Freedom as Noninterference

The new genealogy and foundationalism alike conceive of freedom in terms of noninterference. Religious freedom is constituted by noninterference in religious matters. Freedom is doing as one pleases, and interference violates freedom. Law, therefore, limits freedom—ideally to preserve more freedom. Religious freedom is a matter of balancing freedom lost with freedom gained, typically while treating some losses, those proscribed by rights, as impermissible. Where foundationalism pursues such freedom, the new genealogy regards it impossible and ideological, noting its bias toward individuals over communities. But both agree that this is what freedom is.

Consider a remark from Winnifred Fallers Sullivan's The Impossibility of Religious Freedom. This basic view of freedom seems central to why the book regards religious freedom as “impossible”:

Freedom, here, is noninterference. It is “bounded” when the state imposes restrictions, “controls,” or interferes. Such interference limits freedom, opposes it. Thus, “religious freedom is not absolute” precisely because it is bounded by regulations (IRF, 155). Limitless freedom—to be free in more than “a manner of speaking”—in contrast, would mean the absence of all interference, the capacity to do whatever one pleased.

This vision of freedom also seems apparent in the book's conception of religion: “It is the peculiar nature of religion itself to restrict freedom” (IRF, 155). This restriction is a matter of “devotion to the rules of [some] religion,” “non-negotiable things being demanded of one.”Footnote ¹⁶ Notice the conception of freedom operative here: religions impose rules and non-negotiable demands, therefore and thereby religions restrict freedom.Footnote ¹⁷ Rules necessarily curtail and violate freedom. Freedom, then, consists in noninterference. This is why it is that “to be religious … is to live without a certain amount of freedom” (emphasis in original, IRF, 156). This is why “to be religious is not to be free” (my emphasis). “To be free,” Sullivan remarks, “is today understood to mean to have the ability to choose for oneself in all areas of life without restraint” (IRF, 156). A clearer articulation of freedom as noninterference would be hard to find.Footnote ¹⁸

The new genealogy, for its part, assumes freedom is noninterference and that this represents the ideology of modernity, secularity, and liberalism. The critique is familiar: Freedom dichotomizes belief and practice, making belief the zone of noninterference, claiming all else for itself. It is hyper-individualistic, positing an atomistic, autonomous self for whom communities and commitments degrade freedom. Religious freedom “is articulated, through … liberal rights, as a set of discrete freedoms claimed by autonomous units (individuals or groups) from an assumedly neutral state” (JLR, 361). “Individual conscious and belief as the proper locus of religion” are intrinsic to “the modern conception of religious liberty” (PRF, 143). Indeed, “liberal individualist notion[s make] the modern conception of freedom of [religion] possible” (PRF, 143). Such “individual rights” are one part of “the modern nation-state[’s]” “aim to transform political difference into sameness” (PRF, 143). Most simply, “religious freedom … is built around a particular notion of the ‘free’ … human” (PRF, 51). And this “human of secular liberalism,” “the subject of religious freedom is an autonomous individual defined by his or her freedom to choose to believe or not” (PRF, 51).Footnote ¹⁹ The new genealogy equates freedom with hyper-individualistic liberal autonomy.Footnote ²⁰

If this is what religious freedom means, no wonder the new genealogy rejects it. This noninterference vision grounds criticisms related to freedom and neutrality alike. I consider neutrality-related critiques shortly; those concerning freedom first. These chiefly concern the new genealogy's claim that religious freedom is impossible or incoherent but also bear directly on its other four criticisms. Religious freedom purports to offer freedom, noninterference, but actually involves the state in all sorts of interference. Religious freedom is thus inherently unattainable because always limited. Further, it inevitably involves governmental interference, even if only in the legislation, adjudication, and scrutiny that pursuit of religious freedom involves.

More fundamentally, the necessity of defining religion constitutes profound unfreedom. The question and adjudication of what counts as religion involves ongoing, obvious, and pervasive interference—freedom's antithesis. And, such judgments, the new genealogy says, are anything but neutral: “The right to religious liberty authorizes the state to intervene in what appears to be mere expressions of religious belief but in fact involves the state in making substantive judgments about religion, a domain toward which it claims to be neutral. … [Religious liberty] maintains that religious belief is immune from intervention [that is, free] while at the same time sanctioning its outward expression.”Footnote ²¹ Religious freedom requires that religion be defined in some way but this involves interference, and “Who decides? Who decides what counts as a [religion] deserving of special protection and legal exemption rather than as some other [thing]” (PRF, 51)?Footnote ²² The state decides, of course, and insofar as the state decides, the state interferes, violating freedom. “As Talal Asad has argued, the very acts of … legal systems have constituted [‘religion’], demarcated certain practices and institutions as religious, and recognized specific tokens as religions. Hence both the general nature of ‘religion’ … and the recognition of specific ‘religions’ … are objects of the legal gaze” (PRF, 289). Religious freedom is thus birthed in and perpetuates unfreedom. The state or “the secular is always entangled with the religious” (PRF, 210). Religious freedom, therefore, is necessarily interference and unfreedom all the way down: “religious freedom … necessarily involve a dividing of legal religion from illegal religion—good religion from bad religion … along an ongoing set of dichotomies … individual or communal, private or public, spiritual or material, belief or practice, chosen or given, Protestant or Catholic, Western or Eastern, peaceful or violent, utopian or locative, universal or particular” (PRF, 7). This is religious freedom as ideology, promising one thing; giving the contrary. It is a tool by which religion is “tamed to fit the requirements of law” (JLR, 362).Footnote ²³ It is a regime of oppression and unfreedom. And this is so, in large part, because following its foundationalist opponents, the new genealogy identifies freedom with noninterference.

Neutrality as Ideological, Impossible, and Inevitably Biased

We have already begun to see the new genealogy's second key assumption: a utopian, “God's-eye view” conception of neutrality borrowed from its foundationalist opponents. Religious freedom must arise from and express principles that are “Neutral,” in the sense of being universal, unassailable deliverances of Reason, natural law, freestanding political conceptions, and the like. The new genealogy of religious freedom regards such Neutrality as a fiction, and effectively shows how it implicates all sorts of controversial normative stances. Yet, it imagines that, either singly or jointly, (1) the impossibility of this sort of neutrality along with (2) particular legal cases that display prejudice, show that religious freedom can only be sheer partiality and bias. Collapsing principled evenhandedness into foundationalist Neutrality, the new genealogy mistakenly imagines that Neutrality's bankruptcy means that evenhandedness is impossible or that it is no different from bias or caprice.

