Objects in manifold times: Deleuze and the speculative philosophy of objects as processes

James Williams

The University of Dundee

Abstract: This essay shows how real objects must be processes for Gilles Deleuze. These processes are determined by his account of time as a nine-fold manifold of processes deduced from Deleuze’s account of three interconnected syntheses of time in his Difference and Repetition (Différence et repetition, henceforth DR). It will also be argued that Deleuze’s philosophy of time is speculative in a broad sense and that Deleuze’s account of the real is opposed to forms of abstraction which associate objects with conceptual, perceptual or transcendental identity. In order to demonstrate the radical and systematic nature of Deleuze’s account of process, there is a discussion of a basic process underlying his manifold of time. This process is opposed to Markov chains, in order to set up an opposition to interpretations of Deleuze’s philosophy that deny its metaphysical and speculative approach in favour of scientific realism.

Keywords: Gilles Delezue; Philosophy of Time; Objects; Realism

Time as a manifold in Deleuze’s philosophy of time

In order to understand Deleuze’s critique of approaches to the object through identity in a concept (DR, 26) or through identity in a representation (DR, 179-80), it is necessary to follow how he uncovers processes of becoming behind these conceptual abstractions. In turn, these processes depend on the definition of time as a multiple process. His philosophy of time is a process philosophy of time where time is made through processes. This is in contrast to views where time acts as a container for events, or where time is a representation according to which events are situated. In Deleuze’s Difference and Repetition, time is constructed around three interdependent syntheses, labelled first, second and third syntheses of time (DR 96-165).

The three syntheses are not abstract or cognitive but rather real synthetic processes; they are more like a chemical fusion than a combination presupposed by, occurring in, or posited by a mind. Here, the syntheses will respectively be called contraction in the present, synthesis of the pure past and eternal return of difference. These names, though, are only short-hand for more complex processes and they must not be mapped on to the ordinary division of time into present, past and future, or associated with a timeline on which events can be indexed. Deleuze’s philosophy of time is a radical and revolutionary metaphysics of time that calls into question nearly all such ordinary assumptions because it claims that time is made as multiple by many processes. This multiplicity contradicts notions of the unity of time and of its unique direction from past to future.

The interdependence of the syntheses is very important, since each one, as primary process, includes the others as dimensions. Primary here indicates an independence from the dimensions (DR, 111). For example, the first synthesis is a contraction of the past and of the future, in different ways, where the past and the future depend on a selective process determining the present which itself does not depend on the contracted past and future. This relation of dimensions means that each synthesis is in fact three-fold: there is the prior process and then the two dimensions. For the present, we have the present as selective process and the past and future as contracted dimensions of that selection. (DR, 101-102) The syntheses therefore give us a nine-fold set constituting time as a manifold of processes. We have the present as primary synthesis, as dimension of the past and as dimension of the future. This tripartite distinction also holds for the past and the future.

Any isolation of any one of these processes is incomplete due to their interdependence. Nonetheless, it is possible to treat the syntheses separately, so long as we understand that an incomplete picture has been given and so long as the consequences of this incompleteness are taken into account. This raises a difficult question with respect to the notion of primary synthesis which reveals important factors in Deleuze’s treatment. A primary synthesis is independent with respect to its dimension under the primary characteristic process. It is not, however, independent either of its role as dimension according to other syntheses, or of other processes determining it as primary yet passive with respect to these processes. So primary does not indicate absolute independence as a particular synthesis of time, or within a particular synthesis of time, or as pure activity. The present as a selection determining contractions of the past and of the future is itself determined by the second synthesis and by the third. It is also determined as a passive synthesis and as an active synthesis in the present (DR 105-6). It is therefore more accurate to speak of an indivisible and complex manifold of times.

