1In order to understand the status of the political in the thought of René Girard, we shall attempt to address two closely connected questions. We shall first attempt to show that there is scant space for the political in the most recent version of Girard’s “mimetic theory:” the tripartite system of “mimetic desire,” the “victimage mechanism,” and “Judeo-Christian revelation.” We shall then attempt to determine the conditions under which the core of the system might nevertheless contribute to a theory of the political or, in other words, the sacrifices and amendments that mimetic theory would have to undergo in order to be applicable to a great swathe of social life, which at present it tends to pass over or dissolve rather than explain. These are actually two aspects of one and the same problem, the solution to the former bringing with it the solution to the latter.

2Due to considerations of space, we shall assume a general familiarity with the main tenets of Girard’s anthropology. [1] Let us simply recall that, from his third book onward, [2] Girard believes he can explain all of the characteristics specific to individual and collective human life on the basis of a single principle: the mimetic excess that distinguishes the human species from other primates. For Girard – and this is essential to our thesis – mimesis is the principal source of all the conflicts that can tear apart and poison a community, for it causes the desires of both individuals and groups to spontaneously converge on the same objects. At the same time, however, it also constitutes the cure for the ills to which it gives rise, insofar as its capacity to spontaneously focus violence on a single target – the “surrogate victim” – allows it to temporarily reconcile the community that it has divided, by way of a unanimous and cathartic act. Without this salutary, self-regulating “victimage mechanism,” human societies and, ultimately, humanity as a whole, would succumb to a mimetic crisis that could only end in humanity’s self-destruction. [3]

3Furthermore, we shall take as a given the classical definition of the political as “the totality of activities, mechanisms, and institutions through which a society maintains its inner harmony and external security.” [4] For harmony and security are not immediate and stable givens. They must be safeguarded by an authority or a system that is capable of regulating the conflicts that arise within a community or between it and other communities. It is this capacity for conflict management that constitutes the essence of political power. Let us note here that it is not necessarily the preserve of one part of the community, one that is endowed with a coercive power and that looks down on the rest from on high; it can also be embodied in the very fabric of certain social structures. [5]

A Paradoxical Absence

4Given these two definitions-that of the victimage mechanism and that of the political system as such – we might expect to find in Girard a renewal of political anthropology as much as one of religious anthropology. And yet, this is not at all the case. Although Girard brought to light a principle of conflict management which allowed him to rethink the existing theories of sacrifice and its associated rituals, and even to account for certain practical developments such as the domestication of animals, he never attempts to apply this principle to the political domain, or more generally to the analysis of relations of domination and subordination between human beings. [6] His writings contain no extended study of, nor any elaboration on, all that bears on the political. Although he often pillories social contract theories, he does not show whether and how the victimage theory might explain what they cannot, and this is the case throughout his work. Nowhere does he offer a novel analysis of political power and the forms it assumes, nor of the divisions and conflicts that it creates, nor of their possible resolution. One looks in vain through Girard’s works for his precise views on the state and its functions, the status of nations, the relations between the religious and the political, the functioning of democracy, etc. When he does broach these questions, it is only in the various interviews he has granted, which have not all been published. But there he only touches on these questions without ever subjecting them to a systematic treatment. At most, he sketches an account of the genesis of centralized power, in suggesting that the sacred king – originally a sacrificial victim granted a reprieve – was able to take advantage of the delay between the assumption of his position and his ritual immolation in order to gradually transform himself into a political leader and thereby turn the “burden of royalty” into a form of coercive power. [7]

5The absence of a political theory in his work is all the more surprising considering that Girard himself has always been concerned with life and with concrete political problems. His early works considered the relations between France and the United States, and his Battling to the End [8] betrayed a deep knowledge of the political history of Europe, a fact which surprised many readers. If, then, his mimetic theory – to which nothing that is human is supposed to be incomprehensible – passes silently over such an important aspect of social life, this cannot be due to mere negligence on his part, nor to the inevitability of some sort of lacuna in a domain that does not admit of exhaustive treatment. It is likely that there are deep-seated theoretical reasons for this silence. What these reasons are, we shall now attempt to discover.

