You are here


It may be advisable to add here one or two observations on certain important topics which present themselves at more than one stage in the argument of these lectures, and could not therefore be disposed of in footnotes. I select, in particular, three such topics, because I think it possible I may be thought to have treated them, in different places, in inconsistent ways. I do not believe there has been serious real inconsistency, and I would ask the reader who suspects it at least to suspend his judgement until he has weighed the remarks now to be made.

A. The Rationality of the Universe

In some places I have spoken of the conviction that reality is a rational whole as the fundamental postulate alike of true science, true philosophy, and true religion; in other places I have spoken of the “rationalisation” of the universe as a task which, from the nature of the case, can never be finally achieved. The apparent consequence might be stated epigrammatically by saying that I maintain, in effect, that there are “irrationalities” which are not unreasonable. If this sounds like paradox, the paradox, I believe, is only apparent and arises from what Plato calls1 τὸ τῶν λόγων ἀσθενές, the inadequacy of language to convey the whole of a speaker’s meaning and nothing beyond that meaning.

By the “irrational” we may mean (1) that which is in conflict with the first principles of coherent thinking, the inherently unreasonable, as I should prefer to call it. It would be irrational in this sense to maintain that there are integers which are at once odd and even, are not, and yet also are, divisible by 2 without a remainder. To say that the real, or the universe, may be irrational in this sense would be to say that it is not only a riddle, but a riddle to which there can be no answer, because it is a question with no genuine meaning. A riddle which has no answer is not even a riddle. If reality were a pseudo–enigma of this sort, manifestly science, philosophy, religion would be alike worthless; all would be vain attempts to solve a conundrum which, ex hypothesi, has no solution, to translate “gibberish” into sense. But “gibberish” which could be rendered into sense would not be “gibberish”.

But we also speak of the “irrational” in a very different sense to mean (2) that to which we can find an approximate answer, or even a series of ever more closely approximate answers, but not a complete answer. “Irrationality” in this sense means only that we are dealing with a problem which we are always on the way to solving, but never have solved and never shall have solved. This is what we mean when we speak, in the language of the discipline from which the very word “irrational” has been borrowed, of an “irrational” magnitude or number. When we say that √2̅  is an “irrational,” we do not mean that the question “What number, multiplied by itself, will give the product 2?” is insoluble in the sense in which Lewis Carroll’s conundrum “Why is a raven like a writing–desk?” is presumably insoluble. For, as we know, we can readily find an unending series of fractions such that the product of any term of the series by itself is more nearly equal to 2 than that of any of its precursors by itself. We have a simple rule for constructing this series, and by travelling far enough along it, we can find a number of which the product by itself differs from 2 by a fraction smaller than any we please to assign. What we cannot do is to get to the end of this unending series, or, again, when we “extract the square root of 2” by the more rough–and–ready familiar arithmetical method, to come to a last “decimal figure,” or a group of recurring “decimal figures.” That is, we cannot answer our question “What number, when multiplied by itself, gives the product 2,” by pro ducing a fraction which has finite integers for its numerator and denominator. If x2/y2 = 2, x has not to y the λόγος or ratio of an integer to an integer, and this is why √2̅  has been called an ἄλογον, or “irrational.” But there is nothing unreasonable in the statement that some integers have “irrational square roots”; the unreasonableness would lie in denying this. For by denying it we should be asserting one or other of two propositions, (a) that there are actually pairs of integers which satisfy such equations as x2 = 2y2, x2 = 5y2, or (b) that if we consider the pairs of values of x and y yielded by the integral solutions of the equations x2 = 2y2 ± I, x2 = 5y2 ± I, though the “absolute difference” between 2 or 5 and the fraction x2/y2 steadily diminishes as we consider higher and higher values of x and y, it always remains greater than some assignable rational fraction σ. And both these propositions are at variance with the foundations of coherent thinking. The example will explain what I mean by a reasonable irrationality.

I hold, then, that because our intellect is not creative of the universe, but receptive of a reality which it has to understand but does not freely create, our problem of interpreting that reality by theory is in principle like the evaluation of a “surd.” We may, and should, make persistent efforts to carry our valuation a “place” further than any we have actually reached, but we can never expect to write down the “last decimal figure,” or the “last convergent,” if I may so express myself. This is what the rationalist pur sang, whether he is confession–ally as orthodox as Descartes at least meant to be, or as fanatically anti–orthodox as the contributors to the “Rationalist Press,” assumes that we can do, and this is why rationalism of that kind is inherently unreasonable. On that point, at least, I may claim to be loyal to the central thought of Kant. I would add that, so far as I can see, the case would be the same with a “separated intelligence,” supposing that intelligence not to be itself the Creator of the world. Even for the angels, the “works of the six days” remain the “unbegreiflich hohen Werke”.

B. Freedom and Contingency

To prevent misunderstandings I should like to state briefly what I take to be the essentials of such a doctrine of “choice and avoidance” as seems to me indispensable if our moral accountability for our voluntary actions is to be regarded as more than illusory.

