Abstract
This article argues that an ethic applying the technology assessment (TA)-method is only feasible as a risk ethic, since the consequences of technical action are ambivalent and uncertain. It first distinguishes possible strategies of justification for a risk ethic, that is (a) deontological, (b) teleological and (c) procedural approaches. On the basis of the critique of both (a) and (b), a central problem for the integration of discourse ethics in the TA-method is highlighted by reverting to a game theory modeling: As it turns out, mechanisms of self-selection undermine the representativity of discourses and thus its legitimation. In view of this dilemma, the article pledges for a complementary approach selectively integrating (a)-(c) into a “provisional moral”