Thanks to Nicholas Davies, Timothy Endicott, Brian Flanagan, Leslie Green, Christopher Hinchcliffe, Nicos Stavropoulos, Bosko Tripkovic, and Frederick Wilmot-Smith for comments on earlier drafts.

1. Quine, WV, Ontological Relativity and Other Essays (New York: Columbia University Press, 1969 Google Scholar) at ch 3 [Quine, Relativity]. The chapter stands alongside ‘Two Dogmas of Empiricism’ as exemplifying the more programmatic element of Quine’s work.

2. ‘[This is a project which] studies a natural phenomenon, viz., a physical human subject. This human subject is accorded a certain experimentally controlled input—certain patterns of irradiation in assorted frequencies, for instance—and in the fullness of time the subject delivers as output a description of the three dimensional external world and its history.’ Ibid at 82-83. Although, in his later years, Quine relaxed the view. See Quine, WV, Pursuit Of Truth, revised ed (Cambridge: Harvard University Press, 1992) at 19–21.Google Scholar

3. See, in particular, Kim, Jaegwon, ‘What Is “Naturalized Epistemology?”’(1988) 2 Google Scholar Philosophical Perspectives 381 [Kim, ‘Epistemology’] and Putnam, Hilary, ‘Why Reason Can’t Be Naturalized’(1982) 52 Google Scholar Synthese 3.

4. We might, however, note in passing that even if we come to find Quine’s argument persuasive— that is, if we come to view ourselves as having reason to accept it—we will, by virtue of this very process, remain shackled to the possibility of normativity in the sphere of epistemology—that is, by the prospects of there being good reason to accept a particular conclusion—and it will again call for philosophical explanation. The normative project seems inescapable. See Nagel, Thomas, The Last Word (Oxford: Oxford University Press, 1997 Google Scholar) at ch 2.

5. These articles are collected in Leiter, Brian, Naturalizing Jurisprudence: Essays on American Legal Realism and Naturalism in Legal Philosophy (Oxford: Oxford University Press, 2007 CrossRefGoogle Scholar) [Leiter, Naturalizing Jurisprudence]. See, in particular, ch 1-2, the postscript to part one and ch 4.

6. Ibid at 117.

7. Ibid at 39-40.

8. Ibid at 9.

9. Ibid at 40.

10. Leiter discusses a general model of this form: see ibid 187-89. There exists a burgeoning literature. See Miles, Thomas J & Sunstein, Cass R, ‘The New Legal Realism’ (2008) 75 Google Scholar U Chicago LR 831 [Miles & Sunstein, ‘New Realism’].

11. See Moore, Michael S, Educating Oneself In Public: Critical Essays In Jurisprudence (Oxford: Oxford University Press, 2000 Google Scholar) at ch 1. See also Greenberg, Mark, ‘Naturalism in Epistemology and the Philosophy of Law’ (2011) 30 Google Scholar Law & Phil 419; Greenberg, Mark, ‘Implications of Indeterminacy: Naturalism in Epistemology and the Philosophy of Law II’ (2011) 30 Google Scholar Law &Phil 453 [Greenberg, ‘Implications’]; Green, Michael Steven, ‘Leiter on the Legal Realists’ (2011) 30 Google Scholar Law & Phil 381; Green, Leslie, ‘Law and the Causes of Judicial Decisions’, online: (2009)Google Scholar Social Science Research Network 14 http://ssrn.com/paper=1374608. Leiter responds to Moore and Greenberg in Leiter, , Naturalizing Jurisprudence, supra note 5 at 103–08, 112–17 Google Scholar. Greenberg’s and MS Green’s claims are treated in Leiter, Brian, ‘Naturalized Jurisprudence and American Legal Realism Revisited’ (2011) 30 Google Scholar Law & Phil 499 [Leiter, ‘Realism Revisited’].

