1. During the greater part of the last lecture which was concerned with the defence of the notion of self-activity and with the classification of its main species we were operating on the very threshold of the problem of Free Will; and in its later stages particularly in connection with the analysis of moral-decision activity we may perhaps be judged to have passed beyond the threshold. The present lecture in which we address ourselves formally to the Free Will problem is in fact so closely continuous with its predecessor that I should wish the two lectures to be regarded as constituting in a real sense a single unit.
In the later more constructive part of my programme today this intimate dependence upon what has gone before will become very apparent. My initial task however must be one of elucidation and definition. The general question I have to try to answer a question which is very far indeed from admitting of a ready answer is What precisely is the Free Will problem?
It is something of a truism that in philosophic enquiry the exact formulation of a problem often takes one a long way on the road to its solution. In the case of the Free Will problem I think there is a rather special need of careful formulation. For there are many sorts of human freedom; and it can easily happen that one wastes a great deal of labour in proving or disproving a freedom which has almost nothing to do with the freedom which is at issue in the traditional problem of Free Will. The abortiveness of so much of the argument for and against Free Will in contemporary philosophical literature seems to me due in the main to insufficient pains being taken over the preliminary definition of the problem. There is indeed one outstanding exception Professor Broad's brilliant inaugural lecture entitled ‘Determinism Indeterminism and Libertarianism’1 in which forty-three pages are devoted to setting out the problem as against seven to its solution! I confess that the solution does not seem to myself to follow upon the formulation quite as easily as all that:2 but Professor Broad's eminent example fortifies me in my decision to give here what may seem at first sight a disproportionate amount of time to the business of determining the essential characteristics of the kind of freedom with which the traditional problem is concerned.
Fortunately we can at least make a beginning with a certain amount of confidence. It is not seriously disputable that the kind of freedom in question is the freedom which is commonly recognised to be in some sense a precondition of moral responsibility. Clearly it is on account of this integral connection with moral responsibility that such exceptional importance has always been felt to attach to the Free Will problem. But in what precise sense is free will a precondition of moral responsibility and thus a postulate of the moral life in general? This is an exceedingly troublesome question; but until we have satisfied ourselves about the answer to it we are not in a position to state let alone decide the question whether ‘Free Will’ in its traditional ethical significance is a reality.
Our first business then is to ask exactly what kind of freedom is it which is required for moral responsibility? And as to method of procedure in this inquiry there seems to me to be no real choice. I know of only one method that carries with it any hope of success; viz. the critical comparison of those acts for which on due reflection we deem it proper to attribute moral praise or blame to the agents with those acts for which on due reflection we deem such judgments to be improper. The ultimate touchstone as I see it can only be our moral consciousness as it manifests itself in our more critical and considered moral judgments. The ‘linguistic’ approach by way of the analysis of moral sentences seems to me despite its present popularity to be an almost infallible method for reaching wrong results in the moral field; but I must reserve what I have to say about this for the next lecture.
2. The first point to note is that the freedom at issue (as indeed the very name ‘Free Will Problem’ indicates) pertains primarily not to overt acts but to inner acts. The nature of things has decreed that save in the case of one's self it is only overt acts which one can directly observe. But a very little reflection serves to show that in our moral judgments upon others their overt acts are regarded as significant only in so far as they are the expression of inner acts. We do not consider the acts of a robot to be morally responsible acts; nor do we consider the acts of a man to be so save in so far as they are distinguishable from those of a robot by reflecting an inner life of choice. Similarly from the other side if we are satisfied (as we may on occasion be at least in the case of ourselves) that a person has definitely elected to follow a course which he believes to be wrong but has been prevented by external circumstances from translating his inner choice into an overt act we still regard him as morally blameworthy. Moral freedom then pertains to inner acts.
The next point seems at first sight equally obvious and uncontroversial; but as we shall see it has awkward implications if we are in real earnest with it (as almost nobody is). It is the simple point that the act must be one of which the person judged can be regarded as the sole author. It seems plain enough that if there are any other determinants of the act external to the self to that extent the act is not an act which the self determines and to that extent not an act for which the self can be held morally responsible. The self is only part—author of the act and his moral responsibility can logically extend only to those elements within the act (assuming for the moment that these can be isolated) of which he is the sole author.
