Chapter 1
Non-Linear Historical Materialism; Or, What is Revolutionary in Deleuze and Guattari’s Philosophy of History?
Eugene Holland
My aim today is to lay out the main features and some of the potential political implications of Deleuze and Guattari’s philosophy of history, which is the best formulation of a non-linear historical materialism that I know of. My points of departure are the two books that cleared ground for the view developed here: one is Manuel De Landa’s A Thousand Year of Non-Linear History, and the other is Jay Lampert’s Deleuze and Guattari’s Philosophy of History. The two books have reciprocal strengths and weaknesses: De Landa’s book is strong on science, but almost willfully incompetent on question of politics; Lampert’s book provides significant insight into politics, but is practically silent about science. I hope to remedy the shortcomings of both in what follows. Since I have already published a long review essay on De Landa in Culture, Theory, Critique, I will briefly mention here only two things, and develop them later only parenthetically, as it were (Holland, 2006, pp. 181–96). First of all, it is symptomatic of De Landa’s approach that, in summing up the five systems comprising the version of non-linear historical materialism he wants to derive from Deleuze and Guattari, starting with the solar system and including the biosphere and language, De Landa completely leaves out the sixth system: capital – which is, of course, a vital topic of concern for Deleuze and Guattari from their very first work of collaboration through to the very last (De Landa, 1997, pp. 261–62). It is impossible, in other words, to do justice to the politics of Deleuze and Guattari’s non-linear historical materialism without taking capital, and hence Marx, into account – as we shall see. The other thing to be said about De Landa, though, is that he is not really interested in politics in the first place and that he therefore refuses (or fails) to distinguish between science and politics, as Deleuze and Guattari do with great care.
About Lampert, I will have more to say, although I have also published a review of his book: Lampert’s study of Deleuze’s theory of time as it bears on Deleuze and Guattari’s philosophy of history, difficult though it is, merits closer examination and provides much of the framework for what follows (Holland, 2008, pp. 156–59). Yet, illustrations of contemporary science and complexity theory drawn from De Landa (as well as others) will be required to supplement Lampert’s account in crucial respects. In its basic outlines, this essay explores three inter-related topics: 1. emergence (as a category of non-linear complexity theory), 2. time and history (as they appear in Deleuze’s solo works and in the collaborations with Guattari, respectively) and 3. revolution (which is a persistent theme in all of Deleuze and Guattari’s works).
Emergence
Emergence is a key concept in non-linear mathematics, complexity theory, and contemporary science: it refers to the spontaneous self-ordering of physical as well as social systems. Order emerges from chaos, without that order being imposed from above or pre-determined from before. In Deleuze and Guattari’s more philosophical idiom, order arises immanently instead of being imposed transcendently. In a very interesting, later essay in which Deleuze is referenced by name, Louis Althusser coined the term “becoming-necessary” to characterize this kind of immanent self-ordering (Althusser, 2006, pp. 163–207). Let me say in passing that, in many respects, Deleuze’s debts to Marx pass through Althusser’s work; yet it must also be said that Althusser, particularly in the later works, acknowledges important debts to Deleuze’s work. In any case, the issue Althusser is addressing in this later essay, with help from Deleuze, is the systematicity of the capitalist system or the capitalist mode of production. It is one thing to explain how a given system works, when you take it as given; it is quite another thing to explain how it arose in the first place.
Taking this question seriously means reading Capital Volume 1 backwards, as it were: it entails prioritizing the emergence of capitalism through so-called “primitive accumulation” – the subject of the closing chapters of Capital Volume 1 – over the abstract systematicity of an already-constituted capitalism as embodied in the commodity form – the topic of the famous opening chapters. One of the outcomes of reading Capital this way reflects and reinforces a long-standing preoccupation in Althusser’s work: the perennial problem of reproduction, the sheer difficulty of assuring the system’s internal consistency and ability to perpetuate itself. So-called “primitive” accumulation, in this light, turns out not to be so “primitive” after all: it becomes a continual, if often overlooked, accompaniment and precondition of capital accumulation proper. And this will turn out to have important consequences or implications for political strategy, as we shall see.
