Until recently the study of written genres focused on textual patterns. When researchers wanted to examine a particular genre, they looked across multiple texts for regularities of form and effect. Over the last decade, scholars in composition studies have been reinterpreting genre as social action: a complex pattern of repeated social activity and rhetorical performance arising in response to a recurrent situation. This reinterpretation presents a dilemma for the researcher who wishes to observe a particular genre in a specific setting. When conceived as social action, what, in addition to texts, are the observable constituent elements of a genre? And what are the relationships among the elements?