Part I of this book rests on a key premise: that technical artefacts are a distinct type of entity and that their distinct ontology matters. But what if this is wrong? Do we really need to ask what technologies do to properly examine their role(s) in harms? This chapter examines two perspectives that respond in the negative: actor- network theory and Bourdieu's (1977) theory of practice. These are, in many respects, radically different approaches. While actor- network theory is often lauded for bringing material objects (back) in to sociological inquiry, Bourdieu's theory of practice is notable for its almost complete neglect of material things – a neglect actor- network theory's star player, Latour (2005), was at pains to emphasize. While ostensibly these theories could not be more different, they do, I argue, share some common ground in how they approach material artefacts. Both argue that technology is constituted in practice and cannot be understood outside of it. For the actor- network theorist, technology is – like every other entity that makes up an actor- network – performatively constituted by the network itself. And for Bourdieu, like some other practice theories, technical artefacts are constituted in practice; their effectivity resides only in the actions of humans.

As I show in this chapter, these arguments have at least three major limitations. First, they preclude us from examining the distinct contributions that artefacts, people and other entities make to a harmful event. Second, they prevent us from distinguishing between different types of relations with technologies. And third, they render artefacts neutral by denying them durable dispositions or powers to which we can attribute any degree of causal responsibility for harmful events.