12. Leiter defines rational indeterminacy as follows: ‘the law on some point is rationally indeterminate if the Class [of legal reasons] (on some conception) is insufficient to justify only one outcome in that case.’ Leiter, Naturalizing Jurisprudence, supra note 5 at 9. He has a further story to tell about the realist take on the sources of indeterminacy, but this does not matter for the purposes of our argument. See ibid at 74-75.

13. Ibid at 40.

14. Dworkin, Ronald, A Matter of Principle (Oxford: Clarendon Press, 1986) at 128 Google Scholar [Dworkin, Principle].

15. This, at any rate, was his later view. See Dworkin, Ronald, Justice for Hedgehogs (Cambridge: Belknap Press of Harvard University Press, 2011 Google Scholar) at 400-09.

16. See Dworkin, Ronald, Taking Rights Seriously (London: Duckworth, 1978 Google Scholar) at ch 13, 331-38 [Dworkin, Rights].

17. Dworkin, Ronald, Law’s Empire (Cambridge: Belknap Press of Harvard University Press, 1986 Google Scholar) at 273 [Dworkin, Law’s Empire],

18. A similar argument is made in Greenberg, ‘Implications’, supra note 11.

19. As Leiter notes, to the extent that the realists could be said to have held an explicit theory of law and adjudication, it was often a crude one. He notes, for example, the tendency on their part to assume that only legislation and precedent count as sources of law to the exclusion of legally recognized customary practices. See, Leiter, , Naturalizing Jurisprudence, supra note 5 at 45 Google Scholar, n 138.

20. Ibid at 45.

21. Ibid at 46.

22. It does not follow, of course, that the theory of validity completely determines ones understanding of adjudication. Indeed, it was one of Dworkin’s least charitable criticisms of Hart’s work that his conception of legal validity was to be read as a complete theory of adjudication. See Hart, HLA, The Concept of Law, 3d ed by Green, Leslie, Raz, Joseph & Bulloch, Penelope A (Oxford: Oxford University Press 2012 CrossRefGoogle Scholar) at 238-44 [Hart, Law].

23. Dworkin, in his later work, came to doubt the importance of the concept of legal validity for a developed jurisprudence. The question of what standards were appropriately classified as legal—validity being the mode of existence of legal norms—was, he thought, philosophically uninteresting. And so he might have refused our characterisation of his view. Dworkin’s earlier arguments, on which our attribution is premised, suggest a less revisionist approach. Cf Dworkin, Rights, supra note 16 at ch 2 and Dworkin, Ronald, Justice In Robes (Cambridge: Belknap Press of Harvard University Press, 2006 Google Scholar) at 4-5, 232-34 [Dworkin, Justice],

24. Quine’s epistemological position exhibits a similar flaw. His arguments against philosophic theories of knowledge were themselves philosophical in the ordinary sense. He asked us to leave behind normative explanation by engaging in normative argument.

25. Dworkin, Ronald, ‘No Right Answer?’ (1978) 53 Google Scholar NYU LR 1 at 32.

26. See, for the new argument, Leiter, , Naturalizing Jurisprudence, supra note 5 at 184–89 Google Scholar. For worries of a similar sort to the ones aired here see Dickson, Julie, ‘On Naturalizing Jurisprudence: Some Comments on Brian Leiter’s View of What Jurisprudence Should Become’ (2011) 30 Google Scholar Law &Phil 477. For Leiter’s response, see Leiter, , Naturalizing Jurisprudence, supra note 5 Google Scholar.

27. See Segal, Jeffrey Allan & Spaeth, Harold J, The Supreme Court and the Attitudinal Model Revisited (Cambridge: Cambridge University Press, 2002 CrossRefGoogle Scholar).