For Althusser, like Fernand Braudel and Deleuze and Guattari in this respect, capitalism arose literally by accident. At some point, conditions were such that there happened to arise a critical mass – in the chemical or thermonuclear sense – of so-called “free” labour available for hire, and at around the same time and place, there happened to be a critical mass of the liquid wealth available for investment. A fortuitous encounter between these two critical masses created a specific reaction – commodity production by means of commodified labour-power – and the reaction eventually became self-sustaining. The “laws” of capitalist production were not necessary to begin with: they became-necessary – in Althusser’s felicitous phrase – as the system consolidated itself. Crucial to this process of consolidation, as Robert Brenner has usefully pointed out, was an unusually high (that is to say, practically unheard of) degree of market dependency for both groups (Brenner, 1982, pp. 16–113). The owners of what soon would become capital could not establish or maintain their social position outside the market economy; the owners of what would soon become labour-power could not maintain their social position either – or even their bare existence – without recourse to that same market economy. Indeed, so-called primitive accumulation is in large part a misnomer, since, as Marx says, its key feature is not accumulation at all but rather a kind of “dis-accumulation” or dispossession: that is to say, a forced separation of masses of people from their means of life. In the process called primitive accumulation, the chance encounter of these two critical masses gave rise to a system or mode of production that gradually became capable of sustaining or reproducing itself.
Now, imagine in the place of these two critical masses of human agents, two masses of chemical agents, brought together at a given temperature and in the presence of a certain catalyst. The fortuitous encounter creates a reaction, which results in the formation of a new chemical compound. If the chemical reaction continually produces more of the catalyst as a by-product of producing the new compound, the reaction will become “auto-catalytic”, and the new compound will spread throughout the solution, in a manner somewhat like crystallization spreading through a super-saturated solution. Now let us say that heat turns out to be another by-product of the auto-catalytic reaction: the solution may then pass a threshold beyond which it starts oscillating back and forth between two different forms of the new compound – creating what is called a chemical clock. It may happen, furthermore, that the chemical reaction produces not just an auto-catalyst but various alter-catalysts, which break down the chemical agents in the surrounding environment and make them susceptible to incorporation into the initial reaction. Then again, as this critical mass continues to develop, incorporating more agents and generating more heat, it may pass yet another temperature threshold or tipping point beyond which everything suddenly stops and there are no longer any chemical reactions whatsoever. The solution has transited a bifurcation point, or passed from one phase-space or basin of attraction to another: that one additional degree of heat has completely transformed or undone the systematicity of the system. Looked at strictly within the bounds a given phase-space, the systematicity of the system of chemical reactions appeared absolute and necessary; in fact, however, such systematicity arises out of the pre-existing conditions and parameters of an encounter: the system becomes-necessary upon entering that phase-space – and may become equally unnecessary if certain conditions are no longer met or successfully reproduced, and the system transits into a different phase-space.
This is what I am trying to get at here: the abstract systematicity captured in the opening chapters of Capital Volume 1 is actually a result of what Althusser wants to call a ‘becoming-necessary’ of the encounter evoked in the closing chapters of the same volume. The concept of becoming-necessary is meant to emphasize (among other things) the importance, the omnipresence, the contingency and the fragility of system reproduction: despite impressions generated by the opening chapters, the systematicity of a mode of production is never absolutely necessary and never guaranteed: it is always only becoming-necessary, like an asymptote approaching without ever reaching the line, or the lonely hour of a last instance that never comes. Note the essential role of market dependency as the key catalyst of capitalism: capital accumulation and wage-slavery are now seen not as the “laws” of capitalism, but as products of the material conditions (not of our own choosing) under which we live our lives and make history. The oscillation of a chemical clock, moreover, resembles the historical periodicity of capital accumulation described by Marx, or the ups and downs of the business cycle. Note, also, the role of imperialism as an indispensable alter-catalyst: so-called primitive accumulation is not only not really accumulation, but rather dispossession, it is also not really primitive, but rather always ongoing, as Rosa Luxemburg insisted in her landmark study of capital accumulation (Luxemburg, 1913, 1951). But note, finally, that the potential for widespread social change in the context of a non-linear historical materialism may reside in the slightest variation of conditions at a tipping point. Non-linear revolution, in other words, may not require the kind of massive force of resistance aimed directly against the massive power of capital prescribed by the dialectic of labour and capital located squarely within near-equilibrium capitalist phase-space, but only require some slight and apparently inconsequential shift in the balance of social relations at a far-from-equilibrium bifurcation point on (what becomes) the edge of capitalism.
