Philosophy and Classical Determinism
by Milic Capec and Brenton Stearns
Milic Capek is Professor Emeritus of Philosophy, Boston University. J. Brenton Stearns teaches philosophy at the University of Winnipeg, Winnipeg, Manitoba, Canada . The following article appeared in Process Studies, pp. 190-198, Vol. 11, Number 3, Fall, 1981. Process Studies is published quarterly by the Center for Process Studies, 1325 N. College Ave., Claremont, CA 91711. Used by permission. This material was prepared for Religion Online by Ted and Winnie Brock.
Capek: Their Incompatibility
In his article "Becoming: A Problem for Determinists?" (PS 6:237-48) J. Brenton Stearns raised some objections against the basic thesis of contemporary process philosophy according to which strict determinism, when consistently thought through, implies a complete elimination of becoming or, at best, a relegation of succession to the realm of appearances. The latter view, the so-called "mind-dependence of becoming," proposed by Donald Williams, A. Grünbaum, and J. J. Smart, is rejected by me; but I also reject the view that there is any incompatibility between the classical form of determinism (i.e., the strict predeterminination of the future) and the reality of becoming and/or succession. Stearnsís view is so much more interesting because he restates the process view rather accurately; he even seems to be attracted to it. He agrees with Bergson that any kind of spatialization of time eliminates effectively time itself, i.e., genuine succession. He writes explicitly: "I shall assume that the process thinkers are right in holding that an adequate metaphysics of time must recognize dynamic becoming in nature as well as in human consciousness. The question then is whether time as the process philosophers think of it is incompatible with causal determinism" (PS 6:238). This is precisely what he questions.
What are his arguments for the compatibility of classical determinism and genuine succession? He points out that "the law-like statements contain within them temporal succession, the concept of events taking place at different times." In other words, "the laws that sustain the predictions make use of the concept of temporal successioní (PS 6:239).
He is certainly right on this point; nothing is more familiar than the notion of causally necessitated process gradually unfolding itself through time. But mere familiarity is not enough, for it does not answer the fundamental question: if the causal relation is identified with the relation of logical implication in which the premises imply the conclusion, is it possible to obtain any temporal sequence at all without surreptitiously borrowing it from experience? If the future is completely and in all its details determined now, what prevents it from being present? For the conclusion is contained, or rather pre-contained, in the premises. This is especially obvious in the classical syllogism when we symbolize the mode "barbara" by Vennís diagrams: All S are M, all M are P; therefore, all S are P. It is obvious that the inclusion of the class S into the class P is simultaneous with two "previous" inclusions, that of S into M and M into P. There is not a trace of any succession; there is no temporal sequence at all -- except on the psychological level, in the mind of a student groping for the conclusion, not in the logical structure itself. This is a rather elementary example; but the same is true of any deductive reasoning, no matter how complex.
When a student writes a left side of the equation while leaving its right side blank, he knows very well that this blankness is only in his mind and that the result which he will obtain by a series of successive mental operations is as definite as the expression given on the left side. Unfortunately the expression "it follows" has two different meanings in nearly all languages. We say that the effect follows its cause and also that the conclusion follows from its premises; but while in the first case there is a genuine temporal sequence, there is no succession whatever in the latter case; yet, the false identification of these two very different meanings is the source of unceasing confusions. It began with Spinoza when he claimed that from the nature of God the things follow with the same necessity as from the nature of triangle the theorem about the sum of its internal angles: thus the successive causal action is assimilated to a timeless implicative pattern. But if it is so, then the question which process philosophers ask is only natural: "If the effect is precontained in its cause and, more generally, the future in the present, why is it not already here? It should be present; but it is not; why?"
Stearns answers, "because it is not; experience tells us that it is not." But this is precisely the point. In order to smuggle succession into the classical causal relation, we must borrow it from experience; for it cannot be obtained from the conceptual model itself. Let me repeat what I wrote some time ago:
If the rigid form of determinism is supposed to account for all the features of any future event and at the same time fails to account for its most fundamental one: its "posteriority," its "futurity," its "not yet," then it certainly does not merit its name. Furthermore, if the same doctrine tries to incorporate into its own body the element which is ex definitione excluded, then it ceases to be logically coherent.