The awkward implications of this apparent truism will be readily appreciated. For if we are mindful of the influences exerted by heredity and environment we may well feel some doubt whether there is any act of will at all of which one can truly say that the self is sole author sole determinant. No man has a voice in determining the raw material of impulses and capacities that constitute his hereditary endowment and no man has more than a very partial control of the material and social environment in which he is destined to live his life. Yet it would be manifestly absurd to deny that these two factors do constantly and profoundly affect the nature of a man's choices. That this is so we all of us recognise in our moral judgments when we ‘make allowances’ as we say for a bad heredity or a vicious environment and acknowledge in the victim of them a diminished moral responsibility for evil courses. Evidently we do try in our moral judgments however crudely to praise or blame a man only in respect of that of which we can regard him as wholly the author. And evidently we do recognise that for a man to be the author of an act in the full sense required for moral responsibility it is not enough merely that he ‘wills’ or ‘chooses’ the act: since even the most unfortunate victim of heredity or environment does as a rule ‘will’ what he does. It is significant however that the ordinary man though well enough aware of the influence upon choices of heredity and environment does not feel obliged thereby to give up his assumption that moral predicates are somehow applicable. Plainly he still believes that there is something for which a man is morally responsible something of which we can fairly say that he is the sole author. What is this something? To that question common-sense is not ready with an explicit answer—though an answer is I think implicit in the line which its moral judgments take. I shall do what I can to give an explicit answer later in this lecture. Meantime it must suffice to observe that if we are to be true to the deliverances of our moral consciousness it is very difficult to deny that sole authorship is a necessary condition of the morally responsible act.
Thirdly we come to a point over which much recent controversy has raged. We may approach it by raising the following question. Granted an act of which the agent is sole author does this ‘sole authorship’ suffice to make the act a morally free act? We may be inclined to think that it does until we contemplate the possibility that an act of which the agent is sole author might conceivably occur as a necessary expression of the agent's nature; the way in which e.g. some philosophers have supposed the Divine act of creation to occur. This consideration excites a legitimate doubt; for it is far from easy to see how a person can be regarded as a proper subject for moral praise or blame in respect of an act which he cannot help performing—even if it be his own ‘nature’ which necessitates it. Must we not recognise it as a condition of the morally free act that the agent ‘could have acted otherwise’ than he in fact did? It is true indeed that we sometimes praise or blame a man for an act about which we are prepared to say in the light of our knowledge of his established character that he ‘could no other’. But I think that a little reflection shows that in such cases we are not praising or blaming the man strictly for what he does now (or at any rate we ought not to be) but rather for those past acts of his which have generated the firm habit of mind from which his present act follows ‘necessarily.’ In other words our praise and blame so far as justified are really retrospective being directed not to the agent qua performing this act but to the agent qua performing those past acts which have built up his present character and in respect to which we presume that he could have acted otherwise that there really were open possibilities before him. These cases therefore seem to me to constitute no valid exception to what I must take to be the rule viz. that a man can be morally praised or blamed for an act only if he could have acted otherwise.
Now philosophers today are fairly well agreed that it is a postulate of the morally responsible act that the agent ‘could have acted otherwise’ in some sense of that phrase. But sharp differences of opinion have arisen over the way in which the phrase ought to be interpreted. There is a strong disposition to water down its apparent meaning by insisting that it is not (as a postulate of moral responsibility) to be understood as a straightforward categorical proposition but rather as a disguised hypothetical proposition. All that we really require to be assured of in order to justify our holding X morally responsible for an act is we are told that X could have acted otherwise if he had chosen otherwise (Moore Stevenson); or perhaps that X could have acted otherwise if he had had a different character or if he had been placed in different circumstances.
I think it is easy to understand and even in a measure to sympathise with the motives which induce philosophers to offer these counter-interpretations. It is not just the fact that ‘X could have acted otherwise’ as a bald categorical statement is incompatible with the universal sway of causal law—though this is to some philosophers a serious stone of stumbling. The more widespread objection is that it at least looks as though it were incompatible with that causal continuity of an agent's character with his conduct which is implied when we believe (surely with justice) that we can often tell the sort of thing a man will do from our knowledge of the sort of man he is.
We shall have to make our accounts with that particular difficulty later. At this stage I wish merely to show that neither of the hypothetical propositions suggested—and I think the same could be shown for any hypothetical alternative—is an acceptable substitute for the categorical proposition ‘X could have acted otherwise’ as the presupposition of moral responsibility.
Let us look first at the earlier suggestion—‘X could have acted otherwise if he had chosen otherwise’. Now clearly there are a great many acts with regard to which we are entirely satisfied that the agent is thus situated. We are often perfectly sure that—for this is all it amounts to—if X had chosen otherwise the circumstances presented no external obstacle to the translation of that choice into action. For example we often have no doubt at all that X who in point of fact told a lie could have told the truth if he had so chosen. But does our confidence on this score allay all legitimate doubts about whether X is really blameworthy? Does it entail that X is free in the sense required for moral responsibility? Surely not. The obvious question immediately arises: ‘But could X have chosen otherwise than he did?’ It is doubt about the true answer to that question which leads most people to doubt the reality of moral responsibility. Yet on this crucial question the hypothetical proposition which is offered as a sufficient statement of the condition justifying the ascription of moral responsibility gives us no information whatsoever.
