To many, the notion of toleration as a morally praiseworthy stance is deeply paradoxical. If an act or practice is morally wrong, should it not be condemned and punished? We never encounter claims that murder, rape, or theft should be tolerated, and rarely do we encounter claims that we should tolerate acts or practices manifesting racism or sexism. Claims for “toleration” are usually directed at acts or practices that the claimant believes are morally insignificant or even morally good. But then, what is there about those acts or practices that requires toleration? That is the so-called “paradox of toleration.”1
I believe there is space on the map of the moral terrain for toleration, and that the “paradox of toleration” is overstated. Nevertheless, I shall argue that this space is smaller than many “tolerant” liberals believe, and that in order to enlarge this space, some liberals attempt to ascend to a “neutral,” above-the-fray position that does not exist.
In what follows I shall draw my map of the moral terrain and locate the spaces for toleration on it. I shall then briefly comment on the arguments of Smith,2 Morgan,3 and Forst4 in terms of locating them on the moral map.
Demands for tolerance arise when people who interact hold conflicting views of right and wrong and good and bad. I shall call this situation a clash of first-order moral views. When this clash occurs, the call for toleration is addressed to partisans of one of the first-order moral views and, if heeded, results in various possible forms of forbearance—forbearance from punishment of those holding and acting upon the clashing first-order views, forbearance from other forms of tangible publicly imposed penalties, forbearance from public condemnation, and, in some cases perhaps, forbearance from private shunning and discrimination.
Toleration is possible when the values within each of the clashing first-order views lead to what I shall call “non-engagement.” What I mean by this is best conveyed through examples. Consider the following clashes of first-order moral viewpoints:
(1) Group B holds a religious belief that dictates the holding of sexual orgies among the faithful—all consenting adults—including group sex, both hetero- and homosexual. Group A finds Group B’s practice morally abhorrent and degraded. Nonetheless, Group A, which holds Millian moral beliefs regarding the proper use of state coercion, believes it would be wrong to punish or penalize Group B for its orgies. And Group B in turn believes that it would be wrong to compel participation in its orgies by non-consenting adults and children. Both Group A’s and Group B’s moral beliefs, therefore, although different, do not result in any legal or social conflict. Although as beliefs they are opposed, they fail to engage each other conflictually in action, a result that follows from the other beliefs the groups hold.5
(2) Group A is in control of the state but is super-pacifistic. That is, it believes that it is morally wrong to use coercion of any sort to enforce moral requirements. When members of Group B, who hold different moral views, do things that Group A regards as morally wrong, Group A forbears from punishing or penalizing them because of Group A’s own beliefs about the immorality of coercion. Again, Group A’s and Group B’s moral beliefs, though opposed as beliefs, do not result in conflict. As I am using the term, they do not “engage.”
(3) Group A believes homosexual acts are wrong and should be punished. Group B believes the opposite. Group A is in control of the state and could punish homosexuality. It believes, however, that if it were to do so, there is a good chance either that Group B would rebel, and that property loss, injuries, and deaths would occur, or that Group B would gain enough converts to its views to gain control of the state and impose its views on Group A. Group A believes, therefore, that it is morally prudent to forbear from punishing homosexuality. Put differently, Group A’s second-order moral views about what should be done when not all of its first-order moral views can be realized except at great cost or risk dictate that Group A forbear from imposing its particular first-order moral views about homosexuality. This morally prudential forbearance results in a modus vivendi in which opposed moral beliefs do not engage conflictually.
In these and similar examples, views that Group A holds to be morally wrong are “tolerated” by Group A because Group A’s own moral views dictate a stance of non-engagement.6 Toleration—avoidance of conflictual engagement—is a position within a partisan moral point of view. There is no stronger sense of toleration that is possible. Forbearance due to doubts about one’s own moral views is not tolerance. Nor is forbearance due to moral skepticism or beliefs in moral relativism. (In any event, neither skepticism nor relativism entails forbearance.)
When conflicting moral views do engage, toleration is not an option. If Group A believes abortion is the murder of an innocent human being, and Group B believes that it is on a par with the killing of microbes and should be within the scope of the right of a woman to control her body, then neither group can tolerate the acts of the others—that is, unless they are like the pacifists in (2) above, or decide that a modus vivendi is the best moral strategy, as in (3) above.
