English Department
University of Victoria
P.O. Box 3070 STN CSC
Victoria, BC V8W 3W1

A hundred years ago, Émile Durkheim (1995 [1912]) argued that the origin of humanity begins with the distinction between sacred and profane. All human societies make use of this distinction, though the cultural forms used to enforce it vary widely across different human communities.

Why must the distinction be enforced? Why must human societies worry about what is sacred and what is profane? It is telling that most people—at least most modern, secular people—would say that the distinction is in fact quite unnecessary. Why indeed must the tribesman gather around his totem pole and beat his drum? Or lest the question appear too ethnocentric, why must the Catholic believe that the bread is the body of Christ? Referentially speaking, such beliefs are absurd. The totem is a piece of wood. The bread is not the body of Christ, it is bread.

We usually associate this kind of reaction with hardheaded Voltaire types—people like Richard Dawkins or Christopher Hitchens, for example. Let’s call them the scientific fundamentalists. What is their point? Their point is the same as that made by the Enlightenment philosophes, who held that humanity wouldn’t be free until the last king was strangled in the entrails of the last priest. (I’ve seen this line attributed variously to Diderot, Voltaire, and Jean Meslier—no matter, it nicely captures the attitude of scientific fundamentalism.) Basically, the idea is that once liberated from the superstitions legitimating the tyranny of kings and priests, people would be free to be rational. Rational here means: free to follow the authority of individual experience and test that experience against nature. If the authority of your concepts doesn’t come from your identity, from your status in a hierarchy or Great Chain of Being, your proximity to the sacred totem pole, then it must come from somewhere else. Where? Why, from nature of course. This is the basic idea behind both rationalism and empiricism. Reason must learn to stand on its own two feet without any outside influence from “transcendental” sources.

But let’s return to our original question. Why must the distinction between sacred and profane be enforced? The Enlightenment philosophes believed that it didn’t have to be. They hoped to take the transcendental off its pedestal in order to release humanity from servitude to its obnoxious and irrational demands. But the curious thing is that humanity seems predisposed to servitude. The philosophes prepared the way for the revolution; but the revolution led not to liberty, fraternity, and equality, but to a servitude that seemed at times more violent and terrible than the coercion of divinely sanctioned kings. Nor has subsequent history shown the French case to have been an aberration. Quite the contrary.

The Czech-born and British-educated philosopher and anthropologist Ernest Gellner has an interesting theory of why human beings are predisposed to submission before their gods—or, to put it less dramatically, of why humans are predisposed to constraint by the sacred. He points out that, genetically speaking, humanity is enormously plastic. Human infants will acquire the characteristics of their specific cultures quite effortlessly. Once they are fully incorporated into their culture, however, the capacity to acquire new or different linguistic and cultural habits becomes remarkably more difficult, sometimes impossible. In other words, human beings seem to be defined by a paradox. They are born with a remarkable genetic plasticity, but the particular cultures they acquire are remarkably constraining. As Gellner (1995: 48) puts it, “Man is born genetically free but is everywhere in cultural chains.” Why?

Gellner’s answer to this question is very Durkheimian. Without constraint, human society would be impossible. All types of social organization, whether human or animal, require a degree of homogenization, standardization, and discipline in order to function. From insects to wolves, animal societies have come up with various genetic solutions to the problem of social organization. This is pretty obvious among bees and ants. Individual behaviors or “functions” are genetically given. It is inconceivable for the worker bee to throw off her genetic chains and usurp the resident queen. Bees appear to have solved, rather admirably, the problem of maintaining peaceful coexistence within a single colony. On the other hand, the displacement of the alpha is a fairly routine occurrence among higher social animals, such as wolves and chimpanzees. Like humans, these animals demonstrate a far greater degree of genetic plasticity. But even here there are limits. Chimpanzee societies are a good deal more flexible than insect societies, but they do not come close to demonstrating the cultural diversity of human societies. I’m therefore skeptical of the claim, made by some primatologists, that chimpanzees have culture in the same sense that humans do. But this point does not affect the main argument made here.

Gellner proposes that human culture is a response to human plasticity. Given the fact of genetic plasticity, human society is possible only if a new cultural system of constraint is superimposed on the preexisting biological system that has, so to speak, thrown off its genetic chains. Cultural chains substitute for genetic ones. Here I must emphasize something that remains only implicit in Gellner’s argument, namely, the idea that human social interaction is extremely volatile. The superimposition of human culture onto human nature is a consequence of the volatility of human nature. Gellner takes this volatility to be the plasticity of human cognition. But this is really to beg the question. The human brain is “plastic” because it has co-evolved over millions of years with culture. But plasticity in itself does not explain the origin of culture. The answer to that question must rather be found in Gellner’s idea of the volatility implicit in all forms of social organization, whether human or animal.

