In the days when the Verifiability Principle was accepted by its devotees as a secure philosophical truth, one could understand, though one might not agree with, the sweeping claim that many of the traditional problems of philosophy had been shown to be mere ‘pseudo-problems’. It was easy to see how, given the Principle’s validity, most of the leading questions which agitated our forefathers in metaphysics, in ethics, and in theology, automatically become nonsensical questions. What is perplexing, however, is that despite the pretty generally acknowledged deterioration in the Principle’s status to that of a convenient methodological postulate, the attitude to these same questions seems to have changed but little. To admit that the Verifiability Principle is not an assured truth entails the admission that a problem can no longer be dismissed as meaningless simply on the ground that it cannot be stated in a way which satisfies the Principle. Whether or not a problem is meaningless is now something that can only be decided after critical examination of the particular case on its own individual merits. But the old antipathies seem in large measure to have survived the disappearance of their logical basis. One gets the impression that for at least many thinkers with Positivist sympathies the ‘liquidation’ of a large, if unspecified, group of traditional philosophic problems is still established fact. If that impression is mistaken, well and good. One may then hope for an early recrudescence of interest in certain problems that have too long suffered the consequences of an unhappy tabu. If the impression is correct, a real service would be done to philosophy if it were plainly stated which of the traditional problems are still regarded as pseudo-problems, and what are the reasons, old or new, for passing this sentence upon them. The smoke of old battles, perhaps understandably, darkens the philosophic air, to the considerable inconvenience of all concerned.
Fortunately, however, the obscurity complained of is not totally unrelieved. We do know of one traditional problem that is definitely on the black list of the avant garde — the problem of ‘Free Will’: and we do have pretty adequate information about the reasons which have led to its being placed theron. This, so far as it goes, is satisfactory. A plain obligation now lies upon philosophers who still believe that ‘Free Will’ is a genuine problem to explain just where, in their opinion, the case for the prosecution breaks down. To discharge this obligation is the main purpose of the present paper.
There will be a clear advantage in making our start from the locus classicus of the ‘pseudo-problem’ theory, if locus classicus there be. And I think that there must be something of the sort. At any rate, the casual, and indeed slightly bored, tones in which so many contemporary philosophers allude to the traditional problem, and their contentment to indicate in only a sketchy manner the reasons why it no longer exists, strongly suggest that somewhere the matter has in their eyes been already effectively settled. At least one important ‘document in the case’ is, I suspect, Chapter VII of Moritz Schlick’s Problems of Ethics, first published in 1931. This chapter, the title of which is ‘When is a Man Responsible?’ and the first section of which bears the heading ‘The Pseudo-problem of Freedom of the Will’, presents in concentrated form, but with some show of systematic argument, most of the considerations upon which later writers appear to rely. It will be worth our while, therefore, to try to see just why Professor Schlick is so sure (and he is very sure indeed) that ‘Free Will’, as traditionally formulated, is a pseudo-problem, begotten by mere confusion of mind.
I shall first summarize, as faithfully as I can, what I take to be the distinctive points in Schlick’s argument.
The traditional formulation of the problem, Schlick points out, is based on the assumption that to have ‘free will’ entails having a will that is, at least sometimes, exempt from causal law. It is traditionally supposed, quite rightly, that moral responsibility implies freedom in some sense: and it is supposed, also quite rightly, that this sense is one which is incompatible with compulsion. But because it is further supposed, quite wrongly, that to be subject to causal or natural law is to be subject to compulsion, the inference is drawn that the free will implied in moral responsibility is incompatible with causal continuity. The ultimate root of the error, Schlick contends, lies in a failure to distinguish between two different kinds of Law, one of which does indeed ‘compel’, but the other of which does not. (Problems of Ethics, Ch. VIII. Section 2.) There are, first, prescriptive laws, such as the laws imposed by civil authority, which presume contrary desires on the part of those to whom they are applied; and these may fairly be said to exercise ‘compulsion’. And there are, secondly, descriptive laws, such as the laws which the sciences seek to formulate; and these merely state what does as a matter of fact always happen. It is perfectly clear that the relation of the latter, the natural, causal laws, to human willing is radically different from the ‘compulsive’ relation of prescriptive laws to human willing, and that it is really an absurdity to talk of a species of natural law like, say, psychological laws, compelling us to act in this or that way. The term ‘compulsion’ is totally inept where, as in this case, there are no contrary desires. But the traditional discussions of Free Will, confusing descriptive with prescriptive laws, fallaciously assume ‘compulsion’ to be ingredient in Law as such, and it is contended accordingly that moral freedom, since it certainly implies absence of compulsion, implies also exemption from causal law.