Indeed this hypothetical substitute for the categorical ‘X could have acted otherwise’ seems to me to lack all plausibility unless one contrives to forget why it is after all that we ever come to feel fundamental doubts about man's moral responsibility. Such doubts are born surely when one becomes aware of certain reputable world-views in religion or philosophy or of certain reputable scientific beliefs which in their several ways imply that man's actions are necessitated and thus could not be otherwise than they in fact are. But clearly a doubt so based is not even touched by the recognition that a man could very often act otherwise if he so chose. That proposition is entirely compatible with the necessitarian theories which generate our doubt: indeed it is this very compatibility that has recommended it to some philosophers who are reluctant to give up either moral responsibility or Determinism. The proposition which we must be able to affirm if moral praise or blame of X is to be justified is the categorical proposition that X could have acted otherwise because—not if—he could have chosen otherwise; or since it is essentially the inner side of the act that matters the proposition simply that X could have chosen otherwise.
For the second of the alternative formulae suggested we cannot spare more than a few moments. But its inability to meet the demands it is required to meet is almost transparent. ‘X could have acted otherwise’ as a statement of a precondition of X's moral responsibility really means (we are told) ‘X could have acted otherwise if he were differently constituted or if he had been placed in different circumstances’. It seems a sufficient reply to this to point out that the person whose moral responsibility is at issue is X; a specific individual in a specific set of circumstances. It is totally irrelevant to X's moral responsibility that we should be able to say that some person differently constituted from X or X in a different set of circumstances could have done something different from what X did.
3. Let me then briefly sum up the answer at which we have arrived to our question about the kind of freedom required to justify moral responsibility. It is that a man can be said to exercise free will in a morally significant sense only in so far as his chosen act is one of which he is the sole cause or author and only if—in the straightforward categorical sense of the phrase—he ‘could have chosen otherwise’.
I confess that this answer is in some ways a disconcerting one. disconcerting because most of us however objective we are in the actual conduct of our thinking would like to be able to believe that moral responsibility is real: whereas the freedom required for moral responsibility on the analysis we have given is certainly far more difficult to establish than the freedom required on the analyses we found ourselves obliged to reject. If e.g. moral freedom entails only that I could have acted otherwise if I had chosen otherwise there is no real ‘problem’ about it at all. I am ‘free’ in the normal case where there is no external obstacle to prevent my translating the alternative choice into action and not free in other cases. Still less is there a problem if all that moral freedom entails is that I could have acted otherwise if I had been a differently constituted person or been in different circumstances. Clearly I am always free in this sense of freedom. But as I have argued these so-called ‘freedoms’ fail to give us the pre-conditions of moral responsibility and hence leave the freedom of the traditional free-will problem the freedom that people are really concerned about precisely where it was.
4. Another interpretation of freedom which I am bound to reject on the same general ground i.e. that it is just not the kind of freedom that is relevant to moral responsibility is the old idealist view which identifies the free will with the rational will; the rational will in its turn being identified with the will which wills the moral law in whole-hearted single-minded obedience to it. This view is still worth at least a passing mention if only because it has recently been resurrected in an interesting work by Professor A. E. Teale.3 Moreover I cannot but feel a certain nostalgic tenderness for a view in which I myself was (so to speak) philosophically cradled. The almost apostolic fervour with which my revered nursing-mother the late Sir Henry Jones was wont to impart it to his charges and hardly less his ill-concealed scorn for ignoble natures (like my own) which still hankered after a free will in the old ‘vulgar’ sense are vividly recalled for me in Professor Teale's stirring pages.
The true interpretation of free will according to Professor Teale the interpretation to which Kant despite occasional back-slidings adhered in his better moments is that ‘the will is free in the degree that it is informed and disciplined by the moral principle.’4
Now this is a perfectly intelligible sense of the word ‘free’—or at any rate it can be made so with a little explanatory comment which Professor Teale well supplies but for which there is here no space. But clearly it is a very different sort of freedom from that which is at issue in the traditional problem of free will. This idealist ‘freedom’ sponsored by Teale belongs on his own showing only to the self in respect of its good willing. The freedom with which the traditional problem is concerned inasmuch as it is the freedom presupposed by moral responsibility must belong to the self in respect of its bad no less than its good willing. It is in fact the freedom to decide between genuinely open alternatives of good and bad willing.