Gellner rightly stresses that a certain degree of homogeneity and standardization is necessary for any society to exist. As we have seen, insect societies solve this problem by genetically hardwiring their members from the start. The division of labor is part of their genetic makeup. But this genetic determination is precisely what cannot be relied upon in human societies. Now the cause of this breakdown of the genetic system cannot be sought in the genetic system itself. It must come from outside it. The plasticity-culture tandem must arise from a specific problem within human, or protohuman, social organization. The most plausible theory I know for the source of this volatility comes from René Girard (1977, 1987) and Eric Gans (1985, 1990, 1993, 1997). The plasticity or “volatility” of the human brain is in the first place a response to the volatility of human social organization. More specifically, it is a response to the danger of mimetic conflict in the human, or protohuman, group. The breakdown of the genetic system of stabilization occurs when this system is no longer capable of controlling the mimetic lability of protohuman interaction. In contrast to humans, conflict among chimpanzees is controlled by ontogenetic ritualization. Behaviors such as snarling, hooting, and violently shaking branches are indices of real conflict. They are connected contiguously, in space and time, to a larger pattern of behavior that may in fact take place, namely, a real fight between rivals. These indexical signs are not learned from scratch, as in Skinner’s conditioning experiments. As Terrence Deacon (1997: 330-32) has shown, they are examples of genetically assimilated behaviors. In this sense, they are like our own repertoire of “gesture-calls” (Burling 2005: 24), such as laughing and crying. These gesture-calls are narrowly constrained by our biology, and it is therefore highly implausible to regard them as prototypes or precursors of human culture and language.

Gellner has no concept of mimetic conflict in his theory, but his anthropological intuitions are nonetheless very sharp. He sees clearly that culture originates as the non-genetic means of constraining an otherwise highly volatile protohuman community. And he also sees that this system of constraint is modeled minimally on the arbitrary signs of language. “Language,” Gellner (1995: 50-1) says, “is, initially and basically, a system of prohibitions. Am Anfang war das Verbot. In the beginning was the prohibition.” Language originates not as an attempt to refer to the empirical world—a mistake made by virtually all empiricist philosophers. The existing indexical sign system already does that very effectively. Instead language originates as a radically new way of constraining potentially volatile behavior. The subsequent use of language for exclusively referential purposes—in a word, for modern science—is a secondary and historically very late development. But given the secondary nature of this development, it is no longer possible to interpret scientific rationality as the basic condition of all humanity. This is the lesson the Enlightenment philosophes failed to grasp. Though we can certainly appreciate their optimistic belief that they had transcended the irrationality of religion, it is rather more difficult to excuse the same error made by their avatars and imitators two centuries later. This is one of Gellner’s most fundamental points. The originary function of language is not reference but constraint. The collective focus of attention on a central object is a form of prohibition. It constrains individual behavior in what would otherwise be an extremely labile and unstable situation. Communities can survive only because of this constraint, which is why in most societies ritual prohibition takes precedence over empirical truth.

The question then becomes: What enabled the shift to empirical truth upon which modern industrial societies depend for their economic and technological success? For Gellner, this very recent transition must be understood in the context of previous modes of cognition. First, representation had to be freed from the multi-stranded functions of traditional hunter-gatherer societies. The first step in this direction occurred with the neolithic revolution and the origin of agrarian societies. Here, specialization of functions and the division of labor became vitally important for social organization. In particular, the creation of a food surplus produced, for the first time, political centralization and large-scale coercion because the central storehouse had to be protected. Proximity to the center creates hierarchy. The other important function was the priesthood, the class of legitimizers. The priests were responsible for the codification of doctrine that endorsed the hierarchy. The final crucial ingredient was the origin of writing. When writing is used by the clerisy to produce a doctrine, you have what Gellner calls “generic Platonism.” The important thing about generic Platonism is the codifying of norms and behavior that had previously existed and been enforced at the ritual level. When ritual is codified in writing, ethical concepts are lifted from their communal face-to-face existence to take on a universal, trans-communal application. “The central intuition of generic Platonism,” Gellner (1989: 76) writes, “is the independent existence of concepts, ‘Ideas.’ These entities simultaneously constitute logical and moral models for reality. The transcendent receives formal recognition. Reality does not constitute a check on Ideas: on the contrary, they are the norms by which reality is to be judged and guided.”