The Neutrality assumption animates the new genealogy's critique in two ways. First, religious freedom is not Neutral because it involves pronouncing on substantive matters concerning what constitutes religion and in that and other ways taking sides in religious disputes. Yet such matters are highly contestable, and there is no value-free grounds or “view from nowhere” from whence to resolve them. Why should a court's view count as Reason's deliverance any more than some disputant's view? Strictly every decision unfolds from some contingent position and context, a court's or government's no less than some religionist's. Thus, Neutrality—and so religious liberty—are a sham.

Second, religious freedom not only fails to be Neutral, it is inherently and radically biased in particular ways. Recall: “religious freedom … necessarily involve[s] a dividing of legal … from illegal religion—good religion from bad … along an ongoing set of … dichotomies … individual or communal, private or public, spiritual or material, belief or practice, chosen or given, Protestant or Catholic, Western or Eastern, peaceful or violent, utopian or locative, universal or particular” (my emphasis, PRF 7). The former of each pair here is “good” religion, what religious freedom privileges: belief over practice, individual over community, Christian over non-Christian, soul over body, elite over lived, majority over minority.Footnote ²⁴ If this were so, religious freedom would not simply fail to attain the mirage that is Neutrality, it would fail to be fair or evenhanded in any sense at all, much less any sense worth caring about.

Grant that religious freedom sometimes exhibits such bias.Footnote ²⁵ Yet the new genealogy insists this is all it ever can do: “‘Religious freedom’ … is … partial and contingent, often discriminatory, always partisan” (my emphasis, PRF, 103); “Again and again it passes off an inevitably partisan narrative as a universal one” (PRF, 102); “No secular state is or can be merely neutral or impartial among religions, for the State defines the boundaries within which neutrality must operate” (my emphasis, PRF, 167). Indeed, “the secular state … is a distinctive kind of theological state” (PRF, 333). While considering only scattered cases, it pronounces not merely on those cases but, in modal and essentialist terms, on religious freedom as such: it is necessarily and essentially biased.

There are two discernible grounds for these claims: historical and conceptual. Neither withstands scrutiny.

Consider the historical grounds: the genealogically suspicious reading of some legal case(s). Logically, no analysis of a few cases, however much bias they display, could tell us that religious liberty is essentially or necessarily biased. Imagine we could examine every religious freedom case in history and unanimously agree that all exhibited the biases the new genealogists claim. Even in this religious freedom wasteland, the new genealogy's conclusion would not stand, for it would remain entirely possible both that (a) all of the failures had to do with the actors in question, not the guiding principles or conceptions of religious freedom, and (b) that an even slightly revised vision of religious freedom, let alone a more radically reimagined conception, might generate different results. To approach the conclusion that religious freedom is inherently biased based on empirical evidence or historical analysis, our data set would have to include the attempt to actualize every possible conception of religious freedom and have each of them confronted with every possible case.

This is very far from what the new genealogy of religious freedom has shown. Not only has it surveyed only a handful of extant religious liberty cases, and not only is there little consensus concerning even all of those it has considered, but many insightful and learned scholars believe that religious freedom has sometimes been honored in ways not expressive of the biases that the new genealogy claims inhere in the concept. Further, the cases the new genealogy does consider seem chosen and interpreted with the aim of showing that religious freedom is inherently biased. There is nothing wrong with choosing cases to substantiate one's claims, but it is something else to suppose such cases tell us what religious freedom necessarily or essentially is. Given that when it comes to religious freedom we are imagining how we might order our political communities, deploying strategies and ideals that, at least to some degree, are ours to shape, revise, reimagine, and mold to our interests, I doubt religious freedom is necessarily or essentially anything.

This brings us to its second grounds, the alleged “architecture” or logic of the concept religious freedom (or religion). Due to the concept's alleged origins and the suspect ends it has served, the concept necessarily embeds and entails commitments to problematic principles like “Enlightenment Reason,” “Kantian autonomy,” “panoptical surveillance,” “Christianity as best religion,” “religion as belief,” “religion as dangerous unless privatized.” These (alleged) facts about the concept's origin and history mean its ongoing use necessarily perpetuates the bias and violence present from its inception.Footnote ²⁶ So the new genealogy claims. But this is mistaken.

Take any genealogy you like. The fact that a concept has some unsavory history does not entail that all contemporary usage extends or implicates that history. Here we need to begin with a basic logical point and consider first a mundane example before moving to cases with higher stakes. Many use the “cc” feature in emailing. That practice and concept initially arose in reference to carbon paper. Today, many do not know what carbon paper is or even that cc abbreviates (or once abbreviated) “carbon copy.”Footnote ²⁷ Their use of the concept does not express bias toward carbon paper over against other media, nor anything of the sort. That some concept has a history, even an ethically suspect history, does not by itself tell us that ongoing use of the concept perpetuates that history. The broader logical point is just that: for any concept or practice C, that C arose in some circumstance or has some history does not in itself entail that contemporary deployment of C perpetuates or entangles one in that origin or those circumstances. To think otherwise is a form of the genetic fallacy.

