Manuel DeLanda's Art of Assembly
Theorists have devoted more interest to questions of "the virtual" recently. This is due, in part, to growing familiarity with the scientific concepts necessary to its interrogation, as well as the philosophical writings of Gilles Deleuze and those of philosophers he has resurrected, such as Spinoza and Bergson. But this interest is also the result of growing dissatisfaction with current theoretical approaches that rely on "top-down" methods unable to effectively account for the emergence or mutation of systems. Manuel DeLanda, for instance, has referred in his writing to oversimplifications that attribute causes to posited systems such as "late capitalism" without describing the causal interaction of their parts, which would change in different contexts. In his introduction to Parables for the Virtual, Brian Massumi argues that cultural theory's over-reliance on ideological accounts of subject-formation and coding has resulted in "gridlock," as the processes that produce subjects disappear in critiques that position bodies on a grid of oppositions (male-female, gay-straight, etc.). In one of his more exceptional examples, Massumi argues that Ronald Reagan's success as the "Great Communicator" was not due to his mastery of image-based politics to hypnotize an unwitting public. The opposite was the case. Reagan's halting speech and jerky movements were the source of his power, the infinite interruptions in his delivery so many moments of indeterminacy or virtual potential that were later made determinate by specific receiving apparatuses, such as families and churches. In short, interactions among non-ideological parts produced ideological power. Critiques that consider only the ends of ideology are unable to examine the very processes that create constraining subject-formations in the first place.
In his previous book, A Thousand Years of Nonlinear History (reviewed in ebr by Geoffrey Geoffrey Winthrop-Young), DeLanda developed a bottom-up approach to historiography that posits a system only if an account of its "system-generating process" can be given. Impoverished bottom-up approaches (such as neoclassical microeconomics) tend to "atomize" individuals by looking at causal processes or decisions in isolation from each other, and often ignore synergistic interactions between many parts (actors, causes, etc.). Top-down approaches begin with an assumed system, such as society as a whole, and cannot account for the emergence of the system itself. In either case, the generative processes that govern systems, as well as complex interactions among their parts, remain hidden under the shadow of final products. As Massumi and DeLanda have shown, theories that begin with the extensive - the "actual" physically perceived systems that are, in fact, only the final product of interactions among many parts - do not adequately explain the origins of those systems. Nor do such theories provide an alternative to transcendent forms of ontology (in which matter is said to receive form), as they lack the proper "mechanisms of immanence" to explain how matter may generate form.
Now, in Intensive Science & Virtual Philosophy, DeLanda gives us a way to approach these problems through the often-misunderstood concept of "the virtual." [ link to Paisley Livingston's 1996 essay on Stanislaw Lem and the history and philosphy of Virtual Reality] Synthesizing Deleuzian ontology with key concepts of contemporary science such as chaos and complexity theory, differential geometry, and post-Darwinian evolution theory, DeLanda demonstrates that the virtual is not merely an effect of new image technologies, but a deeply materialist concept, not the non-real, but the more-real. In turn, his study maintains that a philosophy based on virtual power, with an emphasis on self-organizing tendencies in dynamical processes, is not only a way to understand becoming - it also gives us a more complete conception of reality. Finally, DeLanda does more than reconstruct Deleuze's ontology in scientific terms. Writing about speed, symbiosis, attractors, and affects, DeLanda develops a philosophy of generation, an art of assembly.
What we commonly refer to as the "real" is merely the experience of the actual, a realm of extensities or metric space. The geological crust on which we walk, or the bodies that do the walking, seem real enough. Yet these extensive forms are merely the temporary slowing down, or coagulation, of flows. They appear as the end-results of intensive processes animating virtual properties and generating spatial structures through abstract assembly patterns. For example, in the process of "embryogenesis" the body emerges out of intensive processes, such as cell differentiation or muscle growth, through attachment to topological points. Bodies have flexibility of size and shape but maintain the same basic structure or assembly pattern, whereas in industrial assembly, which assembles extensive forms from other extensive forms, Tab A must fit Slot A and so on. The "real" thus includes, in addition to the physically perceived, sensuous world of the actual, the virtual properties inherent in matter and the intensive processes that select and animate them. This composes a sort of passage from the virtual to the actual. The virtue of a Deleuzian ontology of processes is that, in considering "the real" alongside "the actual," it is able to explain morphogenesis (the birth of metric space and matter's generation of form), and bridge virtual and actual viewpoints.