Steams is aware of this difficulty, but he believes to escape it by his distinction between "definite" and "actual." These terms, according to him, are not synonymous; in other words, the events can be fully determined without being actual. (Or, perhaps more accurately, the entities can be fully determined before becoming concrete events.) This view is hardly tenable, as we shall see; but even if we accept it for the sake of argument, it does not help a classical determinist a bit. For such a view leads to the most grotesque bifurcation of reality which is much worse than that criticized so convincingly by Whitehead: on one side, the realm of timelessly valid propositions, including those referring to future events, while on the other side the temporal realm of nature and mind in which the timeless propositions are being gradually embodied. The same unanswered question crops up again: why is this diaphanous realm of ghostly timelessly valid propositions only gradually and step by step embodied into the series of successive events? Again, no answer is possible; for to say that "experience tells us so" means to concede again the impotence of the necessitarian scheme from which the notion of "gradual realization" cannot be obtained. Indeed, experience tells us so; but the conceptual model itself does not! Furthermore, experience tells us literally nothing about the diaphanous realm of fully determined entities waiting, so to speak, for their gradual embodiment; they are invented by our two-value logic whose applicability to the future events was questioned already by Aristotle. For him propositions referring to future events are neither true nor false, because their truth-value remains undecided prior to their actualization. Thus the difficult notion of fully determined, though nonactual, event is avoided.
Can such a notion be meaningful at all? What differentiates such events from actual events? Should we say that they are merely "possible"? But this is what a determinist denies, for he insists that these events are timelessly necessary, the alternative possibilities being eternally excluded. It would be thus inconsistent to regard the future necessary events as unrealized possibilities. In truth, a consistent determinist must deny the realm of genuine possibilities; the only dichotomy which he admits (besides the bifurcation mentioned above) is that between necessity and impossibility which he equates with the dichotomy between reality and unreality. Whatever is necessary is real, and whatever is impossible is unreal -- and vice versa; therefore, the future events, being necessary, must be somehow real, and if they do not appear so, it is due to our ignorance. This was the view of all truly consistent necessitarians such as Spinoza and Laplace. Time is thus reduced, in Bergsonís words, to a human incapacity to know everything at once.
Stearns, like many other empirically minded determinists, wants to have the future both fully determined and unreal; hence his difficult notion of a completely determined but nonactual event. But what then differentiates future determined events from those equally determined in the present? Nothing at all; as Charles Hartshorne says: "An absolutely definite but merely possible entity is a contradiction; for if the thing is definitely all there, there is definitely nothing which it can lack, not even Ďexistenceí" (BH 228). Classical determinists can take a refuge in the view that existence is not a predicate, but a mere positing in Kantís sense. (Curiously, Stearns proposes the very opposite solution, but then finally gives it up.) Then such positing would provide a differentiating feature between two equally determined successive events -- one already existing, the other not yet in existence. But this would not explain why such positing is successive, i.e., taking place in time -- and we are back to the same unanswered question. Try as hard as you can, you cannot get rid of succession -- and succession means irreducible novelty, if it means anything at all.
Furthermore, the Kantian view presupposes the causal inefficacy of time, i.e., the view that time is merely a homogeneous container in which changes take place. Only thus would it be possible to have two qualitatively identical, i.e., fully determined, events whose only difference would be their "position in time." But this Newtonian view of time and space, as de jure empty homogeneous receptacles which are contingently filled up by concrete events, is given up by contemporary physics as I tried to show elsewhere; the space-time framework is now merged with its dynamical physical content. Moreover, the notion of the preexisting future events gradually entering into our consciousness has no physical basis whatever in the relativity theory, notwithstanding the repeated claims to the contrary.2
The indifference to contemporary physics is probably the most serious shortcoming of Stearnsís article. It is true that he discusses the problem of the relation of time to entropy; but after pointing out serious difficulties of the view which identifies "time direction" with entropy, he dismisses them with a simple declaration (PS 6:246). An even more important problem which he does not even mention at all concerns strict determinism as it appears in physics today. lie seems to be unaware of the growing cumulative evidence against it since 1927. Yet, to be aware of this evidence is the only way to confront what I would call argumentum ad familiaritatem, the argument based on familiarity, formulated by Ralph Barton Perry against Bergson long ago: what is more familiar to us than the experience of a physical process aging according to law? To use Perryís own illustration: a simple mechanical movement of a particle whose future positions are completely and accurately predictable allegedly contradicts the assertion that succession and rigid necessity are incompatible (PPT 251f.).