Professor Teale of course is not unaware that the freedom he favours differs from freedom as traditionally understood. He recognises the traditional concept under its Kantian title of ‘elective’ freedom. But he leaves the reader in no kind of doubt about his disbelief in both the reality and the value of this elective freedom to do or forbear from doing one's duty.
The question of the reality of elective freedom I shall be dealing with shortly; and it will occupy us to the end of the lecture. At the moment I am concerned only with its value and with the rival view that all that matters for the moral life is the ‘rational’ freedom which a man has in the degree that his will is ‘informed and disciplined by the moral principle’. I confess that to myself the verdict on the rival view seems plain and inescapable. No amount of verbal ingenuity or argumentative convolutions can obscure the fact that it is in flat contradiction to the implications of moral responsibility. The point at issue is really perfectly straightforward. If as this idealist theory maintains my acting in defiance of what I deem to be my duty is not a ‘free’ act in any sense let alone in the sense that ‘I could have acted otherwise’ then I cannot be morally blameworthy and that is all there is to it. Nor for that matter is the idealist entitled to say that I am morally praiseworthy if I act dutifully; for although that act is a ‘free’ act in the idealist sense it is on his own avowal not free in the sense that ‘I could have acted otherwise’.
It seems to me idle therefore to pretend that if one has to give up freedom in the traditional elective sense one is not giving up anything important. What we are giving up is quite simply the reality of the moral life. I recognise that to a certain type of religious nature (as well as by an odd meeting of extremes to a certain type of secular nature) that does not appear to matter so very much; but for myself I still think it sufficiently important to make it well worth while enquiring seriously into the possibility that the elective freedom upon which it rests may be real after all.
5. That brings me to the second and more constructive part of this lecture. From now on I shall be considering whether it is reasonable to believe that man does in fact possess a free will of the kind specified in the first part of the lecture. If so just how and where within the complex fabric of the volitional life are we to locate it?—for although free will must presumably belong (if anywhere) to the volitional side of human experience it is pretty clear from the way in which we have been forced to define it that it does not pertain simply to volition as such; not even to all volitions that are commonly dignified with the name of ‘choices’. It has been I think one of the more serious impediments to profitable discussion of the Free Will problem that Libertarians and Determinists alike have so often failed to appreciate the comparatively narrow area within which the free will that is necessary to ‘save’ morality is required to operate. It goes without saying that this failure has been gravely prejudicial to the case for Libertarianism. I attach a good deal of importance therefore to the problem of locating free will correctly within the volitional orbit. Its solution forestalls and annuls I believe some of the more tiresome cliches of Determinist criticism.
We saw earlier that Common Sense's practice of ‘making allowances’ in its moral judgments for the influence of heredity and environment indicates Common Sense's conviction both that a just moral judgment must discount determinants of choice over which the agent has no control and also (since it still accepts moral judgments as legitimate) that something of moral relevance survives which can be regarded as genuinely self-originated. We are now to try to discover what this ‘something’ is. And I think we may still usefully take Common Sense as our guide. Suppose one asks the ordinary intelligent citizen why he deems it proper to make allowances for X whose heredity and/or environment are unfortunate. He will tend to reply I think in some such terms as these: that X has more and stronger temptations to deviate from what is right than Y or Z who are normally circumstanced so that he must put forth a stronger moral effort if he is to achieve the same level of external conduct. The intended implication seems to be that X is just as morally praiseworthy as Y or Z if he exerts an equivalent moral effort even though he may not thereby achieve an equal success in conforming his will to the ‘concrete’ demands of duty. And this implies again Common Sense's belief that in moral effort we have something for which a man is responsible without qualification something that is not affected by heredity and environment but depends solely upon the self itself.
Now in my opinion Common Sense has here in principle hit upon the one and only defensible answer. Here and here alone so far as I can see in the act of deciding whether to put forth or withhold the moral effort required to resist temptation and rise to duty is to be found an act which is free in the sense required for moral responsibility; an act of which the self is sole author and of which it is true to say that ‘it could be’ (or after the event ‘could have been’) ‘otherwise’. Such is the thesis which we shall now try to establish.
6. The species of argument appropriate to the establishment of a thesis of this sort should fall I think into two phases. First there should be a consideration of the evidence of the moral agent's own inner experience. What is the act of moral decision and what does it imply from the standpoint of the actual participant? Since there is no way of knowing the act of moral decision—or for that matter any other form of activity—except by actual participation in it the evidence of the subject or agent is on an issue of this kind of palmary importance. It can hardly however be taken as in itself conclusive. For even if that evidence should be overwhelmingly to the effect that moral decision does have the characteristics required by moral freedom the question is bound to be raised—and in view of considerations from other quarters pointing in a contrary direction is rightly raised—Can we trust the evidence of inner experience? That brings us to what will be the second phase of the argument. We shall have to go on to show if we are to make good our case that the extraneous considerations so often supposed to be fatal to the belief in moral freedom are in fact innocuous to it.
