Charles Hartshorne
“Being . . . is undeducible. For our intellect it remains a casual and contingent quantum that is simply found or begged. May it be begged bit by bit, as it adds itself? Or must we beg it only once, by assuming it either to be eternal, or to have come in an instant that co-implicated all the rest? Did or did not ‘the first morning of creation write what the last dawn of reckoning shall read?”’ William James, Some Problems of Philosophy, p. 189.
It is an old view, much defended in recent books and articles, that “freedom” in the moral sense is compatible with causal “determinism.” The motivation inspiring such discussions appears to be this: moral freedom is an indispensable idea, strict causality is also indispensable, or at least science might need to assert it; hence it is important to show that the two ideas do not conflict. In addition, it is held that even for ethics we need insight into the probable results of our acts, viewed as causes, and that a responsible person is he whose acts are effects of a good character, while an irresponsible person is he whose acts spring from an unstable or bad character.
On two points one may well agree with the views just expressed. Moral freedom is certainly indispensable, and nothing could justify denying it: men do choose, and in some cases their choice is influenced by ethical principles. Second, it is true that, even in ethics, we require causality. But what I wish to urge is that the universal validity of causality is not the same as determinism. The point is simple: there is a deterministic definition and an indeterministic definition of “cause,” and the difference does not turn upon whether or not “every event has a cause.” It turns rather upon how causes are thought to be related to their effects, or antecedent conditions to subsequent happenings. Here is where determinism and the (relative) indeterminism I wish to defend part company. One may, if one wishes, use the word “determinism” for the mere assertion that every event has its cause or causes, and “indeterminism” to mean that at least some events have no causes, thus making the issue a choice between two absolutes (so far as these events are concerned), instead of a triadic choice between two absolutistic views and one relativistic view. But if this is done, then it is arguable that indeterminism is a doctrine no one defends, a fictitious position invented for purposes of controversy. By ostentatiously burning this straw man, deterministic writers frequently distract attention from the real opponent who is trying to defend something else. I have read scores of defenses of determinism (including under this description writings which maintain that determinism at least could be true), and how few of them there are which do not, at some point, solemnly commit this lifeless effigy to the flames!1
Every event has its cause or causes; so far we nearly all agree. And to avoid quibbling, if anyone objects to the word “cause” as obsolete in science, let him substitute “antecedent conditions,” so that the argument may proceed. But not every event — indeed some of us would say, not any event in its concrete actuality — is fully and absolutely determined by its causes. In other words, an indeterminist (as conceived in this article) rejects a certain definition of “cause,” namely that it is a condition, or set of conditions, from which only one outcome is possible, or from which, in principle or ideally, the outcome is wholly predictable. To be substituted for this is a definition which, whatever else it includes, involves the following requirement: the cause is a state of affairs such that when granted something more or less like what happens subsequently was “bound to happen,” or (if you prefer) could safely have been predicted. Given dry TNT, a confined space, and a lighted fuse, there will inevitably, or with practically infinite probability, be an explosion; but it does not follow, and indeterminism denies, that the exact details of the explosion, the behavior of each atom and particle, will be the only possible ones (in principle, uniquely predictable) under the circumstances. Read fifty arguments for the tenability of determinism (beginning, say, with David Flume, Part III of the Treatise), and there will perhaps be one or two that address themselves consistently to the issue as thus defined; and probably not one which shows much comprehension of the reasons which have led various philosophers and scientists, including Clerk Maxwell, Whitehead, and other distinguished thinkers, to adopt the indeterminist view.2
Our question, then, is whether causes or conditions determine happenings absolutely, or whether they merely limit more or less sharply what can happen. “More or less sharply” will perhaps seem hopelessly vague. However, one may state the indeterminist or relativist view more subtly, as follows: events are always to an appropriate degree determined by their causal antecedents. And what is “an appropriate degree”? I think we can give at least a rough answer to this question. A human being, in full possession of normal intelligence, surveying wide alternatives of action under general conceptions whose very meaning is that they admit highly divergent possible instances, must dispose of a wider range of possible reactions to a given situation than there is reason to attribute to a molecule reacting to its situation. The scale of chemical and biological types of individuality, from atoms (or lower) up to man, may reasonably be looked upon, and by some scientists and philosophers has been looked upon, as a hierarchy of degrees of freedom in possible responses to given causal conditions (including among these conditions the past history of the individual). Inanimate nature involves the least scope of alternatives — and here the “more or less determined” means “more”; man involves the widest scope — and here it means very much “less.” Thus we need not make man an arbitrary exception to the general principles of nature; he is but the intensive case of the general principles of creative action, of which causality is an aspect.
I have now betrayed my secret, in that word “creative.” Moral freedom, as we all know, requires the exercise of rational reflection and decision; but what many philosophers fail to see is that this exercise of higher powers involves a creative leap beyond anything made inevitable or predictable by the causal conditions. The creative act is influenced by its conditions, and requires them, but it cannot (I wish to argue) be required or precisely determined by them — much less, even, than an electronic event according to the Uncertainty Principle. This aspect of creativity is what the determinist overlooks or denies. How shall we convince him of his oversight?
First, we may ask where the burden of proof lies. Truths in general are relative, not absolute; they are matters of degree, subject to quantitative limitation of some sort. Determinism, however, is an absolute, holding that conditions unqualifiedly restrict the outcome to but one wholly definite sort of event, entirely excluding any creative leap, any novelty not specifiable in advance, given ideal knowledge of conditions. It is somewhat amusing to converse with persons who tell you that they are suspicious of “all absolutes,” and that they accept absolute determinism. They may not use the word “absolute” in the second case, but what does that matter, if they affirm the unqualified causal determination of events? Surely we know by this time that no accumulation of scientific observations could establish, even as probable, an unqualified regularity. Observation does not have that kind of exactitude.
