More than other facets of contemporary liberal theory, discussion of toleration tends to combine historical inquiry with philosophical analysis. The reason may be that toleration—both as concept and practice—is evidently historically prior to full-blown liberalism; this opens the door to a Whiggish inclination to see in seventeenth-century toleration the roots of the liberal thought that followed it by a century and half. A further reason to bring history into our analyses of toleration may lie in the availability of a convenient just-so story about why government adopted a theory articulated by philosophers: “sheer exhaustion brought about by incessant religious war.”1 If we are to see the rise of political liberalism as a process of gradual enlightenment in which peoples, countries, and eventually—one hopes—the world come to see the evident benefits of putting aside violence as a mechanism for resolving disputes, then there is something very appealing about identifying the conditions under which real, power-wielding institutions actually make progress toward the desired goal. For professional intellectuals, especially, the progressive story affords inspiration and hope: ideas really can make a difference! All you need are some wars of religion, and the politicians will come to see the inevitable attractions of reasoned discourse.
Let me hasten to insist that the turn to history in discussions of toleration is, on the whole, an excellent thing, not because it provides a how-to guide for getting power to listen to truth, but because it really can illuminate the relationship between political institutions and political theories. In fact, as I shall suggest, the different motivations of philosophers and politicians that emerge clearly in an account of the history of toleration can shed light on similar divergences in our own intellectual and political worlds, and can do much to chasten the aspiration to see a convergence between what morality requires and what the dynamics of power allow. By all means, let us explore the relationship between the Act of Toleration of 1689 and the Letter Concerning Toleration2 of the same year.
But a word of warning is in order, if it can be urged by an academic who has spent considerable time trying to negotiate between liberal ideas and political realities in Iraq: their ways are not our ways. It is not that politicians (statesmen and stateswomen, if you prefer) are deaf to the power of moral reasoning. To the contrary, many of them can deploy moral arguments with an astonishingly nuanced ear for how they will be heard and received. Rather, the conditions for the exercise of political power—namely that the politician can never completely forget that his authority rests on the ability to direct and control violence—relentlessly ensure that moral arguments acknowledge and respond to pragmatic ends.
Moral values can nonetheless play a crucial role in shaping what people understand as pragmatically possible. I am not saying that in the political realm morality is purely subordinated to power politics, or condemned to the status of empty rhetoric. But in the sphere of politics, morality also never trumps or transcends the basic reality that the state exists through its coercive force and the popular beliefs legitimating that coercion. If morality did have that transcendent capacity, we would no longer be dealing in politics. We would have entered the messianic realm of pure moral reasoning, in which each person conceives of her self-interest solely in terms of what is morally right for all. Only in such a world would we not need to fear the backsliding of the bad man who will get away with anything he can so long as he can avoid punishment that costs him more than what he stands to gain.
It is possible and maybe even conventional to distinguish a moral conception of toleration from a pragmatic or expedient one.3 Toleration by definition relates to beliefs or conduct which one believes to be wrong. According to this view, if I tolerate your beliefs out of the pragmatic or expedient hope that you will tolerate mine in return, or if a government tolerates your beliefs on the assumption that it will function more efficiently if it does, this would not count as moral toleration.4 To qualify as moral toleration, there must be some principled reason to tolerate beliefs or conduct considered wrong: a concern, say, for the victims of persecution considered in themselves,5 or for the wrongfulness of trying to coerce people to give up certain beliefs or conduct to which they are deeply committed.
Which came first? If we look to the early history of toleration, there is reason for uncertainty. Consider the Act of Toleration of 1689. It states that, “Forasmuch as some ease to scrupulous consciences in the exercise of religion may be an effectual means to unite their majesties’ protestant subjects in interest and affection,” various requirements of church attendance shall be lifted by the Act. The text of the preamble gives as the reason to ease “scrupulous consciences in the exercise of religion” not the value or importance of conscience, but the political goal of uniting political subjects in common interest.