Certainly, some concepts do advance histories of oppression—and those histories may be difficult or nearly impossible to avoid extending, even if their users try. Many racial slurs are like this. The racist who uses the “n-word” or the gay-basher who uses “queer” extends the concept's poisonous legacy by carrying on in ways saliently like certain precedent usages. The user here intends to and does use the concept in ways that stand in continuity with these precedent usages. Yet even these maximally toxic concepts can undergo the transformation we call resignification. They can come to take on new valences that are not only other than those associated with prior usages but even directly opposed to them. Thus, a gay man who proudly describes himself as “queer” is neither allying with gay-bashers nor engaged in self-hate. Though this is just what we would have to think if we fell prey to the genetic fallacy and imagined that the history of some concept entailed that contemporary deployments perpetuated or entangled its user in that history. On the contrary, he is deploying a vitally distinct conception as compared to the bigot—the two conceptions are applied to the same people (namely, homosexuals). They share a common circumstance of application, picking out the same group of people. But the consequences that application licenses, the ways of relating and thinking about that to which it is applied, are profoundly different: the one authorizes and enacts abuse; the other, celebration and pride.Footnote ²⁸ That a term as hateful as “queer” was (and often still is) admits of such profound transformation and such distinct senses in such a short time, suggests something about just how much change concepts can undergo and just how many more-or-less related senses or conceptions may develop.Footnote ²⁹

Suppose a concept's meaning is constituted by its circumstances and consequences of application—the circumstances under which it is appropriate to deploy it and the practical or intellectual consequences that follow upon its use, what its application authorizes. Insofar as one or either of these undergoes change, the concept's meaning changes. If the changes become sufficiently significant, we can identify distinct concepts (or distinct senses/conceptions). So, a single term or concept—like queer, chair, religion, or religious freedom—may have multiple senses/conceptions. Even if one or more of these conceptions does carry forward problematic features of some problematic ancestral use, it hardly follows that all extant usages or senses do. It certainly does not when one is clear what conception one is using and explicitly rejects problematic ones.Footnote ³⁰

While I find the genealogies of religion on which the new genealogy draws altogether unpersuasive in virtue of neglecting major swaths of history and many of modernity's most important social movements (such as abolitionism, women's suffrage, or India's movement for independence), my point is that even if we accept whatever genealogies the new genealogy offers, they hit home only insofar as they avoid the genetic fallacy and, even then, only in regard to conceptions saliently like the problematic ones—that is, like them in virtue of their problematic features. If even concepts as poisonous as slurs can sometimes undergo transformation such that we can speak of different conceptions, how much more concepts like religion or religious freedom, whose legacies, however checkered, are by no means as univocally poisonous.Footnote ³¹ We will consider those concepts in some detail below. What is important to note now, however, is that even if the new genealogy is right about intractable bias relative to some conceptions of religious freedom, it has not considered or shown anything about all conceptions, least of all the holist, dialectical conception I sketch later in this article.

While the new genealogy successfully undermines foundationalist pretensions of Neutrality, it wrongly imagines this critique tells against religious liberty generally. Once we reject this elision of Neutrality and fairness, of unjust bias and principled evenhandedness, and once we recognize this critique as pertaining only to certain corrupt forms and conceptions of religious liberty, we clear space for alternate visions.

Power as Unethical

For the new genealogy, religious liberty is, above all, a site of state power. The revelation that power is and always has been and will be at work in religious liberty is supposed to count against religious liberty. Fundamental to the critique is the assumption that power is itself ethically problematic. State power is exclusively unethical, oppressive, and dominating. As a paradigmatic expression of that power, religious freedom is too.Footnote ³²

Religious freedom, the new genealogy claims, is a tool of oppression that bifurcates religion into a private, interior essence beyond state control, and a bodily domain subject to state power (SAQ, 3–4). Beyond subjugating bodies, the very bifurcation epitomizes state power. While claiming “to be neutral … religious liberty [is] a technology of [the] modern state … deeply implicated in the regulation of religion” (SAQ, 4). Indeed, “the secular state is necessarily involved in regulating … religion and often prescribing substantive content while … claiming to treat … its citizens equally regardless of their religious commitments” (my emphasis, SAQ, 6). Religious liberty is not only interference and thus unfreedom, nor merely profoundly biased, it is an ideological operation of state power—and, thus, inherently oppressive and dominating. The divisions and controls it effects are “a—if not the—distinguishing feature of modern secular power” (BRF, 11). It “essentially allows the state simultaneously to uphold the immunity and sanctity of religious belief even as it regulates the manifestation of these beliefs” (SAQ, 5). It is a “technology of modern governance that ensures the state's sovereign right to regulate all domains of social life,” and “privileges the beliefs, values, and practices of the majority religious tradition in any given polity” (SAQ, 5). So, rather than a possible means to protect the disempowered, religious freedom is a means whereby the many can control the few and the state can manipulate and oppress everyone (SAQ, 6). “A contingent power arrangement of the modern West” (BRF, 59), it is “cover for … suppression” and “central to a politics of local containment … neo-colonialism and imperialism” (JLR, 360, 362) by which the state would “transform … difference into sameness” (PRF, 143). The irony and ideology of religious liberty, then, is that it harms those it claims to help, enforces sameness rather than allowing diversity, and maximizes state power instead of restraining it. Moreover, none of this is a “corruption of the right to religious liberty [but is] internal to the conceptual architecture of the right” (PRF, 147), “internal to the concept of religious liberty itself” (SAQ, 4).

Thus, as an “operation of modern secular power [that] generat[es] new forms of religious prejudice and enhanc[es] old ones” (SAQ, 6), religious liberty not only exacerbates the violence and conflict that plague our world, it actually creates such conflict.Footnote ³³ “Enormously productive and transformative of religious identity, often hardening existing lines of religious difference,” religious liberty effects the very divisions and antipathy it is supposed to heal (SAQ, 7). “Creating new forms of communal polarization,”Footnote ³⁴ it scripts diverse individuals and communities into subjects whose agency and particularity are disappeared by the procrustean religious liberty narrative: “‘Religions’ begin to perceive themselves as they are portrayed … as hidebound communities, static bodies of convention. … Boundaries are settled. Orthodoxy is established. … You can be this or that, but not both” (PRF, 53). This marginalizes those who do not conform and generates and sustains pitched, often bloody, battles. Ironically, religious liberty actually causes so-called religious violence: “far from being an instrument of neutrality that protects religion or its practitioners, the right to religious liberty also helps create new identities, reifies religious difference, furthers state regulation … and … facilitates the hegemony of powerful geopolitical actors” (SAQ, 7). Unmasking this as religious liberty's “performative and productive capacity” is one of new genealogy's “basic arguments” (SAQ, 7).

Consider this argument: There are complex patterns of oppression and violence directed against some socially or politically weaker party that we are tempted to identify as somehow implicating religion. We imagine that legal intervention prohibiting religious persecution and empowering the persecuted will help. Yet the very categories we are tempted to deploy were birthed in and embody oppression. Their use extends and deepens that oppression and distorts the identities of those we meant to protect—marginalizing them further. Religious liberty only makes things worse and births new problems. Best then to abolish religious liberty and laws prohibiting religious discrimination—and better to dissuade so-called religious minorities from understanding themselves in those terms or deploying the tools or rights associated with religious freedom to protect themselves against oppressors or to hold states accountable.