As with abortion, so too if Group A believes homosexuality is wrong and merits punishment, while Group B believes homosexuality lies within a general right to choose the forms and objects of intimacy and should be immune from state sanction. And so too if Group A and Group B agree that V is a virtue, but Group A believes there is a moral duty to teach V to all school children, while Group B believes there is a moral duty not to teach V to children in the public schools (because of the rights of dissenting parents). And, a fortiori, so too if Group A believes that certain religious truths should be propagated by the state, while Group B either rejects the religious truths or rejects that aspect of them that dictates that they be propagated by the state.
Certain views of political morality—particularly the “liberal” political moralities of Mill, Nozick, and Rawls7—either because they do not take stands on matters of the Good and religious duties, or because, if they do, they dictate forbearance from direct coercion—potentially leave open large domains of moral non-engagement, which I have claimed toleration must be. I say that those views potentially leave open large domains of moral non-engagement, and not that those views actually do so, because whether non-engagement is possible depends not only on whether one is a Millian, a Nozickian, and so forth, but also on the views of those whom one opposes. For whether Group A can tolerate Group B, or whether instead their views engage conflictually, is jointly determined by both groups’ views. Millians and Nozickians can tolerate religious groups like Group B in (1) above. But that is because Group B does not believe in coercing children or non-consenting adults. Millians cannot tolerate groups that would punish self-regarding conduct. Nozickians cannot tolerate groups that would redistribute property or harm some as a means to aid others. When groups’ views conflictually engage, toleration is impossible. Force or conversion are the only alternatives remaining. Converting non-Millians to Mill will surely eliminate engagement. But conversion—leading to the abandonment of the view one opposes—is surely not toleration of that view. And to repeat, Millian, Nozickian, and Rawlsian views result in the potential for toleration by taking no view on certain moral and religious matters other than that non-engagement by everyone is required vis-à-vis those views; and of necessity they cannot tolerate acts inconsistent with that requirement of non-engagement.
So the call for toleration speaks to someone from within a partisan view, either as a reminder of what is entailed by that view for those who hold it, or as an attempt to convert people to that view. It cannot be a view from nowhere, from above the partisan fray of contending moral and religious views, as liberal views are sometimes wont to characterize themselves. There is no such “neutral” position.8 If A believes abortion is murder, and B believes it is a right, the argument that A may be right but should nonetheless tolerate B is incoherent. It entails a rejection of A’s views, which means that it cannot claim to be “neutral” about those views. Liberal views are not held at some higher epistemic level than the conflicting partisan moral views on whose partisans the liberal urges tolerance. The liberal is just another partisan. That is not to say that the liberal is not correct in terms of what political morality requires. But if the liberal is correct, she is correct at the same epistemic level on which the non-liberal is then necessarily wrong.
Thus, if the non-liberal believes—perhaps because of statements in the Bible coupled with a belief in scriptural inerrancy—that homosexuality is wrong and should be punished, and the Millian, Rawlsian, or Nozickian liberal believes that homosexuality should not be punished, then the liberal must necessarily believe that the non-liberal has misinterpreted Scripture, or else that Scripture is wrong. Moreover, the liberal cannot tolerate the non-liberal’s punishing homosexuals because the non-liberal’s view that homosexuals must be punished engages conflictually the liberal’s view to the contrary. If the liberal tolerates homosexuality—because her views do not engage those of homosexuals—then she cannot tolerate the non-liberals who would punish homosexuals. She may try to convert them to her brand of liberalism. But if she fails, she then must oppose them, coercively if necessary.
If I were to characterize Steve Smith’s rich and provocative essay in terms of the moral map I have described, I would say that Smith’s principal claim is that liberalism should understand that it is a partisan first-order moral position, not a “neutral” second-order position above the fray of contending first-order moral positions.9 Smith endorses liberalism, at least generically, and endorses liberalism’s toleration of non-liberal views of the Good, most notably non-liberal religious views.
Smith’s focus is on the state and its forbearance from coercion of those with heterodox moral views, not on moral agents more generally.10 When the state confronts the various moral views persons hold within a pluralistic society, it will find that they fall into three categories.11 First, there are views that are consistent with the views expressed by the state’s policies. These views are within the prevailing orthodoxy. Second, there are views that are inconsistent with the prevailing orthodoxy but are nonetheless tolerable. And third, there are views that the prevailing orthodoxy cannot tolerate.
Smith is quick (and correct) to point out that the lines separating these three domains will be blurry and shifting.12 Some moral views will be absolutely intolerable. To a (non-pacificistic) liberal, a religious view demanding the killing of infidels or the sacrifice of virgins will be absolutely intolerable. Other views may be tolerable insofar as the criminal law is concerned, but intolerable insofar as whether, say, public school students should receive instruction or training in them. Think of prostitution, for example.13 Still other views may be deemed immoral (by the orthodoxy the state reflects) but fully tolerable. Thus, a small anti-liberal religious group may be fully tolerated in its teachings that liberalism should give way to theocracy, so long as it abjures coercion of its members and violence against others, and complies with laws controlling the education of its children.
