Back in 1988, at a time when my old friend Doug Collins used to invite me regularly to speak at the University of Washington in Seattle, I gave a talk entitled “The Anthropological Idea of God.” Thirty-two years later, I feel I should revisit this idea, but unlike in preparing the second edition of The Origin of Language (TOOL), I don’t think it useful to return to the old text.

My basic ideas about the origin of language have not fundamentally changed. The sign is the originary solution to the problem posed by the potential violence of mimetically enhanced appetites, and culture can still be summed up as “the deferral of violence through representation.”

In contrast, as is perhaps inevitable as one reaches one’s later years, my thinking about religion has evolved, not in the sense of a conversion to belief (Pascal’s j’ai pleuré et j’ai cru), but of a rapprochement between the perspectives of GA and that of (particularly Judeo-Christian) religion.

Even if, as I wrote at the start of TOOL, mysteries should not be multiplied beyond necessity, the mystery of faith cannot be eliminated. For faith is paradoxical; it is mystery itself, since could it be either demonstrated or refuted, it would no longer be faith. But GA can understand in its own terms faith in supernatural events such as God’s creation, virgin birth, or the Resurrection, even if believers cannot accept this understanding. There is no point in attempting to reduce the supernatural to worldly terms. The paradox is always there, whether we let faith solve it, or leave it as an unprovable point of departure.


Whether we are to take the supernatural literally or figuratively is secondary to our understanding of it as the mark of transcendence. This is the key to the paradoxical nature of the sacred/significant. Those who disrespect religion because of its myth-like elements and incompatibilities with empirical reality (such as fixing the origin of the universe in 4004 BC) treat these disparities as simply the difference between fact and fiction. Data tell us that the earth was formed billions of years ago, and that our universe itself, whether or not there are others, is so many more billions of years old—a long time by our standards, although hardly “eternity.” But to respond to the atheist by alleging his ignorance of “why there is something rather than nothing” does not demonstrate the existence of anything like the divinities we worship. Such polemics are not only unwinnable but unproductive.

GA offers an entirely different perspective, untainted as far as is possible by the metaphysical postulate on which philosophy and subsequently natural science is founded: that (mature) language is a neutral, timeless medium of communication “downloaded” by humans when they discerned a need for it, and that the details of the downloading process, formerly the object of religious just-so stories, can be fully understood by seeking the prehuman origins of its cognitive, physiological, and social components.

Once language itself, as opposed to its worldly appearance, need not be explained, its use to describe the universe requires no reference to the divine or transcendental, although the latter reemerges in Western philosophy under the influence of Judeo-Christian transcendentalism. The Greeks had gods as well, but as Xenophanes’ writings make clear, their transcendentality lacked the radical necessity of the One God. The philosopher’s claim that horses would have gods who looked like horses would be meaningless within the Hebrew tradition, where it is illegitimate to either question the reality of the One God or, not coincidentally, to attempt to figure it.

I am certain that the reason GA has survived for 40 years is the fact that, despite the increasing hostility to “speculative” thinking of the institutions that academics depend upon, GA’s supporters share the intuition that its message is of epochal importance. As with the Jews, GA’s chutzpah turns people against it, and until we make our own Israel, it will always lack the institutional backing needed to forcer l’admiration. But as the last decades of language origin literature demonstrate, none of this carefully researched material shows any sign of grasping the categorical difference between the evolutionary emergence of language and that of the opposable thumb. At best we are told that “there is something profoundly distinct about human language” (see Chronicle 611). As for religion’s coevality with language, the less said the better.

Over the years since TOOL, I have come to realize that to appreciate GA’s uniqueness, it is essential to focus on the idea that at the origin, the sacred and the significant are one. This is the core of the originary hypothesis. It is not congenial to our common sense, because the two terms have diverged through history—indeed, we can say that “history” is nothing but the story of their divergence. But we can only hope to understand them if we seek to reestablish the originary mindset that would grasp them as synonymous: that as opposed to using a signal, using a sign for something is in itself taking it as sacred, even if, once the “taking as” becomes a praxis, it can and will detach itself from the attitude that inspired it—whence, in the terms of TOOL, the emergent distinction between formal and institutional representation.

The idea that signs “substitute” for things, as though signs were objects like those Swift’s Laputans carried in their backpacks, is impossible to avoid so long as one refuses to accept the transcendental nature of language, which transcends any linguistic system that purports to “include” it. GA differs from other ways of thinking in recognizing the paradoxical nature of language and thought, and the consequent impossibility of containing them within a metaphysical system that refuses to understand language as an anthropological reality rather than an eternally existing possibility—as is, at least in a first approximation, the case for mathematics.

