Theorizing and Empirical Belief

Abstract

It is generally held that induction contributes little to the appraisal of competing theories. Certainly our expectation that a theory which has adequately explained observed facts of a certain kind will continue to explain facts of the same kind may be considered to be inductive, in a broad sense of the term. However if, as is widely held, theories are underdetermined by the empirical facts for which they account, there must be other criteria which play a major role in selecting between competing theories. In recent decades the historical variability and the social dependence of the criteria that usually are applied has been emphasized and this presents a problem about the rationality of theory selection. Larry Laudan has recently offered an account of theory appraisal in which induction plays a much more substantial role. Specific methodological rules are accepted according to their past success in selecting theories that exhibit the epistemic virtues we seek in them (Laudan 1987). According to this account, criteria of appraisal do change but change rationally. Induction plays a crucial role, yet it is only external; it selects the criteria we employ, but these criteria are in no way implicit in induction itself. I will argue that induction actually contributes much more than this. When an adequate specification of induction is provided (a task that is not as simple as is sometimes assumed) it is found to include criteria which play a substantial role in the appraisal of theory.

I would like to express my gratitude to the anonymous referee for this volume whose comments led me to revise the paper considerably.