It follows that the problem of Free Will, as traditionally stated, is a mere pseudo-problem. The statement of it in terms of exemption from causal law rests on the assumption that causal law involves ‘compulsion’. And this assumption is demonstrably false. Expose the muddle from which it arises and the so-called `problem’ in its traditional form disappears.
But is it quite certain that the freedom which moral responsibility implies is no more than `the absence of compulsion’? This is the premise upon which Schlick’s argument proceeds, but Schlick is himself well aware that it stands in need of confirmation from an analysis of the notion of moral responsibility. Otherwise it might be maintained that although ‘the absence of compulsion’ has been shown not to entail a contra-causal type of freedom, there is nevertheless some other condition of moral responsibility that does entail it. Accordingly Schlick embarks now upon a formal analysis of the nature and conditions of moral responsibility designed to show that the only freedom implied by moral responsibility is freedom from compulsion. It was a trifle ambitious, however, even for a master of compression like Professor Schlick to hope to deal satisfactorily in half a dozen very brief pages with a topic which has been so extensively debated in the literature of moral philosophy: and I cannot pretend that I find what he has to say free from obscurity. But to the best of my belief what follows does reproduce the gist of Schlick’s analysis.
What precisely, Schlick asks, does the term ‘moral responsibility’ mean in our ordinary linguistic usage?’ He begins his answer by insisting upon the close connection for ordinary usage between ‘moral responsibility’ and punishment (strictly speaking, punishment and reward: but for convenience Schlick virtually confines the discussion to punishment, and we shall do the same). The connection, as Schlick sees it, is this. In ordinary practice our concern with the responsibility for an act (he tells us) is with a view to determining who is to be punished for it. Now punishment is (I quote) ‘an educative measure’. It is ‘a means to the formation of motives, which are in part to prevent the wrong-doer from repeating the act (reformation), and in part to prevent others from committing a similar act (intimidation)’. When we ask, then, ‘Who in a given case is to be punished?’ — which is the same as the question ‘Who is responsible?’ — what we are really wanting to discover is some agent in the situation upon whose motives we can bring to bear the appropriate educative influences, so that in similar situations in future his strongest motive will impel him to refrain from, rather than to repeat, the act. ‘The question of who is responsible’ Schlick sums up, ‘is … a matter only of knowing who is to be punished or rewarded, in order that punishment and reward function as such — be able to achieve their goal’. It is not a matter, he expressly declares, of trying to ascertain what may be called the ‘original instigator’ of the act. That might be a great-grand-parent, from the consequence of whose behaviour vicious tendencies have been inherited by a living person. Such ‘remote causes’ as this are irrelevant to questions of punishment (and so to questions of moral responsibility), ‘for in the first place their actual contribution cannot be determined, and in the second place they are generally out of reach’ .
It is a matter for regret that Schlick has not rounded off his discussion, as one had hoped and expected he would, by formulating a precise definition of moral responsibility in terms of what he has been saying. I think, however, that the conclusion to which his argument leads could be not unfairly expressed in some such way as this: ‘We say that a man is morally responsible for an act if his motives for bringing about the act are such as we can affect favourably in respect of his future behaviour by the educative influences of reward and punishment’.
Given the truth of this analysis of moral responsibility, Schlick’s contention follows logically enough that the only freedom that is required for moral responsibility is freedom from compulsion. For what are the cases in which a man’s motives are not capable of being favourably affected by reward and punishment? — the cases in which, that is, according to Schlick’s analysis, we do notdeem him morally responsible? The only such cases, it would seem, are those in which a man is subjected to some form of external constraint which prevents him from acting according to his ‘natural desires’. For example, if a man is compelled by a pistol at his breast to do a certain act, or induced to do it by an externally administered narcotic, he is not ‘morally responsible’; or not, at any rate, in so far as punishment would be impotent to affect his motives in respect of his future behaviour. External constraint in one form or another seems to be the sole circumstance which absolves a man from moral responsibility. Hence we may say that freedom from external constraint is the only sort of freedom which an agent must possess in order to be morally responsible. The ‘contra-causal’ sort of freedom which so many philosophers and others have supposed to be required is shown by a true analysis of moral responsibility to be irrelevant.
