Open and closed models of understanding in conditional reasoning

Conditional reasoning is studied by assuming that subjects reason according to logical rules, while understanding premises in different ways, depending upon the circumstances. The process of understanding will lead to an internal propositional model of the premises and the task which can be more or less closed. In a closed model alternative antecedents and consequents are disregarded. This is not the case in an open model. Two studies which employed constructed conditional reasoning tasks with several premises were used to explore this hypothesis. In the first, performance i in a judgement task was studied. Results showed that the subjects judged the conversion of an implication to be true to a significantly greater extent when premises were consistent with a closed model than when they were not. In the second, the reasoning process was investigated in a conclusion task with a thinking aloud instruction. Several different interpretations of the premises were found. So, for instance, negatives often were transformed to a positive form, either directly or by invoking alternatives within the text. Both these interpretations lead to a closed model.