COEXISTENCE

AND THE FLOW OF TIME

Elie During
University of Paris Ouest Nanterre / Institut Universitaire de France

PREPRINT VERSION

Forthcoming in Japanese translation in S. Abiko, H. Fujita, Hisashi & Y. Hirai (eds), The
Anatomy of Matter and Memory: Bergson and Contemporary Theories of Perception, Mind
and Time, Tokyo, Shoshi Shinsui, 2016.


A plea for the philosophy of time
Let me start with a very blunt question. Why do we need a philosophy of time? What is it
good for? And by the way: which philosopher, among all those who cared to share their
views about time, can be said to have introduced a genuine philosophical concept of time
as opposed to a mere formalization or criticism of common sense intuitions regarding
temporal existence, the persistence of things, or the elusive experience of the moving
now? Plato? Saint Augustine? Kant? Heidegger? Whitehead? Bergson? McTaggart?
Michael Dummett? We all have our favorites. But it is one thing to criticize the folk
conceptions of time, to hold new and possibly counter-intuitive views about temporal
mattersarguing, for instance, that time does not pass, that there is no such thing as
universal time, or that the now is a matter of perspective, etc., it is quite another to
come up with a proper understanding of the function fulfilled by the concept of time in
philosophical discourse. If philosophizing about time is any worth, it is because we expect it
to shed some light on the real issues: for example, those concerning identity and
persistence, continuity and emergence. What would be the point of arguing about the
reality or ideality of time, if it did not change the way we relate to the world itself? Why
should we care about the meaning of passage, if it does not give us a better grip on the
underlying perplexities of becoming?
Those involved in the philosophy of time often speak of time as if it were some sort of Ur-
phenomenon in need of conceptual analysis. But the truth is that time does not lend itself to
scientific or philosophical inquiry as an objective natural property, or a bundle of such
properties universally attached to phenomena empirically given in time. Otherwise, there
wouldnt be much we could say a priori about time. I believe one of Kants great
achievements in that regard consisted in claiming with utmost vigor that time properly

1
understood is not an object in any meaningful sense, but something like a formal intuition,
or a second-order concept. My contention is that this insight is still relevant within the
framework defined by Bergsons particular brand of empiricism, if only to avoid raising
pseudo-problems in connection with his views on the true nature of duration.

Time and duration
Whitehead and Bergson deserve special mention in that respect, because both argued
along different lines, but acting on similar concernsthat it is essential to distinguish time
from duration. By introducing this semantic distinction, they were not merely registering
the obvious difference between lived and measured time; they were trying to make a
conceptual point.
Bergson in particular, far from equating time with duration, brought out the specificity of
time by explicitly contrasting it with lived duration. What was the motivation behind this
move? The underlying problem, as it turns out, is at once epistemological and metaphysical,
or more precisely cosmological, because it relates to the way we can bind together the local
flows attached to particular processes into a single whole, namely the particular object we
refer to as the universea universe that itself endures, as Bergson reminds us in the
opening pages of Creative Evolution. Now if all we had in store was the concept of duration,
there would be no way we could achieve this kind of totalization. Bergson, as is well known,
criticizes the illusion whereby time is collapsed with its relative measures, and consequently
looses touch with lived duration, which alone confers its distinctively temporal character
on time. It may seem that contrasting time with duration in that way comes down to
redefining real time in terms of duration, while leaving aside measured time as a mere
spatialized residue. But this is only half of the story. For once the criticism of spatialized time
has been carried out, it is equally important to understand how heterogeneous durations
unfolding together across the universe come to be woven into the fabric of the evolving
universe. And that is where time comes into play once again, not as another name for space,
or for some sort of hyper-cinematograph, but as the medium of coexistence that must be
recovered beneath spatial perspectives and measured durations. The crucial concept here is
that of simultaneity. Two distant events, separated in space, can nevertheless be
simultaneous because they occur at the same time, at least under some relevant
perspective. Simultaneity more generally refers to the complex phenomenon of being
together in time. Thus construed, it constitutes the core of the 1922 essay on Einsteins
theory of relativity, Duration and Simultaneity. And as Kant already acknowledged (against

2
Leibniz), there is more at stake with simultaneity than mere non-succession. What is needed
is a dynamicand hence genuinely temporalunderstanding of simultaneity. Seen in this
light, time itself cannot be reduced to the symptom of an incorrigible forgetfulness, of a
tendency to translate becoming according to extensive schemes. In fact, the real challenge is
to recover an adequate picture of the propension of duration to extend beyond the here-
nowthat is, not only beyond the now, but beyond the here as well. Commentators have
put so much emphasis on the link between time and measurement that this critical
involvement of duration with extensionnot necessarily with space considered as
quantitative multiplicityhas been more or less overshadowed by the tiresome textbook
opposition between clock time and conscious time.

Bergsons twofold understanding of time
Clearly, the criticism of time initiated in Time and Free Will is only the prelude to a
reconstruction of its concept along philosophical lines, a reconstruction that calls for an
active confrontation with scientific theories. But what is the critical function of duration in
that respect? In what way does the doctrine of duration positively contribute to a
reconstruction of time? Is it a mere placeholder for consciousness (a reminder that one
should not leave out so-called subjective time from the picture)? Let us forget the
somewhat sterile opposition of lived time and clock time for a second. Let us consider,
rather, what the concept (or image) of duration achieves in Bergsons overall philosophical
strategy. I believe the answer is rather straightforward: subjective time owes its importance
from the fact that it is the most immediate access we can find to a special kind of
multiplicity: a purely qualitative, non-serial multiplicity that is best characterized,
topologically, by the reciprocal penetration of its parts. More profoundly, as we shall see,
duration shifts the focus to what Deleuze described as the most fundamental operation of
time, namely: the true ontological split between present and past, actual and virtual. I shall
devote the first half of my lecture to this particular problem. This will be the occasion of
clarifying a few points regarding Bergsons singular position inand unlikely contribution
tothe current presentism/eternalism debate that rages in the analytic districts of the
philosophy of time.
From there, I will move on to the topic of coexistence, which I will deal with in a more
programmatic way. How do we make sense of the parallel unfolding of durations, not only in
experience, from the situated perspective of a living consciousness, but formally, on the
plane of thought? I surmise: by elaborating an adequate conception of simultaneity. The

3
chief reason why this should concern time is that contrary to time which by nature is one,
duration always comes in the plural. There is no Duration (capital D) to embrace all
durations. Even cosmic duration is in the end one duration among othersthe duration of
the particular object known as the universe. To repeat once again, it is the business of
timeor rather, the concept of timeto give a precise formulation to the cosmological
problem of the unification of durations, or what Bergson calls the unity of real time. This
concern is very much apparent in the last chapter of Matter and Memory, when it comes to
dealing with matter itself along temporal lines, as the most relaxed and extended state of
duration. This issue is also one of the main motivations behind Bergsons rather unsuccessful
confrontation with relativity theory. My intention is not to address this particular topic once
more, but to focus rather on the larger issue regarding the meaning and philosophical
relevance of time.

