On cognitive busyness: When person perceivers meet persons perceived.

Person perception includes three sequential processes: categorization (what is the actor doing?), characterization (what trait does the action imply?), and correction (what situational constraints may have caused the action?). We argue that correction is less automatic (i.e., more easily disrupted) than either categorization or characterization. In Experiment 1, subjects observed a target behave anxiously in an anxiety-provoking situation. In Experiment 2, subjects listened to a target read a political speech that he had been constrained to write. In both experiments, control subjects used information about situational constraints when drawing inferences about the target, but cognitively busy subjects (who performed an additional cognitive task during encoding) did not. The results (a) suggest that person perception is a combination of lower and higher order processes that differ in their susceptibility to disruption and (b) highlight the fundamental differences between active and passive perceivers. Many of us can recall a time when, as students, we encountered a professor at a party and were surprised to find that he or she seemed a very different sort of person than our classroom experience had led us to expect. In part, such discrepant impressions reflect real discrepancies in behavior: Professors may display greater warmth or less wit at a party than they do in the classroom. However, just as the object of perception changes across situations, so too does the perceiver. As passive perceivers in a classroom, we are able to observe a professor without concerning ourselves with the mechanics of social interaction. At a party, however, we are active perceivers , busy managing our impressions, predicting our partner's behavior, and evaluating alternative courses of action. Of all the many differences between active and passive perceivers, one seems fundamental: Active perceivers, unlike passive perceivers, are almost always doing several things at once ( Gilbert, Jones, & Pelham, 1987 ; Gilbert & Krull, 1988 ; Jones & Thibaut, 1958 ).