Abstract
Social scientists have traditionally assumed that the only appropriate response to the falsification of a theory is to reject it. That we do not suggests that the behavior of social scientists is irrational (or at least nonrational).
This assumption is a mistake based on an overly simplistic view of what a theory is and what is involved in a test of a theory. Theories are in fact very complex intellectual constructions, involving concepts, definitions, statements of contingent relation, and statements of scope. Tests of theories are even more complex, involving instantiations of concepts in both substantive statements and scope statements, hypotheses, and observation statements. Falsification indicates a problem is present somewhere in that intellectual structure. The appropriate response in such situations is to revise the theory (or the test of theory), not to scrap the theory altogether. The process of revision and retesting may continue indefinitely. There is no specifiable point at which it becomes more rational to reject the theory than to continue modifying it.