Whiteheads process time
I will not be lingering on Whitehead too much. After all, this is a celebration of Matter and
Memory. Nevertheless, Whitehead is a vital resource when it comes to assessing Bergsons
singularity in the philosophical landscape. Things may seem a bit muddled here because
Whitehead and Bergson do not take the time/duration polarity in the same way and may
seem at times to be talking at cross-purpose. More forcefully perhaps than any other
philosopher, Whitehead insisted that time extends in some sense beyond the spatio-
temporal continuum of nature (Science and the Modern World, 181). In other words, the
philosophical meaning of time runs deeper than that of a mere arena or dimension of
changewhich it certainly was in Kant, and still is to a large extent in Bergson. Because time
expresses the process of realization of the universe as a wholeextensive becoming, it
touches on an inherently non-dimensional, non-serial aspect of reality in the making.
William James put it nicely: concrete time, or change, comes in drops. What obscures this
fact is that we cannot help thinking of the process of realization itself as a continuous
phenomenon modeled after local motion, or the propagation of influences from place to
place. It is as if things had to be brought into existence in a continuous fashion, instant by
instant, within the medium of time conceived as a continuous flow. Accordingly, inverting
the Bergsonian idiom (see The Concept of Nature, 54), Whitehead claims that we must not
proceed to conceive time as another form of extensiveness. [] [T]he divisibility and
extensiveness is within the given duration. [] Temporisation is not another continuous

4
process. It is an atomic succession. Thus time is atomic (i.e. epochal) though what is
temporalised is divisible. (Science and the Modern World, 185)
One cannot help thinking that what Whitehead describes as time is very close to what
Bergson describes as duration, contrasting it with the extensive schemes of measured
time. The truth, as Bachelard suspected; is that Bergson may be more attached than he
would think to the geometrical intuitions of continuous motion. On the other hand, it is not
by accident that he insists on the continuous nature of duration. The qualification
heterogeneous (hetereogenous and continuous) is not enough to turn this continuity
into the kind of atomic time entertained by Whitehead: a process operating in bursts,
involving finite units or lapses of durations, and ultimately, a notion of intermittent
existence which cannot be easily reconciled with Bergsons overall view of nature. I do not
wish to develop this point further. It is enough to emphasize the more obvious point that the
doctrine of atomic time runs against a central tenet of Kants doctrine of time in the
Transcendental Aesthetics: the notion that time is an essentially continuous magnitude
just like space. This disagreement over continuity is an indication of the gap separating the
transcendental approach from the kind of speculative empiricism favored by Whitehead. But
it should not overshadow deeper commonalities. For the claim that time must not be
conflated with duration is typically one that Kant too would have endorsed.

The form of time: a detour through Kant
In fact, it is a recurrent theme of the first Critique. Time is not a concept but a form; not a
sensible datum but a pure intuition. We are familiar with these statements. Maybe less so
with their immediate consequence, namely that time does not pass, nor flow in any rigorous
sense. Time does not flow, while everything flows within it (A 144). Even the pure form of
succession is in itself immutable: time, Kant writes, is what remains and does not change
(B 225). No wonder that substance (or permanence), the first determination of time
according to the categories of relation, plays such a critical role in Kants doctrine of time. It
all comes down to this: time itselfthe form of time, time as formmust not be confused
with the temporal phenomena or processes unfolding in time. Time as a transcendental
condition of change is not a special kind of duration, only more general or abstract. It is in
fact nothing like duration. Its position in philosophical discourse is reminiscent of
Wittgensteins idea of formal concepts in the Tractatus. Even as we acknowledge that time
does not flow, we should not mean it in the sense that it stands still. What we should mean,
rather, is that time is not the kind of thing that may flow or not. (In the same way, to say

5
that time is not made of instants doesnt mean that it is made of finitely small stretches of
duration Let us keep that in mind when we consider Whiteheads doctrine of atomic
time). Time does not flow, not because a concept in general does not flow (just as the
concept of a dog does not bark), but simply because the concept of time is not a thing-
concept in the first place. Kant, for that matter, believes it is not a concept at all. At root, it is
a form, and in two respects: first, it is a formal intuition which is required to make sense of
the continuity and connectedness of experience; second, it is a concept that may come to be
elaborated and given objective meaning in relation with the postulates of empirical thought,
as well as very general connecting principles such as: permanence, succession, and
simultaneity, to name those listed in the analogies of experience, following the categories
of relation1.
To repeat, time cannot be treated along the same lines as temporal phenomena, or concrete
durations. Speculations regarding the passage of time have suffered much from this
confusion, and I believe there is a general lesson to be taken from Kant here. Despite the
sophisticated ontological frameworks designed to make sense of the elusive now, we are
still struggling with all too familiar but ultimately misleading metaphors such as the river of
time, the moving spotlight, the growing block and the like. Unpacking their underlying logical
structure does not render them any more adequate. The phenomenological approach to
time-consciousness has brought its own load of metaphors in connection with the so-called
stream of consciousness. And similarly, Bergsons intensive characterization of the
innumerable concrete durations making up the evolving universe has certainly reinforced
the feeling that talk about time or duration in general should be relinquished in face of the
multifarious diversity of actual change. This view, however, must be resisted. There being no
single concrete duration embracing all local durations does not entail the uselessness of the
concept of duration. Nor the uselessness of time, for that matter. Quite the contrary, the
time / duration polarity remains an essential tool for philosophical analysis. As we have
suggested earlier, it is the only way we can give a precise formulation to the cosmological
issue of the coexistence of concrete durations.
Accordingly, in Bergson no more than in Whitehead or Kant do we find such thing as a
flow of time (or duration) that may be speeded up or slowed down. Neither time nor
duration properly understood can possibly behave in that way. Physicists sometimes speak
loosely of the slowing down of time due to kinematic or gravitational factors or the like.