I analyzed the superficial plausibility of Perryís example before in pointing out that it is borrowed from the macroscopic realm which is for all practical purposes deterministic; the predictability of any macroscopic particle is only approximate and does not alter the basic contingencies of the elementary microphysical events of which a "particle" consists. Furthermore, on the microscopic level the very concept of localizable particle as well as the continuity of trajectory loses its meaning. Thus what appears to our macroscopic myopic perspective as a sharp and precise line is really a tube , although a very tenuous one; its transversal dimensions correspond to the quantic indeterminacies of its "positions." What appears to us as a predetermined "path of the future" is really the whole range of multiple possibilities which only in our inaccurate and simplifying macroscopic -- and macrochronic -- perspective shrinks to what it looks as an infinitely thin line of "the only possible future path." It is obvious that in dealing with subtle metaphysical questions, the "ordinary language," which is based exclusively on the limited macro-scopic experience, is thoroughly inadequate.
The previous analysis also takes care of Stearnsís criticism of the distinction between predictable and unpredictable aspects of reality (PS 6:243). This distinction is the very essence of statistical predictions according to which the predictability of the general trends is compatible with the unpredictability of individual events. While we can accurately anticipate that a future individual event will fall within a certain range of possible events, its complete specificity remains "undecided" until its own happening. The idea of complete specificity of a future individual event is, as Charles Hartshorne observed, a meaningless assemblage of words. Or, as Bergson said, it is a "retrospective illusion" based on the tendency of our imagination to project the actual present event into the past where it allegedly existed prior to its actualization as a "fully specified possibility" (!) (CM, part I).
This is not the place to discuss Stearnsís claim that if the future is unreal, so is the past. It only indicates his complete unawareness of what various process philosophers wrote about the status of the past, on the asymmetry of time, on the basic difference between "having existed" and "not yet existing," and on the intrinsic difference between memory and anticipation. Aristotle and St. Thomas were already clearly aware about the indestructibility of the past. It was such a nonmystical philosopher as Bertrand Russell who coined the term immortality of the past,"3 which Whitehead adopted and of which he made one of the cornerstones of his process philosophy. Moreover, the phenomenological analysis of our direct experience of time shows that the emergence of the present novelty and the retention of the past are two complementary aspects of one and the same process (BMP II.8). Not being aware of all this, Stearns wrongly assumes that the process philosopher advocates a completely miraculous indeterminism which would exclude any causal influence of the past. He concedes that he does not recall "A. N. Whiteheadís presenting the argument in the Bergson-Weiss form" (PS 6:243). Whiteheadís words may be different, but their meaning is the same:
The nature of any type of existence can only be explained in reference to its implication in creative activity. . . The alternative is the reduction of the universe to a barren tautological postulate, with a dream of life and motion. (MT 126f.)
If the universe be interpreted in terms of static actuality then potentiality vanishes. Everything is just what it is. Succession is mere appearance, being due to the limitations of our perception. (MT 136)
This is certainly not different from William Jamesís claim that determinism implies the static "block universe" nor from Bergsonsís identical claim that in the universe of Spinoza and Laplace "everything is given" ("tout est donné").
The resistance to the central thesis of process philosophy is due to the entrenched traditional habits of thought as they were shaped by our limited macroscopic experience, classical physics, and the powerful Eleatic component of our intellectual tradition; hence the obstinate belief that the only alternative to classical mechanistic determinism is absolute miraculous indeterminism, negating any kind of mnemic influence of the past on the present. Such view, needless to say, has not been upheld by any process philosopher.
BH -- Charles Hartshorne, Beyond Humanism. Lincoln: University of Nebraska Press, 3rd edition.
BMP -- Milic Capek, Bergson and Modern Physics. Dordrecht: Reidel, 1971. Boston Studies in the Philosophy of Science, VII.
CM -- Henri Bergson, Creative Mind. New York: Philosophical Library, 1946.