From Triangular Desire to Mimetic Desire

6As is well known, Girard has used the expressions “triangular desire” and then “mimetic desire” in order to indicate that a subject’s relation to an object is always mediated by another subject. We do not desire things because they are desirable but because another’s desire, real or imagined, makes them appear so. But this third party is not only a model for the desiring subject. As soon as the latter starts to follow in his footsteps, the model also becomes an obstacle and a potential rival. This is the driving force of Girard’s anthropology, and the source of the continual resurgence of violence in human relations. On this point, Girard’s view has never altered, which is why commentators have generally failed to note that the notions of triangular desire and mimetic desire are not entirely equivalent, and that the slide from the former to the latter does not represent a mere terminological shift.

7At the very beginning of Deceit, Desire, and the Novel, Girard distinguishes between two forms of triangular desire: one involving external mediation, in which the mediator is an imaginary or real transcendent model (e.g. Amadis de Gaule, Napoleon, or a distinguished contemporary) with whom no rivalry is possible, and one involving internal mediation, in which the model is similar to oneself and, ipso facto, a possible rival. Although Girard himself does not remark on it, this early analysis of desire conceives of imitation as a secondary phenomenon. Desire is not triggered by mimesis, but rather, to use the classical vocabulary of the passions, in the one case by admiration for the model, and in the other by envy or jealousy. It is these passions that lead to mimetic behavior.

8Furthermore, as Marxist thinkers such as Lucien Goldmann have observed, Girard does not treat these passions themselves, or these systems of desire, as basic elements, but traces them back to distinct political and social systems. In accordance with de Tocqueville’s analyses, he associates external mediation with aristocratic societies and internal mediation with democratic societies. [9] Even if the seeds of mimetic theory were already there in his first book, [10] the fact remains that mimesis is not yet there seen as a blind drive operating in an undifferentiated space. Snobbery, which, following Hocart, Girard discovers to be as important for sociology as it is neglected by sociologists, amounts at heart to the imitation of the great figures of the world. [11] If Girard sees it as a characteristic feature of democratic societies, [12] this is because it takes on an intensified form there due to the absence of the transcendent mediator to which it aspires. Although it remains implicit, the primacy accorded to external mediation is essential, for the latter is not simply a prior condition that is destined to disappear in the imitative domino effect that it engenders, even though Girard himself will later give this impression. Indeed, we must not forget that, as Descartes and Freud rightly emphasized, we are necessarily children before we are adults and that, without the protection of adults, children would perish. Thus, whether we live in a hierarchical society or not, whether our upbringing is authoritarian or not, our first experience is of an absolute dependence on adults. By the same stroke, the latter are our first models, and perhaps the mold from which all the others are cast. External mediation is not, then, a contingent feature of the human condition; it is constitutive of it. [13]

9It is precisely this basic truth that fades into the background in Things Hidden, losing its place at the heart of mimetic theory. As soon as man is defined as a mimetic animal, and as soon as this property is supposed to govern all the others, the mimetic relation is immediately conceived as reciprocal and symmetrical. There is no more reason to be interested in the fact that A imitates B than there is to be interested in the fact that B imitates A. The two processes occur at the same time and, what is more, sustain each other. Girard therefore now sets out from mimetic rivalry [14] in order, without further ado, to describe its deleterious effects and discover its regulatory mechanism. Right from the start he focusses on the negative reciprocity of violent exchange, just as Lévi-Strauss looked from the outset to the positive reciprocal play constituted by the exchange of women. Reciprocity is the keyword of his “fundamental anthropology,” as it was for structural anthropology. He therefore runs the same risk as his predecessor. The latter had planned a study of the “structures of subordination” [15] which would have formed a counterpart to his study of the “structures of communication,” but never saw the light of day. More prudently, or perhaps simply because he is less sensitive to this lacuna, Girard passes over the issue of relations of domination and subordination in complete silence. It is not until the last part of Things Hidden that the notion of external mediation very discreetly reappears, and then only within the framework of an “interdividual psychology.” This neologism, created by Girard in order to characterize such psychology, is by itself highly indicative of his almost exclusive interest in what he calls “relations between doubles,” or the twin brothers of mimetic rivalry – or, to put it more precisely still, the conjoined twin brothers of mimetic rivalry. [16] At the end of his analysis, only one form of external mediation remains capable of resisting the erosion of differences and the contagion that occurs in relations between doubles. Between the imitation of Jesus Christ and a deadly mimetic dynamic, there is no longer any middle path toward the salvation of humanity.

