more than on another, for the agent even to consider deviating from the course which his formed character prescribes and he most strongly desires, unless that course is believed by him to be incompatible with what is right.(2)
Even within that one situation which is relevant to free will, our view can still recognize a certain basis for prediction. In that situation our character as so far formed prescribes a course opposed to duty, and an effort of will is required if we are to deviate from that course. But of course we are all aware that a greater effort of will is required in proportion to the degree in which we have to transcend our formed character in order to will the right. Such action is, as we say, ‘harder’. But if action is ‘harder’ in proportion as it involves deviation from formed character, it seems reasonable to suppose that, on the whole, action will be of rarer occurrence in that same proportion: though perhaps we may not say that at any level of deviation it becomes flatly impossible. It follows that even with respect to situations of moral temptation we may usefully employ our knowledge of the agent’s character as a clue to prediction. It will be a clue of limited, but of by no means negligible, value. It will warrant us in predicting, e.g., of a person who has become enslaved to alcohol, that he is unlikely, even if fully aware of the moral evil of such slavery, to be successful immediately and completely in throwing off its shackles. Predictions of this kind we all make often enough in practice. And there seems no reason at all why a Libertarian doctrine should wish to question their validity.(3)
Now when these three considerations are borne in mind, it becomes quite clear that the doctrine we are defending is compatible with a very substantial measure of predictability indeed. And I submit that there is not a jot of empirical evidence that any larger measure than this obtains in fact.
Now this criticism, and all of its kind, seem to me to be the product of a simple, but extraordinarily pervasive, error: the error of confining one’s self to the categories of the external observer in dealing with the actions of human agents. Let me explain.
It is perfectly true that the standpoint of the external observer, which we are obliged to adopt in dealing with physical processes, does not furnish us with even a glimmering of a notion of what can be meant by an entity which acts causally and yet not through any of the determinate features of its character. So far as we confine ourselves to external observation, I agree that this notion must seem to us pure nonsense. But then we are not obliged to confine ourselves to external observation in dealing with the human agent. Here, though here alone, we have the inestimable advantage of being able to apprehend operations from the inside, from the standpoint of living experience. But if we do adopt this internal standpoint—surely a proper standpoint, and one which we should be only too glad to adopt if we could in the case of other entities—the situation is entirely changed. We find that we not merely can, but constantly do, attach meaning to a causation which is the self’s causation but is yet not exercised by the self’s character. We have seen as much already in our analysis of the situation of moral temptation. When confronted by such a situation, we saw, we are certain that it lies with our self to decide whether we shall let our character as so far formed dictate our action or whether we shall by effort oppose its dictates and rise to duty. We are certain, in other words, that the act is not determined by our character , while we remain equally certain that the act is determined by our self.
I conclude, therefore, that those who find the Libertarian doctrine of the self’s causality in moral decision inherently unintelligible find it so simply because they restrict themselves, quite arbitrarily, to an inadequate standpoint: a standpoint from which, indeed, a genuinely creative activity, if it existed, never could be apprehended.
From In Defence of Free Will ( London: George Allen & Unwin LTD, 1967).
1. Clearly, on Campbell’s view, a person who exerts the effort needed to overcome the balance of his desires has acted freely. Is the action of a person who chooses not to exert this effort and so does what he most desires to do also a free action (so that people would not be morally responsible only when they do the right thing)? Why or why not? What if the desires opposed to morality are very strong (see Annotation 93 and the associated text)?
2. Does the idea of agent causation, in general and in Campbell’s specific version, really make sense? How is the choice fixed or arrived at if it is not a result of formed character (and also not random)?
3. On Campbell’s view, a person is free and morally responsible only in situations where there is a conflict between desire and morality (and perhaps only when he follows morality and not desire). Is this a plausible view of freedom? Is there any way to expand Campbell’s view so as to allow a wider range of free choices?