(1) It is a fact that we, sometimes at least, really choose between alternative courses of action. It is not true that when we think we are choosing, the real fact is always that we are discovering that there is no choice open to us. (Whether there are some occasions when we fancy ourselves to be choosing, but are mistaken, I am not called upon to decide, but I am not concerned to deny that it may be so. My only concern is to maintain that sometimes at least all of us really do choose, and that the fact must not be explained away. All our choices are not “Hobson’s choice”.)

(2) Again deliberation is a real process, not a mere illusion. Sometimes, at any rate, we really weigh the goodness whether of alternative acts, A and B themselves, or of their consequences, before making our choice, and the weighing, sometimes at least, affects the choice. Deliberation is neither, as Hobbes thought it was, a mere oscillation between conflicting “appetites”2 nor yet a pretence of looking for reasons for an act which we are already “determined” to do. It is genuine “practical” thinking.

(3) Further, there is no reason to doubt that we can, and sometimes do, come to this process of practical thinking with minds not already prejudiced for or against either of the alternatives under examination, just as we sometimes consider the evidence for or against a statement of alleged matter of fact without secret prepossession either way. A man may come to the estimation of evidence with an “open mind,” devoid of any antecedent bias other than a desire to reach the truth about the matter under examination.3 Similarly he may weigh the alternative courses of action A and B with no prepossession beyond the intention to adopt the course which shall, on examination, approve itself to him as the “right,” or the “better.” If many men mistakenly suppose themselves to be impartial in deliberation when they are not really so, men also often suppose themselves to be weighing testimony or arguments with an open mind, when this is not actually the fact. Yet a man can be, and ought to be, candid and open with himself in deliberation, as he can be, and ought to be, candid and open in the balancing of testimonies, or the scrutiny of arguments.

(4) When the conditions thus laid down are fulfilled, it is strictly true to say that during the process of deliberating a man is “indetermined” ad utrumque: in fact, it is the deliberation itself which puts an end to this “freedom,” and “determines” him to one of the alternatives. Until he has deliberated he is “free” to take either course, to do a proposed act, say A, or not to do it.

(5) Such “freedom” does not mean that a man is ever “free” to take just any course he pleases. The alternatives between which I am effectively “free” to choose in a given case will always be limited in number, partly by my present situation, partly by my “past.” I am not “free,” at the moment of writing these lines, to choose whether I will go on with my writing in Edinburgh or spend the evening with a friend in Westminster, since I cannot transport myself forthwith to Westminster. Nor am I “free” to lay down my writing and read the Chinese classics; I have not in the past learned the Chinese language and so could not read a Chinese book, even if I had one at command. But I can choose either of the alternatives to go on with my writing or not to go on with it. If I could not, it would be equally futile to express moral approbation of my conduct if I stick to my work, in spite of the temptation to lay it aside for a diverting romance, and to express disapproval if I abandon my work for the story. Life would not be an education into morally stable character for us if it did not present situations in which we are confronted with the real alternative of doing the act A or not doing it, both courses being really open to us until one of them is blocked by our deliberation itself. Genuine morality would be impossible if it were true that when we take a decision, or suppose ourselves to do so, we are in a position like that of an engine–driver at the point of divergence of two sets of tracks, one of which is already closed against him by an invisible pointsman. I am my own pointsman, as well as the driver of my own engine.

(6) It follows that when I really deliberate and decide, my decision and the ensuing act, though largely conditioned by the past, which restricts the range of effective alternatives open to me (as in the supposed example, it excluded the dropping of my work to read a Chinese classic, though not the dropping of it ἁπλῶς), are not wholly determined by it. And therefore, when we prescind from the question of the range of effective alternatives, and consider simply the choice “to do A or not,” the “past” leaves the issue truly undetermined. To put the point in quasi–mathematical language, if my act is to be considered as a function of my “past,” it must be regarded as a many–valued, not as a one–valued, function of it. This, not the mere difficulty of obtaining sufficiently minute information about the events of another man’s past history, is the reason why it must always be impossible to calculate a man’s future unambiguously from knowledge of his past, and why there could never be such a science as the “ethology” contemplated by J. S. Mill (Logic, bk. vi. c. 5).

(7) It does not follow from these positions that it must always be open to a “free” rational agent who is not the Creator to make a morally evil choice (so that we should have to say that if men, or angels, are free agents, any man, or any angel, may at any moment commit any conceivable sin). For the discipline of the past closes many paths, though it may not close all. Our choice is not always between a morally right and a morally wrong, not necessarily always between a good and a better; it may perfectly well sometimes be between two courses equally good, but different.4 There is thus no inconsistency between such “freedom” as is implied in moral responsibility and the attainment of a stable character from which the discipline of the past has eliminated all possibility of effectively preferring the morally evil, or even the morally less good, alternative. It would even be possible, humanly speaking, that God Himself should always have open “alternatives,” though, if He has, they cannot differ as a morally better and a morally less good. But it is not necessary to make this assertion about God, since always to see the absolute best and to follow it because it is best is to enjoy a “freedom” far transcending our human “freedom of choice” between a bad and a good, a best and a less good. A man who loves his wife is not the less free because his love forecloses any effective possibility of deserting her.5