In the light of what was said in the last lecture about the self's experience of moral decision as a creative activity we may perhaps be absolved from developing the first phase of the argument at any great length. The appeal is throughout to one's own experience in the actual taking of the moral decision in the situation of moral temptation. ‘Is it possible’ we must ask ‘for anyone so circumstanced to disbelieve that he could be deciding otherwise?’ The answer is surely not in doubt. When we decide to exert moral effort to resist a temptation we feel quite certain that we could withhold the effort; just as if we decide to withhold the effort and yield to our desires we feel quite certain that we could exert it—otherwise we should not blame ourselves afterwards for having succumbed. It may be indeed that this conviction is mere self-delusion. But that is not at the moment our concern. It is enough at present to establish that the act of deciding to exert or to withhold moral effort as we know it from the inside in actual moral living belongs to the category of acts which ‘could have been otherwise’.
Mutatis mutandis the same reply is forthcoming if we ask ‘Is it possible for the moral agent in the taking of his decision to disbelieve that he is the sole author of that decision?’ Clearly he cannot disbelieve that it is he who takes the decision. That however is not in itself sufficient to enable him on reflection to regard himself as solely responsible for the act. For his ‘character’ as so far formed might conceivably be a factor in determining it and no one can suppose that the constitution of his ‘character’ is uninfluenced by circumstances of heredity and environment with which he has nothing to do. But as we pointed out in the last lecture the very essence of the moral decision as it is experienced is that it is a decision whether or not to combat our strongest desire and our strongest desire is the expression in the situation of our character as so far formed. Now clearly our character cannot be a factor in determining the decision whether or not to oppose our character. I think we are entitled to say therefore that the act of moral decision is one in which the self is for itself not merely ‘author’ but ‘sole author’.
7. We may pass on then to the second phase of our constructive argument; and this will demand more elaborate treatment. Even if a moral agent qua making a moral decision in the situation of ‘temptation’ cannot help believing that he has free will in the sense at issue—a moral freedom between real alternatives between genuinely open possibilities—are there nevertheless objections to a freedom of this kind so cogent that we are bound to distrust the evidence of ‘inner experience’?
I begin by drawing attention to a simple point whose significance tends I think to be under-estimated. If the phenomenological analysis we have offered is substantially correct no one while functioning as a moral agent can help believing that he enjoys free will. Theoretically he may be completely convinced by Determinist arguments but when actually confronted with a personal situation of conflict between duty and desire he is quite certain that it lies with him here and now whether or not he will rise to duty. It follows that if Determinists could produce convincing theoretical arguments against a free will of this kind the awkward predicament would ensue that man has to deny as a theoretical being what he has to assert as a practical being. Now I think the Determinist ought to be a good deal more worried about this than he usually is. He seems to imagine that a strong case on general theoretical grounds is enough to prove that the ‘practical’ belief in free will even if inescapable for us as practical beings is mere illusion. But in fact it proves nothing of the sort. There is no reason whatever why a belief that we find ourselves obliged to hold qua practical beings should be required to give way before a belief which we find ourselves obliged to hold qua theoretical beings; or for that matter vice versa. All that the theoretical arguments of Determinism can prove unless they are reinforced by a refutation of the phenomenological analysis that supports Libertarianism is that there is a radical conflict between the theoretical and the practical sides of man's nature an antimony at the very heart of the self. And this is a state of affairs with which no one can easily rest satisfied. I think therefore that the Determinist ought to concern himself a great deal more than he does with phenomenological analysis in order to show if he can that the assurance of free will is not really an inexpugnable element in man's practical consciousness. There is just as much obligation upon him convinced though he may be of the soundness of his theoretical arguments to expose the errors of the Libertarian's phenomenological analysis as there is upon us convinced though we may be of the soundness of the Libertarian's phenomenological analysis to expose the errors of the Determinist's theoretical arguments.
8. However we must at once begin the discharge of our own obligation. The rest of this lecture will be devoted to trying to show that the arguments which seem to carry most weight with Determinists are to say the least of it very far from compulsive.