Robert Nozick (1938–2002) taught philosophy for many years at Harvard and was a major figure in twentieth century philosophy. He made important contributions to many areas, but was perhaps best known for his book Anarchy, State, and Utopia (from which two of the selections in Chapter 6 are excerpted).
an agent assigning weights to reasons favoring various alternatives (rather than simply acting in accord with the antecedently most weighty reasons). This could be viewed as a version of agent causation, since the assignment of weight is brought about by the agent, but is not determined by any specific features of the agent (including his own formed character). But Nozick makes no explicit appeal to the idea of agent causation and attempts to explain in other ways why such an assignment of weights is not merely random.In the following selection, Nozick defends a version of libertarianism. In his view, free actions occur as a result of
Choice and Indeterminism, from
Making some choices feels like this. There are various reasons for and against doing each of the alternative actions or courses of action one is considering, and it seems and feels as if one could do any one of them. In considering the reasons, mulling them over, one arrives at a view of which reasons are more important, which ones have more weight. One decides which reasons to act on; or one may decide to act on none of them but to seek instead a new alternative since none previously considered was satisfactory.
After the choice, however, others will say we were caused to act by the considerations which were (or turned out to be) more weighty. And it is not just others. We too, in looking back at our past actions, will see which reasons swayed us and will view (accepting) those considerations as having caused us to act as we did. Had we done the other act, though, acting on the opposing considerations, we (along with the others) would have described those considerations as causing us to do that other act. Whichever act we do, the (different) background considerations exist which can be raised to causal status. Which considerations will be so raised depends upon which act we do. Does the act merely show which of the considerations was the weightier cause, or does the decision make one of them weightier?
… Psychology, sociobiology, and the various social sciences, on this view, will offer causal explanations of why something is or is not a reason for a person (in a situation). They will not always be able to explain why the reasons get the precise weights they do….What picture of choice emerges if we take seriously the feeling that the (precise) weights to be assigned to reasons is “up to us”? It is causally undetermined (by prior factors) which of the acts we will decide to do. It may be causally determined that certain reasons are reasons (in the one direction or the other), but there is no prior causal determination of the precise weight each reason will have in competition with others.
A which are reasons for doing act A and over RB which are reasons for doing act B, it is undetermined which act he will do. In that very situation, he could do A and he could do B. He decides, let us suppose, to do act A. It then will be true that he was caused to do act A by (accepting) RA. However, had he decided to do act B, it then would have been RB that caused him to do B. Whichever he decides upon, A or B, there will be a cause of his doing it, namely RA or RB. His action is not (causally) determined, for in that very situation he could have decided differently; if the history of the world had been replayed up until that point, it could have continued with a different action. With regard to his action the person has what has been termed contra-causal freedom—we might better term it contra-deterministic.It is neither necessary nor appropriate, on this view, to say the person’s action is uncaused. As the person is deciding, mulling over reasons R
 If this were correct,... then causality necessarily would involve causal determination: under exactly the same conditions repeated, exactly the same thing would have (again) to happen. According to the view that distinguishes causality from causal determination, an act can be done because of something and have a cause even though in exactly the same conditions another act could have been done. It is common, in retrospect, to see what caused us to act as we did. Although we can retrospectively identify a cause, this does not mean our action was causally determined; had we acted differently in that situation (as we could have) we retrospectively would have identified a different cause— RB instead of RA.Thus, we draw a distinction between an action’s being caused, and its being causally determined. Some philosophers would deny this distinction, maintaining that whenever one event causes another, there holds a general law in accordance with which it does so: some specification of the first event (along with other conditions which hold) always is and would be followed by an event of the same type as the second. It is a metaphysical thesis that the root notion of causality, producing or making something happen, can operate only through such lawlike universality.
The weights of reasons are inchoate until the decision. The decision need not bestow exact quantities, though, only make some reasons come to outweigh others. A decision establishes inequalities in weight, even if not precise weights.
