Hostname: page-component-74656f9884-j7cmm Total loading time: 0 Render date: 2026-11-09T18:30:42.076Z Has data issue: false hasContentIssue false
  • English
  • Français

THE PLURALIST PREDICAMENT:

Contemporary Theorizing in the Law of Religious Freedom

Published online by Cambridge University Press:  07 June 2004

Steven D. Smith
Steven D. Smith
Affiliation:
University of San Diego School of Law
Rights & Permissions [Opens in a new window]

Abstract

An abstract is not available for this content. As you have access to this content, full HTML content is provided on this page. A PDF of this content is also available in through the ‘Save PDF’ action button.

Information

Type
ARTICLES
Information
Legal Theory , Volume 10 , Issue 2 , June 2004 , pp. 51 - 76
Copyright
© 2004 Cambridge University Press

“Theory” means many things, but in law the term typically denotes an effort to impose order on an unruly collection of phenomena—of seemingly conflicting decisions, or doctrines, or legal arguments—and to do so by self-consciously articulating and elaborating the foundational truths, or principles, or policies that govern the subject in question. The modern law of religious freedom is celebrated for its unruliness; hence the need for domesticating theory may seem urgent. Unfortunately, the conditions in which we live actively embarrass efforts to provide such theory. This essay considers the nature of our current embarrassment.1

The essay was written in response to an invitation for a discussion, in approximately 10,000 words, of the “state of the art” in contemporary theorizing about the law of religious freedom. Any such undertaking necessarily must make heavier-than-usual use of selection and distillation. These streamlining methods have their risks and their costs, but they might also have benefits; readers will judge.

Unlike their founding-era ancestors and to a greater extent than their European counterparts, lawyers and citizens in the United States today typically understand religious freedom in terms of two constitutional commitments—to “nonestablishment” or “separation of Church and State,” and to “free exercise”—that are at least somewhat independent or even conflicting. The first commitment has been expressed in legal doctrines requiring government to confine itself to the domain of the “secular,” to be “neutral” toward religion, and to avoid saying or doing things that send messages either “endorsing” or disapproving of religion.2

See, e.g., Lemon v. Kurtzman, 403 U.S. 602 (1971); Allegheny County v. ACLU, 492 U.S. 573 (1989).

Decisions under these nonestablishment doctrines are notoriously confused and conflicting. The second commitment—to the “free exercise” of religion—has sponsored judicial vacillations between doctrines purporting to require government to avoid burdening the exercise of religion and doctrines that merely forbid persecution of or discrimination against religion; and the scope and contours of these doctrines are exquisitely murky as well.3

See, e.g., Sherbert v. Verner, 474 U.S. 398 (1963); Employment Division v. Smith, 494 U.S. 872 (1990). For an argument showing that these apparently different doctrines collapse into each other, see Larry Alexander, Are Smith and Hialeah Reconcilable? 13 CONST. COMM. 285 (1996).

So the call for ordering theory is clear. And yet little that deserves to be called “theory” has been forthcoming. There are, to be sure, recognizable orientations—towards “strict separation,” for instance, or “substantive neutrality.” But these orientations are most often supported by what we may call (with no condescension intended) “lawyers' arguments”—recitation of precedents, appeals to legal text or “framers' intentions” or entrenched mythologies about the same, ad hoc rummaging among an assortment of ostensible principles or policies. There seem to be at least two reasons why the modern discourse of religious freedom has been relatively theory-lite. The more superficial reason has been that U.S. judges and lawyers seem to have supposed that questions of religious freedom could be answered simply by invoking the constitutional text or its “original meaning.” So history-heavy, theory-thin analyses and polemics have played a larger role here, in both case law and scholarship, than in some other fields.

The deeper reason for the avoidance of theory—and one that arguably motivates efforts to squeeze more out of text and history than those sources can fairly be asked to supply—is, in a word, pluralism. Judges and scholars have consciously or unconsciously supposed, not without reason, that the rampant religious pluralism of our times is an impediment to the sort of foundational theorizing that seems more feasible in other fields. Religious pluralism in a sense provides the subject matter of, or at least the conditions for, theorizing about religious freedom; were it not for the fact that citizens and officials adhere to diverse religious faiths, or to none, issues of religious freedom would not arise. But pluralism also places daunting obstacles in the way of would-be theorists.

WHY HISTORY CANNOT OBVIATE THEORY

Let us start with the first of these reasons. One might almost suppose that the relative paucity of theory in the modern law of religious freedom derives from the accident that the Supreme Court's seminal modern nonestablishment decision happened to be written by a Justice who vaunted his loathing for theory (or for anything that in his loose but capacious conception smacked of “natural law”) and his love for text and history. Hence, in Everson v. United States,4

330 U.S. 1 (1947).

Justice Black's majority opinion purported to rely entirely on an analysis of “original meaning”—and the argument about what “the Framers intended” was on. In comparison with some areas of constitutional law (such as free-speech jurisprudence) in which text and “original meaning” have not played such a prominent role and even self-styled originalists proclaim the need for theory,5

See Robert Bork, Neutral Principles and Some First Amendment Problems, 41 IND. L.J. 1, 22 (1971) (“We are…forced to construct our own theory of the constitutional protection of speech. We cannot solve our problems simply by reference to the text or to its history”).

the law of religious freedom has often been treated by both judges and legal scholars as a straightforward product of history—history manifesting itself in the constitutional text.

Some of the historical work—though not, alas, Justice Black's—has been responsible and illuminating. But even careful history cannot escape the need for at least implicit theorizing. The reason, tersely put, is that history is not self-interpreting and even at its most obliging it delivers to us concepts or principles that are not self-defining or self-limiting. Even the most sanguine of originalists seem to acknowledge that historical research can at most supply us with some general principle that the religion clauses were supposed to embody. (Less sanguine researchers often conclude that the original meaning cannot be ascertained at all, or that it is ambiguous and uncertain, or that the religion clauses were initially conceived as a purely jurisdictional measure assigning authority over religion to the states rather than the national government and that this jurisdictional allocation has long since been forgotten or repudiated.)6

I have argued at length for this jurisdictional interpretation in STEVEN D. SMITH, FOREORDAINED FAILURE: THE QUEST FOR A CONSTITUTIONAL PRINCIPLE OF RELIGIOUS FREEDOM 17–50 (1995).

So interpretation of the religion clauses requires us, first, to figure out what “principle” or principles are embodied in the clauses, and second, to elaborate the meaning of such principles in a way that makes them presently useful and attractive.

The second of these tasks—elaborating the meaning of the constitutional principles—most obviously calls for theoretical work. Contrary to what “separationist” advocates sometimes seem to suppose, it does not automatically follow from a principle forbidding specific subsidies of Christian ministers that the state cannot include religious schools in a general voucher program; a good deal of theoretical labor is need to get from Point A to Point B.

But realistically, the first task also calls upon us for theorizing, because the text and historical context simply do not yield any ready-made principle of religious freedom, even at a high level of abstraction.7

According to the jurisdictional construction noted earlier, this failure to select among principles was not an oversight or shortcoming but rather was quite deliberate; the whole point of the clauses was not to make such a selection but rather to leave the matter to the states.

So if we are going to gather some principle from the text or original meaning, it seems we have little choice but to resort to Dworkinian-style interpretation—to read the clauses as standing for the candidate principle that will make them “the best they can be.” We can limit the candidates, if we choose, to those that we think might have seemed plausible at the time of the founding; even so, selection will be necessary. And that selection will call for argumentation that seeks to show that some particular principle or principles is more attractive than the alternatives.

Consequently, beyond the general illumination it provides, good history serves mainly to counter the dubious inferences that advocates draw from bad history and perhaps to set some outer parameters for the principles with at least a prima facie claim to be “the Constitution's principle.” But even the good history cannot tell us what religious freedom under the Constitution should mean.

As an example, take Philip Hamburger's recent study.8

PHILIP HAMBURGER, SEPARATION OF CHURCH AND STATE (2002).