In order to understand how this could be so, we need to take a closer look at both Deleuze’s theory of time and the philosophy of history Deleuze and Guattari developed on the basis of it.
Time and History
Both Deleuze’s theory of time and Deleuze and Guattari’s philosophy of history are extremely complicated – and doing justice to the relationship between these two complexities is one of the great virtues of Lampert’s book, which constitutes the necessary starting point for any discussion of the topic. The book’s primary virtue, however, is recognizing and insisting that Deleuze and Guattari indeed have a philosophy of history in the first place. There could be legitimate doubts about this, inasmuch as Deleuze and Guattari consistently express a preference for what they call becoming over what is conventionally considered “history”: Lampert’s book lays such doubts to rest, by explicitly addressing the relationship between becoming and history. The book’s second virtue is that it thereby lays the groundwork for what I want to do today, which is to explain the potential for revolution in history, when capitalism makes history universal by intensifying its potential for becoming, thereby altering the ratio of merely “performing” history to actually changing history, in favour of the latter.
We turn first to Deleuze’s philosophy of time. In his magnum opus, Difference and Repetition, Deleuze presents a view of time in terms of what he calls three passive syntheses, those of present, past and future (Deleuze, 1994). For brevity’s sake, let me say that these time-binding syntheses are considered by Deleuze to be strictly passive because of his concern to avoid the transcendental subjectivism of phenomenology: the syntheses of time are not the operations of an active self or ego managing or processing its experience, they are passive operations, which in fact give rise to all experience in the first place, including our experience of the self.
In his exposition, Lampert describes these syntheses as a ‘phenomenology of the present’, an ‘ontology of the past’ and a ‘pragmatics of the future’ – which is good as far as it goes. It might be said that “phenomenology” is a rather curious choice of name for the synthesis of the present, given the general antipathy to transcendental subjectivism I have just alluded to, but I think the choice is tenable in this particular sense: Deleuze’s account of the temporal syntheses in Difference and Repetition seems to take as its point of departure the way we experience time and deduces from that the way the syntheses must operate, as we shall see.
The conventional, linear depiction of time – as old as Newton, at least, and probably far older – presents it as a straight line in which each passing moment recedes behind the present, just as each approaching moment arrives from a future stretched out in front of us along the line we are travelling. It is surprising how pervasive and apparently convincing this depiction is at first blush – given that it is simply not true to our experience of time at all. For the past exists for us as a whole, not strung out along a line: to retrieve a past moment from two weeks ago, we do not have to rewind the entire chain of events to get there: we jump immediately to the first day of March. And we can jump from there to any other past moments, without having to trace out or locate those moments on any linear timelines. The past is, if you will, omnipresent to itself. At least that’s the way it seems to us.
But then the question becomes: is this true only of our experience of the past? Or is it true of the past itself? Reverting to Lampert’s terms: how do you get from phenomenology and how things appear, to ontology and how things actually are? To be sure, past events co-exist in memory – we can scan the past and access this event or jump to that event, without having to replay the entire succession of moments between them. But how do we get from this psychological experience/recollection of the past to the notion that past events themselves co-exist ontologically?
