The idea that religion is a functional component of human society is well known to evolutionary psychologists such as Pascal Boyer and religious sociologists such as Rodney Stark. Although none of these social scientists would dream of accepting GA’s (or René Girard’s) originary hypothesis, they recognize that the “irrational” aspects of religious discourse, taking “rationality” in a logical-positivist sense, are highly functional in bringing people together and cementing their loyalties through collective and individual ritual acts and interdictions, keeping away free riders by rigid and onerous requirements of professed belief and behavior, and generally by requiring serious commitment as the price for the benefits of community solidarity. Stark points out, for example, that Christian communities tended to fare better than others during plagues because their members cared for each other even at the risk of infection rather than leaving sick but potentially curable members to die—a human version of what population biologists call “group selection.”

Given that serious social scientists offer grounds for affirming religion’s usefulness, in contrast with the spate of popular books a few years ago denouncing religion in what can only be called Voltairean terms, one might think that nothing more needed be said in its defense. But from the standpoint of GA, which emphasizes the derivation of the rational language of the declarative sentence from simpler utterance forms, and ultimately from the more fundamental human need for deferring violence, there are further points to be made that might hopefully serve to bring believers and non-believers a step closer to each other.

The rational claim that, for the reasons stated above, religious discourse is useful because of its very irrationality, cannot avoid an element of demystification. Explaining to a believer in the Virgin Birth or in the Mormon doctrine that American Indians are the Lost Tribes of Israel that his belief has sociological value in maintaining the solidarity of his community, or that it has psychological value in tapping unconscious mental-emotional resources, nevertheless treats this belief in the rational context of scientific discourse as a delusion.

While respecting the distinction between rational and religious discourse, GA puts the question in a new light by denying the metaphysical principle that the enunciation of true propositions is the primordial function of language. Thus rather than beginning from the standpoint that language is “naturally” rational and that the supernatural affirmations of religion require, so to speak, “strict scrutiny” to make sure that social benefits outweigh deviation from empirical truth, we begin from the hypothesis that the originary purpose of language is to defer violence by making/discovering as sacred the objects of common desire, or in secular terms, situating them on the scene of representation where their appropriation is deferred.

But we would claim not only that this theoretical position about language origin is implicit in religious discourse itself, but that certain variants of religious discourse are not merely compatible with but actually conducive to the development of rational discourse alongside it. The development of logical and scientific reasoning, which as we know was above all successful at first among the Greek philosophers and later in the Christian West (not to deny the achievements of Islamic astronomers or Indian mathematicians), has until recently taken place superficially in conflict but more broadly in symbiosis with religious discourse. In particular, the insistence on the miraculous that is far more central to Christianity than to the other “Abrahamic” religions (and as far as I can tell, to Buddhism and Hinduism and their variants), where founding narratives are understood more as parables or even myths than as supernatural events, is paradoxically associated with the one society that gave birth to modern science. Just as Max Weber theorized that the quest for personal salvation characteristic of the Protestant Reformation was homologous to the “spirit of capitalism,” so it is likely that the paradoxical fideism of Western Christianity, both Protestant and Catholic, made it more congenial to the emergence of the hypothetical constructions of empirical science in the early modern era than, for example, Islam as practiced in the Ottoman Empire at the moment when its economic and military power made it competitive with the great nations of Europe.

I would venture to say that the Christian insistence on credo quia absurdum, on Virgin Birth and Resurrection, on scandal to the Jews, folly to the pagans, all of which derive from the “category error” of Anselm’s Cur deus homo, God’s taking on a human persona—in a word, Christianity’s insistence on faith beyond reason—can be understood, setting aside all considerations of substantive truth in allegorical, anagogical, etc. form, as in the first place a reminder of the originary purpose of language: the avoidance of conflict through deferral of appropriation and the institution of the sign as a mutual promise of solidarity. The originary act of faith in this promise, the first moral act, precedes hope of appetitive reward, and can be said to serve as a model for the hypothetical nature of scientific discourse as well. As Jacques Derrida eventually realized, language itself is based on faith. (See “Frère Jacques,” Chronicle340.) Whence Girard’s achievement in making Christian faith the source of a fundamental, pre-metaphysical anthropology. (Whether or not the admixture of a Jewish element was necessary in order to adapt Girard’s theory to the existence of language as a logos not simply assimilable to Jesus is not my subject here.)

My point is not to “bridge the gap” between rational discourse based on empirical and confirmable observation and faith in Christian or other doctrine; these are indeed two different usages of language, even if the persistence of paradox, or the Gödelian incompletion of infinite logical systems, suggest the limits of rationality (and who knows, to read from the layperson’s perspective the bizarre doctrines of modern physics, to what extent they are simply difficult to grasp without the requisite mathematical training and to what extent they conceal their own forms of paradox). No, the Virgin Birth and e=mc2 cannot be made part of the same discourse. But that the former is more fundamental to language than the latter is an anthropological fact that the post-Enlightenment world tends too easy to forget. Faith in the “absurd” is, not least, a lesson in the originary nature of language, that, put in everyday human terms, might be expressed as: “language must discover/invent a sacred to ensure peaceful communication before it can express propositional truths,” or more simply, “worldly facts are less fundamental than human love.”


