“Congress shall make no law respecting an establishment of religion.” In 1789, some recent thought has asserted, at least this first of the two religion clauses in the First Amendment to the US Constitution was solely jurisdictional because nonestablishment sought solely to locate in the several states power over government's proper relation to religion. In this way, the clause intended to confirm that enumerated powers of Congress granted by the US Constitution did not include that power. Hence, so-called incorporation of this clause against the states, as the Fourteenth Amendment to the US Constitution has been interpreted, becomes especially problematic. How could power over government in relation to religion be extended to the states if that power was already reserved solely to them? I first summarize two presentations of the solely jurisdictional reading—those of Steven D. Smith and Akhil Reed AmarFootnote 1—that define, and thereby limit, my understanding of it, after which I offer a critique of that view and then outline an alternative. On this alternative, religious freedom is essential to a democracy, that is, to governance in which “we the people” are the final ruling power and determine the activities of government through a full and free discourse, inclusive of religious or comprehensive convictions.
There is a further limitation on this discussion: these pages consider the solely jurisdictional reading of nonestablishment philosophically. I do not offer (or intend) an interpretation of US constitutional law. In other words, the focus of attention is on what Smith calls a principle or theory of religious freedom. Although, when apparently helpful, I make reference to US politics, I have little relevant historical knowledge and do not argue historically for the conclusions I commend. Because both Smith and Amar speak of constitutional law and thus include historical arguments, my critique insofar fails fully to consider their proposals. Still, I assume that US politics and my article have this in common: both claim to be democratic—where democracy means, whatever else it means, popular sovereignty or consent of the governed in the “active” sense, and “active consent” means that “we the people” are the final ruling power.Footnote 2 Hence, neither US politics nor this article can be convincing unless democracy with religious freedom is possible or makes sense.
In what follows, I propose its coherence; that is, I propose, pace Smith, a principle or theory of religious freedom. But I take a solely jurisdictional reading thereof to be incoherent as a candidate for that principle or theory. The reading of Smith and Amar requires that comprehensive convictions cannot be objects of public discourse or discussion and debate among “we the people”—and, given that account of such convictions, the solely jurisdictional view is, I argue, internally inconsistent. To be sure, Smith defends that reading because, for him, there can be no principle or theory of religious freedom; that is, he argues, in his own way, for an incoherence. But his historical argument for what the original nonestablishment clause was designed to achieve is in service to this conclusion only because he adds a theoretical argument against any such theory—and it remains to ask whether one can hold the solely jurisdictional reading if one also seeks to affirm the principle or theory of religious freedom Smith denies.
As mentioned, I argue that a principle of religious freedom is essential to democracy as the form of government in which “we the people” are the final ruling power; that is, I argue that popular sovereignty entails a principle of religious freedom. Thus, any theorist who assumes the internal consistency of what the solely jurisdictional view attributes to the original Congress and ratifiers of nonestablishment insofar cancels her or his own interpretation of the First Amendment. If popular sovereignty is possible, the current public and, in the US system, the Supreme Court require some other understanding of the first religion clause in that Amendment—because neither the public nor the Court can apply an incoherent theory. In a word, the burden of this article is to show the grounds on which politics with religious freedom makes sense.
The Solely Jurisdictional Reading
In Foreordained Failure, Smith argues historically that the original religion clauses “were purely jurisdictional in nature; they did not adopt any substantive right or principle of religious freedom” (17). Those clauses “were an exercise in federalism” (18), which “kept the national government out of religion not because governmental support for religion was generally regarded as improper . . . but rather because the religion question was within the jurisdiction of the states” (21). In this respect, the First Amendment represented “essential federalism,” that is, the law has no substantive meaning independent of its federalism” (24)Footnote 3—and thus did not answer the “first-order” or “substantive” question about “the proper relationship between government and religion” but only the “second-order” question about “[w]hich level of government, state or national, should be responsible for addressing the first-order question” (19). Smith then explicitly draws the following conclusion: “it seems nonsensical or incoherent to suggest that a provision representing ‘essential federalism’ has a substantive meaning independent of its federalism . . . that can be ‘extended’ to the states” (24).
Smith commends this historical reading in part because “[t]he founders did not answer the religion question because they could not have done so” (26). If the nonestablishment clause required an answer to the first-order question, that clause would not be accepted because differing positions on religious establishment were widespread in the several states. The “traditional position . . . [that had been] almost universally held in Western societies for centuries,” namely, that a stable social and political order requires (at least) a largely common religion, and such commonality requires governmental inculcation, “enjoyed widespread support in this country” (19). On the other hand, “the ‘voluntarist’ position,” which “agreed that a [common] religious foundation was vital to the political and social order” (20) but opposed coercive governmental inculcation because religious commitment should be voluntary, was also widely held (most voluntarists apparently believed that governmental coercion was politically unnecessary because voluntary commitment to Jesus as the Christ would be largely common).Footnote 4 “Whether any significant body of opinion” rejected the requirement of a common religion—a third position, at least logically speaking—is, Smith says, “less clear” (20).
