return to religion-online

Immediate and Mediate Memory

by Milic Capek

Milic Capek is Professor of Philosophy Emeritus of Boston University. The following article appeared in Process Studies, pp. 90-96, Vol. 7, Number 2, Summer, 1977. 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.

In his fair and generally sympathetic review of my book Bergson and Modern Physics, David Sipfle raised some important and significant questions which clearly show how extremely complex the questions concerning the nature of time are and how difficult it is to agree on their solutions even for those who share a basic philosophical view. The question which he raised is as follows: "What I ask Capek is further explanation of the distinction between my immediate memory of the earlier portions of my specious present and my memory of my former present" (PS 2:311). This is, indeed, a fundamental question. That there is a difference between these two kinds of memory is beyond doubt, and neither Bergson nor Whitehead denies it. All that Bergson asserts -- and we shall see that in many of his texts Whitehead implicitly adopts the same view -- is that this difference, real as it is, is nevertheless only that of degree, not that of nature, for one simple reason: that the past is always totally immanent in the present, even though in different degrees of vividness. More specifically, the integral immanence of the past in the present is differentiated in the following way: the more remote a certain phase of the past is, the less vividly its influence within the present is felt. But, once more, the totality of the past is present -- not as a homogeneous bloc, but in the form of qualitative continuum of different degrees of vividness.

The assertion that the totality of the past is immanent in the present follows as an inescapable consequence from the immortality of the past. The past is immortal by its own nature: what has happened, happened, and its happening cannot be unmade; it cannot be erased from the total texture of existence: as St. Thomas (following in this respect Aristotle) observed, "However, for the past not to have been involves a contradiction" (S.T., q. 24, a. 4). This means that the past, precisely in virtue of its indestructibility, cannot be a mere nothingness, a mere existential void; in this respect it differs from the future; while it does not exist in the same sense as the present, it still subsists, and by its own everlasting subsistence it contrasts with a sheer nonexistence of the future. The very nature of time, that is, its asymmetry, is based on this contrast. Only in this way is the mnemic influence of the past, that is, memory, in its most general as well as in a psychological sense, possible. If the past were a mere nothingness, the whole distinction between the faithful and false memory, between history and myth would collapse.1

But this causal influence of the past (i.e., its immanence in the present) is diversified: the more remote the past event is, the less intense its effect in the present is. Unlike the indestructibility of the past which follows logically from the law of contradiction, this "fading of the past" is the result of empirical observation. In our direct introspective experience it is perceived as a gradual fading of the present moment into its own recollection; every sensation and feeling gradually acquires the character of the pastness even when its sensory vivacity remains quite pronounced; this character of the pastness is correlated with the emergence of a new present which, metaphorically speaking, pushes its predecessor pastwards. (Of course, we must be careful not to confuse "recollection" and image : the true recollection has always the character of the pastness, and it is usually "dated" [i.e., "located"] in the past, while the image is not. Phenomenologists are correct in stressing this important difference.) In the physical sciences the diminishing impact of the remote past is well known; the farther an event is in the past, the lesser is its action on my Here-Now.

Prior to Einstein this was regarded as a consequence of the inverse square law, that is, as an effect of the increasing spatial distance; today when we know that there are no mere spatial distances, but only spatiotemporal distances, we also know that to be remote in space from any particular Here-Now means also to be also distant in time, i.e., to be in the past with respect to that particular Here-Now. It would be otiose to give examples: a distant thunder is in the past as much as a distant star; but no matter how far in time-space a star or galaxy is, it is always faintly immanent in my Here-Now even when its action is below the threshold of human perception; its action can be made visible by a combination of lenses or a prolonged photographic exposure. Thus the diversification of the mnemic influence of the past according to its remoteness seems to be a general law. (Perhaps with our deepened insight into the nature of time even this empirical law can be justified in an a priori fashion; but I prefer not to go into this problem now.)