Now, this entire argument could equally well be turned against legal concern with race and protections against racial discrimination. I do not see on what grounds the new genealogy could avoid this entailment: its arguments seem to oblige it to remove race as a protected class or as figuring in law at all.Footnote ³⁵ For race is no less a constructed and contested category than religion. As Jay Carter, Willie Jennings, Cornel West, and others have argued, it was birthed in and extended colonialism, violence, marginalization, and oppression—much more clearly and unambiguously than religion.Footnote ³⁶ Far more obviously, consistently, and unambiguously, race has functioned in law as a tool of oppression and state power, marking out some for privilege, others for subjugation. Arguably, until relatively recently, that has been its primary function. Yet even if we suppose that legal categories of race necessarily involve certain forms of reification, division, and constraint, can we seriously imagine that racial minorities—such as blacks in the United States—would be better off without legal protections against discrimination, oppression, marginalization, exploitation, and violence directed at them in virtue of their race? That the law should not attend to the ways in which patterns of oppression, violence, and discrimination by the state, police, and majoritarian communities and individuals are disproportionately directed against blacks? That a movement like Black Lives Matter would be better empowered to pursue justice and equity without the legal mechanisms such as those that allowed the US Department of Justice to investigate and effect change within the Ferguson Police Department?Footnote ³⁷

People act for reasons. Among those reasons are matters to do with perceived realities of race, religion, sex, sexuality, ethnicity, and the like. We need not think that every instance of apparent racist or religious violence admits of a simple, mono-causal explanation to recognize that sometimes and perhaps often this perceived difference and the category associated with it are reason enough to make an outsider or minority a target.Footnote ³⁸ All kinds of factors cause communities and individuals to act hatefully toward those they perceive as different. Medieval Christians targeted Jews for abuse and terror long before “modern religious liberty” came along. Sometimes this was motivated by identification of Jews as guilty of deicide or host desecration. This happened frequently enough that Jewish leaders regularly petitioned the Catholic Church for protection.Footnote ³⁹ In other cases, the proximate occasion of persecution was not obviously “religious”—a missing child, say. But Jews were an easy target for blame and attack in such cases in virtue of their manifest otherness and lack of effective legal protection. Precisely their subordinate, disempowered status and religious identity—expressed through their “strange” practices and obvious rejection of Christianity—made them ready targets even in these cases. Likewise, blacks under Jim Crow were sometimes targeted as an explicit act of racial terrorism, with the very intent of terrorizing and further subjugating them, thereby extending white supremacy. Sometimes they were attacked because a white teenager became pregnant and a trumped-up charge of rape followed—but, here too, it was racism and “blackness” that rendered them targets. In both cases, no analysis of what happened or why it happened would be complete that did not take account of race. The same must be said for religion when it comes to medieval Jews amidst Christian Europe. It did not take the nation-state or religious freedom to constitute Jews as religiously other; Christian practice and theology took care of that just fine. Nothing in history or scholarship suggests that such patterns of behavior are unique to Christianity.

Among the reasons individuals and communities have found to marginalize, abuse, and attack one another are religious reasons: targeting some on the basis of one's own or their religious convictions, practices, or identity, or regarding some as outsiders or dangerous to one's religious community. None of this suggests that other factors are not frequently implicated, that religious reasons are necessarily most salient, nor that we can always understand all of the reasons and relations among the factors. Yet sometimes we can understand them relatively well: the agents tell us and their behavior confirms that they are targeting some group because they are idolaters or heretics and so on. Even when we cannot grasp exactly how various reasons are figuring, we can often recognize that religious reasons are implicated, if only because the victims are stigmatized religious minorities and unprotected by law. What laws concerning religious liberty are, at their best, aiming for and doing are basically what laws against racial discrimination are, at their best, aiming for and doing: recognizing a history of oppression and domination and the majority's tendency to ostracize and harm those considered “others,” and offering law's protection to clothe the nakedness of their social vulnerability.

This hardly suggests that religious liberty is a panacea. What it suggests is that laws protecting religious freedom are among the best tools we have for preventing the targeting or domination of individuals or communities on the basis of their religious otherness. Laws can indeed encode and extend oppression. That is just what Jim Crow did. That is just what medieval Christian laws concerning Jews did. But that is not the only thing law can do. The US Voting Rights Act sought to secure blacks against domination and to give them a voice in governance. Long before that, the thirteenth amendment to the US Constitution abolished slavery. In both cases, these legislative acts were backed up with the threat—and sometimes the mobilization—of state power. Neither is fully thinkable without the concept of race and its deployment in law. At their best, laws outlawing religious persecution seek to protect and give voice to religious minorities, putting them on equal footing with majoritarian communities. The new genealogy collapses the difference between racist and religiously prejudiced regimes of law and those that, however imperfectly, seek to protect minority communities from oppression. And it would rob such communities of some of the best tools they have for resisting persecution and standing before fellow citizens and their government without fear.

There is no doubt that such legislation can be abused, enforced in biased ways, and harden identities. But there is a great distance between these things and the real goods it seeks and often secures. And there is a far greater distance still between its incidentally doing such things and the legislation of oppression, racism, and discrimination that states sometimes pursue. The new genealogy does not adequately acknowledge these differences, collapsing all pursuit of religious liberty into one oppressive, ideological operation of state power.

It would be absurd to suggest that mid-twentieth-century US laws enacted to protect blacks, laws demanded by blacks and their allies, were the cause of the lines of racial difference and discrimination they sought to address. Those differences and that discrimination long preceded those laws, even as racist laws helped make them necessary in the first place. Citing racist backlash against civil rights laws or unintended bad effects as reason not to pursue them was the very argument “white moderates” gave as to why blacks should stop protesting and avoid “creating tension.” While many of us would welcome a post-racist, post-religious bigotry world, that is not the world that actually exists, least of all for the most vulnerable. Communities continue to act against each other for reasons related to religious and racial identities. So long as they do, those who would argue for abandoning race and religion as legal categories have a heavy burden to bear. They will need to show that the harms done to minority communities by legislation they support and help craft, laws aimed at protecting them and equipping them to hold governments and fellow citizens accountable for wrongs done on the basis of racial and religious identity, are greater than the benefits such legislation secures. The new genealogy of religious freedom has not begun to make that case.

