This paper by K&M appeared in Jurisprudence 12 (2021): 1-16. The main contribution of the paper is in section 4, which argues, quite plausibly, for the view that law is dependent on (at least some) people having the concept of law. This is a significant improvement on a related point that Alex Langlinais and I made in defense of Hart’s methodology in “The Methodology of Legal Philosophy,” in T. Gendler et al. (eds), The Oxford Handbook of Philosophical Methodology (2016). I will return to section 4, below.

The other parts of the paper are less satisfactory. Section 2 gives a rather idiosyncratic characterization of “disagreement about law.” We are told that “positivists believe that what the law is, is a matter of nonevaluative reading of the legal materials; whereas nonpositivists insist that the legal materials must always be interpreted in the light of general moral principles which are also part of the law” (2). This is not what positivists believe, and the framing already suggests Dworkin’s trademark confusion between an account of what law is and a theory of how a court should interpret law to decide a particular case. Hartian positivism (which is their reference point, reasonably enough) says that where law exists, there exists a rule of recognition, that is, a practice of officials of relying on certain criteria for deciding which norms are legally valid (i.e., norms of the legal system), and believing they ought to be using those criteria. It is because law is fundamentally conventional in this way that it can guide conduct outside the courts, as it does in every functional legal system. The criteria of legal validity might or might not include evaluative criteria (depending on what exactly is meant by “evaluative”). For example, many constitutions protect fundamental rights within a “balancing” framework, where rights can be overridden for the sake of “public order” or “morality.” Legal positivism has no problem with recognizing such a legal system. Of course, exclusive legal positivists may treat the “morality” clause as an invitation to discretion, but it is still the case that a court deciding a particular case will have to avail themselves of evaluative considerations. The nonpositivist needs an argument that these are part of the law, beyond the fact that a court applies them. (K&M say in footnote 2 that they are “ignoring” differences between “inclusive” and “exclusive” positivism, but I’m not sure they are entitled to do so. After all, Hart, an “inclusive” positivist, argued in the Postscript that Dworkin was simply doing specific jurisprudence, i.e., describing the rule of recognition of some legal systems. I think that is the most charitable thing to say about Dworkin, but it would defeat the claim that there is some deep disagreement.)

Le’ts grant, for the moment, that K&M are correct that the debate between Dworkin (and, more importantly, natural law theorists) and legal positivists has been going on for awhile. They claim it is “intractable,” and perhaps it is. But that is a data point that requires explanation: absent that, no inferences can be drawn from it. Sometimes apparently intractable disagreements are due to ignorance, cognitive defect, dishonesty, or bias, not to there not being a fact of the matter about the subject. Dworkin, as everyone knows, was a lawyer who wanted to win, not a philosopher who thought he could be wrong (as Simon Blackburn pointed out memorably many years ago, although his critique, alas, no longer appears to be available online). As we know from Nicola Lacey’s biography, The Nightmare and the Noble Dream (2004, 350; 330), Hart was frustrated by “Dworkin’s fluid and sometimes elusive analytic style” and came to feel “that there was something willful or even lacking in honesty about Dworkin’s reading of his work.” There is also the possibility that the “debate” persists because students of Dworkin (or of Hart or Raz) are loyal to their teacher, or the possibilty that young scholars find professional advantage in continuing to defend implausible positions. These are real possibilities, which probably explain many apparently intractable disagreements in philosophy, but I will not pursue them. I mention them just to make clear that what inference is to be drawn from the supposed “intractability” of the debate is a complicated empirical question, in part.*

Here, in any case, is a simple explanation for the supposed “persistence” of the Hart-Dworkin dispute: there is a disagreement about the data to be explained by the theory. Hart claims to be explaining what an ordinary person familiar with a modern municipal legal system knows about law, however inchoately. Dworkin claims nothing of the sort–he denies the relevance of such data. Dworkin treats as the central data point how courts, especially appellate courts, decide cases.

Yet Hart’s entire theory is organized around the obviously correct observation that law exists primarily outside the courts, as a way of guiding people’s conduct. (No one consults a lawyer about a contract, a purchase of real property, or a will to end up in court.) Decisions by appellate courts are a very small piece of the legal landscape. Massive agreement about the law in functioning legal systems, as I have argued previously, is one of the facts that a theory has to explain (and Dworkin’s has a lot of trouble doing so). In any case, people will appear to disagree if they are appealing to different data points. Of course, someone might think that the disagreement about data points is evidence of a disagreement about the nature of law. Perhaps it is, but since one side of the debate (the Dworkinians) never acknowledge the data question, it’s not at all clear.

Section 3, on disagreement about morality, asserts that “the suggestion [that moral disagreement means] that we [should] treat moral propositions on which plausible moral theories give divergent verdicts as having no truth value seems preposterous” (8). I won’t spend any time on this dogmatic assertion, since it is obviously unresponsive to why one might think disagreement about morality supports a skeptical conclusion about moral objectivity and truth. (Sidenote to readers outside the Anglosphere: many professional philosophers in the Anglophone world believe that their moral intuitions are objectively true. It helps, of course, that most of them know almost nothing about history or other cultures: see generally.)

