Premise Interpretation in Conditional Reasoning

2 Premise Interpretation in Conditional Reasoning Conditional sentences are pervasive in human communication. Many arguments contain conditional premises or a conditional conclusion. The expression conditional reasoning could appropriately be applied to such arguments. However, there is a tradition in psychological research, which will be followed here, to restrict the use of this expression to refer to two-premise arguments that have one conditional premise (called the major), and a second premise (called the minor) made of either the antecedent of the conditional or its negation, or the consequent of the conditional or its negation. This leads to the following four arguments: Two deductively valid arguments: Modus Ponendo Ponens (henceforth MP): if P then Q; P; therefore Q. Modus Tollendo Tollens (MT): if P then Q; not-Q; therefore not-P. Two deductively invalid arguments: The fallacy of affirming the consequent (AC): if P then Q; Q; therefore P. The fallacy of denying the antecedent (DA). if P then Q; not-P; therefore not-Q. Conditional reasoning is, on par with Wason's selection task, the most investigated paradigm in the psychology of reasoning. The general trend of performance with formal material (in which arguments are presented in their 3 symbolic form) seems robust, even though there is some variability due to factors such as instructions and response format. The conclusion of MP is endorsed by nearly eveybody, that of MT by around two thirds of the participants only, while both fallacies are committed by around one half of them (for a review, see Evans, 1982; Evans, Newstead & Byrne, 1993). It is hard to have as clear a picture of performance with thematic material, that is, material in which symbols P and Q are instantiated by short meaningful sentences. In some early studies (e. Staudenmayer, 1975) it was already observed that the rate of endorsement was a function of the semantic content of the sentences; but this observation was rather incidental. More recent studies of conditional reasoning, with planned observations, have shown that the effects may be of considerable magnitude; they will be reviewed below. These results are not surprising given that investigations of conditionals have regularly shown differences in given greater emphasis in the context of research on the thematic version of Wason's selection task where they soon became the main center of interest. "Content effect" is the name traditionally given to the variation of responses and consequently of reasoning performance as a function of …