Two forms of secularism

Charles Taylor, ‘How to define secularism’, in Stepan/Taylor, Boundaries of Toleration, Columbia University Press, 2014, pp. 60-78.

EVERYONE AGREES TODAY that modern, diverse democracies have to be “secular” in some sense of this term. But what sense? The term (along with the corresponding French term laïcité (and its derivatives) has more than one sense. There are in fact many different meanings, but I believe that we can get to a crucial issue if we single out two key conceptions.

On one view (A), secularism is mainly concerned with controlling religion. Its task is to define the place of religion in public life and to keep it firmly in this location. This doesn’t need to involve strife or repression, provided various religious actors understand and respect these limits. But the various rules and measures that make up the secularist (or laïque) regime all have this basic purpose.
On the other view (B), the main point of a secularist regime is to manage the religious and metaphysical-philosophical diversity of views (including non- and antireligious views) fairly and democratically. 1 Of course, this task will include setting certain limits to religiously motivated action in the public sphere, but it will also involve similar limits on those espousing non- or antireligious philosophies. (For instance, the degree to which either can discriminate in certain relations like hiring). For B, religion is not the prime focus of secularism.

The case I would like to make here is that B is much superior to A, at least for our time. The popularity of A is to be explained by certain Western histories of struggle in which secularist regimes came to be. But our present predicament is for the most part rather different than the one which generated these conflicts. It is above all one of growing diversity in all Western democracies. For these reasons, B is more appropriate.

Let’s look at what B involves a little more closely. In fact managing diversity involves a complex requirement. There is more than one good sought here. We can single out three, which we can class in the three categories of the French Revolutionary trinity: liberty, equality, fraternity. (1) No one must be forced in the domain of religion or basic belief. This is what is often defined as religious liberty, including, of course, the freedom not to believe. This is what is also described as the “free exercise” of religion, in the terms of the U.S. First Amendment. (2) There must be equality between people of different faiths or basic belief; no religious outlook or (religious or areligious) Weltanschauung can enjoy a privileged status, let alone be adopted as the official view of the state. Then, thirdly, (3) all spiritual families must be heard, included in the ongoing process of determining what the society is about (its political identity), and how it is going to realize these goals (the exact regime of rights and privileges). This (stretching the point a little) is what corresponds to “fraternity.”

These goals can, of course, conflict; sometimes we have to balance the goods involved here. Moreover, I believe that we might add a fourth goal: that we try as much as possible to maintain relations of harmony and comity between the supporters of different religions and Weltanschaungen (maybe this is what really deserves to be called fraternity, but I am still attached to the neatness of the above schema, with only the three traditional goods).
Why do I think that this diversity model (B) is superior to the religion-focused model (A)? One reason is that it is more evenhanded. If we look at the three goals, they are concerned respectively, with (1) protecting people in their belonging and/or practice of whatever outlook they choose or find themselves in; with (2) treating people equally whatever their option; and (3) giving them all a hearing. There is no reason to single out religion as against nonreligious, “secular” (in another widely used sense), or atheist viewpoints.
Indeed, the point of state neutrality is precisely to avoid favoring or disfavoring not just religion positions but any basic position, religious or nonreligious. We can’t favor Christianity over Islam, but also religion over against nonbelief in religion or vice versa.
One of the ways of demonstrating the superiority of the three-principle model of secularism over that which is fixated on religion is that it would never allow one to misrecognize the regime founded by Atatürk as genuinely secular, making light as it does of the fundamental principles and even of the separation of state and religious institutions.
This also shows the value of the late-Rawlsian formulation for a secular state, which cleaves strongly to certain political principles: human rights, equality, the rule of law, democracy. These are the very basis of the state that must support them. But this political ethic can be and is shared by people of very different basic outlooks (what Rawls calls “comprehensive views of the good”). A Kantian will justify the rights to life and freedom by pointing to the dignity of rational agency; a utilitarian will speak of the necessity to treat beings who can experience joy and suffering in such a way as to maximize the first and minimize the second; a Christian will speak of humans as made in the image of God. They concur on the principles, but differ on the deeper reasons for holding to this ethic. The state must uphold the ethic, but must refrain from favoring any of the deeper reasons.

The idea that secularism makes a special case of religion arises from the history of its coming to be in the West (as does, indeed, the name). To put it briefly, there are two important founding contexts for this kind of regime, the U.S. and France. In the U.S. case the whole range of comprehensive views, or deeper reasons, were in the original case variants of (Protestant) Christianity, stretching to a smattering of Deists. Subsequent history has widened the palette of views beyond Christianity and then beyond religion. But in the original case the positions between which the state must be neutral were all religious. Hence the First Amendment: Congress shall pass no law establishing religion or impeding the free exercise thereof (or something like this).