Power and Freedom as Non-Domination

In this section, I sketch an alternate, democratic “republican” vision of power and freedom, on which the two are intertwined. Here, unfreedom is subjection to a certain power asymmetry. But not every power asymmetry is ethically problematic or freedom-destructive. Indeed, some power asymmetries are necessary for and help realize social goods. On this view, freedom is constituted by certain sorts of relationships. It concerns not atomistic individuals and noninterference but persons as always, already embedded in right relationships and their obligations.

Recall the claim: “To be free … is … to have the ability to choose for oneself in all areas of life, without restraint” (IRF, 156). Call this vision of freedom as noninterference—shared by foundationalism and the new genealogy—“liberal freedom.” The new genealogy assumes liberal freedom represents an animating ideal of the state and secularity. Religious freedom is and must be paradigmatic liberal freedom. This fuels the new genealogy's insistence that religious freedom is both impossible, thanks to the way defining religion constitutes interference, and inherently biased, thanks to liberal freedom's hyper-individualism. Republicanism, in contrast, rejects liberal freedom's individualism and public/private, secular/religious dichotomies. It requires diverse, robust communities and regards religious interchange, activism, and debate as helping constitute freedom. Far from being particularly modern or secular, its roots run from Cicero through the Italian city-states, the seventeenth-century Dutch republic, English commonwealthmen, American founders, abolitionists, suffragists, the settlement house and civil rights movements up to contemporary broad-based community organizing.Footnote ⁴⁸

Republican liberty is security against domination, freedom from mastery. The master/slave relation is the paradigm of domination. While the relational paradigm itself and questions of resemblance thereto take priority in discerning whether some arrangement constitutes domination, we can also venture ad hoc formulations and principles of assessment. Thus, we can say that a person or community is dominated when someone is in a position to exercise arbitrary power over that person or community. If power is “the capacity that an individual, group, or institution has to produce effects that people would have reason to care about,” what it means for power to be arbitrary is complex and contestable.Footnote ⁴⁹ Still, we can identify paradigm instances and offer various ad hoc principles. So, power is arbitrary when it fails to meet a reciprocity test (for example, dominator would not exchange places with dominated); it is not governed by norms constitutive of relations of mutual respect; the power differential lacks a rational basis, unlike that between teacher/student or parent/child; the dominator is not accountable to the dominated; the roles (for example, tyrant, serf, detainee) constituting the relation are such that no one would willingly occupy them; and so on.Footnote ⁵⁰

Power differentials alone do not constitute domination. Domination involves arbitrary power. Other things being equal, parents, bosses, and teachers do not count as dominating employees, students, and children in virtue of their superior power. Rather, domination pertains to power differentials that are arbitrary in ways like those detailed above.

A slave lives at the master's mercy, in perpetual terror of what the master might do. As eighteenth-century republican Richard Price explains, those “under the power of masters cannot be denominated free, however equitably and kindly they may be treated,” for even under a “benevolent” master, a slave suffers the crippling knowledge of her total vulnerability.Footnote ⁵¹ They depend on nothing more stable than a master's will for their security—and that will may change from benevolence to indifference or hostility at any time. Whether the master actually beats, chains, or otherwise interferes, he can and this constitutes a deep form of unfreedom. This is domination. On a liberal view, in contrast, so long as the master does not actually interfere, the slave is free.

Where the new genealogy blurs or elides the distinction between power asymmetry and oppression, republicans agree that no human community is bereft of power asymmetries but insist on distinguishing between just and unjust species of power, violence, and politics. To begin to grasp the point, consider the following relations: slave/master, athlete/coach, advisee/advisor, young child/loving mother, parishioner/priest, UK citizen/MP, North Korean citizen/president, citizen/state. If we fail to distinguish between just and unjust forms of power, we cannot ethically distinguish among these relations. Each will count as ethically similar, alike in respect of power and thus oppression. Yet why think the young child/loving mother or teacher/student relation is saliently like the master/slave relation—or that either is like the citizen/state relation?

Instead, we can distinguish at least three types of ethically distinct power relations and roles: just (such as faithful daughter/loving mother), unjust (such as master/slave, subject/tyrant), and indifferent, those too vague to morally evaluate without further detail and contextualization (such as citizen/state, employee/employer).Footnote ⁵² New genealogy effectively reduces all to the second category. Yet who really supposes these relations are saliently alike? That the power by which the child is sent to timeout for hitting is ethically like that by which the slave is chained for sass? That the power by which the justly convicted murderer is imprisoned is ethically like that by which a Falun Gong adherent is jailed in China? Moreover, who would claim that efforts to bring it about that someone is secure against such abuse and empowered to compel the powerful to heed their interests, are saliently like the attempt to get the upper hand or seek mastery for oneself? That these two kinds of power-seeking are ethically alike?

Power and its pursuit can be just or unjust, good or problematic. That states tend to concentrate and conceal power is not sufficient for us to know whether some state is dominating – any more than knowing that someone is an employee, citizen, borrower, athlete, advisee, apprentice, or student suffices for us to know whether she is dominated. Likewise, that some community seeks power does not suffice for us to know whether they seek domination or security against it. When it comes to judging the state or such relations, power asymmetries are both intrinsic to them and insufficient for us to pronounce on their justice. Power and its pursuit are ingredients in domination and non-domination alike.