Now how is all this related to the question raised by Professor Sipfle? The answer is obvious: if the process of "fading of the past" is continuous, there cannot be any radical difference between the immediate and mediate memory. The apparently paradoxical character of this assertion will disappear after the apparently most serious objections are analyzed.

Let us first analyze the very language in which Sipfle formulates his question. He speaks of the immediate memory of "the earlier portions" of my "specious" present. Is it legitimate to speak of "the portions" of our psychological present? Is it possible to divide the present moment into two successive subintervals one of which is "immediate recollection" and another is "pure present"? There is a great temptation to do so because we are naturally inclined to symbolize any time-interval by a geometrical segment and, of course, every segment can be bisected. But such geometrical symbolizations of time are the source of endless confusions and vicious distortions which from the time of Zeno of Elea have interfered with our true insight into the nature of time. Bergson’s great merit, still not sufficiently appreciated and understood, was to show how seriously "the fallacy of spatialization" prevents us from grasping the authentic nature of time. On this point Whitehead explicitly agreed with him in several places, although he did not believe that the spatializing tendency is inevitable (SMW 74, 212; PR 126, 174, 319). (Neither did Bergson -- otherwise his criticism of this tendency would be pointless.)

Let me mention at least three instances of distorting spatialization. First, to symbolize time by a geometical line tends to obscure the dynamic and essentially incomplete character of temporal process. A geometrical line consists of simultaneous, juxtaposed parts; it is essentially static. Second, any geometrical line is continuous in a mathematical sense, i.e., infinitely divisible, which is certainly not true of any concrete temporal process: there are indivisible minima of duration not only in psychology (the so called "specious present"), but in the light of quantum phenomena, even in physics. The third instance of vicious spatialization is merely another form of the previous one: the belief that every temporal process consists of a dense succession of durationless instants in the same sense as a geometrical line consists of dense continuum of dimensionless points. But there are no "infinitely thin" instants either in psychology or in physics.

For the same reason, the atomistic theory of time is equally inadequate. Time consists neither of durationless instants nor of sharply delimited segments; either of these views is inspired by a false analogy between time and a geometrical line. It is ironical that "the atom of time," conceived of as antithetical to the concept of durationless instant, really presupposes the very concept which it purportedly negates; for what else but the durationless instants can constitute the boundaries of a sharply delimited temporal "atom"? We must not be confused by Whitehead’s atomistic language as some of his disciples apparently are; the so-called "epochal theory of time" is atomistic only in name. We have seen that the concept of "atom of time" requires that of instant -- the very same concept which Whitehead always rejected as early as in 19.19 (PNK 2f, 6-8; SMW 54, 172).2 Both concepts -- "atom of time" and "instant" -- presuppose the notion of simple location in time which Whitehead denounced as the most dangerous fallacy (SMW 84f, 98, 132).3 His whole doctrine of prehensions is incompatible with the doctrine of external relations which the atomization of time implies.

From this point of view it is futile to look for instants anywhere since they simply do not exist. Instants are merely virtual stops in the stream of our experience which goes on without ever being arrested. Consequently, there is neither the initial nor the final instant of the present moment; nor is there any instant bisecting it. Only when we begin to map the temporal progress geometrically on the axis of abscissa do we start looking for such fictitious entities. But then we are dealing not with time, but with its static and inadequate symbolization.

Now let us return to Professor Sipfle’s question: "What is the distinction between my immediate memory of the earlier portions of my specious present and my memory of my former present?" If by "former present" he means my "immediately preceding present" then the only possible answer is: none. For the former present acquires its character of immediate pastness or anteriority in virtue of its being immediately remembered; the content of immediate memory is constituted by the former present which has just acquired the character of immediate recency. These two terms refer to the same phase of the duration. But, of course, this phase of duration is qualitatively different from the previous phase when the former present was not yet "former" and when the new present was still nonexistent. But this relation of succession cannot be symbolized by a single static diagram because any diagram consists of simultaneous parts and thus cannot convey adequately what is non-simultaneous.

