A. The Relation of Freedom to Ethics

IT is commonly said that moral conduct, though it must conform to an objective principle, must also be ‘free obedience’. The agent must will the act for no other reason than that he believes it to be, in the circumstances, the right act. His will must have no determinant save the goodness of the moral principle on which he is acting. The act must be his act, and not the act of something else ‘compelling’ him. In some important sense it must be true that he can do either the right or the wrong act; and that, in choosing the right, he accepts a principle which he could have rejected.

This assertion, that the moral agent, is a free agent, does not deny that the behaviour of the good man is predictable theoretically in every detail, and to a large extent ‘even in practice. For it is essential to morality that there should be an objective and universal principle or system of principles according to which all right. acts, in whatever circumstances, must be regulated. The right act, there- fore, in any situation, can theoretically be described beforehand; and this is the act that the good man will choose. But though the behaviour of the good man is thus theoretically predictable in every detail, it is not to be called moral conduct unless it is free behaviour, unless his acceptance of the moral principle springs from nothing but his own ‘good will’. in fact, though his conduct is strictly determined in relation to the demands of the objective situation, it must be actually effected (if it is to be moral conduct) by nothing other than the free moral agent, who accepts the moral principle in general, and therefore freely chooses a determinate course through diverse circumstances.

That all the acts of the good man are theoretically predictable would, no doubt, be denied with horror by many whose opinion deserves very great respect. It would be admitted that many of his acts are predictable; but also it would be insisted that only those are predictable which consist in the fulfilling of definite obligations, which are, in fact, acts of simple justice. Those acts of supererogation which go beyond mere justice, which do more than fulfil the rights of others, which express positively and subtly the personality of the agent, would be denied to be predictable. Great works of art, for instance, and the morally creative lives of spiritual persons, would be said to be the spontaneous expression of emergent characters in human personality. To claim that all this is predictable, it might be said, is to fail to take emergence seriously. For the essence of emergence is that the behaviour of the whole is ‘the unique expression of the nature of the whole, and therefore is not predictable from any knowledge but knowledge of the behaviour of the whole itself. And even so, knowledge of the whole’s past behaviour is no adequate guide to its future behaviour, since novel situations may arise which call forth novel emergent behaviour. Thus, though in theory at least it is possible to predict what the good man will not do, in that will not be unjust, or mean, or cowardly, and so on, all those acts which more positively express his unique personality are unpredictable. Even he himself cannot predict them. They have to be done before they can be known.

This view certainly calls for very careful consideration. All hangs on the meaning of the phrase ‘theoretically predictable’. We must, of course, admit that the acts of the good man are not all predictable in actual practice. What then do we mean by saying that they are theoretically predictable? We mean that anyone with full knowledge of the man’s circumstances, and of his nature as a highly developed organism, could predict all his acts, even those of creative imagination. We mean, in fact, that the man’s acts are not arbitrary, that they follow from, or are expressions of, the intercourse of the man and the environment. If this doctrine seems not to take emergence seriously, either the doctrine has been interpreted to mean more than it does mean, or emergence itself has been misunderstood. We must not forget that, although the emergent behaviour of a whole cannot be predicted merely from knowledge of the nature of the parts unorganized, yet admittedly the behaviour of the whole is determined by the nature of its parts. .In organization they reveal (or if it be preferred, they assume) a ‘new’ nature; and anyone getting to know this new nature could predict the emergent behaviour.

