Expressed at a fairly high level of generality, legal positivism is a rather simple doctrine. It tells us that what the law is, is a matter of fact. Despite this apparent simplicity, it has elicited an enormous amount of philosophical debate and controversy. Much of that debate has had to do with the “rather than” clause that seems to be begged by this initial characterization of the position. What positivists are at pains to distinguish law from is “morality” (however that contested concept is itself characterized). Thus, the debate over legal positivism becomes the debate over whether the connection between law and morality is, as philosophers say, contingent or necessary. Granted that there is a lot of moral content in law, the question is whether that content is one in the absence of which law would cease to be law.

I do not want to contribute to that debate here. Indeed, it seems to me to be one of those debates about which every possible position has been staked. It is indeed hard to say how the debate between positivists and “natural lawyers” could be moved forward. Instead, I want to consider legal positivism not so much in its contrast with natural law theories, but rather as an epistemological theory about how we come to know what law is. In other words, I am less interested with legal positivism as a theory about what legal facts are, but rather as one which provides us with indications about how we come to know those facts.

What do I mean by that? Well, consider H. L. A. Hart’s famous critique of Austin. Austin, it will be remembered, viewed law as being made up of three basic elements: command, obedience, and sanction. Hart, famously, and probably unimpeachably, noted that law just doesn’t function that way. It comes closest to looking like what the Austinian model suggests in the area of criminal law, though even there the picture occludes at least as much as it reveals. Austin’s picture, Hart tells us, does not capture some of the most important things that modern legal systems do, such as empower people to create new kinds of relations among themselves. It tells us nothing about how law remains when the commander changes, or how law changes. And so on.

One way of understanding Hart’s critique is to accuse him of a kind of idealism, understood here in the epistemological rather than the normative sense. The accusation can be formulated in the following way: you come to the study of a domain with a preconceived notion of what that domain is, and being insufficiently self-aware of these assumptions, you come to project them onto the world. Hart may be taken as saying that Austin, in his desire to appear hard-nosed against those who would view law as emanating from morality, or from the mind of God, decided in advance of actual observation that law had to do with some people managing to elicit a habit of obedience on the part of other people through their issuing of credible threats of sanction. In so doing, that’s all he saw when he set out to put forward a “fact-based” conception of law. He projected his preconceptions onto the world, rather than looking and seeing.

The main point I want to make here, and one that, unless I am mistaken, has been insufficiently explored, is that Hart’s critique of Austin is much more corrosive than Hart himself might have realized. Indeed, it can be turned against Hart himself.

Hart famously provided us with something like an epistemic principle in order to distinguish law from non-law—that of the rule of recognition. If you want to know what the law is in a given community, then observe the behaviour of legal officials who bear an “internal attitude” to the law and to its justificatory apparatus. When you do so, you will find the law to be a much more diverse set of phenomena than Austin had allowed.

Now, the exact nature of the rule of recognition has been debated at great length in the philosophy of law literature. It has been understood in at least two ways, which I will refer to as Wittgensteinian and Kelsenian. The Kelsenian interpretation is that of an ultimate rule of validity, a rule about which further questions of legal validity cannot be asked. It is presupposed by officials in a legal system that have the appropriate “internal attitude” toward that system. The Wittgensteinian interpretation sees the rule of recognition as constituted (rather than simply recognized) by the practice of legal officials. Fortunately, I do not have to wade into these interpretive controversies in order to make the argument I want to make here. Whether we interpret the rule in one of the other of these manners, it is in either case to the behaviour of public officials that we must look if we want to determine what the law is among the members of a certain community, at a specific time. And this is what Hart would have us do in order to determine what valid law is—observe the behaviour of a population that has already been divided for us into “legal officials” on the one hand, and the rest of the population on the other, whose behaviour is less relevant to determining what the law is than that of legal officials, except insofar as it betokens a requisite level of conformity to laws as enacted in conformity with the structure that culminates with the rule of recognition.

