Studies in social psychology inadvertently call for a subject's reconstruction of past behaviors when using interviews, questionnaires, or personality inventories. Since subjects' past behaviors are unobservables, subjects reconstruct their past retroductively. However, since the behaviors are not perceptually observed, such inquiry is decontextualized and probabilistic. Hence, reconstructions are frequently organized in terms of commonsense plausibility and personal accountability rather than causality. It is proposed that such inquiry may be improved by having subjects not only endorse preformatted material, but also by (...) providing warrants for subjects' endorsements. The provision of warrants has been shown to structure past reality perceptions of laypersons . The provision of warrants not only recontextualizes recall of past behaviors but also discloses historical continuity which, in turn, suggests higher probabilities of future behaviors. (shrink)

Discursive social psychology is used here to study the reconstruction of an event, a homicide, by lay people. Fourteen propositions are outlined to guide discourse analysis, since the epistemological basis of such analysis is somewhat different from that of formal experimental inquiry. An actual discourse is then analyzed, with special emphasis on the evidence used to support the final conclusion of guilt.