Hamburger's book is a carefully researched, skillfully executed effort directed against the assumption that the founding generation erected, as Jefferson wishfully put it, a “wall of separation” between Church and State. Hamburger argues that at the time of the founding, “separation” was viewed as a noxious doctrine that pontificators and politicians of all sorts tried to pin on their opponents, and he traces how support for separationism developed later, in the nineteenth and twentieth centuries, and largely in connection with unsavory nativist and anti-Catholic movements such as the Know-Nothing Party and the Ku Klux Klan.

Hamburger's is a valuable study that effectively subverts what we might call the simplistic separationist story told by Justice Black and, in only slightly less simplistic terms, by Justice Souter.9

See Rosenberger v. Rector, University of Virginia, 515 U.S. 819, 863, 868–872 (1995) (Souter, J., dissenting).

Still, Hamburger cannot tell us what religious freedom should be understood to mean today. Indeed, his study does not even foreclose the possibility of a somewhat more sophisticated “separationist” interpretation. In fact, the “nonestablishment” views that Hamburger says were widely held in the founding period sound not so different from modern separationist interpretations;10

See, e.g., HAMBURGER, supra note 9, at 94–95. For a review elaborating on this difficulty, see Douglas Laycock, The Many Meanings of Separation, 70 U. CHI. L. REV. 1667 (2003).

it is easily imaginable that a sophisticated advocate might find in those founding-era positions the basic “principle” of which modern separationism is a logical development.

In a similar vein, consider Michael McConnell's helpful study of the meaning of religious “establishment” in the founding era. McConnell shows that although “establishments” differed, they typically involved some combination of six features:

(1) [governmental] control over doctrine, governance, and personnel of the church; (2) compulsory church attendance; (3) financial support; (4) prohibitions on worship in dissenting churches; (5) use of church institutions for public functions; and (6) restriction of political participation to members of the established church.11

Michael W. McConnell, Establishment and Disestablishment at the Founding, Part I: Establishment of Religion, 44 WM. & MARY L. REV. 2105, 2131 (2003).

In light of McConnell's depiction, modern descriptions of comparatively innocuous arrangements as “establishments of religion” may seem almost frivolous. Recently, having newly moved to San Diego from northern Indiana, I listened as my neighbor, a longtime resident, described the “winter” we had just endured (in which the daily temperatures frequently did not rise above the low 60s) as “brutal.” He had no grasp, I thought, of what winter even is. In a similar way, when a federal judge concludes that a long-standing lease between a city and the Boys Scouts is “an establishment of religion,”12

Barnes-Wallace v. Boy Scouts of America, 275 F. Supp. 2d 1259 (S. D. Cal. 2003).

the onlooker fresh from reading McConnell's account is likely to have a similar reaction: “With all due respect, your Honor, you don't have the faintest idea what an establishment of religion is.”

And yet McConnell's history still cannot actually tell us what should count as an establishment of religion today. Clearly not all six features are necessary because, as McConnell himself makes clear, even regimes considered to be “establishments” in the founding period did not necessarily exhibit all of these elements. So is one element—say, financial support—enough to make a religious “establishment”? Financial support plus use of religious institutions to perform public functions, as in the currently fashionable “faith-based initiatives”?

The point is that history by itself cannot answer such questions. In order to determine which elements are required, it seems, we would want to know why the Constitution prohibits establishments of religion in the first place. We would want something like a theory of nonestablishment—which presumably might be part of a more general theory of religious freedom.

HOW PLURALISM FRUSTRATES THEORY

But here we encounter the more serious problem: the conditions of modern pluralism have severely restricted the resources and options available to would-be theorists. So we need to consider just how pluralism has affected theorizing about religious freedom.

Pluralism and “the Constraint”

Often the story is told more or less in this way:13

See, e.g., Charles Taylor, Modes of Secularism, in SECULARISM AND ITS CRITICS 31 (Rhajeev Bhargava, ed., 1998); JOHN RAWLS, POLITICAL LIBERALISM xxiv–xxviii (paperback ed., 1996).

In the Middle Ages, Western peoples were united under a common religion, and so when issues about the proper relation between religion and government arose (as they often did), princes and popes and scholars naturally addressed those issues by appealing to what they took to be foundational truths as expressed in Christian scripture and theology.14

See almost any of the arguments collected in BRIAN TIERNEY, THE CRISIS OF CHURCH AND STATE, 1050–1300 (first published 1964). On the heavily Christian cast of early-modern theorizing, see, generally, ANDREW R. MURPHY, CONSCIENCE AND COMMUNITY: REVISITING TOLERATION AND RELIGIOUS DISSENT IN EARLY MODERN ENGLAND AND AMERICA (2001).

But the Protestant Reformation and its aftermath shattered this medieval unity, so that now the ultimate truths themselves are the subject of profound disagreement. And it follows—or seems to—that governments (and theorists who seek to explain and justify and advise governments) can no longer appeal to those contested truths. To do so would rob government of its legitimacy, at least for those citizens who reject the ostensible foundations. Moreover, to base public decisions on contested religious premises would arguably manifest disrespect for those citizens who do not embrace such beliefs; it might seem to treat them, to borrow a phrase, as “outsiders, not full members of the political community.”15

Lynch v. Donnelly, 465 U.S. 668, 688 (1984) (O'Connor, J., concurring).

By this account, in sum, the condition of modern pluralism effectively restricts government and its theorists from appealing to (contested) foundational beliefs or “comprehensive doctrines.” Let us call this restriction the “constraint on contested foundations” or simply, for short, “the constraint.”

And just what is the content of the constraint? In what we can call its “narrow” version, the constraint forbids government to base its authority and at least its most important decisions16

Theorists disagree about the domain of the constraint. The constraint might apply to all public decisions, or only to decisions enforced coercively, or it might apply only to matters involving “constitutional essentials and questions of basic justice”; RAWLS, supra note 14, at 214.

on contested doctrines (a category that presumptively includes religious doctrines as a central instance). The affirmative corollary of this proscription commands government to act in these vital matters on the basis of shared grounds—a category thought not to include religious beliefs, at least in a modern pluralistic society.

So government may not accept and invoke (presumptively contested) religious beliefs. But may government reject religious beliefs? Although the question is debatable, what we can call the “wide” version of the constraint, reflected in current establishment doctrine requiring neutrality and prohibiting government from sending messages either endorsing or disapproving of religion, suggests that it may not; rejection of a religious belief surely sends a message of disapproval. More generally, the constraint is derived from the perception that government forfeits legitimacy and manifests disrespect if it acts on religious beliefs that some citizens do not hold. But the same conclusion would seem to apply—a fortiori, perhaps—if government rejects religious beliefs that some citizens do hold.17

Some theorists seem of late to have embraced the view that government may reject some religious views so long as it does so implicitly, not explicitly. See, e.g., Steven H. Shiffrin, Liberalism and the Establishment Clause, 78 CHI-KENT L. REV. 717, 726 (2003); Andrew Koppelman, No Expressly Religious Orthodoxy: A Response to Steven D. Smith, 78 CHI-KENT L. REV. 729, 733–734 (2003). For my criticisms, see Steven D. Smith, Barnette's Big Blunder, 78 CHI-KENT L. REV. 625, 645–647 (2003).

In sum, the “narrow” version of the pluralist constraint maintains that within some specified domain government must act on shared grounds and hence must not invoke or rely on religious beliefs to justify itself or its important decisions. The “wide” version, which seems at least on first reflection to follow from the same logic, adds that government also must not reject religious beliefs.

REFINING THE PLURALIST STORY

Told in this way, the story about pluralism and the constraint provokes doubts. The depiction of pluralism as a modern, post-Reformation challenge is suspect. Was it a happy consensus on essential ideas that led Thomas Aquinas to undertake a monumental project of carefully listing and evaluating pro and con arguments on a vast array of the most significant theoretical questions in philosophy, law, and politics? And what about the large-scale movements of what the medieval Church viewed as heretics, as well as (internally) Jews and (externally) Muslims, who very much claimed the attention of Christian thinkers and political officials of the period? In short, any suggestion that pre-Reformation theorists could appeal to ultimates because nobody really disagreed about them seems vulnerable.