Here is a grid of the manifold which will also serve as the grid for explaining how objects must be processes in relation to time. The grid is an imperfect representation for the following reasons. First, each square is not a sufficient definition of each time and synthesis. Second, each line and each column must not be seen as a set of three independent processes. Third, the grid as whole should not be understood as divisible according to its squares. Instead, the grid and squares are a starting point for an account that must show the relations between squares in an account aiming at completeness, rather than self-sufficient analytical distinctions. On the grid, primary processes are on the left-to-right, top-to-bottom diagonal, all other processes are dimensions of the primary processes on the corresponding column:

First synthesis of time (synthesis in the present)

Second synthesis of time (synthesis of the past)

Third synthesis of time (synthesis for the future)


As prior selection

As made to pass as the most contracted state of the pure past

As incapable of returning and as caesura, assembly and seriation


As dimension contracted into the present through a singular selection

As synthesis of the pure past

As selected to return as pure difference and as symbolic process


As dimension contracted into the present as a range of possibilities assigned given probabilities

As freedom and destiny

As eternal return of difference

The caveats about the interconnected nature of this grid and about the correct way to approach it are important from an explanatory point of view—as already discussed—but they are also important from a practical point of view. It would be a mistake simply to apply the grid to objects according to its distinct squares in order to give a version of Deleuze’s philosophy of time as relevant to objects or indeed to any given phenomenon. Any such application might well start with such a move, but it must go beyond it to demonstrate the complexity of relations between any identified processes.

Deleuze’s basic process and Markov chains

Though time is a complex multiplicity of processes for Deleuze, there are common patterns across them. The most basic and important one is partly inherited from Lacan, though not without important changes. The most significant locus for this inheritance is Logic of Sense (Logique du sens, henceforth LS; see LS 50-56 for a study of Deleuze’s account of double series and seriation in relation to Lacan’s reading of Poe) though it carries through to Difference and Repetition. The reading of Lacan in relation to Roussel and Poe in Logic of Sense is split in Difference and Repetition into a discussion of Lacan in relation to the virtual object (DR 134-5, see also LS 264-5) and Roussel in relation to dark precursors and double series (DR 154-60). Deleuze’s process also bears loose connections to Markov chains (according to Lecercle, taken via Ruyer in Difference and Repetition, though the connection is not entirely evident in DR itself, see DR 279). These influences and differences will be made more precise here in relation to Markov. The relation to Lacan is more complex and is discussed by Lecercle in close context to his work on Markov discussed below; however, this large and important connection to Lacan is beyond the scope of this essay.

The basic process is a disjunctive synthesis across at least two series (LS 265-7). This synthesis has two important functions that can seem paradoxical when considered together. First, there is a synthesis of states of the two series as they are related to one another through linked terms in each. Second, though, that synthesis implies disjunctions between states of each series when taken in abstraction from the other series and the relating terms. So there is a relation between two series synthesising their different states, yet the effect on each series of this synthesis is such that states of that series are independent from one another. The synthesis is hence disjunctive within each series, yet synthetic across two or more series (LS 267).

It is important to stress that the series must not be viewed sequentially, but rather that their transformation occurs instantaneously for each state of the series (LS 194). It is also essential for Deleuze’s model that sequences should not be impervious to transformation since, as will be shown below, sequence imposes a linearity on models that prejudges how they can be transformed and therefore conditions the forms of time they then imply. This conditioning can occur, for instance, when earlier members in the sequence are labelled ‘past’ and later ones ‘future’ in relation to a ‘present’ member, or when a concept of distance can be used to indicate separation of elements in the series according to the measure of a timeline corresponding to the sequence.

For example, given series A’ and B’ related respectively by twinned elements a’ and b’, the introduction of new relating elements a’’ and b’’ instantaneously transforms A’ and B’ into A’’ and B’’ with no predictability in how A’ becomes A’’ or B’ becomes B’’, hence the serial disjunction. A’ and A’’ are independent as members of the series of A series, but they are related when taken with B’ and B’’ through the relating elements a’, b’, a’’, b’’. A synthetic relation connects A’, B’, A’’ and B’’ through the relation of a’/b’ to a’’/b’’. Yet states of the A series are independent, as are states of the B series. This independence and disjunction is an effect of the synthetic relation.

