Adopting and Adapting: When diffusion-of-innovation models meet the sociology of translation

We discuss the use of simulation models as tools for rethinking innovation in a way more compatible with Science and Technology Studies (STS) than that encoded in standard models of the diffusion of innovation. In the standard conception of innovation, often dubbed the "linear model", innovation processes are divided into phases, including an origin and a diffusion phase. Key components in this conception include word-of-mouth epidemic models with their characteristic S-curve shaped growth data. However, studies in the literature of sociology of translation, or Actor-Network Theory (ANT), and the social construction of technology (SCOT) have drawn attention to how there is a complexity among the determinants of the success of a particular technology, product or project. To incorporate this very different conception of innovation diffusion we employ computer simulations, but rather different ones to those traditionally used. Innovation is represented less as a disease epidemic and more as multiple problems in constraint satisfaction, tackled in parallel in socio-technical networks. Mathematical analyses, curve fitting and forecasts are seen to be of little use. Qualitative data and qualitative model outcomes become more important, as we seek to communicate better an understanding of innovation. Issues raised under this revised understanding include ownership of innovation and the rewards for an identified "innovator", the degree to which would-be users / adopters should be constrained in their manipulation and reinterpretation of an innovation, and the extent to which organisations can plan for innovation. Our models, used as tools for rethinking innovation, bring together work on complex adaptive systems and that of STS. We invite further discussion on how best qualitative studies can be combined with simulation modelling.

Tuesday, 2 November, 2010 - 12:30 to 14:00
Christopher Watts