In any event, Smith joins his historical argument with a theoretical one, which runs as follows: “any account of religious freedom will necessarily depend on—and hence will stand or fall along with—more basic background beliefs concerning matters of religion and theology, the proper role of government, and ‘human nature,’” including “what is valuable among human beings” (63, 68). Accordingly, every theory of religious freedom, Smith holds, takes “a preferred religious or secular position” and “might be called a theory of” governmental toleration, “which implies a preferred or orthodox position that ‘puts up with’ other, less-favored views” (73).
If Smith embraces a “purely jurisdictional” reading of the original First Amendment's nonestablishment clause, much the same can be said of Amar in The Bill of Rights: The “mandate that Congress shall make no law ‘respecting an establishment of religion’ . . . prohibited the national legislature from interfering with churches . . . established by state and local governments” (32).Footnote 5 Hence: “The original establishment clause, on a close reading, is not antiestablishment but pro-states’ rights; it is agnostic on the substantive issue of establishment versus nonestablishment and simply calls for the issue to be decided locally” (34). The clause was, as Amar also says, “a pure federalism provision” (246)—where “pure federalism,” we may assume, is synonymous with Smith's “purely jurisdictional.” Further, he also draws the relevant conclusion with respect to incorporation: “The special pinprick of the point is this: the nature of the states’ establishment-clause right against the federal government makes it quite awkward to mechanically ‘incorporate’ the clause against the states . . . . [I]ncorporation of the establishment clause has precisely this kind of paradoxical effect; to apply the clause against a state government is precisely to eliminate the right to choose whether to establish a religion—a right clearly confirmed by the establishment clause itself” (33–34).Footnote 6
The Solely Jurisdictional Reading: A Critique
Perhaps many in the congressional and ratifying discussions understood the nonestablishment clause with Smith and Amar. But however widely that clause was at the time taken to be “purely jurisdictional” or a “pure federalism provision,” it nonetheless stipulates how the national government relates to religion. On this national account, a political body and its government can be stable without a common religion and, indeed, can provide for indeterminate religious diversity. To be sure, the original First Amendment proscribed to Congress any law “respecting an establishment of religion” (emphasis added) and, therefore, did indeed exclude such laws from the enumerated powers of the national government. But this very exclusion has in principle the following consequence for the new national government: it cannot depend on a common religion. With respect to the national society and its political order, in other words, the nonestablishment clause stipulates the third position Smith briefly reviews, that is, it rejects the requirement of a common religion. Accordingly, we have every right to ask: Was this stipulation recognized, and however that question is answered, does it makes sense constitutionally to affirm religious freedom?
Transparently, the first part of this compound question is historical. Conceding my wants as an historian, I nonetheless doubt that James Madison—author of the “Memorial and Remonstrance against Religious Assessments” in Virginia; responsible for guiding Jefferson's Bill for Establishing Religious Freedom through the Virginia legislature in 1777; and pivotal advocate for both a national government with power in 1787 and, in 1789, what would subsequently become the First Amendment—failed to recognize how the central government he so thoroughly thought necessary was properly related to religion by that amendment, even if he did not make a point of it.Footnote 7
Be that as it may, whether it makes sense constitutionally to affirm religious freedom asks a theoretical question about the national body politic, the answer to which depends on how religion is understood, that is, how distinctively religious activities are different from other kinds of human activity. On my accounting, interpreting the nonestablishment clause requires politically some definition of “religion,” implicit or explicit, because otherwise one cannot know what Congress is prohibited from establishing or what freedom of religion protects.Footnote 8 As far as my reading extends, virtually all academic proposals for understanding national politics in the United States define religions to include comprehensive human convictions, that is, convictions about human activities as such and thus about the ultimate terms of evaluation for all human activities—in distinction from more specific terms in which differing human activities are evaluated. This is so in the later work of John Rawls, who includes religions among what he calls “comprehensive doctrines.” Such doctrines can be “partially” or “fully” comprehensive, and in the latter case a doctrine applies “to all subjects universally” and “to our life as a whole” (I assume that religions, at least typically, include fully comprehensive doctrines).Footnote 9
For Ronald Dworkin also, religions include comprehensive convictions: “The value part of a conventional . . . religion offers a variety of convictions about how people should live and what they should value.” Thus: “Any judgment about meaning in human life . . . relies . . . finally on more fundamental value judgments.”Footnote 10 In addition, Jürgen Habermas, for whom a religion is a “totalizing . . . form of faith,” which a person “taps into performatively to nurture her whole life,” agrees insofar with both Rawls and Dworkin.Footnote 11 With all three, Smith, too, holds that religions are comprehensive convictions: on his theoretical argument, as already mentioned, “any account of religious freedom will necessarily depend on—and hence will stand or fall along with—more basic background beliefs concerning matters of religion and theology, the proper role of government, and ‘human nature,’” that is, “what is valuable among human beings” (63, 68).Footnote 12 Smith's designation of beliefs about human nature (referring, one can assume, to humans as such) and what is valuable among human beings entail that such background beliefs are both implied by and imply the account of religious freedom in question because they are comprehensive and thus include ultimate terms of evaluation.