In terms of the first synthesis in the manifold grid of time the disjunctive synthesis takes place through the following properties. Each novel selection and contraction through a first synthesis is a disjunction with the selection and contraction of other first syntheses, yet these are related through their passing as dimension of the pure past and return as caesura, assembly and seriation as dimension of the third synthesis. Primary syntheses communicate as dimensions of other syntheses, though they are independent as primary syntheses. In the language of the living present (DR 97), adopted by Deleuze in his study of the first synthesis in Difference and Repetition, any living present is a disjunction with any other, yet they are related as dimensions of the pure past and of the eternal return of difference. This form of relation and communication of processes, defined as non-communicating or independent under some primary process or definition, is one of the most fundamental problems of Deleuze’s philosophy. It can be found in Logic of Sense in the discussion of the communication of events (LS 198-207). It returns in Deleuze’s work on Leibniz’s Monadology in The Fold: Leibniz and the Baroque with the communication of monads ‘with no windows’ on to each other (Le pli: Leibniz et le baroque, henceforth LP, LP 38-44).

Take a series A’ given by a pattern of beads hung on a string between two poles. Let there be another series of beads, B’, parallel to A’ and also hung on poles. Let the series be moved by a flow of air moving the beads such that the flow around A’ changes the flow around B’ and vice versa. If the two series accord with Deleuze’s disjunctive synthesis, when the relation of the flow of air to the resistance of the beads changes we know that there will be a change in the pattern and the regularity of the flow. We can describe this change by describing and tracing the new flow and new patterns, A’’ and B’’. But, crucially and counter-intuitively, according to Deleuze’s model and against law-driven or probabilistic accounts, when there is a disjunctive synthesis we can never fully predict the new patterns based on the old patterns and on the new and old flows. Nor can we fully explain their relations retrospectively. There is a disjunction of patterns although they follow from relations between the two series, yet they are synthesised through their relation to one another through the change introduced into each one of them. Syntheses are cross series; disjunctions are internal to series.

I have given this example to highlight the radical and problematic nature of Deleuze’s basic process. If we are dealing with physical objects, it appears that he is committed to a position running counter to areas in which we have reasonably secure theoretical and practical knowledge; aerodynamics in this case. So it seems that Deleuze is committed to a speculative denial of well-established knowledge and the scientific practices and theories it is based upon. This is not the case. The basic process underlying Deleuze’s philosophy of time is consistent with established forms of knowledge, which it can ascribe to series if we abstract them from transformative syntheses. In this case the abstraction would operate through the tautological assumption ‘given no transformative disjunctive syntheses, then…’ The example is then not designed to extend Deleuze’s claim to physical objects and knowledge about their behaviour, but rather to give a model explaining Deleuze’s process as it applies to disjunctions in established models that call into question knowledge of them based on an abstraction from real processes.

Deleuze’s philosophy is therefore compatible with accounts from the sciences. However, his concept of the real extends from an actual abstraction, as given by a form of knowledge, into a speculative model designed to explain transformations which show such forms to be incomplete. His basic process and his manifold philosophy of time are designed to underpin this complete speculative and metaphysical version of the real. Abstraction is therefore defined here as a cut away from the processes constituting the real through two moves. First, a restricted field of those processes is taken, for instance, by considering a series in abstraction from others. Second, that restricted series is itself considered in a limited manner, for instance within set boundaries. Abstraction should not be seen as a purely negative term here. It is necessary and is itself explained as a real process. However, if it is taken as providing either the last word on the real, or a sufficient critical basis for denying the reality of other processes and broader metaphysical frames, then serious damage is done to our capacity to explain real processes.

It is at this point that a significant counter-position to my reading can be introduced. In describing the difference between Deleuze’s model and one based on an abstraction, where abstraction is itself defined in relation to a speculative account of the real, I twice used a vague term: ‘fully’. The term was first used in relation to prediction and second in relation to explanation. It could be claimed that the use of this ill-defined term is a sign of a sleight of hand which allows my account to retain predictability and explanation while also denying them in an ill-defined but supposedly important ‘full’ situation. This in turn is a sign of a failure to account for the basis for the limits set for prediction and explanation, which turns out to be illegitimate, in the sense of based on an ill-founded speculative move. The wider background to this critique is that speculative moves are illegitimate unless they are based on theories taken from the natural sciences or from formal disciplines such as mathematics. There either has to be a scientific ground for incompleteness in prediction or explanation (say, from chaos theory) or a formal one (say from Gödel’s incompleteness theorem). Otherwise, a speculative move is in fact based on illegitimate idealist grounds. This is a charge made against me by Manuel DeLanda in response to my earlier critique of his position on Deleuze and the sciences.[1]