But, it will be objected, the new nature is in principle unknowable, save from observation of the emergent behaviour itself. This contention I should reject. It is true that, for prediction of the emergent behaviour, some- thing is needed beyond knowledge of the reducible nature of the parts; but that something is theoretically obtainable. What is needed is knowledge of: (a) the reducible nature of the parts, (b) the environment, (c) certain principles which have been called the law or laws of hierarchical emergence. These last are theoretically discoverable by induction from experience of the whole range of emergent levels; and, once discovered, they would enable us to predict the emergent nature of any given whole, provided that we had knowledge of its parts and its environment.84 It may be objected that these laws of hierarchical emergence are figments merely of a pious hope. At present we know nothing of them; and we may be told that we are not even entitled to assume that emergence is systematic at all. Yet, while in comparison with what remains to be discovered we do indeed know extremely little about the principles of emergence, on the other hand in certain limited fields we have fragmentary but important inductive knowledge of the kind of whole that does as a matter of fact emerge in certain conditions, and of the kind of behaviour that is to be expected of it. Thus we do as a matter of fact know something about the kinds of behaviour to be expected from men of specific stock and specific circumstances. And the more we study, the more we can predict. To set theoretical limits to this advance were wholly unjustified. No doubt many extravagances have recently been committed by psychologists; but we must not forget that their science is still in its infancy. Though in practice their analysis, and still more their prediction, are extremely uncertain, they seem at least to have established certain very general principles which cover schematically the whole field of mental activity. Whether, indeed, psychology will ever advance so far as to be able to predict, for instance, the character of the next work of a particular artist (of course only after incredibly minute study of the particular case), is irrelevant to this argument. The point is that, though such prediction would entail incomparably more knowledge and skill and patience than the predictions which we effect to-day, and though some of the subtler principles which it would involve may as a matter of fact lie wholly beyond the grasp of human intelligence, yet our psychological experience strongly suggests that prediction of behaviour is limited only by our ignorance and lack of insight, not by anything arbitrary or incoherent in behaviour itself.

To sum up this matter, then, when we say that the good man’s behaviour is theoretically predictable, we mean only that even in its most splendid achievements it is a systematic expression of his nature, and that his nature itself is systematically connected with other things. We do not, for instance, deny that one factor in his nature is a certain determinate degree of the capacity of spontaneous loyalty to the great and remote things at the expense of the minor and intimate things. Nor do we deny that the more complex and creative kinds of behaviour demand for their interpretation knowledge which could never be derived from study merely of the less complicated kinds of beings. We claim only that the faith in systematic connexion is scarcely less justified in the mental than in the physical sphere. All prediction, indeed, rests upon the conviction that laws will hold good in the future just because, so far as we know, they have always held good in the past; and this conviction is incapable of strict logical defence. But the problem of induction applies alike to psychological and physical prediction, and need not trouble us here.

The advocates of ‘free will’ against ‘determinism’ are apt to lose sight of the fact that in practical life the regularity of the good man’s behaviour, and the predictability of his actions are far more important than the fact that he is a free agent, and that in a special sense he could be irregular and unpredictable.85 It matters little to his neighbours whether he is free or not so long as, freely or not, he does what is right. And surely we would rather our governors were automata who inevitably must act so as to achieve the good, than that they were less mechanical but more erratic centres of indeterminate ‘free will’. And for ourselves, would we not willingly sacrifice our freedom to choose right or wrong if we could thereby ensure that for the future we should invariably do right, though as automata? To wish otherwise would be to care more for our own righteousness than for the objective ideal.

Morality, indeed, at least in one sense of the word, depends on the freedom of moral agents. Were there no kind of freedom there could be no obligation in any ordinary sense. But morality depends not only on freedom but also on the distinction between good and evil; and this is quite independent of the question of freedom. It is easy to conceive a world in which, though there were no free agents, there were yet good things and bad things, and possibilities of goods and bads. In such a world acts might still be good and bad instrumentally; and in virtue even of their form alone they might still have intrinsic value. For instance, those acts might be intrinsically best which were fulfilments of the highest rank of emergent tendencies of agents. In such a world the good ‘ought’ to be achieved although it could not be achieved by free agents.

On one view, indeed, freedom is fundamental, namely on Kant’s view that nothing can possibly be good without qualification save a good will.86 Such a will, he held, is good not because of what it effects, but good in itself.87 All other goods are but instrumental to the creation of the good will. And when we are told to treat human beings always as ends and never as means, it is implied that the end which is their fulfilment is that they should fulfil themselves by freely willing the good will.