The Victimage Mechanism and Morphogenesis

10Nevertheless, in the Girardian system there is one way out of the cycle of reciprocal violence and its attendant destruction of differences [effets d’indifférenciation]. This is the surrogate victim mechanism, which is both cathartic and morphogenetic. The process of its discovery and its fundamental properties were first expounded in Violence and the Sacred; its description was then taken up again and its effects more fully described in the first part of Things Hidden.

11In both of these works, Girard’s approach is that of a naturalist attempting to move from ethology to ethnology, i.e. to form plausible conjectures on the emergence of the process of hominization and the genesis of the institutions specific to human societies on the basis of observation of the animal world. Nevertheless, he does not follow quite the same path in each work. As we know, in Things Hidden, he defines the human species via two distinctive properties: on the one hand, the absence of those natural relations of domination found in other primates; on the other, a greater aptitude and propensity for mimesis, which is linked to human beings’ larger brain volume. He develops these ethological arguments, however, only after having already constructed the essential elements of his theory of the victimage mechanism [17] – which he deduces from the mimetic capacities intuitively attributable to human beings, and which he shows to be capable of generating the whole range of cultural phenomena: rituals and prohibitions, monarchical institutions, trading rules, the domestication of animals, and so on. In Violence and the Sacred, he starts out from an observation often made by experts on animal behavior; namely, the ease with which an animal that is prevented from satisfying its aggressive drives can find an alternative object for them. It is this observation which leads him to discover in human beings an analogous cathartic process, also based on the transference of violence to a substitute victim, but this time operating at the level of the community: the victimage mechanism.

12Let us imagine a society in the grip of widespread violence, a society breaking apart along a plethora of fault lines. And suppose that at a particular moment, for one reason or another, all of the resentments felt by the various members of the community come to be focused on a single victim – the surrogate victim – and that the community then rains down on him in unison. There is little doubt that such a unanimously approved, collectively committed murder, which distills the violence of the whole community and unleashes it on a single victim, should calm and reconcile the community, temporarily purging it of its poisonous atmosphere. Now suppose that, once calm has been restored and unity regained, the community, aware of having reached the brink of implosion, wishes to eliminate the risk of returning to that state. In that case it will not only introduce prohibitions to prevent violence from erupting and spreading again, but because violence cannot be completely contained, the community will also periodically provide a public and legitimate outlet for it, by means of appropriate rituals. For Girard, this is the original function of sacrifice, the efficacy of which is based on a double substitution: first, a “founding act violence” substitutes a unique victim for all the members of the community; then an act of ritual immolation substitutes a “sacrificable” victim, taken from the margins of or outside the community, for the surrogate victim of collective murder. [18]

13Catharsis and substitution are then the two pillars of the victimage mechanism and its byproducts. They are the two properties that are attributed to it from the start of Violence and the Sacred, and are, in our view, the least contestable. It is only at a later stage that Girard returns to mimetic desire in order to claim that it is the sole generative principle of both the crisis that tears the community apart and the mechanism that spontaneously heals this division. In doing so, he attributes to the mechanism itself, as to the mimesis from which it is supposed to issue, a seeming excess of morphogenetic properties. Furthermore, he holds that its cathartic and regulatory power is dependent on a strict condition which, in truth, is perhaps not essential: human beings’ misrecognition of the mechanism that they unwittingly act out. It thus remains for us to examine these two crucial aspects of the theory, in order to discern its limits.