The consensus among Rawls, Dworkin, Habermas, and Smith is, on my reasoning, correct: a religion, whatever else it may be, includes a conviction about humans as such in relation to the entirety in which we are set and, thereby, includes some or other ultimate terms of evaluation.Footnote 13 To all appearances, moreover, both the “traditional” and “voluntarist” views, which Smith finds at the outset of the Republic, assert that a stable political order requires a common religion precisely because a religion typically includes a comprehensive conviction: unless a common religion is widespread within the political order, those views agree, members of that body politic would have diverse ultimate terms of evaluation and could not be civilized. Let us, then, call the question to which religions and all comprehensive convictions are so many differing answers the comprehensive question—at least some of whose answers are religious.
On the reading here attributed to Smith and Amar, an answer to the first-order question of how government relates to religion is left to each state, at least in part because the people throughout the several states disagreed about whether a political community requires an established religion. Here, I take establishment to mean, as Smith and others outline, an official religion whose comprehensive beliefs are somehow taught to the state's citizens—because, so the establishment theory goes, a common religion is essential to a stable political order, and governmental support for that religion is essential to its commonality.Footnote 14 So understood, an established religion should indeed be stipulated in the state's constitution, precisely because it provides the terms in which strictly all human activities—and, a fortiori, all political activities—should be evaluated. Accordingly, proper citizenship includes a commitment to evaluate the state's actual and proposed political activities in accord with the ultimate terms of evaluation given with that religion.Footnote 15
But, then, a philosophical problem appears if the solely jurisdictional reading is advanced as a candidate for the relation to religion characterizing the national government—and the problem is this: Because a religion includes a comprehensive conviction, an established religion in any given state stipulates the ultimate terms in which to evaluate strictly all human activities, including necessarily the evaluation of all political claims in national politics. But the nonestablishment clause, as I discuss above, stipulates religious freedom nationally, and the religion established in a given state is not necessarily the proper basis for the evaluation of claims in national politics. Perhaps some will say that what is established in a given state may then be advocated in national politics. But this response fails to reckon that ultimate terms of evaluation apply to all human activities. Precisely because it includes comprehensive terms of evaluation, an established religion in some or other state does not permit the state's citizens to confine that establishment to state politics. To the contrary, a citizen of the given state is bound to assess actual and proposed decisions in national politics as if the same religion were established there, that is, without heed to the nation's practice of religious freedom and, if religious freedom makes sense, whatever political process it requires.
More often than not, perhaps, “religion” is used in both academic and public discussion to designate, not all comprehensive convictions but, rather, some among that class. In other words, “religion” designates in what we may call the conventional sense, namely, those comprehensive convictions on which, roughly speaking, the ground for ultimate terms of evaluation is human relation to some or other reality transcendent to the world and, thereby, such religious convictions are distinguished from secularistic comprehensive convictions. To the best of my reading, something like the conventional sense of “religion” is called to mind when both Smith and Amar use the term. As far as I can see, moreover, something like this conventional sense also defines religion for Rawls, for whom religions are one kind of comprehensive doctrine, along with philosophical and moral doctrines. But this conventional understanding of religion reads into the nonestablishment clause a certain kind of answer to the question about ultimate terms of evaluation—such that other kinds of answers (in Rawls's account, moral or philosophical answers making no appeal to a transcendent reality) are not protected. As a theoretical question, then, why constitutional protection is so limited requires a theoretical explanation beyond the difference between comprehensive and more specific terms of evaluation.Footnote 16
As far as I can see, the implied explanation is the following: religions in the conventional sense include suprarational convictions about what is transcendent to the world. In other words, a religion is inherently the expression of what is finally a suprarational belief in the transcendent reality,Footnote 17 where “suprarational” designates the belief as somehow exempt from validation and invalidation by the giving of reasons.Footnote 18 In that respect, among others, contemporary political theory is indebted to Kant, for whom there can be no rational knowledge of a transcendent reality—and Rawls, Dworkin, and Habermas all consider themselves indebted to Kant. Given such suprarational beliefs, in any event, a democratic process should separate or exclude all religions from political discourse—or to say the same thing, public discourse or public reason cannot include any ultimate terms of moral and political evaluation that depend on something transcendent to the world.