The word secularism didn’t appear in the early decades of American public life. But this was the sign that a basic problem had not yet been faced. Because the First Amendment concerned the separation of church and state, it opened the possibility of giving a place to religion, which no one would accept today. Thus, in the 1830s, a judge of the Supreme Court could argue that while the First Amendment forbade the identification of the federal government with any church, since all the churches were Christian (and in effect Protestant), one could invoke the principles of Christianity in interpreting the law.
For judge Joseph Story, the goal of the First Amendment was “to exclude all rivalry among Christian sects,” but nevertheless “Christianity ought to receive encouragement from the state.” Christianity was essential to the state because the belief in “a future state of rewards and punishments” is “indispensable to the administration of justice.” What is more, “it is impossible for those who believe in the truth of Christianity, as a divine revelation, to doubt, that it is a special duty of government to foster, and encourage it among the citizens.”
This primacy of Christianity was upheld even later in the nineteenth century. As late as 1890, thirty-seven of the forty-two existing states recognized the authority of God in the preambles or in the text of their constitutions. A unanimous judgment of the Supreme Court of 1892 declared that if one wanted to describe “American life as expressed by its laws, its business, its customs and its society, we find everywhere a clear recognition of the same truth … that this is a Christian nation” ( Church of the Holy Trinity v. United States, 143 U.S. 457 at 471).
In the latter part of the century, resistance began to build to this conception, but a National Reform Association was founded in 1863 with the following goal: “The object of this Society shall be to maintain existing Christian features in the American government … to secure such an amendment to the Constitution of the United States as will declare the nation’s allegiance to Jesus Christ and its acceptance of the moral laws of the Christian religion, and so as to indicate that this is a Christian nation, and place all the Christian laws, institutions, and usages of our government on an undeniable legal basis in the fundamental law of the land.” After 1870 the battle was joined between the supporters of this narrow view, on one hand, and those who wanted a real opening to all other religions and also to nonreligion. These included not only Jews but also Catholics who (rightly) saw the “Christianity” of the NRA as excluding them. It was in this battle that the word secular first appears on the American scene as a key term, and very often in its polemical sense of non- or antireligious. 2
In the French case, laïcité came about in a struggle against a powerful church. The strong temptation was for the state itself to stand on a moral basis independent from religion. Marcel Gauchet shows how Charles Renouvier laid the grounds for the outlook of the Third Republic radicals in their battle against the church. The state has to be “moral et enseignant.” It has “charge d’âmes aussi bien que toute Église ou communauté, mais à titre plus universel.” Morality is the key criterion. In order not to be under the church, the state must have “une morale indépendante de toute religion,” and enjoy a “suprématie morale” in relation to all religions. The basis of this morality is liberty. In order to hold its own before religion the morality underlying the state has to be based on more than just utility or feeling; it needs a real “théologie rationnelle,” like that of Kant. 3 The wisdom of Jules Ferry, and later of Aristide Briand and Jean Jaurez, saved France at the time of the Separation (1905) from such a lopsided regime, but the notion stuck that laïcité was all about controlling and managing religion.

If we move, however, beyond such originating contexts, and look at the kinds of societies in which we are now living in the West, the first feature that strikes us is the wide diversity not only of religious views but also of those that involve no religion, not to speak of those that are unclassifiable in this dichotomy. Reasons (1), (2), and (3) require that we treat evenhandedly all of these.
This fixation on religion is complex, and it is bound up with two other features we often find in the debates on secularism: the first is the tendency to define secularism or laïcité in terms of some institutional arrangement, rather than starting from the goals as I proposed earlier. And so you hear mantra-type formulae, like “the separation of church and state” or the necessity of removing religion from public space (“les espaces de la République,” as in the recent French debate). The second follows from the first or may easily seem to. If the whole matter is defined by one institutional formula, then one must just determine which arrangement of things best meets this formula, and there is no need to think further. One cannot find oneself in a dilemma, as will easily happen if one is pursuing more than one goal, because here there is just one master formula.
Hence one often hears these mantras employed as argument stoppers, the ultimate decisive response that annuls all objections. In the U.S., people invoke the “Wall of Separation” as the ultimate criterion, and hyper-republicans in France cite laïcité as the final word. (Of course, if one consulted the First Amendment of the U.S. Constitution one would find two goals mentioned, the rejection of establishment and the assurance of “free exercise.” It is not inconceivable that these could conflict.)
