1. It is not, of course, my intention to imply that the ideas that comprise the conventionalist package are definitive of the conventionalist position or that they are the most important among the theses advanced by legal conventionalists.

2. Shapiro, S., Law, Plans and Practical Reason, 8 Legal Theory387, 441 (2002)Google Scholar. Kutz, Christopher, The Judicial Community, 11 Phil. Issues442 (2001)Google Scholar, makes an analogous point. Michael Bratman has also expressed doubts whether it is warranted to say that contemporary theories in this tradition are about conventions, properly so called; Bratman, M., Shapiro on Legal Positivism and Jointly Intentional Activity, 8 Legal Theory 511, 515–516 (2002).Google Scholar

3. It should be clear from this summary that my reconciliatory project is limited in scope; it does not concern the compatibility of the legal-conventionalist package with non–natural-law theories of law.

4. H.L.A. Hart, The Concept of Law (2nd ed. 1994), esp. ch. VI.

5. It is a further question that I do not pursue here whether Hart would endorse those elaborations. An indication is the fact that in the Postscript to The Concept of Law, he explicitly embraces the view that the rule of recognition is conventional. See Hart, supra note 4, at 255–256. But this indication is not universally taken as conclusive or in line with Hart's original account. See Dickson, J., Is the Rule of Recognition really a Conventional Rule, 27 Oxford J. Legal Stud. 1 (2007)Google Scholar; and see, more generally, J. Gardner, Some Types of Law, in Common Law Theory (D. Edin ed., 2007).

6. There are also hybrid theories, like Shapiro's, that seek to combine the ideas of convention and authority within a legal-positivist framework.

7. See, e.g., Hart, supra note 4, at 114–115.

8. Id. at 101–102. In the Postscript, Hart revisits the issue by saying that the rule of recognition “is in effect a form of judicial customary rule existing only if it is accepted and practiced in the law-identifying and law-applying operations of the courts.” Id. at 256 (emphasis added). But again, I am not sure whether Hart means to restrict the rule of recognition to courts or is just taking the most straightforward case.

9. Id. at 117.

10. Id. at 115 (emphasis added).

11. Id. at 116. This is not to deny that ordinary citizens may also “acknowledge an obligation to obey” the law as a public common standard in the same way that legal officials must of necessity do. Hart of course allows for this possibility. But it is an interesting question if he also thinks that they thereby become actual participants in the social practice of the rule of recognition or that they merely vicariously partake in it.

12. Shapiro, supra note 2, at 419. Legislatures are, of course, not absent in other accounts, although they are not awarded such a prominent role. Thus, for Postema, legislatures provide a way of facilitating coordination, his version of the point of legal practice, by rendering certain solutions to coordination problems salient. See Gerald, Postema, Coordination and Convention at the Foundations of Law, 11 J. Legal Stud. 164, 185–186 (1982)Google Scholar.

13. Postema, supra note 12, at 189.

14. Id. at 191. Interestingly, even Shapiro, who insists that “it strains ordinary usage to describe [ordinary citizens] as ‘participants in the legal system”’ (Shapiro, supra note 2, at 418) nevertheless concedes that “[i]n democracies, for example, citizens are generally entitled to vote for officials and hence they do participate in the operations of the legal system”; id. at 418 n. 43.

15. For an illustration of this point, see H.L.A. Hart, Essays on Bentham 158 (1982).

16. Hart, supra note 4, at 203.

17. Apart from Postema, other theorists who have been at times attracted to the coordination model include Jules, Coleman, Authority and Reason, in The Autonomy of Law: Essays on Legal Positivism 278 (R. George ed., 1996)Google Scholar; Gans, C., The Normativity of Law and Its Coordinative Function, 16 Isr. L. Rev.333 (1981)Google Scholar. The coordination model in law has suffered heavy blows from criticism by Green, L., Positivism and Conventionalism, 12 Can. J. L. & Jurisprudence35 (1999)Google Scholar; and Marmor, A., Legal Conventionalism, 4 Legal Theory509 (1998)Google Scholar.