If, however, by "former present" Sipfle means a moment which is not immediately preceding, then the question becomes different. Then, by definition, we have two distinct degrees of pastness which are qualitatively different, two recollections, one of which refers to the immediately preceding past, another to a more "distant" past. (We do not have in our language a comparative for the adjective "past’ -- something like "paster" -- thus our language again has to use a metaphor of spatial distance: "more remote in the past.") Needless to say, our experience is not restricted to the memory of two successive phases only. In the perception of memory there is the mnemic awareness of the multiplicity of successive phases. This is what Husserl calls "temporal horizon." But we have again to be very much on guard against the misleading terminology. This is not the arithmetical multiplicity of externally related and mutually exclusive units.

Logicians will probably never like the Bergsonian terms "qualitative multiplicity" or "heterogeneous continuity" with the possible exception of Brouwer and mathematical intuitionists in general whose concept of "one-manyness" points in the same direction (cf. BMP 150, 180). As James stressed long ago: "The experience is from the outset a synthetic datum, not a simple one; and to a sensible perception its elements are inseparable, although attention looking back may easily decompose the experience and distinguish its beginning from its end" (PP, I 610). He might have added that this distinguishing would be impossible if the original datum were completely homogeneous, that is, if it were not qualitatively diversified (not divided!) into successive phases (not "parts"!). Nobody will accuse Whitehead of irresponsible Hegelianism when he speaks of "self-diversity of actual occasions"; by this term he obviously tries to express the synthesis of unity and diversity, of continuity and difference, which constitutes the very nature of succession.

One question naturally comes to our mind: why is the temporal horizon limited? If the whole past is immanent in the present, why do we not remember our whole personal past? The temporal span of the immediate memory is clearly limited and it obviously varies according to the degree of wakefulness and attention. But again we would be looking in vain for any precise point-like limit; this is why James speaks of the "vaguely vanishing backward fringe of the specious present." What is temporally beyond its "rearward fringe" is the imageless feeling of our whole personal past, what James in the later period of his thought identified as the feeling of our "full self": "the whole is somehow felt as one pulse of our life, -- not conceived so, but felt so."4

The amazing facts of hypermnesia, whether spontaneous or artificially induced, indicate that no phase of our past is ever destroyed. An impressive number of testimonies suggest that in some instances of extreme danger the totality or quasi-totality of our past is glimpsed "at once, more accurately, in a present moment which is contemporary with a very short interval of public time. Such "panoramic vision of the past" (la vision panoramique du passé) had been discussed extensively at the turn of the century; Bergson took notice of it in Matter and Memory, and the same problem was discussed more recently by George Poulet, the author of Études sur le temps humain.5

Thus Professor Sipfle is quite correct when he writes:

If what makes an isolated unit of time impossible is the presence of the past and if the remote past is present in the same sense that the immediate past is present (differing only in the degree of influence), what is to keep the enduring present from expanding not merely to 12 seconds, but to a lifetime, and beyond that to all time? (PS 2:311)

This is, indeed, how Josiah Royce and Henri Bergson defined the divine consciousness, trying, both in different ways determined by their different philosophical outlooks, to do justice to both time and eternity.6 Whitehead was moving in the same direction when he insisted that in God novelty is present without the fading of the past, without "perpetual perishing" (PR 134, 517, 524f).

Yet the same question still remains: what in a normal human consciousness prevents the totality of the past from being integrally present? The only answer which is plausible can be given by the biological theory of knowledge: in the same way as our perception carves Out of the whole physical reality only that zone which has practical importance for our organism, only those recollections which are relevant to our present situation are transmitted into our present moment. The limiting, selecting factor is our "attention to life," embodied in our normally functioning nervous system. Since, as Whitehead observed, "the relevance of the remote past is practically negligible" (PR 58f),7 it usually remains outside of our immediate memory, subsisting mainly in the form of imageless, virtual awareness.

