A Computational Logic Approach to Human Syllogistic Reasoning

A recent meta-analysis (Khemlani & Johnson-Laird, 2012) about psychological experiments of syllogistic reasoning demonstrates that the conclusions drawn by human reasoners strongly deviate from conclusions of classical logic. Moreover, none of the current cognitive theories predictions fit reliably the empirical data. In this paper, we show how human syllogistic reasoning can be modeled under a new cognitive theory, the Weak Completion Semantics. Our analysis based on computational logics identifies seven principles necessary to draw the inferences. Hence, this work contributes to a computational foundation of cognitive reasoning processes.