Abstract
How does knowledge production during crises challenge or maintain contemporary institutions? This article investigates expert knowledge on nuclear accidents in France in the aftermath of the Fukushima Daiichi accident in March 2011. It examines the French and European ‘lessons from Fukushima’ learning exercise, using document analysis, interviews, and ethnographic observation of institutional experts’ working practices. The article shows how nuclear safety experts produce limited commensurability. Specifically, through a comparative exercise, they render certain features of the Fukushima accident commensurable by relating them to a diversity of pre-existing evaluation scales and metrics of nuclear safety, while preventing different hypotheses, methods, and data from being brought to bear on nuclear safety. These operations rely on and consolidate experts’ epistemic leeway, their discretionary ability to choose from several incommensurable epistemic resources. The article enhances our understanding of the politics of (in)commensurability for expert communities in the context of the pluralization of expert systems. It opens up questions about knowledge on crises, seeing these crises as episodes in which experts redefine the acceptable states of the world.