This kind of move amounts, from the standpoint I’m adopting here, to a fetishization of the favored institutional arrangements, whereas one should start from the goals and derive the concrete arrangements from these. It is not that some separation of church and state, some mutual autonomy of governing and religious institutions will not be an inescapable feature of any secularist regime. And the same goes for the neutrality of the public institutions. These are both indispensable. But what these requirements mean in practice ought to be determined by how we can maximize our three (or four) basic goals.
Take for example the wearing of the hijab by Muslim women in public schools, which has been a hot issue in a number of Western democracies. In France, pupils in public schools were famously forbidden the headscarf, seen as a “signe religieux ostantatoire,” according to the notorious Loi Stasi of 2004. In certain German Laender, pupils can wear it, but not teachers. In the UK and other countries there is no general interdict, but the individual schools can decide.
What are the reasons for this variation? Plainly, in all these cases, legislators and administrators were trying to balance two goals. One was the maintenance of neutrality in public institutions seen (rightly) as an essential entailment of goal (2): equality between all basic beliefs. The other was goal (1), ensuring the maximum possible religious liberty or, in its most general form, liberty of conscience. Goal (1) seems to push us toward permitting the hijab anywhere. But various arguments were made to override this in the French and German cases. For the Germans what was disturbing was that someone in authority in a public institution should be religiously marked, as it were. In the French case an attempt was made to cast doubt on the proposition that wearing the hijab was a free act. There were dark suggestions that the girls were being forced by their families or by their male peers to adopt this dress code. That was one argument that was frequently used, however dubious it might appear in the light of the sociological research carried out among the pupils themselves, which the Stasi Commission largely ignored.
The other main argument was that the wearing of the headscarf in school was less an act of piety than a statement of hostility against the republic and its essential institution of laïcité. This was the meaning behind the introduction of the concept of “signe ostantatoire.” A smaller discrete sign would be no problem argued the Stasi Commission, but these attention-grabbing features of dress were meant to make a highly controversial statement. It was in vain that Muslim women protested that “le foulard n’est pas un signe.”
So on one level we can see that these different national answers to the same question reflect different takes on how to balance the two main goals of a secular regime. But on another level the dilemma and its resolution remain hidden under the illusion that there is only one principle here, say, laïcité and its corollary of the neutrality of public institutions or spaces (“les espaces de la République”). It’s just a matter of applying an essential feature of our republican regime; there is no need or place for choice or the weighing of different aims.
Perhaps the most pernicious feature of this fetishization is that it tends to hide from view the real dilemmas that we encounter in this realm, and which leap into view once we recognize the plurality of principles at stake.
We should be aware that this fetishization reflects a deep feature of life in modern democracies. We can see why as soon as we ponder what is involved in self-government, what is implied in the basic mode of legitimation of states that they are founded on popular sovereignty. For the people to be sovereign, it needs to form an entity and have a personality.
The revolutions, which ushered in regimes of popular sovereignty, transferred the ruling power from a king onto a “nation” or a “people.” In the process, they invent a new kind of collective agency. These terms existed before, but the thing they now indicate, this new kind of agency, was something unprecedented, at least in the immediate context of early modern Europe. Thus the notion “people” could certainly be applied to the ensemble of subjects of the kingdom or to the nonelite strata of society, but prior to the turnover it hadn’t indicated an entity that could decide and act together, to whom one could attribute a will.
But for people to act together, in other words, to deliberate in order to form a common will on which they will act, requires a high degree of common commitment, a sense of common identification. A society of this kind presupposes trust, the basic trust that members and constituent groups have to have, the confidence that they are really part of the process, that they will be listened to and their views taken account of by the others. Without this mutual commitment, this trust will be fatally eroded.
And so we have in the modern age a new kind of collective agency. It is one with which its members identify, typically as the realization/bulwark of their freedom and/or the locus of their national/cultural expression (or most often, some combination of the two). Of course, in premodern societies, too, people often “identified” with the regime, with sacred kings or hierarchical orders. They were often willing subjects. But in the democratic age we identify as free agents. That is why the notion of popular will plays a crucial role in the legitimating idea. 4
This means that the modern democratic state has generally accepted common purposes or reference points, the features whereby it can lay claim to being the bulwark of freedom and locus of expression of its citizens. Whether or not these claims are actually founded, the state must be so imagined by its citizens if it is to be legitimate.
So a question can arise for the modern state for which there is no analogue in most premodern forms: what/whom is this state for? whose freedom? whose expression? The question seems to make no sense applied to, say, the Austrian or Turkish Empires—unless one answered the “whom for?” question by referring to the Habsburg or Ottoman dynasties, and this would hardly give you their legitimating ideas.