There is idealism in the sense that I have described above in this way of proceeding, just as surely as there was in Austin’s case. Just as Austin came to the supposedly empirical study of law with a predetermined sense of what legal relations consist in, Hart approached his study of law with a predetermined set of assumptions about where to look in order to find law, and an accordingly predetermined way in which to divide up the population between legal agents and legal subjects. Very roughly, Hart’s idealism consists in the assumption that law is state law, and that legal agents are those legal officials who occupy official functions within the state’s system of positive law.

If positivism is about looking and seeing with as little theoretical pre-commitment as possible, then I want to suggest that Hart’s doctrine falls foul of the same positivist requirement as does Austin’s.

Saying this, however, would seem to place the legal positivist before an impossible task. We already know from Kant that we do not come to our understanding of the world in a presuppositionless manner. We project a certain kind of order onto the world, lest we experience it as orderless and intelligible. Maybe it is inevitable that we have such proto-theories about the kinds of things that we want to identify before we can identify them. We need to know where to look before we can accept the later-Wittgensteinian injunction to look and see.

So we seem caught in a dilemma. We want to avoid the kind of idealism that led Austin (and arguably Hart) to neglect or to exclude from their study of law, on the basis of unacknowledged theoretical commitments, phenomena that might have been included given other, equally plausible, theoretical commitments. Or, we attempt to approach the study of law in a maximally presuppositionless way, which leads to our not having the tools with which to pick legal phenomena out at all.

There are a number of ways in which to emerge from this dilemma. The first is just to restrict the range of the positivist inquiry, but to do so in a reflective rather than an unselfconscious way, and to say that what we are interested in is state law. When the positivist says that law is a matter of fact, they are restricting their inquiry to facts about, and facts created by, state institutions. The theoretical challenge is thus eased: the question of how to distinguish law from non-law is transmogrified into the question of how to distinguish state from non-state.

I believe that a positivist who is true to the idea that the law is a question of fact must be dissatisfied by this shortcut. There is no a priori reason to restrict the scope of the inquiry into what law is in the manner suggested by the state positivist other than the fact that the inquiry is easier. The positivist who would accept this simplification would find themselves in the position of the person looking for their keys under the streetlight not because they are there, but because there is more light there.

More concretely, such a restriction would not allow us to account for the kind of “stateless law” that even fairly methodologically conservative philosophers and theorists of law would want to include in the ambit of what we count as law—the law that is produced by non-state institutions at the international level, for example.

This leads me to the surprising conclusion, one also endorsed by Victor Muñiz-Fraticelli, to whose work the present argument can be seen as a long footnote, that there is a prima facie affinity between positivism—understood as a methodology for identifying what the law is rather than as a set of substantive theses that might be generated on the basis of that methodology—and legal pluralism—the idea that human agents routinely find themselves at the intersection of a number of different legal orders, rather than being entirely subsumed under one. At the very least, the positivist must not, lest they fall victim to a kind of idealism that is antithetical to positivism, reject on a priori grounds the idea that such multiple legal orders might exist.

Now, clearly, facts don’t entirely dictate categorizations. Once the positivist has observed a range of practices, and identified resemblances and differences between them, the decision to group some of them together as law while eschewing others as “non-law” or “quasi-law” is just that—a decision, one that is answerable to a range of different criteria. “Law” is not a natural kind. Grouping certain ranges of activities and practices in the area of social phenomena like law reflects human interests and concerns, rather than the pressures exercised by entities themselves to be grouped in one manner rather than another. Thus, for example, a sophisticated state positivist might justify their monism by invoking rule of law considerations. If human agents are truly subject to multiple legal orders, it would follow that they would not be able to view law as providing them with clear and non-contradictory directives. Pluralists might respond that monism purchases clarity at the price of doing violence to the multiple identities and community engagements of agents, and that the rule of law must be achieved at a more “meta” level, in the way in which we adjudicate conflicts and tensions as they arise.