This completes the argument that ‘Free Will’, as traditionally formulated, is a pseudo-problem. The only freedom implied by moral responsibility is freedom from compulsion; and as we have rid ourselves of the myth that subjection to causal law is a form of compulsion, we can see that the only compulsion which absolves from moral responsibility is the external constraint which prevents us from translating our desires into action. The true meaning of the question ‘Have we free will?’ thus becomes simply ‘Can we translate our desires into action? And this question does not constitute a ‘problem’ at all, for the answer to it is not in doubt. The obvious answer is ‘Sometimes we can, sometimes we can’t, according to the specific circumstances of the case’.
Here, then, in substance is Schlick’s theory. Let us now examine it.
In the first place, it is surely quite unplausible to suggest that the common assumption that moral freedom postulates some breach of causal continuity arises from a confusion of two different types of law. Schlick’s distinction between descriptive and prescriptive law is, of course, sound. It was no doubt worth pointing out too, that descriptive laws cannot be said to ‘compel’ human behaviour in the same way as prescriptive laws do. But it seems to me evident that the usual reason why it is held that moral freedom implies some breach of causal continuity, is not a belief that causal laws ‘compel’ as civil laws ‘compel’, but simply the belief that the admission of unbroken causal continuity entails a further admission which is directly incompatible with moral responsibility; viz. the admission that no man could have acted otherwise than he in fact did.Could one ever have acted otherwise?
Now it may, of course, be an error thus to assume that a man is not morally responsible for an act, a fit subject for moral praise and blame in respect of it, unless he could have acted otherwise than he did. Or, if this is not an error, it may still be an error to assume that a man could not have acted otherwise than he did, in the sense of the phrase that is crucial for moral responsibility, without there occurring some breach of causal continuity. Into these matters we shall have to enter very fully at a later stage. But the relevant point at the moment is that these (not prima facie absurd) assumptions about the conditions of moral responsibility have very commonly, indeed normally, been made, and that they are entirely adequate to explain why the problem of Free Will finds its usual formulation in terms of partial exemption from causal law. Schlick’s distinction between prescriptive and descriptive laws has no bearing at all upon the truth or falsity of these assumptions. Yet if these assumptions are accepted, it is (I suggest) really inevitable that the Free Will problem should be formulated in the way to which Schlick takes exception. Recognition of the distinction upon which Schlick and his followers lay so much stress can make not a jot of difference.
As we have seen, however, Schlick does later proceed to the much more important business of disputing these common assumptions about the conditions of moral responsibility. He offers us an analysis of moral responsibility which flatly contradicts these assumptions; an analysis according to which the only freedom demanded by morality is a freedom which is compatible with Determinism. If this analysis can be sustained, there is certainly no problem of ‘Free Will’ in the traditional sense.
But it seems a simple matter to show that Schlick’s analysis is untenable. Let us test it by Schlick’s own claim that it gives us what we mean by ‘moral responsibility’ in ordinary linguistic usage.
We do not ordinarily consider the lower animals to be morally responsible. But ought we not to do so if Schlick is right about what we mean by moral responsibility? It is quite possible, by punishing the dog who absconds with the succulent chops designed for its master’s luncheon, favourably to influence its motives in respect of its future behaviour in like circumstances. If moral responsibility is to be linked with punishment as Schlick links it, and punishment conceived as a form of education, we should surely hold the dog morally responsible? The plain fact, of course, is that we don’t. We don’t, because we suppose that the dog ‘couldn’t help it’: that its action (unlike what we usually believe to be true of human beings) was simply a link in a continuous chain of causes and effects. In other words, we do commonly demand the contra-causal sort of freedom as a condition of moral responsibility.