PPT -- Ralph B. Perry, Present Philosophical Tendencies. New York: Longmans, Green, 1916.
1"The Doctrine of Necessity Re-examined," The Review of Metaphysics, 5 (1951), 34; see also BMP 106-12.
2M. Capek, The Philosophical Impact of Contemporary Physics (New York: Van Nostrand, 1969), passim; "Relativity and the Status of Becoming," Foundations of Physics 5 (1975), 607-17.
3Bertrand Russell, Mysticism and Logic. (Baltimore: Pelican Books, 1953), p. 59. See also my essay on "The Elusive Nature of the Past" in Experience, Existence, and the Good: Essays in Honor of Paul Weiss, ed. by Irwin C. Lieb, (1961), pp.126-42.
Stearns: Their Compatibility
Milic Capek, eminent disciple of Bergson, supports my contention that indeterminism is the sine qua non of process metaphysics, or at least is thought to be so by the most able and inspiring process thinkers. We are told that if indeterminism is rejected the whole package must go. We would be left with the static bloc universe in which everything is given. I write not as a doctrinaire determinist but as a friend of process philosophy who thinks that many of the insights of process thought are logically independent of indeterminism and might well survive even if determinism should be true. I object to the idea that determinism is incompatible with a reflective belief in time as process, becoming, or genuine succession.
Yes, if causal determinism be true, then all true propositions about the future can be formulated as conclusions of deductive arguments in which the relevant causal laws and the initial conditions form the premises. Capek comments correctly that the conclusion "follows from" the premises in a timeless sense, that becoming cannot be derived from that deductive relationship. We must remember, however, that covering law explanations are abstractions from nature, and we must not expect every feature of nature to be derivable from them. We are familiar with the way in which temporal order can be given a spatial rendering with time lines and arrows. That is a similar abstraction from nature, and we cannot expect becoming to be derivable from it. We are familiar with the mistake of trying to force from these abstract models every feature of the world they are modeling. Because temporal order can be given a spatial mapping, we are led to think of a temporal dimension which is like a spatial dimension. We are encouraged further in this by our ability to transform observed spatial measurements into temporal measurements and vice versa as we seek to describe the facts as observable from various inertial frames of reference. We are thus tempted toward the "spatialization of time" in which time is regarded as having nothing but spatial characteristics.
I believe Capek would agree with me that mapping time spatially has its useful functions, but that this is an abstract rendering of time from which becoming cannot be derived. I maintain similarly that the deductive modeling of prediction is an abstraction from which there is no reason to expect becoming to be derivable. The fact that a timeless deductive model for rendering prediction cannot yield becoming must not be used as a proof for the thesis that the future cannot be deductively modeled in this fashion.
The Parmenidean mistake is one that Capek and I both seek to combat. Zeno asked that an infinitesimal analysis of space and time yield temporal becoming in its concreteness. But that was not possible, because the infinitesimal analysis came as the result of the application of the analytic intellect to space and time. The analysis did not capture the concreteness of what was being analyzed. Bergson himself exposed the Parmenidean mistake in this fashion in his treatment of Zenoís paradoxes. The mistake was to assume that the mathematical rendering of some aspects of the facts ought deductively to yield the facts. If we abandon that assumption in the case of the issue at hand, we do not have to give up determinism simply because the pattern of argument in which a determinate future can be modeled does not yield becoming.
Capek charges me with a surreptitious appeal to experience to establish becoming. This charge seems to concede that belief in becoming does have empirical foundations, and I should think that would be strong foundation. We can make appeal to experiences of planning, frustration, hope, fulfillment, promise, disappointment, etc. We have no instance of being deceived in our belief that these are appropriate categories of thought for interpreting our experience. Becoming is given in experience, yes, and then we try to give a rational account of it. But concepts are timeless, propositions are timeless, and the rational relation between propositions is timeless. That is why there is a puzzle. That is why Augustine thought he knew what time is but had difficulty in saying what it is.
