Abstract
The problematization of the relations between science and politics in science and technology studies implies that scientific methods generally, and risk assessments specifically, are not neutral tools. Risk assessments have in the last years become a common arrangement for scientifically assessing the possibly adverse health effects of contaminants in food. Inspired by actor-network theory the article will explore the following: What do risk assessments do to food matters? How do risk assessments shape their objects, and what objects and knowledge do they bring into politics? The article follows the work of an interdisciplinary group of scientists in Norway who, in a charged political and cultural context, were given the task of conducting a balanced assessment of the health effects of eating fish, involving a new understanding of fish as potentially both risky and healthy food. The balanced assessment will be interpreted as involving a meeting between the epistemic cultures of toxicology and science of nutrition. It provided a way of studying what happens when a toxicological risk assessment methodology was partly applied to an issue, healthy food, which is not usually viewed in terms of risk. My argument is that the practices of risk assessments contributed to shaping fish as food in a specific way: as an issue of risk or non-risk. But the demand for balance ensured that the positive health effects of eating fish were also considered, and gave rise to a new way of conceptualizing food.