Generic Platonism is still an ethical doctrine, but unlike the multi-stranded ritual practices of traditional hunter-gatherer societies it provides a coherent and ordered world picture. Ethical concepts are theorized as part of a unified ontology. The possibility of modern science occurs when ethics is finally divorced from this unified ontology, when, as Hume famously put it, is becomes distinct from ought. Transcendence retreats from the world and nature becomes the sole basis for an experimental and revisable ontology. Gellner is extremely self-conscious about the difficulty of explaining this transition from agrarian to industrial society, a shift which is without precedent in all of human history up to the point in question. What on earth motivated people not simply to overthrow their kings but to resist the temptation of becoming kings themselves? “How can it happen,” Gellner (1989: 158) asks, “not merely that the weak, the swordless, overcome the swordsmen, but that the whole organization and ethos of society changes, that Production replaces Predation as the central theme and value of life?” Given the existing logic of the agrarian order, it is extremely hard to see how it could produce this shift. The transition requires a wholesale reorganization of society, a reorganization that occurred only once in human history. Modern industrial societies have escaped from the cycle of often brutal coercion that characterizes agrarian societies. As Gellner (1995: 59) puts it, whereas agrarian societies are “inescapably Malthusian,” modern industrial societies are affluent and can afford “the luxury of a marked relaxation of coercion.”

Figure 1 provides a schematic representation of Gellner’s (1989: 77) theory of the cognitive evolution of humankind.


Our society—i.e., modern industrial society—is represented by C, in the upper right quadrant. We inhabit a world that is single-stranded and referential. In contrast, hunter-gatherer societies, represented by A, in the lower left quadrant, are multi-stranded and largely non-referential. What Gellner means by “multi-stranded” is that there is no attempt to unify the concepts that guide behavior into a unified logical picture of the external world. Reference is possible but opportunistic rather than logical. The “fingers” or “periscopes” that extend into the world are not logically connected to one another. Instead they are connected to the social, non-referential world that guides behavior. Referring to E.E. Evans-Pritchard, Gellner (1989: 40-42) cites the Nuer people’s “identification of bulls with cucumbers.” Anthropologists have either taken the view that this “indicates a special, pre-rational kind of mentality, radically discontinuous with our own,” or entertained the “benevolent or charitable thesis” that the “apparent empirical content is really a re-affirmation of the social order.” For Gellner, both views are flawed. The first, because it can’t explain the otherwise excellent abilities of the Nuer and other tribal peoples to negotiate their physical environment perfectly rationally; the second, because it attributes to the tribesmen the same “logical fastidiousness” as the modern philosopher-anthropologist. The first sees only a massive gulf between primitive and modern; the second, no gulf at all. For Gellner, there is a gulf, but the gulf is not explained by referring to modern standards of empirical rationality. Rather the difference must be explained sociologically, by referring to differences in social organization, in particular, to the difference between societies based on single-strand activities (C) and those based on multi-strand activities (A).

The purpose of a philosophical anthropology, at least as conceived by Gellner, is to explain the historical passage from A to C. Thus, you can’t get from A to C without passing through B. This is the period of unification and codification represented by “generic Platonism.” Here the referential content is still small (Gellner 1989: 77), but unlike in A there is an attempt to bring reference under the umbrella of an overarching and logically coherent social ontology. The shift to C represents the inversion of this logical system. The unification remains but receives its sanction not from the ethical order but from the external world. This is the world of modern science with its experimental and provisional ontologies based on reference to nature. The implication of this inversion for the social order is a notable weakening of the concepts governing social ontology (i.e., religion).

What are the implications of this model of human history for our understanding of “transcendence in a secular age”? In his discussion of the political and moral order of modern civil society, Gellner (1994: 94) claims that “a free order is based in the end not on true and firm conviction, but on doubt, compromise and doublethink.” What does this rather enigmatic statement mean? Is Gellner saying that our politicians must be janus-faced? In order to understand Gellner’s statement, we need to compare it to the options that preceded it, in particular, to the options of the two previous stages of human history, the hunter-gatherer stage and the agrarian stage. Both hunter-gatherer and agrarian societies depend upon constraint for their existence. It is the job of culture—and, in particular, of ritual—to enforce this constraint. Industrial society is exceptional in the sense that, for the first time, ethical and conceptual constraint is significantly relaxed. The cognitive ethic of agrarian society is rigidly hierarchical. Social stability is prized above all else. In contrast, industrial society is highly mobile. Individual identities are not rigidly tied to kinship or social status. Instead, they are open and changing. This mobility reflects the opening of the cognitive and economic spheres to experimentation and choice. What is sacralized is method rather than concepts. Indeed, the sacralization of concepts now takes a backseat to method. Gellner (1995: 59) associates this liberation of cognition and production with “a rather special new and inwardly imposed restraint.” This new form of internal constraint is a “second-order sacralization of procedural propriety,” which Gellner (1995: 60) describes as “the rule of treating like cases alike, of conceptual tidiness, of the unification of referential concepts in an ideally unified system, and of their separation, to a remarkable extent, from the markers delimiting social conduct.” The consequence of this liberation of cognition from cultural constraint is that our social rituals are no longer taken very seriously. Serious cognition today is associated with science, not with moral philosophy or theological doctrine. However, science cannot tell us how to live. So we still use the old rituals and ethical concepts, even if these no longer have the authority they used to. As Gellner (1994: 94) puts it, the link between serious cognition and daily life is “wobbly,” because “the superior kind of truth available in science is both unstable and largely lacking in any clear social implications.” It is worth quoting Gellner (1994: 95) at length on this:

