Home >> Volume 1, Issue 01


Lydia McGrew

There is a peculiar separation of religion and public policy that many, including some Christians, believe we are required to maintain. And such a separation can be made to sound speciously plausible. The reasoning runs something like this: “You wouldn’t want to be ruled by someone else’s religious beliefs, so you shouldn’t try to rule other people by your own religious beliefs. You should try instead to make all your public policy recommendations in such a way that even non-religious people or people from other religious backgrounds can see that they are reasonable.” Such intuitions can be made into fairly strong restrictions on virtuous public action. Consider, for example, the following principle, which we may call a Generic Naked Public Square (GNPS) principle:

GNPS: The promotion (by vote or other means) of a given public policy is civically virtuous only if the policy in question is sufficiently supported by non-religious reasons.

Before we say more about what is wrong with this principle, we should find the most plausible version of it. And in order to do that, we shall have to define its most important term—‘religious’. For it is obvious that a great deal turns on that. There could easily be invidious or overly-broad definitions of ‘religious’ that would make the principle ludicrously false, and we should set these aside before we go any further.

Robert Audi, for example, says that a reason or argument is “historically religious” if the cognitive causal chain that lies behind one’s holding it involves belief in some proposition that has obviously religious content (Audi, 2000, 74). The causal chain need not be a chain of reasons; Audi also includes beliefs and associations. But this is obviously wrong. Suppose that a drug addict hears a stirring proclamation of the Christian gospel and feels guilty about his drug abuse as a result. He worries that there may be a God who disapproves, and he goes out, gets his life together, and begins to research arguments against drug legalization. He becomes convinced that hard drugs should be illegal because of all the misery they cause, and he begins to advocate stiffer penalties for drug dealers. But he would not be making these arguments were it not for a causal chain that began with a religious sermon. Still, the arguments he is making are entirely secular, and there is no reason to think of them as anything else. This purely causal sense of ‘religious’ is a non-starter.

Another attempt to define ‘religious’ in such a way as to cut out arguments without reference to their content or cogency might be to point out that in some cases an audience would not find a given argument persuasive if they did not hold some religious beliefs (Audi, p. 75); hence, the speaker may be counting on the audience’s religious beliefs to help make his presentation rhetorically convincing. In those circumstances one might argue that the argument—meaning not a set of propositions supporting a conclusion but an argument as made at a particular time--is religious. But again, this sense of ‘religious’ appears to involve an obvious error. Suppose, by analogy, a young man is explaining to his rather unintellectual girlfriend, using impeccably good arguments, the scientific point that simultaneity is relative to one’s frame of reference. He may know that the conclusion is counterintuitive and that his girlfriend would dismiss it as nonsense if she were not romantically attracted to him, but this does not make it sensible or useful to call his scientific argument “romantic.”

Yet again, it might be proposed that one’s argument should be deemed religious if one’s motive in making it is religious—say, to promote the spread of Christianity—regardless of the nature, cogency, or content of the argument (p. 73). But this, too, seems completely wrong. If a man makes a knock-down argument in physics in order to win a bet which, if won, will garner him a free dinner, his argument should not on those grounds be called “gastronomic.” And since Christians are supposed to do everything they do to the glory of God, this motivistic definition of ‘religious’ would make all their activities religious. While such a designation may well be right in a devotional sense, it is hardly useful in the present context, especially if we are concerned (as Audi is) with proposals that our fellow citizens may resent as religious impositions. A non-religious citizen has no more right to resent one’s policy proposals solely because one’s motives are religious than a student has a right to resent his Christian professor’s teaching him the Pythagorean Theorem in order to glorify God.