Each carefully chosen word of the Act underscores the sense that no fundamental right to liberty of conscience is being created or acknowledged, but that the Act instead advances the interests of successful coercive government. Consciences are not “protected” here; they are simply “eased.” The consciences in question are not ordinary but “scrupulous,” which is to say particularly sensitive, and therefore not so much deserving of being eased as being granted easement as a boon. The gracious conferral of toleration is restricted to “protestant subjects” and by no means expanded to Catholics or to dissenters unwilling to swear to a test oath also prescribed by statute. What is more, conscience is eased “in the exercise of religion”—a phrase which made its way into the First Amendment to the U.S. Constitution a century later—and not generally, which is to say that a claim of conscience would offer no protection against coercion in any non-religious realm.6
This entire formulation is a classical statement of the expedient or pragmatic conception of toleration. The justification for toleration offered by the Act is a reason of state in the classic, Florentine and French7 sense: scrupulous consciences will be eased for the pragmatic reason that it will enable their majesties’ subjects to get along, and thereby benefit the stability of the state and the viability of William and Mary’s revolutionary rule. The state deliberately refuses to acknowledge the liberty or the right of the individual or indeed any moral reason at all for tolerating, and opts instead to tolerate only in order serve its own interests in avoiding civil discord.
It is not that more expansive conceptions of the protection of conscience were unavailable to the drafters of the Act. Fifty years before, the Westminster Confession had declared that:
God alone is Lord of the conscience, and hath left it free from the doctrines and commandments of men which are in any thing contrary to his Word, or beside it in matters of faith on worship. So that to believe such doctrines, or to obey such commandments out of conscience, is to betray true liberty of conscience; and the requiring an implicit faith, and an absolute and blind obedience, is to destroy liberty of conscience, and reason also.8
That earlier formulation, to be sure, still left room for the state to regulate religious worship and teaching; but it proceeded from the assumption that conscience had its own inherent value, and required protection on that ground, not to serve the interests of government. Without quite calling for toleration, the Confession offered an argument against “requiring” faith and obedience: to do so would be to destroy the liberty of conscience that the individual holds by the grace of God. That argument may be recognized as a moral one insofar as it hints that the liberty of conscience is a God-given right that cannot be violated without harming the person who holds it.
To see clearly how the Act of Toleration stands for the pragmatic conception of toleration, not the moral one, contrast its compressed reasoning to that of Locke’s roughly contemporaneous Letter Concerning Toleration. For Locke, the primacy of individual conscience derives, as it did for the Puritans who crafted the Westminster Confession, from the moral and religious truth that God has left the conscience free in order for the individual to form beliefs efficacious for salvation.
It is true that Locke argues that it would be irrational for the government to try to coerce people against conscience—a famous argument that Jeremy Waldron has noted might not itself be a moral one, insofar as it focuses on the irrationality of coercion, not its consequences.9 Locke also argues, in the Letter, that the commonwealth exists to serve civil ends, not religious ones.10 But both of these claims depend, I think, on the logically prior argument that it would be illogical for the individual to alienate to the state his right to form religious beliefs. The reason such alienation would make no sense, for Locke, is precisely that religious belief reached by means other than free, faithful choice has no salvific value: “No way whatsoever that I shall walk in against the dictates of my conscience, will ever bring me to the mansions of the blessed. … In vain, therefore, do princes compel their subjects to come into their church-communion, under pretence of saving their souls.”11 In response to Jonas Proast’s argument that state coercion might in principle lead to the formation of true beliefs, Locke could offer the straightforward (and indeed traditional) response that acting against one’s own conscience was inherently sinful, regardless of whether one’s conscience was in error.12
It follows, then, that Locke’s argument in favor of toleration is a moral-religious one with some pragmatic offshoots. What is ultimately wrong with coercion against conscience is that it forces the coerced individual to act against his conscience and therefore in a way that would bring him the harm of damnation and certainly could not bring him the good of salvation. Thus, the core argument of the Letter, as Waldron has recently acknowledged, “does have to rest on its distinctively Christian foundations.”13 Those foundations are not only religious in the sense of deriving from beliefs about the nature of salvation, but definitively moral in their concern for the well-being of the soul of the one coerced. Locke’s rationale for toleration, then, differs markedly from the rationale offered by the preamble to the Act of Toleration: the moral argument imposes an obligation on the state to respect individual conscience, while the pragmatic argument suggests that tolerating conscience serves the expedient interests of the state. Juxtaposing the legislators’ version of toleration with the philosopher’s strongly suggests, at a historical level, that the state is promoting its own pragmatic interests in peaceful coexistence over and against a moral theory which the statesmen of the Glorious Revolution were loath to adopt.
