Support First Things by turning your adblocker off or by making a  donation. Thanks!


Nietzsche claimed that if men took God seriously, they would still be burning heretics at the stake. In the same spirit, one supposes, are the notions that if men really cherished moral truth, they would suppress all beliefs that they considered wrong, and that if men still cared about the sanctity of the marriage bed, they would go back to making adulterers wear the scarlet A.

Today two different groups of people agree with conditional statements of this sort. In the first group are the ordinary bigots, who are always among us. The second are a kind of modern backlash—call it the reaction—found principally among the “cultural elite.” For instance, whereas the bigots respond to Nietzsche’s conditional by saying, “Yes, that’s why we should burn heretics,” the reactionaries respond to it by saying, “No, that’s why we should suppress the public expression of belief in God.”

These reactionaries claim to love tolerance, but, misunderstanding it, they strangle it in their embrace. Their creed is that intolerance is born at the same moment as public moral commitments; that morality must therefore be a “private” affair; that in order to say that tolerance is a good, we must forbear to say aloud that anything else is good or evil. Their god is Neutrality. In certain intellectual regions he travels under other names such as Autonomy and Rights.

We meet this jealous and negating god on the philosophic right, where conservatives like Michael Oakeshott tell us that the specific and limited activity of “governing” has “nothing to do” with natural law or morals. We encounter him on the philosophic left, where liberals like John Rawls and Marxists like Jurgen Habermas invent devices like the Veil of Ignorance and the Ideal Speech Situation to convince us that if we wish to understand truly the principles of justice, we must pretend to forget not only who we are, but also everything we ever thought we knew about good and evil. We meet this god in law, where many jurists treat ethical distinctions such as “family” vs. “non-family” as “invidious classifications” that deny citizens the equal protection of the law. We meet him in education, where elementary school children are offered books like Daddy’s Roommate, Heather Has Two Mommies, and Gloria Goes to Gay Pride. In fact, we meet this god everywhere: in the university, in the movie theatre, in many churches and synagogues, and, it goes without saying, on the even more ubiquitous altar of the television.

It might seem remarkable that people who insist that tolerance means moral neutrality should themselves be so earnest in ridiculing those who aren’t neutral. But of course, they themselves aren’t neutral either. The scandal of Neutrality is that its worshipers cannot answer the question “Why be neutral?” without committing themselves to particular goods—social peace, self-expression, self-esteem, ethnic pride, or what have you—thereby violating their own desideratum of Neutrality. Yet even this is merely a symptom of a deeper problem, namely, there is no such thing as Neutrality. It isn’t merely unachievable, like a perfect circle; it is unthinkable and unapproachable, like a square circle. Whether we deem it better to take a stand or be silent, we’ve offended this god in the very act of deeming.

To see the folly of neutralism is one thing; to escape from it is another. Many who understand perfectly well that tolerance cannot be defended by suspending judgment about goods and evils have difficulty defending it in any other way. They suspect the worst: that if neutrality is a square circle, then so is tolerance, along with all of its component virtues like objectivity and fairness. They fear that by leaving the reactionaries, they will join with the ordinary bigots. They are right to fear this trap, but make the difficulty of avoiding it greater than it is.


Properly understood, what is this thing called tolerance? What does it really demand of us? What do we need in order to come by it? Let’s see if we can work this out.

The truth is, we already know the answer in part. To tolerate something is to put up with it even though we might be tempted to suppress it. The next step, then: which things are we tempted to suppress? Here, too, we know the answer: we are tempted to suppress those things that we deem mistaken, painful, wrong, harmful, offensive, or in some other way unworthy of approval.

Now, we use the term “temptation” when our hearts solicit us either to do something we ought not do or to forbear from doing something we ought to do. But shouldn’t we suppress the things that we deem mistaken, painful, wrong, harmful, offensive, and so on? The answer is, sometimes we should, and sometimes we shouldn’t. For instance, it’s certainly not acceptable to tolerate the act of rape. But it is certainly right to put up with the profession, by rational argument, of opinions that we deem mistaken.