18. Postema, supra note 12, at 187.

19. Id. at 187–190.

20. This is the barest outline of the position that of course does not do justice to its sophistication. But it does not seem to me to be necessary for present purposes to go into more detail. For the full account, see Bratman, M., Shared Cooperative Activity, 101–102 Phil. Rev.327 (1992)Google Scholar; Bratman, M., Shared Intention 104 Ethics97 (1993)Google Scholar. Jules Coleman, who initially sided with the coordination-based understanding of the rule of recognition, later came to reject it in favor of a Bratman-like understanding. See Jules Coleman, The Practice of Principle: In Defence of a Pragmatist Approach to Legal Theory 92 (2001). Other theorists who have propounded similar views include K. Himma, Inclusive Legal Positivism, in The Oxford Handbook of Jurisprudence and Legal Philosophy (J. Coleman and S. Shapiro eds., 2001); and Shapiro, who nevertheless denies that the joint activity of law necessarily includes anything like a commitment to mutual support; see Shapiro, supra note 2, at 429–431.

21. Shapiro, supra note 2, at 407.

22. Id. at 414.

23. Postema, supra note 12, at 178:

24. See Coleman, supra note 20, at99.

25. Shapiro, supra note 2, at 440, note 62.

26. Id. at 415.

27. M. Bratman, Shared Cooperative Activity, at 340.

28. The term was coined by Christopher Kutz; see Kutz, C., Acting Together, 61 Phil. & Phenomenological Res.1 (2000)Google Scholar. Kutz has extended this conception to the case of law in Kutz, supra note 2.

29. Kutz, supra note 28, at 10.

30. Id. at 464.

31. It is controversial whether Dworkin's interpretivism should be counted among natural-law theories, at least as traditionally understood. Dworkin himself seems to be comfortable with the label. Cf. Dworkin, R., Natural Law Revisited, 34 U. Fla. L. Rev.165 (1982)Google Scholar. One of the reasons I insist on using the generic label is that the arguments I consider in this paper are directed at natural law in general.

32. See, generally, R. Dworkin, Law's Empire (1998). Interpretivism is not the only theory that espouses both theses. We can imagine, for instance, a robust natural-law theory that carves up the test for true propositions of law in two stages. At the first stage, “legal standards” are identified in a value-free manner. At the second, these standards are sifted by appeal to some moral condition that is thought to be essential to law. Only standards that pass the second hurdle are “genuine” legal standards. On this, see N. Stavropoulos, Interpretivist Theories of Law, in The Stanford Encyclopedia of Philosophy (E. Zalta ed., 2003), available at http://plato.stanford.edu/archives/win2003/entries/law-interpretivist.

33. John Finnis, Natural Law and Natural Rights 14 (1980).

34. See, e.g., id. at 281–290.

35. Id. at 355.

36. Id. at 354.

37. Id. at 357.

38. Shapiro, supra note 2, at 437.

39. Id.

40. One possible exception might be the type of natural-law theory advanced by Deryck Beyleveld and Roger Brownsword. They start from Alan Gewirth's moral philosophy to claim that practical reason in general presupposes moral reason. See D. Beyleveld &Brownsword, R., Law as a Moral Judgment 119–150 (1994)Google Scholar. For Gewirth's own take on the issue, seeGewirth, A., The Ontological Basis of Natural Law, 29 Am. J. Jurisprudence95–121 (1984)Google Scholar.

41. Dworkin has challenged this claim on a number of occasions and more recently in Dworkin, R., Thirty Years On, 115 Harv. L. Rev.1655 (2002)Google Scholar, reprinted in R. Dworkin, Justice in Robes 187–222 (2006).

42. Shapiro, supra note 2, at 438.

43. Id.

44. For illuminating discussions of this point, see Greenberg, M., How Facts Make Law, 10 Legal Theory157 (2004)Google Scholar; Stavropoulos, supra note 32.

45. Finnis, supra note 49, at 290.

46. Recall the distinction Dworkin draws in Model of Rules II between concurrent and conventional morality. See Dworkin, R., Model of Rules II, in Taking Rights Seriously 53 (1977)Google Scholar. In both cases there is some agreement over what to do, that is, a social fact. What differs is the ground of that agreement. Conventional morality treats the fact of the agreement itself as part of the reason why we ought to do such-and-such. Concurrent morality does not.

47. Finnis, supra note 49, ch. 10.

48. Shapiro, supra note 2, at 421 (emphasis in the original).

49. Id. at 439.

50. Id. at 440.

51. Id. at 439 (emphasis added).

52. Id. at 414.

53. Id. at 439.

54. As we see above, Shapiro agrees that the notion of joint activity is compatible with the existence of widespread disagreement. See Section II.B.