This is the sense in which a modern state has what I want to call a political identity, defined as the generally accepted answer to the “what/whom for?” question. This is distinct from the identities of its members, that is, the reference points, many and varied, which for each of these defines what is important in their lives. There better be some overlap, of course, if these members are to feel strongly identified with the state; but the identities of individuals and constituent groups will generally be richer and more complex, as well as being often quite different from each other. 5
In other words, a modern democratic state demands a “people” with a strong collective identity. Democracy obliges us to show much more solidarity and much more commitment to one another in our joint political project than was demanded by the hierarchical and authoritarian societies of yesteryear. In the good old days of the Austro-Hungarian Empire, the Polish peasant in Galicia could be altogether oblivious of the Hungarian country squire, the bourgeois of Prague, or the Viennese worker, without this in the slightest threatening the stability of the state. On the contrary, this condition of things only becomes untenable when ideas about popular government start to circulate. This is the moment when subgroups that will not, or cannot, be bound together start to demand their own states. This is the era of nationalism, of the breakup of empires.
I have been discussing the political necessity of a strong common identity for modern democratic states in terms of the requirement of forming a people, a deliberative unit. But this is also evident in a number of other ways. Thinkers in the civic humanist tradition, from Aristotle through to Arendt, have noted that free societies require a higher level of commitment and participation than despotic or authoritarian ones. Citizens have to do for themselves, as it were, what otherwise the rulers do for them. But this will only happen if these citizens feel a strong bond of identification with their political community and hence with those who share with them in this.
From another angle again, because these societies require strong commitment to do the common work, and because a situation in which some carried the burdens of participation and others just enjoyed the benefits would be intolerable, free societies require a high level of mutual trust. In other words, they are extremely vulnerable to mistrust on the part of some citizens in relation to others, that the latter are not really assuming their commitments—e.g., that others are not paying their taxes or are cheating on welfare or as employers are benefitting from a good labor market without assuming any of the social costs. This kind of mistrust creates extreme tension and threatens to unravel the whole skein of the mores of commitment that democratic societies need to operate. A continuing and constantly renewed mutual commitment is an essential basis for taking the measures needed to renew this trust.
The relation between nation and state is often considered from a unilateral point of view, as if it were always the nation that sought to provide itself with a state. But there is also the opposite process. In order to remain viable, states sometimes seek to create a feeling of common belonging. This is an important theme in the history of Canada, for example. To form a state, in the democratic era, a society is forced to undertake the difficult and never-to-be-completed task of defining its collective identity.
Thus what I have been calling political identity is extremely important in modern democratic states. And this identity is usually defined partly in terms of certain basic principles (democracy, human rights, equality) and partly in terms of their historical, or linguistic, or religious traditions. It is understandable that features of this identity can take on a quasi-sacred status, for to alter or undermine them can seem to threaten the very basis of unity without which a democratic state cannot function.
It is in this context that certain historical institutional arrangements can seem to be untouchable. They may appear as an essential part of the basic principles of the regime, but they will also come to be seen as a key component of its historic identity. This is what one sees with laïcité as invoked by many French republicans. The irony is that in the face of a modern politics of (multicultural) identity they invoke this principle as a crucial feature of (French) identity. This is unfortunate, but very understandable. It is one illustration of a general truth: that contemporary democracies, as they progressively diversify, will have to undergo redefinitions of their historical identities, which may be far-reaching and painful.
At this point, I would like to discuss an interesting point that Habermas reminds us of in his paper “Das Politische”: originally political authority was defined and justified in cosmic-religious terms. It was defined within the terms of a “political theology.” But Habermas seems to think that modern secular states might do altogether without some analogous concept, and this seems to me not quite right.
The crucial move that we see in the modern West from the seventeenth century, the move that takes us out of the cosmic religious conceptions of order, establishes a new “bottom-up” view of society, as existing for the protection and mutual benefit of its (equal) members. There is a strong normative view attached to this new conception, which I’ve called the “modern moral order.” 6 It enshrines basically three principles (on one possible enumeration): (1) the rights and liberties of the members, (2) the equality among them (which has of course been variously interpreted and has mutated toward more radical conceptions over time), and (3) the principle that rule is based on consent (which has also been defended in more and less radical forms).
These basic norms have been worked out in a host of different philosophical anthropologies and according to very different concepts of human sociability. It very soon transcended the atomism that narrowed the vision of its early formulators, like Locke and Hobbes. But the basic norms remain and are more or less inseparable from modern liberal democracies.
The rejection of cosmic-religious embedding thus was accomplished by a new conception of “the political,” a new basic norm, which, as Lefort suggests, involved its own representation of political authority, but one in which the central spot remains paradoxically empty. If the notion of sovereignty is retained, no one person or group can be identified with it.