I think it will be apparent that these positions do not involve any unreasonable version of Indeterminism, and that they are fully consistent with the acceptance of the Socratic and Platonic dictum that to be in assured and unclouded apprehension of the “best” would always entail following it. And, so far as I can see, such “freedom of the will” as I am here maintaining is equally in harmony with the teaching, e.g., of St. Thomas Aquinas. If a man likes to say that he means something more than this by “freedom,” and therefore regards Plato and St. Thomas as “determinists,” I can, of course, have no objection to his saying the same of myself. But the doctrine here laid down is so different from anything which was taught by the inventors of the word “determinism,” or the scientific men who have adopted it as a badge of their profession, that I believe nothing but confusion can come of such a careless use of terminology. I may add a remark or two about “contingency” to make the position adopted still clearer.

C. Contingency in Nature?

There are writers for whom I have a deep respect who would, I believe, on consideration, accept all, or most, of the foregoing seven propositions, but would, at the same time, reject the whole conception of any real “contingency” in the course of events.6 As will have been visible from more than one passage in these volumes, I am compelled to take a different view, and to agree with James Ward that any interpretation of the world which is to make room for real history, real morality, real religion, must “let contingency into the heart of things”.7 Accordingly, though I do not appeal to the return of so many eminent physicists at the moment to the assertion of a “principle of Indeterminacy” in the physical at large as an argument for our moral freedom, I believe it to be an important step in the direction of a sounder metaphysic and cosmology. The opposing view, which regards contingency as an illusion begotten of our ignorance of the details of becoming, seems to me to rest in the end upon a misunderstanding of the meaning of the “contingent.” It is taken to mean the capricious occurrence of events which have no sufficient “why and wherefore” in the plan of reality, and might “just as well never have occurred at all”; such meaningless “random” occurrences are then truly said to be incompatible with a genuine theistic faith in the divine government of the world. Or it is also said that they are excluded by the divine omniscience; “if God eternally knows the whole course of history, how can any of the events so known be contingent?” And yet an intelligence which does not know the whole course of history cannot be the God demanded by religious men, for of it it could not be said without reserve, “Trust in the Lord with all thy heart, and lean not to thine own understanding”.

Now here there is, I believe, a bad confusion of thought. It is antecedently most unlikely that such philosophers as Plato and St. Thomas—if I do not add Aristotle, my reason is that the famous αὑτὸν ἄρα νοεῖ seems intended to exclude the course of events from God’s knowledge—should have believed with equal conviction in divine omniscience and divine government of the world, and also in contingency (the πλανωμένη αἰτία of the Timaeus), without seeing the glaring contradiction, if it really does “glare.” And I think it not hard to satisfy one’s self that the contradiction is no more than apparent. As St. Thomas is careful to explain,8 a contingent event does not mean an event which has no cause, or is not “determined” relatively to the supreme (in the older terminology the “superessential”)9 cause, the divine purpose, but one which is not unambiguously determined by its more “proximate” causes. (Thus, to take the standing example, it was held that the motions of the heavenly bodies are “necessary” causes of certain effects, e.g. of the alternation of day and night. But among the effects of the motion of these bodies we have also to include the growth and ripening of crops on earth. Now in this particular case the effect of a “necessary” cause is a contingent event, because there may be some debilitas in the seed which has been sown, and in that case the effect, the ripening of the harvest, does not follow. In fact, in this case, the revolutions in the heavens are not the proximate, but a remote (though not the ultimate and “superessential”) cause of the result considered, and it is therefore not fully “determined” by them.)

It would, no doubt, be hard to defend this doctrine of contingency to–day in the precise form in which it was used by the great schoolmen, who inherited Aristotle’s unfortunate and perverse crotchet of a radical distinction between terrestrial and celestial “matter” and their respective dynamics.10 We tend at once to meet the Thomist example of the harvest which is “contingent” because it is sometimes abundant and sometimes fails, by saying that the presence or absence of a debilitas in the seed is itself a part of the whole cause of the effect—so far, of course, the scholastic could concur—and that it is a neither more nor less “necessary” cause than the motus solis; if the scholastic thinks otherwise on this last point, that, we say, is because by a cause he means an agent, and he mistakenly supposes the “seed” not to be an agent in its own growth, but to be simply and purely passive, a view made impossible to us by our conception of reciprocal interaction in physics. And, as already said, he is also unfortunately imbued with the Aristotelian fancy of the contrast between the immutability of the “heavens” and the mutability of the sublunary region of the universe. If we are to retain the distinction between necessary and contingent causation, we shall be driven to say that the “superessential” cause, God, is the only cause which causes with complete necessity, all other causes, remote or proximate, “celestial” or “terrene,” being infected with contingency.

