1. See, e.g., H.L.A. Hart, Essays on Bentham (1982), ch. 10; Joseph Raz, The Morality of Freedom (1986), at 35–37; Leslie Green, The Authority of the State (1988), at 36–48; Marmor, Andrei, Authorities and Persons , 1 Legal Theory 337 (1995)CrossRefGoogle Scholar; May, Thomas, On Raz and the Obligation to Obey the Law , 16 Law & Phil . 19 (1997)Google Scholar; Christopher W. Morris, An Essay on the Modern State (1998), at 207; Shapiro, Scott, On Hart's Way Out , 4 Legal Theory 469 (1998)CrossRefGoogle Scholar; Einar Himma, Kenneth, H.L.A. Hart and the Practical Different Thesis , 6 Legal Theory 1 (2000)CrossRefGoogle Scholar; Gardner, John, Legal Positivism: 5 1/2 Myths , 46 Am. J. Juris . 199 (2001)CrossRefGoogle Scholar; Perry, Stephen, Law and Obligation , 50 Am. J. Juris . 263 (2005)CrossRefGoogle Scholar; Klosko, George, Are Political Obligations Content Independent? , 39 Pol. Theory 498 (2011)CrossRefGoogle Scholar; Gur, Noam, Are Legal Rules Content-Independent Reasons? , 5 Problema 175 (2011)Google Scholar; Viehoff, Daniel, Democratic Equality and Political Authority , 42 Phil. & Pub. Aff . 337 (2014)CrossRefGoogle Scholar. As this list shows, issues surrounding content-independence are also important in a wide variety of debates beyond political obligation, especially in philosophy of law.

2. A. John Simmons, Moral Principles and Political Obligations (1979).

3. See Klosko, supra note 1, at 516.

4. Green, supra note 1, at 226, claims content-independence is necessary “in any argument purporting to establish the existence of a political obligation”; Hart, supra note 1, at 268, notes that it is one of the “elements which are crucial to the understanding of law.”

5. Markwick, P., Law and Content-Independent Reasons , 20 Oxford J. Legal Stud . 579 (2000)CrossRefGoogle Scholar; and Markwick, P., Independent of Content , 9 Legal Theory 43 (2003)CrossRefGoogle Scholar; Sciaraffa, Stefan, On Content-Independent Reasons: It's Not in the Name , 28 Law & Phil . 233 (2009)CrossRefGoogle Scholar.

6. In contrast to practical reasons, theoretical reasons count in favor of believing rather than acting. I am concerned only with practical reasons in this paper and should be understood to be referring to practical reasons when I use “reason” in this paper. That said, theoretical reasons can also be content-independent, which is important for characterizing some phenomena such as expertise and testimony (see also Hart, supra note 1, at 261, although Marmor disagrees). I think the account of content-independence presented here can be applied equally well to theoretical reasons, but that goes beyond the scope of this paper.

7. On reasons as facts, see Joseph Raz, Practical Reasons and Norms (2d ed. 1999), at 17–19. My view that reasons are facts makes a significant difference to my arguments below. If reasons are propositions or mental states, as some claim, then it is not clear whether I have a case against Markwick in Section III, for example. I think the best view of reasons is that they are facts, but clearly I cannot engage that debate here.

8. See also Hart, supra note 1, at 254–255.

9. Raz, Morality, supra note 1, at 35.

10. R.P. Wolff, In Defense of Anarchism (1970).

11. Hart, supra note 1.

12. Id. at 255; Raz, Morality, supra note 1, at 35.

13. Raz, Practical Reasons, supra note 7, at 70.

14. Raz, Morality, supra note 1, at 36; Shapiro, supra note 1, at 493; Marmor, supra note 1, at 355.

15. Raz, Morality, supra note 1, at 54.

16. Id. at 35; see also Marmor, supra note 1, at 345; and Joseph Raz, The Authority of Law 234 (2d ed. 2009).

17. Hart, supra note 1, at 254–255; and H.L.A. Hart, Legal and Moral Obligation, in Essays in Moral Philosophy (A.I. Melden ed., 1958). Among others, Shapiro and Green similarly focus on the nature of the act.

18. Hart, Legal and Moral, supra note 17, at 100.

19. See Sciaraffa, supra note 5, at 242–243, for some worries about this notion of a direct connection.

20. Marmor, supra note 1, at 345.

21. In his notebooks, Hart calls the insight that rules could bind in virtue of their “source” rather than their content “revelatory.” Nicola Lacey, A Life of H.L.A. Hart: The Nightmare and the Noble Dream (2006), at 227.

22. Thomas Hobbes, Leviathan (Richard Tuck ed., 1991), at 176.

23. Thus it is not the case that content-dependent reasons are simply “reasons that are not content-independent reasons,” pace Sciaraffa, supra note 5, at 234.