Second, if process is not (in some degree) always creation, what is it? This was Bergson’s point, still poorly digested in philosophy. If it is only ignorance of causes which prevents us from mentally seeing the effect in advance, then the entire character of “coming events” is real beforehand. Before events happen, they lack nothing except a totally transparent, featureless something called “actual occurrence.” To some of us this is truly an absurdity. If becoming does not create new quality and quantity, new determinateness, then, we argue, it creates nothing, and nothing ever really becomes. (I pass over a possible counter-argument from “precognition.”) A causal destiny, no less determinate than what happens, is just the happening twice over, once as already true but not yet real, and then as true and also real. But truth is “agreement with reality,” in some sense, and if the truth is already there, then so is what it agrees with. How can there be a wholly determinate relation prior to its term? In short, creativity is an essential aspect of the idea of becoming or process. Conditions do indeed — condition, that is they establish and limit the possibilities for otherwise free or creative activity.
And this phase “creative activity,” or “creative becoming,” only escapes redundancy because there are degrees of creativity, implying the zero case as a lower limit of thought, a necessarily fictitious entity, like “perfect lever,” or “wholly isolated particle.” Thus one may, in a relative sense, speak of “uncreative process,” and at the opposite extreme, by a more violent exaggeration, of “wholly free” or causally undetermined action, where the creative leap is maximal, though in any real case by no means uninfluenced by its conditions. (Consider Kubla Khan, and the known sources for the images and ideas of Coleridge’s poem, together with what is known of his antecedent poetic habits and character.)
Third, the conception of law now actually operative in the sciences is not a deterministic one. To be sure, many are still telling us that the Uncertainty Principle is irrelevant to the problem of freedom. It does not seem to occur to those reiterating this statement that the disconnectedness of basic principles is far from being an axiom of scientific method. Rather, something very like the contrary is axiomatic. What such people may have in mind is: (a) the human organism, viewed statistically, involves large numbers of quanta, so that any indeterminacy due to quantum laws must practically approach zero;3 (b) the freedom of the electrons, if it be called that, is their freedom, not ours; (c) we know that we are morally free, and it is silly to suppose that we need quantum mechanics to justify this principle of everyday life; (d) indeterminacy as experimentally justified expresses only our ignorance of ultimate causes, which may in an undetectible manner fully determine even electrons. I grant all but the last point as probably sound. But as to the last, it must be said that many scientists and philosophers would reject the notion of unobservables which it involves, and that some competent investigators would go further, and hold that there is positive evidence that real randomness is at work, that the laws are genuinely laws of chance. This is so (according to my understanding) apart from the Uncertainty Principle, in the case, for example, of radio-active atoms. However, my main objection is that the four points together do not, even if granted, establish the irrelevance of quantum mechanics to the understanding of human freedom as a fact in nature. For it remains true that quantum mechanics explains the world not through causal laws which contradict the notion of creative leaps or unpredictable aspects of happenings, but rather through laws which are in principle compatible with that notion. The ground laws of the world, as we seem ever likely to know them, are thus essentially statistical, in the sense that their demonstrable exactitude is due to the presence of large numbers of similar events, and not to any knowable precise causal determination of the events taken singly. An exact regularity supposed to be hidden behind these statistical laws is at best irrelevant to scientific explanation. It apparently can have no technological consequences, nor any appreciable experimental significance. Thus the contention, so much and so long insisted upon, that science requires us to think purely deterministically, has been concretely, and probably definitively, discredited. Einstein and Planck resisted this conclusion to the last; but how many among the younger scientists take this resistance as more than a slightly pathetic example of scientific conservatism? Long ago Clerk Maxwell, who with Gibbs, Darwin, Mendel, and Bolzmann was a founder of the statistical approach which emerged in the second half of the Nineteenth Century, suggested what is now perhaps the prevailing scientific view, that the statistical notion of law is the fundamental one. Nearly a century ago this superb thinker calmly set out some of the essentials of this matter.4
A famous physicist once said to me, “The uncertainty principle is not relevant to the question of moral freedom — except [he added] by analogy.” The exception is significant. For if submicroscopic particles are (for our possible knowledge, at least) somewhat indeterminate causally, and if the exactitude of the laws we verify is due to our dealing with large numbers of particles, why should we suppose that particular human individuals are subject to exact laws? A man is not identical with super-billions of particles. He is one human individual. Indeed, is not our very notion of unity, or individuality, based on our own experience as one? The statistical laws of particles can hardly tell the whole story of human behavior, for there is at least one more entity present in that behavior besides particles, namely the human personality, or the stream of human experiences. It is at best sheer assumption that the presence of the higher-level unity makes no difference to the behavior of the particles.5 Quantum mechanics need not be thought the entire and exact truth concerning particles in organisms or in the human brain. Subsidiary principles are presumably needed, if human thoughts and purposes are not “idle wheels in nature,” to use Whitehead’s phrase. These subsidiary principles may be expected to reintroduce the merely statistical character of causal regularity at the higher or human level, so that the single human being will not be bound by the causal conditions to a single determinate course of action. (We shall consider presently the contention that particles are not existent entities.)
Fourth, we must evaluate a familiar deterministic tenet which is that chance or randomness is one thing, and moral freedom quite another. Dice, it is caustically pointed out, are not free. However, that moral freedom is not the same as chance is no proof that it does not require an element of chance. Moral freedom, we have held, is a special, high-level case of the creative leap inherent in all process, the case in which the leap is influenced by consciousness of ethical principles. The leap itself always involves “chance” — meaning simply that the causal conditions do not require just the particular act which takes place. They make it possible, but they make other acts also possible. Moral freedom is chance plus something; no one (except a man of straw) identifies the two.
Fifth, the notion that ethics requires absolute determinism is an oddity in an age when relative standards of value (and relative almost everything) are commonplaces. Moreover, the contention is quite baseless. Ethics requires relative determinacy, and this is precisely what opponents of strict determinism are glad to affirm. For “relative determinacy” and “relative indeterminacy” say the same thing, with a mere difference of emphasis. (This is forgotten when it is sometimes said that determinism and indeterminism are equally erroneous. On the contrary, while the one affirms an absolute, the other by no means asserts the contrary absolute, but rather, it recognizes relativity.) No one would deny, though some popular proponents of indeterminism underemphasize and gravely neglect, the causal conditioning in behavior. For instance, with respect to the argument that responsibility involves character as a cause guaranteeing certain kinds of acts, all that any careful writer disputes is that the guarantee is absolute, or that the “kinds” of acts reduce, at a given moment, to just one precisely-determined course of action. Must ethics be more severe in its demands for law than even physics?