Smith’s first and second categories—the orthodox and the tolerable—are, in terms of my moral map, views that do not conflictually engage with the dominant views. An orthodox view does not, of course, conflict with itself. And the tolerable are, by definition, those views that do not engage the orthodox views conflictually. What Smith emphasizes is that there may be degrees of tolerability.14 Some acts that from the state orthodoxy’s perspective are immoral may nonetheless be legal (because they do not engage the orthodoxy’s moral views if one includes moral views regarding the limits of the criminal law). They may not get taught or subsidized, and they may not escape official criticism, but they will not be punished.
As I said, Smith’s arguments for toleration are largely consistent with my moral mapping. He rejects the ultraliberal “neutral” position above the fray of partisan moral positions as incoherent, as do I.15 And Smith argues that a fullblooded moral case for toleration cannot be based on moral indifference, moral skepticism, or moral fallibility.16 He briefly discusses the modus vivendi version of toleration but declares it not terribly inspiring.17 And he points out that arguments for toleration based on the futility of coercing agreement are rather limited. The case for toleration, he concludes, resting as it does principally on practical limitations, is neither powerful nor stable.18
Liberalism—necessarily of the partisan type à la Mill, Rawls, Nozick, et al., given the impossibility of non-partisan, “neutral” liberalism—is distinguished by its deeming the Right to be prior to the Good and its concomitant elevation of moral autonomy to supremacy among values. Liberals need take no affirmative position on matters that fall within the domain of the Good. They need only reject as false any views of the Good, religious or secular, that would conflict with the priority of the Right. All “reasonable” views of the Good on this account recognize the right of autonomous persons to choose the Bad so long as the Bad is not the Wrong (violative of the right of autonomous persons to choose the Bad, or unjust regarding the distribution of wealth or liberty). So liberalism is tolerant of—does not engage conflictually—all “reasonable” views of the Good, even if those views are wrong. However, it is, and must be, intolerant of “unreasonable” views of the Good.19
Smith appears doubtful that liberalism’s sharp distinction between the Right and the Good can be maintained.20 For example, what provides the ultimate ground for the liberal’s elevation of autonomy in choosing (“reasonable”) conceptions of the Good? Would that not be some conception of the Good itself? If it is just the thin value of freedom to choose, how can it outweigh Truth, Beauty, Dignity, and so forth? And cosmopolitanism as a Good is dependent on the existence noncosmopolitan ways of life from which to be a cosmopolitan sampler.21 But if autonomy derives its value from a thicker conception of the Good, then the liberal’s case for tolerating choosing the Bad is considerably undermined, and the liberal seems driven to accept a more perfectionist liberalism, such as that of Raz,22 Galston,23 or Moore.24 And perfectionist liberalism may engage conflictually with more moral views and thus leave less room for toleration.
As I read Glyn Morgan’s original response to Smith,25 Morgan is a latter day Millian liberal who is attracted by the rhetoric of ultraliberalism, even if ultraliberalism—the non-partisan, neutral, above-the-fray liberalism—is an impossibility. What is not clear is whether Morgan is a Millian liberal out of first-order moral conviction, or whether instead he holds some less tolerant secular position that, for strategic reasons, is best furthered by adoption of a Millian framework. In other words, toleration that for Mill or Millian liberals like Joel Feinberg26 might be a matter of moral principle might be for Morgan a matter of the second-best.
There are strong indications that Morgan views Millian liberalism as a second-best, strategic choice. For example, he declares that the second part of his paper “seeks to show that enlightenment liberals would be tactically wise to opt for a politically liberal polity rather than the tolerant polity that Smith recommends.”27 This passage, and the part of his paper it refers to, imply that Morgan is an “enlightenment (Millian) liberal” as a matter of moral conviction, and that political liberalism is a strategy to secure enlightenment liberalism.
On the other hand, it is not at all obvious to me what the operative differences are between Millian, enlightenment liberalism and a “politically liberal polity” that Morgan urges enlightenment liberals to endorse as a tactical matter. If there are no differences, then either Morgan’s endorsement of Mill is itself tactical, leaving unanswered in service of what first-order moral views is Millian liberalism “tactically wise,” or Morgan endorses Millian liberalism as a matter of principle, leaving it unclear in what sense opting for a liberal political polity is “tactically wise.”