24. The cup is importantly different from the command. Perhaps most importantly, commands claim to be content-independent obligations, not just content-independent reasons generally. There is probably no feature of a cup that could ever be a content-independent obligation, even if there are features that are commonly content-independent reasons. These reasons will often be outweighed by content-dependent reasons; for example, if the cup contains poison, the fact that the poison will kill you is a conclusive content-dependent reason not to drink from the cup.

25. Thanks to anonymous reviewer for encouraging me to clarify this.

26. I am using “weight” in a very broad sense to capture what it is about a reason that binds, which is also expressed by other metaphors, such as the “strength” of a reason, and includes Raz's controversial idea of exclusionary force as a special kind of defeat.

27. We could perhaps say that a reason with no weight is not a reason at all, and so subsume the reason's existence to its weight. I think this is probably correct, but for purposes of clarity I am keeping existence and weight distinct.

28. See also William A. Edmundson, Three Anarchical Fallacies (1998), at 13.

29. This specification matters, for as Markwick, Law, supra note 5, at 594, points out, we surely cannot make sense of a reason that was independent in all ways from its content: this is “an important respect in which legal reasons are content-dependent rather than content-independent.” I am specifying other important respects in which reasons can be independent from content, and these respects can be kept distinct in order to capture a property of some reasons that it still makes sense to call “content-independence” even if it is not independent in the most absolute sense. See Gardner, supra note 1, for some further ways that content matters for the reason.

30. A reason is content-dependent iff it has a container as a constituent part and changing the content of that container determines whether the agent has a reason or determines the weight of that reason.

31. Features of the speech act other than the source can also constitute content-independent reasons. If you have sensitive hearing, the volume of the speech act can be a content-independent reason for you to cover your ears, for example. Volume is a feature of the container itself, not the content of the container (whether you are being berated or praised, the volume of the speech act constitutes a reason) nor its source (whether your employer or your partner is being loud, the loudness gives you a reason).

32. Not all laws command, of course, as many function to empower or permit, for example, rather than obligate. Obligation is a particular case of an authority's directive changing normative standing, i.e., one way in which authority understood as a Hohfeldian power can be exercised. The puzzling features of content-independence arise for all the various ways authorities can bind subjects and change normative standing. I focus on the case of obligation because it is most familiar and most relevant to the literature, but my analysis of content-independence applies mutatis mutandis to all kinds of authoritative directives.

33. Perhaps all laws would bind on views of political authority that take, for example, any actual monarch to be the representative of god's will and exercising god's authority. Yet even these views (and secular views like them, as with Hobbes) impose some limits on the injustice of particular laws; see Klosko, supra note 1, at 504–505.

34. As noted in Green, supra note 1, at 47, content-independence must also be made consistent with the fact that “grossly immoral promises do not bind.” See also Hart, Legal and Moral, supra note 17, at 102. Ultimately, however, I think these problems (and their solutions) will be distinct, as they rest on the distinct natures of promising and command rather than on content-independence per se.

35. Klosko, supra note 1, at 500; see also id. at 518 n.5.

36. See also Raz, Morality, supra note 1, at 62. It seems to me that Raz's explanation of the puzzle, which focuses on its consistency with the preemptive character of commands rather than their content-independence, fails because it appeals to the psychological states of the subject to explain what is a justificatory, normative issue. As Raz himself notes in his discussion of Hart, the preemptive force of commands is not about what is going on inside subjects’ heads.

37. Klosko, supra note 1, at 516ff.

38. Pace id. at 504.

39. See Raz, Morality, supra note 1, at 61–62, 78–79, for the general point about authority and the particular case of political authorities respectively.

40. Klosko uses this language and related analogues, but in doing so he is being skeptical about the notion of content-independence.

41. Green, supra note 1, at 51.

42. Of course, the same fact can constitute more than one reason and could constitute both a content-independent reason and a content-dependent reason. If we have a general political obligation, then we have both content-independent and content-dependent reasons to conform to the law proscribing murder: a content-independent reason because it is the law, and a content-dependent reason because we have extralegal moral reasons to refrain from murder simply because it is murder.

43. The debate over political obligation is so difficult in part because it is about not just whether laws constitute content-independent reasons to act, which they often do in a variety of ways, but whether we have a general content-independent obligations to obey.

44. That said, even theories of legitimacy that are not as thoroughgoingly instrumentalist as Raz's often include issues of reliable performance as a requirement on the justified exercise of political power, e.g., in the guise of a “reasonable likelihood of success” requirement. Insofar as this is true, they may be able to appropriate the solution I use here.

