In the last three decades social scientists had to learn that assumptions concerning the robustness and stability of social processes and structures cannot be held as self-evident any longer since the outcomes of social actions are far more unpredictable than is usually assumed and that events and their surprising occurrence have to be taken into account in making sense of social reality. If this is so, then social theory has to rethink some of its basic methodological and theoretical assumptions. The paper will address the following questions: Which theory of action do we need in order to come to terms with the contingent features of social life? Which types of accounts can social scientists give in order to grasp this very reality? Can social scientists still speak self-confidently of laws and/or general social mechanisms or must they rather rely on narrative methods? The main answer given to these questions will be that social theory indeed will have to take on the problem of “narrativity” which at the same time also means that main stream social theory will necessarily have to thoroughly historicize all its categories.