The Varieties of Normativity: an Essay On Social Ontology

I. Hart and Soft Positivism

We first set the stage for our more detailed treatment of Searle’s views on normativity with a discussion of Hart and Rawls. The work of Hart, especially, forms part of a famous debate between natural law theorists and legal positivists, a debate which reveals that the question of the ontological status of laws has historically been linked to normative issues of moral philosophy. Natural law theorists affirm that immoral law is not law; that is, they believe that the ontological status of laws is determined by their relation to morality, in accordance with the famous motto: “Non videtur esse lex quae justa non fuerit ”. Legal positivists, on the other hand, insist that law is law independently of whether or not it is moral. According to the classical legal positivism of John Austin, for example, the issue of the legal status of law is an entirely empirical affair, to be established primarily through the determination of pedigree and enforceability. Was the entity or institution created and maintained in existence in accordance with the right sorts of rules? Is the entity such that the state can coerce people into complying with it?

According to Austin, we are to understand the nature of a legal system by starting out from the case of someone forcing someone else at gunpoint to hand over his wallet. The normativity of the law differs from the normativity of the highway-man only in this: that the law normally functions on the basis of threats alone; only in extreme circumstances is it necessary to bring guns into play.

InThe Concept of Law , Hart deploys a sustained attack on traditional legal positivism. His criticism of Austin is both elegant and persuasive.[^4] Hart himself still defends a positivistic conception of the ontological status of the law, but he rejects traditional positivism, above all because of its superficial treatment of rules. The rules the gunman imposes upon his victim - “Hand over your wallet”, “Don’t do anything stupid” - are all of the same type: they demand certain sorts of conduct. The law, however, operates on the basis of two types of rules, which Hart calls primary and secondary. Primary rules are duty-imposing; they demand conduct in just the way in which the gunman’s actions do. Secondary rules are power-conferring; they make certain sorts of situations possible - they are rules about rules. A rule that states that a judge is entitled to decide how to interpret a primary rule is a secondary rule; it gives the judge the power to settle disputes by establishing what the correct interpretation of a law is.

It is possible, perhaps, to imagine an entire society in which there existed only primary rules. But such a society would be profoundly inept when it comes to resolving controversies about the laws themselves or about their interpretation. A situation, on the other hand, in which secondary rules would arise in relation to highway-men robbing stagecoaches belongs, at best, to the world of Monty Python.

With only one kind of rule in its conceptual armoury, Hart argues, traditional positivism is unable to distinguish between two crucially distinct phenomena: (1) beingde facto obliged and (2) having a genuinely normative obligation. If a gunman puts a gun to your head, you might indeed be, as a matter of empirical fact, obliged to hand over the money. For you to have a normative obligation, in contrast, it is necessary that you accept not only the empirical fact of your being obliged but also the rightness of thesystem which makes this so (even if you do not accept specific rules in this system). You accept that to do this or that is yourduty ; that it isthe right thing to do This notion finds no purchase in the realm of actions performed in response to gunmen’s threats.

Hart refers to this dimension of acceptance as the “internal aspect” of obligations, to which he opposes an “external aspect” - the only one that traditional positivism is capable of explaining. He asks us to imagine someone describing the functioning of a street light in a busy intersection in the following way: when the street light becomes red in the direction of the cars, the likelihood that cars will stop, and that pedestrians will cross the street is very high; when the street light becomes green in the direction of the cars, the likelihood that cars will move forward and pedestrians will stay put increases. Obviously, Hart points out, such a description fails to mention a fundamental element of what is really going on. The red light is not merely asign that allows us to predict that drivers and pedestrians will behave in this or that way; rather it is areason which gives rise to this or that behavior. The red light indicates not simplythat I stop, but that Iought to stop. This notion of areason is not available to traditional legal positivism.

Since Hart is himself a positivist, it might look as if by introducing normative elements into his determination of the ontological status of laws he concedes too much to natural law theory. After all, for Hart as for natural law theorists, whether a given entity is or is not law depends on normative factors. He insists, however, that he has carved out an intermediate theoretical space between natural law and traditional positivism, which he calls “soft positivism”.[^5]

Hart’s strategy - though he does not himself admit it - is to distinguish between two types of normativity. On the one hand is the robust normativity of the natural law theorist, illustrated for example by the Ten Commandments On the other hand is Hart’s own brand of normativity - what we might call soft normativity - which is what is necessary, in his view, for the existence of laws. Soft normativity is the sort of normativity that flows logically from the very nature of secondary rules. Secondary rules create institutions, and these institutions in turn create the very possibility of certain sorts of acts. Hart himself appeals to the example of games in order to illustrate this point.[^6] A group of people can play football without requiring the presence of a referee of any sort. But when a referee is present and disputes arise, then the referee will have the last word in resolving such disputes. His appointment is however possible only insofar as the players accept the secondary rules that make the institution of refereeing possible. That the referee has the last word is part of the content of the corresponding secondary rule, and it is this same rule which gives rise to the normative component in the referee’s decisions. When a referee declares “penalty kick”, for example, he is not merely providing an indication of what is likely to happen next (any more than a traffic light is providing an indication of likely traffic flows). Rather, his declaration is the very reason whichexplains what happens next, because it explains whatought to be done.

But there is a problem with Hart’s approach. The sense of ‘ought’ as expressing soft normativity, the sense of ought that is involved in rules of games like chess or football, is radically different from the sense of ought that is involved, for example, when someone says that we ought to treat other human beings with respect, or that we ought not to gratuitously harm them. We believe that any ontology of legal institutions that does not do justice to the distinction between these types of normativity is doomed to fail.