If all religions in the conventional sense include a suprarational conviction, we can then provide provisional sense for the solely jurisdictional reading of the nonestablishment clause. Even if that clause does provide for religious freedom in national politics, the suprarational character of religions allows the stipulation that each state has jurisdiction over the proper relation between government and religion because religions can be separated from the democratic determination of political decisions through discussion and debate among “we the people.” A solely jurisdictional reading of the first religion clause would be (provisionally) compatible with this exclusion because the consequent national nonestablishment would be then inconsequential to politics, and religious freedom would merely designate the freedom to accept any suprarational (and nonpolitical) faith one pleases. Accordingly, the apparent contradiction between a state in which some religion is established and the stipulation of religious freedom nationally would disappear. Whether a state does or does not establish a religion would make no difference to national politics; an official or voluntary religious commitment in a given state would be irrelevant nationally.
As an explanation for the solely jurisdictional reading of nonestablishment, however, the suprarational account of religion is only provisional because self-contradictory. Any given religion in the conventional sense claims that human relation to something transcendent to the world is the sole authorization or ground for all evaluation, that is, provides the terms in which to evaluate strictly all human activities, including political activities. To suppose that religious convictions are suprarational and, therefore, separate from political discourse is incoherent—because whatever is taken to be the terms for evaluating all human activities cannot be separated from the evaluation of any human activity. Hence, the exclusion of these religions from democratic political determination is so far from protecting their freedom as to deny them—and to assert, by implication, that politics does not depend on ultimate terms that relate humans to something transcendent to the world.Footnote 19
Once one allows the suprarational character of religions in the conventional sense, moreover, all answers to the comprehensive question, whether religious in the conventional sense or secularistic, must be placed outside of politics. In other words, all comprehensive convictions must be separated or excluded from the democratic political discourse because none of them can be validated or invalidated by public reason. If any given one could be so validated, all contrary others would thereby be proven false, and if any given one could be invalidated by public reason, the criterion for assessment would be rational. Hence, the comprehensive question, to which differing comprehensive convictions are so many differing answers, must itself be a suprarational question.
The suprarational view of comprehensive convictions then becomes something very like the political liberalism advocated by the later Rawls—at least if, as is to all appearances the case, he separates all “comprehensive doctrines” from the domain of politics. In Rawls's book, Political Liberalism, the basic distinction is between a political conception and a comprehensive doctrine.Footnote 20 Indeed, this distinction is, in truth, a separation because politics is entirely circumscribed, at least insofar as the “basic structure” of society is the object, by the political domain, and thus that domain is separated from all comprehensive doctrines (or convictions). Principles of justice, then, are (at least insofar as the “basic structure” is the object of politics) “freestanding,” that is, independent of any particular comprehensive doctrine. This does not deny the importance of such doctrines, each of which gives to one or more members of the society their complete conception of the good to be pursued and without which there would be no reason for justice. Thus, the basic distinction is repeated in Rawls's insistence that being “reasonable” is distinct from being “rational,” where a citizen is rational in pursuit of her or his complete conception of what is good and is reasonable insofar as that conception allows her or him to tolerate other comprehensive doctrines and thus adhere to principles of justice independent of any oneFootnote 21—and the “right” has “priority” to “ideas of the good”Footnote 22 or justice is prior to any comprehensive doctrine. Conceptions of the good should be consistent with whatever principles of justice are determined by public reason.
To the best of my reading, however, Rawls never defends the separation of political conceptions from comprehensive doctrines. This is because his basic distinction cannot itself be derived from a particular comprehensive doctrine and thus must itself be within the political domain—in which case, the separation is simply reasserted or assumed. Moreover, Rawls cannot defend the separation, I expect, without deriving it from a particular comprehensive doctrine and thus some ultimate terms applicable to all human activities, political or otherwise. In other words, the political domain that Rawls seeks to keep separate from comprehensive doctrines cannot be defended without comparing it to all other human activities, and the only possible terms for this comparison are the ultimate terms of evaluation.
In his later work, Rawls grounds the principles of justice “in terms of certain fundamental ideas seen as implicit in a public political culture of a democratic society”Footnote 23—and this apparently means the absence of any universal principle or principles of justice. But if universal principles are absent anywhere, they are absent everywhere, so that justice can only be in all respects relativistic or contextual, dependent in all respects on the specific cultural and social situation in question. As many have noted, however, this relativism is itself involved in self-contradiction because, in its own way, dependent on the universal principle that political terms of evaluation are always and everywhere limited in all respects to context—an assertion that implies a comparison with all possible human activities and thus ultimate terms of evaluation.