Can we imagine an actually existing state adopting a policy of toleration on moral grounds? One can perhaps imagine a Christian state tolerating on Lockean principles,14 or an Islamic state tolerating Jews and Christians on the ground that the Qur’an prohibits coercion in religion—a variant on Locke’s concern for the well-being of the souls of the tolerated person. Rawls, however, proposes a model of moral toleration not grounded in religion. In it, “toleration is not derived from practical necessities or reasons of state,” but from the principle of the equal liberty of conscience.15 Under this approach, the state adopts a constitutional scheme “that guarantees an equal liberty of conscience regulated solely by forms of argument generally accepted”16—roughly, what the later Rawls calls public reason.
According to Rawls’s account in Political Liberalism, the overlapping consensus that provides among other things the grounds for toleration “is quite different from a modus vivendi,” both because the conception of justice that is the subject of the overlapping consensus is moral, and because (more importantly for our purposes) “it is affirmed on moral grounds.”17 The moral basis for toleration on which an overlapping consensus may emerge is that every person has an equal right to form his own beliefs; and the state adopts the policy of toleration in recognition of this principle. The relation between tolerance and public reason is that non-public, comprehensive doctrines may give strong reasons for intolerance—Rawls mentions both Catholics and Protestants of the sixteenth century as having such reasons18—but public reason does not provide such reasons, according to Rawls, except in the limited case where toleration of intolerance would lead to the downfall of the tolerant state.19
When considered in the light of what we know about the real world, Rawls’s attractive vision of a constrained politics runs against our intuitive (and arguably empirical) sense of politics as the realm in which interest groups seek to express their preferences both in shaping the way institutions of government deploy power (constitutional politics) and in distributing the resources of the state (ordinary politics). At the same time, of course, we can also observe in many instances a type of political argument that appeals to impartial interests and uses “forms of argument generally accepted”—whether in the arenas of civil rights, environmentalism, or, increasingly, international affairs. The aspiration to a politics that would replace an appeal to the common interest (or in the fashionable jargon, public choice with public interest) can thus be understood as the wish to infuse the sordid politics we know with the elevated politics we occasionally glimpse.
One standard reaction to this aspirational vision of commoninterest politics is simply that it is unrealistic—that politics in this world can never achieve the goal of impartiality, because nothing short of a transformation in human nature can ensure that citizens will not band together to seek their partial self-interest and will not make arguments for it in terms that are less than generally acceptable or accepted.20 Without active state regulation of political culture, what is to stop interest groups from making arguments that rely not on impartiality but naked particularism and on assumptions not generally acceptable? In a moment, I intend to make just such a realist (or maybe cynical) response, although I hope to add to it at least a brief distinguishing account of why moral aspirations in politics nevertheless can bring about practical consequences even in a universe of living, breathing, self-interested, power-seeking humans. Before I do, however, I first want to pause and consider a possible moral objection to the argument that the right way to instantiate moral toleration lies through a political sphere characterized by an overlapping consensus that nevertheless is not a comprehensive conception of the good.
First, consider the position of persons who believe as a matter of their comprehensively held doctrine that toleration is only appropriate on expedient, modus vivendi grounds. Rawls says that “so long as such views … are very much in the minority, and are likely to remain so, they do not significantly affect the moral quality of public life and the basis of social concord.”21 Of course, if they were to become something more than a small minority, “the moral quality of political life will also change in ways that are obvious and require no comment.”22
What is the tolerant state to do with the peaceful efforts such minorities will make to use the apparatus of the state to convince the rest of the citizens of the state to adopt their comprehensive conceptions of the good? I am not now speaking of the intolerant, whose claims may plausibly be dealt with—as Rawls deals with them—by noting both the practical necessity of limitation and also their weakened claim to toleration.23 I am speaking, rather, of those who tolerate and favor toleration, but who as a matter of principle can do so only on non-moral grounds. Such people do not participate in the overlapping consensus insofar as they cannot adopt the state’s policy of toleration on moral grounds without compromising their comprehensive beliefs.