In answering this point, I will first give a definition of speculative as it is used in this essay. I will call the construction of a philosophical system speculative when it introduces a concept or set of concepts to a system which, though they may be taken from other subjects and practices, have a meaning, or consistent place, or justification proper to the philosophy. For example, a philosophy of mind is speculative when it takes a concept of mind from cognitive sciences, but then sets that concept into a system such that at least one of these condition holds: the concept of mind departs from the one found in the science, or is assigned a consistent role within a system with non-scientific elements, or is justified in terms of claims to truth or validity that are not scientific. This definition is deliberately very wide, such that almost all philosophy must be speculative. However, it matters how philosophy is speculative: through the creation of new meanings, consistent systems and claims to truth and validity.

In line with this definition, the answer to the critique about the speculative aspect of Deleuze’s philosophy is that it is not a matter of concern that a philosophy is speculative. Any philosophy that is not slave to another subject will be—Indeed so will such subservient philosophies at the point where they justify their submission. It is matter of concern, though, if the speculative aspect is poorly done on its own terms and in relation to the history of philosophy. So the deeper concern of the critique outlined above is that my account of the systematic aspect of Deleuze’s speculative moves is vague and ill-defined, as shown by the loose use of ‘fully’ with no well-defined boundaries, in relation to prediction and explanation. The counter to this is that we can define the limits of explanation and prediction in Deleuze’s system accurately according to the concept of disjunction. Where there is a disjunction, the validity of prediction is broken and the type of explanation changes. This marks a shift from established explanations to necessarily experimental ones seeking out new modes of explanation in accordance with syntheses between series that are internally forced into a disjunction by their relation. I will now clarify this further by distinguishing Deleuze’s process from work on Markov chains.

Jean-Jacques Lecercle has discussed the combination of disjunction, or independence, and synthesis, or partial dependence, in relation to language and to Markov chains in mathematics (and their applications in other sciences). Markov chains are generated by the addition of novel states independent of earlier ones in a series: ‘[the state] retains no memory of where it has been in the past. This means that only the current state of the process can influence where it goes next’.[2] For example, in a game of snakes and ladders, each throw of the dice is independent from earlier throws. There is no memory in the dice of earlier outcomes or later ones (of earlier ‘luck’ or ‘misfortune’). The novel state is independent of past ones, yet probabilities can be assigned to future states, hence the partial dependence. The probable outcomes of a game of snakes of ladders can be assigned after each throw of the dice, but they also change with each throw and with its independence from earlier ones.

Lecercle explains Deleuze and Guattari’s interest in Markov chains according to the following properties: they are ‘linear’, ‘aleatory’ and ‘partially dependent’.[3] Lecercle’s example for this is in the generation of words completing a sentence: ‘To use my example again, if I feel a strong compulsion to finish the Markov chain ‘Pride comes before a …’ with the word ‘fall’, nothing actually prevents me from engaging in a creative exploitation of the proverb by replacing the expected noun with another one: ‘Pride comes before a vote of impeachment’—so I was talking about Nixon all the time’.[4] The sense of linear used here is of a series that states are added to. The sense of aleatory is that the choice is free. The sense of dependence is that a new sense emerges for the series with the new word. None of these are quite accurate enough for an understanding of the difference between Markov chains and Deleuze’s basic cross-serial process.

Though Lecercle’s essential study of Deleuze’s work on language captures something important about Deleuze and Guattari’s references to Markov chains and the opportunity they provide for a critique of Chomsky’s linguistics, the three properties outlined above need to be modified or at least explained further in order to understand Deleuze’s work on process and time in Difference and Repetition and Logic of Sense. This is because the concept of linearity leads too rapidly to the idea of a line of time unfolding in a linear manner from past to future. An indication of this implication can be found in James Norris’s definition of Markov chains through the concept of ‘past’, as given above. For Deleuze, the most basic process is rather a relation across series and a radically aleatory disjunction between different states of single series, rather than a serial unfolding. For Deleuze the basic states are states of series, rather than elements added to a sequence. The notion of the aleatory does not depend on chance selection within a finite set (as it does in Markov chains) but rather in the introduction of a novel ‘transformer’ within a series in relation to a novel different ‘transformer’ in another series. This is why I use the adjective radical here in order to distinguish Deleuze’s aleatory ‘dice throw’ from probabilistic definitions of chance. For Deleuze’s disjunction, a die morphs when it is thrown and loses its prior relations of probability. If it did not, there would not be disjunction in a fundamental sense.

