Obligation

The two most salient deontic realms are law and morality. But how are legal and moral obligations related?

Positivism is the doctrine that our legal and moral obligations are quite distinct, that what the law is is a separate issue from what it ought to be. In the form originally propounded by Jeremy Bentham and Austin, legal obligations are generated by commands of the sovereign, or that entity whom society is in the habit of obeying, who is not in the habit of obeying anyone else, and who has the power to sanction noncompliance. H. L. A. Hart showed that this form of positivism is subject to fatal objections. It can account for neither the continuity of legal authority across sovereigns nor the persistence of legal obligation after a given sovereign is gone. Moreover, he noted that it assumes that normative necessity must be instrumental necessity, which, as discussed above, is a mistake. What is needed are nonoptional rules, and sanctions are neither necessary nor sufficient for these (such as in the chess example). Part of the problem, Hart thought, was that Austin used criminal law, in which the law is a barrier to doing certain things, as his model for the whole of the law. Civil law (such as the law of contracts), in which law is a solution for various problems that would otherwise exist, is a much better model.

The law, Hart argued, is "a union of primary and secondary rules" (1994, p. 107). Primary rules direct each how to act ("don't eat shellfish," "no one may sit higher than the king," and so on). These rules impose genuine obligations, but not in the way that Austin thought. For Austin, rules are little more than means by which one can predict when the state will harm you. To see a rule in this way is to take an "external" view of it: conforming to the rules has no part in defining one's membership in the society. The first step in having a legal system is to have a system of rules that most take an internal view of, or see as standards of criticism and justification of one's behavior.

Hart argued that problems will inevitably arise in a system composed only of primary rules. First, in such a system, there will be uncertainty about whether a rule exists covering many situations and what to do when rules conflict. Second, the system will be too rigid to deal with changing circumstances. How will new rules come into and go out of existence? Third, the system will be inefficient in enforcing the rules and determining whether they apply in any case. What is required is a system of secondary rules, or rules concerning how to determine, introduce, abolish, change, and apply primary rules. Most importantly, what is required are rules of recognition that identify those features in virtue of which a rule is a rule of the group, to be supported by social pressures of various sorts. Rules of recognition are the sorts of rules that define lawmaking in U.S. state and federal legislatures. Thus, a rule exists as a rule in some system if its pedigree can be traced to rules of recognition defining legality for that system. Notice that, in contrast to Austin's view, rules can exist even if nobody ever obeys them. As long as a rule is enacted in the right rule-defining and -creating way, it exists. Hart insisted, however, that the rule of recognition itself need not be backed in any sense by any standard of authority, legal or moral. A rule of recognition exists in the sense of an "external statement of fact," that is, if it is generally not disregarded in defining and creating law.

Positivism's stark separation of legal and moral obligation has been challenged most forcefully by Lon L. Fuller and Ronald Dworkin. Fuller argued that if a given system of directives is immoral in a formal sense, then it will fail to be a legal system because it will fail to be a genuine system of rules. A system of directives is arbitrary if it does not conform to principles such as treating like cases alike, making its directives public, or not introducing directives after the fact. And an arbitrary system is not a system of rules. Moreover, a system possessing precisely these nonarbitrary features is also considered to be formally just. Thus, a system of directives that is not formally just is not a system of rules. If legal systems are systems of rules, then it follows that a formally unjust system of directives is not a legal system at all and does not generate legal obligations.

Dworkin's view challenges Hart's positivism by arguing that the law requires more than merely formally moral features. He argues that the actual practice of judges shows that legal systems contain elements from the moral traditions within which they arise and to which a judge's decisions must appeal to be valid. In cases that are not clearly covered by an existing law, judges in his view actually legislate, reaching outside of the law to make laws on the basis of good social policy. Dworkin argued that this does not match either actual or good judicial practice. Legal systems are composed of more than rules. They include principles invoking moral rights and obligations that judges are bound by in deciding cases. Indeed, what litigants are typically claiming is that some principle protecting a right is on their side. For Dworkin, there is always a determinate answer to the question of who has a right. Thus, there is always a determinate solution to legal conflicts to which litigants have a right and the judge is bound to find and deliver. Dworkin's own account of the pedigree of law therefore requires invoking moral principles and so implies that judges do not have discretion to create legal obligations as required by Hart's account. They must find the answer to which litigant's rights to protect, and not by intuition, but by articulating and defending a view of rights. Thus, for Dworkin, there is no sharp separation of legal and moral obligations.