Is the determinist guilty of a bifurcation between logic and nature? Indeterminists had better take care in laying charges like this, since they wish to make some propositions about the future exceptions to the law of the excluded middle. I do not recommend bifurcations, simply the modest admission that a true account of some aspects of nature need not be expected to yield all truths. Why are timelessly true propositions gradually embodied? I would not say that the propositions are being embodied, Simply that the states of affairs they describe are gradually coming about. Anyway, it sounds almost like a theological question. Maybe not everything God wanted to make could be made at once. Perhaps God wanted to make a world suitable for moral beings where there should be planning, regret, consequences of action, praise, blame, etc. Maybe God wanted a causally contingent world, and temporality is a necessary feature of causal contingency. In any event, experience tells us that there is gradual unfolding of history.
In a deterministic world, are all true propositions logically necessary? If stated tenselessly, all true propositions about matters of fact are timelessly true. The events they describe could logically have been otherwise, although causally they could be only what they are They could not have been otherwise given their causal antecedents, and since the causal antecedents could not causally have been otherwise, there never were any open possibilities for the state of affairs in question other than the one which does in fact occur. Since only x is causally possible, x is causally necessary. If that is what is meant by saying that events in a deterministic world are necessary, I have no objection. Propositions which describe them are nevertheless logical contingencies, propositions which can be denied without contradiction.
I do not see how lam committed to the view that time is causally inefficacious. The hypothesis that an event is a space-time configuration or distortion is, I should think, compatible with the eventís being predictable. And I certainly do not accuse indeterministic process thinkers of holding that there is no causality whatsoever in the world.
But the crux of the matter lies in the challenge which Capek offers. What distinguishes the future nonactual event from the same event actual and present if they are equally determinate? Existence cannot be the differentiating characteristic, since existence is not a determining property. True enough, there is no difference between a possibility when not actualized and the identical possibility when actualized. And yet we distinguish possible worlds according to which possibilities are actualized. Genuine succession can then consist in the transition from one possible world to another one.
There is no difference in determinacy between the present and the past, or between the near past and the remote past. Yet the past differs from the present in that it is not actual (except in memory). In my article I showed that by reasoning like Capekís we can show that there is no past even in the event that indeterminism is true. I will elaborate a bit more on this point. What are we to suppose that the past has lost since it was present? Is it existence that is lost? But existence is not a predicate, and so we must conclude that the past cannot have lost existence. The past event has exactly the same characteristics it had when it was present. Since the past is just as determinate as the present, it cannot have lost anything in becoming past. The argument we are examining requires us to say that the past cannot be past. We must say that whether we are determinists or indeterminists, since on both views the past maintains the determinacy of the present. Since Capekís line of reasoning cuts against the indeterminist as well it provides no ground for rejecting determinism.
Certainly there remains the problem of saying in what the asymmetrical temporal ordering of events consists. Process thinkers should find attractive the idea that it consists in the asymmetrical internalization of one event into another. The asymmetricality of the relative past and the relative future would then have to do with the distinction between memory and anticipation. That has nothing to do with an order of becoming determinate. In becoming what is vaguely anticipated becomes vaguely remembered. That is a clear alternative to becoming determinate.
Capek might reply that if a proposition about my remembering in 1995 an event that happened to me in 1990 is true in 1985, then there is nothing holding us back from saying that the remembering in 1995 exists in 1985. Aristotle maintained that all actuality is determinate; Capek says in addition that all determinates are actual. Then, since memory is by hypothesis determinate in advance, memory cannot become, grow, or accumulate.
I answer that anticipation and memory essentially involve tensed expressions. Can we say that we anticipate that the Yankees (timelessly) win the World Series in 1985? Perhaps, but part of the sense of "anticipate" is that the event is regarded as in the future. Can we say that we remember that John F. Kennedy (tenselessly) dies from an assassinís bullet? We would not be likely to speak that way, and in any case the sense of "remember" here contains the conviction that the event in question already happened. So in a deterministic world, "will be" expressions change from true to false, and "has already happened" expressions change from false to true. Process consists in that change and in the changes in nature which go with it. What then becomes? Memory becomes in the transition of facts from future to past. Memory must become; memory is not reconcilable with a bloc universe. It may not be determinate that my memory of my leaving home will be poignant and nostalgic, but the memory has not yet become since at the moment I still but vaguely anticipate my leaving home. Memory is compatible with determinism, and yet requires becoming, process, and genuine succession. That should be sufficient to prove that a determinist can be a process philosopher.