It is more or less clear how our perceptual organs operate the selection of "the zone of the middle dimensions" from the physical universe; but any attempt to explain the mode of selecting action of our brains from the totality of our pasts would involve us in a lengthy, complex, and untraditional discussion of the mind-body problem which is clearly beyond the scope of this short article. Let me only say that the problems touched upon in this article simply do not exist for any present naturalistic or even idealistic doctrine; for materialism, epiphenomenalism, double-aspect theory, or even idealism, only the present exists, and what we call the past is nothing but a less vivid present "trace" which is interpreted -- sometimes correctly, sometimes not -- as "belonging to the past." In this respect the influence of Hume’s claim that the recollection is nothing but a weakened perception unfortunately still persists.


Let me then restate my answer to Professor Sipfle as concisely as possible:

(a) Durationless instants are mere fictions; consequently, it is otiose to look for any instantaneous boundaries of the present moments or for any instant bisecting these moments.

(b) Two apparently rival theories -- that of mathematically continuous time as well as that of the atomistic time -- presuppose the existence of durationless instants and thus cannot be adequate representations of our experience of time.

(c) The absence of instantaneous cuts in any temporal process does not exclude qualitative diversity of successive phases. "To be qualitatively different" and "to be sharply separated by instantaneous cuts" is not the same.

(d) Novelty of the present is possible only on the background of the immediate memory of its antecedent past.

(e) Immortality of the past is the basis of memory in general, whether immediate or mediate. These two kinds of memory do differ, but in degree rather than in nature; a more detailed analysis of these differences can be done only within the framework of a comprehensive theory of the mind-body relation.



BMP -- Milic Capek, Bergson and Modern Physics. Dordrecht: D. Reidel, 1971.

PP -- William James, The Principles of Psychology. Two volumes. New York: Henry Holt & Co., 1890.



1 I have dealt with this topic extensively in two articles: "The Elusive Nature of the Past" (in Experience Existence and the Good, Essays in Honor of Paul Weiss, ed. by Irwin C. Lieb, Southern Illinois University Press, 1961, pp. 126-41) and "Memini Ergo Fui?" (in Memorias del XIII Congresso Internacional de Filosofia 5 [Mexico. 1963], 415-26).

2 See also my article "The Fiction of Instants," Studium Generale, 24 (1971), 31-43.

3 See also Whitehead’s criticism of durationless instants in CN 72f. The fallacy of simple location although not yet named so, is criticized in the same book (CN 145f.) Professor Sipfle himself established that the contrast between Bergson’s and Whitehead’s theory of duration has been greatly exaggerated: "Henry Bergson and The Epochal Theory of Time." in Bergson and the Evolution of Physics. ed. by P. A. Y. Gunter (The University of Tennessee Press, 1971), pp. 275-94.

4 A Pluralistic Universe (New York: Longman’s, Green Co., 1947), p. 289. On this point see my article "The Reappearance of the Self in the Last Philosophy of William James," The Philosophical Review 62 (1953). 526-44, esp. p. 539f. Compare with James’s view, quoted above, the following passage of Charles Hartshorne: "If it be asked how the individual can be aware of this infinite range if his experience is finite, the answer is that it is only the distinct or fully conscious aspect of human experience which is finite; while the faint, slightly conscious background embraces all past time" (Beyond Humanism. Essays in the Philosophy of Nature [Chicago: Willett, Clark & Company, 1937]. p. 122; italics mine).

5C. Poulet, "Bergson et la vision panoramique des mourants," Revue de theologie et de philosophie 10 (1960). 23-41.

6 Cf. my article "Time and Eternity in Royce and Bergson," Revue Internationale de Philosophie, No. 79 (1966), 22-45.

7 This is, so far as I know, the only passage where Whitehead comes close to a biological theory of knowledge.

Viewed 9608 times.