Fortunately a good many of the arguments which at an earlier time in the history of philosophy would have been strongly urged against us make almost no appeal to the bulk of philosophers today and we may here pass them by. That applies to any criticism of ‘open possibilities’ based on a metaphysical theory about the nature of the universe as a whole. Nobody today has a metaphysical theory about the nature of the universe as a whole! It applies also with almost equal force to criticisms based upon the universality of causal law as a supposed postulate of science. There have always been in my opinion sound philosophic reasons for doubting the validity as distinct from the convenience of the causal postulate in its universal form but at the present time when scientists themselves are deeply divided about the need for postulating causality even within their own special field we shall do better to concentrate our attention upon criticisms which are more confidently advanced. I propose to ignore also on different grounds the type of criticism of free will that is sometimes advanced from the side of religion based upon religious postulates of Divine Omnipotence and Omniscience So far as I can see a postulate of human freedom is every bit as necessary to meet certain religious demands (e.g. to make sense of the ‘conviction of sin’) as postulates of Divine Omniscience and Omnipotence are to meet certain other religious demands. If so then it can hardly be argued that religious experience as such tells more strongly against than for the position we are defending; and we may be satisfied in the present context to leave the matter there. It will be more profitable to discuss certain arguments which contemporary philosophers do think important and which recur with a somewhat monotonous regularity in the literature of anti-Libertarianism.
These arguments can I think be reduced in principle to no more than two: first the argument from ‘predictability’; second the argument from the alleged meaninglessness of an act supposed to be the self's act and yet not an expression of the self's character. Contemporary criticism of free will seems to me to consist almost exclusively of variations on these two themes. I shall deal with each in turn.
9. On the first we touched in passing at an earlier stage. Surely it is beyond question (the critic urges) that when we know a person intimately we can foretell with a high degree of accuracy how he will respond to at least a large number of practical situations. One feels safe in predicting that one's dog-loving friend will not use his boot to repel the little mongrel that comes yapping at his heels; or again that one's wife will not pass with incurious eyes (or indeed pass at all) the new hat-shop in the city. So to behave would not be (as we say) ‘in character’. But so the criticism runs you with your doctrine of ‘genuinely open possibilities’ of a free will by which the self can diverge from its own character remove all rational basis from such prediction. You require us to make the absurd supposition that the success of countless predictions of the sort in the past has been mere matter of chance. If you really believed in your theory you would not be surprised if tomorrow your friend with the notorious horror of strong drink should suddenly exhibit a passion for whisky and soda or if your friend whose taste for reading has hitherto been satisfied with the sporting columns of the newspapers should be discovered on a fine Saturday afternoon poring over the works of Hegel. But of course you would be surprised. Social life would be sheer chaos if there were not well-grounded social expectations; and social life is not sheer chaos. Your theory is hopelessly wrecked upon obvious facts.
Now whether or not this criticism holds good against some versions of Libertarian theory I need not here discuss. It is sufficient if I can make it clear that against the version advanced in this lecture according to which free will is localised in a relatively narrow field of operation the criticism has no relevance whatsoever.
Let us remind ourselves briefly of the setting within which on our view free will functions. There is X, the course which we believe we ought to follow, and Y, the course towards which we feel our desire is strongest. The freedom which we ascribe to the agent is the freedom to put forth or refrain from putting forth the moral effort required to resist the pressure of desire and do what he thinks he ought to do.
But then there is surely an immense range of practical situations—covering by far the greater part of life—in which there is no question of a conflict within the self between what he most desires to do and what he thinks he ought to do? Indeed such conflict is a comparatively rare phenomenon for the majority of men. Yet over that whole vast range there is nothing whatever in our version of Libertarianism to prevent our agreeing that character determines conduct. In the absence real or supposed of any ‘moral’ issue what a man chooses will be simply that course which after such reflection as seems called for he deems most likely to bring him what he most strongly desires; and that is the same as to say the course to which his present character inclines him.
Over by far the greater area of human choices then our theory offers no more barrier to successful prediction on the basis of character than any other theory. For where there is no clash of strongest desire with duty the free will we are defending has no business. There is just nothing for it to do.
But what about the situations—rare enough though they may be—in which there is this clash and in which free will does therefore operate? Does our theory entail that there at any rate as the critic seems to suppose ‘anything may happen’?
Not by any manner of means. In the first place and by the very nature of the case the range of the agent's possible choices is bounded by what he thinks he ought to do on the one hand and what he most strongly desires on the other. The freedom claimed for him is a freedom of decision to make or withhold the effort required to do what he thinks he ought to do. There is no question of a freedom to act in some ‘wild’ fashion out of all relation to his characteristic beliefs and desires. This so-called ‘freedom of caprice’ so often charged against the Libertarian is to put it bluntly a sheer figment of the critic's imagination with no habitat in serious Libertarian theory. Even in situations where free will does come into play it is perfectly possible on a view like ours given the appropriate knowledge of a man's character to predict within certain limits how he will respond.