Rainer Forst acknowledges that the case for toleration “must not rest on a skeptical epistemology or on a ‘neutralist’ standpoint that—in a self-defeating way—doubts the truth of its own moral stance.”28 But having said that, Forst essentially takes it back. For he relegates the claims of religion to a separate epistemological realm, the realm of “faith,” and contrasts religious claims to secular claims, the latter of which fall within the epistemological realm of “reason” and “knowledge.”29 Disputes over religious matters cannot, therefore, be settled by reason, and disputed religious claims are, says Forst, “not good reasons for forcing others to act in a certain way (or believe certain things).”30 For Forst, “any use of force has to be mutually justifiable”; and religious justifications will fail that test.31
Forst not only rejects the “unity of epistemology” that I and others have argued for in the past, a view that would put religious claims on a par with other knowledge claims. He also rejects Smith’s contention that religious believers will be forced into schizophrenia by having to accept the Forst principle of mutual justification, which puts religious claims off the table as not based on reason, along with the truth of their own religious beliefs.32 But it is Forst, not Smith, who is wrong here, at least if religious believers’ own religious beliefs do not entail Forst’s principle. For if religious believers take their beliefs to be true—which is tautological—they will not perceive the distinction between faith and knowledge that Forst perceives, nor will they take their beliefs to be other than “reasonable.” If they have had a vision on the road to Damascus, and the vision instructed them to extirpate certain practices or views, they will surely not be moved by Forst’s principle. Of course, they may be deluded; and perhaps Forst can convince them that they are. But that would be a substantive, partisan attack on the truth of certain religious views—an attack that may be justified, but an attack nonetheless.
In the end, both Forst and Morgan may be engaged in an enterprise of conversion—with Forst, conversion to his principle of mutual justification and its denigration of religious “knowledge”; with Morgan, conversion to Millian liberalism. If either succeeds, the realm of potential conflictual engagement will be greatly narrowed because the space of moral agreement will be increased. What follows will not necessarily be toleration, especially in Forst’s case, for toleration entails moral disagreements, and the enterprise of moral conversion aims at eliminating moral disagreement.
For me, what is important is that we get clear about just why we should not conflictually engage moral views that we deem erroneous, if it is true that we should not do so. What kind of moral view is self-effacing to that extent, and why is such a view compelling? Moral prudence—the modus vivendi account of toleration—is comprehensible, if uninspiring. Arguments from the self-defeating nature of coercion are less convincing wholesale and equally uninspiring retail. Arguments that go to the moral wrongness of imposing what is morally correct promise a sturdy foundation for toleration, if only those arguments can be shown to be convincing.33
1. The “paradox of toleration” is oft noted. See, e.g., Alex Tuckness, Locke and the Legislative Point of View (Princeton: Princeton University Press, 2002), 17–19; Susan Mendus, Liberalism and the Limits of Toleration (London: Macmillan, 1989), 18–19.
2. Steven D. Smith, “Toleration and Liberal Commitments,” this volume.
3. Glyn Morgan, “How Impoverishing Is Liberalism?” this volume.
4. Rainer Forst, “Toleration and Truth,” this volume.
5. For a recent approach to toleration that is Millian in nature, see Hans Oberdiek, Tolerance: Between Forbearance and Acceptance (Lanham: Rowman and Littlefield Publishers, 2001).
6. There is a fourth form of toleration as non-engagement that is identified by Steven Wall. Sometimes Group A tolerates a practice of Group B that Group A regards as immoral, not out of respect for Group B’s rights to autonomous decision making, as in (1) above, and not because of super-pacifistic or strategic reasons, as in (2) and (3), but because the immoral practice is inextricably a part of a broader practice that has moral value. Steven Wall, “The Structure of Perfectionist Toleration,” in Steven Wall and George Klosko, eds., Perfectionism and Neutrality: Essays in Liberal Theory (Lanham: Rowman and Littlefield Publishers, 2003), 231 et seq.
7. See John Stuart Mill, On Liberty (New York: F. S. Crofts & Co., 1947); Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974); John Rawls, Political Liberalism (New York: Columbia University Press, 1993).
