How do we learn models? Introducing the supposed range vs. real range hypothesis

Abstract

We present a general learning model explaining in more depth how we learn (or fail to learn) models and theories. It specifically addresses the phenomenon of preconceptions introduced by the constructivist pedagogical approach. Nature of cognitive conflicts as well as existence, birth and resolution of preconceptions are clarified through a two-stage model based on the formalization of validity conditions of models and theories. Illustration of our hypotheses is provided by various examples, from limited-range models to wide-range scientific theories. Specific consequences on research are also discussed.