This is where Deleuze draws on Bergson, Nietzsche, and Leibniz. The past for Bergson is not the repository of a linear series of passing presents, but an a-temporal bloc where each and every passed event co-exists with all the others. For Bergson, it is not just in memory that one event can be connected with any other, irrespective of their respective places on a timeline: in the Bergsonian past, past events themselves co-exist and ‘undergo constant combination’, as Lampert insists; and this in turn entails – and here he quotes Deleuze – ‘ “the ever-increasing co-existence of levels of the past within passive synthesis” as time passes’ (Lampert, 2006, p. 48; Deleuze, 1994, p. 83). So pastness, as Lampert puts it, ‘is the logical capacity to undergo inexhaustible transformations; it is the virtuality of the event’ (Lampert, 2006, p. 48).
Here, with the introduction of “virtuality”, we need to take a detour and a short cut. Virtual and actual are central concepts in Deleuze. For present purposes, think of the language-system as Saussure describes it, as an illustration. It is the accretion or sedimentation of countless actual speech-acts extended over time, but it exists as a synchronic structure completely omnipresent to itself, to revert to the term that I used a moment ago. The language-system as virtual structure is to speech-acts as the past as virtual structure is to historical acts. The past as a virtual whole (or as a bloc) is the condition for actual events to take place in the present, just as the language-system as a virtual whole (or as a structure) is the condition for actual speech-acts to take place in the present. The Bergsonian past, then, is a realm of virtuality.
Now to account for the actuality of the event, Deleuze draws on Nietzsche’s anti-Platonic elevation of becoming over being and on Leibniz’s principle of sufficient reason. Being is merely a momentary, subsidiary and largely illusory suspension of becoming, according to Nietzsche; becoming is always primary and fundamental. This means not merely that each and every thing has a history – rather, each and every thing simply is its history: apparent being is always the temporary but actual culmination of real becoming; it is the present actualization of antecedent conditions contained in the virtual past. But to say that any entity is “its” history is not quite right either: each entity or state of affairs is not just its own self-contained history, but in fact the history of the entire universe, the entire past contracted via passive synthesis (as Deleuze puts it) from the perspective of that present thing or monad. This philosophical view aligns directly with contemporary science as informed by non-linear mathematics and complexity theory: basins or islands of linear determinacy certainly exist in the universe, but they emerge out of non-linear dynamics of the kind illustrated a moment ago with chemical reactions. Determinate being does emerge occasionally from becoming, but it arises always from a broader context of non-linear indeterminacy.
However – and this is crucial – the determination of any and every actual being by the virtual past in its entirety remains contingent for Deleuze: it only has determinacy when read retroactively; it could always have happened otherwise. (Think of evolution as a model of this retroactivity: rewind evolution and replay it one hundred times, as Stephen Jay Gould once put it, and you could end up with one hundred different results.) Couched in terms of the Deleuzo–Guattarian opposition between Royal or State science and nomad science or philosophy, we can say that State science tries to narrow down any thing’s antecedent conditions to the point where virtual becomings succumb to actual being and the thing appears to obey the eternal “laws of nature”; nomad philosophy, by contrast, retains the complexity and non-linearity of antecedent conditions, so that a thing’s present being is understood as a more or less temporary and unstable, contingent contraction of its becomings (Deleuze and Guattari, 1987; Holland, 2006, pp. 191–206).
This is the place to open a parenthesis to address one of De Landa’s two failings: his refusal to distinguish clearly, as Deleuze and Guattari always do, between science and philosophy, which amounts in his case largely to neglecting politics altogether. As Deleuze and Guattari explain in What is Philosophy?, the relations between the virtual and the actual in science and philosophy are the inverse of one another (Deleuze and Guattari, 1994). Entities and states of affairs come into being when a set of virtual conditions gets actualized in a specific way. By controlling variables and repeating experiments, science focuses squarely on actualized being, turning its back on multiple virtuality to define singular actuality as precisely as possible. Philosophy, by contrast, moves in the reverse direction: away from a given state of affairs, philosophy turns towards the virtual conditions from which it emerged. The task of philosophy is to extract from a state of affairs a map of the virtual of which it is an actualization – for any state of affairs is but one among many potential actualizations of its virtual conditions. The virtual is always richer in potential than the actual.
To be fair to De Landa and do justice to Deleuze and Guattari, it must be said in closing this long parenthesis that the foregoing stark contra...