As a result, the problem invading Rawls's theory is general to any democratic theory that separates principles of justice and, thereby, public reason from the comprehensive convictions religious freedom protects: warrant for that separation can be provided only by a particular comprehensive conviction. For this reason, every separation of politics or justice from whatever convictions religious freedom protects is, against itself, required to place within the political community's constitution a stipulation of this separation, and because the constitution constitutes the political process, all citizens are required to accept it. Thereby, the constitution contradicts religious freedom because even contrary religions in the conventional sense assert, as do all comprehensive convictions, that politics cannot be separated from what is (in that religion) affirmed to be the ultimate terms of evaluation. To be repetitious, whatever is taken to be the terms for evaluating all human activities cannot be separated from the evaluation of any human activity.Footnote 24
In apparent contrast to Rawls, for instance, Dworkin's liberalism asserts universal principles of justice, even while separating these from conceptions of the good. “The first principle . . . holds that human life has a special kind of objective value. . . . The second principle . . . holds that each person has a special responsibility for realizing the success of his own life, a responsibility that includes exercising his judgment about [that is, determining by decision] what kind of life would be successful for him.”Footnote 25 In other words, moral and political principles should be separated from what Dworkin calls, in an uncommon distinction, the ethics of a given person or the good for which a person strives.Footnote 26 Although the principles are universal, justice in Dworkin's liberalism remains separated from comprehensive convictions or conceptions of the good.
In somewhat similar fashion, Habermas asserts that democracy separates political norms from the encompassing values that individuals pursue. “The universalization principle acts like a knife that makes razor-sharp cuts between evaluative statements and strictly normative ones, between the good and the just.”Footnote 27 Thus, justice is separated from one's own values, and “only secular reasons count beyond” what Habermas calls “the institutional threshold” of “parliaments, courts, ministries, and administrations”Footnote 28— where “secular reasons” designates reasons consistent with “the equalitarian individualism of modern natural law and universalistic morality.”Footnote 29 Notwithstanding the universality asserted by both Dworkin and Habermas, as far as I can see, this separation itself depends on a comprehensive conviction on which justice is independent of any ethics or encompassing values for which a person strives—and thus contradicts every contrary comprehensive conviction, for each of which it is implied by every principle of justice. Both authors, in other words, assert the priority of this independent justice and, at least by implication, require its constitutional stipulation, whereby democratic citizens are bound to accept it, such that every contrary comprehensive conviction is delegitimized.Footnote 30
Some will perhaps hold that separating justice from religious or comprehensive convictions is sufficient to democratic discourse because it works in practice. Pursuit of an “incompletely theorized”Footnote 31 consensus on some more or less specific political outcome, that is, an agreement diverse participants join for differing fundamental reasons, may often be a wise counsel to democratic citizens. The time and capacities of humans are limited, and special demands are involved if critical reflection turns to the most fundamental convictions. These facts are pertinent in politics, where often a common decision is required in the more or less short term. Given extensive and profound pluralism, citizens may be well advised in many situations of discourse to abstract from basic disagreements insofar as possible and to seek more specific values and purposes each takes to be authorized by her or his comprehensive conviction. But incompletely theorized agreement cannot be a successful theory of politics with religious freedom. Taking the wise counsel to constitute public reason simply presents another separation of political principles from the diversity of comprehensive convictions.
I conclude that a suprarational understanding of the convictions protected by the nonestablishment clause requires, against itself, an established comprehensive conviction authorizing this separation. Perhaps this establishment is itself a religion in the conventional sense—for instance, the separation between God's eternal law (which is said to be suprarational) and the natural law (which is said to be rational) advocated by Aquinas.Footnote 32 But if not a religion in the conventional sense, then an established secularistic conviction that denies all others—for instance, perhaps the secularistic conviction implied, and separation advocated, by Rawls or Dworkin or Habermas. If religious freedom is also constitutionally provided, so that differing comprehensive convictions are present in the political community and all such convictions are suprarational, they can relate to each other only strategically because each takes itself to be the sole grounds for any evaluation. A democratic political order can be no more than a modus vivendi. Life together is, in that sense, without any principle and is a civil war waiting to happen, and union depends on governmental coercion—one form of which is the constitutional separation of justice from any contrary comprehensive conviction, all of which are, in contradiction to the provision for religious freedom, thereby denied.