If we make this modification, the doctrine seems to me to be perfectly intelligible. It means, in effect, that while everything that happens in cosmic history happens as God ordains or permits, no event is a perfectly determinate “one–valued function” of other specific events, and that when we say that the occurrence of X may certainly be inferred from the occurrence of A, B, C, … there is always an understood Deo volente. It may be that the ultimate “pattern of the whole” demands a divergence from the most uniformly exhibited “routine of sequence,” and if it does, the sequence will not occur; the sun will, at need, “stand still upon Gibeon.” But whether the sun stands still or “hastes to go down,” it is certain that there is a “pattern of the whole” and that it will not be violated. No “innovation” will be a capricious departure from it. But it is impossible in principle to calculate from data already in our possession whether and when an “innovation” will take place, because the “pattern of the whole” is not and cannot be a datum. (Or, to take an illustration from human action, it would be manifestly fallacious to argue that a phrase found in the published work of a writer must be an “error of the press” because the same writer has published many thousands of lines, but has nowhere else used that particular phrase. If it is the specially right and appropriate expression of the thought in his mind at the moment of writing he may use it, though he never used it before and will never use it again. A man’s habits of speech have a great deal of influence on his choice of phraseology, but they never absolutely dictate it.)

It would thus be wholly consistent with theistic belief in the government of the world by God to recognise a genuine element of contingency in all historical events. You may in a sense resolve this contingency into defect of knowledge on our part, but only if you mean that we are not fully acquainted with the divine purpose. The defect could not be removed by any extension of our acquaintance with the details of past cosmic history, since the fullest acquaintance with them would not put us in possession of the “whole counsel of God.” There is thus, so far, no reason to take up a priori an attitude of opposition to physicists who tell us they are led by their own special studies to admit a “principle of Indeterminacy” pervading the whole physical order. They may be right, or they may be wrong, but they are not saying anything which conflicts either with the inherent reasonableness of the universe, or with theistic faith. Professor Eddington, for example, is not maintaining that Δῖνος βασιλεύει, τὸη Δί᾽ ἐξεληλακώς.

Nor do I see that the admission of contingency conflicts with belief in the divine omniscience, as is often supposed. It would do so, if we impiously thought of God as inferring our future from our past much as an astronomer calculates the future positions of a planet from a record of positions it has occupied in the past. But no theologian, I take it, ever thought of God’s knowledge in this fashion. To quote James Ward, “How God knows, or even what knowledge means when attributed to the Supreme Being, few of us will pretend to understand”.11 But, as Ward is arguing in the context of the remark, at least it will not be imagined that He calculates the course of events, like a “Laplacean demon,” from a multitude of differential equations. Whatever omniscience is, it is not this.12

These observations leave it still an open question whether it is requisite for human freedom of choice that there should be “contingency” in nature at large. May we not accept all the seven theses we began our discussion of choice by formulating, and at the same time deny that any natural event really is contingent? (Perhaps nature, at all events, really is bound “fast in fate”?) Clearly, of course, the denial of contingency, if it is to leave human moral freedom unaffected, must not be extended to those physical events which are the expressions of our responsible choices, the actus imperati which carry the actus elicitus which is my decision over into the physical order. If it is true that the movement of my hand is ever the result of my choice, then that movement cannot be a determinate one–valued function of previous events of the physical order; these events must leave it an open issue whether my hand is to move or not. Consequently, the same consideration must apply to all events of the physical order which depend causally, no matter at how many removes, on the choice of a moral agent. And of how many actual events, if of any, could we say that no actual choice by any moral agent is conceivably to be found among their causal antecedents? Theoretically, however, we might, I conceive, say that events of the physical order which have no acts of choice by moral agents among their causal antecedents, if there are any such events, might be regarded as wholly non–contingent without any compromise of the positions upon which the reality of man’s moral freedom depends. Even if God be needlessly assumed to have bound “nature” fast in fate, our moral freedom may be none the less real, provided that by “nature” we only mean whatever in the actual physical order is entirely independent of causation by the choice of a moral agent, if anything is so independent. To assert moral freedom, one need not assume the omnipresence of an element of “indeterminacy” in physical processes as such. There are apparently good grounds for this assumption, but they are of a different order. (I should perhaps add that I should regard it as very rash to assume that there is a single physical event which is wholly independent of the causality of some moral agent, since I see no reason to suppose that men are the only such beings in the universe. And in this context, when I speak of “moral agents,” I am, of course, intending created moral agents, whether human or otherwise.)