This means that Deleuze’s model goes beyond even complex and interconnected Markov chains. First, this is because it is not a selection within a finite set. Second, it is because though it is not amenable to assigning probabilities in series when they are considered in terms of the transformative relations between series. Of course this does not mean that we cannot think of series in terms of probability when they are treated in abstraction. Partial dependence must therefore be understood as dependence within a series until it is disrupted through its relation to another one which introduces an independent disruption through both series without exception. Deleuze’s philosophical model can therefore be seen as an attempt to go beyond the theorems of Markov chains. The standard definition of a Markov process is at odds with Deleuze’s account of time in its most basic process (I would argue that this is also the case in terms of the third synthesis of time and Deleuze’s concepts of caesura and eternal return, but that is beyond the scope of this essay—for a full account of these process see my Gilles Deleuze’s Philosophy of Time: A Critical Introduction and Guide).[5] The process relating series in Deleuze’s model therefore has to be defined in such a way as to be inconsistent with Markov chains. The process does not depend on a closed set, it is not linear and it is resistant to the assigning of probabilities according to the mathematical theorems for Markov chains. The occurrence of a new state in a Markov chain is not a disjunction or real event in Deleuze’s sense of the terms.

However, the discussion of Markov chains is important and helpful here because it allows for a general specification of the properties of Deleuze’s most basic process. It is a full transformation of at least two series by at least two related factors whose effects instantaneously run the length of the series and retain no memory of prior states of the series. In his conclusion to his insightful essay on Deleuze and chaos, Gregory Flaxman therefore underestimates the differences between Deleuze’s account of the ideal game from Logic of Sense (LS 74-82) and Markov chains, as referred to in Deleuze’s Foucault and Deleuze and Guattari’s L’anti-œdipe ‘proches d’une chaîne de Markoff’.[6] Flaxman rightly notes Deleuze’s insistence on a single rule of ‘variability, variation and variety’.[7] This underestimates the rigidity of Markov chains, which map many cases of games and process with fixed degrees of probability for given states, in contrast to Deleuze’s most basic processes denying the possibility of assigning such probabilities in relation to new states and transformations. The challenge, in studying Deleuze’s philosophy of time, is then to explain how he can present us with a consistent manifold of time without returning to regularity or to a mathematical model such as Markov chains, yet retaining coherent structures and processes.

Objects in manifold time

Having outlined the manifold of times deduced from Deleuze’s three syntheses, as well as the basic process underlying the manifold, I will now suggest how both transform our conceptions of objects. In this section I will situate objects in the manifold. In the concluding section, I will combine this with the basic process in response to critical objections.

When Deleuze considers objects and subjects as repeated, in relation to the first synthesis of time in Difference and Repetition, he makes the claim that neither an objective nor a subjective account is satisfactory for an understanding of repetition and time: ‘In considering repetition in the object, we remained short of the conditions that render an idea of repetition possible. But in considering change in the subject, we are already beyond them, in the general form of difference’. (DR 97) The processes of time as repeated variations are neither repetition of objects, not repetition for a subject. Instead, objects and subjects must be thought of in terms of the syntheses of time, as opposed to grounding those syntheses on objects and an objective realm or subjects and their intentions and acts. Objects and subjects are processes of the manifold of times. When objects or subjects are posited outside such processes the object or the subject is an abstraction from the real.

For the purposes of this essay, I will set aside the question of the subject except where it impinges directly on reflection about the object in time. This leaves the problem of how objects must be thought of in relation to Deleuze’s account of a manifold of time as multiple processes, and in terms of the basis process outlined in the previous section. In this section, I will consider the manifold, leaving the basic process of dual related series to the next. Here is the time grid again, expressed according to a generic object W defined as process:

First synthesis of time (synthesis in the present)

Second synthesis of time (synthesis of the past)

Third synthesis of time (synthesis for the future)


An object W is a selection in the present

W is made to pass as the most contracted state of the pure past

W is incapable of returning and institutes a caesura, assembly and seriation of events