14The first chapter of Violence and the Sacred contains a weak spot that its author has never amended. He claims that in societies without a state – which he wrongly defines as societies lacking a judicial system [19] – sacrifice would be the only form of legitimate violence. Having characterized sacrifice as a form of violence that does not risk vengeance, he then adds that, in such societies, it would also be the only effective way of eliminating the mortal danger that they face. While the victimage mechanism spontaneously puts an end to the violent excesses of the cycle of reprisals, the sacrificial ritual deliberately takes the initiative: it prevents the explosion of violent desires and the ensuing spiral of vengeance by periodically throwing these desires a bone. Because every violent act invokes another, and because all reprisals invoke further reprisals in turn, without sacrifice the whole society would rapidly be engulfed by violence.

15But is this really the case? With good reason, Raymond Verdier has raised the objection to Girard that vengeance itself might be ritualized. [20] Far from being blind, excessive, and sans foi ni loi, the vengeance that has been observed in societies without a state is on the contrary targeted, calculated, and governed by a code of honor. Far from destroying a society, it reestablishes a balance between the groups that compose it, and at the same time prevents any one group from wiping out another. For Verdier, Girard remains blind to the existence of a civilized form of vengeance, present in “vindicatory systems,” which he confuses with a notion of savage vengeance borrowed from theorists of the state of nature – which is an imaginary construct of no use to either the ethnologist or the historian. Verdier might have added that Girard reduces the vindicatory to the vindictive, just as others have lumped together sacrifice and murder.

16What is the root of Girard’s error, and why has it not been rectified either by him or by his overly zealous disciples? The cause of his error lies in a form of reasoning by association, the faulty character of which might easily go unnoticed. As he writes, “Vengeance, then, is an interminable, infinitely repetitive process. Every time it turns up in some part of the community, it threatens to involve the whole social body.” [21] As this passage makes clear, Girard surreptitiously slides, without the least proof, from a thesis concerning time to one concerning space. Later on, his mimetic hypothesis will help to make this slide smoother and even to give it legitimacy. That vengeance is an interminable process does not imply that it is contagious. Two families, lineages, clans, or nations, can bear each other an eternal hatred, passing it on from generation to generation, without thereby pulling other families, lineages, clans, or nations, into the fray, nor, a fortiori, causing the community as a whole – or more apocalyptically, the whole world – to be engulfed by violence. A segmental societal structure is, on the contrary, a barrier to such proliferation, [22] and the vindicatory logic elucidated by Verdier is a means of buttressing this firewall.

17If Girard and the majority of his commentators have failed to grasp this objection, it is not just due to lack of information; it is also because the savage vengeance and the generalized violence described in Violence and the Sacred are not purely imaginary, as Verdier claims, but are potentially present in all societies. The study of vindicatory systems itself shows that they are not perfectly stable and self-regulating, but are often on the verge of tipping over into the savage vengeance that they are meant to contain. What is interesting is that in these dangerous situations, it is usually by means of sacrificial mechanisms and rituals that the looming disaster is averted and balance restored. It is as though sacrifice were, in the final analysis, the vindicatory system’s regulatory principle or safety valve. [23]

18It is nonetheless true, however, that the vindicatory system is a particular form of legitimate violence, and though it is neither a form of sacrificial nor of state violence, it possesses no less regulatory power. To ignore it is to disregard societies without a state, and to disregard these is to deny oneself the possibility of properly understanding the state, of comprehending the specificity in which it consists. In Violence and the Sacred, the latter is implicitly reduced to the criminal justice system. Now, as the death penalty undeniably has its origins in sacrifice, [24] it follows that the decline of sacrificial practices will ultimately lead to the decline of the state. Thus, in the postsacrificial world described by Girard from Things Hidden onward, there is no longer a place for the state, nor for politics in general. Following the devaluation of all institutions, there remains an undifferentiated space which offers no resistance to mimetic rivalries, themselves now deprived of any form of internal regulation.