But ‘probable’ prediction in such situations can I think go further than this. It is obvious that where desire and duty are at odds the felt ‘gap’ (as it were) between the two may vary enormously in breadth in different cases. The moderate drinker and the chronic tippler may each want another glass and each deem it his duty to abstain but the felt gap between desire and duty in the case of the former is trivial beside the great gulf which is felt to separate them in the case of the latter. Hence it will take a far harder moral effort for the tippler than for the moderate drinker to achieve the same external result of abstention. So much is matter of common agreement. And we are entitled I think to take it into account in prediction on the simple principle that the harder the moral effort required to resist desire the less likely it is to occur. Thus in the example taken most people would predict that the tippler will very probably succumb to his desires whereas there is a reasonable likelihood that the moderate drinker will make the comparatively slight effort needed to resist them. So long as the prediction does not pretend to more than a measure of probability there is nothing in our theory which would disallow it.
I claim therefore that the view of free will I have been putting forward is consistent with predictability of conduct on the basis of character over a very wide field indeed. And I make the further claim that that field will cover all the situations in life concerning which there is any empirical evidence that successful prediction is possible.
10. Let us pass on to consider the second main line of criticism. This is I think much the more illuminating of the two if only because it compels the Libertarian to make explicit certain concepts which are indispensable to him but which being desperately hard to state clearly are apt not to be stated at all. The critic's fundamental point might be stated somewhat as follows:
‘Free will as you describe it is completely unintelligible. On your own showing no reason can be given because there just is no reason why a man decides to exert rather than to withhold moral effort or vice versa. But such an act—or more properly such an ‘occurrence’—it is nonsense to speak of as an act of a self. If there is nothing in the self's character to which it is even in principle in any way traceable the self has nothing to do with it. Your so-called ‘freedom” therefore so far from supporting the self's moral responsibility destroys it as surely as the crudest Determinism could do.’
If we are to discuss this criticism usefully it is important I think to begin by getting clear about two different senses of the word ‘intelligible’.
If in the first place we mean by an ‘intelligible’ act one whose occurrence is in principle capable of being inferred since it follows necessarily from something (though we may not know in fact from what) then it is certainly true that the Libertarian's free will is unintelligible. But that is only saying is it not that the Libertarian's ‘free’ act is not an act which follows necessarily from something! This can hardly rank as a criticism of Libertarian-ism. It is just a description of it. That there can be nothing unintelligible in this sense is precisely what the Determinist has got to prove.
Yet it is surprising how often the critic of Libertarianism involves himself in this circular mode of argument. Repeatedly it is urged against the Libertarian with a great air of triumph that on his view he can't say why I now decide to rise to duty or now decide to follow my strongest desire in defiance of duty. Of course he can't. If he could he wouldn't be a Libertarian. To ‘account for’ a ‘free’ act is a contradiction in terms. A free will is ex hypothesi the sort of thing of which the request for an explanation is absurd. The assumption that an explanation must be in principle possible for the act of moral decision deserves to rank as a classic example of the ancient fallacy of ‘begging the question’.
But the critic usually has in mind another sense of the word ‘unintelligible’. He is apt to take it for granted that an act which is unintelligible in the above sense (as the morally free act of the Libertarian undoubtedly is) is unintelligible in the further sense that we can attach no meaning to it. And this is an altogether more serious matter. If it could really be shown that the Libertarian's ‘free will’ were unintelligible in this sense of being meaningless that for myself at any rate would be the end of the affair. Libertarianism would have been conclusively refuted.
But it seems to me manifest that this can not be shown. The critic has allowed himself I submit to become the victim of a widely accepted but fundamentally vicious assumption. He has assumed that whatever is meaningful must exhibit its meaning-fulness to those who view it from the standpoint of external observation. Now if one chooses thus to limit one's self to the rôle of external observer it is I think perfectly true that one can attach no meaning to an act which is the act of something we call a ‘self’ and yet follows from nothing in that self's character. But then why should we so limit ourselves when what is under consideration is a subjective activity? For the apprehension of subjective acts there is another standpoint available that of inner experience of the practical consciousness in its actual functioning. If our free will should turn out to be something to which we can attach a meaning from this standpoint no more is required. And no more ought to be expected. For I must repeat that only from the inner standpoint of living experience could anything of the nature of ‘activity’ be directly grasped. Observation from without is in the nature of the case impotent to apprehend the active qua active. We can from without observe sequences of states. If into these we read activity (as we sometimes do) this can only be on the basis of what we discern in ourselves from the inner standpoint. It follows that if anyone insists upon taking his criterion of the meaningful simply from the standpoint of external observation he is really deciding in advance of the evidence that the notion of activity and a fortiori the notion of a free will is ‘meaningless’. He looks for the free act through a medium which is in the nature of the case incapable of revealing it and then because inevitably he doesn't find it he declares that it doesn't exist!