Having set aside these attempts to define arguments as religious without reference to content, we should consider one somewhat more interesting but still misguided attempt to define a policy argument as religious, this time with reference to its content. Someone might argue, or at least imply, that a policy proposal is religious if it depends crucially upon a deontological moral prohibition rather than upon consequentialist or utilitarian considerations.[1] This concept of what count as secular and religious reasons is rarely stated explicitly but often seems implicit in policy discussions, sometimes even on the side where one might most expect to see it challenged. For example, when political conservatives discuss comprehensive sex education, they focus almost entirely on its bad consequences. Undoubtedly, such indoctrination of the young has devastating consequences—physical, psychological, and social—and there is nothing wrong with emphasizing them in presenting the conservative side of the issue. But often it seems that conservatives are assuming (partly because they are sure that their opponents will assume) that such arguments from consequences are the only ones they are allowed to make, that they are not allowed to bring up, for example, the simple point that sexual intercourse outside of marriage is morally wrong.

One particularly unfortunate application of the idea that we must always argue from consequences is the use of slippery-slope arguments against actions that are in themselves gravely immoral. When scientists propose that we abandon the dead donor rule and take vital organs from undeniably living human beings, with safeguards to guarantee that only willing donors are involved, our first cry should not be, “How can we know that the safeguards will be applied?” To make that our first line of response is tacitly to give up on fighting the initial proposal on its own ground, to imply that it is bad chiefly insofar as it leads to something worse, a hypothetical bottom of the ethical barrel that day by day recedes before us. But if nothing is bad in itself, then there is no point in arguing that some policy is bad because it will lead to some worse evil. By never saying, “This itself is wrong and must be opposed whether it does or does not lead to worse,” we guarantee that our fighting line is always shifting.[2]

There may be more than one reason for avoiding bare statements that certain acts are intrinsically wrong in policy debates.[3] But if the worry is that absolute moral pronouncements are tacitly religious, it should be discarded. To say that something is wrong aside from its consequences does not amount to and need not depend epistemically upon a religious proposition. Absolute statements of moral wrong can be seen to be true in and of themselves. Indeed, secularists are likely to have some things which they themselves consider wrong apart from consequences—e.g., rape, lynching, or slavery—so the attempt to force consequentialism upon religious citizens would represent a double standard.

If we are to have an interesting discussion about religious reasons in the public square, much less about the possibility that their role should be a limited one, there is really only one definition of ‘religious’ that has any plausibility or potential relevance, and it is one that goes approximately like this:

An argument for a policy is religious if and only if it depends for its cogency either explicitly or implicitly upon at least one proposition about God, gods, life after death, or supernatural beings or forces.

On this definition of ‘religious’, the GNPS principle is saying that virtuous citizens must have sufficient support for their policy proposals from arguments which do not rely on religious considerations, where ‘religious’ is understood in more or less the ordinary sense of the term.

Why would anyone think that this principle is true? The rough sketch of an argument I gave for it at the outset appears to rely on the Golden Rule: If you would not want someone else to impose his religious standards on you, you should not try to impose your religious standards on someone else. But this argument is open to an immediate objection. A major reason, perhaps the only reason, why many of us would not want other people to impose their religious standards on us is that we think their religions false, not that there is something special about religion.[4] The argument cannot get any traction at all if we do not assume that we are talking about standards that differ from one religion to another. For example, Muslims believe that they are obligated to fast during Ramadan; Christians do not. But if we are simply talking about differences of opinion about what is true and false, then the principle becomes trivial: “You would not want other people to impose their unnecessary codes of conduct, which they falsely take to be obligatory, upon you; therefore you should not impose your unnecessary code of conduct, which you falsely believe to be obligatory, upon others.” But if we believed our own religion to be false, we would not be making an argument from it to some policy proposal. And almost any policy proposal, regardless of the reasons that support it, will find some dissenters who believe that others’ false standards are being imposed upon them. So the Golden Rule argument turns out to have very little to tell us about religion, specifically.

Robert Audi lays out a number of considerations that he regards as reasons for treating religion specially in its influence upon public policy (2000, 100-105), but they are all exceedingly unsatisfying as justifications for even a relatively weak naked public square thesis, much less Audi’s much stronger version.[5] Almost all of them rely on generalizations about what religious people tend to be like, combined with the assertion that these characteristics mark differences between religious people and secular people. Audi lists the danger of fanaticism, the willingness to coerce even the virtuous to conform, condemnatory tendencies, the dangers of an inflated sense of self-importance among their leaders, passionate concern with making others act in accordance with their religious norms, and anger with children who leave or do not fully conform with the faith in adulthood.