19In order to understand this idea of a tabula rasa, it is necessary to return to the principle of sacrifice as it is interpreted in mimetic theory. It is because the victimage mechanism is supposed to have generated everything else that it brings everything else down with it in its fall. This postulate, however, appears rather rash. For if mimesis, being the main architect of things, could therefore destroy everything, it does not follow that it could also rebuild everything by itself. One can use a hammer to destroy a computer but not to repair it, or to kill someone but not to bring them back to life. We must therefore tread carefully here.

20We can accept Girard’s definition of what he first called the “sacrificial crisis” and then the “mimetic crisis” as a crisis of a lack of differentiation [crise d’indifférenciation]: the loss of any difference between the salutary violence of sacrifice and the homicidal violence of murder, or between friends and enemies, parents and children, and masters and disciples – all of these differences are swept away in the whirlwind of mimetic rivalry. We can also concede that the unanimous alliance against the surrogate victim puts an end to this crisis, not only by providing an outlet for the pent-up violence and overcoming the divisions within the community, but also by reestablishing the difference between legitimate and illegitimate violence, and the difference between the sanctified victim and the lynch mob that has now come to worship him, once its members have been reconciled with one another through him. However, the victimage mechanism cannot achieve any more than this. It allows the differences that had been effaced by the crisis to reappear, but it neither creates nor recreates them. This is clear where generational or sexual difference is concerned, but it is no doubt also true for many other differences and hierarchies. The victimage mechanism creates a climate in which they can flourish, but it does not itself generate them.

21Girard is confronted with the same problem as that faced by social contract theorists. He improves on their analyses, but runs into no less intractable difficulties. Mimetic rivalry amounts to a more radical version of what in Leviathan Hobbes calls “the state of nature” and what Rousseau, in the Second Discourse calls “a new state of nature” or in The Social Contract, “our primitive state.” A state of war is no longer potential or hypothetical, but inevitable and necessary. It is not only brought about through competition for indivisible objects or by an accidental transformation of self-love [amour de soi] into egoism[amour-propre]. Rather, because human beings are mimetic, they inevitably covet the same objects, and when two rivals meet, they are necessarily more concerned with “the obstacle to be overcome than the object to be attained.” [25] As Hobbes says, in this state of general war, human beings seek peace as a condition of their survival. For Rousseau, the human species would perish were it not to change its way of being. But in such a situation, how are human beings to come to an understanding that would allow them to substitute peace or harmony for war? It is to Girard’s credit that he addresses this question. A unanimous act of violence against a surrogate victim is precisely “the act by which a people is a people,” the act which transforms a multiplicity into the unified totality [totalité solidaire] sought by Rousseau. And in the sanctified victim we can glimpse the dawn of the “mortal god” who for Hobbes was to take on the role of the sovereign.

22However, spurred on by, and almost as though he were intoxicated with, the success of his analysis, Girard immediately compromises it by seeking to derive all of the institutions and essential characteristics of human societies from mimesis alone. Rousseau was both more modest and more lucid when, in his Constitutional Project for Corsica, or in his Considerations on the Government of Poland, he did not base his reasoning on an abstract and general notion of humanity but on particular peoples, each endowed with a specific national tradition.

23Let us be clear: Girard is quite justified in reviving the state of nature tradition. As Hobbes and Rousseau knew, yet today’s social scientists tend to forget, thought experiments are not only legitimate but are in fact indispensable. In order to establish the principle of inertia, it is first of all necessary to imagine frictionless movement; in order to establish the law of gravity, it is first of all necessary to imagine bodies falling in a vacuum, even if nature abhors a vacuum. It is because he understood and heeded this necessity that Galileo is rightly considered the father of modern physics. Einstein proceeded in the same way in establishing the theory of relativity. Girard is then perfectly justified in asking what humanity would be if the individuals composing it were ruled exclusively by mimesis, and in establishing certain wide-ranging principles on the back of this hypothesis. His only mistake was to stop there. His mimetic theory, we might say, ignores friction. It proceeds as though all of physics followed from the law of inertia alone, as though we could describe the falling of a body without taking into account the resistance of the air. His reading of Clausewitz demonstrates this. Girard takes from him only the notion of an “escalation to extremes.” At no point does he look for a mechanism that might slow down this escalation. If Clausewitz himself appeared to seek out such a mechanism, this was only because he shrank back from his discovery, hesitating to draw all of its consequences. This is why, for Girard, it is necessary to complete his work.