As singular selection in the present W contracts all past series but to greater or lesser degrees

W presupposes but does not have effects on a synthesis of the pure past

W returns only as pure difference and as symbolic process


All future series are contracted by W according to a different ranges of probability

W has a degree of freedom and a destiny

W presupposes but does not have any effects on the eternal return of difference

If we study the first column on the first synthesis of time, a number of important and surprising features of objects come to the fore. An object is not a well-delimited thing in space-time, but rather a process of selection. An object is not a thing that is picked out or selected. The object is a selection that picks out and selects past and future series within a process. For the past, this selection gives greater and lesser degrees of significance to series within a contraction in the present. The processes operating from the present on the past are selection and contraction. Selection means changes in degree. Contraction means a transformation of a past series according to a change in the present. The present here is therefore not an instant or an eternal present, but rather a transformation across series with a focal zone. Loosely we can call this zone the living present. An object is therefore a process concentrating past and present through its living present, where life must not be associated with human life or even biological processes, but rather simply with a selection.

For the future, the selection in the first synthesis alters ranges of possibilities through changes in probability. The selection picks out an object in the living present and contracts past series. It also contracts later series by assigning different probabilities to them. (DR 97) This selection is not by a subject, but indicates a mere change in a pattern, this could be a pattern in the brain of something picked out as a human subject, but it could just as easily be a morphological change in a biological series, or a crack appearing in a material, or a change in a weather pattern. The key is that the change selects series differently. For the past this is in terms of degrees of significance of series in terms of their contraction in the living present; for instance, when a past series becomes unimportant due to present selection.

For example, a selection occurs in the living present with changes in territory or climate which then concentrate series of evolutions differently, making some lines lead to extinction and others ongoing success stories. For the future, change in the living present selects different probabilities to assign to possible futures. A different pattern in the brain might make some futures for that human being very unlikely. The crack in a vase might make its future as an effective water carrier extremely dubious or a change in weather patterns might make successful wine growing in Perthshire more likely. There is no concentration of the past in that crack that is not also a concentration of the future. However, these processes are asymmetrical since one assigns different degrees to actual series whereas the other assigns different probabilities to possible series.

Two types of remark can now be made in relation to objects in relation to the past and the future. First, it is striking how underdeveloped Deleuze’s work is in terms of its detail. For instance, there needs to be a lot more research on the exact form of probability and possibility implied by the idea of the future as dimension of the living present. Are some things impossible? Are others incommensurable? What is the right formal account of probability? How does the use of the possible and the probable in the first synthesis of time work with the critique of the category of the possible in Deleuze account of the real and the virtual in Difference and Repetition? It is not that Deleuze’s texts fail to consider such questions, but rather that the places where they are tackled are very different and disparate; for instance in terms of incommensurability and incompossibility in Logic of Sense (LS, 199-203) and The Fold: Leibniz and the Baroque (LP, 79-86).

Second, the first synthesis of time can be a trap for readers because it is somewhat close to our grooved intuitions about past, present and future. For example, the difference between past and future dimensions fits well with accounts of the asymmetry of time implied by a difference between the uncertainty of the future and the settled nature of the past. It is therefore very important to stress the incompleteness of the living present and its contractions of the past and of the future in relation to questions about, past present and future in Deleuze’s thought. The present is never simply the living present, or the past the contracted past, or the future the probable future. This is not only because there are other processes that must be taken account of, such as the pure past, the passing present and the future as eternal return. It is because these processes play necessary roles within the first synthesis of time, its prior process and dimensions.

A first understanding of Deleuze’s extension of the present, even when defined as a contraction, can be found in its definition as made to pass by the pure past in the second synthesis of time (DR 105) and as never to return by eternal return in the third. It is important not to view these as separate from the present as selection and contraction. Instead, the present is a selection and, inseparably, also a passing away never to return. This means that Deleuze’s account of objects can itself be extended in two original and deeply challenging ways. An object is a selection across series, but as such it is immediately a falling away into a pure past, that is, not as a representation but rather as something that cannot be represented. This does not mean that it becomes nothing in relation to the pure past, but rather that the effect on the present lies in a change of dimensions and degrees of pure relations, that is, of relations that are not between things or terms but rather strictly to other relations as degrees of relatedness. When a present is made to pass, its passing changes the relations of degrees and significance for everything else that has and will pass.

