Hart: The elements of a legal system
Hart begins by asking us to imagine a society very like Mbuti society. There are many rules that govern Mbuti life, rules that are recognized and largely obeyed by most Mbuti people.
But there are no officially organized sanctions for breaches of these rules. People who disobey them regularly will be criticized and, perhaps, in the end, ostracized. But there are no judges, no police officers, no courts. These basic rules—rules that are necessary if people are to live together in a society at all—Hart calls “primary rules.” They say what a member of the society may or may not do. Typically, there will be primary rules against taking other people's property, against using unnecessary violence in disputes, and against breaking one's freely made promises. Primary rules include more than the precepts of morality: for example, morality does not determine exactly how property should be transferred between generations. But many of the primary rules will be moral rules: rules against murder and lying, for example. According to Hart, this minimum structure of primary rules captures the truth in natural law theories; any group of people that failed to recognize even these basic rules would hardly constitute a society at all.Primary rules are not enforced by officials; as in the case of the Mbuti, there may be no state to enforce them. And, in a society with only primary rules, there is plainly no legal system.
Now, the Doulians certainly have more than primary rules, because they do have some officials—what they call “police officers,” for example. But, as we have seen, they still do not have a legal system. Hart argues that what we need to add to the system of primary rules in order to create a legal system is not merely a set of sanctions enforced by officials—otherwise the Doulian system would be a system of law—but a number of other kinds of rules.
These other rules he calls secondary rules.Secondary rules, Hart says, “are in a sense parasitic upon or secondary to” primary rules.
For they provide that human beings may by doing or saying certain things introduce new rules of the primary type, extinguish or modify old ones, or in other ways determine their incidence or control their operations.
Hart sees secondary rules as introduced to meet a number of deficiencies in the system of purely primary rules that the Mbuti have—deficiencies that would need to be remedied if the Mbuti were to move from a society organized in small groups to the larger scale of society in which almost all human beings now live.
The first deficiency that Hart identifies is that a system of primary rules is uncertain. What he means by this can be made clear enough in the Mbuti case. A system of primary rules has two kinds of uncertainty. One kind of uncertainty arises when it is not clear, on the basis of the evidence available, which of two rules actually applies in a given case.
Suppose, for example, that the Mbuti held that a man's bow and arrows should be inherited by the son who is the best hunter. And suppose they also held that a person could give away (or sell) his own bow and arrows. Then when a man died, it would not always be clear whom his bow and arrows should go to.
Now suppose that in some particular case everybody knew that the best hunter in a certain family was the eldest son. If one of the younger sons claimed that he had been given the bow and arrows before his father died, then this younger son could claim that the rule of inheritance need not be invoked. For, at the moment of death, the bow and arrows no longer belonged to the father. There would now be a dispute between the two sons about who owned the bow, and there would be no mechanism for deciding who should get it.
But systems of primary rules are open to another sort of uncertainty, an uncertainty of an even more troubling kind.
For in a system of primary rules, even if the facts are agreed, there is no way of deciding, in a disputed case, what rules actually apply.For example, if the eldest son claimed that there was a rule that said that a father could not give his bow and arrows away on his deathbed—that it was wrong, by Mbuti custom, to do so—there would be no way of checking to see whether this was, in fact, a rule of their society. There would also be nobody who could decide definitively whether the oldest son was right and then enforce that decision.
The first kind of secondary rule, therefore, that Hart argues a legal system must have is what he calls a “rule of recognition.” A rule of recognition is a rule that tells us how the question whether a rule is a law in our society is to be decided. In the United States, for example, as in all modern societies, there is a highly complex set of rules of recognition. The rules of recognition of the United States say, very roughly, that a rule is a federal law if it is either
a) a constitutional provision or
b) a law created by the constitutionally defined process of lawmaking, or
c) a rule that was established by the courts in the common law tradition that grows out of the legal tradition that predated the Constitution and which has not been explicitly cancelled or superseded by rules made under the Constitution.
Similar considerations determine whether a rule is a law in the states. It also tells us which laws are to be applied in cases where there is conflict; in some matters, federal laws take precedence, and in others, state laws do.
The rules of recognition of a society, even of a modern industrial society, do not need to be written; British judges do not rely on a written document telling them to apply laws made by the British parliament and signed by the queen.