Democratic societies are organized not necessarily around a “civil religion,” as Rousseau claimed, but certainly around a strong “philosophy of civility,” enshrining the three norms, which in contemporary societies are often expressed as (1) human rights, (2) equality and nondiscrimination, and (3) democracy.
But, in certain cases, there can be a civil religion: a religious view incorporating and justifying the philosophy of civility. This was arguably so for the young American republic. It was adopting a form which was clearly part of God’s providential plan for mankind (“We hold these truths to be self-evident, that men were created equal …”). Or it can alternatively be part of a non- or even antireligious ideology, as with the First French Republic. One can even argue that all-englobing views of this kind seem more “natural” to many of our contemporaries. After all, the principles of our civil philosophy seem to call for deeper grounding. If it’s important that we agree on the principles, then surely things are much more stable if we also accept a common grounding. Or so it may appear, and the centuries-long tradition of political life seems to testify to this idea.
For indeed the overlapping consensus between different founding views on a common philosophy of civility is something quite new in history and relatively untried. It is consequently hazardous. And, besides, we often suspect that those with different basic views can’t really subscribe to these principles, not the way we do! (Because, as “we” know, “atheists can’t have principles or, as [another] “we” knows, “religions are all against liberty and /or equality.”)
The problem is that a really diverse democracy can’t revert to a civil religion or antireligion, however comforting this might be, without betraying its own principles. We are condemned to live an overlapping consensus.
We have seen how this strongly motivated move to fetishize our historical arrangements can prevent seeing our secular regime in a more fruitful light, which foregrounds the basic goals we are seeking and which allows us to recognize and reason about the dilemmas we face. But this connects to the other main cause of confusion I cited previously, our fixation on religion as the problem. In fact, in many Western countries we have moved from an original phase in which secularism was a hard-won achievement warding off some form of religious domination to a phase of such widespread diversity of basic beliefs, religious and areligious, that only clear focus on the need to balance freedom of conscience and equality of respect can allow us to take the measure of the situation. Otherwise we risk needlessly limiting the religious freedom of immigrant minorities, on the strength of our historic institutional arrangements, while sending a message to these same minorities that they by no means enjoy equal status with the long-established mainstream.
Think of the argument of the German Laender that forbade the headscarf for teachers. These are authority figures, surely; but is our idea that only unmarked people can be authority figures? That those whose religious practices make them stand out in this context don’t belong in positions of authority in this society? This is maybe the wrong message to inculcate in children in a rapidly diversifying society.
But the fixation on religion as the problem is not just a historical relic. Much of our thought and some of our major thinkers remain stuck in the old rut. They want to make a special thing of religion, but not always for very flattering reasons.
What are we to think of the idea, entertained by Rawls for a time, that one can legitimately ask of a religiously and philosophically diverse democracy that everyone deliberate in a language of reason alone, leaving their religious views in the vestibule of the public sphere? The tyrannical nature of this demand was rapidly appreciated by Rawls, to his credit. But we ought to ask why the proposition arose in the first place. Rawls’s point in suggesting this restriction was that everyone should use a language with which they could reasonably expect their fellow citizens to agree. The idea seems to be something like this. Secular reason is a language that everyone speaks and can argue and be convinced in. Religious languages operate outside this discourse, by introducing extraneous premises that only believers can accept. So let’s all talk the common language.
What underpins this notion is something like an epistemic distinction. There is secular reason that everyone can use and reach conclusions by, conclusions, that is, with which everyone can agree. Then there are special languages, which introduce extra assumptions, which might even contradict those of ordinary secular reason. These are much more epistemically fragile; in fact you won’t be convinced by them unless you already hold them. So religious reason either comes to the same conclusions as secular reason, but then it is superfluous, or it comes to contrary conclusions, and then it is dangerous and disruptive. This is why it needs to be sidelined.