What makes these two cases different? In one respect, of course, they are just the same. Whichever we endeavored to suppress—rape, or the profession of false opinions—we would presumably be trying to avert particular evils, or, if you prefer, to protect particular goods. The goods that are injured in the act of rape include the dignity of the woman, her physical and emotional well-being, and the integrity of a certain pattern of relationships between men and women, a pattern that depends on trust rather than fear. Similarly the goods that are injured in the act of professing false opinions include the clear knowledge of the truth, the public recognition of its value, and the integrity of a certain pattern of conduct, a pattern that depends on right judgment rather than error. By suppressing rape we would be trying to protect the first set of goods; by suppressing the profession of false opinions, the latter.

And so it is with every case. People may not agree about what is good and what is evil; or they may be mistaken about what is good and what is evil. They may even call evil good, and good, evil. But every time someone wants to suppress something, we can be sure he is attempting to prevent what he thinks, rightly or wrongly, to be evil; alternately, to protect something he thinks, rightly or wrongly, to be good. Why, then—to return to the question—do we sometimes tolerate an evil, or put up with an injury to a good?

The answer is not, as some people hold, skepticism. Take the case of conducting a debate, a practice notorious for tolerating the evil of false opinions. There are three kinds of skeptics: the utter skeptic, who is in doubt about all things; the partial skeptic, who is in doubt about some things; and the non-skeptic, who is in doubt about nothing whatsoever. Can the utter skeptic say, “Because all is in doubt, all should be heard”? No, he has rather to say, “The rightness of hearing and the rightness of silencing are equally in doubt; I cannot tell you which to choose.” Can the partial skeptic say, “I don’t believe in debate”? Yes, but he might instead say, “I am in doubt about what is true in general. But I’m sure that truth is good, I’m sure that exchange of discursive reasoning will help to find it, and I’m sure about the kind of manners that such exchange requires.” And third, can the non-skeptic say, “To put up with falsehood is wrong”? Yes, but he might instead say this: “Although I am sure of every truth, one such truth is that the exercise of rebutting error will sharpen the insight I already possess, and another is that it may convert my opponent.” We see that only the non-skeptic and the partial skeptic can deal with debate. Neither does it because of what he doubts; each does it because of what he doesn’t doubt. Each tolerates falsehood for the sake of truth.

Once we see this, the real reason that we sometimes tolerate evils or put up with injuries to good becomes clear: We do it to prevent graver evils, or to advance greater goods. For there is a certain paradox in this business of suppressing evils: the act of suppression itself may give rise to evils. In fact it often does. Because this is so, we must always put the two evils, the evil that suppression engenders and the evil that it prevents, on a scale. When the evil that suppression engenders equals or exceeds the evil that it prevents, we ought to put up with the thing in question instead of suppressing it.

This paradox is the very basis of the virtue of tolerance. But on closer examination we see that there are two cases, in one of which the paradox is more pronounced than the other.

The less paradoxical case is one in which the goods that are protected by suppression and those injured by suppression are different. For example, at the same time that we consider suppressing the profession of false opinions for the sake of truth, we might consider tolerating their profession for the sake of peace. Truth and peace are both goods, but of course they are not the same good. First we have to decide which of the two goods is of higher order, because that one trumps the other. If they are of the same order, then we must resort to judgments of degree. One need not suppose that judgment has mathematical precision; only that there is such a thing as judgment.

In the other, more paradoxical, case, the goods that are protected by suppression and the goods that are injured by suppression are the same. For example, at the same time that we consider suppressing the profession of false opinions for the sake of truth, we might consider tolerating their profession—at least their profession by rational argument—also for the sake of truth. On the side of suppression we might plead, “After all, the opinions in question are false, aren’t they? Then isn’t it a gain to get rid of them?” But on the side of toleration, we might ask, “But what better engine have we for honing truth than to try it against error in a fair fight?” In this case we don’t have to decide which of two goods is of higher order, because there is only one good at issue. But we do have to compare different hypotheses about what really promotes that good.