In addition, the inference from the fact of pluralism to the constraint against relying on contested “foundational” propositions should raise doubts. What we might call the premodern (or perhaps prepostmodern) response to pluralism ran just the other way: it acknowledged the fact of disagreement, inferred that among inconsistent basic beliefs at least some of them must be wrong, and strove in a Cartesian spirit to develop better methods of figuring out what the true beliefs were. Hence more than a century and a half after Luther posted his famous theses and decades after an English king was beheaded in the course of a civil war provoked in part by deep religious differences, John Locke was busy attempting to demonstrate the reasonableness of Christianity and justifying religious toleration (and political equality)18

See JEREMY WALDRON, GOD, LOCKE, AND EQUALITY (2002).

on the basis of (contested) Christian premises. A century later still, James Madison's famous Memorial and Remonstrance argued for religious freedom largely on (contested) theological grounds.19

See JOHN T. NOONAN, JR., THE LUSTRE OF OUR COUNTRY: THE AMERICAN EXPERIENCE OF RELIGIOUS FREEDOM 72–75 (1998).

And Jefferson's celebrated Virginia Statute for Religious Freedom began by declaring its essential (and controversial) premise—that “Almighty God hath created the mind free” and that coercion in religion was “a departure from the plan of the Holy author of our religion”20

Reprinted in THE SUPREME COURT ON CHURCH AND STATE 25 (Robert S. Alley, ed., 1988). For a discussion of the shift in theoretical frameworks for thinking about religious freedom from religious to more secular, see Michael W. McConnell, Why Is Religious Liberty the “First Freedom”? 21 CARDOZO L. REV. 1243 (2000).

It is arguable, in short, that pluralism (albeit in different degrees and forms) has been the usual condition of Western peoples and that the normal and natural response to pluralism is not to shun contested foundational truths but rather to strive with even greater care and energy to figure out what those truths are. And indeed, this response would seem natural enough today in many fields. Faced with disagreements, do physicists or biologists or economists conclude that any appeal to contested premises is taboo?

But these observations point us to a more precise description of what is distinctive about contemporary religious pluralism: it is not the bare fact of deep pluralism so much as the widespread acceptance of pluralism as permanent and natural and even desirable. If I disagree with you about a proposition of physics (perhaps I believe in the “Steady-State theory” rather than the “Big-Bang theory”), you will presumably think that I am mistaken and you will likely try to persuade me to change my opinion. But if I disagree with you about a religious proposition (I am a pantheist, perhaps, while you are a trinitarian), the contemporary ethos tells us that this disagreement is legitimate and enduring and that we will probably get along better if we agree, at least tacitly, to respect and defer to each other's beliefs.

Given a general acceptance of pluralism as a permanent and legitimate condition, in sum, most theorists of religious freedom are understandably loathe to base their theorizing on beliefs that are themselves contested. But these are the very beliefs that would provide the natural foundations for thinking about the proper relation between government and religion (and that at least until relatively recently did provide the foundations for such thinking). So the contemporary situation is peculiar; it is as if we were to tell a textbook writer: “Please include a chapter explaining the vast variety of life forms. Oh, and don't make any reference to notions of evolution or creation, because those notions are controversial.” In such a situation, what can theorists appeal to? How are they supposed to do their job?

ANTIFOUNDATIONALISM AS FOUNDATION?

Perhaps surprisingly, most scholars of religious freedom have not perceived their task to be nearly as hopeless as I have just made it seem. On the contrary, attempting to turn adversity to beneficial use, they have typically taken “the constraint,” which I have presented as a formidable obstacle to theorizing, as itself the proper foundation for and even the essential content of a theory of religious freedom—or at least of the nonestablishment commitment. And this turning of the tables has typically led them to adopt one or more among the ever proliferating variations on either of two themes (or perhaps two labels for one theme) that also figure prominently in modern legal doctrines: neutrality and secularity.21

The discussion in this section is my attempt to distill the key themes and objections in what has become a voluminous literature. For two especially careful and contrasting analyses, see CHRISTOPHER J. EBERLE, RELIGIOUS CONVICTION IN LIBERAL POLITICS (2002); and ROBERT AUDI, RELIGIOUS COMMITMENT AND SECULAR REASON (2000).

Neutrality embraces the constraint in its “wide” formulation: it suggests that government should avoid either invoking or disapproving religious beliefs by simply staying away from—remaining “neutral” with respect to—religion. Secularity is responsive to the constraint's “narrow” formulation: it tells government (and theorists) what shared and nonultimate purposes and grounds they can appeal to—namely, “secular” ones. Though the terms are used in various senses, in their core meanings they are usually viewed as complementary, or perhaps mutually entailing, or maybe even different labels for the same basic idea. The standard assumption among judges and legal scholars seems to be that a secular government is neutral toward religion and that a government that wants to be neutral toward religion can do so by limiting itself to the domain of the secular. Hence in a good deal of judicial and scholarly discourse the terms seem almost interchangeable.

The secular neutrality position is attractive—almost irresistible, it seems—in part because, as noted, it seems to solve the problem of pluralism and the constraint by adopting the constraint itself as the foundation for and content of thinking about religious freedom. If modern pluralism means we can no longer appeal to contested foundations, why not make the absence of foundations itself the foundation for our thinking on the subject? It is an ingenious gambit, and the list of theorists who have exploited some variation of it could make up a central chapter of a Who's Who of elite political theorists of our time: Rawls, Gutman, Macedo, Ackerman, Nagel, and Audi just for starters.

But the move also provokes suspicions. Isn't it plain that antifoundational notions such as “neutrality” and “secularity” are themselves contested, just as the more traditional foundations are? If so, what do we gain—and what might we lose?—by shifting from one kind of contested foundation to a different and more elusive kind of contested foundation? Indeed, is not the notion of an antifoundational foundation a contradiction in terms? The strategy, it seems, might depend on theorists being able to flash dexterously back and forth between the antifoundational and foundational sides of the secular neutrality card—showing the antifoundational side to slip past “the constraint,” then flipping to the foundational side when theory must be done or actual decisions made, then flipping rapidly back to the antifoundationalist side whenever pluralist objections arise. The theoretical or juridical hand will have to be quicker than the eye. So, then, what happens when the hand tires or grows clumsy—or when the audience starts paying closer attention?

These motifs and rationales and doubts have by now been played out in a vast and intricate (and, truth be told, often tedious) apologetical and critical literature; in a short essay it would be impossible even to begin to sketch the full flowchart of the debates. In particular, it would be foolhardy here to plunge into the treacherous currents surrounding “political liberalism” as a general theory or position.

Still, trying to limit ourselves to the field of religious freedom, we can perhaps say this much: in the midst of the sound and fury, the critique of secular neutrality presses the central claim that secular neutrality is not truly neutral. Purporting to respect pluralism and to ground itself in the constraint, secular neutrality in fact shuts its eyes to pluralism and wantonly violates the constraint. And the apologetical responses to this core objection amount to attempts—futile ones, in my view—to reframe and qualify the central position so as to avoid this central criticism.

The Impossibility of Neutrality

The central objection asserts that neutrality—genuine neutrality—is impossible. This assertion must be immediately qualified. The term “neutrality” is a label for many things, and nearly everyone will like some of those things.22

Recently, for example, “neutrality” has often been invoked by proponents of broadly based school voucher programs that include religious schools on a nondenominational basis along with nonreligious schools or of “faith-based initiatives” programs that allow religious service providers to receive government funding along with more secular providers. The approach may be attractive—I am inclined in my waffling to think so—and there is nothing to prevent the use of the term “neutrality” to describe it. But we should also note the very limited nature of this “neutrality”; it is a “neutrality” that self-consciously rejects the views both of no-aid “separationists” (both religious and secular) and also of those who favor public support for a more selective range of religion (Christian and Jewish but not Muslim, perhaps).

For example, the term is sometimes used in an essentially procedural sense: it means merely that whatever the governing rule or principle (or “baseline”) is, it should be applied consistently without respect to religion. If the law prohibits jaywalking, the police violate neutrality if they ticket jaywalking atheists but not jaywalking Christians. This sort of procedural neutrality is unobjectionable. Indeed, it is probably tautological; it merely asserts that whatever rules or criteria ought to govern, those rules or criteria ought to govern.23

For elaboration of this point, see Steven D. Smith, Symbols, Perceptions, and Doctrinal Illusions: Establishment Neutrality and the “No Endorsement” Test, 86 MICH. L. REV. 266, 325–331 (1987).

