Abstract

reasoning lies through the career of similarity. There is evidence for effects of specific content in deductive reasoning . A large number of studies based on Wason's (1968) classic selection task have shown that while normal subjects typically fail in the application of conditional inference rules to abstract and unfamiliar material (in particular, they fail to apply Modus Tollens), their performance can be greatly improved if concrete and far 11 miliar material is used . In the selection task, subjects see four cardse.g., "A," "C, 17 &'4," and "3"and are told to turn over the necessary cards to test the following rule : "If a card has a vowel on one side, then it has an even number on the other side." Only a small number (about 10%) choose the correct two cards ("A" and "3") ; the majority choose cards "A" and "4," suggesting that the participants, rather than following a rule of inference to test the conditional statement, were simply basing their choice on surface matches between the rule as stated and the cards (i .e., choosing the card that matched the antecedent (the vowel) and the card that matched the consequent (the even number) . However, it appears that performance is considerably better when the ,rule is stated in terms of a familiar domain in which the contingencies are accessible to the subjects . Wason & Shapiro (1971) used a thematic rule instead of an abstract conditional rule ("Every time I go to Manchester I travel by car") and a set of cards representing various destinations and modes of transport . They obtained 62% correct selection . Griggs & Cox (1982) found substantial improvement by using a familiar rule such as "If a person is drinking beer, then the person must be over 19." Johnson-Laird, Legrenzi and Legrenzi s) (1972) found improvement (to 81% correct) using a familiar postal rule ("If a letter is sealed, then it has a 50 lire stamp on it") and a set of envelopes (sealed and unsealed, with a 50 lire or a 40 lire stamp) . When the rule concerned an arbitrary relation between symbols, instead of a realistic relation between concrete objects, the subjects relapsed to the