In the real world of democratic politics, such minorities, like everyone else, will vie to use the state’s institutions, including its schools, to teach children their version of the good life. If the state were to prohibit them from trying to achieve this goal, then the state’s moral version of toleration would in effect function as a limit on the conscientious political participation and action of people who are themselves committed to toleration, albeit on expedient grounds. This is a troubling result, for reasons that, as Rawls would say, “are obvious and require no comment.”
Alternatively, the state could simply allow such minorities to try to use the state’s institutions to promote comprehensive views that include the view that one may tolerate only for expedient reasons. This approach, however, opens the door to the abandonment of the moral ground for toleration. The problem is especially acute in democracies, where the plurality of interests will drive each sub-group to seek to benefit itself proactively, lest other groups take advantage. In the modern state, with its plethora of incommensurable interests, taking the morally tolerant position in politics may well mean losing out to alternative visions. Moral toleration, it turns out, is far easier to accomplish when an absolute ruler exclusively occupies the political field and in that capacity takes moral account of the interests of all his subjects. Although Rawls in Political Liberalism tells a genealogical story in which expedient toleration gives rise to constitutional consensus and then genuine overlapping consensus, it is at least possible to tell a story that eventually runs in the other direction—from moral toleration to the growth of minorities who eventually capture the state and return it to toleration on expedient grounds alone.
Recognizing that the democratic state with its plural values and identities is a historically contingent mode of organizing political power leads the way to describing the serious practical problems with grounding toleration in a moral theory. The most basic problem is that, in practical terms, we have little reason to expect that either individuals or groups who would like to pursue their own interests will refrain from seeking to advance those interests in constitutional or ordinary politics.
If we do not want to regulate the political realm to restrict forms of discourse that embrace toleration on expedient grounds only, how we are to expect moral toleration to emerge? Leaders and their states may condemn the worst excesses of intolerance. But this is not enough to make the goal of moral toleration seem more than utopian. Politicians’ incentives are structured by the way they get their bread. In an established democracy, that means they will need to get reelected; in a more fluid, non-state situation (for example, a transitional occupation), politicians make their living by asserting that they can command constituencies capable of shaping the affairs of the country. In either case—the latter more obviously than the former, of course—politicians deploy a certain type of power, power derived from their capacity to shape and direct violence. The legislator passes laws that crystallize the coercive power of the state into a command potentially backed by violence, while the communal or religious leader may have to put bodies on the streets to create a coercive threat to other groups or institutions. But the business is the same for each. In both cases, actual violence is a last resort and leaders matter precisely because they coordinate popular action in a way that reduces coordination costs and legitimates the exercise of power.
From political leaders’ function one can deduce the mechanism for getting them to adhere to principles of toleration: in a word, self-interest. In general, politicians will negotiate solutions that seem to them mutually advantageous, just like any other relatively rational actors.24 This is the historical explanation for the emergence of toleration in the wake of the wars of religion: not exhaustion or enlightened proto-liberal revelation, but rather a complexly negotiated settlement in a situation where statesmen calculated that their states’ interests (and their own) would be served better by coexistence than coercion. The Act of Toleration has more in common with the Peace of Westphalia than with the philosophy of John Locke. Iraqis in their constitutional negotiations settled on a principle of toleration to the extent that Shi‘i, Kurdish, and Sunni Arab leaders concluded that their particular interests would be better served by avoiding the political turmoil that would follow from intolerance. (As it turned out, the best pragmatic intentions of the politicians have not been enough in the Iraqi case.) To hear Rawls tell it, expedient reasons are replaced by moral ones because people see the success of the expedient model, begin to trust their fellows, develop a constitutional consensus,25 and then, most important to the transformation, develop political conceptions to appeal to others. But why, exactly, should the institutionalized self-interest of the last stage give rise to moral reasons for toleration? The reason there is something dissatisfying about a morally-based account of toleration that focuses on the creation of political sphere of overlapping consensus is that such an account slights the role of self-interest in making toleration into a practice of continuing utility at every stage in its existence.