Some may avoid the need to establish a comprehensive conviction authorizing the separation of justice from any such conviction by asserting that US politics is nothing other than a modus vivendi, that is, a process to which all comprehensive convictions relate strategically. That view implies the absence in US politics of any moral claim because justice is not a moral obligation. Strategic relations are sufficient to politics because one is bound to consult nothing other than hypothetical imperatives consequent on what individuals or particular groups want or what is merely asserted to be important in an individual's or a particular group's life. But any view to that effect in fact delegitimizes all comprehensive convictions—at least if each such conviction purports to be true or makes a claim to validity for its ultimate terms of evaluation. Given that claim, a comprehensive conviction asserts the supreme moral law and purports that justice is a moral term, derivative from that supreme law.
In this discussion, I assume that comprehensive convictions, whether religious in the conventional sense or secularistic, claim to be true—at the very least, some religious or comprehensive convictions make that claim. I am aware of the view that all supposedly moral claims are nothing other than emotive or can never be true or false, but I am aware of no theorist of religious freedom, whether he or she takes the comprehensive question to be suprarational or rational, who accepts that account. As far as I can see, it implies that popular sovereignty is impossible because there is finally nothing to discuss and debate—and I expect, although do not here pursue, that a denial of all moral claims can be convincingly defeated.Footnote 33
An Alternative Reading
If the foregoing is, on the whole, correct, the solely jurisdictional reading of nonestablishment receives provisional grounds as a theory of democracy with religious freedom only if religions are suprarational. The indeterminate religious diversity with which a national body politic is constituted is then excluded from, and the establishment of religion in any given state is then irrelevant to, the national democratic discourse. On that theory, moreover, the comprehensive question itself is suprarational. To repeat the point expressed earlier: if any given answer to that question could be validated by public reason, all contrary others would be proven false, and if any given one could be invalidated by public reason, the criterion for assessment would be rational. The grounds are provisional because, if religions answer a suprarational question (as do all comprehensive convictions), nonestablishment is incoherent. To continue being repetitious: a religion includes some or other ultimate terms of evaluation, and whatever is taken to be the terms for evaluating all human activities cannot be separated from the evaluation of any human activity.
Accordingly, justice in a democracy with religious freedom must be separated from comprehensive doctrines or ethics or encompassing values, and a comprehensive conviction authorizing this separation must be constitutionally stipulated. All contrary comprehensive convictions are thereby denied by the constitution, and there is, in truth, no religious freedom at all. This consequence, then, leads to an alternative account of the nonestablishment clause, namely, that the comprehensive question is rational. If a solely jurisdictional theory of that clause is incoherent because it depends on the suprarational character of comprehensive convictions, the alternative is that such convictions can be assessed by public reason and some answer or answers redeemed by sufficient argument.
As far as I can see, democracy as popular sovereignty, such that “we the people” are the final ruling power, makes sense if and only if “we the people” can engage in full and free political discourse. “Full” here means that every political claim, including those for comprehensive convictions in their pertinence to politics, can be contested and, if contested, requires validation by argument, even if the person who makes the claim cannot supply the argument; and “free” means that all adult citizens are together as equals. In other words, popular sovereignty is consistent with political order or with a people united in principle only because action-as-one occurs through the way of reason. Absent the possibility of full and free political discourse, at least in sufficient measure to overcome the mere assertion of comprehensive beliefs and diverse actions derived therefrom, differing comprehensive convictions are, as mentioned above, reduced to solely strategic deliberation—and if two or more parties in conflict take the political issue at stake to be sufficiently important, they can only fight.
In this respect, we may recur to the eloquence of Thomas Jefferson: as he wrote in concluding the rationale behind his Bill for Establishing Religious Freedom in Virginia, “and finally, that truth is great and will prevail if left to herself; that she is the proper and sufficient antagonist to error, and has nothing to fear from the conflict unless by human interposition disarmed of her natural weapons, free argument and debate; errors ceasing to be dangerous when it is permitted freely to contradict them.”Footnote 34 In any case, democracy as popular sovereignty is possible if and only if the question about true ultimate terms of evaluation is a rational question, that is, the true answer can be established by good argument. A truly democratic constitution should stipulate religious freedom or Smith's “third position” because and only because reason is left free to combat all errors of comprehensive conviction. Popular sovereignty is a coherent possibility only as a product of the Enlightenment.
I do not seek at length to defend the rational character of our comprehensive question, that is, some answer to that question can be established by public discourse and its several answers can be validated or invalidated thereby. Given how widely religions in the conventional sense are thought to assert suprarational convictions, however, a word about this defense is in order: the true answer includes the ultimate terms of evaluation or the supreme moral law, and individuals cannot be moral or immoral unless they understand those terms. In other words, the Kantian dictum, “ought implies can,” has two meanings: On the more familiar account, no prescription can be valid for any person to whom it applies unless she or he includes among the alternatives for decision what ought to be chosen. But the dictum also means this: no prescription can be valid for any person to whom it applies unless she or he is aware of what ought to be chosen and thus can choose it because (for the reason that) she or he ought to do so.