I suspect that the reason why some excellent writers who seem to assert freedom of choice in express terms yet describe themselves as “determinists” is that they assume that Libertarianism is necessarily committed to this admission of contingency as a cosmic principle; and that they regard such a conception as “unscientific.” I would urge on any reader of my own who takes this point of view, two considerations: (1) In point of fact there is apparently reason to believe that contingency is actually making its way back into scientific thinking on strictly theoretical grounds, as forced upon us in the interpretation of experimental results13; (2) in any case, this is not the issue really at stake between Libertarian moralists and the “scientific determinists.” It is not contingency in “nature,” but choice which the determinists of the nineteenth century were anxious to explode as a superstition, and they have left their representatives behind them. If anyone doubts this, I recommend to his notice an address on “The Nature of Life,” delivered by Professor L. Hogben to the British Association at Cape Town on July 25, 1929. Mr. Hogben, at least, makes no attempt to disguise his conviction that all human moral purpose is an illusion; the whole social and moral life of man consists of “conditioned reflexes” which have no purpose, and a “new school of psychologists,” with whom the speaker clearly sympathises, “has come into being with the express object of … relieving Man, the celestial pilgrim, of his burden of soul”.14 It is surely a pity that moralists who would regard this reduction of the spiritual life to “conditioned reflexes” as the death of all morality should mark their dissent, where it exists, from those of us who believe in contingency in “nature” at large, by adopting a label which confounds them with the “scientific” enemies of responsibility and practical reason.

It is hardly necessary to add that the “Libertarian” is left by his theory perfectly free to recognise that the full character of human “free” action is only to be found in acts of conscious deliberate choice. How far impulsive acts can be said to be done with freedom, and, again, how far my choice is free when my own past misconduct or negligence has closed alternatives which would otherwise remain open, is another question.

D. Free Will of Indifference

A reader of the preceding paragraphs may conceivably ask whether I mean to assert or to deny the reality of what has been called “free will of indifference.” Do I, or do I not, mean that we can, and sometimes do, choose between alternatives without a “motive” for our preference? I should reply (1) that if there are such “unmotived” choices, they must surely have no significance for our moral life, since they do not express the character of the agent supposed to be making the choice. It is just the choices which are rooted in our personal moral quality and give expression to it with which the moralist is concerned. If “motiveless choice” occurs at all, it may fairly be taken to occur only in connection with the kind of insignificant movements regularly treated by the schoolmen as their standing examples in discussing the possibility of morally indifferent acts (barbam vellere, festucam de terra tollere, and the like). Or, to put the point differently, “motiveless choice,” if really possible at all, would be a grave abuse of our liberty in any matter of the slightest moment, because it would mean refusing to deliberate in a case where we ought to deliberate.

Further, it is not clear that there is, even in these apparently trivial cases, anything we can properly call unmotived preference. This becomes clear, I think, if we define our terms with a little care. A motive, we must remember, is not the same thing as a mere impulse which releases, or discharges, an act. To act with a motive is not merely to be impelled to act in a certain way, but also to regard one’s act as justified by a certain consideration. When I say that I act thus and with this motive, I mean both that the considerations I allege are truly those which impel me to act as I do, and also that they make my acting as I do the right and reasonable thing for me to do. A motive is always something which, at the time of acting, the agent regards as a reasonable incentive. It is a “reason” in the double sense that it explains why the agent does what he does, and that, so long as he does not repent, it is held by him to justify his behaviour. It follows that a man’s “motives” are rarely, if ever, present to his own mind at the moment of action in “clear and distinct” apprehension; they are usually very largely “subconscious,” or “habitual.” But this does not detract from their rationality. A driver who has learned the British rule of the road “drives to the left,” because he has learned that this is the established rule, and that it is dangerous to disregard it. He does not actually recall these considerations—if he has really “learned how to drive”—as he steers himself through the traffic. If he is at all practised, he regulates himself “automatically” by the rule. But it is a rule, and it is because he has knowledge of the rule so deeply ingrained in him that his “secondary automatic responses” are what they are. His whole conduct is an example of rational choice; it does not issue from what some writers are fond of calling the passional nature, but from intelligence. In the vast majority of those voluntary acts which are not preceded and conditioned by explicit deliberation, scrutiny will, I believe, reveal “motives” as rational as the driver’s preference for the recognised rule of the road. In most cases there is intelligent “justification” for the course adopted, and the agent would not have taken that course if he had not been acquainted with that justification, though he was not actually thinking about it at the moment of acting. (I do not, of course, mean that the “justification” will always bear strict investigation; in the case of our morally wrong acts it will not. I mean that there are considerations which the agent regards as justification, and to which he will sincerely appeal, if the morality of his act is disputed. The man who has taken a human life will at once plead, if his act is impugned, that “it was his life or mine,” and this is meant, and is felt by the homicide to be a rational justification of the fatal shot or blow, though it is another question whether the plea will satisfy the “impartial spectator”.15)

I believe this analysis applicable to almost all the normal acts of human beings, when free from external constraint. There are grounds which, in the opinion of the agent at the time of acting, make his act the reasonable one to be done. Those grounds are not commonly before his mind, since most of his acts are done without explicit deliberation between alternatives. But they are in his mind, as is shown by the readiness with which they are produced in reply to any suggestion that his conduct has been unreasonable. I am, therefore, convinced that it is a mistake to attack the standing doctrine of Greek moralists, that the sinner does wrong because he is misled by a false judgement of good, on the ground that it over–rationalises human action. If unmotived, or unreasonable, choice occurs at all, it only occurs, I would submit, in connection with alternatives which are taken to be morally indifferent. It might be alleged that it occurs here. “Where you can take either of two courses, A1 or A2,” it may be urged, “and there is no reason for regarding either as in any way more or less good than the other, clearly the fact that you take the course A1 shows that you are making a choice, and yet, ex hypothesi, you know of no reason why A1 should be chosen rather than A2. Here, then, there must be unmotived choice”.