Again, we do ordinarily consider it proper, in certain circumstances, to speak of a person no longer living as morally responsible for some present situation. But ought we to do so if we accept Schlick’s essentially ‘forward-looking’ interpretation of punishment and responsibility? Clearly we cannot now favourably affect the dead man’s motives. No doubt they could at one time have been favourably affected. But that cannot be relevant to our judgment of responsibility if, as Schlick insists, the question of who is responsible ‘is a matter only of knowing who is to be punished or rewarded’. Indeed he expressly tells us, as we saw earlier, that in asking this question we are not concerned with a `great-grand-parent’ who may have been the ‘original instigator’, because, for one reason, this ‘remote cause’ is ‘out of reach’. We cannot bring the appropriate educative influence to bear upon it. But the plain fact, of course, is that we do frequently assign moral responsibility for present situations to persons who have long been inaccessible to any punitive action on our part. And Schlick’s position is still more paradoxical in respect of our apportionment of responsibility for occurrences in the distant past. Since in these cases there is no agent whatsoever whom we can favourably influence by punishment, the question of moral responsibility here should have no meaning for us. But of course it has. Historical writings are studded with examples.
Possibly the criticism just made may seem to some to result from taking Schlick’s analysis too much au pied de la lettre. The absurd consequences deduced, it may be said, would not follow if we interpreted Schlick as meaning that a man is morally responsible where his motive is such as can in principle be favourably affected by reward or punishment — whether or not we who pass the judgment are in a position to take such action. But with every desire to be fair to Schlick, I cannot see how he could accept this modification and still retain the essence of his theory. For the essence of his theory seems to be that moral responsibility has its whole meaning and importance for us in relation to our potential control of future conduct in the interests of society. (I agree that it is hard to believe that anybody really thinks this. But it is perhaps less hard to believe to-day than it has ever been before in the history of modern ethics.)
Again, we ordinarily consider that, in certain circumstances, the degree of a man’s moral responsibility for an act is affected by considerations of his inherited nature, or of his environment, or of both. It is our normal habit to ‘make allowances’ (as we say) when we have reason to believe that a malefactor had a vicious heredity, or was nurtured in his formative years in a harmful environment. We say in such cases ‘Poor chap, he is more to be pitied than blamed. We could scarcely expect him to behave like a decent citizen with his parentage or upbringing.’ But this extremely common sort of judgment has no point at all if we mean by moral responsibility what Schlick says that we mean. On that meaning the degree of a man’s moral responsibility must presumably be dependent upon the degree to which we can favourably affect his future motives, which is quite another matter. Now there is no reason to believe that the motives of a man with a bad heredity or a bad upbringing are either less or more subject to educative influence than those of his more fortunate fellows. Yet it is plain matter of fact that we do commonly consider the degree of a man’s moral responsibility to be affected by these two factors.
A final point. The extremity of paradox in Schlick’s identification of the question ‘Who is morally blameworthy?’ with the question ‘Who is to be punished?’ is apt to be partially concealed from us just because it is our normal habit to include in the meaning of ‘punishment’ an element of ‘requital for moral transgression’ which Schlick expressly denies to it. On that account we commonly think of ‘punishment’, in its strict sense, as implying moral blameworthiness in the person punished. But if we remember to mean by punishment what Schlick means by it, a purely ‘educative measure’, with no retributive ingredients, his identification of the two questions loses such plausibility as it might otherwise have. For clearly we often think it proper to `punish’ a person, in Schlick’s sense, where we are not at all prepared to say that the person is morally blameworthy. We may even think him morally commendable. A case in point would be the unmistakably sincere but muddleheaded person who at the cost of great suffering to himself steadfastly pursues as his ‘duty’ a course which, in our judgment is fraught with danger to the common weal. We should most of us feel entitled, in the public interest, to bring such action to bear upon the man’s motives as might induce him to refrain in future from his socially injurious behaviour: in other words, to inflict upon him what Schlick would call ‘punishment’. But we should most of us feel perfectly clear that in so `punishing’ this misguided citizen we are not proclaiming his moral blameworthiness for moral wickedness.
Adopting Schlick’s own criterion, then, looking simply ‘to the manner in which the concept is used’, we seem bound to admit that constantly people do assign moral responsibility where Schlick’s theory says they shouldn’t, don’t assign moral responsibility where Schlick’s theory says they should, and assign degrees of moral responsibility where on Schlick’s theory there should be no difference in degree. I think we may reasonably conclude that Schlick’s account of what we mean by moral responsibility breaks down.