But a purely procedural neutrality also fails to address the live questions of religious freedom—or, for that matter, to ensure compliance with the constraint on contested foundations. Government could satisfy the demands of a purely procedural neutrality, for instance, by mandating daily recitation of the Apostles' Creed or the payment of tithes to the Orthodox Church and enforcing these requirements impartially against Christians, Jews, Muslims, agnostics, and everyone else. So the constraint calls for a more substantial neutrality; and it is that sort of neutrality that theorists attempt to provide—and that critics believe to be impossible, either on a “meta” level or on a more operational level.

On the “meta” level, we need merely note that the demand that government be “neutral” toward religion already effectively repudiates competing views that have been passionately advocated over the centuries and that still have their proponents. Many have thought that government should support and uphold religion; this view was widely held, for example, in founding-era America (both before and after adoption of the First Amendment). Recent controversies over, say, the Ten Commandments or the words “under God” in the Pledge of Allegiance arise from at least pale vestiges of this view. Others have thought—and continue to think—that government should actively discourage or even prohibit religion. Communist regimes have taken this position, for example, and Graham Greene's classic The Power and the Glory narrates a dramatic example from Mexico; Stephen Macedo's suggestion that in the interest of democracy government ought to promote “wishy-washy” religion24

See Stephen Macedo, Transformative Constitutionalism and the Case of Religion: Defending the Moderate Hegemony of Liberalism, 26 POL. THEORY 56, 61, 63 (1998).

while discouraging more fervent faiths is a kinder and gentler instance. Viewed against these alternatives, it appears that “neutrality” itself is a contested and indeed a highly partisan position.

But set aside these “meta” concerns; at the level of actual operation, neutrality in law is equally elusive. To be sure, there are matters that simply do not call for any sort of governmental response. Suppose two Christian denominations disagree about, say, the efficacy of infant baptism. It seems possible—and highly desirable—for government to stay benignly detached from that controversy. But other matters have a “for me or against me” aspect. Suppose, for example, that a religious pacifist argues that he should be exempted from a general conscription law; a government that invokes “neutrality” and declines to consider the claim will in fact have rejected it. In addition, government may conceive of its own business in a way that makes nonpartisanship and noninvolvement impossible on an operational level.

For example, if government does not take it to be a public function to provide education, or if the public education system does not choose to teach anything about biology or human origins, then government might manage to avoid taking any side in the long-standing disagreements between the evolution theory and creationism. Conversely, if government does establish schools that address such subjects, it becomes well-nigh impossible for government to remain neutral with respect to that controversy. The state may say that because government must be neutral in the matter it will teach evolution and not creationism—indeed, that is basically what the Supreme Court has said—but this assertion is approximately as plausible as a president's claim that because he favors a nonpartisan judiciary he will appoint only pro-life conservative Republicans to the bench.25

See SMITH, FOREORDAINED FAILURE, supra note 7, at 77–97. Supporters may try to defend this position by arguing that even though the teaching of evolution admittedly has the effect of contradicting and hence opposing biblical literalist views of creation, that is not the motive or purpose of such teaching. Schools teach evolution, presumably, not because they are trying to stamp out Christian fundamentalism but simply because evolution is supported by the available scientific evidence. But this description, even assuming its complete accuracy, does nothing to rescue a curriculum teaching evolution and not creationism from the suspicion that it fails to achieve religious neutrality either in its effects or in the purposes and considerations that generate the policy. After all, as Stephen Carter has argued, the assumption that the school curriculum should be based on what scientific evidence indicates rather than what the Bible (literally) teaches, however sensible, plainly rejects one method of ascertaining truth (and of structuring the curriculum) in favor of a different method. Stephen L. Carter, The Resurrection of Religious Freedom? 107 HARV. L. REV. 118, 140 (1993).

These kinds of issues present what William James in a famous essay described as a “forced option”:26

William James, The Will to Believe, in THE WILL TO BELIEVE AND OTHER ESSAYS IN POPULAR PHILOSOPHY AND HUMAN IMMORTALITY 1, 3, 11 (Dover, ed., 1956).

to refuse to take sides is effectively to take sides. Suppose the evangelist says, “You must believe in Jesus or lose your salvation,” and the agnostic responds, “You might be right. Far be it from me to disagree. But I prefer to suspend judgment—to remain neutral in the matter.” In reality and for all practical purposes, the agnostic is making a judgment and taking a position: he is rejecting the evangelist's appeal (and, if the evangelist is right, forfeiting his salvation). So if he imagines he is actually avoiding a choice, James's essay suggests, he is merely fooling himself. To a large extent, modern pretensions of neutrality reflect a similar self-deception, and modern theorizing about neutrality serves to provide the murkification that permits such deception to flourish.

Sometimes the fog briefly lifts, and the spectacle becomes faintly embarrassing. Observe Edward Foley's attempt to work out the implications of John Rawls's liberalism for religious questions. Foley relates how Rawls shows that in the modern pluralistic world the state must be neutral toward religion; and as a devotee of Rawls, Foley wholeheartedly agrees. He notices but somehow is not troubled by the corollary that the state cannot be neutral toward “illiberal” religions.27

Edward P. Foley, Political Liberalism and Establishment Clause Jurisprudence, 43 CASE W. RES. L. REV. 963, 965–966, 973–974 (1993).

But then Foley goes on to worry that the liberal state will necessarily reject a whole array of familiar religious propositions.28

Id. at 973–978.

Quoting a passage in which Rawls denies that political liberalism reflects a judgment about the truth of even “unreasonable” religious views, Foley remarks: “I confess I am not altogether sure about the meaning of this passage. It seems to me that liberalism does reject as false, as well as unreasonable, the idea that the state must endorse a particular creed if the people are to avoid eternal damnation.”29

Id. at 975 n. 40. What Foley may fail to appreciate is Rawls's attempt to separate political philosophy from questions and claims about “truth” altogether. But that strategy produces serious problems of its own. For critical discussion, see Jody S. Kraus, Political Liberalism and Truth, 5 LEGAL THEORY 45 (1999).

There is something almost plaintive about Foley's confession, but it is in reality merely an atypically self-conscious instance of the predicament of “neutrality” theory (and law) generally. Some claims by their very nature authorize a sort of dissenter's veto; they render sincere objections self-validating. “Our decision is unanimous” or “This policy is desirable because it is acceptable to everyone” are cases in point; a single “I disagree” is alone enough to defeat such claims. In the same way, the modern discourse of religious freedom presents an unseemly display of judges and theorists insisting that some law or public program is “neutral” toward religion in the face of vociferous objections that “It is antithetical to my religion.” The judicial and academic partisans of neutrality in effect respond, over and over, “You may think this policy is inconsistent with your religious faith, but you're wrong; we tell you it's neutral!” The refrain is not so much theory as effrontery, and it calls to mind C.D. Broad's comment on an argument he disliked in Kant: “This is of course absolutely indefensible, and charity bids us turn our eyes from the painful spectacle.”30

C.D. BROAD, FIVE TYPES OF ETHICAL THEORY 128 (1959).

“Shared” Grounds: Secularity without Neutrality?

The critique of neutrality suggests that the secular-neutrality position does not satisfy the “wide” version of the constraint—the version that forbids government either to invoke or to reject religious beliefs. But might government, by confining itself to the domain of the secular (whatever that means),31

Both “secular” and “religion” are problematic concepts, but the difficulties cannot be examined in the course of this short essay. For further discussion, see Steven D. Smith, The “Secular,” the “Religious,” and the “Moral: What Are We Talking About? 36 WAKE FOREST L. REV. 487 (2001).

at least satisfy the “narrow” version that enjoins government not to rely on religious premises and instead to act on “shared” grounds? Might there be virtue in a secular government, in other words, even conceding that it is not neutral toward religion?