Consider the role of “motives,” of which the determinists have made so much. A motive, in a human being at least, is always more or less general, e.g., the wish to “make a good impression”; and an act is never general but always particular. What spans the gap between the general and the particular? The character? But this is something general also: an established type of action, not a determinateness of agency with respect to the particular act now to be performed for the first time. The creationist holds that our very power to form general conceptions (in a sense in which these are beyond the reach of the other animals) is the same as our being not determined by irresistible impulse, habit, or antecedent character, to but one mode of acting in a given case. The openness to alternatives, the flexibility, of our response is the behavioristic aspect of our knowledge of the universal, as that which can be indifferently instanced by this particular or by that. Such instancing, by its very meaning, must have wide ranges of freedom. Freedom in the indeterministic sense is thus inherent in rational understanding as such, understanding through universals.
Let us now consider the following “paradigm case.” “Suppose two young persons, normal, of opposite sex, with obvious congruence of background and nature, no available third person calculated to attract either, no obstacle to marriage, and no inclination to ‘free love’; could not a psychologist (or perhaps anyone) predict that they will marry? And will such predictability deprive their decision of its freedom ?“ So determinists sometimes argue. But they fail to make the requisite distinctions. Does “predictable” mean with certainty, or with probability? If the former, then will any careful scientist accept it? If the latter, then, since probability only refers to what will happen in many similar cases, the individual case remains unpredictable. Again, it is one thing to say, “The two will marry,” and quite another to foretell a single concrete action that either will take. Every act of either party might be strictly unpredictable, and yet it might reasonably be foretold that the marriage would (barring death or other accident) take place, for this only means that, whatever series of actions the two might perform, the abstract features of “proposal” and “acceptance” would be involved. And indeed, many writers tend to forget that the particular acts we feel to be free are concrete, while what is predicted is not. Moreover, it is common enough for people in such situations as we have imagined above to declare that it is fate or necessity, not free will, that has brought about their union, that they “could not have helped themselves.” Granting this, their behavior may still have had moral character. For while they perhaps could not have avoided marrying, there are thousands, or an infinity, of ways in which this could have been done, so that at every moment they may have exercised moral choice. Moreover, their conscientiousness may have been such that had there been any cogent moral objection to the marriage, they would have recognized and yielded to it and thus have removed the step from the class of features “common to the real possibilities.” Would it then have been excluded from all the possibilities, so that not marrying would have been inevitable? Even supposing this to be so, still, moral choice may have been involved farther back, in the creative acts, from infancy onwards, by which such a firm ethical character bad been developed. When we feel morally responsible for an act, this need not mean that we think that, as we were the instant before, we had the psychological capacity to refrain from it. It may only imply that at some time previously we could have entered upon a different course from the one which led to the act. Until we learn to do justice to the subtle relativities of this problem, we shall merely beat the air (or each other) with verbal alternatives.
Sixth, the idea that to know is to be able to predict, so that any inability to predict must mean partial ignorance, is, I hold, untenable. The ultimate function of knowledge is not to foresee, but to create. The two are by no means coincident. We predict most perfectly astronomical phenomena where we have no creative control. We control most perfectly our bodily behavior, but how vague and uncertain are our predictions of this behavior! We must ask whether by “control” is meant settling things in advance, and once for all, or — step by step? To predict what is to happen tomorrow is to deprive tomorrow of the right to decide for itself. I predict where I shall be weeks in advance when I admit or establish a conclusive obligation or reason to be there; but in this case, I renounce the right to make the decision later on. I must already have made it. Is the goal of knowledge thus to put all decision into the present, leaving for the future the mere execution? But the life of decision-making is life itself. We should be dead, from then on. Does not science aim most basically at power, rather than foresight? Is the goal of psychology to be able to write the poet’s poem beforehand by predicting it, to foresee, and thus make, the creative decisions, and their expression in speeches and laws, by which statesmen, or an entire people, resolve some political tangle? I suggest that, since creative activity is that which leaps unpredictably from its causal base, the proper function of psychology in regard to creation is to increase our freedom from the compulsions and fixations which often make behavior only too largely predictable. When therapy (or a good upbringing) has freed a man to the full use of his constructive powers, it is not for science to say how in particular he will use these powers. That would be sheer impertinence and redundance. Statistical laws applying to large numbers of similar cases, yes, but the numbers and the similarity are more limited than in physics, and hence the laws cannot be so precise. Where behavior is essentially unconscious or mere habit, as in the knee-jerk, or in all cases where the cells of the body act with no appreciable interposition of thought, there statistical laws applying to the numerous bodily parts may give highly predictable modes of action. But all truly individual activity, especially on high levels, means the impossibility of precise prediction.
The idea of predictability, when turned into an absolute, like so many other ideas when so treated, loses all sense. Scientific prediction is not an end in itself, whose absolute attainment is our goal, but rather (a point insufficiently emphasized in this article) we use prediction as a means of finding out errors or blind spots in our knowledge of the statistical possibilities open to creative action. Discovering, by verified prediction, the principles of atomic fission did not make the future of things on this planet more precisely foreseeable. What it did was to show us hitherto undreamt of, or inadequately discerned, potentialities for the future. Primitive man knew well enough in rough outline what he himself was going to do; but he knew very little indeed of what could be done by other men in other circumstances with other means at their disposal. What upset the expectations of the primitive Australians was the coming of white men, with their unpredictable ways. But who can fail to see that while the white men were not so unpredictable to themselves as to the natives, they were at least as unpredictable to themselves as the natives had been to themselves. Science probably decreases the predictability of human events, in the absolute sense of “predictable”; and the race to catch up by conditional scientific prediction seems a losing race. Moreover, the more power men acquire over nature, the more closely prediction is assimilated to social and political prophecy. If “we” could control the planets, who would predict their movements? Fortunately, to foresee the future is not the purpose of life. The purpose is rather to maximize the opportunities for good in the future; it is not to be able to say now exactly what use will be made of these opportunities. Deciding that will be the future; it is no task for the present.