Each of these cases reveals a different element in the practice of tolerance: the first, in which the good that suppression protects is different from the one it injures, holds up right judgment in the protection of greater ends against lesser ends; the second, in which the goods protected by suppression and injured by suppression are the same, holds up right judgment in the protection of ends against mistaken means.

In the end, element two goes more to the heart of the matter. The ends with which a given end comes into conflict are largely a matter of circumstance. The question of by which means a given end cannot, by its nature, be pursued, is a constant.

If the constant element in the practice of tolerance is right judgment in the protection of ends against mistaken means, then the constant element in intolerance is false judgment in the protection of ends against mistaken means.

Right away we see that intolerance shows itself in two different ways, for we can err in either of two different directions. One way is by an excess of indulgence—putting up with something we should suppress. Let’s call this the error of softheadedness. The other way that we can err is by a deficiency of indulgence—suppressing what we should put up with. Let’s call this the error of narrowmindedness. Each of these two opposite errors is a deviation from true tolerance; each of them therefore has the same claim to the name of “intolerance.”

This may at first sound odd, because our language has so far reserved the word “intolerance” for narrowmindedness. By contrast, consider courage, which is easier to talk about than tolerance. Although we do sometimes forget that rashness as well as cowardice is opposed to true courage, at least we aren’t burdened by a term, say “in-courage,” that could be applied only to cowardice and not to rashness. It seems that either we prefer the error of softheadedness to the error of narrowmindedness, or else we don’t realize that softheadedness is an error at all. But this is a deep confusion. It is just as much a deviation from true tolerance to put up with rape as it is to suppress the profession of false opinions advanced by rational argument.

Thus far we have three possibilities: two kinds of wrong judgment in the protection of ends against mistaken means, and one kind of right judgment in the protection of ends against mistaken means. Expressed in this way, what we have are three points that appear to be floating in space. Of course there is a flaw in this mental diagram. The three need to be arranged along a continuum which is bounded by two extremes, the one excessive, the one deficient; the one softheaded, the other narrowminded. Because just how indulgent we are toward something—just how much we put up with it, just where we fall between the two extremes—is a matter of degree. For instance, we might make some act a crime. Or we might shun those who do it, but without going so far as criminalization. Or we might try to persuade the wrongdoers to change their ways without going so far as to shun them. Or we might ignore them. Or we might encourage them. We might even reward them. The truly tolerant point will always be somewhere between the two endpoints of the continuum, its location depending on the act in question and on the circumstances. But precisely where it is along this line will vary. The location of true tolerance can be determined only by the exercise of case-by-case judgment about the goods and the evils involved. Just as true courage is a mean between rashness and cravenness, and true friendliness is a mean between obsequiousness and boorishness, so true tolerance is a mean between softheadedness and narrowmindedness.

This leads us to the following propositions.

(1) Tolerance cannot be neutral about what is good, for its very purpose is to guard goods and avert evils.

(2) Tolerance is not a moral rule, a moral attitude, a moral feeling, or a moral capacity, but a moral virtue. Further, although tolerance is not one of the moral virtues that Aristotle discussed, it is a moral virtue of the Aristotelian type. For it is a mean between two opposed vices, one of them characterized by excess and the other by deficiency, its location to be discovered in the case-by-case exercise of practical wisdom.

(3) The circumstantial element in the practice of tolerance is right judgment in the protection of greater ends against lesser ends. This is no different from any exercise of practical wisdom, except insofar as its constant element, right judgment in the protection of ends against mistaken means, makes it special.

To be sure, this is only a formal answer to the question of what tolerance is. It directs us toward the exercise of practical wisdom—of well-founded judgment about the goods and evils involved in putting up with things. To give a substantive answer to the question of what tolerance is, however, would be no small matter; even if one had the necessary wisdom to do it, the space of a single article could never suffice.

Still, the merely formal definition of tolerance does do certain work for us. For instance:

The fact that tolerance is a moral virtue of the Aristotelian type tells us a great deal about its relation to the others. For those of us who wonder how, if at all, tolerance might be taught, this relation carries powerful implications.

