Paradox of Self-Amendment by Peter Suber Section 7: Hart's Theory of Acceptance A. Salient points from Hart's text Hart's theory of acceptance is complex. He can banish the specter of Ross's paradox only if he succeeds in arguing that acceptance and usage can confer legal status on new rules of change. This is equivalent to arguing that not all legal change occurs through inference-like transfers of authority from prior norms, or that the people and officials of a legal system can accept as law a self-amended AC that contradicts its predecessor. Hart undoubtedly urged his acceptance theory for different programmatic reasons than to make possible the legal change of the rules of change. But that he urged the theory at all allows him, within a systematic philosophy of law, to answer Ross and provide an alternative to immutability and paradox. This is not the place to assess every aspect of Hart's complex theory. But we should ask exactly what acceptance is. Hart defines acceptance primarily via negativa or by denying its identity with similar phenomena. One of Hart's first concerns in The Concept of Law is to show the inadequacy of John Austin's theory that laws are commands or coercive orders backed by threats given by a person or body whom the populace habitually obeys. Hart rejects almost every element of this theory, and therefore is careful to avoid identifying acceptance with habitual obedience or mere convergence of behavior (54ff, 73).[Note 1] He is equally concerned to distinguish his theory from the natural law tradition that makes the final validity of positive law depend on its conformity to moral law. Hence acceptance is not normative in the moral sense. It is independent of consent (168). It is normative only in that it reflects an attitude —called the "internal" point of view— that conduct is prescribed by rule rather than accidentally convergent. It avoids the descriptive position that looks only to public force, private feelings of obligation or compulsion (56, 85-86), and psychology (136), as well as the "external" point of view that sees rules as binding only on some other group that accepts them. It is not acceptance of a moral ideal, but only of legal authority (98, 198-99). Moreover, acceptance must differ from custom because, in Hart's theory, custom has no law-making effect until and unless it is itself recognized or accepted as law (44f, 98). Acceptance may be comprised, at least in part, of social pressure or expectation to conform (56, 84, 85, 165). But social pressure and expectation cannot be the whole story or else the distinction between law and morality would be blurred (165ff). Hart is proud of the fact that acceptance and usage, while they are distinct from habit, convergence of behavior, moral approval, custom, feelings, and mere social pressure, are nevertheless matters of fact, not of metaphysics or mystery (62, 101, 107). Acceptance validates or authorizes the ultimate rule of recognition (92, 93, 97, 106, 205, 229). The rule of recognition is accepted as common, shared, or communal to all members of the society, even if obedience is "each for his own part" (112-113). Acceptance also validates some other types of law, such as standards of judicial review (142), cases of first impression and other judge-made rules (149-50). These examples show that Hart's acceptance model does not require the authority of a law always to precede the law in time (see Section 6.C). Acceptance can, but need not, catch up with acts that precede or exceed their authority. Even when acceptance bypasses or overrules the rule of recognition, it makes law, not merely an excuse for putting up with illegality or a climate that makes any inquiry into legality futile. In primitive legal systems without secondary rules acceptance directly validates or authorizes all law (230). In international law, where there are some secondary rules but, Hart says, no general rule of recognition, he is less sure (230-31). However a "transition towards acceptance" is possible in international law today (231). If any rule of international law applies to a nation that does not assent to it (for example, by treaty), then international law must have a rule of recognition that so held; but Hart doubts that any rule binds nations without their assent and therefore doubts that there are any international legal rules that acceptance could validate even in the absence of a rule of recognition (230-31). However, he says that treaties bind only because rules external to the treaties say they bind (219, 231); he applies the same principle to contracts (42, 220). I will call this the irreflexive view of treaties and contracts, as opposed to the reflexive view that they bind ex proprio vigore or by their own strength. Hart's assertion of the irreflexive view would be vitiated if he believed that the external rule that validated treaties could be brought into effect merely by the agreement of the parties to a treaty. Yet he does not say whether it is validated by acceptance, a rule of recognition, or something else. The problem of reflexive and irreflexive validation will not go away and will be considered at greater length below. When accepted, the rule of recognition is made law, not merely an assumption, postulate, or hypothesis (105) and not merely a "pre-legal" or "meta-legal" rule, or mere "political fact" (108). Acceptance, therefore, does more than support law by assuring compliance and preventing revolt; it makes or defines law. Legal rules bind if and only if they are accepted as binding (226, 230). The rule of recognition is "fact" (107, 108) in that it exists only in the actual practice of officials (105, 106). The law made or defined by acceptance is not a brooding omnipresence, a prediction of future judicial behavior, or a formal ideality: it is fact and practice, an observable social reality. The rule of recognition is usually tacit (98) and is ascertainable only by empirical observation of the practice of officials of the system (98). Incidentally, Hart does not address the question whether the ultimate rule of recognition is law only because it is accepted, or whether it also recognizes itself in self-application (but see 104, 108). But while the rule of recognition is law, it is the only law that cannot be called valid or invalid; it is "simply accepted as appropriate for use" (105, see 106). In context Hart implies that it cannot be valid because it cannot be said to conform in the usual way to the criterion of validity, which is itself, the rule of recognition. To use his analogy, we could say that the standard meter bar in Paris is one meter long by its own standard; nevertheless, we know it is a meter, not from self-application, but from agreement or acceptance to 35