From the self-interest story, I propose, though, it does not follow that there is no room for a moral account of the value of toleration in shaping political reality. It is just that in such a story, morality will not provide a transcendent basis for toleration within the self-interested activity of politics. Politicians are people like any others (only more so, one is tempted to add). That means they, and the people whom they govern or represent, form beliefs that guide their conceptions of their own self-interest. Here morality, whether individual or collective, enters the picture, shaping the norms that politicians promote, and occasionally even the means they consider legitimate for promoting them. Martin Luther King, Jr.’s political morality need not be seen solely as a self-interested attempt to advance the interests of African-Americans; it can be seen as part of a sincerely motivated moral project of advancing universal interests. The same can even be said of Lyndon Johnson’s advocacy of the Civil Rights Act of 1968. No matter how well-deserved Johnson’s reputation for political cynicism, there is reason to think that the civil rights movement changed his moral evaluation of segregation. What cannot plausibly be asserted, however, is that the political strategies of King in Selma or Birmingham or of Johnson in manipulating Congress depended solely or even primarily on the deployment of impartial public reasons. To get what they wanted, these statesmen played politics. What they did was no different than what corporations do when they want environmental laws weakened, or for that matter what environmentalists do when they want the same laws strengthened. They play the game of politics: they appeal to self-interest and they manipulate the levers of power.
What is then so appealing about a moral conception of toleration, in which I tolerate you or your views on principle, rather than because I hope you will tolerate me, as well? Rawls seems to suggest that the principle seems more likely to be durable than the contingent fact that I seek to be tolerated myself: unlike overlapping consensus, he says, the stability of a modus vivendi toleration “does depend on happenstance and a balance of relative forces.” This view, however, if intended normatively, would seem to place the moral argument in the service of (expedient) self-interest in stability. Furthermore, self-interest would seem to be at least as durable as abstract principle; so even if we were after the theory of toleration most likely to protect us, it seems entirely possible that we would prefer the reassurance of mutual self-interest to the shaky guarantees of moral commitment.
To put the question the other way around, what would be dissatisfying about concluding that, in the end, self-interest and expediency are the only basis we have for being tolerant? The answer seems to be twofold. First, there is the problem of cheating. If we adopt a general scheme of toleration because it serves our interests, circumstances will arise in which we think we can get away with an occasional defection from the principle without endangering our own right to be tolerated. (Perhaps our exception will not be much noticed, or the party bearing the brunt of our intolerance will lack the capacity to get revenge.) This consequence of a game theoretic approach could perhaps be obviated if we were guided by principled moral toleration, which ought to have few exceptions, or none.
From this perspective, a moral theory of toleration might turn out to be a useful supplement to an expediency theory: suspenders and a belt to make sure that our arrangements hold up. Pragmatists might want not only to be tolerant, but to promulgate a moral theory in support of their toleration, in order to assure that toleration would exist as much of the time as possible. The idea might be that convincing people to internalize a belief is far more efficient than enforcing a social practice like toleration against self-interested potential back-sliders. One vision of the relation between moral and pragmatic toleration, then, would be that the moral approach does indeed ultimately serve pragmatic ends.
Alternatively, expedient toleration might take a shape that would turn out to be morally unjustifiable under some set of conditions; and if we want our political arrangements to satisfy the demands of morality, we might discover that we had to reimagine toleration in moral terms in order to justify it. According to this approach, it is not enough for toleration to be useful to everybody—it must actually be morally just. If we were to assume that toleration was morally neutral, perhaps we could allow it on that basis. But it seems relatively unlikely that a practice which permits the maintenance of morally reprehensible beliefs and actions would turn out to be morally neutral. On this view, we want a moral theory of toleration precisely because we suspect, as a matter of initial intuition, that toleration is immoral.