8. See Larry Alexander, “Liberalism, Religion, and the Unity of Epistemology,” San Diego Law Review 30 (1993), 763. For a somewhat Millian view of what should and should not be tolerated, but one with a more conservative cast than most contemporary Millian views, see John Kekes, The Illusions of Egalitarianism (Ithaca: Cornell University Press, 2003), 168–86. Kekes recognizes that the toleration he endorses is and must be the product of decidedly partisan views of Right and Wrong, Good and Bad, and the limits of justified coercion; it cannot be the product of some ultra-liberal neutrality. Whatever moral views one holds to be correct, “one cannot treat with equal concern [neutrally] their protectors and violators.” Ibid., 186.
9. Smith, “Toleration and Liberal Commitments,” 243–80.
10. Ibid., 245.
11. Ibid., 244.
12. Ibid., 246.
13. See Scott A. Anderson, “Prostitution and Sexual Autonomy: Making Sense of the Prohibition of Prostitution,” Ethics 112 (2002), 748.
14. Smith, “Toleration and Liberal Commitments,” 246.
15. The ultraliberal attempt at universal toleration of all positions reminds me of the joke about the two Jewish scholars arguing about a point of religious doctrine. They agree to take their dispute to the rabbi. The rabbi listens to the arguments of the first disputant, at the end of which he declares, “You’re right.” But the second one then presents his arguments for the contrary position, at the end of which the rabbi declares, “You’re right.” Perplexed, the first disputant objects: “Rabbi, you said that I was right, and then you said that he was right. But we can’t both be right.” To which the rabbi replies, “You’re right.”
16. Smith is less certain about the relation between fallibility and tolerance than he is about the lack of connection between tolerance and either skepticism or indifference. If one believes an act is morally wrong, but also is aware of the possibility that his moral views on the subject are mistaken, does he not have a reason to tolerate the act? I believe that the answer is complex and turns on the degree of confidence one has in one’s views and, as importantly, on what is at stake if one tolerates the act and if one does not tolerate it. Suppose John believes abortion is morally equivalent to murder but realizes that he may be wrong in that judgment. Should he tolerate Mary’s abortion, assuming he can coercively prevent it? If he tolerates it, he believes he is, more likely than not, allowing a murder. If he fails to tolerate it, he believes he is risking to some (lesser) extent a serious intrusion on autonomy, bodily integrity, family planning, etc. Awareness of one’s moral fallibility might not result in toleration in that case, nor, arguably, should it. In other cases—say, deciding whether to tolerate discrimination by a private club that one believes is morally wrong and not shielded from coercion by any moral right—the stakes are (somewhat) lower and one’s sense of moral fallibility may lead to a different result regarding toleration.
17. Smith, “Toleration and Liberal Commitments,” 251.
18. Ibid., 256.
19. See Jürgen Habermas, “Religious Tolerance—The Pacemaker for Cultural Rights,” Philosophy 79 (1) (2004), 5.
20. Smith, “Toleration and Liberal Commitments,” 260.
21. See Larry Alexander, “Illiberalism All the Way Down: Illiberal Groups and Two Conceptions of Liberalism,” The Journal of Contemporary Legal Issues 12 (2002), 625.
22. Joseph Raz, The Morality of Freedom (Oxford: Clarendon Press, 1986), 380–422.
23. William A. Galston, Liberal Pluralism (Cambridge: Cambridge University Press, 2002). For a critique of Galston’s attempt to derive liberalism from value pluralism, see Robert B. Talisse, “Can Value Pluralists Be Comprehensive Liberals? Galston’s Liberal Pluralism,” Contemporary Political Theory 1 (2004), 1.
24. Michael S. Moore, Placing Blame (Oxford: Clarendon Press, 1997), 763–95. See generally Larry Alexander, “The Legal Enforcement of Morality,” R. G. Frey and Christopher Heath Wellman, eds., A Companion to Applied Ethics (Oxford: Blackwell Publishing, 2003), 128–41.
25. Glyn Morgan, “Can Liberals Tolerate Religious Minorities?” Paper presented at the American Society for Political and Legal Philosophy, Atlanta, Georgia, January 2, 2004 (on file with editor).
26. See, e.g., Joel Feinberg, The Moral Limits of the Criminal Law, vol. I: Harm to Others (Oxford: Oxford University Press, 1984); Joel Feinberg, The Moral Limits of the Criminal Law, vol. IV: Harmless Wrongdoing (Oxford: Oxford University Press, 1988).
27. Morgan, “Can Liberals Tolerate Religious Minorities?” 7–8 (emphasis added).
28. Forst, “Toleration and Truth,” 282.
29. Ibid., 285–86.
30. Ibid., 286.
31. Ibid.
32. Ibid., 288.
33. I thank Steve Smith for commenting on drafts of this paper.