A final consideration against determinism is that it confronts us with the dilemma: either admit with Hume that events have no logical connection with their conditions, or accept an unqualified rationalism, according to which temporal succession, if “real” at all, coincides with logical implication. Indeterminism enables us to explore a third possibility, which is that conditions logically imply not any particular subsequent event as their inevitable outcome, but only a class of “really possible” outcomes (excluding as impossible a vacuous outcome of nothing at all happening). The features which all the really possible outcomes have in common are then necessary, i.e., they are bound to be actualized whatever happens. This is the precise meaning of “necessary,” namely, “common to all the possibilities.” What is common to most of the possibilities is probable. Thus temporal succession is quite different from unqualified logical implication: yet there is implication. Something like what happens next was bound to happen — so far, there is “necessary connection — but the precise particulars just do happen, quite without necessity or implication. To be sure, all this is no “answer to Hume,” until or unless it is made intelligible how conditions can limit the scope of “possible outcomes” to a definite class, with traits in common. I have elsewhere tried to explain, following Peirce, James, Bergson, and Whitehead, how this can be understood.6 The point here is that the sole answer to Hume which has not been refuted (it has rather been ignored) is one which, according to its leading expositors, implies indeterminism, or — the same thing — relative determinism.
In contrast to this relativism, contemporary thought tends to combine two oddly-associated absolutistic notions: (1) at each moment anything could conceivably happen next; but (2) if we are rational we shall expect only what the conditions and laws imply. We can think anything following upon anything else (thus, as the poet had it, lowing herds might “break from the starry skies”); but we ought not to think of anything but the one thing which is causally implied. Thus an idle unlimited freedom of conception is combined with a theoretical narrowing down of conception to a single course of events. In this view, events and the laws connecting them are absolutely distinct; the laws, though perfectly strict, are not characters of events, but — of what? They are, as it were, between events, not in them. This merely separate reality of causal relations which Russell (following Hume) has done so much to fasten upon the learned world (partly by excess of reaction, perhaps, to Bradley’s equally excessive monism) is in strange company with the unqualified unity of pattern asserted for the world process throughout time. Forget “laws,” and events have simply nothing to do with each other; remember laws, and there is really but one vast complex event or event-system stretching into the remotest past- and future. Thus we have absolute pluralism and absolute monism, equally by fiat or by magic. The two contrary and similarly excessive contentions taken together make the universe unintelligible. No wonder it is contended that philosophy can tell us nothing about reality: certainly it cannot when it is proceeding thus in absolutistic fashion, and in two opposite directions at once. What we need is a qualified or suitably relativized doctrine of the logical independence and dependence of events. The key notion is that events logically and precisely require the very antecedent conditions which they have, but not precisely the results which issue from them.
An English physicist, influenced by Wittgenstein, tells us that there is no such thing as physical, but only logical, necessity; yet he also claims that logical necessity is simply a relationship among our symbols (if I have not misunderstood him).7 So then when an overtaken rabbit finds he cannot escape from but must go into the fox’s jaws, this necessity derives from our way of talking about the world! Alas, I cannot make it out. I think there is physical necessity, and it is also logical. The logic, however, is that of events, and not simply of our thought about events. (God’s thought, perhaps, but not essentially ours.) But there are also open possibilities in nature, and this is what determinism in its pure form denies.
To prevent misunderstanding, it should be said that my contention is not that we must either accept a certain brand of indeterminism or give up causal explanation and prediction. This last we cannot do, for every animal must have its expectations modified by past experience; and above all man must exhibit such modification. But it remains a genuine alternative whether we can understand an objective counterpart in nature to our capacity to predict (statistically or approximately) or must be content to exercise the capacity without further curiosity. Those who disparage such curiosity because we can make our predictions without satisfying it are forgetting, I think, that the wish to know transcends pragmatic needs and is a principal glory of our natures. In addition, such thinkers often suppose that the alternative to Hume can only be a strict rationalism according to which causal laws suffice to render events (ideally) deducible in all detail. The irony of this is that Hume himself held the rationalistic view! His skepticism attached not to the Newtonian or deterministic concept of causality, but only to our means of demonstrating it to be true of nature. That we should assume it as valid he had no doubt. Our cultural situation calls for very great doubt on just that score. Indeed, we have apparently much to gain and little or nothing to lose by adopting the properly qualified or relativized indeterministic theory. Creative freedom of individuals as such, with statistical regularities expressive of the influence of causal conditions, gives us what we need, whether in ethics or in science.
One of the most painstaking and detailed discussions of our topic is a posthumous essay, characteristically acute and learned, by Ernst Cassirer.8 In view of the deserved prestige of this author and his extraordinary grasp of the history of science and philosophy, let us see what we can glean from him. Cassirer defends what he terms “critical determinism,” which he sharply contrasts with mechanism, or the Laplacean conception of the complete predictability of the future, given absolute knowledge of the past. Absolute knowledge, our subtle author reminds us, is not a human affair, and if the knowledge is thought of as divine, then there should be no question of prediction, since a divine intelligence is to be conceived of, if at all, as a single eternal intuition of all times. (I have frequently defended a different concept of omniscience, as did the Socinians hundreds of years ago; their well-formulated and pointed arguments on this topic seem to have escaped our learned author’s attention.) Critical determinism is the acceptance of the regulative principle that we are to look for “strict” laws, in seeking to explain nature. If, as Cassirer seems disposed to grant, the only strict laws we can find are statistical, that is, applicable to collectives, not to individual things or events, this does not abrogate the strictness of the laws or imply any compromise with indeterminism. (In other words, “determinism” and its denial are redefined so that one may remain attached to the former word, but give up the chief point which it has been used to express!) Moreover, he holds, where properties of supposed individuals elude observational detection (for instance, the simultaneous location and velocity of electrons), there simply are no determinate properties of the kind in question. The law of excluded middle has, it seems, no application at this point.9 Nature is what she is humanly knowable as, nothing more. Nevertheless, there is talk of individual events, which must have definite properties, and it is hard to see how there can be collectives unless there are singulars. Yet these singulars, whether construed as single human perceptions, or as events independent of our knowledge, are in either case, as Margenau reminds us toward the end of his preface to the book, subject to no knowable laws except statistical ones.
