John Rawls has advised us that definitions cannot settle fundamental questions. The merit of a definition depends on the theory in which it is embedded. Analyses of meaning, therefore, have no privileged status but must be assessed as elements of a larger theoretical 'enterprise.1 As applied to law, Rawls' advice is well taken. Disagreements over definitions of law are invariably correlated with differences in fundamental elements of the theories which contain them. It does not follow, of course, that the disagreements are empty exercises over matters that are 'merely verbal'. Rather, the definition serves as a proxy for the deeper issues at stake. A more direct approach would be to examine each alternative theory taken as a whole. On the other hand, one test of the soundness of a theory of law could be the adequacy of the definition it yields, especially if adequacy is tested by certain pretheoretical criteria. We might agree with Morris Cohen, for example, that a good definition organizes the subject matter in such a way as to help us grasp more clearly the fundamental patterns in our field of study. "In this respect so[me] definitions are certainly more helpful than others."2 It may make sense, therefore, as a preliminary exercise, to develop criteria of definitional adequacy even before elaborating a complete theory. That is the premise of the following exploration. What motivates this inquiry is Lon Fuller's elaboration of a con-