This attitude once established, language origin becomes, in the context of Darwinian evolution, like the origin of any other form of behavior, and can be investigated as such. Such investigations are not illegitimate in themselves; but they are like the blind men’s descriptions of an elephant. They can attain their objective only by being situated in a context of genuine ontological change—as the emergence of a new kind of being, to put it in necessarily inadequate philosophical terms, which could not fail to be accompanied by a new type of mental function. What I have called the scene of representation must have a physical counterpart in the individual brains of the first exchangers of the sign, but cannot be reduced to its material reality, because it necessarily bears the trace of the communal scene in which language arose. The reader may be shocked to realize that claiming that “language is a gift of God,” however unsatisfying this may be from a scientific viewpoint, is of a higher level of truth than the metaphysical agnosticism that denies that language’s origin is pertinent to the discussion of the ideas it expresses.

Unlike the relevant components of animal societies, language, and culture as a whole, are not reducible to the contents of individual neurosystems. They involve unpredictable interactions that cannot be reduced to the chemical formulas that have been so successful in regulating the behavior of social insects. Language is a means of communication in a sense qualitatively different from that of the signal-exchanges characteristic of insects—or of vervet monkeys or chimpanzees.


In the first version of “The Anthropological Idea of God,” I defined this “idea” as the subsistent center of the scene of representation. This is true enough from an etiological standpoint, but clearly the substance of the idea “God” in any religious system is not equivalent to its reduction to a memory trace of the originary scene. The believer who encountered this “definition” would rightly reject its reductionism as incompatible with his and with every other believer’s experience of God. And since God has no objective worldly extension, but is only (universally understandable as) an Idea, the believers would have every right to agree in this rejection.

There is no way to get around this objection other than to recognize it, which is possible only from GA’s extra-metaphysical perspective. Although we use declarative sentences as do the philosophers, we recognize, as they do not, language’s anthropological rather than “logical” blind spot.

Paradox is not a mere curiosity, nor a defect that could be cured by additional rigor. Paradox exists because language can operate only by pretending it does not. Hence although we may assume that the aura of the sacred, after inhabiting the original central object, is afterward transferred to the locus of its immolation and consumption, this is neither a logical consequence nor an inference from our experience of holy places. It need not even be true. But what must be true is that the sense of the sacred, which Kierkegaard called “fear and trembling,” once born in this originary event, is never lost, not simply as the memory of an event, but as a world-transcending (although derivatively, individually experienced) Being. This Being cannot be designated by the metaphysical term concept, but neither should we consider it bound permanently to anything so specific as a “subsistent center.” The sacred, in the first men’s experience and in our own, is that which imposes the need for différance, the néant that separates us from our appetite for it without negating this appetite, as would a Pavlovian inhibition. The superiority of the term différance/deferral to designate our sense of the sacred reflects its relative clarity as a philosophical term that has nonetheless a sufficient element of deconstruction, that is, of what was in its origin an anti-metaphysical use of quasi-anthropological language within metaphysical discourse, to confer on it sufficient specificity.


The absence of any “name” for the sacred-in-general is most simply illustrated by the custom of Orthodox Jews to simply refer to God (written “G-d”) as “the name” (Hashem). Whatever role YHVH played in ancient Hebrew history, the word is never pronounced, and referred to only by a substitute, in prayers commonly Adonai (Lord); but Hashem is the usual way to speak of God in ordinary language. This is a paradox in a single word—what has no name can be designated only as “the name”—a tighter “definition” of the One God even than ehyeh asher ehyeh. Coupled with the interdiction on “graven images,” it embodies a maximal effort to separate the sacred from the simply nameable, the signifié of Saussure, not because God is a “spiritual” rather than a “material” being, but because the ontology of the sacred is transcendental and cannot be encompassed on either side of Saussure’s “sheet of paper” with signifiant on one side and signifié on the other.

“Belief” has the outward appearance of a metaphysical entity surrounding a transcendental core. What one “believes” is a discourse, a consecrated account of a revelation of divine action in the world, not a product of reasoning or deduction, although much may be deduced from it, which is what we call theology. The possibility of such deductions, like the possibility of deducing anthropological truths from artworks, is inherent in language because the metaphysical language of truth is there made to reveal its pre-declarative origin. For example, the interpretation of what appear to me the central Hebrew and Christian revelations found in the text of the two Testaments make up the core of Science and Faith.