For the new genealogy and foundationalism, law as such necessarily violates freedom, for law is nothing if not interfering. On this view, the compromise of freedom that law constitutes is tolerated because it prevents still more or worse interference. So, law compromises my freedom to steal to preserve everyone's freedom to enjoy her property. For republicans, in contrast, just law does not compromise freedom at all. So long as some law is just—so long as it is nonarbitrary, as a law against theft is—there is no violation of freedom, for there is no domination. On this view, just laws do not compromise freedom; they help constitute it. There is a vital point here. It concerns the difference between republican and liberal or foundationalist attitudes toward families, religious communities, and the like. These are high interference relations—binding us in webs of obligation, prescribing limits, demanding sacrifice. For liberals they are thus relations of unfreedom. In such a regime, they are, at best, tolerated. The new genealogy rightly unveils liberalism's individualist, antireligious bias. But for republicans these high-interference relations need not be relations of unfreedom at all, for they need not be dominating.Footnote ⁵³ Such relations are not tolerated but celebrated: they foster a society that prizes liberty and they realize the good of mutuality.Footnote ⁵⁴ Republicans seek the flourishing of such relations, for by enjoying good relations in families, religious communities, and the like, citizens not only learn habits of identifying and resisting dominative relations, they actually manifest the social good of freedom.Footnote ⁵⁵

Non-domination is a common good: enjoyment by one requires enjoyment by many. Imagine a society where women suffer the arbitrary power of men. Now imagine a well-positioned woman whose husband happens not to arbitrarily interfere with her, though he could. To liberals, she is free.Footnote ⁵⁶ She may help battered wives and advocate laws limiting male interference—but she does so as an individual helping other individuals, individuals who, unlike her, are unfree.Footnote ⁵⁷ For republicans, in contrast, she is herself unfree, for the community of women is dominated. Concern with her own freedom is concern for the community and vice-versa: “There is no way of removing the domination without altering the conditions under which women generally relate to men.”Footnote ⁵⁸

So things go for any other identity, not least religious identities. Ideally, one moves outward in concentric circles of communal identity: parishioners → Methodists → Christians → religionists, for instance.Footnote ⁵⁹ Republican liberty is an ongoing communal task secured not only by laws but by dispositions to recognize other individuals and communities as similarly deserving of recognition and security against domination. In its democratic form, which we have implicitly had in view, republicanism is committed to ever greater degrees of inclusivity when it comes to who gets to enjoy the benefit of liberty: not merely straight white men, for instance, but strictly all adults in some community no matter their sex, class, education, race, sexual orientation, immigration status—or even their religion.

Beyond Essentialism, Dichotomies, and the Very Idea of Religious Freedom

When the new genealogy claims that “to be religious is not to be free” (IRF, 156) or “What religion is not is freedom” or “the very idea of freedom of religion is paradoxical; it is the freedom to be unfree” (PRF, 298), we confront the inadequacy of its noninterference vision. It neither captures the way many believers understand freedom nor why they might pursue it.Footnote ⁶⁰ Ironically, in so conceiving of freedom, the new genealogy actually deepens and reifies a religion/secular dichotomy and essentializes each: religion is unfreedom; the secular is freedom. Religion opposes freedom; states celebrate it (though, so much the worse for them). One must choose: religious practices or political participation, communal identity or civic standing. Freedom can only be secular ideology, wholly alien to religious self-understandings and communities.

Regarding freedom as non-domination dissolves this story, this essentialism, and these dichotomies. It allows for continuity between religious visions and politically pluralist ones and for imagining that religious freedom might echo some value within a given tradition. The new genealogy forecloses that possibility altogether: the very idea of freedom—modern, liberal, and secular—and of religion are essentially opposed.

From God's-Eye Neutrality to Principled Evenhandedness

As the new genealogy has it, liberal states purport to be neutral and agnostic in regard to religion and religious truth. Yet the very idea of religious freedom requires the state to make controversial judgments concerning what religion even is. Thus, the liberal state fails to be neutral; it is “a distinctive kind of theological state” (PRF, 333). Moreover, we here behold its totalizing, ideological essence.

This story depends on the foundationalist identification of Neutrality with neutrality, the premise that neutrality requires not only noninterference but governance according to “God's-eye-view” principles or unassailable deliverances of Reason, natural law, the “original position,” and the like. The new genealogy accepts this foundationalist conception of neutrality—and debunks accordingly. But why accept that conception?

The neutrality worth caring about and wanting in governance is a species of justice: a commitment to giving each person and community its due, attending both to desert or what each is owed, and equity or treatment of like cases in like ways. Neutrality names especially a commitment to preventing biases, prejudice, or extraneous interests from distorting this work. A neutrality worth prizing is one that pursues principled, sensitive, humble, evenhandedness, coupled with publicity, openness to rebuke, and procedures for self-correction, contestation, and accountability. Reciprocity principles like the golden rule, mechanisms of accountability, and, once deracinated from totalizing ambitions and dubious questing for foundations, even devices like veils of ignorance and reflective equilibriums can help in pursuing neutrality without foundations. For instance, we imagine being at the mercy of the powerful so that they could do with us as they please. And we imagine, in contrast, what it would be like if they were constrained to listen to us and accountable to us. We prefer the latter. If we would not want to be under unconstrained power, can we, in fairness, seek it ourselves?

Yet, because this vision is dialectical, it is not appropriate to permanently enshrine some necessary and sufficient criteria or formalized principles derived from such tests and simply apply them to each novel case. Foundationalists seek the perfect universal principles so that particular cases can get processed through them.Footnote ⁶¹ The new genealogy unmasks the contingency and historicity of principles and cases alike. A dialectical view stresses that such contingency and historicity does not mean our efforts are bound to be un-principled, arbitrary. It further adds that each application of a principle changes that principle.

Think of principles enshrined at the founding of the United States, principles many saw as compatible with slavery and patriarchy. If meaning is a matter of use—the claim “all men are created equal” at that time meant something other than we now take it to mean.Footnote ⁶² As slaves are freed and women enfranchised—it is not that a once-set principle has been applied to new cases while itself remaining static. Rather, those and other new applications of principles are, simultaneously, transformations of the principle. For the principles’ meaning is their use: what they are applied to and the consequences that follow. Legal precedents offer a helpful analogy: the new application of some precedent—in that very application—affects that precedent's meaning, both now and for future cases. And the meaning of the precedent qua precedent is the history of its application to date.

The dialectical departure from foundationalism is not just a matter of a different view of the origin and status of principles but of their role and of the relation between them and particular cases.Footnote ⁶³ When we undertake politics in view of such considerations we can pursue justice and evenhandedness, no matter that these pursuits are ad hoc, contestable, open to doubt, and so on. This pursuit is distinct from seeking mastery or abandoning principles. Let those who doubt that there is a real difference here—who claim rights or principles are only tools for state domination—agree to live under a regime without them.