45. Raz, Morality, supra note 1, at 53. I personally endorse an instrumentalist theory of authority that significantly modifies the Razian approach, which allows a more direct and more clearly motivated solution to the puzzle of content-independence. But the service conception is widely known, widely influential, and sufficiently relevantly similar to my preferred view, so I use it as the basis of my discussion here.

46. Id. at 58–59.

47. Goldman, Alvin I., Experts: Which Ones Should You Trust? , 63 Phil. & Phenomenological Res . 85 (2001)CrossRefGoogle Scholar.

48. How can novices judge track records, if that is a function of content? Some kinds of successes will be relatively obvious: doctors curing diseases, for example. Other successes become obvious in retrospect, like the predictions of astronomers about the time of an eclipse. See id. at 106–108. As an anonymous reviewer pointed out to me, the egregious injustice is not quite the same as a track record considered as a diachronic pattern of past performance. Instead it is more akin to a physician who negligently kills a patient. The point in either case is that the quality of the result can be assessed even by nonexperts and matters (a great deal) for how we judge the quality of the agent's capacities more generally.

49. Consider the boy who cried wolf.

50. The details of this “infection” would have to be carefully spelled out, especially in relation to the piecemeal nature of authority on a Razian account. Although I cannot address this issue here, on my view the extent and limits of the infection are demarcated by issues of trust and trustworthiness, analyzed along the lines of Baier, Annette, Trust and Antitrust , 96 Ethics 231 (1986)CrossRefGoogle Scholar.

51. Notice that the way the decision would poison our assessment of the government's judgment as a whole would be a function of the clear egregiousness of the decision, not the mere fact that it would be an injustice. This is consistent with our considered judgments (slightly unjust laws can bind) as well as with our understanding of authority (some mistaken commands must bind).

52. We could, I think, imagine cases where a mistake was a blip in A's judgment and she rehabilitates herself to the point where she has authority over B again (especially in the case of large and complex institutional authorities, where a rogue official might explain a blip). But this could not happen immediately; the correct description of the case would be that she lost authority for a while and then regained it, not that she had it throughout. Further, the more egregious the mistake, the less likely that it could be a blip. For a variety of reasons, legalizing slavery again could not reasonably be considered a blip.

53. We can also imagine an analogous case where the content of a cup infected the container and resulted in content-independent reasons not to drink from the container, as when a cup becomes irradiated by some radioactive content. The container's features are changed such that even if you filled it with fresh water, you would have reasons not to drink from it because the cup itself would harm you. The radioactive content was indirectly relevant to whether you have a reason to drink from the container, but only because it changed the container itself; features of the container still determine the existence and weight of the reason.

54. I think a similar solution works for those who worry about authority's scope; see, e.g., Sciaraffa, supra note 5, at 240; and Klosko, supra note 1, at 506. Content is indirectly relevant for considerations of scope insofar as content determines the favored action, but this does not implicate content-independence properly understood. This topic is complex, and I cannot address it here.

55. Markwick, Law, supra note 5; and Markwick, Independent, supra note 5.

56. Sciaraffa, supra note 5.

57. Notably, I accept one of Markwick's other arguments showing that the favored action must be determined by content. Also, Markwick is careful to note that he does not consider all possible understandings of content-independence and was trying to reconstruct the idea from brief and often vague discussions. One way of seeing my argument here is as proposing an understanding that Markwick does not consider and is immune to his challenges.

58. Markwick, Law, supra note 5, at 591; and Markwick, Independent, supra note 5, at 58.

59. Confusion about the nature of the content is not restricted to Markwick. As shown below, Sciaraffa's theory leads him to restrict his attention to the content of intentional acts, and Gur, supra note 1, restricts attention to the content of requirements. A virtue of my account is how it captures the breadth as well as the limitations on the contexts in which content-independence can be usefully applied.

60. Raz, Morality, supra note 1, at 35.

61. Markwick, Independent, supra note 5, at 44 n.6 and 45 n.9. See also Markwick, Law, supra note 5, at 580 n.5, although there is no explicit reference to Raz in this case.

62. Markwick, Law, supra note 5, at 587 n.22; and Markwick, Independent, supra note 5, at 45 n.9.

63. Raz argues that threats are content-independent reasons to believe, not practical reasons at all. See Raz, Morality, supra note 1, at 36. The correct analysis of threats is not our point, however; our point is that the notion of content-independence is still being used in the context of a container.

64. Note my qualifier, supra note 57.

65. Raz, Practical Reasons, supra note 7, at 48 and 51; see also Shapiro, supra note 1; as well as Himma, supra note 1, at 27, where he ascribes content-dependence to moral norms.