1
Kants suggestion is that these principles operate through the temporal schematization of the three
categories of relation: substance (permanence), causality (succession), community or reciprocal
action (simultaneity). The relevant section in the Critique of Pure Reason starts in A 176/B 218.

6
Since they cling to the metaphors of passage, the safest thing to say is still that time flows
equably (in Newtons parlance), at the uniform rate of one second per second. What is
sometimes described as a slowing down or dilation of time in relativity is only a fancy way of
expressing discrepancies in elapsed durations. Local measures of time keep desynchronizing;
there are delays and dephasings, as illustrated by the famous twin paradox in relativity
theory, because proper time turns out to be a local, path-dependent magnitude2. There
are shortcuts in space-time, if you prefer. But nothing like a mysterious time substance
that may stretch or contract by virtue of speed or acceleration. This manner of talking is
positively medieval and should be banned from popular introductions to relativity theory.
Nothing happens to time or duration as a result of relative speed or acceleration: it is only a
matter of reconciling various measures of time within a space-time framework, and more
importantly, to harmonize conflicting figures of time itself (such as parameter time and
coordinate time, local time and global time), while doing justice to the more general
underlying issues (persistence through time, coexistence, etc.).
As far as the philosophy of time is concerned, the morals of the situation is quite simple. To
repeat, time is not itself a process, but the form of any process. It is a formal concept. What
this means is that time is essentially the name of a problem, or a cluster of problems,
rather than a general denomination subsuming a class of particular objects (whether we call
these durations or changes). As a result, any particular figure of time calls for
philosophical interpretation, rather than direct application. Such an interpretation must be
carried out according to the demands that the form of time makes on us. Following Kants
characterization of the order of time in the analogies of experience, I would suggest that
the horizon of expectation associated with time points to three major directions:
1 Duration (permanence), according to the category of substance
2 Order (succession), according to the category of causality
3 Coexistence (simultaneity), according to the category of community (or reciprocal
action)
1 and 2 are typical ingredients of the view of time as the dimension of change. I need not
emphasize that point here. We have seen in what sense the notion of continuity traditionally
associated with that of spatial and temporal dimension clashes with the intuition of the
discrete nature of actual processes. But equally important is the third aspect of time,
described under the heading of coexistence: time as a sheaf or envelope of becoming,
extending over space, holding together bundles of durations that may flow together, overlap

2
More on this below.

7
or coexist at a distance, according to simultaneity relations that extend beyond the here-
now of coincident events. 2 and 3, taken together, raise the issue of times relation with
the causal order as a whole.
My contention is that we cannot choose between these three profiles of time: they are
inseparable, they come as a package, and that is what makes of time, more than a relational
concept, a genuine structure of thought. Time itself points at a bundle of related,
intertwining issues. If a philosophy of time does not manage to address all of them in some
way or other, it is not worth one hours effort, elle ne vaut pas une heure de peine, as
Blaise Pascal would say. But the real philosophical task consists in weaving the different
dimensions of time together, rather than reviewing them and analyzing them in turn only to
leave them in their original state of dispersion. Let this be a blueprint for any future
philosophy of time.

Beyond presentism and eternalism
Back to Bergson. The vocabulary may be different from Whiteheads, but the initial impulse
is very similar, as well as its bearing on the philosophy of time in general. I mentioned
several underlying issues so far: persistence (or identity through change), continuity,
emergence. Give up these metaphysical concerns and the whole discussion about the
passage of time, the debates between tensers and detensers, presentists and eternalists,
becomes a rather frivolous affair. As it turns out, the disagreement between the contenders
is not really about time, as much as it is about existence. As the debate is usually framed, it is
clear that the two parties mainly differ on the scope of the existential quantifier: presentists,
as they are often presented, hold that only what is present exists, while eternalists believe
that everything (past, present or future) exists (which, obviously, is not to be taken as:
exists now, unless now ranges over all possible times, past, present or future, to the
extent that they were, are or will be now).
The debate can be widely entertaining, if you are in the mood. It involves quirky questions,
such as: How do we know it is now now3? But is the discussion really about time? What
relevant aspect of time does it address, and to what end? Besides, if there is certainly much
to be learnt there about the grammar of exists, as well as about the way matters of tense
or persistence parallel those of identity across possible worlds, the issue of realization is
quite a different story. Realization does not reduce to the mere happening of things, to
the formality of taking place, as Eddington put it. If it were the case, the world as a whole

3
Or if you prefer: What epistemic grounds do we have for claiming that what we hold as our present
is not already past with respect to the absolute present singled out by the moving spotlight?

8
could be recapitulated within each moment, as if it were created anew: The world the
mathematician deals with is a world that dies and is reborn at every instantthe world
which Descartes was thinking of when he spoke about continuous creation (Creative
Evolution, 23-24).
Bergson has point when he suggests that metaphysicians are closer to mathematicians than
they would like to believe. Think of the way both presentists and eternalists deal with past
and present moments: from an ontological point of view, each moment, it seems, can be
considered apart from all others, as if it were alone, or the first. The only sense in which the
past can be said to continue in the present is by means of actual memories, retentions or
material traces4. All there is (now) is supposed to be given in actuality, laid out in full view,
including what remains from the pastanything, that is, but the past moments themselves,
which are irremediably gone. But the world of the mathematician is not the world we live in,
it is not a universe that endures, infused as it is with buzzing life. Nor is it exactly the
physicists world, by the way. For even matter, which Bergson defines as a present that is
always beginning again (Matter and Memory, 178), is only so if we suppose it to be entirely
extended in space, which Bergson acknowledges is only an ideal limit to which the real can
only strive. Even in the limited realm of matter, presentism remains a fiction: it has only
tendential validity, it can only be true at the limit, when we identifyby a leap of the
understandingmatter with instantaneous space.