The significance of DeLanda's work is that it radically changes the ontology of a properly materialist philosophy. In using "ontology," I will follow DeLanda's definition of it as the "the set of entities [the philosopher] assumes to exist in reality" (2). In an ontology of essences morphogenesis refers to the physical appearance of ideal forms in matter. Matter is a passive receptacle for transcendent forms inherited from a transcendent realm. It follows then that these realizations of forms, or essences, are judged upon their resemblance to transcendent norms. In a Deleuzian ontology of processes morphogenesis refers to the ability of matter's own virtual properties to generate forms that are not copies of a transcendent ideal, but the result of anexact (precise, but not metrically so) processes realized differently across different forms of materiality. The difference between essence and process may be illustrated by an example involving soap bubbles and salt crystals. An ontology of essences describes a soap bubble based on its sphericity, and a salt crystal based on its cubicity. Each has different finished forms and is thus classified in different sets. Yet each has similar morphogenetic processes in which their components try to solve the same form problem, in this case seeking "a point of minimal free energy," minimizing bonding energy for the salt crystal, minimizing surface tension for the soap bubble (15). In a virtual ontology of processes, such apparent differences are merely the result of the same assembly pattern becoming realized in different kinds of matter.
This is not merely a different way of classifying physical entities. Intensive genesis creates forms through difference, activating phase transitions when critical differences between indivisible intensive processes such as temperature, speed, and pressure, occur. For example, one can take an extensive quality like length and metrically divide it in half, thus changing its form. But if one divides a quantity of water in the same manner, it retains the same temperature. Whereas essences are measurable finished products, an ontology of intensive processes refers to non-quantifiable, productive differences in which forms become other forms at critical points, as when water turns into ice, or when cell differentiation produces organisms. A philosophy of intensive interactions thinks of difference not as the failure of one object to be another, but as an intrinsic part of any object. Deleuze's concept of multiplicity is thus crucial for DeLanda, and he devotes an entire chapter to unpacking multiplicity's "technical background," lucidly working out three mathematical concepts vital for understanding the virtual: the manifold, the vector field, and the theory of groups
The manifold originated within the differential geometry of Friedrich Gauss and Bernhard Riemann when the former tapped into differential calculus to express rates of change and find instantaneous values. Applying these techniques to geometry, Gauss made it possible to study a two-dimensional curved surface as a space in its own right, without reference to a higher dimension. Whereas in the previous Cartesian methodology, such a space had first to be plotted onto a three-dimensional space, making its coordinates definable only in relation to the unity of the higher dimension (N+1), Gauss showed how to "coordinatize" the surface itself. Riemann further radicalized this "new way of posing spatial problems" by extending it to the study of spaces of many dimensions. The immanent properties of N-dimensional surfaces, or manifolds, could now be studied without reference to a unity-imposing transcendence.
Scientists use manifolds as abstract spaces in which they model the dynamic behaviors of physical objects. First, they add dimensions representing an object's degrees of freedom. Then, through experimentation and observation, the state of the object is mapped moment-by-moment, capturing its variations at points in a modeling space. If we follow this process one instant at a time, the points become trajectories. We can describe these trajectories as asymptotically drawn toward singularities (two-dimensional manifolds) that act as attractors for matter in the system. Though never "actualized," they represent the long-term tendencies of a system not excessively affected by outside forces.