In the occurrence of a moral or immoral decision, ignorance of the supreme moral law is impossible for moral (in distinction from nonmoral) creatures or subjects as such; immorality chooses the false even while aware of what ought to be decided. Because all participants in public reason are subjects, they are necessarily aware of the moral law and ask and answer the comprehensive question, at least implicitly, whenever they evaluate alternative courses of their activity. In other words, the moral law is at least implicitly understood by them or is a part of common (adult) human experience, as is its ground or the backing for why the moral law is prescribed. For this reason, explicit answers to the comprehensive question can be included within public reason because the true answer is included within common (adult) human experience.
As a consequence, I propose the following: the term “religious” in “religious freedom” (and thus “religion” in the nonestablishment clause) should be understood to mark any explicit conviction about the ultimate terms of evaluation and, at least typically, the ground for those terms. Among the great merits of Rawls's proposal is his recognition that democratic government ought to protect all comprehensive doctrines. To be sure, as noted above, Smith argues that every theory of religious freedom implies certain “background beliefs” that privilege one among the many answers about ultimate terms of evaluation to which the government supposedly should be neutral (63). But a question about those terms can be given an explicit meaning, even if that question implies a given answer—and, thereby, establishing the question is neutral to all answers. To the best of my reasoning, Smith ignores precisely a difference between the explicit meaning of the question and the “background beliefs” it may imply. To ask about the ultimate terms of evaluation may well be to imply the valid answer—because those terms, which obligate all humans in all of their activities, will be implied even by the activity of asking the question. But whether the valid answer is religious in the conventional sense (and, if so, what religious answer is implied, and how it is warranted) or secularistic (and if so, what secularistic answer is implied, and how it is warranted) may itself be debated. Explicitly to formulate the comprehensive question can leave its valid answer to a full and free political discourse.
It follows that “religious” in “religious freedom,” as constitutionally stipulated, is rightly defined in an extended or broad sense—and with respect to national politics (or to state politics, following “incorporation”), that stipulation establishes nothing other than the comprehensive question because it stipulates that any explicit answer to that question is legitimate. In other words, a constitution providing religious freedom says nothing explicitly about the ultimate terms of evaluation, leaving to a full and free discourse among “we the people” what answer to the comprehensive question is true—and if religious freedom is morally valid, the answer implied by even that stipulation, that is, by the comprehensive question itself. Accordingly, citizens so constituted are free to affirm—or do so legitimately, without compromising their citizenship—any such ultimate terms each takes to be true, and neither the constitution nor statutory law may stipulate any such conviction. This, on my accounting, is the only principled constitutional proscription on government that could be stipulated by a nonestablishment clause, namely, that neither a constitution nor a government shall make a law “respecting an establishment of religion.”
On this accounting, moreover, “religious” or “religion” should be understood in this broad sense because only so is popular sovereignty or government by active consent of the governed consistently affirmed. The people can be the final ruling power, having a “sovereignty original and unlimited,”Footnote 35 only if each is religiously free in the broad sense, that is, free to believe any ultimate terms of evaluation she or he finds convincing, and, therefore, sovereign over every political claim. Religious freedom in this sense is compromised whenever the constitution or statutory law constrains the people's power by stipulating the ultimate terms of evaluation, thereby providing that citizens cannot assess those terms and, to the contrary, should simply accept and deliberate from them.