Laplaces ghost
One is reminded here of Laplace, who derived his ideal of determinism from the assumption
that everything there will be can be strictly derived from what there is at the present
moment, as if it were somewhat compressed and encapsulated in it. In order to apply
deterministic laws, becoming needs to be spelled out in terms of successive instantaneous
nows. Picture these as ideally infinitesimally thin layers stringed together by external
relations of succession: each now appears entirely determined by its past, while at every
stage the past as a whole appears recapitulated or projected unto the immediately
preceding momentthat is, a moment that can be taken as close as we wish to the now

4
We are inclined to think of our past as inexistent, and philosophers encourage this natural
tendency in us. For them and for us the present alone exists by itself: if something of the past does
survive it can only be because of help given it by the present, because of some act of charity on the
part of the present, in shortto get away from metaphorby the intervention of a certain particular
function called memory, whose role is presumed to be to preserve certain parts of the past, for which
exception is made, by storing them away in a kind of box.This is a profound mistake! (The Creative
Mind, 177)

9
elected as temporal index. According to this conception, nothing would be changed if the
world were destroyed and created afresh at every moment, as Descartes imagined.
Suppose, however, that something of the past leaks into the present, and moreover, that
past and present differ in kind (in nature), just as the actualized image of past events differs
in kind (in nature) from its nebulous counterpart in pure memory. Then there is no reason to
think that the present does not contain more than what can be derived from the past: we
have made room for genuine creation. This is probably one of the most subtle views held in
Creative Evolution, but it is already at work in the discussion of the relation between present
and past in Matter and Memory. The same claim can be made from different perspectives.
In a way, Bergson argues that, irrespective of what is actually happening in time, the present
introduces genuine novelty by the mere fact of enduring. This implies of course that the
present is not strictly speaking instantaneous. Being itself a unit of change, it has some
breadth, it is thick, a fact well confirmed by experimental psychology. But more profoundly,
Bergsons claim is that the past continuously grows upon itself, forcing the present to pass.
There is no point denying that the present flows in some sense: the fact that it endures,
rather than being given in a flash, makes it less incomprehensible. However, the present
does not flow because it is in the nature of time to do so, it does not pass by virtue of
conforming to an impersonal and abstract template of becoming in general (Creative
Evolution, 324). Bergson sometimes explains that the present passes because it is infused
with the past that pushes against it. This is a striking image, but it is no more than that. In
order to spell it out more clearly, we would need to consider what is really involved in the
continuity of real change: not so much the mathematical continuity attached to continuous
magnitudes, but the notion that past and present overlap in an indivisible process, so that
the past is genuinely preserved as past within the present.
To put it in yet a different way, following Bergsons intensive description of the planes of
consciousness (which parallels those of reality), we may say that the present moment
contracts the whole past from its own, unsubstitutable perspective, as a result of the
interplay of memory and perception of that makes up the actual present. For a present
stripped from its concrete content is nothing but the empty form of awareness, or which
amounts to the same, the mere form of our activity or attention to life. In actual fact, to
perceive is to remember: the constitutive role of memory in the elaboration of perception is
a central claim of Matter and Memory. As a result of recapitulating the entire past from the
perspective of embodied, situated action, each concrete moment, each slab of experience
appears as new in itselfnew, and to that extent unpredictable, non-derivable from

10
previous moments, by the mere fact that some time has elapsed in between. New, not in
virtue of particular contents resulting from the combined operation of causal
determinations acting from outside, so to speak, nor because it is in the nature of things to
happen in the present, continuously adding new content to reality, but more
fundamentally, in virtue of the form of time itself. The present draws the whole past with
itself and comes to be tainted by it. That is why no two presents can be identical: even if
their objective contents were in fact indiscernible, they would not carry the same amount of
past! That is why there is necessarily more to the present than to the past. Very simply: in
the present moment, there is the present plus the past! There is perception plus memory,
contracting the past from a particular perspective 5. And finally, that is why the future
cannot be simply read off the present, as if nothing was genuinely created in between.
Hence determinism as a general doctrine collapses. Bergson argues that deterministic
schemes are at best approximations with only local validity: as a matter of fact, these
schemesthe laws of natureonly work under certain circumstances, from a particular
perspective that enables the scientist to carve out quasi-isolated material systems within the
universe as a whole. So much for Laplace.

In what sense does the past exist?
It is important to realize that in this plot, the elusive, transient now is by no means the
main character. It is at best a mirage. Both Bergson and Whitehead argued (as Heidegger
did, from a different perspective) that too much emphasis had been laid by philosophers and
psychologists alike on the present as the alleged locus of passage. An important step was
taken when people like William James derived the philosophical implications of the
specious present, reflecting on the fact that the lived experience of the present discloses
something quite different from a knife-edged instant. But the more audacious speculative
move consisted in acknowledging the difference in nature between the presentwhatever
its shape and breadthand the past as such.

5
Incidentally, this means that the notion of an absolute present, a present that would not be the
present of anything in particular, is immediately dubious from a Bergsonian standpoint. Referring to
the present without further qualification, as if it were an objective slab of reality, is not only
problematic because of Einsteins claims regarding the relative character of simultaneity relations; it is
intrinsically misguided in assuming a homogeneous, impersonal becoming made of successive layers
of now stacked up in continuous fashion. When referring to the present, one should always ask:
whose present? From which perspective? Keeping this in mind, the claim of a fundamental difference
in nature between present and past should be less unsettling. The past is not an objective domain
composed of all former presents that are now past: these presents, considered as universal planes of
coexistence, never existed in the first place, they are mere abstractions. There are only concrete, that
is situated, presents. The present is a local or regional concept. So is simultaneity.