 Defenders of the claim do point out other situations (of choice or answering questions) where the relevant preference or motive can be identified; so the truth of the claim in this decision situation is testable, given the assumption that the preference or motive is stable from the one situation to the other. However, if our conception of the bestowal of weights (with a commitment that lingers) holds true, then these independent “tests” are to be interpreted differently. We do not always act on what was a preexistingly strongest preference or motive; it can become strongest in the process of making the decision, thereafter having greater weight (in other future decisions) than the reasons it vanquished. The prior independent test of a preference therefore need not discover one that existed; it may establish a preference which then consistently carries over into a new decision situation. The testing procedure cannot show that we always act on a preexistingly strongest preference or motive.The claim that we always do what we most prefer or always act from the strongest motive is sometimes said to be empty of content, since the preference or the strength of motive is identified by what the person does. If the claim is to have empirical content, it must sometimes be possible to discover what a person’s preference or strongest motive is via some other situation, to independently identify it in order then to check in this situation whether the person is doing what he most prefers or has the strongest motive to do.
 Shall we say, though, that every free decision involves a conflict of some sort, with reasons pulling in different directions? The reasons in conflict need not then have indeterminate weight, for a free decision may “act out” an earlier weighting decision as precedent. (But is there always present a reason of indeterminate weight to reexamine and overturn an earlier precedent, which reason itself must be given a determinate lesser weight in the decision to follow the precedent?) Even though it will include no interesting cases we especially want to judge, still, we may formulate the theory to avoid the uncomfortable consequence that actions in the face of no contrary reasons are not free ones.Only when there are opposed reasons for different actions is it necessary to arrive at a weighting; otherwise, one can just do what all the reasons favor. However, neither group of these opposed reasons need be moral; decisions that involve a conflict of duty or other moral motives with (nonmoral) desires are only a subclass of the free decisions.
How can the giving of weights be other than random? Since (by hypothesis) there is no cause for giving or bestowing these particular weights on reasons rather than other weights, must it be merely a random act when these are bestowed?... If the absence of causation entailed randomness, then the denial of (contra-causal) free will would follow immediately. However, ‘uncaused’ does not entail ‘random’. To be sure, the theorist of free will still has to explain wherein the act not causally determined is nonrandom, but at least there is room for this task.
 The question remains: how is her decision among the alternatives causally open to her (the alternatives it is not causally determined she won’t choose) not simply a random matter?In what way is the bestowal of weights not simply random? There may be causes limiting the reasons on which (nonzero) weight can be bestowed, and the interval within which these weights fall may similarly be limited. However, although it is not a random matter that the weights bestowed fall within this range, neither is that decided by the person.
Understanding and Explaining Free Choices
 We might interpret those theorists who pointed to our choices not as trying to prove that we made free choices but as ostensively explaining the notion, showing its intelligibility. Were they saying that we understand free choice and agency by virtue of making free choices as agents?... Our problem is that we are puzzled about the nature of free choices, so any inside knowledge we may have of such choices due to and in making them obviously hasn’t served to clear up our puzzles about their nature. It is tempting to say our puzzlement stems from supposing we must be able discursively to say or describe what a free choice is like, yet the fact that we cannot, when we are directly acquainted with them, doesn’t interfere with understanding them. But too many ineffabilities spoil the philosophical broth. Since I do not myself have even the feeling of understanding, I will continue the (discursive) attempts at explanation.In what...way...can we understand the process of making free choices? By making them, perhaps.
...We have said already that the decision process (sometimes) bestows weights on the reasons for and against the various alternatives, and that this bestowal of weights is self-subsuming and so to that extent not random. Still, there can be different self-subsuming bestowals of weight. Although after one occurs we will be able retrospectively to give a reason as the cause (though without causal determination), can anything be said about why that one self-subsuming decision is made rather than another? No, the weights are bestowed in virtue of weights that come into effect in the very act of bestowal. This is the translation into this context of the notion of reflexivity: the phenomenon, such as reference or a law’s holding, has an “inside” character when it holds or occurs in virtue of a feature bestowed by its holding or occurring.