In a small earth tremor, the last remaining vase turned by a famous potter is smashed as it totters off a display. The line of creations is concentrated differently, destined to disappear in the quake. Future probabilities change too; no one will discover a long lost piece, though forgers are more likely to try to make one. The vase, though, has not passed into record as a representation in the pure past, since according to Deleuze’s philosophy of time any such record is a concentration in the present selection, as dimension of the present. The pure past makes the present pass by separating the present as process from the past as ever changing reserve of relations of value that return in the future. This is one of Deleuze’s most difficult thoughts but in terms of objects it can be understood in this way. An object is a selection concentrating past and future, but as such it is also a passing away into incapacity to change anything. As it smashes the vase, the hammer can never again smash it in the same way and therefore its capacity to alter the values and significance of that blow also fades away (for a good technical discussion of passing away and the pure past, see Jay Lampert’s Deleuze and Guattari’s Philosophy of History).[8]

Does this mean that the hammer and vase pass into nothingness or into a frozen record representing passed moments? No, that which makes the process pass is not a physical passing but one of relations of values and intensities. A process as living present passes because its significance is altered in relation to all the values it is related to as these values change in the pure past. A moment cannot be reclaimed because the way in which it matters in relation to the way all other moments have and will matter has changed. It’s not that the wedding gift is smashed irreparably. Nor even that you smashed it. It’s that nothing feels or can ever be the same again since you did.

The pure past and eternal return, as future, are therefore very important in relation to objects in Deleuze’s manifold of time because neither one allows the present object as selection and concentration to touch it in any way. This is why the manifold grid for W does not refer to it at the past in the second synthesis or at the future in the third. The pure past and the differences of eternal return are neutral, not in the sense of impotent, but in the sense of impassive. W is not a variable in those parts of the grid. No object returns. No object has a presence in the pure past, in the reserve of values ever changing in intensity and ever calling all objects to pass.

Is this then a deliverance of objects from the past and the future? Not at all: it is a shift from understanding objects in a causal relation to the past and to the future, in any relation of continuity with them, to an understanding of the present process as a break and as a special kind of transformation resisting external structures or laws guaranteeing general types of continuity. Each process becomes singular as break and transformation in relation to the pure past and to eternal return. It is only in that singularity, in singular differences, that there is a passing away and a return.

So if objects are thought of in relation to Deleuze’s manifold of time, they do not only become processes. They also become causally restricted as such. In relation to the pure past, any object is also a moment of freedom: its passing does not condition what follows it, either in terms of appropriate reaction or in terms of necessary effects. Yet the object also has a destiny, but one that always goes beyond it, as a destiny to return differently and hence also as a trace in that difference. In relation to the future, the object as present process becomes a break or caesura within series (DR 120). Everything changes with the process as it ushers in the new. Yet in that change there is also an assembly of all series as transformed by it. In turn, this transformation allows for a seriation into the before and after of the change.

So when the earth trembles, or when a will snaps, something is broken. The break is not only a process changing the actual past and the possible future. It is also an opening, a freedom for different and novel events. The process is also being made to pass and this passage into pure difference, into pure values, is all that can return in the future as novelty. The process has a double destiny in this return. It cannot return as itself but must return as the trace of the transformation of its passing. According to Deleuze’s definition of the third synthesis of time, the break is also a caesura where series—history if you like—are cut asymmetrically into before and after, yet also set into a new assembly (DR 120-1). As process an object transforms, passes away never to return as the same, yet returns as the return of difference. As such it transcends all boundaries in space and time. It is many different processes, each one carrying the object through other series as concentration, change in probability, gift of freedom, cut, seriation, assembly, destiny, oblivion of the same and return of pure differences.