The rebuttal of Schlick’s arguments, however, will not suffice of itself to refute the pseudo-problem theory. The indebtedness to Schlick of most later advocates of the theory may be conceded; but certainly it does not comprehend all of significance that they have to say on the problem. There are recent analyses of the conditions of moral responsibility containing sufficient new matter, or sufficient old matter in a more precise and telling form, to require of us now something of a fresh start. In the section which follows I propose to consider some representative samples of these analyses — all of which, of course, are designed to show that the freedom which moral responsibility implies is not in fact a contra-causal type of freedom.
But before reopening the general question of the nature and conditions of moral responsibility there is a caveat which it seems to me worth while to enter. The difficulties in the way of a clear answer are not slight; but they are apt to seem a good deal more formidable than they really are because of a common tendency to consider in unduly close association two distinct questions: the question ‘Is a contra-causal type of freedom implied by moral responsibility?’ and the question ‘Does a contra-causal type of freedom anywhere exist?’ It seems to me that many philosophers (and I suspect that Moritz Schlick is among them) begin their enquiry with so firm a conviction that the contra-causal sort of freedom nowhere exists, that they find it hard to take very seriously the possibility that it is this sort of freedom that moral responsibility implies. For they are loth to abandon the commonsense belief that moral responsibility itself is something real. The implicit reasoning I take to be this. Moral responsibility is real. If moral responsibility is real, the freedom implied in it must be a fact. But contra-causal freedom is not a fact. Therefore contra-causal freedom is not the freedom implied in moral responsibility. I think we should be on our guard against allowing this or some similar train of reasoning (whose premises, after all, are far from indubitable) to seduce us into distorting what we actually find when we set about a direct analysis of moral responsibility and its conditions.
The pseudo-problem theorists usually, and naturally, develop their analysis of moral responsibility by way of contrast with a view which, while it has enjoyed a good deal of philosophic support, I can perhaps best describe as the common view. It will be well to remind ourselves, therefore, of the main features of this view.
So far as the meaning, as distinct from the conditions, of moral responsibility is concerned, the common view is very simple. If we ask ourselves whether a certain person is morally responsible for a given act (or it may be just ‘in general’), what we are considering, it would be said, is whether or not that person is a fit subject upon whom to pass moral judgment; whether he can fittingly be deemed morally good or bad, morally praiseworthy or blameworthy. This does not take us any great way: but (pace Schlick) so far as it goes it does not seem to me seriously disputable. The really interesting and controversial question is about the conditions of moral responsibility, and in particular the question whether freedom of a contra-causal kind is among these conditions.
The answer of the common man to the latter question is that it most certainly is among the conditions. Why does he feel so sure about this? Not, I argued earlier, because the common man supposes that causal law exercises ‘compulsion’ in the sense that prescriptive laws do, but simply because he does not see how a person can be deemed morally praiseworthy or blameworthy in respect of an act which he could not help performing. From the stand-point of moral praise and blame, he would say — though not necessarily from other stand-points — it is a matter of indifference whether it is by reason of some external constraint or by reason of his own given nature that the man could not help doing what he did. It is quite enough to make moral praise and blame futile that in either case there were no genuine alternatives, no open possibilities, before the man when he acted. He could not have acted otherwise than he did. And the common man might not unreasonably go on to stress the fact that we all, even if we are linguistic philosophers, do in our actual practice of moral judgment appear to accept the common view. He might insist upon the point alluded to earlier in this paper, that we do all, in passing moral censure, ‘make allowances’ for influences in a man’s hereditary nature or environmental circumstances which we regard as having made, it more than ordinarily difficult for him to act otherwise than he did: the implication being that if we supposed that the man’s heredity and environment made it not merely very difficult but actually impossible for him to act otherwise than he did, we could not properly assign moral blame to him at all.
Let us put the argument implicit in the common view a little more sharply. The moral ‘ought’ implies ‘can’. If we say that A morally ought to have done X, we imply that in our opinion, he could have done X. But we assign moral blame to a man only for failing to do what we think he morally ought to have done. Hence if we morally blame A for not having done X, we imply that he could have done X even though in fact he did not. In other words, we imply that A could have acted otherwise than he did. And that means that we imply, as a necessary condition of a man’s being morally blameworthy, that he enjoyed a freedom of a kind not compatible with unbroken causal continuity.