66. Marmor, supra note 1, at 346. I think Marmor is actually mistaken about maps only ever constituting content-dependent reasons. In the normal case, this is true. But imagine that you found the map of a distant ancestor and wanted to follow in her footsteps. In this case you do not care whether the map is accurate, you care only that it was your ancestor's map because you care about the journey and not the destination. It is a feature of the container (provenance), not the content (accuracy), that provides the reason for you to (for example) follow a certain path. Of course changing the content of your ancestor's map would change what trail you should follow, but this is consonant with my understanding of content-independence, as it changes the favored action but not the existence or weight of your reason.

67. Sciaraffa, supra note 5, at 234, 246, and 254.

68. Thanks to an anonymous reviewer for pushing me on this point.

69. Sciaraffa, supra note 5, at 235.

70. Klosko, supra note 1, at 499.

71. The law can, of course, tolerate a certain amount of nonconformity in both theory and practice. The point is that nonconformity cannot be a fundamental and pervasive feature of the system. See A. John Simmons, The Duty to Obey and Our Natural Moral Duties, in Christopher Heath Wellman & A. John Simmons, Is There a Duty to Obey the Law? (2005), at 168–169.

72. Hart, Essays on Bentham, supra note 1, at 253.

73. See Raz, Morality, supra note 1, ch. 3, esp. at 58–60.

74. Id. at 58.

75. This is absolutely central to Raz's view. Klosko, supra note 1, at 499 n.5, claims that Raz's view is actually content-dependent, giving two reasons. One is that political obligations are not compatible with clear injustices, which I address above. The other is that it is limited in scope. Neither of these tells against content-independence. Further, this would be a massive, fundamental revision of Raz's view, given his discussions of content-independence. Raz says that the focus on the speech act rather than its content is of “great importance.” Raz, Morality, supra note 1, at 35.

76. Recall (supra note 26) that “weight” is being used to include exclusionary force. Whether the command constitutes an obligation still depends on features of the law and not features of the content, regardless of whether we regard the bindingness of the obligation in terms of weight or exclusion.

77. See also Hart, Legal and Moral, supra note 17, at 104.

78. See, e.g., Simmons, supra note 2, at 194ff.

79. Again, some restrictions on content are consistent with content-independence as I define it, including, most important, the limit on egregiously unjust laws that I discuss in Section II.

80. Klosko, supra note 1, at 501. It is also worth noting that Klosko relies on Markwick's definition of content-independence, which is strange because Klosko does not claim that the search for content-independent political obligation fails on conceptual grounds but rather that purely legal reasons cannot establish such an obligation for substantive reasons. Yet on Markwick's definition, purely legal reasons (and any other reasons) would fail to establish any coherent and useful content-independent political obligation as a conceptual matter.

81. Klosko, supra note 1, at 511–514, 517.

82. Klosko, George, Content-Independent Obligations: A Reply to Kevin Walton , 42 Pol. Theory 223 (2014)CrossRefGoogle Scholar, at 224–225.

83. See Klosko, Are Political Obligations, supra note 1, at 504–505.

84. Id. at 500–501.

85. Walton, Kevin, The Content-Independence of Political Obligations: A Response to Klosko , 42 Pol. Theory 218 (2014)CrossRefGoogle Scholar argues to a similar conclusion. Walton's main point does not directly address content-independence, instead focusing on a distinction on how reasons arise in relation to the will and actions of agents. My argument revolves around what substantive issues the notion of content-independence highlights and what those issues bring to the debate over political obligation and authority, so it is very different.

86. Christiano, Thomas, The Authority of Democracy , 12 J. Pol. Phil . 266 (2004)CrossRefGoogle Scholar; Kolodny, Niko, Rule over None I: What Justifies Democracy? , 42 Phil. & Pub. Aff . 195 (2014)CrossRefGoogle Scholar; and Kolodny, Niko, Rule over None II: Social Equality and the Justification of Democracy , 42 Phil. & Pub. Aff . 287 (2014)CrossRefGoogle Scholar; Viehoff, supra note 1. Again, these questions may very well be answerable. My contention is just that Klosko's classification incorrectly implies that they are “not overly difficult” to answer.

87. Klosko, Are Political Obligations, supra note 1, at 516. Note, as Gur, supra note 1, at 183, says, it may not be difficult to find content-independent formal reasons, but this is quite distinct from finding content-independent formal obligations, let alone a general content-independent formal obligation.

88. Klosko, Are Political Obligations, supra note 1, at 517.

89. Simmons, supra note 2, ch. 5.

90. Within the category of content-independent views, focusing on formal reasons rather than purely legal reasons seems eminently reasonable to me and also seems to characterize how the literature has generally evolved. Klosko's push to a more piecemeal approach is also plausible to me, but it is orthogonal to the issue of content-independence.