The good of “religion in general.”

In order to offer a coherent account of the religion clauses, I argued in an earlier post, it is necessary to define the Establishment Clause less abstractly than the Court has, in order to permit the special treatment of religion that is mandated by the Free Exercise Clause. I then argued that the answer is, as Justices Rehnquist, Scalia, and Thomas have suggested, that the state may promote “religion in general.” But, I argued, their solution needs modification, because their identification of religion with monotheism is naïve. Religion must be understood more abstractly than this. How ought this to be done?

Religion is a category that is hard to delimit. Most of the best treatments of the problem of defining “religion” for constitutional purposes have concluded that no dictionary definition will do, because no single feature unites all the things that are indisputably religions. Religions just have a “family resemblance” to one another. In doubtful cases, one can only ask how close the analogy is between a putative instance of religion and the indisputable instances.

Any solution to the free exercise/establishment dilemma that allows accommodations must permit the state, at some level of abstraction, to recognize the value of religion. Thus far Scalia is correct. In order to avoid the problem of sectarian preference, however, religion must be understood in a more abstract way than Scalia proposes. This difficulty can be handled in the following way.

Begin with an axiom: The Establishment Clause forbids the state from declaring religious truth. One principal reason why it is so forbidden is because it is incompetent to determine the nature of this truth. Madison wrote that the idea “that the Civil Magistrate is a competent Judge of Religious Truth” is “an arrogant pretension falsified by the contradictory opinions of Rulers in all ages.” The idea of state incompetence is reinforced by evidence that that state sponsorship tends to diminish respect for religion.

This incompetence entails that the state may not favor one religion over another. It also bars the state from taking a position on contested theological propositions, such as whether God exists. (There is an exception for ceremonial deism, confined to public rituals of long standing whose religious content is sufficiently bland, but in its nature this exception cannot permit any new instances.)

A second classic reason for barring the state from declaring religious truth is civil peace: In a pluralistic society, we cannot possibly agree on which religious propositions the state should endorse. The argument for government agnosticism is that, unlike government endorsement of any particular religious proposition, it is not in principle impossible for everyone to agree to it. Nonestablishment makes possible what John Rawls calls “civic friendship,” in which citizens share a conception of justice that all can reasonably be expected to accept, despite their widely differing moral, philosophical, and religious beliefs.

A final reason for getting the state out of the religion business is that the individual’s search for religious truth is hindered by state interference.

It is, however, possible for the state, without declaring religious truth, to favor religion at a very abstract level. The Court noticed this in Texas Monthly v. Bullock when it invalidated a law that granted a tax exemption to theistic publications, but not atheistic or agnostic publications. Justice Brennan’s plurality opinion thought that a targeted exemption would be appropriate for publications that “sought to promote reflection and discussion about questions of ultimate value and the contours of a good or meaningful life.” Justice Blackmun thought it permissible for the state to favor human activity that is specially concerned with “such matters of conscience as life and death, good and evil, being and nonbeing, right and wrong.” What is impermissible is for the state to decide that one set of answers to these questions is the correct set.

The precise character of the good being promoted is itself deliberately left vague, because the broad consensus on freedom of religion would surely collapse if we had to state with specificity the value promoted by religion. “Religion” denotes a cluster of goods, including salvation (if you think you need to be saved), harmony with the transcendent origin of universal order (if it exists), responding to the fundamentally imperfect character of human life (if it is imperfect), courage in the face of the heartbreaking aspects of human existence (if that kind of encouragement helps), a transcendent underpinning for the resolution to act morally (if that kind of underpinning helps), contact with that which is awesome and indescribable (if awe is something you feel), and many others. The establishment clause permits the state to favor religion so long as “religion” is understood very broadly, forbidding any discrimination or preference among religions or religious propositions.

This understanding makes it possible to defend accommodations without running into the free exercise/establishment dilemma. The state is recognizing the value of religion, but it is making no claims about religious truth. It is the making of such claims that violates the establishment clause. But it is possible to use the term “religion” intelligibly without making such claims. We don’t have a precise definition because we don’t want one. The vagueness is deliberate, and it is wise.