Now what is it that is supposed to be wrong with this simple piece of argument? – For, of course, it must be rejected by all these philosophers who tell us that the traditional problem of Free Will is a mere pseudo-problem. The argument looks as though it were doing little more than reading off necessary implications of the fundamental categories of our moral thinking. One’s inclination is to ask ‘If one is to think morally at all, how else than this can we think?’.
In point of fact, there is pretty general agreement among the contemporary critics as to what is wrong with the argument. Their answer in general terms is as follows. No doubt A’s moral responsibility does imply that he could have acted otherwise. But this expression ‘could have acted otherwise’ stands in dire need of analysis. When we analyse it, we find that it is not, as is so often supposed, simple and unambiguous, and we find that in some at least of its possible meanings it implies no breach of causal continuity between character and conduct. Having got this clear, we can further discern that only in one of these latter meanings is there any compulsion upon our moral thinking to assert that if A is morally blameworthy for an act, A ‘could have acted otherwise than he did’. It follows that, contrary to common belief, our moral thinking does not require us to posit a contra-causal freedom as a condition of moral responsibility.
So much of importance obviously turns upon the validity or otherwise of this line of criticism that we must examine it in some detail and with express regard to the ipsissima verba of the critics.
In the course of a recent article in Mind, entitled ‘Free Will and Moral Responsibility’, Mr Nowell Smith (having earlier affirmed his belief that `the traditional problem has been solved’) explains very concisely the nature of the confusion which, as he thinks, has led to the demand for a contra-causal freedom. He begins by frankly recognizing that ‘It is evident that one of the necessary conditions of moral action is that the agent “could have acted otherwise” ‘ and he adds ‘it is to this fact that the Libertarian is drawing attention’. Then, after showing (unexceptionably, I think) how the relationship of ‘ought’ to ‘can’ warrants the proposition which he has accepted as evident, and how it induces the Libertarian to assert the existence of action that is ‘uncaused’, he proceeds to point out, in a crucial passage, the nature of the Libertarian’s error:
`The fallacy in the argument (he contends) lies in supposing that when we say “A could have acted otherwise” we mean that A, being what he was and being placed in the circumstances in which he was placed, could have done something other than what he did. But in fact we never do mean this.’
What then do we mean here by ‘A could have acted otherwise’? Mr Nowell Smith does not tell us in so many words, but the passage I have quoted leaves little doubt how he would answer. What we really mean by the expression, he implies, is not a categorical but a hypothetical proposition. We mean ‘A could have acted otherwise, if he did not happen to be what he in fact was, or if he were placed in circumstances other than those in which he was in fact placed’. Now, thesepropositions, it is easy to see, are in no way incompatible with acceptance of the causal principle in its full rigour. Accordingly the claim that our fundamental moral thinking obliges us to assert a contra-causal freedom as a condition of moral responsibility is disproved.
Such is the ‘analytical solution’ of our problem offered (with obvious confidence) by one able philosopher of to-day, and entirely representative of the views of many other able philosophers. Yet I make bold to say that its falsity stares one in the face. It seems perfectly plain that the hypothetical propositions which Mr Nowell Smith proposes to substitute for the categorical proposition cannot express ‘what we really mean’ in this context by ‘A could have acted otherwise’, for the simple reason that these hypothetical propositions have no bearing whatsoever upon the question of the moral responsibility of A. And it is A whose moral responsibility we are talking about — a definite person A with a definitive character and in a definitive set of circumstances. What conceivable significance could it have for our attitude to A’s responsibility to know that someone with a different character (or A with a different character, if that collocation of words has any meaning), or A in a different set of circumstances from those in which A as we are concerned with him was in fact placed, ‘could have acted otherwise’? No doubt this supposititious being could have acted otherwise than the definitive person A acted. But the point is that where we are reflecting, as we are supposed in this context to be reflecting, upon the question of A’s moral responsibility, our interest in this supposititious being is precisely nil.
