s. 3-4:

In an important early paper Callon and his co-author Bruno Latour under- stood the concept of the actor in semiotic terms/. Drawing on the work of Greimas (1987) they de ned an actor (actant) as ‘whatever unit of discourse is invested of a role’ (Callon and Latour 1981: 301–2).6 In this way an actor could be either a human or a non-human entity and it could have both material and ‘social’ components (Callon 1986). Its constitution as an actor depended on its shifting network of connections with, and differences from, other entities. Rather than assume a distinction between the (social) actor and the economic and cultural environment within which the actor and its actions were embedded, Callon and Latour’s approach took the identity of actors as always bound up with the networks of which they were a part. Instead of speaking of actors and their networks as if they were distinct objects, this approach suggested that it would be more accurate to speak of ‘actor-networks’. Actors were able to have the power to act in so far as they were elements of a network. The notion of the actor and the idea of the network simply referred to different dimensions of the same thing. In effect, the formation of actor-networks generated speci c worlds of actors. The identity, capacity and strength of an actor were relational.

For Callon the semiotic concept of the actor-network should not just be applied to the study of written texts (Callon et al. 1986). The concept always implied a form of material ordering. As a concept it had some similarities to Foucault’s concept of apparatus (dispositif ). It described the constitution of a reality forged from heterogeneous elements (materials, texts, bodies, skills, interests, experimental devices) (cf. Deleuze 1988). The process of innovation was a process of ‘heterogeneous engineering’ (Law 1994; Callon and Law 1995). However, there were signi cant differences between the actor-network theorists and Foucault. For, whereas Foucault identi ed paradigmatic historical forms of apparatus, Callon and others were always interested in the speci city of particular socio-technical arrangements. Moreover, whereas Foucault’s approach bracketed the question of the contestability and mutability of particular apparatuses, actor-network theory put the question of the stability or instability of the network to the fore. In effect, the Foucaultian notion of apparatus suggested something too static and too ‘structural’ for actor-network theory. Innovation was the privileged object of analysis; micro-sociology was the most appropriate method.

In accounting for the dynamism of actor-networks, Callon and Latour had earlier adapted Michel Serres’ (1974) concept of translation: ‘By translation we understand all the negotiations, intrigues, calculations, acts of persuasion and violence thanks to which an actor or force takes, or causes to be conferred on itself, authority to speak or act on behalf of another actor or force’ (Callon and Latour 1981: 279). The notion of translation emphasized the way in which the identity of actors, and their relations, was always in process.7 But it also implied that translation was a political process in which politics was conceived not so much in terms of competing ideologies or interests, but as a calculated Machia- vellian act. Seen in these terms, the process of technical change could not be explained by reference to the kinds of social, political and economic interests which determined it. Rather, technical change was itself a form of politics that both revealed and translated the identity of social and economic actors. Tech- nical change was conceived as a kind of experimentation, which in effecting a transformation of the world also revealed a dimension of what it was. Callon himself made the point through a discussion of a case study of the development of the electric vehicle in France. Criticizing the lack of attention to ‘material’ devices and artefacts in the sociology of Bourdieu and Touraine, Callon demon- strated the weakness of their approach in accounting for the lack of success of the electric vehicle in effecting a transformation of French society. Bourdieu and Touraine, in the sociological tradition established by Durkheim, made a clear distinction between the ‘social’ and the ‘material’. Yet, ‘How can social elements be isolated when an actor-network associates the spin of an electron directly with user satisfaction? How can any interpretation of social interaction be established when actor-networks constantly attempt to transform the identities and sizes of actors as well as their interrelationships?’ (Callon 1987: 99).