Introduction

Writing a book about the concept of assemblage presents various challenges. The easiest one to meet is terminological. The word in English fails to capture the meaning of the original agencement, a term that refers to the action of matching or fitting together a set of components (agencer), as well as to the result of such an action: an ensemble of parts that mesh together well. The English word used as translation captures only the second of these meanings, creating the impression that the concept refers to a product not a process. If this were the only challenge it could be easily bypassed. We could simply take the term agencement to be the name of the concept, the concept itself being given by its definition. But this way out is blocked by the fact that the concept is given half a dozen different definitions by its creators, Gilles Deleuze and Félix Guattari. Each definition connects the concept to a separate aspect of their philosophy, using the terms that are relevant for that aspect, so when taken in isolation the different definitions do not seem to yield a coherent notion. This book is an attempt to bring these different definitions together, introducing and illustrating the terms required to make sense of them. We can begin with the simplest definition, one involving a minimum of additional conceptual machinery:

What is an assemblage? It is a multiplicity which is made up of many heterogeneous terms and which establishes liaisons, relations between them, across ages, sexes and reigns – different natures. Thus, the assemblage’s only unity is that of a co-functioning: it is a symbiosis, a ‘sympathy’. It is never filiations which are important, but alliances, alloys; these are not successions, lines of descent, but contagions, epidemics, the wind.1

In this definition, two aspects of the concept are emphasised: that the parts that are fitted together are not uniform either in nature or in origin, and that the assemblage actively links these parts together by establishing relations between them. The contrast between filiations and alliances gives us a clue regarding the type of relationships needed to hold the parts together. Some relations, such as that between parents and their offspring, or those between brothers or sisters, define the very identity of the terms that they relate. One can only be a father if one is related genealogically to a son or a daughter, and vice versa, so that the identity of the role of father, or of that of son or daughter, cannot exist outside their mutual relation. Traditionally, these relations are designated as relations of interiority. On the other hand, when two groups of people related by descent enter into a political alliance, this relation does not define their identity but connects them in exteriority. It is a relation established between the two groups, like the air that exists between them transmitting influences that connect them but do not constitute them.2 The terms ‘interiority’ and ‘exteriority’ are somewhat misleading because they suggest a spatial relation, a relation internal or external to something. A better choice would be intrinsic and extrinsic, but the intent is clear: if a relation constitutes the very identity of what it relates it cannot respect the heterogeneity of the components, but rather it tends to fuse them together into a homogeneous whole.

The majority of relations in the world are extrinsic. Intrinsic relations tend to be confined to niches, such as social roles defined by conventions. For there to be a convention, there must be alternative ways in which the identity of a social role is defined, and the choice among the alternatives must be arbitrary. Family relations, for example, vary across cultures, as do the rights and obligations attached to the role of parents, offspring, and relatives. So the fixing of one of these alternatives by an arbitrary social code constitutes its very identity. A similar situation arises in biology with respect to the roles that organisms of the same species play relative to one another. When the behaviour of an organism is not learned but is rigidly coded by its genes, and when there exist alternative behavioural patterns that could have performed the same function, its identity can be considered to be determined by relations of interiority. Hence Deleuze’s attraction to the ecological relation of symbiosis, as in the relation between insects and the plants they pollinate, because it involves heterogeneous species interacting in exteriority, and their relation is not necessary but only contingently obligatory, a relation that does not define the very identity of the symbionts. In both the social and biological cases, intrinsic relations are such because they are coded, and because the code arbitrarily selects one alternative over the rest. This suggests that the opposition between the two types of social ensembles mentioned in the previous quotation, those linked by filiation and alliance, respectively, may be captured by a single concept equipped with a variable parameter, the setting of which determines whether the ensemble is coded or decoded.

Chapter 1 explores this possibility. In their exposition of assemblage theory Deleuze and Guattari tend to use a series of oppositions: tree/rhizome, striated/smooth, molar/molecular, and stratum/assemblage. But they constantly remind us that the opposites can be transformed into one another. In particular, the kinds of ensembles designated as ‘assemblages’ can be obtained from strata by a decoding operation.3 But if one member of these dichotomies can be transformed into the other then the oppositions can be replaced with a single parametrised term capable of existing in two different states. This yields a different version of the concept of assemblage, a concept with knobs that can be set to different values to yield either strata or assemblages (in the original sense). The coding parameter is one of the knobs we must build into the concept, the other being territorialisation, a parameter measuring the degree to which the components of the assemblage have been subjected to a process of homogenisation, and the extent to which its defining boundaries have been delineated and made impermeable. A further modification to the original concept is that the parts matched together to form an ensemble are themselves treated as assemblages, equipped with their own parameters, so that at all times we are dealing with assemblages of assemblages. Using this modified version of the concept, Chapter 1 goes on to detail a materialist social ontology in which communities and organisations, cities and countries, are shown to be amenable to a treatment in terms of assemblages.

Chapter 2 uses this social ontology as a context to discuss the assemblage approach to language. As is well known, Deleuze and Guattari were highly critical of orthodox linguistics, and adopted ideas from sociolinguistics to study language in its communal and institutional context. A tightly knit community, for example, is an ensemble of bodies (not only the biological bodies of the neighbours, but also the architectural bodies of their houses, churches, pubs, and so on) in which the fitting together is performed by linguistic acts that create social obligations among the neighbours. A promise between community members must be kept, else the reputation of the member breaking it will suffer, and he or she may be punished with ostracism. Similarly, a military or corporate organisation is an ensemble of bodies (including the bodies of their weapons and machines) in which commands create the bonds that fit them together: after being commanded to do something a subordinate is held responsible for the fulfilment of the command, and punished for disobeying it. Social ensembles held together by enforceable obligations are referred to by the authors as ‘collective assemblages of enunciation’.4 However, their discussion of this important concept is hampered by a social ontology that includes only three levels: individuals, groups, and the social field. A more finely grained ontology, with many levels of social ensembles between the person and society as a whole, will help us to clarify and extend their ideas about language.