Suppose, for example, that all citizens in a given community agree that government may act on the basis of economic considerations, and some but not other citizens also believe (in accordance with their religious faith) that government may or even must act on the basis of theological considerations. A government that limited itself to economic considerations would not be “neutral”; it would have rejected the religious citizens' views at a “meta” level and it would likely reject their more specific conclusions at an operational level as well (if, for instance, taking theology into account would sometimes lead to conclusions inconsistent with those reached on purely economic grounds—a “just price” or “just wage” policy, perhaps). Even so, could we not plausibly say that this single-mindedly economic government, though not religiously neutral, acts on the basis of shared grounds? After all, everyone does agree that government may act on economic grounds, which is just what government is doing.

Advocates of secular government may prefer not to present their position in quite this way, because the position is more appealing if secularity is held out as being benignly neutral toward religion. But it is arguable that the numerous proponents of secular government or of secular “public reason” have something like this position in mind. And indeed, a more open embrace of a “shared though nonneutral secularity” might rescue jurists and theorists from the apparent effrontery, noted a moment ago, of appearing to insist that a policy is “neutral” toward religion in the face of the believers' protest that it is not.

But “nonneutral secularity” provokes its own objections. In the first place, even disregarding the occasional otherworldly ascetic who thinks that all secular values should be shunned and hence assuming that practically all citizens accept some set of “secular” grounds, it is doubtful whether a government that acts only on such grounds can plausibly be described as acting on “shared” grounds. The characterization rests on a description of the religious citizens as accepting two severable propositions: (1) government may act on secular grounds, and (2) government may act on religious grounds. Nonreligious citizens reject proposition (2), but they accept (1), which can hence be said to be “shared.” In reality, though, at least some religious citizens probably embrace something more like a single, complex proposition: “Government may act on both secular and religious grounds,” or perhaps “Government may act on secular grounds so long as religious grounds are also taken into account.”32

See Eberle, supra note 22, at 145.

The nonreligious citizens presumably reject this complex proposition; so there is no proposition in common—nothing that is truly shared.33

For elaboration of the point, see Steven D. Smith, Separation and the “Secular”: Reconstructing the Disestablishment Decision, 67 TEX. L. REV. 955, 1009–1010 (1989):

In more familiar contexts, we would immediately spot the common denominator strategy as fraudulent. Suppose Dad and Daughter are discussing what to have for dinner. Daughter proposes: “Let's just have dessert.” Dad suggests that it would be better to have a full meal, with salad, meat, fruit, cooked vegetables, and then dessert. Daughter responds: “Obviously, Dad, we disagree about a lot of things. But there is one thing we agree on; we both want dessert. Clearly the fair and democratic solution is to accept what we agree on. So let's just have dessert.” Although he might admire Daughter's cleverness, Dad is not likely to be taken in by this common denominator ploy. The argument that secular public discourse provides a common denominator that all citizens share is comparably clever—and equally unpersuasive.

Id. at 1010.

But even if the secular grounds can plausibly be characterized as “shared,” the more essential problem is simply that a nonneutral secular government would not avoid the problems that generated the constraint in the first place. By rejecting much of what is most important to the kind of religious citizens we have been discussing, and by allowing government to be responsive to all of the commitments of secular citizens but only some of the (lower-priority) commitments of these religious citizens, the secular government would still manifest if not outright disrespect then at least something less than equal respect for the religious citizens. Indeed, insofar as some religious citizens cannot address some issues (abortion is the common example) without relying on their religious convictions, a secular public sphere would effectively exclude such citizens from participation in public deliberation and decisions regarding such matters. Or, which amounts to the same thing, it would force them to dissemble or effectively to adopt assumed identities for public purposes.34

This objection is pressed in MICHAEL J. PERRY, UNDER GOD? RELIGIOUS FAITH AND LIBERAL DEMOCRACY 32–33 (2003); Michael W. McConnell, Five Reasons to Reject the Claim that Religious Arguments Should Be Excluded from Democratic Deliberation, 1999 UTAH L. REV. 639, 655–656.

Proponents of secular government or of secular “public reason” typically try to rebut these criticisms by arguing that secular government at least offers all citizens the possibility of a community that is “fair,” or that operates on grounds that all citizens can accept “in principle,” or that all citizens could accept if they were “reasonable.” “Reasonable,” in this usage, is understood to refer not so much either to the substantive content of a belief or to an epistemic virtue but rather to a quality of civility; a “reasonable” citizen is one who is willing to live on terms that are mutually acceptable in a pluralistic community.35

See, e.g., Rawls, supra note 14, at 48–54.

A different way to put the position is that a “reasonable” citizen accepts a principle of “reciprocity.” So religious beliefs are excluded from some part of the public domain not because they are wrong—such beliefs, the theorists of public reason cheerfully concede, might in fact be true—but because they are not shared and cannot be a reciprocal basis of community. If I would not be willing to have you impose your religious values on me, then “reasonableness” and reciprocity demand that I refrain from imposing my religious values on you. Right?

Well, not quite. Once “reasonableness” is defined not in substantive or epistemic terms but rather in terms of civility and the practical demands of community, then the content of “reasonableness” is no longer something that can be decided from within the confines of the professor's office; it becomes, rather, an empirical and sociological matter. What do the citizens in a given community by and large actually view as an acceptable basis for their civic enterprise? On what grounds are they in fact able and willing to live together? There is no way to deduce answers to such questions from abstractions like community, citizenship, “reasonableness,” or reciprocity. Citizens in one community might think that public celebration of the lordship of Zeus is essential to the polis; a city that refused to acknowledge this (to them) supremely important fact would be unworthy of anyone's allegiance. And the occasional citizen who raises a ruckus about what nearly everyone else regards as necessary would be “unreasonable”—using the term, of course, as one of sociability. A different community might think exactly the opposite and hence might regard the polytheocratic zealot who presses for public acknowledgment of Zeus and company as an “unreasonable” troublemaker. And so he would be—in that community and with reference to the values of civility and community.

In short, there is no way for mere theory to generate the content of a civility-oriented “reasonableness”; it all depends, as they say, on “the facts.” Nor does the notion of “reciprocity” alter this conclusion (except, perhaps, through rhetorically powerful obfuscation). Reciprocity (much like “equality” in Peter Westen's well-known analysis)36

Peter Westen, The Empty Idea of Equality, 95 HARV. L. REV. 537 (1982).

is a formal concept; it suggests that if I want you to be subject to criteria or principles X, Y, and Z, then I ought to agree to be subject to the same criteria or principles. No “X-for-you” but “not-X-for-me.” But the bare notion of reciprocity cannot provide the substantive content of X. X might be “majority rule,” or “Rawlsian justice,” or “Benthamite utilitarianism,” or “the true Catholic faith.” Whatever a community adopts as its governing criteria or fundamental commitments, those criteria or commitments can with complete consistency be reciprocally respected. And conversely, an advocate can, by simply framing the question to his advantage, describe any sort of position as a violation of reciprocity. (“If you would object to my imposing my (fascist) political philosophy on you, then reciprocity demands that you refrain from imposing your (liberal democratic) political philosophy on me.”) Reciprocity, in short, can do no work here—no legitimate work, at least.

Dumb Religion? Dumbed-Down Discourse?

One sophisticated variation that arguably escapes these objections (and that hence warrants separate notice) has been developed by Kent Greenawalt. In early work37

See, e.g., KENT GREENAWALT, RELIGIOUS CONVICTIONS AND POLITICAL CHOICE (1988).

Greenawalt powerfully criticized the claim that public decisions should be based only on “rational” or “publicly accessible” grounds—a category usually thought to exclude religious beliefs. Greenawalt argued that some public decisions (such as decisions about abortion or animal rights) cannot be made on such grounds, so nonrational grounds will necessarily be consulted, and in that case there is no reason to discriminate against religious grounds. In addition, Greenawalt argued that any such constraint would effectively exclude some thoroughly religious citizens from full participation in the democratic process.

In a later book38

KENT GREENAWALT, PRIVATE CONSCIENCES AND PUBLIC REASONS (1995).