It is sadly amusing to read this list and to consider how well its negative aspects apply to secular people and movements. Communism, for example, is as fanatical as any conventional religion and demands group-think on an unrivaled scale. Contemporary feminism aspires to control worldview, language, and behavior. The New Atheists are exceedingly passionate about making people behave in accordance with their own beliefs (making sure children are taught Darwinism as unquestioned fact, for example), and Richard Dawkins and P.Z. Myers have an inflated sense of self-importance that would make many a Christian megachurch pastor look modest by comparison. Dawkins is infamous for having repeatedly and insistently called a religious upbringing “child abuse,” and while Dawkins has shied away from the obvious legal implications of this accusation, not everyone who thinks as he does is so cautious. Other secularists, self-styled “comprehensive liberals,” have expressly advocated the use of the power of the state to monitor and limit parents’ ability to transmit their religion to their children (see Hitchcock, 2004). As for the vicious condemnation of children who do not fully conform to their parents’ secular ideology, a good example of this phenomenon is the strange story* of Rebecca Walker, daughter of feminist icon Alice Walker. And, on the other hand, there are plenty of religious people who do not display such negative characteristics. It simply does not appear to be true that we reduce fanaticism, self-important leadership, attempts at thought control, and the like in society by reducing the role of religion in public life.

Audi’s two remaining reasons are that religious liberty is an important and delicate thing and that permitting religious reasons for public policy would plausibly lead to coercing expressly religious behavior—prayer, for example—on the part of people who do not accept a particular religion. As for the second of these, it would certainly be undesirable if people were being coerced to pray to any God, even the true God. But then, secular ideology can and sometimes does demand that we do homage to itself—in the form of changing our language to make it politically correct, for example, or treating two men or two women as “married” in all of our business activities. The problem with forcing people to pray to the true God is that the true God is not truly worshiped in that fashion. The problem with forcing people to pray to false gods and to pledge allegiance to false ideologies is that they are false. You will not avoid the problem of the coercion of conscience by limiting the role of religion in public life. You will only shift that problem so that the unreasonable coercion comes from some quarters rather than others.

But when we come to Audi’s brief discussion of the delicacy and deep importance of religious liberty, we find a rather interesting statement:

Citizens in [a democracy] are naturally and permissibly resentful about coercion by religious factors...in a way in which they are not permissibly resentful concerning coercion by, for instance, considerations of public health. Even the moral errors of others are, for many, easier to abide as supports of coercion than religious convictions having the same result. Perhaps the thought is that one can argue with others concerning their moral or economic or philosophical views in a way one cannot argue with them about their religious convictions.

The first part of this paragraph is simply false, and it is enlightening to see that Audi apparently considers policy based even on true religious premises to be more objectionable than policy based on secular moral errors. A slave owner would not be permissibly resentful of the emancipation of his slaves on the grounds that their emancipation had come about as a result of religious arguments. A parent, on the other hand, would be permissibly resentful of the forcible administration to his perfectly healthy child of mind-altering drugs even if such a policy was argued for from secular premises. How permissible (if one means, as Audi must mean, something like “understandable” or “reasonable”) one’s resentment of some law is depends on how reasonable the law is. It does not depend upon the origin of the considerations that brought about the law but rather upon whether the law is good or bad, merely annoying or outrageous, and so forth.