The two hypothetical propositions suggested in Mr Nowell Smith’s account of the matter do not, however, exhaust the speculations that have been made along these lines. Another very common suggestion by the analysts is that what we really mean by ‘A could have acted otherwise’ is ‘A could have acted otherwise if he had willed, or chosen, otherwise‘. This was among the suggestions offered by G. E. Moore in the well-known chapter on Free Will in his Ethics. It is, I think, the suggestion he most strongly favoured: though it is fair to add that neither about this nor about any other of his suggestions is Moore in the least dogmatic. He does claim, for, I think, convincing reasons, that ‘we very often mean by “could” merely “would, if so-and-so had chosen” ‘. And he concludes ‘I must confess that I cannot feel certain that this may not be all that we usually mean and understand by the assertion that we have Free Will’.
This third hypothetical proposition appears to enjoy also the support of Mr C. L. Stevenson. Mr Stevenson begins the chapter of Ethics and Language entitled ‘Avoidability-Indeterminism’ with the now familiar pronouncement of his School that ‘controversy about freedom and determinism of the will … presents no permanent difficulty to ethics, being largely a product of confusions’. A major confusion (if I understand him rightly) he takes to lie in the meaning of the term ‘avoidable’, when we say ‘A’s action was avoidable’ — or, I presume, ‘A could have acted otherwise’. He himself offers the following definition of ‘avoidable’ – ‘”A’s action was avoidable” has the meaning of “If A had made a certain choice, which in fact he did not make, his action would not have occurred”‘. This I think we may regard as in substance identical with the suggestion that what we really mean by ‘A could have acted otherwise’ is ‘A could have acted otherwise if he had chosen (or willed) otherwise’. For clarity’s sake we shall here keep to this earlier formulation. In either formulation the special significance of the third hypothetical proposition, as of the two hypothetical propositions already considered, is that it is compatible with strict determinism. If this be indeed all that we mean by the ‘freedom’ that conditions moral responsibility, then those philosophers are certainly wrong who hold that moral freedom is of the contra-causal type.
Now this third hypothetical proposition does at least possess the merit, not shared by its predecessors, of having a real relevance to, the question of moral responsibility. If, e.g., A had promised to meet us at 2 p.m., and he chanced to break his leg at i p.m., we should not blame him for his failure to discharge his promise. For we should be satisfied that he could not have acted otherwise, even if he had so chosen; or could not, at any rate, in a way which would have enabled him to meet us at 2 p.m. The freedom to translate one’s choice into action, which we saw earlier is for Schlick the only freedom required for moral responsibility, is without doubt oneof the conditions of moral responsibility.
But it seems easy to show that this third hypothetical proposition does not exhaust what we mean, and sometimes is not even part of what we mean, by the expression ‘could have acted otherwise’ in its moral context. Thus it can hardly be even part of what we mean in the case of that class of wrong actions (and it is a large class) concerning which there is really no question whether the agent could have acted otherwise, if he had chosen otherwise. Take lying, for example. Only in some very abnormal situation could it occur to one to doubt whether A, whose power of speech was evinced by his telling a lie, was in a position to tell what he took to be the truth if he had so chosen. Of course he was. Yet it still makes good sense for one’s moral thinking to ask whether A, when lying, ‘could have acted otherwise’ : and we still require an affirmative answer to this question if A’s moral blameworthiness is to be established. It seems apparent, therefore, that in this class of cases at any rate one does not mean by ‘A could have acted otherwise’, ‘A could have acted otherwise if he had so chosen’.
What then does one mean in this class of cases by ‘A could have acted otherwise’? I submit that the expression is taken in its simple, categorical meaning, without any suppressed `if’ clause to qualify it. Or perhaps, in order to keep before us the important truth that it is only as expressions of will or choice that acts are of moral import, it might be better to say that a condition of A’s moral responsibility is that he could have chosen otherwise. We saw that there is no real question whether A who told a lie could have acted otherwise if he had chosen otherwise. But there is a very real question, at least for any person who approaches the question of moral responsibility at a tolerably advanced level of reflection, about whether A could have chosen otherwise. Such a person will doubtless be acquainted with the claims advanced in some quarters that causal law operates universally: or/and with the theories of some philosophies that the universe is throughout the expression of a single supreme principle; or/and with the doctrines of some theologians that the world is created, sustained and governed by an Omniscient and Omnipotent Being. Very understandably such world-views awaken in him doubts about the validity of his first, easy, instinctive assumption that there are genuinely open possibilities before a man at the moment of moral choice. It thus becomes for him a real question whether a man could have chosen otherwise than he actually did, and, in consequence, whether man’s moral responsibility is really defensible. For how can a man be morally responsible, he asks himself, if his choices, like all other events in the universe, could not have been otherwise than they in fact were? It is precisely against the background of world-views such as these that for reflective people the problem of moral responsibility normally arises.