But if as we surely ought in this context we adopt the inner standpoint then (I am suggesting) things appear in a totally different light. From the inner standpoint it seems to me plain there is no difficulty whatever in attaching meaning to an act which is the self's act and which nevertheless does not follow from the self's character. So much I claim has been established by the phenomenological analysis in this and the previous lecture of the act of moral decision in face of moral temptation. It is thrown into particularly clear relief where the moral decision is to make the moral effort required to rise to duty. For the very function of moral effort as it appears to the agent engaged in the act is to enable the self to act against the line of least resistance against the line to which his character as so far formed most strongly inclines him. But if the self is thus conscious here of combating his formed character he surely cannot possibly suppose that the act although his own act issues from his formed character? I submit therefore that the self knows very well indeed—from the inner standpoint—what is meant by an act which is the self's act and which nevertheless does not follow from the self's character.
What this implies—and it seems to me to be an implication of cardinal importance for any theory of the self that aims at being more than superficial—is that the nature of the self is for itself something more than just its character as so far formed. The ‘nature’ of the self and what we commonly call the ‘character’ of the self are by no means the same thing and it is utterly vital that they should not be confused. The ‘nature’ of the self comprehends but is not without remainder reducible to its ‘character’; it must if we are to be true to the testimony of our experience of it be taken as including also the authentic creative power of fashioning and re-fashioning ‘character’.
The misguided and as a rule quite uncritical belittlement of the evidence offered by inner experience has I am convinced been responsible for more bad argument by the opponents of Free Will than has any other single factor. How often for example do we find the Determinist critic saying in effect ‘Either the act follows necessarily upon precedent states or it is a mere matter of chance and accordingly of no moral significance’. The disjunction is invalid for it does not exhaust the possible alternatives. It seems to the critic to do so only because he will limit himself to the standpoint which is proper and indeed alone possible in dealing with the physical world the standpoint of the external observer. If only he would allow himself to assume the standpoint which is not merely proper for but necessary to the apprehension of subjective activity the inner standpoint of the practical consciousness in its actual functioning he would find himself obliged to recognise the falsity of his disjunction. Reflection upon the act of moral decision as apprehended from the inner standpoint would force him to recognise a third possibility as remote from chance as from necessity that namely of creative activity in which (as I have ventured to express it) nothing determines the act save the agent's doing of it.
11. There we must leave the matter. But as this lecture has been I know somewhat densely packed it may be helpful if I conclude by reminding you in bald summary of the main things I have been trying to say. Let me set them out in so many successive theses.
1. The freedom which is at issue in the traditional Free Will problem is the freedom which is presupposed in moral responsibility.
2. Critical reflection upon carefully considered attributions of moral responsibility reveals that the only freedom that will do is a freedom which pertains to inner acts of choice and that these acts must be acts (a) of which the self is sole author and (b) which the self could have performed otherwise.
3. From phenomenological analysis of the situation of moral temptation we find that the self as engaged in this situation is inescapably convinced that it possesses a freedom of precisely the specified kind located in the decision to exert or withhold the moral effort needed to rise to duty where the pressure of its desiring nature is felt to urge it in a contrary direction.
Passing to the question of the reality of this moral freedom which the moral agent believes himself to possess we argued:
4. Of the two types of Determinist criticism which seem to have most influence today that based on the predictability of much human behaviour fails to touch a Libertarianism which confines the area of free will as above indicated. Libertarianism so understood is compatible with all the predictability that the empirical facts warrant. And:
5. The second main type of criticism which alleges the ‘meaninglessness’ of an act which is the self's act and which is yet not determined by the self's character is based on a failure to appreciate that the standpoint of inner experience is not only legitimate but indispensable where what is at issue is the reality and nature of a subjective activity. The creative act of moral decision is inevitably meaningless to the mere external observer; but from the inner standpoint it is as real and as significant as anything in human experience.5
- 1. Reprinted in Ethics and the History of Philosophy, Selected Essays.
- 2. I have explained the grounds for my dissent from Broad's final conclusions on pp. 27 ff. of In Defence of Free Will (Jackson Son & Co., 1938).
- 3. Kantian Ethics.
- 4. Op. cit. p. 261.
- 5. An earlier, but not in substance dissimilar, version of my views on the Free Will problem has been criticised at length in Mr. Nowell-Smith's Ethics. A detailed reply to these criticisms will be found in Appendix B.