On this essentially moralistic view, then, the supremely important end in the universe is that free agents should choose to act according to an objective, universal and rational principle. Kant’s universal maxim is now generally admitted to be insufficient. ‘I am never to act otherwise,’ he says, ‘than so that I could also will that my maxim should become a universal law.'88 It is not difficult to show that the only universality characteristic of the good will is that, in each particular situation, it is just the will to do that particular act which, in the circumstances, will achieve the greatest good. The mere universality of the good will does not constitute its goodness; though universal it certainly must be, since in any sphere (to adopt Professor Hobhouse’s phrase) principles which are fundamental can admit of no exception. The goodness of the moral principle in fact must consist, not merely in that it can be universalized, but in that it is the universal application of the principle of the fulfilling of needs, in that it is universally the will for that which, in the universal view, is seen to be demanded.

Kant’s aim was to divorce morality entirely from the mere fulfilment of desire, or the mere following of inclination. For in his view the essence of morality was that it was free choice according to a purely rational principle, while inclination and every kind of need was conceived as external to the rational agent himself. To follow inclination, even the noblest inclination, was therefore to sell oneself into bondage, and to cease from being a free rational being. And, of course, it is true that the moral experience is often felt as a resistance of inclination for the sake of a rational principle. But clearly it is a mistake to suppose therefore that morality has nothing whatever to do with need, that it is, so to speak, a needless acceptance of abstract rationality. It is, indeed, as we have seen, a spontaneous espousal of an objective need, but in mere rational coherence there is not any need which can be espoused. Though, according to the theory adopted in this book, motive is objective, not subjective, voluntary action is none the less inconceivable without motive. Kant lost sight of the fact that moral conduct, in transcending one need, merely accepts another and greater. The rational principle in morals is but the principle of embracing the widest field of objective needs.

Kant’s prepossession with freedom was perhaps in part inherited from the free will controversies of protestant Christianity. But it was also the outcome of his own ‘Copernican revolution’ in the theory of knowledge. When the experient was taken to be fundamental in epistemology, it was natural that the attitude of the experient should seem the all-important fact in ethics. But however Kant was led to his theory, he seems to fail through a too exclusive interest in morality itself at the expense of the goodness of ends. It is well, therefore, to remind ourselves that, while without goodness of ends there can be no morality, without any shadow of freedom there may still be goodness of ends. Even if every human act be predestined in the nature of things, and all our moral struggles simply hallucinatory, yet it may still be in strictness better that one act should be done rather than another.

But in a world in which men have a sense that they are free agents confronted by an objective distinction between good and evil, ‘goodness of will’ must seem of great importance. For only through ‘goodness of will’ is the good likely to be achieved. But to suppose that ‘goodness of will’ is itself the ultimate ground of the distinction between good and evil is, if not precisely to set up a means as an end, at least to mistake a part for the whole. If we are to take human beings as ends in themselves, let us do so without evasion. Let us not take them as ends only in so far as they achieve an abstract form of will. Let us rather take them as ends, because they themselves, with their diverse needs, are capable of richer fulfilment than any other beings within our ken. Thus we shall at least avoid an abstract moralism. For the good will, after all, is only good in that it seeks ends that are good in themselves. A man of good will is just a man whose established habit it is to seek ends which seem best by acts which seem most effective

Nevertheless, the question of freedom is important in ethics. For, though freedom is not the ground of good and evil, it is certainly in some sense relevant to moral obligation. If the experience of free choice is illusory, the experience of obligation must be illusory too. If, when a man chooses, he could not possibly have chosen otherwise, there seems no meaning in saying that he was morally bound to choose otherwise. We may still say that the good ought to be achieved, meaning thereby simply that it is good that the good be achieved; but we have no right to say of anyone that he ought to have done otherwise than he did, unless we believe that he could have done otherwise. Where there is no freedom of choice, right acts seem to merit no praise, and wrong acts no blame; save in the sense in which all good things are to be praised, or rather appreciated, and all bad things condemned. Obligation can only occur if the agent can, but also can not, act so as to achieve the best in the circumstances.
B. Introspection of Volition