Before making a few further remarks on this issue, I shall relate for the reader’s entertainment an incident which happened in my presence at a meeting of the Australasian Philosophical Association. A lively young man had read a paper dealing with the philosophy of science; at the close of his reading another lively and still younger man eagerly rose to put a criticism or question which, unluckily for him, he prefaced with the confession, “I don’t know anything about physics . . .,” whereupon the reader of the paper broke in harshly, “Then why do you talk about it”? Therewith that segment of the proceedings came to a dead stop; we never did learn what the criticism was to have been. Warned by this incident, I shall not confess to knowing “nothing” about physics; and it would be incorrect, since I have read works by many excellent authorities with some care, and conversed with others. At worst I could be said to know next to nothing. Yet who can resist talking about these things?
Heisenberg appears to have a strong case for his view that the mathematical beauty (“symmetry”) of the laws stands or falls with the admission that what is described in them is neither the absolute incidence of particles, as observed or unobserved, nor the individual events of observation, but rather the statistical possibilities for the latter. This does not mean that the human mind has to intervene, for a machine can be set to do the “observing.” But it does mean that what physics deals with as strictly lawful is not the singular actualities, observed or otherwise, but the potentialities of nature as expressible in innumerable cases of reaction, such as an experiment is. Attempts to reduce these potentialities to conceived actualities subject to precise laws have so far proved incapable of clarifying the theoretical structure or of leading to new facts. No one can be forbidden to seek an escape from this (to some persons unsatisfactory) situation. But a conclusive argument in the deterministic direction seems out of the question, and there may never again be even a strongly plausible one.
Von Weizsäcker told me once how he quit discussing philosophy with N. Hartmann after he learned that the latter refused to admit a distinction between actuality and real potentiality. Physics apparently needs just this distinction, which — as Peirce and James saw some time ago — is the nub of the determinist controversy. Not what does happen can be strictly lawful, according to the indeterminist, but only what can or would happen in sufficiently many cases.
The point is not, as I see it, the positivistic one that unobservables cannot exist, but rather that we can infer definitely as to what is beyond experience only through laws, and according to the “Copenhagen” view the only definite or actual facts in any way accessible to us are those of everyday experience. If we want to go further and “penetrate into the details of the atomic happenings, then the contours of the objectively actual world begin to dissolve, not into the mist of a new and still less clear conception of actuality, but into the lucid clarity of a mathematics, which expresses the lawful structure of the possible not the factual.”10 Real potentiality is thus the subject-matter of atomic physics, for “possible” here is not simple logical possibility, but is “objective.” Deterministic philosophies, if they understand themselves, deny this objectivity. They want the rigor of law to apply to what actually happens, to facts, not to classes of really possible happenings or facts. The patterns of probability of atomic physics are then subjective devices for dealing with the actual. But it makes at least as good sense to view them as the very framework of law of the objective world, within which, with an element of chance and freedom (the only positive meaning of “chance,” as Peirce pointed out), actual facts are born. Law is exhibited in this theory as a limitation upon chance, not its absence. And thus creativity can be given a pervasive role.
Is it not of some significance that one of the most distinguished scientific representatives of determinism today, the psychologist Skinner, has as his principal working conception that of the “probability” or “strength” of a mode of response in a given organism at a given time, a strength which can be increased or decreased? How can he fail to see that he is dealing with real potentiality, not with inevitable, singular facts? Surely a probability of one is here a mere limiting concept. True, he says the psychologist would be content to “settle” for psychological “indeterminancy” if it were no greater than that which the physicist has had to accept.11 But why assume that the indeterminancy is no greater? If indeterminancy as positive (and every negative fact has a positive side) consists in creativity, then who would suppose an atom to be as creative as a human being? And I see no sign that Skinner has faced the absurdity of the psychologist undertaking to do in advance (i.e., predict) the “creative thinking” for facilitating which his book gives some admirable advice. Nor is there a necessary connection between the fine causal accounts in Skinner’s work and his determinism. What he shows us is that (as all sensible people know) creativity has narrow limits, which can be shifted by stimuli in all sorts of ways which (also within limits) are predictable. As James Clerk Maxwell said, not without a touch of exaggeration, “if there is any freedom it is infinitesimal.” I sometimes think of an individual’s freedom as the fraction of which the numerator is the momentary experience of the individual, and the denominator is the past of the universe, so far as effectively involved. The value of this fraction is small, but still not zero. A pigeon so conditioned that it definitely will perform a certain act within ten seconds, or two minutes, has no creative options as to performing or not performing that act in that time. This does not mean it necessarily suffers from any feeling of “compulsion” or that the act is in no sense voluntary, but it means that the act is not an example of creative freedom. And a man equally narrowly determined with respect to a certain act is in that respect not creative either. But this is quite compatible with indeterminism. A pigeon’s life in the segment of time mentioned is immensely more complex, even though in slight and trivial ways, than the mere “act” describes; much more so a man’s. If the creative leeway is cut off here, it can open up there, and when it has no opening the creature can cease to be, as an actually sentient or conscious individual — in deep sleep, for example.
It takes time to outgrow so deeply fixed a habit as determinism represents for many. But the working physicists seem already largely to have outgrown it. To some persons there is a kind of impiety in this, and in very truth determinism has been almost a deity to multitudes.
Cassirer shows this attitude clearly enough. He agrees that laws have been shown to be essentially statistical. But then he turns special pleader and urges that this is not “indeterminism worthy of the name.”12 “Real” indeterminism would be the doctrine that not merely are there alternative ways in which a singular event can happen, but there are alternative laws which nature might apply to a given case. Taking account of the ultimate time perspective, or of a possible succession of cosmic epochs, each with its own set of laws, this is not necessarily absurd. But, be this as it may, the usual meaning of determinism is that every event in all details is unambiguously specified by its conditions and the causal laws. We shall see how Cassirer’s indifference to the question of singularity avenges itself when he comes to deal with ethics. And it is clear that he has an ax to grind. He wants to make an absolute out of causal order and to drive the defender of relativity into the contrary absolute position, that there is no order, but only chaos. Thus the doctrine of the merely relative validity both of “order” and of its contrary are by -fiat denied access to verbal expression. And in the process something else becomes relative indeed, namely the independence of physical reality from our knowledge! Nature is the lawfulness of human experience. To me it seems that nature is a sublime totality, of which humanity is but a very tiny and dispensable fraction.