Suppose a process of decision can have these features, bestowing weights in a self-subsuming fashion which is reflexive. The decision then does not simply dangle there at random—we can see the many ties and connections it has (including internal ones); the particular decision is not inexplicable—we see it as something that could arise from a process of this sort.
More might be demanded, however; it might be demanded that the theorist of free will show how the decision is causally determined. Otherwise, it will be said, the character and nature of the decision will remain mysterious. But clearing up any mystery in that way would come at the cost of the act’s contra-causal freedom. No adequate condition on explanation or understanding necessitates... causal explanation.... Free will is to be explained differently, by delineating a decision process that can give rise to various acts in a nonrandom nonarbitrary way; whichever it gives rise to—and it could give rise to any one of several—will happen nonarbitrarily. These remarks are independent of the particular process we have delineated here, involving the bestowal of weights, reflexive self-subsumption, and so on. What is inappropriate is to demand that a free choice be explained in a way that shows it is unfree.
 I have more worries about terming this bestowal nonarbitrary and nonrandom because it is self-subsuming and reflexive. This position has too much the flavor of applying shiny new tools and ideas everywhere, as a magic key—except that some of the applications depend, perhaps, upon these ideas being not so well understood, not so shiny. So we should be somewhat wary of this use of the themes of self-subsumption and reflexiveness to delineate the nonarbitrary nature of a free choice. They do have the right flavor, though....The theme of the bestowal of weights to reasons, in a situation of no preexistingly determinate weights, seems to me phenomenologically accurate and proper to emphasize.
Could One Have Bestowed Otherwise?
The process of decision can yield the intentional doing of different actions, and it would have if different weights had been assigned, which could have happened. But does it follow that the person could have done otherwise, that it was within the person’s power to bestow different weights, as opposed to that merely happening? In what way could the person have done otherwise, not merely been the arena in which otherwise happened?
From Philosophical Explanations (Cambridge, Mass.: Harvard University Press, 1981).
1. Is Nozick right that his view seems “phenomenologically accurate”? Think of a realistic example or two and try to spell out just how Nozick’s view of such a case differs from that of either the hard determinist or the two sorts of compatibilists. (The comparison with Frankfurt is especially interesting.)
2. Does the idea of self-subsumption solve the problem of why the assigning of weights isn’t random? Does it even contribute to a solution of this problem? Why or why not?
3. Nozick concedes that he does not “have even the feeling of understanding” with regard to the question of how a free choice really works, how it is different from either determinism or randomness. Assuming that no one is in any better position in this respect, how serious a problem is this for libertarianism? Is it possible that we have good reasons for thinking that there are such choices and perhaps even that we can tell which ones they are, even if we don’t really understand their nature?
 This is a very strong requirement. Can you think of an example of an action that it is plausible to think is caused solely by your self, with no causal influence by anything else? Indeed, does it even make sense to think of a self—as opposed to one or more of its states, such as desires or motives—as a cause?
 Campbell adds the requirement that the person could have done otherwise, which means on his view that the nature of the self or agent did not necessitate the action in question—or else that the nature that does necessitate the action was itself a result of one or more earlier choices, in which the self or agent could have done otherwise.
 This sounds most like Blatchford. But all the other authors so far in this chapter would agree pretty closely that freedom in the sense that Campbell describes is impossible.
 Again, this is a very strong requirement—one that may seem quite impossible to satisfy.
 Of course, none of the compatibilists would agree that moral responsibility requires this self-determining sort of freedom.
 Campbell is referring to the actual experience of choice, especially in a situation in which one is pulled in more than one way.