But will the argument really stand examination? A typical example would be that of a man who is about to play a game of chess and is “offered his choice” of taking the white pieces (and attacking) or the black (and defending). In discussing such a case we need to draw distinctions. It may be that the player to whom the option is given knows himself to be stronger and more practised in attack than in defence, or vice versa, and chooses accordingly. He may do this with clear and full consciousness of the reason for his choice. Or he may not be consciously thinking about the matter, and yet it may be what really decides his option, as is shown by his reply, when asked, e.g., why he chose white, that “I am more accustomed to the white pieces and more at home with them”.16 In neither case can it fairly be said that there is not a rational motive for his choice, though it may be a “subconscious” one.

But what of the case of the man who is equally expert in the attack and the defence, and knows this? He also may be offered his option, and he must make it, or there will be no game. Is not this a clear case of making a choice which must be unmotived? It does not seem to me that it is so. The man in question has, indeed, no motive for choosing White rather than Black, or Black rather than White. But he has a motive for making either option rather than declining to opt, since if both players are equally expert, and both know it, and therefore refuse to make any option, the game, which is what both desire to have, will never begin. I think, therefore, that what really happens in such a case is that the player who is “offered his choice” makes a real choice which has a motive, and a sound one, the choice to foreclose one of the alternatives, but does not really choose as between White and Black. He simply says the word which happens to “come to the tip of his tongue.” In practice we commonly avoid this situation of having to make what appears to be a choice between equally desirable alternatives by enacting a rule that the point shall be decided by “tossing up.” That is, we voluntarily remove the particular decision from the sphere of the voluntary. So again, when I, who am not much interested in such things, am offered a choice between two dishes or two wines, I feel sure that I often make no real option; I say the word which “comes handiest,” merely because I want to get the point decided one way or the other. This is making a real and rational choice between settling the question and leaving it open, but not, as it seems to me, a real choice for one alternative as against the other.

On these grounds I feel very doubtful whether any genuine choice is really without a rational motive, i.e. without what the chooser, at the moment of choosing, regards as a reasonable ground for preference. Even when, to take the old example of the schools, I pick up a straw from the ground, I should probably not do so consciously unless I disliked the look of “litter,” or wanted to exercise a group of muscles, or something of the kind, and these are rational grounds for choice. The nearest approach we make in actual life to “indifferent” choice, I should say, is made in the cases when we rationally will to eliminate one of two alternatives, but do not care which is eliminated. This is not a typical case of morally significant “free choice,” but rather, in the words of Descartes,17 infimus gradus libertatis. It is not in our “indifference” in such a case that we show our freedom, but in our resolution to bring the indifference to an end.

A final word may perhaps find its place here, as a Rechtfertigung against the charge, urged more than once in private correspondence against the present writer by Dr. Rashdall, of clinging to an “unintelligible” Libertarianism. If I have no desire to find the source of responsible moral freedom in a liberty of caprice, why am I not content to treat moral freedom, after the fashion of Leibniz, as spontaneity along with the consciousness of spontaneity?18 Why do I hold that a free man is not adequately described as automaton spirituale? I would reply by reminding my reader of a striking passage in Kant’s second Critique.19 Kant is there admitting the existence of moral “incurables,” on whom all education and discipline is wasted. They manifest utter moral depravity in early childhood, and grow only the more depraved as they grow older. But we are justified, he says, in treating them morally and juristically as no less responsible and accountable than others, and they themselves admit the justice of this attitude, “in spite of the desperate native mental constitution thus imputed to them.” This, Kant pleads, is an argument for his rigid distinction between temporal appearance and eternal reality. The depravity displayed through life by the “incurables” is itself merely the consequence of the “free causality” of their morally evil wills.