In the final chapter, in which our erudite author turns to the question of ethical freedom, his bias becomes especially clear. Thus he repeatedly identifies the denial of strict determinism with “limitless indeterminism,” with the assertion of action which should “simply fall out of the causal nexus,” etc., thereby missing entirely the possibility of “appropriate” degrees of determinacy, proportional to the various levels of being. The vice of thinking in mere dichotomies here raises its ugly head. We are not surprised to find human freedom and worth practically identified with “reliability,” stability of good character,13 as though there were no such thing as creativity, whose measure is precisely its transcendence of any mere permanence or derivability by rule from the past. And what has become of the admission that strict laws apply to collectives, not to individuals or singular events? Is a human experience of decision not a singular — is it, as seems to be hinted, a mere collective of atomic occurrences? Cassirer refers us to the Kantian notion of a noumenal freedom not in space or time. One had thought that this idea, or rather formula, had been sufficiently criticized by now. He also refers to Spinoza almost as though there were no serious objections to metaphysical necessitarianism. He does not refer to the views of Locke, James, Descartes, Lequier, Renouvier, Boutroux, Varisco, Bergson, Peirce, concerning the untenability of determinism.
I entirely agree with Cassirer on one point, that the self-identity of individuals or substances through time is not the ground, but rather an aspect of, the interconnectedness and lawfulness of events. I also agree that electrons are probably not enduring individuals moving through space. But it does not follow that we are free to define the realities of nature simply in terms of humanly knowable regularities or laws. Nature is a stream of interconnected events, and both laws and “things” are functions of these events. Each event is a determinate, but not antecedently determined or predictable, act of concretion, endowed with its proportional spontaneity or possibility of partial self-determination. Lequier’s great saying, “God has created me creator of myself” (not a contradiction, though a paradox), applies to every concrete unit-event. Ethical freedom is merely the kind of self-determination appropriate to events in which there is consciousness of a hierarchy of ideas, hence of alternative possible kinds of actions, not merely alternative details of actions all of one kind. A dog can only act doggishly; but it has a very different meaning to say that a man can only act humanly. For we choose not simply among details, but among categories, of actions, and under more or less conscious reference to a highest principle of good which transcends even the difference between man and other animals, and gives us obligations to the reality including all species.
Cassirer protests against the blurring of boundaries between sciences — for instance between ethics and physics.14 It does not seem to occur to him that science has no intention of permitting man to be regarded as outside nature, and that biology is bound to conceive the scale of animal forms as a single problem. Nor does he see that, as various biologists have pointed out, it would be silly to refuse to take advantage of the fact that in ourselves we have the one individual piece of nature which we know in its individuality from two sides: externally, quantitatively, and by behavioristic observation and test, and also internally, qualitatively, by immediate intuition — whether termed introspection or retrospection (short-run memory) is no matter. Here is our only complete clue, not to noumena out of time, but to concrete spatio-temporal reality. To hold that nature is what is knowable as absolutely determined by strict laws is to imply either that the qualities of individual experiences, whether human or subhuman, are thus determined, for which there is no scientific evidence at all, or else that there are no such qualities. Neither alternative seems to me defensible. What transcends our knowledge through strict laws is simply “experient events” (Whitehead), whether human or otherwise, in their concrete qualities. There is little in present-day science to even suggest exact laws applicable to such events taken singly. Cassirer is trying to have it both ways, to have his determinism absolute or strict, yet not mechanistic or Laplacean; but he can do so only by rendering ambiguous the notion of reality as applicable to individual events, without which “collectives” must remain unmeaning.
Our author’s learning saves him from altogether overlooking the significant truth that it is only through the law of increasing entropy, a statistical law if there is one, that the direction of time can be construed in physics.15 But the drift of his discussion tends to minimize the asymmetry of time. He would like to think of causal order as applying equally and in the same way to retrospective and prospective relations.16 Thus he exhibits what I call the prejudice of symmetry, which was so neatly embodied in the old saying, “The cause must be equal to the effect.” Logical analysis shows that symmetrical ideas like that of equality are derivative: thus “X is equal to Y” merely means that neither is greater than the other; while on the contrary “X is greater than Y” cannot be derived from the denial of their equality. The ultimate relations must then be non-symmetrical or directional. If causality is ultimate, it must be so either in the form, the cause greater than the effect, or in the converse relation. Which is it? The answer is again given by simple analysis of meanings. In a causal transaction in which, from the set of conditions C issues a result E, we have first C alone and then C and E. Thus the total result of causation is in every instance an enhancement of reality, the creation of a new whole. The more comes from the less. This (crudely sketched) is Bergson’s and Whitehead’s idea of process as creation. Any denial of the creative aspect of process leads to antinomies. Thus, if it be said that when E is actual, C the cause has ceased to be, we face the contradiction of an alleged causal relation of E to C, although there is said to be no such thing as C to serve as term of this relation. And if it be said that from an ultimate, or eternal, point of view there is no such thing as “C alone,” but only C as prior to E, or E as subsequent to C, then, since these two complexes are equivalent, one is defying the indications of logic that equality, a symmetrical relation, is derivative. I incline to believe that logicians will eventually, though some of them reluctantly, accept the theory of asymmetrical creativity as ultimate. But this acceptance, when and if it comes, will be a vast intellectual revolution whose consequences can only in part be foreseen. One foreseeable consequence is that absolute determinism will be definitively discredited. Another consequence must be that the doctrine of the timelessness of truth will be limited to necessary truths, and abandoned with respect to truths of fact, truths about particular events or any non-eternal entities. A tense or modal theory of factual truths, such as Aristotle, Lukasiewicz, Prior, have tried, not very successfully, perhaps, to work out, will then become an unavoidable task.17
Cassirer is partly right when, to the notion of an electron as an individual making choices, he objects that physics has now discredited the notion of electrons as enduring individuals identical through time.18 But as a philosophy of creativity sees the matter, he is partly wrong in supposing that this destroys the relevance of electronic indeterminacy to the question of free choice. For the locus of freedom is not in the enduring ego, but in the self here and now, the subject immanent to a unit-event. Cassirer with all his vast grasp of thought has never quite comprehended what a philosophy of creativity, really thought through, is about. The decider of a present issue is not simply identical with the self which resolved a previous issue, but is a new decider.19 This is the only consistent way to maintain the notion of self-creation. It is the precise meaning of freedom, as transcendence of mere regularity or law. It is also the way to answer those, who, like Schopenhauer, argue that a self which has been created (no matter by what) cannot be free, for its creator must, by determining its character, have also determined its acts. For we are as we act and act as we are. The solution of the puzzle is to deny any antecedent character or cause sufficient to determine present action. My present self is not real until I act, and its becoming determinate is the occurrence of the act, not the cause of this occurrence. Effects are more determinate than their causes, causation is creation of new determinacy. Thus there could be electronic decisions even though no electron survives from one moment to another as the same.