The role of the rules of recognition in the British system depends on the fact that judges have learned, in the course of their education and their practice as lawyers, how the legal system decides whether a rule is a valid rule of law.But rules of recognition are not the only secondary rules that are needed to turn a collection of primary rules into a legal system. A second class of secondary rules is needed to remedy a second defect of the Mbuti system, namely, that there is no way for the Mbuti to change their rules explicitly. Rules of this kind—’’rules of change,” Hart calls them—are embodied in the American Constitution in the sections setting out the powers of the president, the legislature, and the judiciary. Once more the position is complex and can only be very roughly described in a brief compass: but one rule of change says, roughly, that if a rule has
a) been through the procedures necessary to be passed by the legislature, and
b) been signed by the president (or returned to the legislature and passed by a majority sufficient to override a presidential veto),
it will be recognized by the courts, provided it is not in conflict with the Constitution. If, in interpreting these laws, the courts declare that certain rules follow either from the statutes explicitly passed by the Congress or from the Constitution itself, then
c) those rules become incorporated in the law also.
Finally, Hart argues, there is one other deficiency in the system of primary rules exemplified in Mbuti society: it is highly inefficient. When there is a dispute about whether a rule applies, there is no settled procedure for determining the issue; and even if it is clear which rule applies, there is no one who is given the job of stopping offenders or punishing them.
The addition of rules of recognition and rules of change would not, by themselves, remedy this deficiency.
The reason is obvious enough. I have already talked of which rules courts recognize; obviously, what is needed to gain the advantages of the other secondary rules is a set of rules that create something like courts. These rules should determine which individuals have the task of deciding, in which cases, which rules apply. This third sort of secondary rule Hart calls a “rule of adjudication.”In most societies it will also be thought necessary to assign to somebody the task of enforcing the decisions in those cases, since there is an obvious advantage in having officials—such as bailiffs, police officers, and prison guards—who make sure both that the decisions of the courts are carried out and that those who ignore the rules are punished. But Hart says that these further officials are not essential to the existence of a legal system. In a small-scale society it might simply be that once the courts had decided that someone was to be punished, anybody could punish them. What is required for a legal system is only that there be officials charged—by the rules of adjudication—with determining what the rules are, and a relatively clear set of principles—the rules of recognition and change—by which they make those decisions.
If you believe that the element of coercion by the government is central to the idea of law, then you will want to add to Hart's claims the thesis that the rules the courts decide are applicable should be enforced by the government, through its agents. And so you might want to add a fourth kind of rule—"rules of enforcement,” I'll call them—that creates a class of officials who have the responsibility of punishing offenders and enforcing the judgments of the courts. But you can still agree with Hart's basic definition of a legal system as “the union of primary and secondary rules.”
In line with Hart's proposals, then, we can thus modify my original definition of laws:
Laws are rules, backed by the threat of force, promulgated by a legitimate government to regulate the behavior of people subject to its authority, and which belong to a system containing both primary rules and secondary rules of recognition, change, adjudication, and enforcement.
Institutional laws, governing the way courts should operate, are secondary rules of adjudication; constitutive laws, such as the laws governing the creation of wills, are, in effect, part of the system of rules of change.
For such laws allow people to create rules—my property should go to my designated heirs—that will then be applied by the courts.If the Doulians were to change their system in such a way as to create rules of recognition, change, adjudication, and enforcement, and if these rules were actually operative in Doulia, then many people would surely say that Doulia had—at last!—achieved a legal system. Once there was such a system, generally directed to the common good, they would say, with Austin, that even a bad law that was not aimed at the common good was nevertheless a valid law of the Doulian legal system. But this would not mean that they had agreed entirely with the positivist tradition, for this second definition makes it a condition of being a legitimate government (and thus a condition of being a source of valid law) that you should have instituted a system of rules aimed at the common good.
This second definition is much closer to the natural law position than is Hart's, because it requires that the system of laws be enforced by a legitimate government; it implies some moral constraints on the content of a legal system because a legitimate government must aim to promote the common good. But some philosophers have argued that this is not the only way in which moral ideals play a part in determining what sorts of rules and procedures can be recognized as part of legal systems. They have argued, following the natural law tradition, that there are certain moral constraints, internal to the idea of law, that mean that the rules and procedures of a legal system must answer to certain moral ideals. So I propose now to examine this claim in the case of one particular kind of procedure, namely, the institution of criminal punishment. If, as I have suggested, any legal system must have rules of enforcement, then any moral ideals that constrain punishment are part of the concept of law.
7.5