28. Leiter, , Naturalizing Jurisprudence, supra note 5 at 188–89 Google Scholar.

29. This may be the best interpretation of Segal and Spaeth’s ‘legal model’ which is, in many senses, crude. Leiter admits that ‘for the purposes of presentation’ he has improved upon it. In fact, he has cleaned up their position considerably. From unsystematic references to ‘statutes’ and ‘precedents’, Leiter implies adherence to ‘hard positivism’, a theory of law developed most persuasively by Joseph Raz. Given the complexity of Raz’s theory and the fact that all theories of law take it upon themselves to explain the importance of precedent and statute, it is hard to see why this particular conception is the one to be attributed. See ibid at 187-89, and 89 n 21.

30. So it is strange to see Leiter suggesting that the predictive power of a socio-scientific theory is the reason to endorse its jurisprudential assumptions. These assumptions are precisely the elements of the theory not being tested. If the question is ‘law or politics’ an answer to the question ‘What is law?’ must be assumed. See, in response to Dickson, Julie, Leiter, , supra note 5 at 511–12 Google Scholar.

31. It is a common complaint on the part of social scientists that the analyses and theories of philosophers of law are not ‘empirically grounded’ and should, after having been formed, be ‘tested’. How one would ‘test’ Rawls’ theory of justice, or Hart’s theory of law, though, is hard to understand. A currently fashionable suggestion is that we question people about philosophical possibilities in survey form; this approach is known as experimental philosophy. The aim is to present the general public with famous hypotheticals which have been taken to motivate particular theoretical positions and record their responses. The proposal suffers a number of problems, two of which are particularly important. First, the use of hypotheticals or ‘thought experiments’ as they are sometimes known forms only one part of the philosophical method and it is hard see how one could ‘test’ other more direct forms of argument. Second, thought experiments that may be taken, unreflectively, to suggest a particular theory are often unclean, i.e., they contain extraneous factors which suggest the result favoured by a theory but not the theory itself. Here again careful philosophical work is required if we are to see the light. Surveys based on un-reflexive judgement are a poor substitute.

32. The reading is a charitable one, and many of the realists appear to go further. For Felix Cohen reference to legal rules as providing reason for judicial decisions was to be thought of as a form of ‘transcendental nonsense’. And he considered it to be of ‘.... some practical importance [that we should recognise] that the traditional language of argument and opinion [does not justify] court decisions.’ Herman Oliphant was similarly dismissive of the ‘over general and outworn abstractions in opinions and treatises.’ The admonition was to leave the jurists ‘haven of concepts’, having realised that it was a place only of appearances, and this is a general claim. See Cohen, Felix S, ‘Transcendental Nonsense and the Functional Approach’ (1935) 35 Google Scholar Colum LR 809 and Oliphant, Herman, ‘A Return To Stare Decisis’ (1928) 14 Google Scholar ABAJ 75.

33. Leiter, , Naturalizing Jurisprudence, supra note 5 at 41 Google Scholar.

34. Ibid at 41. See Leiter, Brian, ‘Explaining Theoretical Disagreement’ (2009) 76 Google Scholar U Chicago LR 1215 at 1226-28.

35. Ibid at 1227.

36. See Leiter, , Naturalizing Jurisprudence, supra note 5 at 175–78 Google Scholar. For general doubts about Quine’s arguments see Hacker, PMS, ‘Passing by the Naturalistic Turn: On Quine’s Cul-de-Sac’ (2006) 81 Google Scholar Philosophy 231.

37. Leiter, , Naturalizing Jurisprudence, supra note 5 at 39 Google Scholar.

38. Ibid at 39.

39. Ibid at 39 [emphasis added].

40. Perhaps the most important modern critic is Ronald Dworkin. See Dworkin, , Rights, supra note 16 Google Scholar at ch 2.

41. Hart, , Law, supra note 22 at 273 Google Scholar. The difference between judge made law and legislation is also explored by Gardner, John, Law as a Leap of Faith: Essays on Law in General (Oxford: Oxford University Press, 2012 CrossRefGoogle Scholar) at 37-43.