Greenawalt revisited the issues. He emphasized, as argued above, that the proper grounds for democratic decisions cannot be deduced from abstract theory; the problem is a sociological one, and hence what nurtures community and mutual respect in one society may be disruptive or offensive in a different kind of society. In late-twentieth-century pluralistic America, however, Greenawalt believed the use of religious grounds for public decisions to be problematic. At the same time, forbidding the use of such grounds would pose the problems noted in Greenawalt's earlier work. So he sought a compromise position or middle ground. But how?

Greenawalt's answer is complex, but perhaps the most controversial aspect of his middle-ground proposal suggests a distinction between the grounds that citizens and officials may properly rely on in making public decisions and those they may properly present in their public justifications.39

Id. at 134–140.

At least in some contexts, he suggested, an official might rely on a religious belief in making her decision, but her public explanation of the decision should be couched in more secular and perhaps generic terms. Not “the Bible teaches that we should protect endangered species,” but rather “it is morally right and in the public interest to protect endangered species.” In this way, Greenawalt hopes, religious believers will not be excluded from politics but the alienation sometimes associated with public invocation of religion may be avoided.

One criticism asserts that Greenawalt's middle ground authorizes dishonesty. But this objection is misconceived: Greenawalt carefully explains that although his proposal contemplates less than full disclosure of the grounds of decision, it does not excuse false accounts of those grounds. To be sure, religious citizens or officials might still be hindered to some extent from wholehearted participation because they will be restrained from unqualified articulation and advocacy of their full views. But in what is self-consciously intended as a compromise position, this much restraint might be viewed as an acceptable price to pay to enhance the civility of public debate.

Perhaps the more serious objection to Greenawalt's middle ground suggests that it would further enfeeble a public discourse that is already, as someone put it, “deeply, deeply shallow.” Public justifications limited to generic claims that such-and-such is or is not in the “public interest” may be inoffensive, but such justifications also give other participants in public debate little understanding of the real reasons that motivate other actors and little substance to consider and perhaps respond to.40

See Perry, supra note 35, at 38–44.

Greenawalt arguably misjudges the more serious deficiencies of current public discourse, which is already too thin, not too richly substantive. Indeed, an observer who has watched a recent election campaign might argue that current discourse combines the worst fears of all sides of the debate; the discourse, one might plausibly conclude, tends to be substantively unrevealing and empty but nonetheless personally offensive, sometimes vicious. In this context, imposing further constraints on the substantive content of the discourse may seem a miscalculation.41

For more extended arguments to this effect, see Douglas Laycock, Freedom of Speech that Is Both Political and Religious, 29 U. C. DAVIS L. REV. 793 (1996); Jeremy Waldron, Religious Contributions in Public Deliberation, 30 U. SAN DIEGO L. REV. 817 (1993).

In this respect, Greenawalt's proposal serves to illustrate a concern raised more generally about the liberalism in which “public reason” proposals are grounded. The constraint disfavoring the public use of contested ultimates reflects a sort of bargain—or perhaps a wager. In a pluralistic society, the constraint assumes, matters of ultimate truth are likely to be controversial, so we should steer public discourse away from them. In effect we try to purchase greater consensus and avoid alienation by making public discourse more innocuous. Whether this exchange will actually realize its intended benefits is doubtful; we might simply be exchanging a discourse based on divisive ultimates for a discourse based on equally divisive nonultimates—and getting an impoverished public conversation in the bargain. In this vein, Ronald Beiner argues that modern liberal theory, with its commitment to an agnostic neutrality, has produced a “reluctance to engage with the kind of large and ambitious claims about human nature and the essence of our social situation that alone furnish a critical foothold for bedrock judgments about the global adequacy or deficiency of a given mode or life.”42

RONALD BEINER, PHILOSOPHY IN A TIME OF LOST SPIRIT 55 (1997).

Taking Pluralism Seriously

Beyond the particular objections and rebuttals, the neutrality and secularity approaches suffer from a central (and ironic) failure: they fail to take their own premise as seriously as it deserves. Starting with an emphasis on pluralism, ironically, the theorists favoring these positions in effect try to wish pluralism away. They assume that underneath the pluralism at one level there is an underlying agreement at another level—on some “neutral” default position, or some method of deliberation, or some set of basic principles—upon which citizens can converge in forming a community. But when the contents of this ostensible “overlapping consensus” are articulated, they invariably turn out to be unacceptable to some—or perhaps to many, or even most—actual citizens.43

Cf. Frank Michelman, Morality, Identity, and “Constitutional Patriotism,” 76 DENVER. U. L. REV. 1009, 1023 (1999):

It is not clear how we can say that a constitutional norm such as “equality of concern and respect” remains invariant—remains one and the same norm—under reasonably contesting major interpretations of it (“color-blindness” versus “anti-caste”). And that threatens disaster to the proposed constitutional contractarian justification of politics. For, obviously, the justification cannot succeed if it turns out that the constitutional “principles and ideals” to which everyone, as reasonable, hypothetically agrees are just forms of words papering over unresolved and deeply divisive political-moral disagreements among the reasonable.

We might put this point in a different way. The standard story about pluralism and the constraint on contested ultimates suggests that modern liberal government is essentially different from premodern government. Premodern government operated on the basis of some assumed orthodoxy. Modern liberal government, by contrast, respects pluralism and hence eschews any such orthodoxy; rather, it remains agnostic regarding “the good” and the sorts of questions addressed by religion. The criticisms considered in this section contest this self-description. Modern liberal governments, by this view, of necessity operate from some orthodoxy, just as premodern government did. The real difference in this respect is that modern governments attempt (with the help of legal doctrines and theories) to disguise the fact.

Perhaps modern pluralism necessitates some such deceit. Contemporary liberal political theorizing evinces an almost desperate effort to deflect, by whatever means necessary, an obvious but fearsome challenge that might be phrased something like this: If “legitimate” government must be based on “the consent of the governed,” as the Declaration of Independence says, and if there is in fact no proposition or truth to which all or nearly all citizens consent, then there is no proposition or truth upon which legitimate government can be based. And it may seem to follow that the alternatives are…anarchy (or at least the theoretical endorsement of anarchy) or…what? Denial and obfuscation, perhaps? If these are our choices, then it is hardly surprising if theorists devote themselves to sophisticated projects in obfuscation—the more sophisticated, the more successfully obfuscating.

But are those the only choices?

DEFYING THE CONSTRAINT: FOUNDATIONAL THEORIZING

One possibility would be to rethink and perhaps reject the constraint—and hence to base theories of religious freedom (and of government generally, perhaps) on foundations that the theorist takes to be true (even though contested). In this vein, attempting to break away from the endless recycling of neutrality and secularity themes, a few theorists have explored the possibility of overtly basing religious freedom on religious foundations or at least on contested truth claims or orthodoxies. This road is currently less traveled by than the multilane highway of secular neutrality and “public reason,” and it is less clear where the road might eventually lead. Its modern travelers have concerned themselves more with free exercise than with nonestablishment.44

There is no reason why religious rationales cannot be given for nonestablishment commitments. Early-modern advocates of nonestablishment often emphasized the importance of purity of the Church, and this rationale is still occasionally invoked.

But we can notice some of the principal efforts.

A Religious Foundation for Religious Freedom?

Perhaps the most prominent attempt by a legal scholar to develop an account of religious freedom on self-consciously religious foundations has been made by John Garvey. Garvey's discussion begins by criticizing familiar rationales for free exercise that attempt to honor the constraint—or, as he puts it, that adopt “the agnostic viewpoint.” Hence the rationale that argues for freedom of religion as an expression of personal autonomy fails to identify anything distinctive about religious beliefs or choices warranting special respect; moreover, the rationale itself violates the constraint by implicitly adopting a view of the person (as constituted at the deepest level by a “free-floating self”) that many religious believers reject. And the rationale that argues for religious freedom as a means of preventing political strife fails to explain why that end should not be achieved, in the case of politically ineffectual groups, by repression.45

JOHN H. GARVEY, WHAT ARE FREEDOMS FOR? 42–49, 56–57 (1996). These objections hardly exhaust the criticisms of rationales based on autonomy and civil peace. In this essay, however, I have not focused on standard rationales for religious freedom based on this and similar generic values largely because I do not think discussion of such rationales has been the most interesting or important theorizing about religious freedom in recent years. And the most serious theorizing on these subjects has tended to criticize these generic value rationales as inadequate. For a discussion of the major rationales and objections, see SMITH, FOREORDAINED FAILURE, supra note 7, at 99–117.