24In short, it is Girard’s blindness to anything exterior to and independent of mimesis that detracts from his anthropology. Two further examples, with which we shall conclude this section, will allow us to demonstrate this, should any further proof be required.

25Despite the fact that it plays a significant role in all societies, Girard says almost nothing about sexual difference. This is detrimental to his account of initiation ceremonies, which focuses exclusively on their sacrificial dimension. Having deduced from the mimetic principle that desire “lacks an object,” [26] he does not see that in most societies initiation ceremonies serve as the locus of a battle of the sexes for control over procreation. [27] This oversight is equally detrimental to his general theory of the social bond, which is likewise exclusively based on the victimage mechanism and mimesis, and supposes that these operate in a neutral and homogeneous social space. Yet it can hardly be doubted that, as Hume saw, the mutual attraction between the sexes is one of the main sources of sociability. It works to institute relations of equality between parents and between children, as well as hierarchical relations between generations, and divides the social sphere into relatively stable and autonomous units. As mimesis is, for Girard, principally a source of violence, it is likely that such family ties could serve as a counterpoint to its destructive effects.

26Girard often speaks of human “communities,” but never of their limits, nor of the territory that they inhabit. [28] More precisely, he does offer a hypothesis on the origin of the point around which any group assembles – this is of course the site where the surrogate victim was killed – [29] but he offers no theory of and, indeed, does not even mention, territorial borders. [30] Nevertheless, many sacrificial rituals are performed on the edges of a community’s territory, and for peoples that lack political bodies, such rituals can serve as a means to extending their territory and absorbing new populations. [31]

The Victimage Mechanism and Ignorance Misrecognition

27Another contentious aspect of mimetic theory is that it sees misrecognition [méconnaissance] as a sine qua non of the successful functioning of the victimage mechanism. It is only because human beings know not what they do, only because they are the pawns of the victimage mechanism, that the latter can reconcile them at the victim’s expense. If they knew that this victim were neither more nor less responsible for the ills afflicting the community than any other of its members, his murder would lose its cathartic power and would even become impossible.

28Girard clings to this hypothesis because it constitutes the fundament of what he calls “Judeo-Christian revelation.” The latter is nothing less than the slow unveiling of the victimage mechanism over the entire course of the Bible, an unveiling which is consummated in the drama of the Passion. It follows that all of the rituals and institutions that have their origins in the victimage mechanism – in other words, all of the known forms of civilization – would begin to totter when irremediably undermined by the discovery of the murders and the corpses on which they rest. This is not the place to discuss this interpretation of the revelation. We shall only attempt to show that the premise on which it is based is extremely fragile.

29We would not deny, of course, that misrecognition plays a significant role in human relations, particularly in violent relations. To this extent, Girard’s account of mimetic rivalry is convincing. The more the conflict between two rival parties escalates, the more they come to resemble each other, and the more they imagine themselves to be different from each other. In general, a third party does not fall for this illusion. But it can happen that the proximity between the opposed values, of which the two parties claim or believe themselves to be the last line of defense, succeeds in going unnoticed by anyone.