Another point which Cassirer is too knowing to miss, but, as it appears to me, too biased to see clearly, is the connection between atomism or discontinuity and indeterminism. Let us start with the most striking case, a living organism. Such an organism reacting to stimuli is subject to a threshold principle. Below a certain minimum of stimulation there is no response; as one gradually increases the stimulus, abruptly a response occurs. This disproportion between continuity and abruptness can easily be spanned by a statistical law. From zero probability of response one goes toward a probability of one. Thus, viewed statistically, the disproportion between continuous increase of stimulus and abrupt emergence of response disappears. But who can imagine a non-statistical law here? In each given case, response occurs or not. This situation is not peculiar to living things: indeed, the point of quantum physics, in this regard, is that the problem of a threshold is general. Individuals are sensitive to their environments, but are they infinitely sensitive?20 Rather there is an all-or-none law. Thus a quantum of light is either reflected or refracted by a prism; there are, as physicists say, only two “degrees of freedom” in the response, while the angle of incidence of the stimulus, or causal condition, is varied continuously. True, the angles of reflection or refraction vary continuously, but nevertheless the abrupt disjunction between going through the glass and returning from its unpenetrated surface is without parallel on the side of the conditions. There we have only continuous variation of angle. Again, suppose a ball rolled through a tube aimed from various angles at the sharp-edged top of a metal sheet. If the angle is at one extreme, a ball will certainly drop on one side of the sheet, if at the opposite extreme, on the other side. In between it becomes more and more uncertain on which side the ball will fall. Again we have an abrupt disjunction between two possible outcomes, and a sheer continuity on the side of the causal conditions. Once more, suppose a drop of water falling upon a knife-edged divide and rolling then into either of two oceans. The drop may split, to be sure, but not into an infinity of fragments, one for each possible angle! Here too, we have the problem of relating discontinuously variable results to continuously variable conditions.
The disproportion is overcome by the continuity of probabilities; how else would it be overcome? It is not a question of our ignorance, but of the mathematical form of the situation. It is no accident that Einstein, with his Spinozistic bias, should have sought to reduce quantum mechanics to a mere corollary of field physics. He was trying to get rid of discontinuity or (as Heisenberg profoundly put it, individuality), with its implication of a creative leap. Cassirer had a different way of seeking to escape the implications of atomicity: he merely denied that individuals, so far as they fail to exhibit regularity of action, exist. The statistical states of collectives constitute the entire reality; definite members of the collectives, seemingly, there are none. Is this a genuine solution?
Cassirer makes one final effort to show that statistical laws are just as incapable of providing for a freedom of indifference, of open alternatives, as classical or “dynamical” laws. Not only are there statistical regularities in the number of suicides, but also, according to some authors, in each kind and method of suicide, down to fine details. Reference is here made to some resounding claims of Qu6telet and Buckle concerning the analogy between physical and sociological laws. Let us grant that “freedom of indifference” may not be the correct phrase; the various sorts of acts which could result from a given set of conditions, including those furnished by the individual’s own antecedent character, are not equally probable. But it does not, so far as I can see, follow that there is a certain act which has a probability of one, and other acts but zero probability. And if it be objected that probability is merely relative to our knowledge, merely subjective, then I maintain, with many present-day physicists, that all of science must on that assumption be equally subjective. The very meaning of “law” is modal; the controversy over contrary-to-fact conditionals, or over dispositional properties, is still seething, and there seems no reason why it should not go on forever, or until objective possibility and probability are conceded. Freedom is choice among really possible acts, acts possible not only given the external situation but also given the pre-existing individual with his history and constitution; however, “really possible” need not mean “indifferently possible,” i.e., equally likely. The whole mystery of probability lies here, in this weighting of possibilities. Popper, who is not a determinist, speaks of “propensities,” and perhaps (I do not know) he would admit degrees of strength in these propensities.21 He says propensities are something like generalized forces, and these certainly vary in strength.
The age-old experiment of absolutizing or “deifying” causal law was an aspect of another deification, that of deductive reason. Surely the derivation of necessary consequences is not the absolute function; this function is rather the production of valuable novelty, of real additions to the definiteness of reality. Not “reason,” in the mathematical or predictive sense, but “creative love” (a theme for another occasion) is the symbol of supreme power. If the evil in the world seems to contradict this, it is perhaps because we forget that were there but one creative individual, its love must take the degenerate form of mere self-love, and since there are many individuals, then because plural freedom cannot be ordered (no matter by whom) save approximately and statistically, a certain element of disorder and hence of conflict is to be expected, even assuming a theistic interpretation. Absolute determinism and a certain alleged notion of “omnipotence” are of one family of doctrines; what a relief it might be if we could rid ourselves of both, as distortions which destroy the meanings of the terms they attempt to expound!
