Jones and Thibaut (1958) argue that information gleaned from social interaction depends on the perceiver's current objectives. Five experiments examined how narrative construction and impression formation objectives influenced interpretation of social behaviors. In Experiments 1 and 2, participants read a series of sentences describing persons performing various behaviors and indicated the cause of the behaviors. Results indicated a reliable tendency for all participants to explain behaviors using situational attributions rather than chronic dispositions; however, this tendency was enhanced for participants instructed to construct a narrative using the behaviors. Experiments 3 and 4 found that participants given impression formation objectives were more likely than those with narrative construction objectives to generalize observed behaviors to future situations. Experiment 5 examined how attributions were rendered without explicit processing goals. Implications for social perception and judgment are discussed.