Religious Freedom and Neutrality Worth Having

What is required for neutrality-as-evenhandedness? We can easily imagine countless things incompatible with it: outright persecution; treating religionists in capricious, discriminatory ways; conceiving religion in a way that (deliberately or not) systematically privileges the majority, renders it privatized, or construes it in Christian terms; and the like. Positively, evenhandedness would require: special sensitivity to the vulnerable, disempowered status of minority religionists; conceptions of religion that err on the side of including self-identified believers of minority communities and heavily weigh their conceptions of what constitutes adherence to their tradition; care in conceiving of religion so as not implicitly to privilege the understanding favored by or best fitting a religious majority; recognition of judicial tendencies to overestimate knowledge of religious traditions; and so on. In the US and European contexts, this would involve attending especially to features of religion that may not figure in dominant forms of Christianity. Thus, a case like Lukumi Babalu Aye, wherein a contested ordinance forbidding animal sacrifice was generally applicable but Santeria practitioners nonetheless prevailed, would exhibit the neutrality I have in mind.Footnote ⁶⁴

Now, suppose such a decision does entangle the court in considering the relative importance of some practice to some religion and in weighing that against governmental interests. Suppose, as the new genealogy contends, that the state judges what constitutes religion and even the relative importance of some practice for some religion. The knowledge by which a court makes such judgments need not be arbitrary, biased, or ideological. It should derive from expert and believer testimony, including parties to the case. Such testimony constitutes some of the facts of the case—as it constitutes some of the facts of the course when I teach students what is central to some variety of Islam, what is less so, where debates lie, and so on. If the importance of some practice is disputed, the fact of the dispute should be taken into consideration, along with the difference made by the various contexts in play. None of this seems to violate norms of neutrality or justice. It seems their presupposition. Call all this attention, weighing, and consideration, “talking theology.”Footnote ⁶⁵ How does it serve the cause of justice, fairness, or evenhandedness not to “talk theology”? Not to attend to these facts concerning conceptions of the religion, its practices, their importance, and disputes about these things?

Imagine various versions of the Smith peyote case before a Supreme Court in a US-like legal system and running along a scale bounded by the following scenarios:Footnote ⁶⁶

• Scenario A. At one end, notwithstanding legitimate governmental interest, the law's general applicability, and absence of animus, the court decides for Smith, granting an exemption in view of the fact that peyote use is central to his religion and religious freedom would otherwise be severely compromised. Any threat to public order is outweighed by his right to religious freedom, the wrong of compromising this individual's or community's capacity to practice their religion.

• Scenario Z. At the other end, the court not only rules against Smith but declares that these practices are not religion. They throw in blinkered commentary about their irrationality, their dissimilarity to Christianity, and their departure from “civilized” ways. The judges show Smith blatant disrespect, failing to attend to or engage arguments on his behalf. It is a mere show of a trial.

We can imagine countless variations running the gamut from A to Z, let alone those beyond Z—such as putting the verdict to referendum, deciding by coin toss, pursuing outright persecution. Who doubts there are vital differences here? That scenario A and Z differ radically in regard to justice, neutrality, and religious freedom? Would someone really want to say that scenario A exhibits the paradox of religious freedom, shows religious liberty to be domination, indicates inevitable bias toward belief over practice, or displays a privatizing impulse somehow essential to “the state” and “the secular”? So long as we can imagine scenario A, the new genealogy is misguided. Nothing proper to “the state” renders scenario A impossible. I see no sense in which scenario A fails to exhibit neutrality, let alone justice.Footnote ⁶⁷

Consider another scenario along the same scale:

• Scenario L. The Court rules against Smith in view of the general applicability of the law, lack of animus, and (relatively minimal) threat to public order. It does so thanks precisely to its refusal to consider or entangle itself in judging the importance of the practice to this religion in the name of foundationalist Neutrality. They weigh only some undifferentiated, generic “religious value.”

Compare scenario L to scenario A. In what way does scenario L exhibit a species of neutrality worth valuing? In what sense does it more deeply exhibit justice or honor religious freedom? I can think of none. If what it displays deserves to be called “neutrality,” I hardly see what service it is to the cause of justice or religious freedom.

Now consider a set of variations on our first scenario:

• Scenarios A₁, A₂, A₃, etc. Here, the facts of the case vary somewhat in each instance: the public interest is more (or less) compelling, the practice less (or more) central to the religion, or both. The verdict changes accordingly in view of the matrix between the relative importance of the public interest and the centrality of the practice—which the Court takes into account. Smith wins or loses accordingly.

Does this seem unjust? A failure of neutrality? Compare this to the set:

• Scenarios A¹, A², A³, etc. Here, the facts of the case vary in identical ways as compared to the other set, but in all these cases the court foreswears considering the importance of the various practices or the teachings of the religion. Thus, their decision varies independently of anything to do with the practice's significance.

In what sense does this represent a gain for justice over the former set? For freedom? Do we imagine scenario A and iterations like it not to be just or principled? Is it no morally different from scenario Z, let alone the horrendous, all too familiar alternatives not even on this scale?

To be very clear, I am not suggesting that most or even many religious freedom cases are like this—nor that governments cannot entangle themselves in theological matters or controversies in ways that would be violative of religious freedom. I mean, instead, to begin to enflesh a vision of evenhandedness unlike that shared by foundationalism and the new genealogy. On this view, attending to the particularities of a religious community or individuals, weighing the relative importance of some practice, discerning the differences made by local and broader contexts, in short, “talking theology,” is not necessarily contrary to religious freedom or neutrality but often essential to it.

Dialectical Religious Freedom: Judging Importance and “Talking Theology”

Pursuing religious liberty involves judgments of importance, discerning how central or weighty some religious matter is. Such judgments seem inevitable. Implicitly or explicitly, hidden or public, ignorantly or informed, uncontested or subject to scrutiny, they will be made. If so, the best chance of their being made justly involves rendering the relevant commitments and reasoning public and contestable. Domination is more, not less, likely when such work proceeds without acknowledgement, scrutiny, or challenge. Pursued well, such activity does not violate religious freedom but helps constitute it. Fairness involves taking all the relevant facts into account. Among these are facts about the relative importance of various practices and commitments. Such weighing of goods and judgments of importance need not be arbitrary or unprincipled. They can be highly objective. In no religious tradition, community, or life is everything equally important in relation to everything else. Were everything equally important, nothing would be important at all.

