Abstract

The question of the “validity” of the “sign-rule” has been a source of continuing disagreement among sociologists. Some of the confusion surrounding this question can be dispelled by focusing on the problem of providing satisfactory interpretations for calculi constructed to represent various commentators' versions of acceptable “sign-rule” arguments. It is shown that the formulae of a calculus constructed to represent standard sign rule arguments must be interpreted in terms of propositions asserting that rankings of objects in virtue of one property are identical to rankings in virtue of another property. A system constructed on the basis of “the assumption of high correlations” is shown to involve distinctions between the logical forms of the premises and the conclusions of key arguments. This sharply limits the deductive power of this system. A third system, constructed on the basis of the “causal modeling” view of the sign rule, is shown to require an interpretation which renders the procedure irrelevant to the construction and testing of general theories.