Religious faith teaches us about the use of language; but it should also teach us about language itself. The philosopher of science Bruno Latour’s Jubiler (Seuil, 2002), in which he insists eloquently that religious discourse cannot be judged according to rational-analytic criteria—using as a notable analogy to religious language the clearly ostensive example of a declaration of love—nevertheless fails to define a formal pre-declarative basis within language itself for the distinctions he so respectfully draws. This lacuna is one more residue of the faith in propositional language that lies at the foundation of Parmenidean-Platonic metaphysics. If one takes for granted that language is essentially, as in Wittgenstein’s Tractatus, a collection of declarative sentences, along with an occasional “defective” imperative or “interjection” (the closest standard grammar comes to recognizing the existence of the ostensive), then religious discourse will always be understood in one way or another as “myth.” But once one realizes that language could only have emerged from animal communication as a new way of dealing with a problem of collective order that had become insoluble, and that the declarative sentence is a complex structure that could not have been the originary means of solving this problem, one is prepared for an, if not the originary hypothesis, and in consequence, to grasp the originary importance of the ostensive.

That we cannot empirically verify the most crucial statements of religious discourse is hardly surprising. But it is worthwhile reflecting that not only is “belief in Jesus’ birth from a virgin” a lesson in originary thinking that forces the believer to accept an idea that can have no empirical foundation, that must be accepted first and then its “referent” sought in the various aspects of one’s experience, but such belief is also a lesson in originary linguistics. What religious discourse teaches us in the domain of language is that declaratives are not originary and that their roots lie in ostensives. For whether one have faith or not, one must realize that the affirmation of the Virgin Birth is not a “proposition” in the rational sense but the translation of the ostensive presence of the sacred into propositional terms, an expression of the paradox that is inherent in any postulation of significance. We can calmly affirm, “the cat is on the mat” because other uses of representation have bought us a time of deferral that grants us the tranquility to dispassionately inspect our surroundings. But in the originary utterance that paid the price for this peace, the cat would be the divinity re-presented by the sign, and the mat, the sacred scene of its representation and sacrifice.


A note on blasphemy

Since religious language is in rational terms “absurd,” a secular culture, one founded on the primacy of rational discourse, cannot reasonably forbid modes of discourse that point up this absurdity, however jejune or vulgar these affirmations may appear to neutral parties and however offensive to believers. Whatever else modern society allows or forbids, it cannot forbid blasphemy, whether against traditional religions, as is today enforced virtually exclusively in Muslim countries, or against political cults such as that practiced in North Korea or Cuba.

Many European countries, notably Germany, have laws forbidding Holocaust denial or antisemitic statements. Speaking personally, to these laws, whatever their historical justification, I prefer the American First Amendment. But whatever one’s objections to them, they are not, as polemicists sometimes claim, selective blasphemy laws whose existence demonstrates that some people are more equal than others. Holocaust denial itself deserves to be called an “article of faith” in an antisemitic cult, but to call “belief” in the reality of the Holocaust a religious doctrine is to deny the rational domain altogether. (Jean-François Lyotard points out in Le différend what might be called the anti-ostensivity of Robert Faurisson’s claim that the gas chambers never functioned: only the victims could give evidence of their functioning and they, of course…)

As for antisemitism, the interesting thing about it is that it is a hatred not of the Jewish religion but of Jews. And if one ask, what is it about Jews that arouses this secular hostility, it is not their affinity for supernatural beliefs of the sort mocked by Dawkins, Harris, Dennett, and Hitchens, but what I have called their firstness, their primary relationship to sacred text, and as the inventors of “monotheism,” which is also, and perhaps primarily, monotextualism, to the very idea of the Sacred Text itself. The “scandalous” genius of Christianity, in deliberately insisting on its absurd faith, is among other things a “supplementary” compensation for the simpler, more originary faith of the Jews in the One God and the (consequently) universal moral truths enunciated in the language associated with him, at first in oral form but more quickly than elsewhere in antiquity in a broadly accessible written tradition.

Blasphemy, insulting God’s name, is listed as a capital crime in Leviticus 24, but God’s naming himself as ehyeh asher ehyeh in Exodus 3 is not simply the silencing of his “proper” name but its denial, making such insult impossible and blasphemy against the Hebrew-discovered One God, effectively nugatory. On the one hand, the One God has no name and cannot therefore becalled names. On the other, to deny the One God is effectively to deny God altogether, and the resulting discourse would be not blasphemous, but simply atheistic. The consequent invulnerability of the first to discover this is one more way of explaining the intensity of the age-old historical resentment of antisemitism.

Christianity has historically been much more sensitive to the idea of blasphemy, and one might expect that the burden of Christian faith that we have been discussing would make Christian authorities more sensitive than Jewish ones to details of dogma. And as a means of expressing hostility to Christianity, the desecration of Christian images has flourished in the West from the French Revolution to Piss Christ. But when all is said and done, the Crucifixion is already the ultimate desecration; what Saul/Paul found out on the road to Damascus was that insulting Christ is just a way of worshipping him. Christian faith, in sum, need have no fear of blasphemy.

Islam, in contrast, has very little dogma. Its hypersensitivity to blasphemy is one with its denial of historical filiation with the older religions from which its traditions derive, a denial whose anti-historicism has been vastly intensified in reaction to the modern era that the Christians and Jews have created. Muslim theocrats who assert the Koran’s timelessness cannot mediate, as can Jews and Christians, between their sacred text and the truly atemporal discourse of scientific reason. This suggests that it is only with the demise of Muslim blasphemy laws and a renewed effort at such mediation that the Islamic world will become integrated into the global world of modernity.