I owe thanks to Bill Edmundson, George Klosko, and Trenton Merricks, each of whom commented closely on earlier drafts of this article. I am also very much indebted to an anonymous referee at CJLJ for his or her extensive and fair-minded criticisms, all of which were extremely helpful to me in revising the piece for publication.

1. Smith, M.B.E., “Is There a Prima Facie Obligation to Obey the Law?” (1973) 82 Yale L. J. 950 CrossRefGoogle Scholar; John Simmons, A., Moral Principles and Political Obligations (Princeton, NJ: Princeton University Press, 1979)Google Scholar; Raz, Joseph, The Authority of Law (Oxford: Oxford University Press, 1979)Google Scholar; Green, Leslie, The Authority of the State (Oxford: Oxford University Press, 1990)Google Scholar. The term “philosophical anarchism” comes from Paul Wolff, Robert, In Defense of Anarchism (New York: Harper and Row, 1970) at 19 Google Scholar.

2. See, e.g., Raz, supra note 1 at 233.

3. I understand the moral obligation here as prima facie, or better, pro tanto. And I do not understand ‘obligation’ in the specific sense in which obligations arise only through voluntary undertaking, or through some particular form of social relationship.

4. Tyler, Tom R., Why People Obey the Law (New Haven, CT: Yale University Press, 1990).Google Scholar

5. Green, Leslie, “Who Believes in Political Obligation?” in Edmundson, William, ed., The Duty to Obey the Law (Lanham MD.: Rowman and Littlefield, 1999) 301 Google Scholar.

6. A reason to ф is content-independent if its status as a reason does not depend on the content of the act of ф-ing. A reason to ф is a protected reason if it is both a first-order reason to ф and a second-order reason not to act on reasons against ф-ing. In appealing to this conception of authority, Green follows Raz.

7. Green, supra note 1 at 86; also Green, supra note 5 at 309.

8. It is by no means clear that Raz and Green are right to hold that the law claims authority for itself in this strong sense. On their view, the law claims to have authority to issue any dictate that it wishes (though it may, of course, choose to limit its own authority through various devices). On an alternative, weaker view, for any dictate that the law issues, it claims the authority to issue that dictate; on an alternative, even weaker view, for any dictate that the law issues, it claims that there are decisive reasons to comply with that dictate. The question of what the law necessarily asserts vis-à-vis its own authority is far from settled. See, for views alternative to Green’s, Einar Himma, Kenneth, “Law’s Claim of Legitimate Authority” in Coleman, Jules, ed., Hart’s Postscript (Oxford: Oxford University Press, 2001)Google Scholar; Dworkin, Ronald, “Thirty Years On” (2002) 115 Harv. L. Rev. 1655 at 1661-67CrossRefGoogle Scholar (reviewing Coleman, Jules The Practice of Principle Oxford: Oxford U.P., 2001 Google Scholar); Soper, Philip, The Ethics of Deference (Cambridge: Cambridge University Press, 2002) at 79–87 CrossRefGoogle Scholar; and Murphy, Mark C., Natural Law in Jurisprudence and Politics (Cambridge: Cambridge University Press, 2006) at 52–56 CrossRefGoogle Scholar.

9. It seems clear that the philosophical anarchists would not take it to be a vindication of their view if the vast majority of (rather than all) citizens in a political community were bound by ties of political obligation. Simmons, for example, does not seem to think that it is a serious strike against gratitude and fair play accounts of political obligation that there would be occasional people who do not benefit adequately from government action to be bound to obedience; while noting that there are such people (Simmons, supra note 1 at 159-60), their existence plays no role in his attack on fair play (at 136-42) and gratitude (at 183-90) accounts of political obligation.

10. Smith and Simmons explicitly frame their investigations in terms of a prima facie obligation of obedience (Smith, supra note 1 at 951; Simmons, supra note 1 at 9). Raz frames his discussion in terms of obligation, which he understands not merely as a prima facie reason for action but a protected one (Raz, supra note 1 at 234-35). But he says that he is concerned to show not only that there is no obligation to obey the law in his strong sense, but also that there is no prima facie reason for obedience for citizens generally (Raz, supra note 1 at 235).