As for Habermas, he has always marked an epistemic break between secular reason and religious thought, with the advantage on the side of the first. Secular reason suffices to arrive at the normative conclusions we need, such as establishing the legitimacy of the democratic state and defining our political ethic. Recently his position on religious discourse has considerably evolved to the point of recognizing that its “Potential macht die religiöse Rede bei entsprechenden politischen Fragen zu einem ernsthaften Kandidaten für mögliche Wahrheitsgehalte.” But the basic epistemic distinction still holds for him. Thus when it comes to the official language of the state, religious references have to be expunged. “Im Parlament muss beispielsweise die Geschäftsordnung den Presidenten ermächtigen, religiöse Stellungnahmen und Rechtfertigungen aus dem Protokoll zu streichen.” 7
Do these positions of Rawls and Habermas show that they have not yet understood the normative basis for the contemporary secular state? I believe that they are on to something, in that there are zones of a secular state in which the language used has to be neutral. But these do not include citizen deliberation, as Rawls at first thought, or even deliberation in the legislature, as Habermas seems to think from the lines I have quoted. This zone can be described as the official language of the state: the language in which legislation, administrative decrees, and court judgments must be couched. It is self-evident that a law before Parliament couldn’t contain a justifying clause of the type “Whereas the Bible tells us that p.” And the same goes, mutatis mutandis, for the justification of a judicial decision in the court’s verdict. But this has nothing to do with the specific nature of religious language. It would be equally improper to have a legislative clause: “Whereas Marx has shown that religion is the opium of the people” or “Whereas Kant has shown that the only thing good without qualification is a good will.” The grounds for both these kinds of exclusions is the neutrality of the state.
The state can be neither Christian nor Muslim nor Jewish, but by the same token it should also be neither Marxist nor Kantian, nor Utilitarian. Of course, the democratic state will end up voting for laws that (in the best case) reflect the actual convictions of its citizens, which will be either Christian or Muslim, etc., through the whole gamut of views held in a modern society. But the decisions can’t be framed in a way that gives special recognition to one of these views. This is not easy to do; the lines are hard to draw, and they must always be drawn anew. But such is the nature of the enterprise that is the modern secular state. And what better alternative is there for diverse democracies? 8
Now the notion that state neutrality is basically a response to diversity has trouble making headway among “secular” people in the West who remain oddly fixated on religion as something strange and perhaps even threatening. This stance is fed by all the conflicts, past and present, of liberal states with religion, but also by a specifically epistemic distinction: religiously informed thought is somehow less rational than purely “secular” reasoning. The attitude has a political ground (religion as threat), but also an epistemological one (religion as a faulty mode of reason). 9
I believe we can see these two motifs in a popular contemporary book, Mark Lilla’s The Stillborn God. On one hand, Lilla wants to claim that there is a great gulf between thinking informed by political theology and “thinking and talking about politics exclusively in human terms.” 10 Moderns have effected “the liberation, isolation, and clarification of distinctively political questions, apart from speculations about the divine nexus. Politics became, intellectually speaking, its own realm deserving independent investigation and serving the limited aim of providing the peace and plenty necessary for human dignity. That was the Great Separation.” 11 Such metaphors of radical separation imply that human-centered political thought is a more reliable guide to answer the questions in its domain than theories informed by political theology.
So much for the epistemological ranking. But then, toward the end of his book, Lilla calls on us not to lose our nerve and allow the Great Separation to be reversed, 12 which seems to imply that there are dangers in doing so. The return of religion in this sense would be full of menace. 13
This phenomenon deserves fuller examination. Ideally, we should look carefully at the double grounds for this stance of distrust, comment on these, and then say something about the possible negative political consequences of maintaining this stance. But in this chapter I shall only really have space to look at the roots of the epistemological ground.
I think this has its source in what one might call a myth of the Enlightenment. There certainly is a common view that sees the Enlightenment (Aufklärung, Lumières) as a passage from darkness to light, that is, as an absolute, unmitigated move from a realm of thought full of error and illusion to one where the truth is at last available. To this one must immediately add that a counterview defines “reactionary” thought: the Enlightenment would be an unqualified move into error, a massive forgetting of salutary and necessary truths about the human condition.
In the polemics around modernity, more nuanced understandings tend to get driven to the wall, and these two slug it out. Arnold’s phrase about “ignorant armies clashing by night” comes irresistibly to mind.
But what I want to do here, rather than bemoaning this fact, is to try to explain what underlies the understanding of Enlightenment as an absolute, unmitigated step forward. This is what I see as the “myth” of the Enlightenment. (One can’t resist this jab, because “myth” is often cited as what Enlightenment has saved us from.)
This is worthwhile doing, I believe, because the myth is more widespread than one might think. Even sophisticated thinkers, who might repudiate it when it is presented as a general proposition, seem to be leaning on it in other contexts.
Thus there is a version of what Enlightenment represents which sees it as our stepping out of a realm in which revelation, or religion in general, counted as a source of insight about human affairs into a realm in which these are now understood in purely this-worldly or human terms. Of course, that some people have made this passage is not what is in dispute. What is questionable is the idea that this move involves the self-evident epistemic gain of our setting aside consideration of dubious truth and relevance and concentrating on matters that we can settle and that are obviously relevant. This is often represented as a move from revelation to reason alone (Kant’s “blosse Vernunft”).