42. Leiter, , Naturalizing Jurisprudence, supra note 5 at 39 Google Scholar. One reason that the point may be overlooked is if the ‘normative theory of adjudication’ is cast, as Leiter sees it being, in terms of ‘reasons’ for decision. A permission to perform some action (φ) does not count in favour of one φing. This contrasts with a rule requiring that one φ which has normative force in relation to its object, and in this sense counts as a ‘reason’ to φ. Does the fact that a permission lacks the force of a mandatory rule render it sterile as part of a normative account? It does not, for there are a number of ways that permissions may constrain action. They may give others reason not to interfere with one’s chosen course of action, and typically in law a judicial permission includes only a limited set of options (a judge may decide A, B, or C, but not D), and in this way it circumscribes possibilities. I thank Les Green for suggesting the point.

43. The former source of indeterminacy forms only a contingent aspect of law, the latter a necessary part. See, on the latter, Endicott, Timothy, ‘Law Is Necessarily Vague’ (2001) 7 Google Scholar Legal Theory 379.

44. For the classic defence of the epistemic theory of vagueness see Williamson, Timothy, Vagueness (London: Routledge, 1994 Google Scholar). It is interesting to note that both Dworkin’s and Williamson’s views are motivated in part by the application of bivalent logic to declarative sentences framed in natural language. See Dworkin, , Principle, supra note 14 Google Scholar at ch 5.

45. This is one of the most important lessons to be learnt from the transition in thought from Wittgenstein’s Tractatus through to his mature view in The Philosophical Investigations. See, on this point, Baker, Gordon P & Hacker, PMS, Wittgenstein: Understanding and Meaning, Part I, Essays, 2d ed (Oxford: Blackwell, 2005 Google Scholar) ch 13-14.

46. The best modern account is Endicott, TAO, Vagueness In Law (Oxford: Oxford University Press, 2000 CrossRefGoogle Scholar).

47. Recently Mark Greenberg has argued that Dworkin’s theory, or at least many of its most central elements remain intact even if his view on legal determinacy is abandoned. Greenberg’s central claim is that the right answer thesis lies downstream of Dworkin’s views about the relation between law and morality, and that one might abandon his holistic understanding of morality without thereby abandoning his conception of the relation between law and morality. This is true enough, but many of the arguments that Dworkin uses against alternate theories and in favor of his own rely on a determinate view of morality. If the grounds of law are indeterminate it follows, for example, that judges have discretion in the application of law—i.e., they must make active choices not determined in advance by legal standards—and one of Dworkin’s central arguments against positivism and in favor of his own conception was that such a picture of discretion in law was unacceptable. See Greenberg, , ‘Implications’, supra note 11 at 466–68 Google Scholar.

48. Formalism requires legal determinacy because if it is a live question whether a particular standard applies (either because there exists a contrary legal standard or because the standard is a vague one) then legal reasoning will also encompass a decision on that matter. Such a decision would not be deductive in nature.

49. Leiter, , Naturalizing Jurisprudence, supra note 5 at 39 Google Scholar.

50. Ibid at 42.

51. On the limits of philosophical understanding of adjudicatory practice see Raz, Joseph, The Authority of Law: Essays on Law and Morality, 2d ed (Oxford: Oxford University Press, 2009 Google Scholar) at 206-07. It is striking that even Dworkin’s expansive conception is not intended as a concrete guide in actual cases. His argument is not that by taking up his theory anew, judges will come to make the right call in circumstances of controversy, but that they are already committed to the form of reasoning that he describes and that they cannot help but try—however imperfect their actual reasoning might be—to find the correct answer in hard cases. Dworkin is quite clear that even on his theory ‘any judge … might well be wrong in his political judgements, and that he should … decide hard cases with humility.’ Dworkin, , Rights, supra note 16 at 130 Google Scholar.

52. Justice Sotomayor, for example, said at her confirmation hearing that ‘The task of a judge is not to make the law, it is to apply the law.’