So if there is a plausible case for religious freedom, Garvey concludes, it will be based (in defiance of the constraint) on religious beliefs—on “the believer's viewpoint rather than the agnostic's viewpoint”—and hence on “reasons that only some people accept.”46

GARVEY, supra note 46, at 54.

So is there a plausible religious rationale for religious freedom? Garvey thinks so; indeed, he sketches four religiously based rationales. Two of these rationales—that coercion of religious belief is futile and unacceptable to God, and that freedom allows for the progressive revelation of God's truth—would operate to protect believers and nonbelievers alike against coercion in religion. The other two—asserting in essence that believers owe duties to God and that violation of these duties will lead to extratemporal suffering—are more “lopsided” (as Garvey puts it) in their implications; they seek to justify “free-exercise exemptions” excusing religious believers from complying with generally applicable laws. Garvey acknowledges that these reasons will be unpersuasive to nonbelievers, but he argues that they are the best available rationales for existing constitutional commitments to religious freedom.47

Id. at 46–57.

In a trenchant criticism of Garvey's analysis, Larry Alexander agrees that the familiar “agnostic” or “neutral” arguments for religious freedom are unpersuasive; hence a satisfactory argument would necessarily be “sectarian.”48

Larry Alexander, Good God, Garvey! The Inevitability and Impossibility of a Religious Justification of Free Exercise Exemptions, 47 DRAKE L. REV. 35 (1998).

But sectarian rationales like Garvey's turn out to be equally unsatisfactory. The problem with sectarian justifications, Alexander maintains, is not just that nonbelievers will find them unpersuasive; the deeper problem is that they will justify religious freedom only for people who embrace what the theorist or the state regards as true religion. Alexander gives the example of Christian Scientists, who believe that God forbids them to get medical help for a gravely sick child. They may believe that they have a divinely ordained duty to forego medical attention and that extratemporal sanctions will follow from the violation of that duty; however, not only agnostics but also religious believers who do not share the commitment to faith healing will conclude that these parents, however sincere, are simply wrong. Hence Garvey's divine duty and extratemporal-suffering rationales will not justify protection for the Christian Scientists.49

Id. at 40–41.

More generally, “[r]eligious believers do not view compliance with imagined duties as a good. Rather, they view compliance with actual duties as a good.” Consequently, “a religious argument for religious freedom will account for only that amount of freedom consistent with the religion's account of Truth. And that amount of religious freedom may be quite small indeed.”50

Id. at 40, 43.

The Centrality of Conscience?

Is Alexander's objection answerable? Consider closely the often-quoted words of Madison's Memorial and Remonstrance: “It is the duty of every man to render to the Creator such homage, and such only, as he believes to be acceptable to him.” Read one way, the assertion hints at a conception of religion that might avoid the objection. It is true, we might reply to Alexander, that freedom of religion protects the exercise only of true religion; but it turns out, happily, that true religion for any given person is just what that person thinks it is. So everyone's religion is protected after all.

Though this conception of religion might avoid Alexander's objection, as just presented the conception also seems untenable. Who actually believes that truth in religion is whatever anyone thinks it is? Or, rather, who can coherently believe this?51

The qualification is necessary because it seems that many religionists today do at least think they believe something that resembles this aggressively latitudinarian position. See, generally, ALAN WOLFE, THE TRANSFORMATION OF AMERICAN RELIGION 67–95 (2003).

And the formulation quickly leads to embarrassment. You ask: “What should I believe about religion?” I answer: “Anything you sincerely believe will be true—for you.” But this doesn't help. “Wonderful!” you respond. “So I can't go wrong. Except … that's only if I actually believe, right? And I still don't know what to believe.”

Taken for all it is worth (and perhaps a bit more), in short, Madison's claim seems untenable. But the claim points to something that many have accepted: the importance of “conscience.” Noah Feldman argues persuasively that a commitment to “freedom of conscience” was the central theme in the Anglo-American tradition of religious freedom from the time of Locke until well into the twentieth century,52

Noah Feldman, The Intellectual Origins of the Establishment Clause, 77 NYU L. REV. 346 (2002).

and even if it has lost its primacy, the commitment is still widely honored. The appeal of conceiving of religious freedom in terms of “conscience” in a pluralistic society is evident: “conscience” makes room for respecting different views of what religion and morality require.

But if pluralism makes freedom of conscience especially attractive, pluralism also tends over time to erode the value of conscience or even to debase the concept. Reflection on the concept can show how this deterioration is likely to happen.

Consider the proposition: “You should always do what you conscientiously think is right.” In one sense, this is merely a truistic proposition—like “You should believe what you really think is true” or “You should bet on the horse that you actually think will win.” Thus understood, conscience is unobjectionable but can claim no special dignity or respect. Invoked in support of anything you or I disapprove of, the plea “I did it from conscience” will elicit the same response that “I sincerely believed ‘Lucky Number’ would win the Derby” provokes: “Too bad. You were wrong.” Conscience can claim some special respect, it seems, only if we think there actually is some virtue or dignity that attaches to doing something because someone believes it is right—even if he or she is mistaken.

But why might that be so? A particular religion or theology might provide a rationale. I have argued elsewhere that in an “ultra-Protestant” theology, Martin Luther's notions of “justification by faith” and “imputed righteousness” might be extended to the creedal component of religion;53

STEVEN D. SMITH, GETTING OVER EQUALITY: A CRITICAL DIAGNOSIS OF RELIGIOUS FREEDOM IN AMERICA 164–175 (2001).

in this way a believer might be able to conclude that even a religion she regards as false may nonetheless be “counted as” true (by a benevolent deity) for those who conscientiously believe and practice it. But this is a religious rationale for respecting conscience (as in fact the classical account of conscience was).54

See Feldman, supra note 53, at 357–372, 424–446.

It is harder to see how a purely secular philosophy could support the same move.

The crucial point, in any case, is that the value of conscience is not just self-evident and freestanding; conscience is something that will be highly esteemed by some encompassing philosophies or “comprehensive doctrines” but not by others. And that observation allows us to appreciate how, although conscience might support pluralism, pluralism might in turn undermine the commitment to conscience.

Suppose the “comprehensive doctrines” that prevail in a society at some point do in fact value conscience. And respect for conscience serves to protect a variety of beliefs—including some beliefs that are incompatible with the comprehensive doctrine or doctrines in which respect for conscience is grounded. As these newer doctrines grow and proliferate, so that the original doctrines become increasingly contested rather than generally accepted, the foundation upon which respect for conscience is based is likely to become increasingly insecure. Even if some commitment to “conscience” persists (perhaps through inertia or tradition), the result is likely to be confusion—as to both why conscience is important and what the meaning and scope of “conscience” are.

Such a declension is apparent in the current rhetoric of conscience. Noah Feldman explains:

[T]he modern understanding of liberty of conscience seems to be that every person is entitled not to be coerced into performing actions or subscribing to beliefs that violate his most deeply held principles. This definition differs fundamentally from that of the eighteenth century in that it is secular. To the eighteenth-century mind, liberty of conscience meant that the individual must not be coerced into performing religious actions or subscribing to religious beliefs that he believed were sinful in the eyes of God and that could therefore endanger his salvation. Indeed, it was, following Locke, literally “absurd, to speak of allowing Atheists Liberty of Conscience,” because conscience necessarily related to one's salvation, in which atheists presumably disbelieved altogether. Because this view seems implausible today, liberty of conscience may require some justification other than the religious justification that underlay the eighteenth-century version of the theory.55

Id. at 424–25. I note, however, that from the “ultra-Protestant” perspective mentioned above, what Locke viewed as “absurd” is not so obviously illogical. To say that salvation is irrelevant to someone who does not believe in it would be no more logical than saying that germs are irrelevant to people who do not believe in them. Hence though atheism might not support the justification for respecting conscience, it would not necessarily follow that atheists would not be protected by it. See SMITH, EQUALITY, supra note 54, at 174.