But the more interesting statement comes in the latter part of the paragraph and points us to what is, I suspect, the real reason that many support something like the GNPS thesis: They think that religion is intrinsically irrational or arational. And here we come to what is really the only remotely plausible argument for anything like the GNPS principle, which we can see in the form of a simple syllogism:

  1. People should not be ruled by irrational beliefs.
  2. Religious beliefs are irrational.


  1. People should not be ruled by religious beliefs.

And one can extend this reasoning to the role of religion in public life:

  1. It is worse to force others to act upon your own irrational beliefs than merely to act upon them yourself.


  1. It is worse to force others to act upon your own religious beliefs than merely to act upon them yourself.

Audi says nothing like this. He attributes the thought that religious people cannot be reasoned with to his hypothetical secularists who are filled, he says, with a permissible resentment of laws based on religious considerations. But given the weakness of the stereotype-based considerations Audi brings up to argue for limiting the role of religion in public life, this argument seems at least interesting by comparison. It is especially interesting if we consider the possibility, which no doubt is something many secular people think, that religious beliefs must be irrational, that they are simply not the sort of thing for which rational argument can be given. If that were true, it would at least make religious beliefs as a class different from non-religious beliefs.

But are religious believers bound to admit such a premise? No doubt some of them do. There are plenty of fideistic believers in many different world religions. But as Audi well knows (and this is doubtless why he does not make the above argument in his own voice), Christianity, at least, has a long and illustrious history of apologetic arguments, of which natural theology arguments for theism and for God as the creator of the world and of man are a subset. And even when it comes to the more particular claims of Christianity, historical apologetics defending Christianity by way of arguments for the resurrection of Jesus has yielded enough excellent works to fill a library. It would be highly invidious, not to mention uninformed, simply to state as obvious that religious belief per se is not susceptible of rational support.

What if a citizen’s religious beliefs are well-supported rationally, and what if they do tend to support some public policy? Why should he not advocate and seek to enact that policy on those grounds? Audi considers this possibility and dismisses it on the basis of another weak argument about what people might resent or not wish to do:

This pathway to moral truths, however impeccable logically, still runs through religious territory, and some rational non-religious citizens in a liberal democracy might be permissibly unwilling to follow it as an essential path to those moral truths....[N]early all of them would resent being forced to take [this path] or to occupy the position it leads to if they see no other route to that position (p. 127).

Audi’s point is that, if we cannot induce others to believe that the policy is a good one without their going through “religious territory,” and if they don’t want to go through that territory, then we cannot advocate the policy on the basis of that argument, even if it is “logically impeccable.” This must be wrong. Audi’s theory implies that, in order to be virtuous, citizens may be obligated not to act on the basis of a whole class of rationally well-supported beliefs which support some civic action—that is, unless they can find an entirely different type of reason for the action that does not fall into the suspect class. But it should be axiomatic that public policy should be informed by all the evidence at hand. Arbitrary causal neutering of good evidence on the grounds that it is religiously tainted is unacceptable from the perspective of trying to discover the best public policies.

But a question does arise for someone who believes, as I do in fact believe, that the important moral precepts needed for a just public order are accessible by way of the natural light. Why, then, should religious reasons—reasons that make reference to God expressly, for example—be important? Why isn’t my disagreement with Audi largely a moot point? The answer is that it is often easier for people to deceive themselves about moral truth when their views on moral issues are not informed by an acknowledgement of religious truth. Audi’s own evident position on abortion is a case in point. Not only is he unconvinced by all secular arguments for the personhood of the unborn child, he also says that it is “religious or philosophical or both” to call the fetus a child, since being a child entails being a person. Why a claim’s being “philosophical” should be a problem even on his own principles he does not say, but his point in the context is that to call the fetus a child is not scientific, that people think that this claim is supported by scientific evidence only because they are tacitly influenced by their religious beliefs, and that their arguments therefore are “religious” in the sense that the arguments would not be convincing absent religious considerations (Audi, 1997, 60-61, note 35). On the other hand, with a rather charming lack of self-consciousness, he states that the wrongness of enslavement can be directly intuited as a foundational proposition (p. 174).

Our perception of ethical truths is often highly colored by our own time and culture, and it therefore behooves us to use all the sources available to us, including those insights we gain from religion, especially when our religious beliefs are themselves rationally defensible. It may indeed be possible to see directly that it is wrong for one man to enslave and sell another, but historically far more people have thought with Aristotle that it is natural that some should be masters and others slaves, and it was only when the notion that all men are created equal was applied to the question that progress towards abolition was made.