Instead, religious communities, like other communities, have more or (often) less settled views on the relative significance of various practices and commitments—however rough, implicit, fluid, and revisable. Certainly, such matters are contested. Indeed, such contestations and the divisions they spur help define boundaries between communities, subgroups, and individuals. Nonetheless, we can find points of agreement, and corresponding judgments of importance, in virtue of which, for instance, we regard Sunnis and Shia as alike Muslims—or Muslims and Jews as, alike, monotheists. Such divisions instruct us about what is more or less important to whom and in what ways. And the same point holds, for instance, within Sunni Islam or Therevāda Buddhism—where such disagreements divide this group from that. Most simply, working carefully, we can accurately discern what is more or less important to some religious community—as various members and observers see it. The possibility of academic study of religion presupposes this point. We can say, for instance, that the Eucharist is more central to Catholicism and much of Christianity than foot washing, or that Passover is more important for most Jews than Chanukah. Yet such judgments are also profoundly context- and community-dependent. For certain Baptists, foot washing is more important than the Lord's Supper; for some Jews, Chanukah more important than Passover. All this can be taken into account as needed in some case.

When it comes to such discernment, we must attend to the practices and priorities of the communities and contexts in question—not just the broader communities of which they are part. It would be mistaken to assume that what is normative or important for some particular Muslim or Muslim community comports with what has been normative or important for most Muslims or Islam more generally. What we want is a sense of the relative importance in some case for this adherent or community. We cannot assume “normative” Islam (whatever that is) and then evaluate. Rather, we need to attend to what this adherent or community identifies as important, including which communities they regard most salient, whom they recognize and seek recognition from. This may be in continuity with “mainstream” strands of some tradition, but it may not. For one parish, the most salient community may be that parish; for another, it may be the “Church catholic.” And the relative importance of some practice may vary accordingly. Thus, among the relevant facts are those related to the community's or individual's vision of their relation to still broader communities—and these may differ dramatically from what “mainstream” teaching about this relation suggests.

None of this suggests simply taking some adherent's “word for it,” but of tacking dialectically between what she says and what we discern to be the case for some community. This fundamentally inductive approach accords local context and communities immense importance.Footnote ⁶⁸ What it rejects is the notion that attending to these matters is impossible or somehow compromises freedom. And it rejects approaches that are top-down, artificially formalized, or context-negligent. Such a priori, anti-dialectical approaches are exemplified in the expert testimony of Daniel Pals and Nathan Katz in a Florida trial concerning religious burial practices (IRF, 147–48).Footnote ⁶⁹ They offer abstract, universalistic formulas as one-size-fits-all tests for judging religious authenticity and importance. As Winnifred Sullivan rightly notes, each predictably favors “mainstream,” “dominant,” institutionalized strands of tradition.Footnote ⁷⁰ Local practices are discounted as heterodox or lacking authority and so as less religious or important. My approach rejects such foundationalism but refuses to abandon efforts to judge relative importance.Footnote ⁷¹

Distinct from but related to these importance issues, religious freedom conflicts also admit of different kinds, depending on what some practice means for some community. Say some state interest impinges on some central practice, and suppose the state interest seems clearly weightier. Still, perhaps the practice is so important that it helps constitute a community as the very community it is. To subtract that practice is not merely to compromise the community but to violate it. And perhaps religious freedom requires security against violation. In this sense, some harms may compromise or even damage, but not every harm violates. Think, for instance, of the difference between a mugging in which someone is badly beaten and a sexual assault at knifepoint. Beyond bodily harm, we can consider the spiritual and psychic damage. Ceteris paribus, the sexual assault but not the mugging seems violative. This has to do with complex features such as the importance, intimacy, or centrality to self of what is attacked, the kind of vulnerability implicated, and so on, that resist formalized articulation. Likewise, whatever the bodily harm, if some mugging's motive were patently racist—the victim targeted for being black—it, too, could be violative, a “hate crime.” And like sexual assault, its violative character will have to do with what precisely it attacks: this person as black. Most obviously, laws exhibiting animus against some religious group seem violative in this way—even if the apparent harm is minimal or the affected practice relatively unimportant. But what I am suggesting is that even apart from animus, compromising certain practices, even if they are not central, can amount to violation and thus be incompatible with religious liberty. As with the hate crime, context and history matter: even if circumcision were not so central to Judaism, its proscription in Germany amounts to something weightier than its proscription elsewhere.

When it comes to burdens and conflicts alike, multiple distinctions are salient for religious liberty. Alongside considerations of importance, we might distinguish among burdens, compromises, restrictions, and violations as naming different losses some religious community might suffer. And we can still further distinguish impacts in terms of scope, rationale, likely consequences, etc.Footnote ⁷² Yet so much as to bring these distinctions in view requires “talking theology,” attending to history and context, and gauging the importance of some practice for some community.

In all this, there is a distinction worth caring about between imperfect and perfect religious liberty. No religious liberty regime can guarantee that every believer is always without just complaint. But this does not mean that religious liberty is impossible or a tool of oppression. The inherent imperfection of religious liberty—inherent given the world we actually live in—is not the same as its absence or violation. Religious liberty comes in degrees. Some things seem altogether incompatible with it. But there are a wide range of arrangements and affairs in which religious freedom can count as present even as it is more or less perfectly realized. Perhaps we can further identify a narrower range wherein we regard a society as enjoying religious liberty. This will mean that there are basic, inviolable guarantees against religious coercion, persecution, discrimination, and the like. And beyond this, the legal system will guarantee that, in the inevitable conflicts and frustrating losses that invariably follow for some communities, the relevant decision-making processes themselves are nonarbitrary, manifesting respect, equity, contestability, and the like. This is what the scenario A–Z thought experiments bring into view.Footnote ⁷³ “Talking theology” and pursuing judgments of importance are certainly forms of interference and entanglement. But they are nonarbitrary and undertaken in order to render each her due under law. They do not compromise religious freedom; they make it possible.