Chapter 3 concentrates on a specific social organisation, the army. One of the earliest illustrations of an assemblage was the warrior–horse–bow ensemble of the nomads.5 This assemblage can become a component part of a larger one, a nomad army, while its own components can also be treated as assemblages: the bow as an ensemble of a flexible arc, a string, and a projectile.6 And similarly for a Second World War army as an assemblage of platoons, themselves composed of soldiers, their rifles, and their radios. Armies are therefore a perfect example of a nested set of assemblages. In the history of armies we can detect transformations that add to their flexibility or that, on the contrary, make them more rigid and obedient. These transformations can be modelled by setting the parameters of the assemblage to the right settings, a task that is greatly simplified if an assemblage’s components have parameters of their own. This way, we can locate the right level in the nested set at which the coding or territorialisation occurred, and do justice to the complexity of the historical record.

Chapter 4 discusses scientific fields, viewed as assemblages of a domain of laboratory phenomena, a community of practitioners, and the techniques and instrumentation that fit one to the other. Unlike other approaches, in which the cognitive items governing scientific practices (concepts, statements, problems, explanations, classifications) are viewed as related to one another in interiority, forming a monolithic paradigm from which there is no escape short of a religious conversion, in this chapter we explore the idea that cognitive tools are not fused into a totality but rather coexist and interact in exteriority. The distinction between strata and assemblages in this case corresponds to what Deleuze and Guattari refer to as major and minor sciences. An example of a major science is an axiomatised version of classical physics, in which immutable truths about nature’s laws are used as self-evident axioms, while deductive logic is used to derive in a uniform way an unlimited number of further truths (theorems). An illustration of a minor science would be chemistry, a field that resists axiomatisation and in which the phenomena in the domain continuously confront partitioners with variation, even as they strive to find constants, posing new problems for them to solve.7 In this case too, we can replace the dichotomy major/minor by a single concept, while deriving the very real distinctions discussed by the authors from the settings of the parameters. And as in the case of armies, scientific fields can also be treated as assemblages of assemblages, allowing us to locate at the right level of the nested set the changes brought about by the conditions created by the settings of the parameters

Chapter 5 tackles the most difficult notion in this approach: the diagram of an assemblage. An ensemble in which components have been correctly matched together possesses properties that its components do not have. It also has its own tendencies and capacities. The latter are real but not necessarily actual if they are not currently manifested or exercised. The term for something that is real but not actual is virtual. An assemblage’s diagram captures this virtuality, the structure of the possibility space associated with an assemblage’s dispositions. But in addition to defining a virtual space already caught up into actual ensembles, trapped into their materiality and expressivity to a degree specified by the parameters, the diagram connects an assemblage with other diagrams, and with a cosmic space in which diagrams exist free from the constraints of actuality. While the ontological status of dispositions that are not currently being manifested is controversial, the existence of the cosmic plane is clearly much more so. Nevertheless, this chapter strives to show that both are compatible with a materialist metaphysics.

Chapter 6 deals with another metaphysical question: all assemblages should be considered unique historical entities, singular in their individuality, not as particular members of a general category. But if this is so, then we should be able to specify the individuation process that gave birth to them. In the previous chapter we had already begun to use examples of assemblages belonging to the natural world, proof that the approach is not confined to social assemblages, an emphasis that continues into this chapter. The individuation processes behind physical atoms and biological species are used as illustrations. Chapter 7, finally, returns to the question of the virtual diagram of an assemblage, but this time to connect this notion to epistemological rather than ontological questions. It is the most technical chapter, because a rigorous discussion of diagrams must proceed using concepts from mathematics, but it introduces all the necessary notions in clear technical and historical detail.

What gets lost in this new version of the concept of assemblage? Not much. The rich descriptions made by Deleuze and Guattari of rhizomes versus trees, or of molecular flows versus molar aggregates, or of smooth spaces versus striated ones, are all recoverable as descriptions of qualitatively different phases of one and the same entity, making these renditions every bit as useful as a detailed portrayal of the differences between a substance in the liquid and crystalline states. Hence, the change in this respect boils down to a matter of emphasis: using strata and assemblages as distinct categories allows one to stress their very important differences, even if it complicates the discussion of their mutual transformations.8 The other change, conceiving of the components of an assemblage as themselves assemblages, is also harmless, as is the idea that the environment of an assemblage is itself an assemblage. The authors introduce further categories of being to define the different kind of components, like the category ‘bodies’ for the working parts of ‘machinic assemblages’, and use words like ‘conditions’ to define the larger context in which an assemblage operates.9 But their tendency to view the world (natural and social) in terms of two (or three) levels makes the expression ‘the larger context’ particularly dangerous, since it ends up engulfing what in reality is a multi-level ontology. Hence, replacing bodies (and other component types) and contextual conditions by smaller and larger assemblages, respectively, allows us to sidestep this difficulty. It also yields a view of reality in which assemblages are everywhere, multiplying in every direction, some more viscous and changing at slower speeds, some more fluid and impermanent, coming into being almost as fast as they disappear. And at the limit, at the critical threshold when the diagrams of assemblages reach escape velocity, we find the grand cosmic assemblage, the plane of immanence, consistency, or exteriority.

Notes