Abstract
In recent discussions about event as the foundation of historical research, it is William H Sewell Jr who has developed the most comprehensive version. However, Sewell’s theory of the event is not adequately articulated and is even one-sidedly dominated by his theory of structure. He did not take the problem of self-reference (and therefore circular causality) seriously enough, with the consequence that his emphasis on eventful temporality and contingency could not be carried through to the end. This article attempts to overcome such shortcomings by introducing Niklas Luhmann’s insights, which take a fully temporalized concept of event as departure, in order to build an adequate theory of event that complements the idea of eventful temporality.