Interaction of arguments and values. Bridges between Artificial Intelligence and the Psychology of Reasoning

Abstract

The argumentation models proposed from Artificial Intelligence offer simplicity and precision to analyze the acceptability of an argument in interaction with others. However, there are doubts when considering their correctness, since the character, more dialectical than logical, of the argumentation prevents having a formal semantics with which to relate it. Here we will discuss the value-based argumentation models by Gabbay and Bench-Capon. Gabbay, for instance, seeks to implement the intuition that confronting arguments that promote the same value (religious, political, legal, etc.) is more effective than doing it from a different, unshared value. Using some examples taken from the literature, I will show the importance of building bridges between the models and the empirical that enable to contrast such intuition. I will argue that there are both conceptual and representational problems that need to be addressed, and I will point out some lines of experimental research in these directions.