Introduction

There is an ongoing debate amongst philosophers of social science about the possibility of laws of the kind Condorcet recommends. The stakes are high, as laws are what science is said to search for and are at the core of traditional accounts of explanation. On the deductive nomological account of explanation (Hempel 1965) laws are required for the logical deduction of the explanandum, the statement of the event to be explained, from statements describing the antecedent conditions. To explain why a particular individual makes certain market choices, a law of the rationality of maximization is required. Alan Nelson, in discussing economics, points out further that “when we work with defective laws, we often wind up with defective accounts of facts” (Nelson 1986: 163). The logical representation of what counts as an explanation has come under strong criticism. It is characteristically replaced by claiming that explanations may not be derivations, rather they appeal to causes. To answer why something occurred, what is needed is to identify the cause of the event. Even with explanation by causes, laws enter the scene, as it is laws that describe the causal relations that hold between events in the world.