What does this amount to, if we have once rejected Kant’s identification of the temporal with mere appearance, but to the doctrine that the “incurable” is created incurable, and then held accountable by his “dark Maker” for the flaw in the Naturbeschaffenheit seines Gemüths? It is the horrible Augustinian notion of the massa perditionis reduced to its simplest terms. The “incurable” is imagined to be sent into the world already “damned,” with a will already and unalterably “wholly averse from God.” And we are expected to acquiesce in the justice of this situation. (I do not dwell on the difficulty of the quaestio facti whether there are such “incurables.” If our failure to strike the right note with some transgressors could be taken as evidence of their incurability, I am afraid, when I consider how helpless candid self–scrutiny seems to prove us all to be against some of our weaknesses, that we may fairly suspect ourselves and all mankind of belonging to the massa.) If we seriously believe in the theory, can our moral theology be anything better than a dishonest attempt to curry favour with a malevolent Maker by flatteries we know to be undeserved? If there were no Creator, or an evil Creator, the difficulty would not arise. But since there is a Creator, and a righteous and merciful Creator, we cannot reconcile determinism with an ethical Theism by assuming that some men have been created already “damned.” And we must not shirk the issue, as Kant tries to do, by saying that the “incurable” is not created “damned,” but damns himself once and for all by a primal free act of wrong choice which is not “in time”.20 This is a rank “unintelligibility.” For a “first act” of the series of my transgressions must have a place in the temporal series to which the rest of my transgressions belong, and thus, on Kant’s own theory, it should be part of the “phenomenal series” of consequences, not the “intelligible” cause of the whole series. If Kant’s language is to have a tolerable meaning, the primal free wrong choice should be taken merely as an imaginative symbol of the character exhibited by all our temporal wrong choices, and in becoming such a symbol it ceases to be an explanation. If it is more than such a symbol, our actual moral life is deprived of the significance Kant in particular is anxious to ascribe to it as a discipline into goodness of will; in the case of the “incurables,” the discipline and struggle must be no more than illusion, and none of us can be sure that he is not himself one of their number.

I see no way out but to strike at the root of the whole conception by insisting on the utter “creatureliness” of all finite agents. Nowhere in them is there any element of character which is unmade, an eternal and unalterable datum. Their being is always a γένεσις εἰς οὐσίαν, never simply οὐσία. And the admission destroys in principle the foundation of all determinism, “hard” or “soft.” The real “unintelligibility” seems to me to be with the determinist who is, consciously or unconsciously, transferring to the creature, or to some ingredient in its composition, the “once–for–allness” in–communicably proper to the Creator. And for that reason I cannot feel certain that there are actually any “incurables”; the notion may have its uses, as a check on moral presumption, but it may be only a “limiting concept”.

  • 1.

    Ep. vii. 343 A.

  • 2.

    Elements of Law, pt. i. c. 12: “This alternate succession of appetite and fear during all the time the action is in our power to do, or not to do, is that we call Deliberation.” Leviathan, c. 6: “When in the mind of man, Appetites and Aversions, Hopes and Fears, concerning one and the same thing, arise alternately … the whole summe of Desires, Aversions, Hopes and Feares, continued till the thing be either done, or thought impossible, is that we call Deliberation.”

  • 3.

    E.g. Did Virgil write (Ecl. iv. 62) “qui non risere parenti,” or “ quoi non risere parentes”? Surely it is ludicrous to suggest that I cannot consider the question without a secret antecedent bias.

  • 4.

    As, for example, when a man considers whether he will spend his holiday in the Scottish Highlands, seeing lochs and mountains, or in Italy, seeing cities and pictures. It may be that, for a given man, either course is as good as the other, though the two goods are different. Or one might have to choose between two different careers without being able to say that one could serve God or man better in the one than in the other.

  • 5.

    Thus God’s freedom should probably not be called “freedom of choice.” (Kant, it will be remembered, denies that we can properly speak of Triebfeder in connection with the divine activity.) “We must not conceive God to be the freest agent, because he can doe and prescribe what he pleaseth, and so set up an Absolute will which shall make both Law and Reason, as some imagine. For as God cannot know himself to be any other than what indeed he is; so neither can he will himself to be anything else than what he is. For this were to make God free to dethrone himself” (John Smith, Of the Existence and Nature of God. c. ii. § 6).

  • 6.

    I am thinking particularly of the avowed “determinism” of such moralists as Dr. Rashdall and Dr. McTaggart, and, again, of the position taken by the Rev. C. J. Shebbeare in his recent Problems of Providence. To judge from the incidental remarks on the subject in Five Types of Ethical Theory, Dr. Broad would probably agree still more closely with the general view I have tried to set forth.

  • 7.

    Naturalism and Agnosticism, ii. 280.

  • 8.

    Cf. S.C.G. i. 85 “requirit autem ordo universalis aliquas causas esse variabiles, cum corpora sint de perfectione universi, quae non movent nisi mota … unde videmus quamvis causa remota sit necessaria, si tamen causa proxima sit contingens, effectum contingentem esse.” S. Th. 19, art.8 resp. “cum igitur voluntas divina sit efficacissima, non solum sequitur quod fiant ea quae Deus vult fieri, sed et quod eo modo fiant quo Deus ea fieri vult. Vult autem quaedam fieri Deus necessario, quaedam contingenter, ut sit ordo in rebus ad complementum universi.” Professed Thomists, I observe, commonly speak of three kinds of effects—“necessary, contingent, and free.” But I presume that “free effects” are not meant to be “contra–divided against” the other two as a third species, but to be understood as a sub–class of the contingent, “contra–divided against” the contingent but unfree.

  • 9.

    Thus R. Bacon in his Commentary on Aristotle, Physics, i.–iv. (Oxford, 1928, p. 249), speaking of the succession of the seasons, distinguishes (1) the superessential cause, the divine dispositio of the universe; (2) the remoter cause (causa longinqua), the revolution of the primum mobile; (3) the proximate cause, “the movement of the sun in his proper circle,” viz. that of the Ecliptic.

