IS DEONTIC REASONING SPECIAL

M. Oaksford and N. Chater (1994), like several theorists before them, attributed cases in which selections in the Wason 4-card task match th e prescriptions of formal logic to a special kind of deontic reasoning . They proposed that a utility model of card selection governs perform ance in deontic contexts. We show that such ''logical'' performance ca n be obtained in nondeontic contexts as well, contexts in which utilit y cannot explain differential card selections. Our results also cannot be explained by Oaksford and Chater's information-gain model without violating a fundamental assumption of that model, the rarity assumptio n. Our data challenge Oaksford and Chater's framework as well as any t heory of selection-task performance that attributes logic-like perform ance to a special kind of deontic reasoning.