Although the concept of an event is widely used as the basic unit in the organization of experience, memory and meaning, little attention has been paid to how events emerge or what determines the boundaries of an event. It is usually taken for granted that one knows what an event is or how events are demarcated. In this paper an explanation is offered for the emergence of events, the cut hypothesis, which states: ''A sub-sequence of stimuli is cut out of a sequence to become a cognitive entity if it has been experienced many times in different contexts'', and three experiments to demonstrate the predictive power of the hypothesis are described. The stimuli in all three experiments were video films, constructed by randomly assembling short excerpts from movies. In the first experiment the cut hypothesis was juxtaposed with the thesis of demarcation at major changes, and it was shown that, after experiencing a certain repeating sequence, subjects hardly considered dividing at an internal point, even if it was a point of maximal change; points of maximal change were determined on the basis of performance by control subjects who did not experience the repeating sequence. In the second experiment the cut hypothesis was juxtaposed with an associationistic explanation; it was shown that subjects who viewed a certain sequence repeating in variable contexts recognized it better than subjects who had viewed the same sequence repeating always in the same context. In the third experiment a prediction of the hypothesis on recall behaviour was tested and it was shown that experience with sequences of stimuli repeating in various contexts results in cohesion of their elements.