11. It seems clear from the writings of the philosophical anarchists that they would take an argument that showed that (say) 85% of the population is under a prima facie obligation to obey the law to be a threat to their positions. For, first, the philosophical anarchists’ arguments, if successful, establish that none, or a trivial percentage, of those living in contemporary political communities are under such an obligation. That is, if one is convinced by the arguments against accounts of political obligation offered in, for example, Simmons’s Moral Principles and Political Obligations while also believing that 85% of the population is under a moral obligation to obey the law, then one will be under rational pressure to give up one of those views: one must either give up the view that 85% of the population is bound by ties of political obligation or the view that Simmons’s attacks on political obligation in that book were successful. Second, it is clear from the philosophical anarchists’ interpretations of the practical upshot of their views that they want to deny that there is even a prima facie (even if not a protected) reason to obey, governing the behavior of most (even if not all) citizens. Consider Simmons’s summary of the situation if philosophical anarchism is true: because “most citizens do not have political obligations,” for them “obedience remains as much in need of justification as disobedience” (Simmons, supra note 1 at 200). To make this claim even approximately true requires not just the denial of the law’s authority in the wholesale way that the law presents its own authority; it requires the denial of the view that there is even prima facie reason for any significant segment of the population to obey the law.

12. One might claim that there is a disanalogy between the common moral belief that it is wrong to disobey the law and the common moral beliefs that it is wrong to lie, or to steal: in the former case, folks may have a hard time distinguishing between duty to obey the law and the myriad duties to do or refrain from doing what the law in reasonably just political communities would have one do or refrain from doing (take reasonable precautions against harming; refrain from killing, assaulting, thieving; etc.). There is no disanalogy, though. One could raise arguments against the view that it is commonly thought to be wrong to lie by claiming that folks can have a hard time distinguishing between the duty not to lie and the myriad duties to refrain from doing what characteristically results from or accompanies lying (causing distress; preventing people from realizing their worthwhile goals; etc.).

13. Does the fact that our belief in political obligation is belief in a limited obligation, one that holds only if the law is (say) not too unjust, present a problem for my argument? For, given that clarification, the most that the argument from the conscience principle an show is that most people are obligated to obey the law if (say) it is not too unjust. I will discuss this objection in more detail below (see section 3.), but for the moment it will suffice to note that if this counts as a devastating objection to the argument from the conscience principle, then it will count as a devastating objection to every important argument for political obligation ever offered. For all such arguments—arguments from consent, fair play, the natural duty of justice, gratitude, coordination, etc.—impose some limitations on the space within which the law can issue authoritative dictates.

14. Simmons, supra note 1 at 3.

15. Smith, supra note 1 at 975.

16. Raz, supra note 1 at 235.

17. That it is invariably wrong to act contrary to one’s conscience does not mean that one is necessarily acting rightly by acting in accordance with one’s conscience. The depraved conscience of the convinced Klansman does not make his persecution of the black or Jew or Catholic morally right. The requirement to act in accordance with conscience in such cases places such persons in moral dilemmas: so long as their consciences are mistaken, they cannot avoid acting wrongly. See Aquinas, Thomas, Summa Theologiae IaIIae, Q. 19, AA. 5-6Google Scholar; also Murphy, Mark C., “The Conscience Principle” (1997) 22 J. Phil. Research at 387–407 and 390-96CrossRefGoogle Scholar.

18. It is clear that one who rejects the conscience principle must accept that it can be perfectly right to act under the description doing what is morally wrong. For if I believe an act of ф-ing to be wrong, then when I ф I act under the description doing what I believe to be morally wrong; but at the point of action doing what I believe to be wrong is indistinguishable from doing what is morally wrong. So if one believes that it can be right to do what one believes to be morally wrong, one is committed to believing that it can be right to act under the description doing what is morally wrong.

19. This is only a test; I am not identifying the requirement to act according to conscience with some weaker requirement to provide a justification for one’s conduct.

20. For defenses of the conscience principle see May, Larry, “On Conscience” (1983) 20 Am. Phil. Q. 57–67 Google Scholar, and Murphy, supra note 17.

21. See Aquinas, Thomas, Quaestiones Disputatae de Veritate, QQ. 16–17, esp. Q. 17, AA. 1-4Google Scholar.

22. See Sidgwick, Henry, The Methods of Ethics, 7th ed. (Indianapolis, ID: Hackett, 1981) at 207–08 Google Scholar. Sidgwick may here be merely pronouncing on what he takes to be the moral sense of all humankind when he says that “no act can be absolutely right, whatever its external aspect and relations, which is believed by the agent to be wrong” (at 207). But he offers an argument in support of it (at 207-08) which he never rebuts, and his distinction between formal and material rightness allows him to affirm (though with reservations about the force of rational egoism) a substantive utilitarianism while leaving this intuition untouched.