53. Consider, by way of comparison, Quine’s argument for the ‘failure’ of normative epistemology. His attack was directed against a theory known in the literature as Cartesian foundationalism, a belief being justified on this picture if either indubitable or entailed by a belief that is itself indubitable. Quine produced a number of arguments that have been taken to show that such a theory is impossible and from here he moved to recommend a naturalistic replacement.

Philosophers have rightly despaired of translating everything into observational and logico-mathematical terms. And some philosophers have seen in this irreducibility the bankruptcy of epistemology. But I think that at this stage it may be more useful to say rather that epistemology still goes on, though in a new setting and [with] a clarified status. Epistemology, or something like it, simply falls into place as a chapter of psychology and hence of natural science.

The basic move is from the failure of foundationalism through to a non-normative ‘epistemology’ of the psychological variety; Quine’s question being how evidence relates to theory in a causal sense. Assuming for the purposes of argument the failure of foundationalist epistemology we may still wonder if this entails the death of normative theorizing. It all depends, one might think, on the contenders. For if there are other plausible conceptions, ones not called into question by the same arguments which have been taken to spell the death of foundationalism, then much more needs to be established. As Jaegwon Kim notes: ‘the Cartesian project of validating science starting from the indubitable foundation of first-person psychological reports …is not the whole of classical epistemology …’ nor, one might add, of modern epistemology. Analogical reasoning, it seems, may be as prone to preserve flaws in argument as it is to transfer insight. See Quine, , Relativity, supra note 1 at 82 Google Scholar and Kim, , ‘Epistemology’, supra note 3 at 388 Google Scholar.

54. Hart, HLA, ‘American Jurisprudence Through English Eyes: The Nightmare and The Noble Dream’ (1977) 11 Google Scholar Georgia LR 969 [Hart, ‘Jurisprudence’]. See also Hart, , Law, , supra note 22 at 124–41 Google Scholar. For Leiter’s response see Leiter, , Naturalizing Jurisprudence, supra note 5 Google Scholar at ch 2.

55. Hart, ‘Jurisprudence’, ibid at 989.

56. In a recent response to Mark Greenberg, Leiter has suggested that he meant the phrase ‘theory of adjudication’ to apply only to theories which do attempt to ground judicial decision-making in a determinate manner. In doing so he preserves—by stipulation—his conclusion that ‘the normative theory of adjudication’ fails, but at a deeper level the argument remains incomplete (why, after all, is a theoretical understanding of legal justification which concedes indeterminacy not equally entitled to the label?). Frank Jackson has noted the danger in cases like this of turning ‘interesting philosophical debates into easy exercises in deductions from stipulative definitions together with accepted facts.’ For Leiter’s argument, see Leiter, , ‘Realism Revisited’, supra note 11 at 506–08 Google Scholar and for Jackson’s worries about this way of proceeding see Jackson, Frank, From Metaphysics to Ethics: A Defence of Conceptual Analysis (Oxford: Clarendon Press, 1998 Google Scholar) at 30-31.

57. Although Leiter himself would want to eschew any such distinction as ultimately plausible. He follows Quine in understanding philosophy as the abstract and reflective branch of the sciences including, in Leiter’s case, the social sciences.

58. Leiter, , Naturalizing Jurisprudence, supra note 5 at 42 Google Scholar.

59. See Miles, & Sunstein, , ‘New Realism’, supra note 10 Google Scholar.

60. Leiter, , Naturalizing Jurisprudence, supra note 5 at 52 Google Scholar.

61. See Dworkin, , Justice, supra note 23 Google Scholar at ch 6-7 and Dworkin, Ronald, ‘Hart and the Concepts of Law’ (2006) 119 Google Scholar Harv LR Forum 95.

62. See, in particular, Leiter, Brian, ‘In Praise of Realism (and against Nonsense Jurisprudence)’ (2011) 100 Google Scholar Georgetown LJ 865.

63. Dworkin, , Law’s Empire, supra note 17 at 90 Google Scholar.