 The morally wrong action, B, is supposed to be the one that your “desiring nature” would choose and thus the one that you want most, all things considered—the one that accords with the overall balance of your desires. You may have some desire to do the morally right thing, A, but not enough, by itself, to overcome the stronger desire for B. Everyone has experienced situations that at least seem to fit this specification, and you should think of an example of your own.
 New information or just a change of mind may alter the balance of desire in favor of A. Acting in accord with the new balance of desire would be simply doing what follows most naturally from one’s “desiring nature” and would not, in Campbell’s view, be a genuine example of free choice.
 Note carefully that Campbell is here describing only what he thinks we all believe to be true in a situation of this sort: that it is within one’s power to exert the effort to overcome the overall balance of desire in favor of B and do A instead. (Is he right that this is what everyone believes about such a situation? Do you believe it? Think about this question in relation to some examples.)
 If a person really does have the ability to put forth this special sort of effort (and also, of course, not to do it), then whether or not he does so seems to be controlled by him alone and not to any degree determined by other causes (since all other causes merely contribute to constituting the situation in relation to which this effort is or is not exerted).
 While formed character determines the balance of one’s desires, the choice as to whether to exert the effort to overcome that balance of desires or not cannot be due to formed character—if such a thing really occurs.
 Campbell claims, first, that a person cannot help believing in such a situation that he does have the ability in question, that whether he chooses to exert the effort or not is not determined by his formed character. Second, he claims is that this is enough to create “a strong prima facie presumption” in favor of libertarianism, thereby putting the burden of proof strongly on the opposing position. (ST Is he right about this second claim, even if the first is granted? Is the fact that in such a situation of choice we cannot help believing that we have this sort of freedom a reason for thinking that the libertarian claim that we do have it is true?)
 Here is an obvious objection to libertarianism, one that is raised by both Blatchford and Hume.
 Freedom, according to Campbell, involves a choice between one’s overall balance of desires and the demands of morality. Since both one’s desires and one’s conception of morality are a part of the formed character that is a result of heredity and environment, there will still be a basis for many approximate predictions: predictions that a person will not do a wide range of things that are favored by neither desire nor morality and that his action will fall within a certain range (one that includes both the desired option and the one favored by morality).
 Thus, if situations of choice between desire and morality are fairly rare, most actions will be predictable on the basis for formed character. This apparently means that we are free and morally responsible on only these fairly rare occasions.
 But if it is very hard to overcome the balance of desire that results from formed character, is a person who fails to do so still fully free and fully morally responsible? Isn’t his action very largely determined by formed character, which is in turn determined by environment and heredity, leaving very little room for freedom?
 There is one obvious way in which a choice that does not accord with the agent’s character seems to be possible: if the choice is simply random. But a random choice would not be the act of the self in any sense relevant to moral responsibility. Thus another way to put the problem is to say that it is hard to understand how a choice can be neither the result of formed character nor random. If it is neither of these, where does it come from?
 Campbell appeals to our experience of actually making choices in this way. But even if he is right (is he?) that our experience seems to support the view that such choices actually occur, this isn’t enough to explain what is going on: what determination by the self but not by the formed character or nature of the self really amounts to. Or how such a thing is possible.
 Again the real problem is not so much whether there is a kind of evidence for the existence of such choices as whether we can understand what such a choice would amount to and how it would be arrived at.
 Think of a situation involving a choice between two or more alternatives, where there are reasons for each of the alternatives, but where it is uncertain how much weight the various reasons should have. One might think that the agent in such a situation is trying to figure out what weights the various reasons already possess, but Nozick’s suggestion is that in the most important cases there are no antecedent weights to be discovered, so that they have to be assigned.
 What reasons an agent finds relevant may be determined by antecedent causes. It is only the weights of the reasons that are claimed, in a case of genuinely free choice, not to be thus determined.
 Thus whichever act he ultimately performs, he could have done otherwise—not just in the weak compatibilist sense that he would have done something different if his desires or character had been different, but in the strong sense that the other act could have occurred with no change in his desires or character prior to the moment of decision.