30We can therefore concede that both the victimage mechanism and the rituals that flow from it are always accompanied by a significant degree of misrecognition; but does it necessarily follow that without misrecognition the mechanism would be completely ineffective and the sacrifice useless? A few examples will suffice to show that that is unlikely to be the case. Let us begin with an example that Girard knows well, that of the Buphonia. In the course of this ceremony, an ox is sacrificed in a very precise ritual manner, the details of which it is not necessary to recall here. However, this sacrifice, whose name means “ox-slaying,” has a very distinctive character: it is followed by the simulacrum of a trial, as though the killing of the ox were a crime for which the guilty party ought to be punished. Indeed, the one who felled the ox immediately fled the scene, as though he had done something wrong. Unable to fall on him, the remaining participants begin to accuse each other of bearing the principal responsibility for the victim’s murder, until the point when the sacrificial knife, [32] accused in turn and unable to defend itself, is found guilty and thrown into the sea. It is evident that the violent origins of this sacrifice – attributed by Girard to all rituals of this kind – do not need to be concealed in order for it to constitute a religious and legitimate act; on the contrary, these origins are made quite clear in the second act of the ritual, when the knife takes the place of the surrogate victim. The ritual is both a perfect illustration of the theory of sacrificial substitution and a refutation of the theory of ignorance.

31We might say the same thing of the Jewish scapegoat ritual (Leviticus, 16), which, curiously, Girard has never analyzed. Two goats that are as similar as possible are chosen; lots are then drawn in order to determine which will be offered to Yahweh and which to Azazel. The first – the ritual victim – is burned as a sin offering; the second – the scapegoat or surrogate victim – having been burdened with all of the sin of Israel, is sent into the desert. The text only states that, on his return, the man who has led the goat to Azazel must cleanse himself. Many commentators claim that his mission was to cast the goat into a ravine. [33] Sacrifice in its literal sense is thus closely associated with an ignominious death. Here again, the ritual ceremony bears not only the trace but, as it were, the ritual memory, of its violent foundations.

32This, however, is not all. We know that Jewish sacrifices completely disappeared during the first century AD. This was not because the realization or revelation of the sacrifices’ violent origins invalidated them, but because since Josiah’s reforms all sacrifices had to be performed in the Temple, and this had not been rebuilt following its destruction by the Romans. Pious Jews have contented themselves with reading the sacrificial prescriptions contained in the Law on the appropriate dates and times of the ritual calendar. When performing the ritual in which we are most interested, however, they do much more than this. As Frazer writes, “Modern Jews sacrifice a white cock on the eve of the Day of Atonement […] The father of the family knocks the cock thrice against his own head, saying, “Let this cock be a substitute for me, let it take my place, let death be laid upon this cock, but a happy life bestowed on me and on all Israel. Then he cuts its throat and dashes the bird violently on the ground. The intestines are thrown on the roof of the house.” [34] Leaving aside its nonliteral use of the verb “to sacrifice,” this passage describes a private version of the collective scapegoat ritual, bringing to light in startling fashion the violence inherent in it. Girard often compares the ritual process to a criminal who continually attempts to efface the traces of his deed. Here, the opposite is the case. The high priest and the sacrificial victim have disappeared, but the ritual brings the surrogate victim and its terrible fate to the forefront. The main byproducts of the victimage mechanism are lacking, but the mechanism itself and its violent mise en scène remain, as though it were more stable than the rituals it has engendered, or even indestructible.

33A great deal of evidence supports this hypothesis. Machiavelli, for example, tells of how Cesare Borgia repaid Remirro de Orco for his good and faithful service. Given carte blanche by Borgia in his mission of pacifying Romagna, this “cruel and ready man” soon drove the brigands from the region and restored order and security, to the general satisfaction of its inhabitants. But his brutal methods nevertheless caused collateral damage and incited much resentment. Borgia therefore, “had him placed one morning in the piazza at Cesena in two pieces, with a piece of wood and a bloody knife beside him. The ferocity of this spectacle left the people at once satisfied and stupefied.” [35] Like a floorcloth that, soiled with all the dirt it has picked up, is discarded once it has fulfilled its function, like the pharmakos that was either put to death or banished from the city, but only after having been made to pass through all of the streets and squares in order to carry away their pestilential odor, the faithful servant is served up for public condemnation, in order that all of the violence and disorder that in one way or another he had a hand in be swept away along with him. Here we are no longer dealing with a ritual, but with the political exploitation of the victimage mechanism. It crudely reveals to us the workings of such rituals, without this revelation damaging the mechanism’s efficacy in the least.

