The new psychology of reasoning: A mental probability logical perspective

摘要

Mental probability logic (MPL) has been proposed as a competence theory of human inference. MPL interprets indicative conditionals as conditional events. While recent probabilistic approaches assume an uncertain relation between the premises and the conclusion, the consequence relation remains deductive in MPL. The underlying rationality framework of MPL is coherence based probability logic. I discuss cases in which the psychological predictions of MPL on human reasoning about conditionals diverge from psychological predictions based on traditional approaches to probability in the context of zero antecedent probabilities. Specifically, I reconstruct a paradox of the material conditional in purely probabilistic terms. Moreover, I report a new experiment on a generalised version of the probabilistic truth table task, which investigates the interpretation of conditionals under incomplete probabilistic knowledge. The data show that-during the course of the experiment-most people shift their interpretation and converge on the conditional event interpretation of conditionals. Finally, I illustrate how the data of this experiment and of experiments on a paradox of the material conditional confirm the psychological predictions of MPL.