In every kind of choice we certainly do feel that we ‘can’ take either course. At this moment I am confident that I can either raise my hand or not ‘as I will’. I do not experience my choice to move it as something foreign to me, compelling me; my choice is simply I, as I am at the moment. If a fly settles on my hand and tickles me, I may still move or not move, as I will. But in this case I do experience an objective and rebellious ‘tendency to move’. And indeed the tickling may become so disturbing that at last my hand may twitch in spite of my will to keep it at rest. The more I am tickled, the more vigorously have I to ‘will’ not to move, in order to restrain the automatic action of my hand. By an act of volition I may interfere with the natural rhythm of my breathing, now accelerating it, now stopping it altogether. And the greater the divergence from the normal rate, and again the more prolonged the interference, the more vigorously must I ‘will.’ But if anyone will try for himself the experiment of holding his breath for a long while, he will find that, save in the last extremity, there is no actual rebellion of the physiological mechanism, and flouting of his command that it should not act. Rather, at every stage but the last extremity, the mechanism is absolutely under his control; but he becomes increasingly reluctant to control it, increasingly reluctant to issue the inhibitory command which, at every stage, he can issue, and at every stage but the last will be absolutely obeyed; but the organic consequences of the command become increasingly undesirable.

Similarly with hunger. A man may choose to starve rather than steal food; but, as the organic impulse to eat increases in urgency, his resolution may waver, and finally collapse. But at every stage, even (in this case) in the last extremity, his behaviour is absolutely under subjective control. When he finally succumbs to temptation, he chooses to give rein to a tendency which, had he but chosen otherwise, would have been curbed.

In a clear moral choice the good does not compel us in the manner of the last extremity of breathlessness. Nor does the evil; though many persons seek to excuse them- selves by believing that their choice of evil was the work of some force outside themselves. When we choose rightly we eagerly take credit for the choice, insisting that by a spontaneous ‘act of will’ we chose for no other reason than that the choice seemed right. But when we choose wrongly we are apt to say that the effort to control ourselves, to master our instinctive or habitual mechanism, was beyond our power. And it is certainly true that in panic-flight, or rash pugnacity, and even in gradual surrender to a subtle temptation, we may feel ourselves unable to resist the drive of a mechanism. But though the irrational choice may be called ‘compulsive’, it is choice all the same. It is not physical compulsion, but a definite psychical surrender to mutinous demands ‘within the citadel of the self’. In one sense, of course, both the reflex mechanism and the mechanism of instinct and habit are within the self’, in that both fall within the mental content. In another sense they are both ‘external to the self’, in that they are both objective to the subjective act of choice or will, which accepts or rejects tendencies in something other than itself, namely in the psycho-physical organism. But whereas pure reflex activity may take place without any volition, strictly psychical activity, even when ‘compulsive’, does involve an act of choice or will. The sense of guilt is evidence that, however we excuse ourselves, we know (or at least believe) that we I could have done otherwise.

It certainly does seem, then, that we direct our behaviour to some extent by spontaneous ‘acts of will’. Perhaps it is all a delusion. Perhaps the experience of choosing is but another case of the conceit of the fly on the axle, an epiphenomenon which has no influence on the course of events. Perhaps in conscious choice we do but cognize the activity of something which is not itself a conscious activity. There is no need to venture into such problems here. We need only insist that, until epiphenomenalism is proved, the experience of choice as efficient must not be ignored. After all, our belief in its actual effectiveness is very strongly confirmed. By a very broadly based induction we are convinced that with certain types of acts, if we do not choose to do them, they do not get done, and if we do cho............