Policy and technology actors seem to focus “naturally” on risk rather than on technology’s social and ethical impacts that typically constitute an important focus of concern for philosophers of technology, as well as for the broader public. There is nothing natural about this bias. It is the result of the way discourses on technology and policy are structured in technological, liberal, pluralistic societies. Risks qualify as “hard” (i.e., objective, rational, neutral, factual), other impacts as “soft” (i.e., subjective, emotional, partisan, value-laden) and are therefore dismissable. To help redress this bias, it is necessary to understand how this distinction between hard and soft impacts is construed – in practice and in theory. How are expected (desired, feared) impacts of technology played out in expert-citizen/consumer interactions? we first discuss online patient deliberations on a future pill for celiac disease (“gluten intolerance”) promising to replace patients’ lifelong diet. By “rejecting” this pill, patients displayed concerns about how the new technology would affect their identity, and the values incorporated in the way they had learned to handle their disease. Secondly, we analyze how experts construct a consumers’ concern with “naturalness” of food: as a private – and invalid – preference that requires no further debate. The point of the analysis is to make available for discussion and reflection currently dominant ways to demarcate public and private issues in relation to emerging technologies, including the accompanying distributions of tasks and responsibilities over experts and laypersons. However, the actors themselves cannot simply alter these demarcations and distributions at will. Their manoeuvring room is co-shaped by discursive structures at work in modern, technological, pluralist, liberal societies. In the third section, we therefore identify these structures, as they provide the hegemonic answers to the three key questions with regard to the possible impacts of emerging technologies: how are impacts evaluated; how are they estimated; and how are they caused? we conclude with some suggestions for further research.