In addition, religion presents the acid test for tolerance. For the loyalty that it concerns is ultimate; if tolerance cannot survive it, then tolerance cannot survive. However, the special role of tolerance in protecting ends against mistaken means gives us the one clue we need to unscrew this inscrutable.

Explaining the first of these two points is the burden of the next section. Explaining the second is the burden of the last.


All moral virtues—or at the very least those of the Aristotelian type—are interdependent. The classical demonstration of this truth, which derives from Thomas Aquinas, pivots on the relation of these virtues to practical wisdom. For every moral virtue depends on practical wisdom; hence if practical wisdom is impaired, then every moral virtue is impaired. But on the other side, practical wisdom depends on every moral virtue; hence if any moral virtue is impaired, practical wisdom is impaired. It follows, then, that through practical wisdom, a flaw in any moral virtue entails a flaw in every other.

Think, by way of illustration, of the virtue of courage. Courage involves a mean between fear and daring—as we said before, enough fear to avoid being rash, enough daring to avoid being craven. But because the right balance between fear and daring varies from case to case, the habit of courage must be informed by practical wisdom. But how, on the other side, does practical wisdom depend on every moral virtue? Again, consider courage. In thinking of its exercises, our imaginations usually go no further than pain and death. Aristotle himself thought no evil could be greater. But this is false: to be held in contempt is more fearful, and vice, if it is not, certainly ought to be. To achieve practical wisdom, one needs enough fear to be vigilant of error and enough daring to risk it in pursuit of truth. To hold onto such wisdom, one needs enough fear to dread its loss and enough daring to risk contempt in its defense.

The same relation exists between practical wisdom and every other moral virtue. Using a bicycle wheel as our model, the moral virtues are to spokes as practical wisdom is to the hub. We all know what happens when we use a bicycle wheel with a damaged spoke. Before long, the others give in, too, and the wheel gets more and more out of true. This is the classical thesis of the unity of the virtues. If one virtue bends, then every virtue bends.

Everyday experience corroborates this. We don’t even have to go through the hub, for without implying their equality we may say that all the moral virtues are connected at the rim. Tolerance is addled in the unfriendly man; friendship is addled in the dishonest man; honesty is addled in the unjust man; justice is addled in the loveless man; love is addled in the hopeless man; hope is addled in the impatient man; and patience is addled in the intolerant man. So the wheel is closed.

Thus tolerance is one of the moral virtues, and depends on all the rest of them. This has implications for the cultivation of tolerant citizens. How so? The unity of the virtues works in only one direction. That is, while impairment of one moral virtue entails impairment of all the rest, progress toward one moral virtue does not in and of itself entail progress toward all the rest. Think of the bicycle wheel again. Beginning with a perfect round, bending one spoke will soon bend all the others, too, but straightening one spoke of a crushed wheel will not simply pull the others back in true. In fact it may cause some spokes to bend even more. With damaged bicycle wheels, there are three alternatives. We can replace the wheel; we can take it apart, straighten each part separately, and put it back together; or we can leave it in one piece and straighten every part at once. With a soul, the first two alternatives are out of the question because we can neither replace it nor take it apart. The only alternative is to leave it in one piece and straighten every part at once.

Here is what follows: if all the virtues depend on one another, then tolerance cannot be taught unless all the rest are taught as well. We cannot compensate for the collapse of all our virtues by teaching tolerance and letting the rest go by, as some educators and social critics seem to think; the only cure for moral collapse is moral renewal, on all fronts simultaneously.

That is a hard adage. For even with crushed wheels, the simultaneous straightening of every spoke is hardly thinkable. With crushed souls—which is what we all are—we’ve no idea how our own efforts might bring it to pass. More than education, we need redemption. For virtues are complicated things: complex dispositions of character, deeply ingrained “habits,” by which one calls upon all of his passions and capacities of mind in just those ways that aid, prompt, focus, inform, and execute his moral choices instead of clouding them, misleading them, or obstructing their execution. This means that virtues cannot be imparted just by encouraging certain feelings or developing certain capacities; feelings and capacities are instruments of the virtues, not their realization.

