Social science has for the most part lost its ambition to be ‘science’, as shown in the recent change in the American Anthropological Association statement of purpose. The new term is expertise. The change points to something fundamental: social science methods are now largely stable; they have well-developed uses for public and policy audiences; because they are user-friendly they are unlikely to radically change, and new problems arise for them to be applied to. New concepts are developed, but they do not develop into ‘theories’ in a scientific sense. So what is expertise? It is something other than fact generation: it is the result of aggregating and assessing research in an area. Experts engage with one another, so there is a community ...