Clearer examples are found in contemporary political thinkers, for instance, Rawls and Habermas. For all their differences, they seem to reserve a special status for nonreligiously informed reason (let’s call this “reason alone”), as though a) this latter were able to resolve certain moral-political issues in a way that can legitimately satisfy any honest, unconfused thinker and b) where religiously based conclusions will always be dubious and, in the end, only convincing to people who have already accepted the dogmas in question.
This surely is what lies behind the idea I mentioned earlier, entertained for a time in different form by both thinkers, that one can restrict the use of religious language in the sphere of public reason. We must mention again that this proposition has been largely dropped by both, but we can see that the proposition itself makes no sense, unless something like (a) + (b) is true. Rawls’s point in suggesting this restriction was that public reason must be couched in terms that could in principle be universally agreed upon. The notion was that the only terms meeting this standard were those of reason alone (a), while religious language by its very nature would fail to do so (b).
Before proceeding farther, I should just say that this distinction in rational credibility between religious and nonreligious discourse, supposed by (a) + (b), seems to me utterly without foundation. It may turn out at the end of the day that religion is founded on an illusion and hence that what is derived from is it less credible. But, until we actually reach that place, there is no a priori reason for greater suspicion being directed at it. The credibility of this distinction depends on the view that some quite “this-worldly” argument suffices to establish certain moral-political conclusions. I mean “satisfy” in the sense of (a) it should legitimately be convincing to any honest, unconfused thinker. There are propositions of this kind, ranging from “2+2=4” all the way to some of the better-founded deliverances of modern natural science. But the key beliefs we need, for instance, to establish our basic political morality are not among them. The two most widespread this-worldly philosophies in our contemporary world, utilitarian and Kantianism, in their different versions, all have points at which they fail to convince honest and unconfused people. If we take key statements of our contemporary political morality, such as those attributing rights to human beings as such, say the right to life, I cannot see how the fact that we are desiring/enjoying/suffering beings, or the perception that we are rational agents, should be any surer basis for this right than the fact that we are made in the image of God. Of course, our being capable of suffering is one of those basic unchallengeable propositions in the sense of (a), as our being creatures of God is not, but what is less sure is what follows normatively from the first claim.
Of course, this distinction would be much more credible if one had a “secular” argument for rights that was watertight. And this probably accounts for the difference between me and Habermas on this score. He finds this secure foundation in a “discourse ethic,” which I unfortunately find quite unconvincing.
The (a) + (b) distinction, applied to the moral-political domain, is one of the fruits of the Enlightenment myth, or perhaps one should say it is one of the forms that this myth takes. It would be interesting to trace the rise of this illusion through a series of moves that were in part well-founded and in part themselves grounded on illusions. In another essay I identified three, of which the first two are relatively well traced and the third requires more elaborate description. 14 I’ll briefly mention the first two here.
First comes (1) foundationalism, which one sees most famously with Descartes. This combines a supposedly indubitable starting point (the particulate ideas in the mind) with an infallible method (that of clear and distinct ideas) and thus should yield conclusions that would live up to claim (a). But this comes unstuck, and in two places. The indubitable starting points can be challenged by a determined scepticism such as we find in Hume, and the method relies much too much on a priori argument and not enough on empirical input.
But even though his foundationalism and his a priori physics were rejected, Descartes left behind (α) a belief in the importance of finding the correct method and (β) the crucial account that underpins the notion of reason alone. He claimed to be prescinding from all external authority, whether emanating from society or tradition, whether inculcated by parents or teachers, and to rely only on what monological reason can verify as certain. The proper use of reason is sharply distinguished from what we receive from authority. In the Western tradition this supposedly external imposition comes to include, indeed to find its paradigm in, religious revelation. As the marquis de Condorcet put it in his account of the progress of the human mind,
Il fut enfin permis de proclamer hautement ce droit si longtemps méconnu de soumettre toutes les opinions à notre propre raison, c’est-à-dire d’employer, pour saisir la vérité, le seul instrument qui nous ait été donné pour la reconnaître. Chaque homme apprit, avec une sorte d’orgueil, que la nature ne l’avait pas absolument destiné à croire sur la parole d’autrui; et la superstition del’Antiquité, l’abaissement de la raison devant le délire d’une foi surnaturelle disparurent de la société comme de la philosophie. 15
Our reasoning power is here defined as autonomous and self-sufficient. Proper reason takes nothing on “faith” in any sense of the word. We might call this the principle of “self-sufficient reason.” The story of its rise and its self-emancipation comes to be seen as a kind of coming of age of humanity. As Kant put it, not long after Condorcet wrote, Enlightenment is the emergence of human beings from a state of tutelage for which they were themselves responsible, a “selbstbeschuldigte Unmündigkeit” (a self-responsible nonage). The slogan of the age was sapere aude! Dare to know. 16
The first crucial move is that to self-sufficient reason. The second (2) was to point to natural science as a model for the science of society, the move we see in Hobbes, for instance. I shall not pursue this further here because reductive views of social science have less credibility today, although they are, alas, still present on the scene.