11
The cinematographic rendering of becoming as a succession of moments coming into
actuality misses the point entirely, not only because it suggests a weird passage of time
itself, which turns it into some kind of hyper-process, but because it takes for granted that
present and past are fundamentally homogeneous. The problem with presentism and
eternalism is that they both suffer from a deeply engrained attachment to actualism: they
share the common presupposition that only what is real and actual can be said to exist; and
in this respect the ontological contents attributed to past and present are fundamentally
similar, despite their modal disparity. The past is merely a present that has passed. One is
reminded of the image discussed in Creative Evolution. Becoming, says Bergson, is made of
distinct states or stages, set side by side like the beads of a necklace. We may say that the
presentist draws our attention to the beads, which he reviews one at a time, while the
eternalist focuses on the solid thread that holds the beads together, in one piece. The
difference is immaterial, because they both believe in temporal beads.
We can put it differently. The presentist insists that the past does not exist. But what would
the past look like, if we made it to exist? The answer is obvious: it would look exactly like the
present that it was. So if the presentist were to say what it is that doesnt exist, his answer
would have to be this: what doesnt exist is exactly the same thing which the eternalist says
does exists, namely, the moments that were, the present that is now past. The difference
between past and present is purely modal.
To measure the shift in perspective introduced by Bergson, we must realize that the
question for him is not whether the past exists, but rather whether it is available. In other
words, the question is how it can be retrieved and whether it can be acted upon. The long
development about unconscious memories in chapter III of Matter and Memory (189 ff)
makes a convincing case: there is no serious reason to deny the continued existence of the
pastno more reason, at any rate, than to deny the existence of external objects beyond
the scope of actual perception. The past exists, in some real sense, below the threshold of
consciousness6. But what this really means is that it is not sufficient to describe the past as a
present that was and is no more: the past is not the present minus actuality; it is something
else altogether. And this difference in nature between past and present is the reason why
the past does not itself pass but conserves itself as a whole. It does so not in a deactivated
form, as a mere duplicate of past presents lacking only actuality, but in a transformed, form.

6
Elsewhere Bergson establishes the continued existence of the past in a more straightforward way by
simply emphasizing the continuity of lived duration: the preservation of the past in the present is
nothing else than the indivisibility of change (The Creative Mind, 182). But what remains to be
established is in what sense the past thus preserved can be said to exist.

12
Thus, the passing moment prolongs itself in the present, while growing by involution into
the virtual, building momentum as it stretches over multiple levels of consciousness or
reality, i.e., coexisting degrees of contraction and dilation in which it is recapitulated as a
whole. In order to get a feeling of the inherently heterogeneous nature of virtual memories,
of all the thresholds of consciousness one must cross to simply fold back one tiny facet of
our past experience onto the present field of awareness7, one only needs to focus on the
experience of having a name on the tip of ones tongue and providing the efforts required to
recall that name by extracting it from its nebulous state and condensing it into a conscious
memory-image (see Leffort intellectuel in Mind-Energy). Considered against the
penumbra of virtual becomings that borders each present moment like a halo, the past is
not a mere duplicate of the present subsisting in ideal form, it is not a neutralized present.
Quite the contrary, it is the present, as usually represented, which is the result of something
being subtracted or filtered out from the past. We think a new present has been substituted
for earlier presents that have become obsolete, when in fact the present is really the most
contracted degree of the past. One might as well say that it does not exist: artificially
separated from the past, reduced to an instantaneous cut in the flow of becoming, it lacks
almost any substantiality. All its substance, it owes to the past.
This idea can be given an even more radical twist by observing that the past is not primarily
what comes after the present. As explained in the eerie text on the memory of the
present, memory is truly simultaneous with perception, to the extent that it does not need
to wait to be formed8. This bold thesis confirms the description of the circles of memory in
Matter and Memory: the shortest, most contracted circuit between perception and memory
already implies such a contemporaneity between the present and its immediate past
(Matter and Memory, 127).
At this point you might want to object that we are not exactly dealing with the past here,
but merely with the experience of the past as memory or remembrance. Bergsons answer
to this worry is rather straightforward: what other access than memory do we have to the
past as such? For it is the proper business of philosophy to investigate the past as such: not
the actual traces of the past, surviving within the present, nor the foregone days projected
backwards on the timeline, expressed in language through the grammar of tenses, but the
essence of the past, its sheer pastness. Yet there is no such thing of the past in general,

7
See also Matter and Memory, 174 and 319.
8
I have commented this idea at length in the critical edition of The Memory of the Present and False
Recognition (Le Souvenir du prsent et la fausse reconnaissance, Paris, Presses Universitaires de
France, 2012).

13
there is no absolute past, anymore than there is an absolute present, if it is not referred to a
particular perspective capable of relating to it. Whatever happens at this present moment,
whatever happened at such and such time, those statements hardly make any sense
unless we view them, say, as short-hands for the physicists planes of simultaneity slicing
across the material universea very special perspective indeed. Again, there is not one past,
anymore than there is a unique duration encompassing all durations. There are rather many
pasts, as many as there are centers of experience in the universe. Thus, memory does not
primarily represent the past; it is the past that is, at root, an ontological memory.

What is it that passes?
All the preceding remarks point to the following idea: passage is an utterly inadequate
metaphor. The past does not itself pass; it grows upon itself, allowing for the present to
pass, and thus for real change to occur. But for the present to pass really means that it is.
Seen from this angle, everything falls into place. For the seemingly phenomenological
descriptions of duration, the peculiar topological nature of pure memory considered as
fuzzy or indistinct multiplicity, are merely expressions of a more fundamental fact: the
indivisible continuity of change, the continuation of the past in the present, in short their
paradoxical coexistence, more essential than their depiction as a succession of distinct
states, now-present and then-past. In that respect, it is no more illuminating to identify
Bergson as the proponent of some variety of the growing block theory of time, rather than
as a presentist or an A-theorist. From his perspective, it doesnt make much difference
which side you take. Once you have laid out the series of moments (infinitesimal instants or
temporal spans) on a single ontological plane, you have committed yourself to the view that
time is merely the dimension of changea dimension that only needs to be cut out or
punctuated in the appropriate way to yield all there is to be said about temporal matters.
And that, of course, is the original sin: For our duration is not merely one instant replacing
another; if it were, there would never be anything but the presentno prolonging of the
past into the actual, no evolution, no concrete duration. Duration is the continuous progress
of the past which gnaws into the future and which swells as it advances. (Creative
Evolution, 4-5) The continuous progress of the past: considering what has just been said,
this has clearly nothing to do with the growing block view of becoming, which is just another
variety of what Bergson criticizes in Creative Evolution as the cinematographic illusion
underlying the view of time and becoming as a global succession of instants or stages.