Elaborating on the justification problem and its implications, Feldman observes that if we:

broaden conscience to include secular matters of deep belief, … the Lockean distinction between the sphere of the church and that of the state evaporates. Suddenly there is no clear rationale for allowing government to take any action of any kind where it violates conscience; or alternatively, all attempts to protect conscience look unjustifiable.56

Feldman, supra note 53, at 426.

Providing evidence for Feldman's assessment, Ronald Beiner suggests that a book on the subject by David Richards demeans the concept of conscience:

The spuriousness of this recurrent appeal to the sacredness of conscience is very clearly displayed in the discussion of pornography. How can this possibly be a matter of conscience? What is at issue here, surely, is the sacredness of consumer preferences.

And Beiner goes on to scoff that “[b]y [Richards's] contorted reasoning, the decision to snort cocaine constitutes an act of conscience.”57

BEINER, supra note 43, at 30.

The difficulties in current conceptions of conscience are apparent in one of the leading treatments of the subject by legal scholars. Christopher Eisgruber's and Larry Sager's analysis in a sense travels a course just the opposite of Garvey's. They begin by arguing at some length that religious rationales for religious freedom or legal doctrines that privilege religion over nonreligion are unduly “sectarian” and hence unacceptable in a pluralistic democracy. So any defense of religious freedom will need to be “non-sectarian,” both in its premises and its applications. Eisgruber's and Sager's own position emphasizes the importance not of privileging conscience as something especially valuable but of protecting it against discrimination to which it is especially vulnerable.58

Christopher L. Eisgruber & Lawrence G. Sager, The Vulnerability of Conscience: The Constitutional Basis of Protecting Religious Conduct, 61 U. CHI. L. REV. 1245, 1250–1270 (1994).

But what is conscience anyway? And if it is not anything particularly valuable, then why should the possibility of discrimination against conscience be so troublesome? Eisgruber's and Sager's commitment to a “nonsectarian” approach severely limits the answers they can give to such questions. They cannot allow the law to value and protect religious but not secular versions of conscience; hence the historical sense of conscience, as noted above by Feldman, is inadmissible. Instead, “conscience” is reduced to a person's “deep concerns”—whatever the source or content of those concerns.59

Id. at 1283, 1285.

“Deep” presumably has at least a positive resonance—most of us would rather be “deep” than “shallow”—but elsewhere the valences are inverted; hence Eisgruber and Sager argue that ostensible religious obligations “in some respects resemble physical disabilities. Both religious obligation and physical disability may make it hard for individuals to comply with otherwise neutral laws.”60

Id. at 1267.

And this incapacity is part of the reason for protecting conscience even though it is not positively valuable:

Nonsectarian judgment comfortably supports the conclusion that for some people under some circumstances the demands of religious belief are “special and important” in the same sense that disabilities are “special and important”: both can have profound effects upon individual well-being. But it does not follow that either religious beliefs or disabilities are “intrinsically valuable.”61

Id.

So “conscience” turns out to be a sort of disability with which some people are afflicted—something akin to blindness or deafness—and it is protected against discrimination for much the same reasons that we prohibit discrimination against disabilities. This conception, arrived at under the acknowledged pressure of pluralism and the constraint, has plainly traveled a long way from the traditional understanding of conscience as something sacred—as the “free response [of] the individual called distinctively by the Divine within.”62

MARIE FAILINGER, WONDERING AFTER BABEL, IN LAW AND RELIGION 94 (Rex J. Adhar, ed., 2000).

Toleration

In Locke's writing, the themes of conscience and toleration are closely interwoven, and the preceding discussion suggests that positive respect for conscience is inherently a doctrine of toleration in its core sense. A tolerant individual or government, that is, is one committed to some base position or orthodoxy that prescribes that even (some) beliefs and practices that are wrong—those based on “conscience” perhaps—nonetheless should be permitted.

In recent years, toleration has received increasing attention from legal and political theorists.63

See, e.g., Steven D. Smith, Toleration and Liberal Commitments, in TOLERATION AND ITS LIMITS, NOMOS (2004); TOLERATION: AN ELUSIVE VIRTUE (David Heyd, ed., 1996).

Within that discussion, two debates are especially relevant to our present subject. The first is the debate between what we can call “liberal tolerance” and “liberal neutrality.” Is a position of tolerance truly “liberal” or is it objectionable because it assumes that some position or orthodoxy is the “official” position of the political community and that those who disagree with this orthodoxy are merely “tolerated”? Second, can a position of toleration be justified on “universal” and abstractly theoretical grounds, or is tolerance of necessity a highly contextual stance based on more local and largely prudential considerations?

It is impossible to rehearse these debates here. But the foregoing discussion already suggests that despite its claims to being more consistent with pluralism and the constraint, liberal neutrality will in the end amount to an exercise in self-deception. Hence if liberal values (especially including religious freedom) are to be respected at all, they will be maintained from a position of toleration. The discussion also suggests that at least some of the ostensibly “universal” rationales sometimes offered for toleration—“reciprocity” is perhaps the foremost—will upon examination prove to be empty and question-begging.64

I elaborate on these points at greater length in Smith, supra note 63 at xx (forthcoming).

More generally, “toleration” is an orientation that by its nature must be understood and justified relative to some prevailing base position or “orthodoxy.” That fact would seem to rule out in advance any “universal” justification for “toleration” per se: the most we might hope for would be a convincing and “universal” argument for some particular orthodoxy—Christianity, perhaps, or maybe a strong version of Kantian autonomy—which would in turn justify a derivative but “universal” commitment to toleration. In a pluralistic society like ours, the prospects for any such two-stage demonstration seem feeble at best.

This is emphatically not to say that more local or perhaps ad hoc arguments based on widely accepted values and views and traditions are futile. I myself have tried repeatedly to make the case for toleration in these terms. But any general, all-purpose theory of toleration is unlikely to be forthcoming anytime soon.

CONCLUSION: KEEPING “THEORY” IMPLICIT?

Given the difficulties of theorizing about religious freedom, it is not surprising that most jurists and scholars have tried to avoid theorizing on the subject altogether. In comparison with some areas of constitutional law, as noted, the law of religious freedom has more often been treated, both by judges and by legal scholars, as a product of history—history manifesting itself in legal decisions expressed in the constitutional text, perhaps, or in an ongoing and evolving constitutional tradition. I have argued above that history affords no refuge from the need to theorize. For better or worse, though, it seems that history can provide a way of keeping our theorizing largely implicit—below the level of public presentation and often, no doubt, below the level of conscious thought. More generally, theory always involves both implicit and explicit dimensions; but the proportions vary from person to person and field to field. And it is fair to say that for religious freedom, the imbalance in favor of implicit over explicit theorizing has been—and will likely continue to be—relatively greater than in some other fields of law.

The self-conscious theorizer may find this situation unsatisfactory. Does not the progress of human understanding consist largely of bringing to light and deliberately scrutinizing what had been merely presupposed or accepted out of habit? Perhaps. But then again, the theorizer's preference for making presuppositions explicit might itself be a prejudice that flourishes best when not closely examined. In any case, it is not clear that in the area of religious freedom modern efforts to make principles more explicit have resulted in better law or better understanding—or even in enhanced self-understanding. Recent efforts to maintain the currently fashionable notion that government must not “endorse” religion without repudiating a long tradition profuse with such endorsements—not only in rituals like the currently controversial Pledge of Allegiance but in cherished political landmarks like Lincoln's Second Inaugural, the Declaration of Independence, and Jefferson's Virginia Statute for Religious Freedom—illustrate the embarrassment.

To be sure, arguments from tradition are like arguments from neutrality in one crucial respect: both fail to harmonize pluralistic divisions under some encompassing principle. But at least tradition need not (though of course it may) deceive or insult us by pretending to have harmonized them. Given our theoretical frustrations, and in a situation of pluralism that raises deep questions of legitimacy, the course of deferring largely to theory-burying tradition65

For a lengthy defense of the traditionalist approach, see Steven D. Smith, Separation as a Tradition, J. Law & Politics 215 (2002).

has its appeal.