Adding the vector field allows us to see a relationship between actual and virtual spaces in a manifold. Once points are mapped, the manifold is "populated" with actual trajectories. Using this "phase portrait" to find the attractors in a system, a vector (direction) field that captures the system's tendencies is generated, and may be used to predict future changes. While the phase portrait captures everything that "actually" happened during observation, the vector field tells us everything that could, potentially, happen. Many more points, representing potential changes in the system, are included in the vector (the line of the vector is made up of a practically infinite number of points), compared to the relatively small number of actual points plotted in the observation of the system. What is original about Deleuze's ontological analysis, DeLanda shows, is the distinction made between the vector field and the phase portrait:
"despite the fact that the precise nature of each singular point is well defined only in the phase portrait (by the form the trajectories take in its vicinity) the existence and distribution of these singularities is already completely given in the vector (or direction) field" (31).
Just as the vector field contains the entire distribution of singularities, including attractors not currently attracting and singularities not yet singled out, in a Deleuzean multiplicity the virtual is "more-real," as it contains additional information that shows how a system would act under non-actualized conditions. The real exists in the virtual in addition, and prior to, its existence in the actual.
While Deleuze's analysis of the vector field expands the concept of "reality" to include all available potentials in a dynamic system, the theory of groups shows how those potentials might be affected to generate actual forms. Breaking with rigidly geometric essences in favor of topological forms, the theory of groups allows mathematicians to classify entities based on how they are affected by transformations, rather than by geometric properties. For example, a cube remains unchanged at 90-degree rotations, but not at 45-degree rotations, whereas a sphere remains invariant under either transformation. Entities in a theory of groups are defined (whether a cube or a sphere), by their levels of invariance under transformation operations such as rotating, stretching, or folding, and not by comparison with fixed ideal forms.
DeLanda's reconstruction of these concepts - the manifold's possibility for modeling processes in abstract spaces; the vector field's ability to reveal the unactualized possibilities open to dynamical systems; and the theory of groups' definition of entities - provides a way of thinking about immanence and assemblages. We have long suffered under a double denigration of reality. Philosophers, who classify the virtual as the "not real," make entities into transcendent forms given in "nature" or into purely social constructions. In either case, they provide no explanation for the becoming of forms. Scientists, on the other hand, have overlooked intensive processes, conflating the extensive with the real. This amounts to little more than mechanistic transmissions of laws applied to final states, or the adding of solutions to solutions. Even the science of thermodynamics, which has allowed us insights into intensive processes, creates an objective illusion by focusing on a system's final equilibrium state. Is it any wonder that such a world gives us industrial assembly's degradation of the assemblage, with its unadaptive rigidly metric parts, utter homogeneity, and repetitive stress disorder?
What is so vital about the Deleuzian ontology DeLanda develops, then, is the way it provides mechanisms of immanence. In other words, not only does Deleuze's virtual philosophy conjure away transcendence, it explains the resources immanent to matter that generate form. DeLanda shows that when we replace essence with multiplicity, we replace a hierarchical ontology of genera, species, and individual organisms (lacking causal connections), with a flat ontology that treats all spatial structures as individuals created through specific causal processes. The only difference between these formerly distinct classes is now one of scale. Species are merely individuals at a larger scale, as a whole species emerges out of interactions between individual organisms, much as an individual organism emerges out of cellular interactions or a city from its inhabitants and flows of goods or energy. DeLanda invites us to imagine a nested set of spaces, in which a symmetry-breaking cascade acts to "unfold spaces which are embedded into one another" (69). These spatial structures are not classified according to qualities or measurable extensities, but by their relative invariance under transformation. What is important about each spatial structure, animal, mineral, or vegetable, is its capacity to enter assemblages and remain unchanged when affected by interacting parts.