Audi considers, rather chillingly, the question of arguments against mutilation from function. In discussing the argument against foot-binding (2000, 73) from the fact that it makes the foot badly suited to its function of walking, he raises the possibility, while leaving the question open in the end, that we may not be able to give a secular defense of a standard of proper physical function. In the movie The Inn of the Sixth Happiness, Gladys Aylward expressly relates proper function to a theological argument, as she examines the feet of little Chinese girls to make sure they are not bound: “If God had wanted little girls’ feet to be stumpy and horrid, he would have made them that way, wouldn’t he?” While it is often rather easy to see that a particular organ of the body has a particular function by the natural light, there is also the question of why we should care what the function of our organs is. The importance of not behaving in a way contrary to nature is, again, accessible to the secular man, but the secular man does not always want to hear the message, whereas the man who believes that God created mankind in a certain way for a purpose will find it harder to ignore that divine plan. Obviously, the issue of natural function has ramifications beyond the issue of mutilating young girls’ feet, and most especially in sexual matters. But in our own day and age, with the coming of Muslim immigrants and the issue of female genital mutilation arising in the West, it may even be necessary expressly to hold a line against horrific child abuse on the basis of a concept of natural function. If Audi has any real doubts as to whether permanent, severe, physical mutilation of children can be ruled out by secular reason, he should perhaps be willing to consider that religious reasons are important, after all, to the right ordering of public policy.

Just as religiously-routed reasons can bolster and prompt our understanding of moral truths, so too our direct perception of moral truths can be originally prompted by a religious upbringing but take on independent status once we see reality clearly. A person might have been raised by his parents to have abolitionist leanings on religious grounds but, once he sees the wrong of enslaving humans, understand that wrong in a way that has epistemic force independent of those religious considerations. Something similar is doubtless true of present-day life issues. While a person might be raised Catholic and hence pro-life, once he realizes that the unborn child is his fellow human being, he will have shaken off the culture’s pervasive and irrational prejudice against the unborn, and he will then understand the wrongness of killing the unborn child just as he would understand it for a born child, in a direct way that does not rely crucially upon premises with religious content. Religious people may think that they believe some moral proposition solely for religious reasons when in fact this is not the case.

This is one reason for not counting the history or origin of a moral intuition against it. An entire social group can be brought by the work of religiously-motivated people to see a moral truth, but it does not follow that any individual in such a culture who holds that truth as a result of their work would be unjustified in believing it absent the religious premises that began the process. Thus direct intuition of the moral law and religious training are mutually reinforcing, and it would be a mistake to try to make a radical cut between them in practice.

Even granting that some religious people are explicit fideists, and even granting, further, that it is not good to urge public policy on the basis of premises for which one has no good reason, it does not follow that such religious fideists have a special duty immediately to abandon their work to bring about public policies that, they believe, are based on their religious beliefs. It would be perfectly legitimate for them to reexamine their beliefs about public action, especially if the public actions in question have strong independent reasons against them. For example, if one thinks that one’s religion requires suicide bombings, this is a strong argument against either that religion or one’s interpretation of it. If one’s religion requires one, in the name of ritual purity, greatly to inconvenience the helpless in one’s line of work—as, for example, when Muslim taxi drivers refuse to accept blind customers with guide dogs considered unclean—the very fact that one is doing harm to an innocent third party who, not being a member of one’s religion, is not even covered by the ritual rule in question, should make one ask whether the regulation is being interpreted too strictly or is simply silly or incorrect.

But it would be a mistake for a religious believer, even one who believes that he lacks good reasons for his religion, simply to abandon his work towards public policies supported by that religion. One problem with doing so is the point made above—he might have multiple forms of access to the relevant moral truth and hence might be mistaken in thinking that his inclination in a certain direction is just a matter of his “personal religious belief.” And he might be able fairly easily either to make explicit or to acquire non-religious reasons for the policies in question.