I don’t want to pursue the issue of what range of values and human concerns are the relevant ones to invoke in order to fix the extension of the concept of law in one way rather than another. What I want to suggest is that a legal positivist—one who is dedicated to what I take to be the foundational positive commitment to be, namely, that “what law is” is determined in the first instance by looking and seeing rather than by bringing thick proto-theoretical commitments to the table in making that determination—shares with the legal pluralist the concern with observing, in as theoretically unencumbered a manner as possible, the range of practices, institutions, and communal behaviour that bear sufficient family resemblance to one another as to constitute plausible candidates for inclusion into a useful definition of law. Positivists and pluralists alike would also engage in careful reflection as to the range of values—ethical, pragmatic, epistemic—that militate for different ways in which to define the extension of the concept in one way rather than another.

Does this mean that at the end of the day, natural lawyers were right all along in claiming that law cannot be identified without reference to values? After all, my account would seem to suggest that we must select certain empirical traits rather than others as warranting the admission of a practice to the status of law, and that this selection must be made ultimately not on the basis of empirical considerations, but rather on the basis of values that point to certain empirical properties as most salient or relevant to the classificatory exercise. It would, if the values on the basis of which the selection gets made were primarily ethical values. If that were the case, then it would turn out that legal positivism could only complete the task of identifying law from non-law by the identification of certain ethical values as having pride of place in the very definition of law.

However, I do not believe that that restriction would be warranted. There are a range of values to which one might appeal to draw the appropriate boundaries. Some of those may be ethical, to be sure, but others may be pragmatic, and others still may have to do with concerns of theoretical parsimony—we don’t want to group too many phenomena together under the same theoretical rubric lest we come to emphasize to too great a degree that which practices share as opposed to that which allows us to tell them apart. Seen in this way, natural law theories might themselves be viewed by the legal positivist as a form of epistemic idealism, since, at least in some of their instantiations, they give pride of place to ethical values in the classificatory exercise in a question-begging way. This might at the end of the day be the crux of Hart’s complaint against Fuller, though, as we have seen, Hart is himself a culprit of the idealist temptation in unreflectively assuming state centeredness.

This is the point at which my account departs from that of Victor Muñiz-Fraticelli’s in The Structure of Pluralism. Muñiz-Fraticelli views intelligibility as the chief criterion distinguishing law from non-law, or, as he puts it in a telling turn of phrase, “law before law.” And such intelligibility on his account, if I understand it correctly, requires the presence of recognized legal officials who can present themselves to the outside world, and in particular to other legal officials, as the sources of authoritative interpretations and the guarantors of intra-systemic legal validity.

I do not want to deny that intelligibility—and the ability of legal systems to enter into communication with each other through the identification of authorized interlocutors such as legal officials—would be one relevant criterion to bring to bear in our attempt at identifying the empirical properties that warrant the application of the designation “law.” But there are others that may have to do with a community’s domestic affairs, as it were, rather than with its foreign relations. It is possible that the requirement of intelligibility and of a certain level of institutional realization arises when the need for mutual recognition makes itself felt. But to make this into the criterion through which law is identified by non-law requires that an argument be provided that makes clear the reasons for which it is more important to emphasize these dimensions of legal systems rather than others in the classificatory exercise. I am struck, for example, that Muñiz-Fraticelli’s classification would end up ruling out of court some of the legal traditions that are included by the late Patrick Glenn in his book. The level of institutionalization and the development of an authoritative class are seen there as a tendency that some legal systems realize to a greater degree than others, with variations over time. Legal positivism must, I would argue, remain alive to the possibility that the optimal realization of all of the values that are relevant to the classificatory exercise we are interested in requires that we draw the classificatory lines in a more capacious manner than Muñiz-Fraticelli allows.

References

Austin, John, The Province of Jurisprudence Determined (London, UK: John Murray, 1832).

Glenn, H Patrick, Legal Traditions of the World: Sustainable Diversity in Law, 5th ed (Oxford: Oxford University Press, 2014).

Hart, HLA, The Concept of Law (Oxford: Clarendon Press, 1961).

Muñiz-Fraticelli, Víctor M, The Structure of Pluralism: On the Authority of Associations (Oxford: Oxford University Press, 2014).