Just as a flat ontology of individuals implies a nested set of individuals at different spatial scales, so too are there nested sets of temporalities. Each temporality's scale "defines what oscillators at that level 'perceive' as relevant change" (90). Oscillators moving relatively too slowly or quickly will not appear to change, and will be unable to interact in a way that would produce assemblages. The objective effect of the existence of relative time scales is easily shown in the concepts of relaxation time and interaction time. Relaxation time refers to the time taken for a trajectory subjected to an external shock to return to its attractor (provided it is not dislodged from its basin and "connected" to another attractor). For instance, some solid materials, known as glasses, "unlike their crystalline counterparts, do not have a well-defined phase transition from the liquid state" (90). The difference between these glasses and liquids, each of which has the same "spatial arrangement of molecules," is their relaxation time, relatively short for liquids and relatively long for glasses. On a shorter observational time scale, the flow of liquid does not appear to be moving in the glass, but on a longer observational time scale, the glass would appear a flowing liquid. Observational discrepancies in such a case are not psychological tricks, but an effect of objective interaction times between time scales. Thus, the ability of different spaces to connect into assemblages, to be affected, not only depends on a space's capacities, but also the interaction of these time scales and their relative accelerations and decelerations. As Deleuze might phrase it, a space can be defined by its singularities and its capacity to be affected, but also by its speeds of becoming and capacities to become.
What DeLanda conceptualizes in combining Deleuze's ontology with contemporary science is a machinic space of temporal rhythms and spatial structures continuously interacting at various scales, gaining and losing in complexity as matter enters into and detaches from various assemblages. The term "machinic space" appears in Deleuze and Guattari's A Thousand Plateaus and refers to the way multiplicities mesh together in an immanent space. This is nearly synonymous with DeLanda's use of the concept "machinic phylum" in War in the Age of Intelligent Machines (New York: Zone Books, 1991). In Intensive Science & Virtual Philosophy, the term seems to take on added meaning, denoting not only the existence of different assembly patterns that emerge from and can connect with multiplicities bound together in a plane of consistency. In addition, DeLanda is clearly developing his own conceptual toolbox. He describes assemblages forming from temporal oscillating relationships in which different structures accelerate or decelerate their pace in order to fit, cellular collectives communicate in networks as an organism unfolds, and symbiosis leads to the emergence of new forms. The picture is one of a constantly communicating, unfolding space oscillating to the rhythm of machinic refrain. DeLanda gets rid of transcendence, but provides a structure in its place. His examples are from the "natural" world, but one cannot help but extrapolate DeLanda's explication of assembly processes to the possibility of thinking new social and political formations. This project would seem to be a necessary synthesis with his previous book, A Thousand Years of Nonlinear History, as we now have the concepts to accurately describe the material formation of systems as well as insert ourselves into those same processes of formation.
This is the synthesis I see DeLanda working toward, or at least a crucial problematic one could extract from his work. Deleuze has said that the job of philosophy is to create concepts and pose problems, not to find solutions. Intensive Science & Virtual Philosophy poses questions about multiplicities and machinic spaces with exceptional rigor, and might well model novel social or political assemblages. Another recent work that utilizes virtual ontologies, the best-selling Empire (Cambridge: Harvard University Press, 2000), mulls the problematic of a "multitude" that would push through politics. Michael Hardt and Antonio Negri define the multitude's virtual potential as DeLanda might, and there are many resonances between the two books. Empire posits the virtual power of action and labor "beyond measure" before its capture by capital, making the question of how to construct a passageway from the "possible" to the "real" essential. DeLanda is concerned with a similar passage, but what is exceptional about his work is the emphasis he places on the passage itself, the intensive processes that unfold the virtual into the actual formation of dynamic systems, which in turn affect further symbiotic transformations.
The question now may be why the intensive passageway between the virtual and the actual cannot be visualized in a philosophical work such as Empire that shares many of the same concerns found in Intensive Science & Virtual Philosophy. Why does the virtual materialize in a work focused on nonhuman forms, and not in a work with an ontology of living labor? In other words, what if we have concrete examples of Empire 's "multitude" in communicating cells or steam? This is, perhaps, why Deleuze found phase transitions so crucial. Grasped in the instant it embodies intensity, a virtual property is poised on the verge of becoming a new form. Hardt and Negri aver that, for our virtual powers of labor to become realized in multitudes, we must learn to construct a new set of machines. Political theory must "assume the language of generation." Although Intensive Science & Virtual Philosophy is not a work of political theory, here DeLanda has provided us with precisely such a language. DeLanda moves us toward a horizon-space in which, even though we still may not see the forms that will accelerate us through Empire, we might notice their intensity when they hit us.