23. It is true that Kant flatly denies at points that we have a duty to act in accordance with conscience ( Metaphysics of Morals, trans. by Gregor, Mary J. (Cambridge: Cambridge University Press, 1996) 6:401 at 160-01Google Scholar). But what he means by “conscience” here is something narrower than one’s overall judgments about what is to be done; he is picturing conscience, rather, as an internal tribunal by which one judges whether one has in fact lived up to one’s overall judgments about what is to be done (cf. Hill, Thomas E. Jr., “Conscience and Authority” in Respect, Pluralism, and Justice (Oxford: Oxford University Press, 2002) 260 at 270-71Google Scholar.) (This is why Kant says, seemingly strangely, that conscience cannot err: Metaphysics of Morals, 6:401 at 529.) But Kant is clear that to avoid moral guilt one must act in accordance with one’s practical judgments (Metaphysics of Morals, 6:401 at 530).

24. See, e.g., Hart, H.L.A., “Are There Any Natural Rights?” (1955) 64 Phil. Rev. 175 CrossRefGoogle Scholar; Rawls, John, “Legal Obligation and the Duty of Fair Play” in Law and Philosophy, ed. by Hook, Sidney (New York: New York University Press, 1964) 3 Google Scholar. Klosko, George’s formulation of the view, which does not rely on voluntary acceptance of benefits, does not fit this mold: see his The Principle of Fairness and Political Obligation (Savage, MD: Rowman and Littlefield, 1992)Google Scholar.

25. Locke, John, Second Treatise, in Two Treatises of Government, ed. by Laslett, Peter (Cambridge: Cambridge University Press, 1988) at §119, 121CrossRefGoogle Scholar.

26. For a discussion of nondetachability, see Dancy, Jonathan, Practical Reality (Oxford: Oxford University Press, 2000) at 70–76 Google Scholar.

27. It is true that the conscience principle is rarely explicitly relied upon in deliberation, but that is not due to its status as a nondetachable reason-statement (if it is such) but just because one characteristically acts on the basis of what one judges to be right, not on the fact that one believes it to be right. But there are cases in which one may act explicitly on the conscience principle. I may believe some action to be required, but recognize that it is not adequately motivating; I may, however, be able to summon up adequate motivation by noting that not only would my failure to act count as doing something independently wrong but would also count as violating my own integrity.

28. For some discussion of this principle with respect to moral beliefs, see Smith, Michael, The Moral Problem (Oxford: Blackwell, 1994) at 195–96 Google Scholar. One might think that we should add a fourth way that the folk might be unreasonable to continue to hold some belief—that they lack cogent arguments in support of it. But this seems to be far too strong—one would hardly want to place a condition on the reasonableness of folk belief that entails that the folk can reasonably believe in the external world or in other minds, only if they can provide arguments that will pass philosophical muster. And it hardly seems that we should place the reasonableness of folk belief in particular moral principles—that lying is wrong, or that doing charity is at least morally permissible—at the mercy of the availability of philosophical argument in this way. Note that the three ways of showing an ordinary belief to be unreasonable that I offer appeal to the ordinary person’s own lights, to showing that there is a tension in his or her own beliefs, or that the belief logically entails something that he or she would find unacceptable, or that the belief is, within his or her economy of belief, based on an argument that does not hold up under scrutiny.

29. See Simmons, supra note 1 at 16-23, 45-56.

30. See Raz, supra note 1 at 237-44; Smith, supra note 1 at 953-69; Green, supra note 1 at 89-121.

31. Simmons, supra note 1 at 194; Raz, supra note 1 at 245; Smith, supra note 1 at 972.

32. Raz, supra note 1 at 247-49.

33. Simmons, supra note 1 at 193-94.

34. Simmons, supra note 1 at 193.

35. A. Simmons, John, “The Anarchist Position: A Reply to Klosko and Senor” (1987) 16 Phil. & Pub. Affairs 269 Google Scholar.

36. Smith, supra note 1 at 975.

37. Simmons, supra note 1 at 195.

38. A. Simmons, John, “Philosophical Anarchism” in Sanders, John T. & Narveson, Jan, eds., For and Against the State (Lanham, MD: Rowman and Littlefield, 1996) 19 at 33-34Google ScholarPubMed.

39. Green, supra note 5 at 306.

40. Simmons, supra note 38 at 33.

41. To assume that the existence of a law requiring ф-ing, if it provides of itself a moral reason to ф in some case, must provide a moral reason to ф in every case, or even a moral reason to ф of the same strength in every case, is to commit the Additive Fallacy. See Kagan, Shelly, “The Additive Fallacy” (1988) 99 Ethics 5 CrossRefGoogle Scholar.

42. Raz, Joseph, The Morality of Freedom (Oxford: Oxford University Press, 1986) at 53–57 Google Scholar.