14
The cinematographic illusion reconsidered
For this is what is really at stake in the analogy: not the fragmented, discontinuous aspect of
the instantaneous film frames, but the global framing imposed by the common usage of
coordinate time. The problem is not so much the artificial, imitation of movement effected
by the juxtaposition of discontinuous stills, but the presupposition of a uniform flow
embracing a myriad of local flows unfolding across the world at different rhythms. This
global temporal frame is illustrated in the mechanism of the cinematograph by the
seemingly automatized run of the film strip at a fixed rate across a beam of light, indifferent
to the contents being projected9. Thus, the stroboscopic view of becoming as a series of
discontinuous instants (the conception of motion criticized in Bergsons discussion of Zenos
paradoxes) derives from a deeper presupposition regarding the serial nature of the time-
dimension. On a different level, the cosmological cinematograph dreamt by the
metaphysician and the mathematician achieves the feat of bringing together a plurality of
heterogeneous durations within a single global time. It does so by pitching them against a
homogenous dimension that renders them commensurate, and most importantly enables
them to share a single sense of simultaneity (at any rate, in the ideal Newtonian version). In
fact, as Bergson argues, this is made possible by a cinematographic mechanism of thought
performing the extraction of a single, simple, impersonal and abstract representation of
becoming in general out of the variety of concrete becomings (Creative Evolution, 324).
Such an indefinite, undetermined becoming is not the becoming of anything in particular: it
acts as the universal medium of change. I suggest we refer to it as frame-time.
Bergsons criticism of this inner cinematograph, which he sees already at work in an
unconscious way at the level of natural perception, should refrain us from identifying
Bergsonism to a late modern variety of Heracliteanism. Celebrating flow or becoming in
general is not an option: quite the contrary, it is explicitly criticized as an illusion! And
Bergson would no doubt have been greatly amused to find himself counted in the ranks of
the proponents of A-time The truth is that Bergsonism is primarily a philosophy of
durations (in the plural), of the coexistence of durations. It challenges us to find a way of
accounting for coexistence that does not rely on the abstract concept of simultaneity
introduced by frame-time. That is the real issue behind the cinematographic illusionnot

9
Interestingly enough, Bergson makes no reference whatsoever to the hand of the operator, at a
time where most projectors still involved manual handling. The kind of cinematographic projection he
has in mind does not involve any cranking device, nor any possibility of adapting the run of the wheels
to the inner rhythm of the filmed action. See my Notes on the Bergsonian Cinematograph, and Vie
et mort du cinmatographe : de Lvolution cratrice Dure et Simultanit.

15
the rehashed complaint lodged against the spatialization of time or the confusion of instants
with points on a geometric line.

Simultaneities: the twins
To sum up, everything I have been saying so far confirms the intuition that passage and
flow, far from being primitive, constitutive features of duration, only have secondary and
derivative status. In the conceptual framework set up by Bergson, durations main
contribution to time consists in shifting our attention to the third aspect or profile of time
identified by Kant, namely: coexistence. Simultaneity, then, rather than the succession of
present and past, but a simultaneity that takes its full meaning once we acknowledge their
difference in nature. For what Deleuze described as the most fundamental operation of
time is the unceasing production of a present contemporaneous with its own past,
conceived as virtual.
Thus, the philosophical concept of time does not fall with the cinematographic illusion. On
the contrary, it gains in precision. Bergsons analysis of the cinematographic mechanism of
thought lays bare a crucial problem which could even be properly formulated on the sole
basis of duration. It is the problem of the coexistence of multiple durations interlocking and
unfolding together in time! In order to simply see what is at stake here, it is necessary to
consider the various forms of extensive simultaneity that the concept of time can
accommodate. There is indeed no reason why philosophical reflection should confine the
meaning of simultaneity to the physicists concept of worldwide instants (what I referred to
earlier as planes of simultaneity). Fixing simultaneity relations between spacelike
separated events by means of appropriate reference frames (i.e., systems of coordinates) is
only one way to achieve distant simultaneity. The global temporal perspective involved in
the use of coordinate time does not exhaust the meaning of perspective. The spacetime
framework also allows for non-global, regional figures of simultaneity. Interestingly enough,
these figures turn out to be intrinsic, as opposed to merely relative. They can be read off
from the invariant topological or causal structure of spacetime, which agrees with
Whiteheads idea that perspectives are not superimposed upon becoming, or imported from
outside, but truly embedded in it, as real features of nature.
Therein lies the significance of the so-called twin paradox discussed in Duration and
Simultaneity, a paradox according to which two accelerated observers age differently along

16
their respective journeys10. Bergson spent much effort on distinguishing real time from
fictitious time within the very domain of measured time. That is one of the most original
insights of the book: real time is not another name for pure duration, it is a variety of
measured time. It is, as he writes, the time measured by a real clock attached to a portion
of matter. Now some philosophers, following Langevins suggestion, have identified real
time with the physicists proper time, a local magnitude measured along timelike curves
in spacetime11. They are right to the extent that real time is a variety of measured time,
which I must insist is the central issue here. Again, duration in the strict sense cannot be
measured, but there is such a thing as time, and time of course can be measured. If this is
not an entirely spatial affair, then surely duration must be involved at some level. Indeed, in
Bergsons view, it is only by virtue of its relation with lived duration that real time can play
the role of a connecting thread, achieving the unification of durations in spite of the
discrepancies in measured elapsed times. Such unification implies more than uncovering
relativistic invariants such as proper time or the space-time interval. It involves the
philosophical elaboration of a genuine concept of coexistence. Hence, referring to the twins,
Bergson writes: Not only do the multiple times of the relativity theory do not destroy the
unity of real time, but they even imply and uphold it. [] Without this unique and
experienced duration, without this time common to all mathematical times, what would it
mean to say that they are contemporary, that they are contained in the same interval? What
else could such an affirmation mean? (Duration and Simultaneity, 118) That is the deeper
issue behind Bergsons confrontation with relativity.
For the twins separate only to meet again, and surely it makes sense to say that while the
traveller was away, cruising in space, his brother got divorced and remarried. The twins
coexistin a temporal senseas they go about their business along separate
spatiotemporal paths. They may be aging differently, but they never cease to be
contemporary, as long as they both live. How can we represent the temporal togetherness
of independent proper times? Bergson claims there is a simultaneity of flows which is in

10
Paul Langevins original tale mentions one travelling observer, who on getting back to Earth turns
out to have aged less than everyone at home: the difference in the overall elapsed durations depends
on the degree to which the travelling observer is accelerated, as well as on the speed at which he
travels. The paradox stems from the apparent symmetry of the observers, who may each consider
themselves as rest while the other is moving. According to the principle of relative motion, it may
seem that they have equal right to claim that they have aged less. The details of this situation are
presented in almost every textbook introduction to special or general relativity. Remarkably enough,
the strategies adopted to present and explain away the paradox are almost as diverse as the ways of
telling the story.
11
Contrasting it with the frame-time of the cinematographic illusion, we may refer to this variety of
measured time as fiber-time.