From Deleuze’s work on time as process, I have shown how objects must also be complex manifolds of processes that can be understood according to a manifold of times. This work is only preliminary though. It only shows the promise and complexity of Deleuze’s work for an approach to objects. The detail of that work remains to be done in deeper practical and theoretical analyses. The promise for such work is two-fold. From a negative point of view, it provides a basis for a critique of theories of the object that make claims for its independence from processes of becoming and of selection, of evolution and conflict. When viewed from Deleuze’s account of time any object is a crossing point for transformative interactions, it is made from and expresses at times conflicting and at other times synthetic encounters. An object can never be assigned final boundaries in time or space. From a positive point of view, Deleuze’s speculative process philosophy gives us a rich frame for explaining and interacting with objects. It reveals our investments in them as mutual processes, rather than one-way projections. It also shows the many ways in which objects must necessarily thwart and overcome those investments. This speculative approach avoids any mystery to be assigned to objects in their resistance to thought. Instead, resistance is the product of the multiplicity and complexity of processes. Their multiple natures mean that any object must be approached through many different dimensions and according to different orders of priority that cannot be reduced to a totalising final view. The complexity comes from the role of singular selections within processes. None are simply genera. Processes cannot be viewed in the abstract without losing something of the real object in each singular series of processes drawn together according to individual syntheses. This means that objects work within multi-layered worlds. These are determined by singular selections, yet also interact with one another as selections, again though with no possibility of either totalising or fully disjunctive solutions to the paradox of singular worlds taking in one another.


Deleuze, Gilles, Difference and Repetition, trans. Paul Patton, London, Athlone Press, 1994.

Deleuze, Gilles, Logic of Sense, London, Continuum Press, 2004.

Deleuze, Gilles, Foucault, London, Continuum Press, 2006.

Deleuze, Gilles, The Fold: Leibniz and the Baroque, London, Continuum Press, 2006.

Deleuze, Gilles and Guattari, Felix, Anti-Oedipus, London, Continuum Press, 2004.

Flaxman, Gregory, ‘The subject of chaos’, in The Force of the Virtual: Deleuze, Science and Philosophy, ed. Paul Gaffney, Minnesota University Press, 2010.

Lampert, Jay, Deleuze and Guattari’s Philosophy of History, London, Continuum, 2006.

Lecercle, Jean-Jacques, Deleuze and Language, Basingstoke, Palgrave Macmillan, 2002.

DeLanda, Manuel, ‘Afterword: The metaphysics of science: an interview with Manuel DeLanda’ in The Force of the Virtual: Deleuze, Science and Philosophy, Peter Gaffney (ed.), Minneapolis, University of Minnesota Press, 2010.

Norris, James R., Markov Chains, Cambridge University Press, 1997.

Williams, James, ‘Science and dialectics in the philosophies of Deleuze, Bachelard and DeLanda’, Paragraph: a Journal of Modern Critical Theory, vol. 29, no. 2, July 2006.

Williams, James, Gilles Deleuze’s Philosophy of Time: A Critical Introduction and Guide, Edinburgh University Press, 2011.

[1]. Manuel DeLanda, ‘Afterword: The Metaphysics of Science: An Interview with Manuel DeLanda’ in The Force of the Virtual: Deleuze, Science and Philosophy, Peter Gaffney (ed.), Minneapolis, University of Minnesota Press, 2010, pp 325-33, p 329. James Williams, ‘Science and dialectics in the philosophies of Deleuze, Bachelard and DeLanda’, Paragraph: a Journal of Modern Critical Theory, vol. 29, no. 2, July 2006, pp 98-114.

[2]. James R. Norris, Markov Chains, Cambridge University Press, 1997, p. xiii.

[3]. Jean-Jacques Lecercle, Deleuze and Language, Basingstoke, Palgrave Macmillan, 2002, p. 96.

[4]. Ibid. p. 95.

[5]. James Williams, Gilles Deleuze’s Philosophy of Time: A Critical Introduction and Guide, Edinburgh, Edinburgh University Press, 2011, pp. 86-134.

[6]. Gilles Deleuze, Foucault, London, Continuum Press, 2006, pp. 71, p. 96. Gilles Deleuze and Felix Guattari, Anti-Oedipus, London, Continuum Press, 2004, p. 42, note that this reference is situated at a point where Deleuze returns to his work on Lacan and Poe.

[7]. Gregory Flaxman, ‘The subject of chaos’, in The Force of the Virtual: Deleuze, Science and Philosophy, Paul Gaffney (ed.), Minnesota University Press, 2010, pp. 191-209, p. 209.

[8]. Jay Lampert, Deleuze and Guattari’s Philosophy of History, London: Continuum, 2006, p. 46.