The riddle of things: Activity theory and actor‐network theory as approaches to studying innovations

Abstract

This article compares cultural‐historical activity theory (AT) and actor‐network theory (ANT) as approaches to studying technical innovations. The concept of nature and society production in the ANT and the concept of activity in the AT have much in common as attempts to transcend the dualism between subject and object, nature and society. The symmetrical (ANT) and the dialectical (AT) interpretations of the concept of mediation are compared. It is suggested that the historically developed, artifact‐mediated structure of human activity is instrumental in studying interaction and coevolution of social and material entities. Three limitations of the concept of generalized symmetry, or symmetrical mediation, become evident when the concept is used in empirical studies of innovation: First, it does not supply any criteria for defining the nature and scope of actors in a heterogeneous network. Second, it leads to an asymmetrical, Machiavellian analysis of innovation in which the contribution of designers, users, and nonhuman entities remains marginal. Third, it does not provide any explanation for the intentionality and competence of humans. It is suggested that these problems can be solved if the innovation network is studied as a network of activity systems. Nonhuman entities are included in the analysis as historically developed arrays of tools and raw materials of the activity systems. This approach is elaborated by analyzing an unsuccessful innovation process, the production of ethanol from wood through the use of cellulose‐degrading enzymes. It is suggested that instead of applying the symmetrical semiotic language proposed by ANT in the analysis, a dialogue that utilizes the historically developed resources and languages of different thought communities is needed.