17
fact more primitive than the simultaneity of instants. But what is the exact scope or range
of the twins perspectives on the becoming which carries them both? What is the locus of
the relational present that they seem to share despite the fact that their respective
durations are in some sense incommensurable? How far does coexistence extend? How
large is a moment of cosmic time? There is no univocal answer to these questions, because
coexistence itself comes in a plurality of modes or regimesa plurality that is not exhausted
by hyperplanes of simultaneity cutting across space-time at various angles.

Coexistence and contemporaneity
Bergson considers the time between the twins separation and their reunion (a time which
is measured differently by each) as an interval defining something like an envelope of
becoming. This can be given precise topological meaning in the spacetime framework,
provided that we do not forget that the disjoint spacetime paths of the twins remain
incommensurable as far as simultaneity is concerned, unless we introduce an inertial frame
somewhere in the picture. Even then, such a frame can yield no more than a relative and
arbitrary perspective on the overall simultaneity of unfolding durations. It is frame-time
once again. The best one can say is that a continuous one-to-one correspondence between
simultaneous events on both paths is available in some frames: thus, the Earth twin can
sweep along his brothers path, plotting the distant proper time of the traveller against his
own. But such a device is necessarily relative (to the choice of a particular reference frame:
his own), as well as disjunctive (because the resulting simultaneity relations are neither
symmetrical nor transitive). There is nothing absolute, nothing real in that sense, in the
kind of simultaneity achieved by the global perspective of frame-time.
This being said, for all its limitations, the regional simultaneity defined by the topological
envelope of becoming nicely illustrates the way a notion of simultaneity can be
accommodated to make sense of the relational coexistence of two observers sharing a
common history. The topological interval embraces a host of virtual durations that are
simultaneous in an intrinsic (i.e. non frame-relative) sense, simply because they are bounded
by the same pair of events (in the twins case: separation and reunion). The very fact that
frame-time and global simultaneity relations are available in some frames is in itself an
absolute (frame-invariant) fact about the situation.
Drawing from the light-cone structure of relativistic spacetime, Whitehead devised another
model for what he called contemporary events or processes: events which are
indeterminate as to their time order because they stand in a relation of mutual causal

18
independence. Contemporaneity in this sense implies disconnection. This model of
coexistence is limited to certain classes of events (space-like separated, in the physicists
parlance) and is not easily generalized to spacetime paths, or stretches of duration.
Nevertheless, the negative definition of coexistence as disconnection manages to capture a
phenomenological feature of our experience of the extended present: the absence or void
which hollows out the time elapsed between the emission of a signal and the reception of a
feedback, as when one is waiting for an answer to a letter, somewhat helplessly. Whitehead
acknowledges the fact that the relation of contemporaneity is only of importance in high-
grade organisms. The twins are a case in point: separated yet contemporaneous, it is as if a
certain measure of disconnection (in space?) was dialectically intertwined with connection
(in time?), as if absence was dialectically incorporated within the overall sense of co-
presence shared by the two observers. Langevins scenario of communicating twins
illustrates this nicely. As they exchange electromagnetic signal (factoring in Doppler effects),
the twins coexist continuously in Bergsons sense, but they are also contemporaneous with
each other in Whiteheads senseat least in part, and in a way that is less than clearly
continuous, since they are still capable of communicating in so-called real time

The importance of perspective and relational time
I could go on reviewing models of non-standard simultaneity: there are several others in
store, such as the active (or interactive) present based on the Alexandroff interval (the
dual image of contemporaneity, in a way12). But my main purpose was to illustrate the basic
issue at the root of all these considerations of coexistence in order to drive home, once
again, this very simple message: we are not dealing with time, properly speaking, unless we
make room for all its relevant dimensions, including simultaneity.
As I said, there is no point in choosing between duration, order and coexistence. To realize
that we need only reflect on the fact that mere succession, devoid of any notion of at the
same time, has nothing specifically temporal about it. It does not help much to introduce
the confused idea of a moving present, with new moments constantly chasing out previous
ones as they come into existence. Succession by itself is barely intelligible, it cannot easily be
distinguished from the continuum of instants itself until some sort of perspective comes into
play, extending beyond the successive here-nows, enabling us to consider time at a
distance, if that makes any sense.

12
On these issues of spatiotemporal coexistence, I recommend Youri Balashov, Persistence and
Spacetime, Oxford, Oxford University Press, 2010.

19
The problem becomes particularly acute in a relativistic context, with the promotion of
proper time. We may think at first glance that this local figure of time is more in tune with
Bergsonian duration because it is measured where the action is taking place, so to speak.
Modeled upon a particular processes, it may seem to massively reintroduce temporal
becoming, distributing it everywhere across space. It is as if time was finally unfrozen and
allowed to flow distributively along every fiber of spacetime, despite the lack of any
objective or absolute lapse of time between distant events (this, remember, was the lesson
of the twin paradox). That was more or less Paul Langevins offer to the philosophers he
addressed in 1911 at the Socit Franaise de Philosophie: what you take to be time, with
the associate phenomenological features (continuous flow, etc.), is really proper time. This
of course was meant as a friendly, reconciliatory gesture from the physical community to the
philosophers. However, as Bergson and Einstein both realized, this could not work at all. For
proper time is an inherently geometric concept associated with the representation of a
timelike path (a worldline) in spacetime. It is, at root, a four-dimensional concept with no
immediate temporal implications.
The advocates of the block-universe view are well aware of that: they have no issues
describing the block as an agglomeration of fibers, and they are quite right that no global
figure of time will help reintroducing an overall sense of passage, because such global
perspectives are of necessity artificial and relative, that is frame-dependent. The block itself
indicates no privileged foliation, and doesnt need to. Hence the uneasy feeling that
something essential has been left out of the picture. It is of course always possible to slide
along a worldline as if its points constituted a continuous series of instants. No hyperplane
is required here, and we may think we are thereby animating a fundamentally static four-
dimensional representation of the universe, but the point is that in the absence of any
notion of the same time, the succession of instants along a worldline is but an ordered
series and nothing more. In order to read earlier-later relations into the series and get a
sense of passage, we need some independent justification. The parameter by itself does not
provide that. You may use the symbol t if you wish, and consider that it can only measure,
but the truth is that the parameter does no more than what it does when describing a curve
in space. To assume that one is measuring durations since one dealing with time-like
intervals in space-time, is merely begging the question. This is not even spatialized time: this
is space, full-stop.
What is lacking here? Quite simply: a perspective on time. The possibility of scanning local
time from a distance. This may involve sweeping along a timelike path by means of a

