Contending Conceptions of the Theory of Rational Action

Advocates of the theory of rational action are divided in regard to the version of the theory they accept. This paper distinguishes between a narrow version, claiming that the kinds of preferences and constraints to be used in explanations must be restricted, and a wide version imposing no such restrictions, and thus including beliefs, altruism, norms and social sanctions in explaining behavior. The paper begins by describing the major assumptions of both versions and then discusses the key arguments that are adduced in favor of a narrow and against a wide version: (1) preferences and beliefs cannot be measured; (2) the wide version is tautological, circular, empty or trivial; (3) predictions with the wide version are difficult; (4) the assumptions of the narrow version are sufficient to explain behavior; (5) when problems are encountered, the narrow version should only be applied to situations where it works. It is argued that those arguments do not provide an adequate defense, and thus suggest that the narrow version is inadequate. The last part of the paper offers some methodological considerations, suggesting how different types of assumptions can be combined.