the contingent nature of the connections between the elements that are brought together in an assemblage. When Deleuze asks “What is an assemblage?” he answers that it is “a multiplicity which is made up of heterogeneous terms and which establishes liaisons, relations between them,” stressing that its “only unity is that of a co-functioning … It is never filiations which are important, but alliances, alloys” (Deleuze and Parnet 2002, 69). Manuel DeLanda, in glossing this passage, stresses the radical mobility of the relations between the elements that are brought into such alliances: “a component part of an assemblage may be detached from it and plugged into a different assemblage in which its interactions are different” (2006, 10). The constituent elements of assemblages are bound together not through a lineage of shared descent, or through any intrinsic connection to the other elements with which they are coassembled, but solely through the contingent mechanisms of connection that characterize particular moments in what are constantly unfolding processes of disassembling and reassembling. While such assemblages may be of varying – and often extended – durations, assemblages are constitutively unstable.
I take the second attribute from Deleuze and Guattari’s characterization of an assemblage as, on the one hand, “a machinic assemblage” of bodies and things, and, on the other, “a collective assemblage of enunciation, of acts and statements” (Deleuze and Guattari 1988, 88; emphases original). This is not, however, a distinction between two different levels or orders, between the order of words and the order of things, or the dualities that such a distinction might subtend:
An assemblage of enunciation does not speak “of ” things; it speaks on the same level as states of things and states of content … the independence of the two lines is distributive, such that a segment of one always forms a relay with a segment of the other, slips into, introduces itself into the other. We constantly pass from order-words to the “silent order” of things, as Foucault puts it, and vice versa.(Deleuze and Guattari 1988, 87; emphasis original)
The concept of assemblage is, in this respect, a development of Foucault’s concept of the dispositif or apparatus as a combination of heterogeneous elements – texts, things, technologies, bodies – whose modes of interaction are, ontologically speaking, all of the same kind rather than being riven by a dualistic distinction between the real and its representations. The concept of the exhibitionary complex takes these considerations on board by interpreting museums as sites where texts, things, technologies, and bodies are brought together in complex relations with one another, and in accounting for the operation of the evolutionary space of representation produced by the exhibitionary disciplines in performative terms as an integrated set of bodily, mental, and visual effects. This is, however, a performativity that is conceived solely in relation to the museum’s exhibition practices. The formulations of assemblage theory have the decided advantage of allowing for a greater pliability of the relations between texts, things, technologies, and bodies that museums orchestrate, and a greater variability in the fields of effect to which this gives rise.
This is particularly so when viewed in the light of the third aspect of assemblage theory I want to comment on: its multiscalar qualities. What is an assemblage of varied elements at one scale of analysis is thus, at another scale, an element that is in its turn a constitutive component of other assemblages – and of many different assemblages at the same time. If the concept of the exhibitionary complex thus contends that the practices of public museums have to be considered in their relations to a wider set of exhibitionary institutions, it simultaneously closes off avenues of inquiry that are needed to explore how museums operate in relation to the other assemblages they have formed a part of. There are a range of different candidates here: their relations to the machineries of state education, and the intersections between museums and the cinematic apparatus, for example.8 Considered from the perspective of governmentality theory, however, museums are perhaps best considered in terms of their relations to two different kinds of governmental assemblages which operate through different mechanisms.
Let me go back to the distinction I drew attention to earlier that Foucault made in his essay on governmentality between governmental practices which work through campaigns that address the population as subjects and those which relate to population as an object that is ignorant of how it is affected by such practices. He elaborates this distinction in a couple of related lectures where he translates it into a distinction between the population as species and the population as public. He thus writes of the late eighteenth-and early nineteenth-century development of governmental forms of power:
From the species to the public; we have here a whole field of new realities in the sense that they are the pertinent elements for mechanisms of power, the pertinent space within which and regarding which one must act. (Foucault 2007, 75)
What is the difference between these two mechanisms? In the case of the public, practices of government relate to the population via educative programs and campaigns which seek to influence conduct by acting on their beliefs, opinions, fears, prejudices, and customary ways of doing things. In the case of biopower, where population is related to as species, it is the milieu that constitutes the point to which power is applied and the mechanism through which it operates where milieu is defined as “a set of natural givens – rivers, marshes, hills – and a set of artificial givens – an agglomeration of individuals, of houses, etcetera … [producing] a set of overall effects bearing on all who live in it” (Foucault 2007, 21). Both mechanisms relate to populations as subjects of wants and needs, but only governance via the public relates to population as subjects whose opinions, views, convictions, and so on, constitute the mechanisms through which they are to be governed.
Museums need to be considered in terms of their relations to both kinds of governmental assemblages, and less as self-contained knowledge/power apparatuses than as switch points in the circuits through which knowledges are produced and circulated through different networks. As such, they play a part in the distribution of the freedom through which liberal forms of government are organized, according a capacity for free and reflexive forms of self-government to some sections of the populations they connect with while at the same time denying such capacities to others. Chris Otter provides an example of this in relation to the Great Exhibition, which aimed at the improvement of the working classes via educative housing and sanitary displays that would bring this about as a consequence of their own activity. It was, however, also connected to other circuits for the distribution of knowledges which, premised on the working class’s intellectual and sensory incapacity to respond to such programs, aimed to transform working-class milieus through sanitation programs which treated the population as an object to be acted on (Otter 2008, 65–67). The history of the relations between museums and colonial practices provides more telling, because they are more sharply polarized, instances of the complex ways in which museums have operated as switch points in the flows between different networks for the production and circulation of knowledge.
Let me give two examples from my own work (Bennett 2004; 2009; 2010; 2013a) focused on the role played by museums in relation to the flow of knowledges out from museums to colonial sites of collection in the form of anthropological fieldwork expeditions; back from the field to the museum as a center of collections and calculation – a place for ordering what is collected and translating it into other forms so that it becomes capable of acting back on the world;9 and then out from the museum to connect with public spheres via the museum’s exhibition practices, or back onto indigenous populations via its connections with the institutions and practices of colonial administration. The Musée de l’Homme during its formative years, from 1928 to 1949, and the National Museum of Victoria during the first three decades of the nineteenth century, provide two cases in point, and significantly contrasting ones in view of their relations to quite different rationalities of colonial governance. Viewed in terms of its relations to the Parisian and broader French public spheres, the Musée de l’Homme translated the collections from its fieldwork expeditions to France’s African colonies into exhibitions and other forms of public pedagogy which tended, albeit imperfectly, to deracialize cultural differences by presenting all cultures as of equal value in ways which resonated with the democratic politics of the Popular Front. However, the Musée