20
reference frame once that path has been traced out across spacetime (analysis is by
necessity retrospective). Or it may require real-time signaling procedures and the concrete
transmission of information across space, as in the twins scenario. In any case, it is not
enough to merely trace or retrace a continuous path by following a parameter as it ranges
over an interval. You can flow along with the parameter t as much as you like, mere
parametrization will not yield time. What is needed is an observation deck of some sort,
whereby the inner duration of a conscious observer can pitch itself against a process that it
can consider as unfolding not only in space, but in time as well.
To repeat, the main difficulty is that as a local (monotonously increasing) parameter, proper
time does not come equipped with a natural notion of simultaneity. Lacking entirely the
depth or thickness attached to the idea of real, extensive becoming, I maintain that it has no
immediate temporal meaning13. It only acquires it when we step back and introduce at least
a second proper time acting as a baseline from which to evaluate duration. Thus, the
coexistence of the twins becomes a truly temporal affair as soon as we consider the way
they can take reciprocal views on their respective durations14. As a matter of fact, the actual
procedure involved in the physical measuring of any duration already involves just this.
What does a clock do, if not plotting a certain motion (or more generally, a sequence of
becoming, such as an hour spent on doing this or that) against another motion (or sequence
of becoming, preferably periodic: beats, oscillations and the like) taken as reference? The
relational account of time derives from this basic insight. It considers the physical concept of
time as a kind of operator fixing the exchange rate between the dynamic quantities involved
in situation. Accordingly, we never deal with time itself, as if it were some kind of underlying
property. Measuring durations is a matter of comparing several other physical variables;
time emerges from the comparison of at least two proper times15.

13
Cord Friebe makes a similar point in Twins Paradox and Closed Timelike Curves: The Role of
Proper Time and the Presentist View on Spacetime, Journal for General Philosophy of Science, 43(2),
2012.
14
Another illustration of this claim can be found in Bergsons comment on Zenos paradox of the
moving rows (also known as the Stadium) in Matter and Memory (252-253), which I take to be an
anticipation of the relativistic paradoxes. The moral of Zenos paradox can be summarized in the
following way: there wouldnt be anything temporal about either of the lines traced by the same
moving body in two distinct reference systems if some consciousness was not capable of surveying
the simultaneous unfolding of relative motions and infuse its own sense of duration to the resulting
trajectories drawn across space. In order to achieve this, it is necessary for consciousness to view the
whole situation from its own vantage point, at a distance, plotting the various temporal perspectives
against each other. That is the fundamental situation to which Duration and Simultaneity keeps
coming back: an interplay of perspectives attached to moving frames, and the pending question of
where the philosopher should stand in relation to them. Bergson suggests, and Merleau-Ponty after
him: neither here nor there, but somehow in between
15
This is the view defended today by physicists such as Carlo Rovelli or Lee Smolin.

21

Conclusion
The bottom line is that time, to the extent that it has something to do with genuine change,
cannot be confined to lines, whether tracks across space or worldlines in spacetime. Along
such curves, nothing flows: neither time itself, nor the consciousness of observers, despite
what Hermann Weyl suggested. How can we retrieve the extensive character of time,
spilling over local happenings to extend across whole regions of spacetime? How can we
recover simultaneity from proper time? Not by relying on reference frames and their ready-
made frame-time. Such frames are relative and in that sense arbitrary; moreover, they are
only approximations in the actual universe. But the real problem is that they offer a hybrid
solution to the problem of coexistence. Relying on them for philosophical amounts to
juxtaposing two distinct determinations of timeproper time and coordinate time
parameter and coordinatethat remain mutually external. What we need, instead, is a
genuine dialectic of fiber-time and frame-time, locality and perspective, a dialectic that may
enable us to account for the joint action of the three fundamental aspects of time: duration,
order and coexistence. Encouraged by Bergson, I believe that part of this endeavor consists
in probing the philosophical resources of non-standard accounts of simultaneity. The war
raging between presentists and eternalists, tensers and detensers, should not distract us
too much from this task.


BIBLIOGRAPHY

BALASHOV, Yuri, Persistence and Spacetime, Oxford, Oxford University Press, 2010.
st
BERGSON, Henri, Matter and Memory, London, Macmillan, 1929 [1 ed., 1911].
Creative Evolution, London, Macmillan, 1911.
The Creative Mind, New York, Philosophical library, 1946.
Mind-Energy : Lectures and Essays, Westport, Conn., Greenwood Press, 1975.
Le Souvenir du prsent et la fausse reconnaissance, Paris, Presses Universitaires de France, 2012.
BERGSON, Henri, Matter and Memory, London, Macmillan, 1929.
DURING, Elie, Vie et mort du cinmatographe : de L'Evolution cratrice Dure et Simultanit , in
Bergson, C. Riquier (ed), Paris, Editions du Cerf, 2012, p. 259-293.
Notes on the Bergsonian Cinematograph , in Cine-Dispositives: Essays in Epistemology across
Media, F. Albera et M. Tortajada (eds)., Amsterdam, Amsterdam University Press, 2014, p.
259-293.
FRIEBE, Cord, Twins Paradox and Closed Timelike Curves: The Role of Proper Time and the Presentist
View on Spacetime , Journal for General Philosophy of Science, vol. 43, n 2, 2012, p. 313
326.
KANT, Immanuel, Critique of Pure Reason, London, Macmillian, 1963.
WHITEHEAD, Alfred North, The Concept of Nature, Cambridge, The University Press, 1920.
Science and the Modern World, New York, The Macmillan company, 1925.

22