REASONING BIASES, FORMAL LANGUAGES AND TEACHING LOGIC

Abstract

Empirical studies from the field of cognitive psychology have shown that human reasoning does not obey the prescriptions of logic. On the contrary, these studies showed that human reasoning relies on the construction of mental models and that there exist certain information processing patterns, called cognitive biases, which are linked to reasoning errors. The aim of this paper is to analyze if these human reasoning features have an impact on the apprehension of the concept of logical consequence. We conducted an exploratory study in a group of university students of an introductory course of Logic, in order to determine if there is any relation between formal languages and argument evaluation. As a result, our study showed that there exists a significant correlation between formal languages, on the one hand, and syntactic methods, on the other hand, and the ability for checking whether an argument is valid. These results would support the hypothesis that formal languages and formal methods are essential tools for teaching logic. Although pedagogical proposals that are close to naïve reasoning could be useful in the first approach to core concepts in logic, our study reveals some negative consequences related to these didactic proposals that are linked to reasoning biases activation.