Furthermore, to the man who has attained this level of reflection, it will in no class of cases be a sufficient condition of moral responsibility for an act that one could have acted otherwise if one had chosen otherwise — not even in these cases where there was some possibility of the operation of ‘external constraint’. In these cases he will, indeed expressly recognize freedom from external constraint as a necessary condition, but not as a sufficient condition. For he will be aware that, even granted this freedom, it is still conceivable that the agent had no freedom to choose otherwise than he did, and he will therefore require that the latter sort of freedom be added if moral responsibility for the act is to be established.
I have been contending that, for persons at a tolerably advanced level of reflection, ‘A could have acted otherwise’, as a condition of A’s moral responsibility, means ‘A could have chosen otherwise’. The qualification italicized is of some importance. The unreflective or unsophisticated person, the ordinary ‘man in the street’, who does not know or much care what scientists and theologians and philosophers have said about the world, sees well enough that A is morally responsible only if he could have acted otherwise, but in his intellectual innocence he will, very probably, envisage nothing capable of preventing A from having acted otherwise except some material impediment — like the broken leg in the example above. Accordingly, for the unreflective person, ‘A could have acted otherwise’, as a condition of moral responsibility, is apt to mean no more than ‘A could have acted otherwise if he had so chosen’.
It would appear, then, that the view now favoured by many philosophers, that the freedom required for moral responsibility is merely freedom from external constraint, is a view which they share only with the less reflective type of layman. Yet it should be plain that on a matter of this sort the view of the unreflective person is of little value by comparison with the view of the reflective person. There are some contexts, no doubt, in which lack of sophistication is an asset. But this is not one of them. The question at issue here is as to the kind of impediments which might have prevented a man from acting otherwise than he in fact did: and on this question knowledge and reflection are surely prerequisites of any answer that is worth listening to. It is simply on account of the limitations of his mental vision – that the unreflective man interprets the expression ‘could have acted otherwise’, in its context as a condition of moral responsibility, solely in terms of external constraint. He has failed (as yet) to reach the intellectual level at which one takes into account the implications for moral choices of the world-views of science, religion, and philosophy. If on a matter of this complexity the philosopher finds that his analysis accords with the utterances of the uneducated he has, I suggest, better cause for uneasiness than for self-congratulation.
This concludes the main part of what it seems to me necessary to say in answer to the pseudo-problem theorists. My object so far has been to expose the falsity of those innovations (chiefly Positivist) in the way of argument and analysis which are supposed by many to have made it impossible any longer to formulate the problem of Free Will in the traditional manner. My contention is that, at least so far as these innovations are concerned, the simple time-honoured argument still holds from the nature of the moral ought to the conclusion that moral responsibility implies a contra-causal type of freedom. The attempts to avoid that conclusion by analysing the proposition ‘A could have acted otherwise’ (acknowledged to be implied in some sense in A’s moral responsibility) into one or other of certain hypothetical propositions which are compatible with unbroken causal continuity, break down hopelessly when tested against the touchstone of actual moral thinking. It is, I think, not necessary to defend the procedure of testing hypotheses in the ethical field by bringing to bear upon them our actual moral thinking. If there is any other form of test applicable, I should be much interested to learn what it is supposed to be. Certainly ‘logical analysis’ per se will not do. That has a function, but a function that can only be ancillary. For what we are seeking to know is the meaning of the expression ‘could have acted otherwise’ not in the abstract, but in the context of the question of man’s moral responsibility. Logical analysis per se is impotent to give us this information. It can be of value only in so far as it operates within the orbit of ‘the moral consciousness’. One may admit, with some qualifications, that on a matter of this sort the moral consciousness without logical analysis is blind: but it seems to me to be true without any qualification whatsoever that, on the same problem, logical analysis without the moral consciousness is empty.