The world in which men think seriously, and to which serious thought refers, is no longer identical with the world in which one lives one’s daily life. The instability, contestability and often incomprehensibility of the serious, respect-worthy kind of cognition, and hence of its object, make it and them altogether unsuitable to be the foundation of a stable, reliable social order, or to constitute the milieu of life. The mechanisms underlying that cognitive and technological-economic growth on which modern society depends for its legitimacy, require pluralism among cognitive explorers as well as among producers, and it is consequently incompatible with any imposition of a social consensus. The attempt to impose it, in Marxist societies, in the end proved catastrophic, and helped bring about its eventual disintegration.

The upshot, as Gellner (1994: 95-96) puts it, is that

Daily life is sacralized at most in an ironic spirit. . . . Social co-operation, loyalty and solidarity do not now presuppose a shared faith. They may, in fact, presuppose the absence of a wholly shared and seriously, unambiguously upheld conviction. They may require a shared doubt. Inner-directedness co-exists with a recognition of the legitimacy or even the obligation of an ultimate doubt. Inner authority is more effective than an external one, but it leads to an ultimate sovereignty of individual inquiry, and thereby also to scepticism. If the Cartesian consciousness is the ultimate court of appeal, it is free to reach the conclusion that the evidence available does not warrant a firm decision.

This, then, is what Gellner means when he says that we moderns must live our lives in a spirit of “doubt, compromise and doublethink.”


To sum up the larger argument, Gellner’s position is basically a restatement of Durkheim’s but with two important modifications. First, he sees more clearly than Durkheim the connection between ritual and language. Language is basically a system of prohibitions; it constrains an otherwise unstable human society. Second, the theory is purely deductive. Gellner does not assume, as Durkheim (1995 [1912]) did, that a global anthropological theory must be based on the empirical observation of an actually existing culture, one that is assumed to be the most elementary available. Instead, he begins with what he takes to be the minimal assumptions necessary to make sense of the overall structure of human history. Without beginning with these assumptions, it would be impossible to make sense of historical change, because the very idea of human history, with its remarkable potential for accumulative cognitive, economic, social, and technological growth, presupposes the prior emergence of these minimal concepts or categories. The question is not whether such an “originary” hypothesis is permissible, but whether we choose to make our presuppositions explicit. It is a mistake to assume that historical analysis begins with pure description. For how do we know what to describe? A description is already a selection. The point of beginning with a hypothesis of human origin is to make this selection explicit from the beginning. If the hypothesis is judged to be inadequate when measured against the available historical facts, then it must be revised. The implementation of this procedure is, of course, no easy task. It requires a range of inquiry and knowledge quite out of fashion these days. But as Gellner (1989: 12) points out, this skepticism is itself paradoxical. Global speculative history is out of fashion among postmodernists, but the ideas of Hegel, Marx, Comte, and Spencer are nonetheless “everywhere in use” in our everyday language. Gellner’s work challenges us to think more critically about the broader pattern of human history. For this reason, he is an excellent interlocutor for those engaged in “originary thinking.”


Burling, Robbins. 2005. The Talking Ape: How Language Evolved. Oxford: Oxford University Press.

Deacon, Terrence. 1997. The Symbolic Species: The Co-Evolution of Language and the Brain. New York: Norton.

Durkheim, Émile. 1995 [1912]. The Elementary Forms of Religious Life. Trans. Karen E. Fields. New York: Free Press.

Gans, Eric. 1985. The End of Culture: Toward a Generative Anthropology. Berkeley and Los Angeles, CA: University of California Press.

———. 1990. Science and Faith: The Anthropology of Revelation (Savage, MD: Rowman & Littlefield.

———. 1993 Originary Thinking: Elements of Generative Anthropology. Stanford, CA: Stanford University Press.

———. 1997. Signs of Paradox: Irony, Resentment, and Other Mimetic Structures. Stanford, CA: Stanford University Press.

Gellner, Ernest. 1989. Sword, Plough and Book: The Structure of Human History. Chicago: University of Chicago Press.

———. 1994. Conditions of Liberty: Civil Society and Its Rivals. London: Hamish Hamilton.

———. 1995. Anthropology and Politics: Revolutions in the Sacred Grove. Oxford: Blackwell.

Girard, René. 1977. Violence and the Sacred. Trans. Patrick Gregory. Baltimore, MD: Johns Hopkins University Press.

———. 1987. Things Hidden Since the Foundation of the World. Trans. Stephen Bann and Michael Metteer. Stanford, CA: Stanford University Press.