But let us end on a more positive note. The relativity and statistical character of the world order do nothing to diminish its majesty. On the contrary, only when viewed as a power inspiring, yet not individually determining, countless acts of partial self-determination, wondrously coordinated to make a coherent world in which frustration and confusion, though real, are secondary, while fulfillment and harmony are primary, can we adequately appreciate the “grandeur of reason incarnate” of which Einstein so nobly speaks.22 Those who, like Cassirer, emphasize the mathematical definiteness of current science, in protest against loose talk about the indefinitely chaotic world sometimes supposed to be implied by it, are well-justified. Yet we should not forget that wholly exact regularities are known only where very numerous (Bernouilli’s “Law of High Numbers”) and closely similar cases of process occur; that, further, the higher the level of being, the fewer are the individuals (in general) and the more pronounced the individual differences; and that, finally, on the highest level known to us (apart from deity), that of human beings, a single individual can, through symbolic means, communicate to others something of what is unique to himself, not merely in details but in basic ideas and attitudes, and thus infect multitudes with his own creative novelty. Solely by keeping all these considerations in mind can we see steadily and as a whole the mixture of order and disorder which is reality. Thomas Jefferson said that without God the world would be a “shapeless chaos.” The relative indeterminist or creationist only wishes to add: with God it must still be chaotic, but how wonderfully shapely a chaos!
Notes
1. An exception is Charles Stevenson, “Ethical Judgments and Avoidability,” see Mind, XLVII (1938), 45-57, reprinted in Readings in Ethical Theory, edited by Sellars and Hospers (New York: Appleton-Century-Crofts, 1952), especially p. 555. However, Stevenson cleverly, and perversely, tosses the burden of proof to the indeterminist, that is, the relativist.
2. To show that I am not thinking of straw men, or even of mediocre men, let us cite an illustrious contemporary example, Bertrand Russell. See Section IV of his brilliant “Elements of Ethics,” in Philosophical Essays (London, New York: Longman’s, Green and Co., 1910), reprinted in Readings in Ethical Theory, pp. 17-23. The distinction between relative and absolute determinism is missed, and the argument proceeds as though one had two absolutes to choose between, one of which no one could defend. (Russell would be somewhat more careful today.)
3. However, see J. C. Eccles, The Neurophysical Basis of Mind (Oxford: Clarendon Press, 1953), pp. 278-279.
4. See the essay quoted in The Life of James Clerk Maxwell by Lewis Campbell and W. Garnett (London: Macmillan and Co., 1882), pp. 434- 444, especially 438, 441, 444. That even classical physics was not rigorously deterministic has been held by high authorities, e.g., by Karl Popper, in “Indeterminism in Quantum Physics and in Classical Physics,” British Journal for the Philosophy of Science, I (1950), 117-133, 175-195; also by Hans Reichenbach, The Direction of Time (Berkeley; The University of California Press, 1956), p. 95. See, too, H. Feigl’s characteristically lucid account in “Notes on Causality,” in Readings in Philosophy of Science, edited by Feigl and Brodbeck (New York: Appleton-Century-Crofts, 1953), pp. 408-418, especially pp. 411-412.
5. That this difference is fully compatible with what we know of neural action is clearly explained by Eccles, op. cit., pp. 271-278.
6. See “Causal Necessities: An Alternative to Hume,” Philosophical Review, LXIII (1954), 479-99, especially 485-489.
7. W. H. Watson, On Understanding Physics (London: Cambridge University Press, 1938), pp. 68-74. For an interesting comparison, showing, I think, that changes in intellectual fashions may be less than pure gain, contrast Watson’s analysis of necessity with that given by Bernardino Varisco, in his neglected but brilliant and powerful book, Great Problems, trans. by Salvadori (London, 1915), especially pp. 158-175. Varisco seems to me to have gone fairly far toward neoclassical metaphysics, without quite completing the transition from the metaphysics of being.
8. Ernst Cassirer, Determinism and Indeterminism in Modern Physics (New Haven: Yale University Press, 1956).
9. Ibid., pp. 189-195, especially 190.
10. See the splendid account, in which this passage is cited, by Oskar Becker in Die Grösse und Grenze der Mathematischen Denkweise (Freiburg/München: Verlag Karl Alber, 1959); also Norwood Russell Hanson, “The Copenhagen Interpretation of Quantum Theory,” The Philosophy of Science, edited by Danto and Morgenbesser, pp. 450-70. The sentence of Heisenberg is to be found in Daedalus, Vol. 87 (1958), No. 3, p. 100.
11. B. F. Skinner, Science and Human Behavior (New York: The Macmillan Company, 1953), p. 17.
12. Cassirer, op. cit., pp. 118-119. W. H. Watson similarly neglects the fact of unpredictable events, burying it under subtle talk about language and symbolic levels. (Op. cit., pp. 79-83). He has no notion of creativity as the pregnant meaning of freedom nor of the problem of objective chance, which in the end is unavoidable, since, in spite of Spinoza, the totality of facts cannot follow from any necessary premise, so that chance either comes into the world in a single cosmic throw of the dice, or piecemeal. For an admirably clear exposition of the basic role of chance in physics see W. G. Pollard, Chance and Providence (New York: Scribner’s, 1958).
13. Cassirer, op. cit., 204
14. Op. cit., pp. 197-198, 205-206.
15. Pp. 75-79.
16. Pp. 63-64.
17. See A. N. Prior, Time and Modality (London: Oxford University Press, 1957) R. Taylor, “The Problem of Future Contingencies,” Philosophical Review, LXVI (1957), 1-28. The best brief analysis of this subject that I know of is Paul Weiss’s “The Semantics of Truth Today and Tomorrow,” Philos. Studies, IX (1958), 21-23. See further P. Wolff, “Truth, Futurity, and Contingency,” Mind, LXIX (1960), 398-402.
18. Cassirer, op. cit., p. 208.
19. See my “Strict and Genetic Identity,” Structure, Method and Meaning: Essays in Honor of Henry M. Sheffer (New York: The Liberal Arts Press, 1951), pp. 24-254.
20. Aloys Wenzl, Die Philosophischen Grenzfragen der Modernen Naturwissenschaft (Stuttgart: Kohlhammer, 1954), p. 104.
21. Karl Popper, “Philosophy of Science: a Personal Report,” British Philosophy in the Mid-century, A. C. Mace, editor (New York: The Macmillan Company, 1957), pp. 155-191, especially 188.
22. Science, Philosophy and Religion: A Symposium, edited by L. Bryson and L. Finkelstein (New York: Conference on Science, Philosophy and Religion in Their Relation to the Democratic Way of Life, 1941), p. 214.
Source:
Charles Hartshorne, The Logic of Perfection, pp. 161-190.
HyC