, but it seems worth raising again, if only as a way of trying to rescue analytic legal philosophy from the inbred sterility of its current research program. Let us begin with a brief sense of the terms of the question: (i) the concept of law and (ii) the Rule of Law.
What is the relation between the concept of law and the Rule of Law? The question has come up before in jurisprudence
(i) The concept of law.
Analytic legal philosophers (like Jeremy Bentham, John Austin, H.L.A. Hart, Jules Coleman, and Joseph Raz) ask and answer questions like these: “What is law?” “What is a proposition of law?” “What is legal validity?” They make a study of the concept of law. They intend that study to help clarify the meaning of straightforward propositions such as “It is against the law to drive faster than 70 kilometers per hour” as well as more controversial propositions such as “It is lawful in certain circumstances for police officers to enter private dwellings without knocking.”
They also intend their study of the concept of law to explicate and illuminate claims about the existence of legal systems in particular societies such as “Mexico has a legal system” (though it has to be said that their interest in statements of the latter kind is mostly for the sake of the contributions they make to the truth-conditions of statements of the former kind).
(ii) The Rule of Law.
Political philosophers (and some legal theorists) study the Rule of Law, which is a political ideal—one of a number of ideals which define free societies in the modern world. (The others include democracy and human rights.) The Rule of Law celebrates features of a well-functioning system of government such as publicity and transparency in public administration, the generality and prospectivity of the norms that are enforced in society, the predictability of the social environment that these norms help to shape, the procedural fairness involved in their administration, the independence and incorruptibility of the judiciary, and so on.
It looks to a world where people in positions of power exercise their power within a constraining framework of public rules rather than on the basis of their own preferences, their own ideology, or their own individual sense of right and wrong.
Our question is: what is the relation between (i) and (ii)? Do those who study the concept of law focus on something different than those who study the Rule of Law? Do they focus on the same thing, but in a different way? Does the enterprise of studying one of these things build upon the enterprise of studying the other?
Grammar suggests that we need to understand the concept of law before we can understand the Rule of Law. “The Rule of Law” is a complex phrase and the word “law” is one of its components. Just as we cannot understand a phrase like “the protection of human rights” unless we understand the smaller component phrase “human rights,” and just as we cannot understand “the spread of democracy” without a grasp of the meaning of “democracy,” so too we cannot understand the meaning of “the Rule of Law” unless we already grasp the concept of law. I will call this the ROL COL position: it holds that talk of the Rule of Law presupposes that we have already made independent sense of the concept of law.
In this paper I will present the case for the contrary position. This we can label as COL ROL: the concept of law already implicates the ideal we call the Rule of Law. To identify a society as having a system of law, as opposed to some other sort of order, is to identify it as satisfying some or all of the requirements associated with the Rule of Law. If grammar is an issue, then perhaps we can understand the occurrence of “law” in “the Rule of Law” as syncategorematic; the phrase should be read as though it were just one word—“ruleoflaw.” Ruleoflaw can then be regarded as something that figures non-tautologously in our analysis of the concept of law itself.
Anyway the exact parsing of the phrase matters less than the substantive position. Everyone agrees that there are a number of ways of ordering a society and that only some of them count as law.But modern jurists – particularly modern positivists—are quite casual about what a system of governance has to be like in order to earn the appellation law. If it calls itself a system of law, they are very reluctant to question that self characterization.
Or if their starting point is not the self-characterizations of actual systems of governance, their own a priori specifications are very broad: basically any well-organized system of centralized order by the use of spoken or written prescriptions and prohibitions counts as law, at least if the prescriptions and prohibitions that feature in the ordering can be recognized as such, and referred to and talked about, by those whose actions and decisions constitute the ordering’s human face. I want to propose in this paper that a philosophy of law should be less accommodating than this.
 University Professor in the School of Law, New York University. I am grateful to Mark Bennett, Jules Coleman, Ronald Dworkin, David Dyzenhaus, Kent Greenawalt, Stephen Perry, Joseph Raz, Scott Shapiro, Tom Campbell, and Ben Zipursky for discussions of this topic over the years. This paper was originally prepared for the Second Congress on the Philosophy of Law, Institute of Law Research, Mexico City, March 27-31, 2006. It benefited there from comments by Mitch Berman, Tom Campbell, Julie Dixon, Imer Flores, Andrei Marmor, Enrique Villanueva,
and Will Waluchow.
 See, e.g., Fuller 1969, Dworkin 2004, Simmonds 2005, and Bennett 2005
 For a forceful opinion of the sterility of modern analytical legal philosophy—particularly the
research program of modern positivism—see Dworkin 2002, 1677-80
 Cf. Hudson v. Michigan, Supreme Court of the United States, No. 04-1360, decided June 15,
 I capitalize the term “the Rule of Law” to distinguish it from the phrase “a rule of law” which
may be used to refer to a particular legal rule such as the rule against perpetuities or the rule in the United States that the President must be at least thirty-five years old.
 The direction of the arrow tells us about the order of implication: so, COL ROL entails ~ROL ~COL. The arrow points in a direction opposite to the order of understanding. So “ROL COL” expresses the claim (which I oppose) that the Rule of Law implies the concept of law, which of course indicates that the concept of law has to be understood first. It is like “Bachelor Unmarried,” which claims that a proposition abut a person being a bachelor implies a proposition about that person’s being unmarried. But as to the order of understanding, we have to understand how to use the term “unmarried” before we can understand “bachelor.”
 Thus most legal theorists admit the possibility of acephalous societies or purely customary “prelegal”societies: see e.g., Hart 1992, 91-4.
 Most rulers try to use the word “law” to describe their mode of rule, no matter what it is, just as most latter-day rulers use the word “democracy.” The difference is that in the case of democracy, political theorists are not afraid to challenge them: we don’t pander to the authoritarians and say that anything that calls itself a democracy is a democracy. I fear that, by contrast, modern analytic legal philosophy has been infected with an ethos of pandering to those who call their system of rule “law.” If it is an effective system of rule and it’s called “law,” then it is law, we say; we do not even look at the detail of its self-presentation, let alone at the way in which rule is actually carried out. It is time to stop this; it is time to be a little more discriminating. This is perhaps a caricature of a theory like that of H.L.A. Hart 1992; but, in my view, it is not as much of a caricature as it ought to be.