Explanation-based Belief Revision: Moving Beyond Minimalism to Explanatory Understanding

The document challenges the principle of minimalism in belief revision, proposing the explanatory hypothesis that suggests people prioritize comprehensive explanations over minimal changes when revising their beliefs, and provides empirical evidence supporting this hypothesis through human-subject studies.

Abstract

In belief revision, agents typically modify their beliefs when they receive some new piece of information that is in conflict with them. The guiding principle behind most belief revision frameworks is that of minimalism, which advocates minimal changes to existing beliefs. However, minimalism may not necessarily capture the nuanced ways in which human agents reevaluate and modify their beliefs. In contrast, the explanatory hypothesis indicates that people are inherently driven to seek explanations for inconsistencies, thereby striving for explanatory coherence rather than minimal changes when revising beliefs. Our contribution in this document is two-fold. Motivated by the explanatory hypothesis, we first present a novel, yet simple belief revision operator that, given a belief base and an explanation for an explanandum, it revises the belief bases in a manner that preserves the explanandum and is not necessarily minimal. We call this operator explanation-based belief revision. Second, we conduct two human-subject studies to empirically validate our approach and investigate belief revision behavior in real-world scenarios. Our findings support the explanatory hypothesis and provide insights into the strategies people employ when resolving inconsistencies.

Method

This document uses two human subject studies to investigate belief revision behavior, to validate the explanatory hypothesis, and to provide strong evidence for the applicability of the proposed framework. The research involves three common inconsistencies in the cognitive science literature, analyzing the data through statistical tests and effect size measurements to validate the explanatory hypothesis. In addition, the document discusses logical premises, experimental design, and results obtained from human subject studies to support the proposed framework. These empirical assessments provide solid evidence for the explanatory belief revision framework and delve into the strategies people adopt when resolving inconsistencies.

Main Finding

The main finding of this document is that individuals tend to engage in non-minimal revisions when resolving inconsistencies, preferring more comprehensive explanatory frameworks that involve altering their existing beliefs to a greater extent than minimalism would predict. This inclination challenges the principle of minimalism in belief revision and validates the explanatory hypothesis, which suggests that people prioritize generating coherent and plausible explanations over making minimal changes to their beliefs when faced with inconsistencies. The empirical evidence from two human-subject studies supports the prevalence of non-minimal revisions in participants' explanations, indicating a natural inclination towards understanding the underlying factors that give rise to inconsistencies rather than merely resolving them in a superficial manner. These findings have implications for belief revision theories and human-aware AI systems, suggesting a potential need for reevaluation of existing frameworks aimed at modeling actual human reasoning patterns.

Conclusion

The conclusion of the document challenges the principle of minimalism in belief revision and introduces an explanation-based belief revision framework, which prioritizes explanatory understanding over minimal changes. The framework is supported by empirical evidence from two human-subject studies, indicating that individuals tend to engage in non-minimal revisions when resolving inconsistencies, favoring more comprehensive explanatory frameworks that involve altering their existing beliefs to a greater extent than minimalism would predict. The findings validate the explanatory hypothesis, suggesting that people prioritize generating coherent and plausible explanations over making minimal changes to their beliefs when faced with inconsistencies. This approach presents a challenge to the proponents of minimalism and has implications for belief revision theories and human-aware AI systems. The document emphasizes the importance of explanations in human belief revision processes and suggests a potential need for reevaluation of belief revision theories aimed at modeling actual humans. The introduction of the explanation-based belief revision framework offers a new perspective in belief revision theory and provides a foundation for further research in the development of practical human-aware AI systems.

Keywords

belief revision, explanatory hypothesis, minimalism, human-subject study, cognitive science literature, propositional language, logical preliminaries, AGM model, coherence model, foundational model, explanation-based belief revision, human-aware AI, model reconciliation, minimal revisions, non-minimal revisions, empirical evidence, explanation-driven revisions, human reasoning, explainable AI, synergistic interactions, transparent AI, human-AI collaborations