What adds to the difficulty is that virtues are much more than readiness to follow the rules. There are, of course, some rules that are true in all circumstances. Murder is always wrong. But virtues are more like a fitness to distinguish true rules from false, and to choose rightly even where there are no rules or where the rules are no more than rules of thumb and seem to contradict each other. To be sure, if rules are applied judiciously, they can help to restrain the most obvious evils. And this in turn is bound to help in the nurture of virtue. But virtue cannot, as we have said, be taught simply by means of an exhaustive list of rules. Not only would such a list be endless, but the vicious would rebel before we even reached the second page.

In sum, we aren’t going to transmit the virtue of tolerance through a quick fix, like a Freshman Orientation Weekend; through a long fix, like a three-volume set of workplace sensitivity regulations; or through a false fix, like a Children of the Rainbow Curriculum. If Plato was right that justice is medicine for the soul, then such instrumentalities are its patent medicine: 5 percent poison, 10 percent flavoring, 85 percent intoxicating spirits, and pure delusion from the first to the last.


Time now to turn to the question of religious tolerance, where even the rules are far from easy to discern.

What is religion anyway? Some people say that all religions depend on faith, while all secularisms depend on reason. But as Chesterton remarked in Orthodoxy, “It is idle to talk always of the alternative of reason and faith. Reason is itself a matter of faith. It is an act of faith to assert that our thoughts have any relation to reality at all.” Other people say that all religions believe in God, while all secularisms do not. But though Buddhists do not believe in God, yet we call Buddhism a religion.

Still others, like Tillich and Niebuhr, hold the mark of religion to be the practice of ultimate concern that orders all other concerns, unconditioned loyalty that trumps all other loyalties. Here we finally hit the mark. For Christians, the ultimate concern is the saving God of Abraham, Isaac, and Jacob who has revealed himself in Messiah. Though Buddhists do not believe in the God of Abraham, Isaac, and Jacob, much less in Messiah, they do have an ultimate concern—escape from suffering, inherent in desire, which, they hold, springs in turn from the illusion of existence.

But if religion is the practice of ultimate concern, then we have another problem. In the first place, even a secularism may be the practice of an ultimate concern. We acknowledge this, for instance, by calling Leninism a religion; similarly we say of a greedy man that “his god is money” and call misplaced devotion “idolatry.” In the second place, even among those secularisms that do not go so far as to identify ultimate concerns, none is without implications as to what could, or could not, count as an ultimate concern. John Stuart Mill could never decide which, if any, of the “permanent interests of man as a progressive being” was deserving of unconditioned loyalty. But one thing he was sure of, that Messiah was not among them.

What all this tells us is that “religious” and “secular” constitute a false dichotomy. We would do better with a trichotomy. An acknowledged religion like Christianity or Buddhism posits an ultimate concern and admits it. An unacknowledged religion like Leninism posits an ultimate concern but denies that so doing is religious. And an incomplete religion like Millianism has not finished ranking its concerns.

Incomplete religion can live only in the dreamworld of thought. In the light of day it must become complete or die. For in every life or way of life—whether lived simply, lived with the guidance of an ethical theory, or even lived in defiance of an ethical theory—given enough time, some concern eventually emerges as paramount. Eventually there is something to which every knee bows. This is the person’s god. As a matter of theory, one may deny that any concern deserves ultimacy. But as a matter of practice, no one escapes ceding ultimacy to something, whether it deserves ultimacy or not. Choices between incompatible urgencies are unavoidable. To prevent the rise of one or another of these urgencies to supremacy, a person would have to practice a truly Stoic discipline of contradiction—and in the end we would have to ask what urgency he served in so disordering himself. In short, one need not be conscious of his god, or even conscious that he has a god. One might think he has no god, or that he is “looking for” or “waiting for” a god. One may even be converted from one god to another. But one will have a god—or at least be on the road to having one.

