Explanatory Coherence and Belief Revision in Naive Physics

A computational model of how experimental subjects revise their naive beliefs about physical motion is described using ECHO, a connectionist computer program that uses constraint satisfaction techniques to implement a theory of explanatory coherence.

Abstract

Abstract : Students of reasoning have long tried to understand how people revise systems of beliefs. We maintain that people often change their beliefs in ways driven by considerations of explanatory coherence. This report describes computational model of how experimental subjects revise their naive beliefs about physical motion. First, instances are presented in which subjects changed their beliefs while learning elementary physics. Each of these cases involved an individual's attempt to explain a surprising observation. Next, we show how their belief revisions can be modeled using ECHO, a connectionist computer program that uses constraint satisfaction techniques to implement a theory of explanatory coherence. The resulting simulations even captured temporal characteristics of the observed changes in beliefs. Finally, the model's representational sensitivity and procedural robustness are discussed and one can show how ECHO can be used to generate empirical predictions about subjects' current beliefs. Keywords: Explanatory coherence; Belief revision; Naive physics; Connectionism.