55. How much disagreement is “just enough”? When does it become impossible to handle? I am not sure we can answer such questions in the abstract. For an excellent analysis of this problem in the context of a legal system see J. Waldron, Law and Disagreement (1999), ch. 9.

56. The locus classicus for a discussion of various senses of discretion is Dworkin, supra note 46, at 31.

57. Id. at 31.

58. Thanks to Mike Giudice for suggesting this qualification.

59. I have mentioned only three of the most common modalities. There may be others. For Shapiro's own take on the idea of law's practical difference, see Shapiro, S., On Hart's Way Out, 6 Legal Theory127 (2000)Google Scholar.

60. Shapiro, supra note 2, at 426.

61. Cf. T. Christiano, The Authority of Democracy, 12–13 J. Pol. Phil. 266 (2004). For an application of this distinction, see Dimitrios Kyritsis, Principles, Policies and the Power of Courts, 20–22 Can. J. L. & Jurisprudence 379 (2007).

62. Second-order considerations also determine the jurisdiction of a certain agent of governance. Jurisdiction is here understood as the range of issues that an agent of governance is recognized as having the power to regulate or the kinds of acts that it may dictate in the exercise of its power. In this sense, jurisdiction may be thought to bear on the content of a decision rather than on the way that decision was produced. However, respect for jurisdiction is not part of what it means for a decision to dictate the right thing, and this is why I do not include it in the first parameter. An agent of governance may have acted intra vires and still have got it wrong. Jurisdiction rather reflects a concern that a certain decision is made by the proper person or institution, which is why I think that it ought to be counted among second-order ideas.

63. Rawls, J., Two Concepts of Rules, in Collected Papers 20, 39 (S. Freeman ed., 1999)Google Scholar.

64. I also assume that the value of democracy is at least partly independent of the outcome of democratic decision-making. Many theorists deny this. I cannot at this point say anything to defend my assumption to these theorists. I only hope that even if we disagree about democracy, there is some other second-order idea that they find more compelling, in which case they can pick the idea they favor and work with that.

65. I owe this label to Mattias Kumm.

66. There are, of course, many other responses to the value of democracy that are also appropriate. Such responses, such as feeling proud of one's legal system, need not detain us here. Besides, democracy is valuable even for people who are not citizens of democratic states. These people value it by refraining from obstructing the normal working of democratic states or by writing pamphlets about its advantages.

67. Rawls, supra note 63, at 31.

68. Carlos Rosenkrantz has pointed out to me that a full account of legal practice along the lines I am suggesting needs to be supplemented by a story that identifies criteria of participation in a specific joint activity. Why must a judge heed the decisions of one democratic legislature rather than another? This is an important point that I cannot address in this paper. Here I limit myself to sketching the structure of the reason that participation in legal practice gives rise to on the natural-law view.

69. Which role the value of democracy and other second-order considerations play in determining the nature and content of judicial duty crucially depends, among other things, on the conception of the proper combination between second- and first-order considerations one happens to espouse. It must be stressed that the purpose of the discussion so far is to suggest not what that conception must look like more specifically but rather how the availability of second-order considerations changes judicial duty in a way that helps us rebut the straight-to-morality assumption.

70. Even after we determine in what ways second-order considerations may be relevant to the practical reasoning of ordinary citizens, we still need to decide which of the ways in which they are relevant may properly be regarded as part of legal practice rather than part of the broader political culture. For a related point, see Raz, J., The Problem of Authority: Revisiting the Service Conception, 90 Minn. L. Rev.1003, 1004 (2006)Google Scholar.

71. I am not precluding that such a system of governance may still be legitimately called “law” in some further sense. For a distinction of various “concepts” of law, see R. Dworkin, supra note 41, Introduction.

72. The standard-bearer of this view is Joseph Raz. See J. Raz, The Morality of Freedom (1986), chs. 1–4; and more recently, see Raz, supra note 70.

73. See J. Waldron, supra note 55, ch. 5.

74. T.R.S. Allan makes a similar point in the context of his critique of the concept of due deference in the theory of judicial review. Here is how he puts it:

Allan, T.R.S., Human Rights and Judicial Review: A Critique of “Due Deference,” 65 Cambridge L.J.671, 688 (2006)Google Scholar.

75. Wollheim, R., A Paradox in the Theory of Democracy, in Philosophy, Politics, Society 71 (Laslett & Runciman eds., 1969)Google Scholar.