Priming exceptions: A test of the scope hypothesis in naturalistic trait judgments

Trait judgments draw on two kinds of memory: (a) trait summaries, which provide information in the form of a generalization, and (b) memories of episodes in which a person behaved in ways that are relevant to the trait. According to the scope hypothesis (e.g., Cosmides & Tooby, 2000; Klein, Cosmides, Tooby, & Chance, in press), a trait summary is most useful when its scope is delimited (i.e., when it is accompanied by information specifying those situations in which it does not apply). Episodic memories that are inconsistent with a trait summary can serve this function, because they encode specific situations in which the generalization fails to predict the outcome. This suggests that judgment procedures should be designed to search for summary information in semantic memory and, upon retrieving it, also search for episodic memories that are inconsistent with that summary. This prediction has been tested and supported in previous experiments using artificial target persons (Babey, Queller, & Klein, 1998). Herein, we present the findings from two experiments supporting this prediction using trait judgments about real people for whom subjects have real world knowledge: the self (Experiment 1) and one's mother (Experiment 2). The experiments also test a subtle prediction of the scope hypothesis: that a trait summary must exist and be retrieved for trait-inconsistent episodes to be primed. The results show that in the absence of a trait summary, trait-inconsistent episodes are not primed, but trait-consistent ones are.