Were the constitution or statutory law to stipulate, say, that democracy is a morally valid form of government, ultimate terms of evaluation from which a citizen advocates monarchy or aristocracy would be themselves delegitimized in the political community, and insofar the people would not be the final ruling power. Indeed, religious freedom is, by implication, violated whenever constitutional or statutory law stipulates any substantive political evaluation.Footnote 36 This is because the point of religious freedom as a constitutive political provision is democracy as popular sovereignty. If government stipulates a substantive political evaluation, the constitution or statutory law would deny the sovereignty of each citizen over every political claim; that is, would deny any ultimate terms of evaluation whose application is said to dissent from the stipulation—and, thereby, would, in effect, stipulate the ultimate terms presupposed by the given political evaluation. Hence, the test for whether any constitutional or statutory law is consistent with religious freedom is whether obedience to that law leaves every citizen free to contest it; that is, whether the law is consistent with a full and free political discourse.Footnote 37
Thus, popular sovereignty means that constitutional provisions, like any statutory law, should be such that even they, too, are objects of assessment by the peopleFootnote 38—and we can make this point by saying that a constitutional stipulation of religious freedom should itself be religiously free, that is, must be something whose provision is itself subject to popular assessment. This means that popular sovereignty itself can be called into question, and the people may decide to assign sovereignty elsewhere; hence, “we the people” cannot continue to be the final ruling power unless they are convinced that popular sovereignty is consistent with the ultimate terms of evaluation. In his First Inaugural, Jefferson exhorted, “If there be any among us who would wish to dissolve this Union or to change its republican form, let them stand undisturbed as monuments of the safety with which error of opinion may be tolerated where reason is left free to combat it.”Footnote 39
The commitment to a rational discourse about the ultimate terms of evaluation in their pertinence to politics is itself implied by religious freedom. In other words, that commitment should not be constitutionally stipulated because a conviction on which those ultimate terms are suprarational is itself legitimate. If the stipulation that all citizens are religiously free should itself be, as mentioned, religiously free, that constitutional provision can be what it should be given only the difference between an explicit establishment of the comprehensive question and the presence only in the political discourse of differing answers. Any citizen for whom the ultimate terms of evaluation are beyond reason is, as it were, invited to argue for her or his belief that evaluation depends on some suprarational termsFootnote 40—although other members of “we the people” need not consider that advocacy important unless it can be validated by argument. The point, in other words, is not whether a religion claims to be suprarational but, rather, whether the comprehensive question is indeed rational. Ideally, then, a democratic constitution should stipulate only (a) the rights inherent in being a member of “we the people,” and (b) a decision-making procedure maximally informed by the discourse and through which governmental activities are effected.
Conclusion
On the best of my perception, the suprarational account of religious belief is now so taken for granted that it is thought no longer to need a defense: all religious belief (“religious” is here intended, typically but not exclusively, in the conventional sense) is said to be transparently a matter of faith in whatever the given religion asserts. At least that account is found or implied, on my reading, not only in the proposals of Rawls and Dworkin and Habermas but also in those of Christopher Eberle, John Courtney Murray, Michael Perry, and Jeffrey Stout.Footnote 41 Often affirmed by one or another religion itself, the suprarational view of faith perhaps derives from the massive success of modern empirical science in explaining details of the world, resulting in a widespread conviction that all rational questions about human life in relation to other things are scientific—the so-called scientific view of the world.
If modernity has included a “war between science and religion” about the world's character, in other words, science has seemingly won, and religions, some say, have purchased continuing relevance to modern life only by asserting that religious beliefs are suprarational. Be that as it may, a solely jurisdictional theory of the nonestablishment clause depends on that assertion. But one cannot deny validation and invalidation by public reason of all answers to the comprehensive question without implying that a true answer must be, because the question itself is, suprarational. Hence, all answers to that question must be excluded from public reason or democratic discourse. This exclusion of those answers in fact denies their comprehensive claims—even while that exclusion can be defended only as a political application of a particular comprehensive claim, which must be constitutionally established.
In contrast, the question about how humans relate to the entirety in which we are set is, as far as I can see, rational—and calls for an answer that implies metaphysics, that is, an account of ultimate reality or existence as such to which humanity and its communities are related and by which the ultimate terms of evaluation are given their ground. Accordingly, ultimate reality is not the object of scientific inquiry because metaphysics so understood defines the possible as such and thus its statements are necessarily true—and if something transcendent to the world is possible, includes the nature of that transcendent reality. On my accounting, popular sovereignty is possible because common (adult) human experience includes, at least implicitly, the truth about human life as such—and explicit comprehensive convictions are so many differing attempts to explicate something all humans experience. This implicit experience, then, is independent of the differing explications and thus something to which the way of reason may appeal in seeking the true answer to the comprehensive question. In other words, the comprehensive truth about human life as such is implied by human life as such, inclusive of its relation to the entirety in which we are set, and false answers to the comprehensive question are pragmatically self-contradictory.
In any event, we can conclude this: popular sovereignty is possible if and only if the comprehensive question is rational. Thus, an affirmation of “we the people” as the final ruling power commits one to assessment of all such religious beliefs by the way of reason.Footnote 42 The solely jurisdictional view of religious freedom implies the suprarational character of religious convictions and, thereby, is inconsistent with their comprehensive character. Accordingly, that view is impossible as a theory of religious freedom. However those who originally affirmed the nonestablishment clause understood their deed, we today who endorse popular sovereignty are bound also to endorse an assessment within public reason or democratic discourse of religious explications—whether “religious” be taken in the conventional or broad sense—and, therefore, a rational assessment of both religious (in the conventional sense) and secularistic comprehensive convictions in their pertinence to politics.
