Based on a conversational analysis of experimental procedures and consistent with the principle of relevance, we predicted that participants' verbal responses will be in¯uenced by their tacit inferences about the researcher's epistemic goals, derived from their knowledge of the researcher's academic aliation. We tested this prediction in a core area of social-personality and cultural psychology, causal attribution. University students provided causal attributions about mass murder cases, while the questionnaire identi®ed the researcher either as a social scientist or a personality psychologist. The results indicated that attributions were overall more situational than dispositional, and as predicted, this main eect was quali®ed by an interaction between conversational cue and type of attribution. Thus, participants gave relatively more situational explanations when the letterhead of the questionnaire identi®ed the researcher as a social scientist compared to when the researcher was identi®ed as a personality psychologist. The reverse pattern emerged for dispositional attributions. Methodological and conceptual implications are discussed.