 Nozick’s view is that the action is still caused by the reason to which the agent assigns the greater weight, even though not causally determined (since it was not determined that this reason would be assigned greater weight).
 Nozick is saying that his view cannot accept this metaphysical thesis.
 But this is somewhat misleading. The cause of the action isn’t really RA by itself, but rather RA with an assigned weight that exceeds that of RB. And there is a general law that a more weighty reason will triumph over a less weighty one. It is the assigning of weights that is the ultimate cause of the action, and that assigning seems to be neither determined nor caused on Nozick’s view.
 In this way, Nozick is suggesting, such a choice can shape a person’s subsequent desires and overall character in at least roughly the way that Edwards thinks is required for genuine freedom and moral responsibility (but also thinks to be impossible).
 Blatchford, among others, claims that a person always acts from his antecedently strongest motive. Nozick questions how this motive is to be identified (since to say that it is just whatever motive he ends up acting on would reduce the claim to the trivial thesis that he acts on whatever motive he acts on—hardly something anyone would deny). Thus some other way of identifying the antecedently strongest motive is needed.
 One solution to this problem is to identify the pre-existing strongest motive by what the person says or does on some other occasion. Nozick responds that this procedure cannot distinguish between a motive that was antecedently the strongest motive and one that becomes the strongest as a result of an assignment of strength by the agent (on the occasion of the question or previous action)—and thus cannot support Blatchford’s claim. (ST Is there any other way to identify an antecedently strongest motive, one that distinguishes it from a motive that becomes strongest only by being at some point assigned greater weight?)
 Here is one important way in which Nozick’s view differs from Campbell’s/
 Nozick does not really explain how this can be done.
 Thus the idea of assigning weights to reasons cannot by itself solve the central problem for libertarianism: that of finding a third alternative to determinism and randomness.
 The idea is that the limitations on the reasons and the interval would themselves be causally determined.
 A random occurrence is sometimes thought of as one that happens for no reason. Nozick is suggesting that there might, in a way, be a reason for the assigning of weights. It might operate via a general principle that is self-subsuming: that is, whose adoption is an instance of itself. Nozick’s example is the general policy or principle of choosing in such a way as to “track bestness,” that is, do whatever is morally best. Adopting that principle is itself arguably an instance of doing what is morally best, so that the principle is self-subsuming. (Another way to put this is to say that the principle adopted may provide a reason for the adoption of that very principle.)
 Such a self-subsuming decision may be as general as a choice of an overall kind of life: live the morally best life, where adopting that principle is part of living the morally best life. C2 But while there is in this odd way a reason for the adoption of such a principle or conception of one’s life, in the most important cases there will be alternative choices that would have similarly been self-subsuming. So why isn’t the choice between these still arbitrary—that is, random?
 This is essentially Campbell’s suggestion.
 Experiencing such a choice doesn’t really explain how it works and in particular how it can be neither determined nor random. Nozick concedes (and it is a very large concession) that he doesn’t really even seem to himself to understand this.
 To say that the assignment of weights is reflexive is to say that it is adopted by virtue of itself: it explains its own adoption. (But does it really?)
 To say that this picture seems “phenomenologically accurate” is to say that it seems to accurately reflect what goes on in an actual experience of free choice.
 If the assignment of weights is a part of creating or constituting the sort of self that one chooses to be, then there will be another (related) way that it is nonrandom: it reflects the very nature of the resulting self. (C2 But there would still be alternative choices that would reflect the nature of alternative resulting selves, still making it unclear why the choice among these different selves isn’t random.)
 Nozick is saying that the choice of self was something the person did, not something that merely happened to him (either as a result of causal determination or randomness). Thus the person could have done otherwise in the sense that he could have chosen a different self with a correspondingly different assignment of weights. (C2 But this still doesn’t explain why the choice between the different weights and the different selves isn’t random.)