A Structuralist Framework for the Logic of Theory Change

Abstract

Ever since the 1980s, the dominant trend in the field within philosophical logic concerning the logic of theory change has been to work with models lying rather close to the well-known AGM approach. In particular, most of the (normative, formal) theories on the subject of theory change that can be found in the literature seem to be in relatively close agreement on how to represent theories; roughly, it is assumed that theories are representable as sets of statements. In this paper, I will try to draw the outlines of a model for theory change in which this assumption is revised: instead of the usual representation of theories, the foundation of the model will be based on the structuralistic notion of a theory net. Structuralism proceeds from the idea that scientific theories have what we may call a deep structure, and aims to provide a formal representation of theories that respects this structure and describes it in as much detail as possible. The result is a formal notion of “theory” which is considerably more fine-grained than the AGM-style representation in terms of logically closed sets of sentences, and my hope is that it will help shed some new light on the problems studied in the preexisting frameworks for theory change, as well as open up new and interesting research problems in the field.