Section 4 is the most illuminating and important section of the paper, and can be profitably read without regard for the preceding. Here K&M make a strong case that the existence of a legal order in a society requires the concept of law. As they note, on Hart’s view, “a socially accepted rule of recognition is necessary for the identification of law; such a rule is precisely a set of criteria to distinguish laws from non-laws [or better: legally valid norms from norms that are not legally valid]…Merely distinguishing instances from non-instances of law may not be sufficient for having the concept of law, but doing so reflectively on the basis of criteria would seem to be a weak sufficient condition for possessing the concept of law” (10). As they go on to observe, not everyone in a society with law needs to possess the concept, only “that some people possess it and to some extent converge in their judgments of application. Without this much, it is hard to see how a distinct normative order of law, as opposed to conventional morality, or etiquette, could emerge” (10) (cf. 11: “It would seem that at least most people involved in legal decision-making, enforcement, and expert interpretation should possess the concept, for otherwise the distinct normative order of positive law would lack stability.”) I think this is right but am curious if any readers have counter-arguments.

K&M also usefully distinguish between three kinds of “social kinds.” First, some are concept-independent (i.e., they can exist even if no one has the concept): examples (following Amie Thomasson) are “racism,” “recession,” and “superstition” (11). Second, some social kinds are concept-dependent: law, for example, but also “money” (they draw on Searle’s famous discussion [12]). Third, some of the latter concept-dependent social kinds are ones in which “each individual instance of the kind is concept-depenent” (11). This is not true of “money” (12), but is true of certain legal categories (e.g., “permanent resident,” “jury”) (see the nuanced discussion at 12-13 of these and other possible examples).

What about “law” itself? They argue that it belongs only to the second (“concept-dependent”) category, rather than the third category (i.e., the category of kinds all of whose instances require possession of the concept) (14). What bearing does this conclusion have on the debate between Hart and Dworkin? I don’t see that it has any: the only question is which theory of law best fits law and legal phenomena.

K&M, however, think the fact that law is a concept-dependent kind (in the second sense) is relevant because of their mischaracterization of the debate from section 2. They write: “people do apply a concept of law to identify grounds of law [in non-Dworkinian language, they presuppose what the criteria of legal validity are in order to identify legally valid norms]. The trouble is that we don’t all have the same view. Some think it is crucial to give a straight, nonmoralized interpretation of the legal materials; others disagree” (14-15). But as I noted at the start, this is not what the disagreement is about. Legal positivists are happy for judges to give a moralized interpretation of legal materials in order to decide a case. What positivists deny is that law necessarily involves moralized interpretations, and they deny that based on the data they are trying to explain, including the fact that law exists primarily outside the courts, that in a functioning legal system there is massive agreement about the law, and that, as Raz noted, everyone recognizes a difference between legal and moral expertise. This is only a partial list. How judges ought to decide particular cases is not the same question as what law is.

K&M observe that “it is possible that the disagreement between positivists and nonpositivists is only superficial. Deep down, we all agree…But it is this hope that apparently intractable disagreement about the grounds of law seems to undermine” (15). That we really agree is one possible explanation for apparent disagreement. But there are others, as noted earlier, and some of them seem very plausible given the actual facts about the supposed disagreement between legal positivists and a handful of Dworkinians.

I am especially interested in comments on K&M’s proposal that law is a concept-dependent kind (in their second sense), which seems to me right. I will approve comments on other topics in this post if they are well-informed and contribute to understanding. If you have not studied legal philosophy at a high level, please do not even try to comment.

* [footnote]. All lawyers I have talked to are legal realists, and all legal realists are necessarily legal positivists, as I argued long ago, so this “intractable” disagreement may mostly be an artifact of a very tiny corner of the academy. How tiny? Who actually defends Dworkin’s view? “Dworkin-lite” is popular with some constitutional law professors who don’t actually understand his view. Among serious legal philosophers, I can think of only a handful of defenders of Dworkin’s view: Nicos Stavropoulos at Oxford most prominently, but also Ronald Machado in Brazil, Dale Smith in Melbourne and maybe George Letsas in London (although relatively little of his work is devoted to defending Dworkin). Jeremy Waldron and Liam Murphy treat Dworkin’s work respectfully, which is unsurprising (Waldron was Dworkin’s student, Murphy his colleague), but neither are Dworkinians. (There are a few additional unserious and feeble defenders of Dworkin [whom I won’t name, to protect the guilty], but not even many of them.) Legal positivists vastly outnumber Dworkinians in legal philosophy over the last quarter-century. A short (Anglophone) list: Adams, Bix, Coleman, Dickson, Enoch (probably), Gardner, Giudice, Goldsworthy, Green, Kramer, Leiter, MacCormick, Marmor, Raz, Shapiro, Waluchow. Most natural law theorists, like Finnis and Mark Murphy, are closer to legal positivism on most issues than they are to Dworkin. These facts are also relevant to figuring out what the best explanation is for this state of affairs.)