Second, in many of the most controversial cases, such as abortion, euthanasia, suicide, and embryo-killing research, we are talking about very serious issues, about what is, according to the conservative position, the direct killing of innocent persons. These are not cases of ritual religious rules such as keeping kosher or avoiding the saliva of dogs. The actions in question are gravely evil for everyone, if they are wrong at all. It is therefore a matter of sheer prudence for a religious believer to be cautious in abandoning his public policy recommendations, even if he thinks both that they are based on his religious beliefs and that his religious beliefs are not well-grounded. It would be better for him both to try to make his religious beliefs better grounded and to consider more carefully his actual understanding of the moral wrongness of an act such as abortion.

And that brings us to the third and most important reason for religious people to refuse to accept the easy out of fideistic belief combined with a naked public square to avoid pushing their merely personal beliefs upon others. This approach encourages a lazy and debilitating contentment with irrationality. The believer who takes the “personally opposed, but...” line is acquiescing in the notion that it is perfectly all right for him to hold beliefs that are silly and irrational so long as he keeps them to himself. But not only is it not all right to believe irrational things, it is also not all right to keep our beliefs about vital questions to ourselves. Murder is not a private matter, and if you think your reasons for considering an act to be murder are so poor and subjective as to make your conclusion merely private, then you would do well to start thinking hard and to decide, on more carefully considered grounds, whether we are talking about murder or not. The issue is too important for woolly thinking. In proposing a naked public square, the secularist patronizes the believer and tells him that he may keep his strange views so long as he keeps them locked up in a box marked “religious,” so that no matter how important his conclusions are, prima facie, to his public actions, they do not actually guide his public actions. But no one should be thus complacent in irrationality, and no one should acquiesce in such a blatantly artificial separation between the sacred and the secular. Religion speaks not only to strictly religious observance but also to matters of justice and right doing between men. If, then, we hold our religion irrationally, we should try to do better, for the guidance of all our actions, both public and private. But we should not rest content to keep our religion childish, unreasonable, and private, while letting the world go to hell unhindered. Rather than cherishing a subjective and poorly-grounded set of religious beliefs and abandoning the body politic to its own devices, we should stand up like men and follow after truth with all our might.


Robert Audi, 1997. (With Nicholas Wolterstorff) Religion in the Public Square. Lanham, MD: Rowman and Littlefield.

Robert Audi, 2000. Religious Commitment and Secular Reason. Cambridge: Cambridge University Press.

James Hitchcock, 2004. “The Enemies of Religious Liberty,” First Things, February, 2004.

[1] Audi does not make this argument. He accepts that some moral prohibitions can be directly intuited (1997, 173-4).

[2] Sometimes this strategy—or its anticipation--becomes darkly ludicrous. When Dutch doctors put forward the Grögen Protocol for the active termination of sick newborn babies, they tried to forestall criticism by saying, “This is not the slippery slope.” To which their critics might well have responded, “No, it isn’t. It’s the bottom of the slippery slope.”

[3] Perhaps one reason for avoiding such absolute claims is an awareness that not everything that is wrong should be illegal. But while this is true, the fact that something is wrong is relevant to the question of whether it should be illegal. If we thought that wife-beating were permissible, we would probably have very different laws about wife-beating.

[4] An adherent of a religion might have arguments against the legal enforcement of all the injunctions even of his own religion; for example, a Catholic might understandably think it a bad idea if fasting on Fridays in Lent were required by law. But the force of the Golden Rule objection seems to lie in the imposition of religious injunctions on those who do not belong to the religion in question. The Golden Rule objection does not appear to be about distinctions among different religiously-supported injunctions, some of which are more appropriate matters for public policy than others, but rather about objections to the imposition of religiously-supported policies because they are come from someone else’s religion.

[5] For example, Audi argues (2000, 96-100, 163-5) that a citizen is not truly civically virtuous, even if he has an adequate secular reason for the policy he proposes (where “secular” means “not religious” in any of Audi’s senses already discussed), unless he would be motivated to promote the public policy by his secular reason alone in the absence of additional religious reasons.