With all of this ultimate concern floating about, how can there be religious tolerance at all? The answer is, there can’t be—unless one’s ultimate concern commands it, or at least allows it. For in this case and this case alone, tolerance toward other claimants to ultimacy is obedience to one’s own.

Thus St. Hilary of Poitiers: “God does not want unwilling worship, nor does He require a forced repentance.” The idea is that although God demands and deserves our unconditioned loyalty, He is of such a nature that nothing exacted by threats could truly serve Him. For He desires sons and daughters, not slaves: His love is inexorable and consumes everything contrary to itself. This is not the Kantian idea that choice is lovable but rather the Christian idea that love is chosen. I do not say that His supposed followers have always practiced the loving tolerance He demands. I do say that intolerance stands under His judgment.

But notice: the same consuming fire that for its own sake demands tolerance, for its own sake sets the limits to what is tolerated. If Hilary was right that God does not want unwilling worship, then Hilary’s tolerance must be absolute with respect to permitting belief in other gods. This does not mean permitting every act of service to these gods. Hilary must claim the right to say that there are evil services which nothing deserving of unconditioned loyalty could demand, and the correlative right to try to stop anyone who attempts them. For instance, whatever claims of conscience Hilary may honor he cannot permit a person to plead them in justification of murder. “God told me to kill anyone who got in my way” cuts no ice with him; nor is the case different when other ultimate concerns, other gods, are pleaded in place of God. The Defense of the Revolution, The Greater God of the Whole, The Purity of the Race, the Hunger of Moloch, The Right to Control One’s Body—neither these nor any other claimants to ultimacy are accepted as justifying the sacrifice of innocents. “Even conceding your God-given right to be left alone by me in your honor to another god,” I imagine Hilary saying, “that right concerns your own soul only. I will not permit you, in its service, to inflict injuries which my own God abhors and forbids.”

My example is Christian because I am a Christian. But the logic works just the same if you posit some other ultimate concern, some other god than mine. For instance, the god of the Benthamite utilitarian is “aggregate pleasure.” Hence if the Benthamite could tolerate other creeds at all, such tolerance would be both ordained and limited by the requirements of such pleasure. Likewise, religious tolerance for the Millian utilitarian would be both ordained and limited by the nature of man’s “permanent interests” as a “progressive being,” and religious tolerance for the Leninist would be both ordained and limited by the needs of “proletarian dictatorship.”

One might suppose that this logic works only for so-called teleological creeds, said to give priority to achieving the good over doing the right. This is not so. No recent writer has more sternly insisted on the priority of right over good than John Rawls. Yet even he has an ultimate concern. His concern is “autonomy,” the conditions for the realization of which are supposedly determined by choices made behind a Veil of Ignorance that obliterates personal memory. But the conclusion is obvious: For the Rawlsian, religious tolerance is both ordained and limited by what people could want who no longer remembered the love of God.

Where does all this leave us? The bottom line is that Neutrality is no more coherent in the matter of religious tolerance than it is in tolerance of any other sort. What you can tolerate pivots on your ultimate concern. Because different ultimate concerns ordain different zones of tolerance, social consensus is possible only at the points where these zones overlap. Note well: The greater the resemblance of contending concerns, the greater the overlap of their zones of tolerance. The less the resemblance of contending concerns, the less the overlap of their zones of tolerance. Should contending concerns become sufficiently unlike, their zones of tolerance no longer intersect at all. Consensus vanishes.

This, I believe, is our current trajectory. The embattled term “culture war” is not inflammatory; it is merely exact. And we can expect the war to grow worse. The reason for this is that our various gods ordain not only different zones of tolerance, but different norms to regulate the dispute among themselves. True tolerance is not well tolerated. For although the God of some of the disputants ordains that they love and persuade their opponents, the gods of some of the others ordain no such thing.

J. Budziszewski, a new contributor to First Things, teaches in the Department of Government at the University of Texas at Austin. He is author, most recently, of True Tolerance: Liberalism and the Necessity of Judgment.