This whole matter deserves much further consideration, more than I can give it here. But I am convinced that this further examination would lend even more credibility to the diversity concept I have been proposing, which amounts to this: What deserve to be called “secularist” regimes in contemporary democracy have to be conceived not primarily as bulwarks against religion but as good-faith attempts to secure the three (or four) basic goals I have outlined. And this means that they attempt to shape their institutional arrangements, not to remain true to hallowed tradition, but to maximize the basic goals of liberty and equality between basic beliefs.

NOTES
1 . Rawls would talk here of “comprehensive conceptions of the good.” See John Rawls, Political Liberalism, exp. ed. (New York: Columbia University Press, 2005).
2 . Christian Smith, The Secular Revolution (Berkeley: University of California Press, 2003).
3 . Marcel Gauchet, La Religion dans la Démocratie (Paris: Gallimard, 1998), pp. 47–50.
4 . Rousseau, who laid bare very early the logic of this idea, saw that a democratic sovereign couldn’t just be an “aggregation”; it has to be an “association,” that is, a strong collective agency, a “corps moral et collectif” with “son unité, son moi commun, sa vie et sa volonté.” This last term is the key one, because what gives this body its personality is a “volonté générale.” Contrat Social (Paris: Garnier Flammarion, 1966), book 1, chapter 6, p. 52.
5 . I have discussed this relation in “Les Sources de l’identité moderne,” in Mikhaël Elbaz, Andrée Fortin, and Guy Laforest, eds., Les Frontières de l’Identité: Modernité et postmodernisme au Québec (Sainte-Foy: Presses de l’Université Laval, 1996), pp. 347–64.
6 . See Charles Taylor, Modern Social Imaginaries (Durham: Duke University Press, 2004).
7 . Jürgen Habermas, Zwischen Naturalismus une Religion (Frankfurt: Suhrkamp, 2005), p. 137. Of course, Habermas is right: official language in diverse democracies must avoid certain religious references (although this shouldn’t be stretched to include assembly debates), but this is not because they are specifically religious, but rather because they are not shared. It would be just as unacceptable for, say, legislation to be justified by a “whereas” clause referring to an atheist philosophy as by such a clause referring to the authority of the Bible.
8 . I am not sure whether I am disagreeing with Habermas or whether the difference in formulation really amounts to a difference in practice. We both recognize contexts in which the language of the state has to respect a reserve of neutrality and others in which freedom of speech is unlimited. We differ perhaps more in our rationales than in the the practice we recommend.
9 . Sometimes the obligation of citizens to address their compatriots in the language of secular reason is grounded in an obligation to make one’s position intelligible to them. “The self-understanding of the constitutional state has developed within the framework of a contractualist tradition that relies on “natural reason”, in other words soely on public arguments to which all persons are supposed to have equal access.” Jürgen Habermas “Religion in the Public Sphere,” p. 5. But what reason is there to think that “natural reason” offered us a kind of ideological Esperanto? Were Martin Luther King’s secular compatriots unable to understand what he was arguing for when he put the case for equality in biblical terms? Would more people have got the point had he invoked Kant? And, besides, how does one distinguish religious from secular language? Is the Golden Rule clearly a move in either one or the other?
10 . Mark Lilla, The Stillborn God: Religion, Politics, and the Modern West (New York: Knopf, 2007), p. 5.
11 . Ibid, p. 162.
12 . Ibid, pp. 305–6.
13 . Habermas is an exceptional figure; in many respects, of course, but here I want to point out that although he is a major thinker in the epistemological distinction religion/reason (for which I criticize him further on), he most emphatically does NOT share the political mistrust of religion which often goes with this.
14 . See “Blosse Vernunft,” in Charles Taylor, Dilemmas and Connections (Cambridge: Harvard University Press, 2011), chapter 11.
15 . Nicolas de Condorcet, Esquisse d’ un tableau historique des progrès de l’esprit humain (Paris: Flammarion, 1988), p. 225. I have learned a gret deal from the interesting discussion in Vincent Descombes, Le raisonnement de l’ ours (Paris: Seuil, 2007), pp. 163–178.
16 . Immanuel Kant, “Was ist Aufklärung?” in Kants Werke, Akademie Textausgabe (Berlin: Walter de Gruyter, 1968), 8:33.