  • 10.

    It cannot be too carefully remembered that the distinction was introduced into cosmology by Aristotle, and that it is, in particular, anti–Platonic.

  • 11.

    Naturalism and Agnosticism1, i. 42.

  • 12.

    Cf. the remark of St. Thomas (S.Th., art. 171, q. 6 ad. sec.) that “divina praescientia respicit futura secundum duo: scilicet secundum quod sunt in seipsis, in quantum scilicet ipsa praesentialiter intuetur; et secundum quod sunt in suis causis, in quantum scilicet videt ordinem causarum ad effectus. Et quamvis contingentia futura, prout sunt in seipsis, sint determinata ad unum, tamen prout sunt in suis causis, non sunt determinata quin possint aliter evenire.” That is, it is eternally part of the divine providential plan that a certain event shall happen: also, it is not the case that this event is what I have called a “one–valued function” of preceding events. Both these truths are known to the divine Mind.

  • 13.

    On this see, e.g., Eddington, Nature of the Physical World, pp. 220 ff.; Whitehead, Process and Reality, p. 30.

  • 14.

    “The modern mechanist,” says Mr. Hogben in the next paragraph of his discourse, “does not say that thought and love and heroism do not exist; he says, show me behaviour to which you apply the adjectives thoughtful or loving or heroic, and we will, one fine day, endeavour to arrive at predictable conclusions with reference to it by following the only method of enquiry which we have learned by experience to trust.” But if the “endeavour” is to be successful, if we are, “one fine day,” to discover that all the acts we call thoughtful, loving, heroic, can be predicted without taking the existence of thought, love, heroism, into account (and not one of the three can be discovered as a “laboratory” fact), how does the position Mr. Hogben accepts on behalf of his “mechanist” differ from the position he disclaims? What is meant by saying that “love exists,” but that there are acts which cannot be predicted without knowing that the agent loves someone, or something? And does Mr. Hogben never count on the good behaviour of his banker, or his servants? If, like other men, he sometimes does so, will he say he has studied banker or servant “by the only method of enquiry” he has “learned by experience to trust?” (My references are to the report of the discussion published by the Cape Times as “revised by the authors”.)

  • 15.

    Or, to take a standing example from St. Thomas, fornication is malum in se, and therefore has no real justification. But it is true that the fornicator—unless he is actually deliberately sinning “in contempt of God”—is taking the means to a delectatio carnalis which is, considered simply as such a delectation, bonum quoddam temporale. He is not wrong in thinking that this bonum is a bonum so far as it goes; but there is a superior bonum with which it is incompatible. The sinner is not alive to the superiority of this other bonum, and hence, from his point of view, his conduct appears to be rationally justified. (This is, of course, why Aristotle says that it συμβαίνει πως ὑπὸ λόγου ἀκρατεύεσθαι.)

  • 16.

    Or, as might be the case, “I am more accustomed to Black, and so wish to take this opportunity of practice in handling White.” Greater familiarity with the pieces of one colour may lead to either choice, according as the chooser cares more about winning on this particular occasion, or about “improving his game.”

  • 17.

    Meditat. iv. “indifferentia autem illa quam experior cum nulla me ratio in unam partem magis quam in alteram impellit, est infimus gradus libertatis, et nullam in ea perfectionem, sed tantummodo defectum sive negationem quandam in cognitione testatur.”

  • 18.

    Though possibly this is an unduly minimising interpretation of Leibniz’s own phrase, “spontaneity along with intelligence” (spontanéité qui devient liberté dans les substances intelligentes), Discours de métaphysique xxxii.

  • 19.

    Kdpr V. I. Th. i. B. iii. Hptst. (Werke, v. 104).

  • 20.

    I put the matter as Kant himself puts it in Religion innerhalb d. Grenzen d. blossen Vernunft. In the Kdpr V. (Werke, v. 106–7) he speaks less pictorially. We are there told that if space and time were more than “appearances,” it would follow that the moral responsibility for the conduct of creatures rests with their Creator, and not with themselves. But a Creator creates only realities, and space and time are merely phenomenal. God is therefore the cause of my existence as a free agent in the intelligible world, but not of my actions in time and space. Surely we must say that this way of relieving my Creator from responsibility for my sins reduces the whole moral life to an illusion. Kant apparently wants to reproduce the scholastic reasoning which argues that God is not the author of my misdeeds, since God created me free, and I freely choose to do wrong. But he ruins the force of the argument by trying to make it turn on the “ideality” of time and space. If that were part of the argument, it should also follow that God is not the cause of any of the observed events of the natural order. The “argument from design” must not only cease to be probative; it must also lose all that right to our respect which Kant himself claimed for it. And it then becomes very hard to understand how Kant’s virtuous man can be entitled to a rational faith that the natural order is controlled by God in the interests of a moral end,—the crowning of virtue with happiness.