To follow this possibility where it goes, one might be tempted to argue that toleration could be found conclusively moral if we were to follow some sort of a veil-of-ignorance procedure and conclude that reasonable persons would agree to it. This raises a puzzle: would our toleration be of the moral variety if we adopted it because reasonable people would agree to it behind a veil of ignorance—but on purely self-interested grounds? Or would it be expedient merely?26
I will not try to solve this puzzle here; rather I would like only to suggest that the categories of moral and pragmatic toleration can be made to collapse into one another in ways that call into question both the practical and theoretical consequences of the distinction. The interplay of pragmatic politics with moral theory has a long enough history to make one doubt that either side of the equation will soon disappear. The social practices of politics run on non-moral fuel; but despite the fond hopes of rational choice theorists, people keep doggedly asserting and believing in moral arguments. Occasionally they even change their behavior as the apparent result of some change in moral belief. Why, then, should it matter so much whether we tolerate despite ourselves or out of self-love? If toleration did not serve self-interest, it would probably not be sustained for long. If, on the other hand, toleration served only self-interest, it seems unlikely that so many people would care so passionately about it.
1. Ingrid Creppell, “Toleration, Politics, and the Role of Mutuality,” this volume at 324.
2. John Locke, “A Letter Concerning Toleration,” reprinted in John Locke: A Letter Concerning Toleration in Focus, John Horton and Susan Mendus eds. (London: Routledge, 1991) [hereinafter Letter].
3. Ingrid Creppell, Toleration and Identity: Foundations in Early Modern Thought (New York: Routledge, 2003). See also “Toleration, Politics, and the Role of Mutuality,” this volume at 315–59. References below are to this essay, not to the book.
4. Ibid., 325.
5. See Waldron, “Locke, Toleration, and Persecution,” Letter, 113.
6. Conscientious objectors, for example, would have no claim to be exempted from the nominally secular activity of making war by this language, which protects only religious activities.
7. Ragione di stato, raison d’état.
8. The Confession of Faith, Together with the Larger and Lesser Catechismes, composed by the Reverend Assembly of Divines, Sitting at Westminster, Presented to Both Houses of Parliament (London: Co. of Stationers, 1658). The Westminster Confession was collectively authored by the Westminster Assembly (1643–52).
9. Waldron, “Toleration and the Rationality of Persecution,” Letter, 120.
10. Locke notes, on scriptural grounds, that although the Hebrew Bible did establish a religious state, “there is absolutely no such thing, under the Gospel, as a Christian commonwealth.” Letter, 30. The difference is significant for the history of liberty of conscience, which derives in part from the doctrine of Christian liberty according to which Christ liberated the conscience from the duty to obey the laws of the Old Testament. See Noah Feldman, “Intellectual Origins of the Establishment Clause,” 77 N.Y.U. L. Rev. 346 (2002) 354–72.
11. See Letter, 32.
12. See Richard Vernon, The Career of Toleration: John Locke, Jonas Proast, and After (Montreal: McGill-Queen’s University Press, 1997), 17–34 (presenting Locke’s argument from belief and Proast’s critique of it); Jeremy Waldron, “Locke: Toleration and the Rationality of Persecution,” Letter, 98, 115–19.
13. Jeremy Waldron, God, Locke, and Equality (Cambridge: Cambridge University Press, 2002), 210.
14. Compare John Rawls, Political Liberalism (New York: Columbia University Press, 2005), 145, n.12.
15. Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1973), 214.
16. Ibid., 215.
17. Rawls, Political Liberalism, 147.
18. Ibid., 148.
19. Rawls, A Theory of Justice, 215–21.
20. Rawls himself is concerned to refute the view that overlapping consensus is “utopian.” Political Liberalism, 158–64.
21. Ibid., 148.
22. Ibid., 149.
23. Cf. Rawls, A Theory of Justice, 218–20.
24. There is a growing rational choice literature on this subject, much of it painfully full of truisms about people agreeing to things just when they decide to agree to them. But the truism is still true.
25. Rawls, Political liberalism, 163.
26. Rawls says that the denizens of the original position “cannot take chances with their liberty by permitting the dominant religious or moral doctrine to persecute or to suppress others if it wishes.” A Theory of Justice, 207.