Christianity goes beyond Judaism to make the point that maintaining a notion of the sacred based on the historical revelations of the Old Testament and of God’s laws as spelled out in the Torah still does not make clear where, beyond a set of propositions in which to “believe,” it is that religion enters my life. The personal relationship of the Christian with Christ, as realized in the Eucharist and personified in the Trinity by the Holy Spirit, must be experienced as a continual presence, so that the individual human being may never for an instant become unaware of his intimacy, as if with a fellow human, with humanity’s transcendental source. The “absurdity” in the credo quia absurdum (more accurately, credible quia ineptum est) for which Tertullian remains uncanonized is the Incarnation of this source in a single historically existing individual, which is another way of saying that we can only “understand” the sacred as just this absurdity. In contrast to seeking to understand the sacred in “logical” terms, congenial to metaphysics, Christianity offers an experiential mode of understanding. Whereas their “elder brothers,” the Jews, refuse this “absurd” understanding and accept Hashem’s mystery. Which does not prevent most religious Jews from frequently “talking” to Hashem beyond reciting prayers, given that Hashem’s “conversations”—with Adam, with Moses—are the heart of the Torah.

Other religions demonstrate other ways of dealing with the transcendental. My limited knowledge of Buddhism allowed me to interpret (in Chronicles 515516) Nagarjuna’s paradoxes, in contrast with Zeno’s, as denying rather than merely problematizing Parmenides’ way of truth, and hence the coherence of the metaphysical universe of ideas. Although the construction of Nagarjuna’s paradoxes appears arbitrary, I think they can simply be understood from the standpoint that a given sentence stands both inside and outside the world it describes—transcendence cannot be reduced to logic.

Hence in Nagarjuna’s Mahayana Buddhism, the scene of representation is one from which all language has been expelled as unreliable, a néant from which not merely all desire but all thought processes are “deferred.” This state of deferral, essential to the various forms of Buddhist meditation and its derivatives, is an experience of abstract sacrality that substitutes in Buddhism for “God.” This explains why Buddhism is so blissfully indifferent to the worship of the gods of other religions, or to the articulation of the sacred/human status of the Buddha himself, who unlike Christ does not link the divine to the human by including different “persons,” but bathes in a sacrality that universal deferral allows to exist within ourselves and can be embodied in an indefinite number of Bodhisattvas. In this sense, Buddhism is “atheistic,” retaining from the originary scene the sacred as disincarnate experience, the prolongation of the originary deferral of appetitive action, experience for which the Buddha remains a human example rather than, or more than, a transcendent source.

Even this limited example shows that the minimal anthropological basis provided by the originary hypothesis allows a given religion to be studied on its own terms, just as it provides a basis for understanding our originary relationship to language. Extending this procedure to the world’s other religions would permit the creation of a universal religious typology.


Much can be learned about cultural phenomena without regard to the insight provided by originary anthropology, but their deepest connection with the human community is thereby lost.

Structural linguistics, which studies language as a set of mature systems, would surely benefit from examining, even in speculative terms, language’s prehistory. The remarkable uniformity of the world’s languages that led Chomsky, albeit not without question-begging, to figure human language as the product of a universal “module” with a set of “switches” whose settings exhaust the possibilities of linguistic diversity, would be seen from an entirely different perspective if the effort were made to derive these structures from an originary hypothesis. The current revolt against the Chomskian model to study “languaging” (see Chronicle 634) as a mode of interaction rather than of syntactical construction would particularly benefit from tracing this activity to a concrete, evenemential origin.

But if linguistics could surely benefit from generative anthropology, the study of religion would be wholly transformed by it. Religious culture cannot be understood without reference to transcendence. The dismissal of the transcendent, scandalously common in the scientific community, whether as a myth concocted by the ruling class as “opium of the people” or the result of man’s “wonder before nature,” reduces religion to a set of superstitious practices that at best convey socially useful values at the cost of falsifying our understanding of both natural and human phenomena.

Whether the natural world suffices to generate the presence of transcendence in the human universe, or whether some extra-worldly force is necessary, is a question we have every right to bracket, so long as we recognize that in the absence of the deferral through sacred “fear and trembling” of our appetitive existence, and of the signs by which we not merely recall but regenerate this state, none of the phenomena uniquely characteristic of the human community could have appeared. Once we grasp this fundamental truth, we are free to explore both the various religions and the worldly “life-philosophies” that would replace them without concern that in either case we are putting the reality of transcendence in jeopardy.