34We also find the mechanism at work in the genesis of the modern world. Let us compare the traditional coronation ceremony of the Mwami – the king of Rwanda [36] – with the speech given by Saint-Just on 13 November 1792, during the trial of Louis XVI. In a crucial respect, these bear a significant resemblance. The Rwandan king is treated as a condemned man. A bull is sacrificed in front of him and finished off like a wild beast. The king then climbs on to the animal’s flank and is drenched with blood in order to emphasize their identification as fully as possible. An official presiding over the ceremony then recites the following lines: “I sentence you to wounding by spear, broadsword, vireton, rifle, club, and machete. If any man or woman has died from an arrow or spear wound…I shall inflict these wounds on you.” As Luc de Heusch has noted, “we can see the Mwami being transformed into a scapegoat, burdened with the crimes of the whole kingdom.” Exactly the same could be said of Louis XVI as he appears in Saint-Just’s violent indictment, particularly when the latter attributes to him, in extensive detail, the guilt for all the blood spilled since the beginning of the Revolution: at the Bastille, at Nancy, on the Champ de Mars, at Tournay, at the Tuileries, and so on. While in the Rwandan ritual the king’s life is spared, first by sacrificing an animal in his place, then by allowing him to escape from a burning wooden hut in which he has taken refuge, the hot-headed revolutionary recommends that the Convention not waste any time on a judicial ritual, to which Louis, guilty of the greatest evils ever inflicted on the nation, has no right. As he says, those whose task it is to “found a republic,” ought on the contrary to do away with the tyrant as the Romans rid themselves of Caesar. Even though his recommendation was not followed to the letter, and though the trial did go ahead, Louis XVI was found guilty of monstrous crimes and decapitated; and though the decision was far from unanimous, [37] his death did in fact serve to irreversibly found the republic. In this way, the Revolution did not strip politics of its sacred dimension but only replaced the king’s sovereignty with the no less mysterious concept of the Nation. Everything that had been royal became national. The nation is the religious complement of the republican state, with its cult, its rituals, and its sacrificial demands inscribed in black and white in the Marsellaise and the Chant du depart [38] – but it took many decades for this cult to become established. For more than a century, the monarchists were able to give the impression, or labor under the illusion, of being capable of regaining power, without ever succeeding in reviving a past that had died [passé révolu]. Nevertheless, it was not a political defeat that definitively took the reins of power from them, but what has fittingly been called the “sacred union” against the common enemy, i.e. the new cult of the nation, consecrated in blood under the auspices of the victimage mechanism once again.

35The common counterargument to this idea is that nations are now themselves in decline and destined to disappear. As widespread as it is in certain western countries, this impression is surely misleading. The least stable political entities are great empires and multination or multiethnic states, which is why, far from disappearing, borders rather have a habit of multiplying. [39]

36Two of the most significant political phenomena of the last century – communism and National Socialism – are also two of the most significant religious phenomena. Both regimes developed similar, imposing rituals; both made scapegoats of the “classist enemies” on the one hand and the “stateless Jews” on the other. However, as both attempted to build a new empire, we should not be too quick to conclude that their failure demonstrates the decrepitude of the victimage mechanism. Among the many causes of their downfall is no doubt a common overestimation of the power of human will, which led on the one hand to such a defiance of the laws of nature as to induce economic misery and on the other to a headlong rush into suicidal wars of conquest. By itself, the victimage mechanism is incapable of halting such self-destructive processes; once again, it is not all-powerful. However, it does not need to be so in order for it to be omnipresent. We have just traversed several centuries in the history of the western world, without seeing it bend or show signs of weakening. Why should it be otherwise in the future? It is unlikely that we have seen the last of it.

A Girardian Theory of Democracy?

37An African custom previously served as the inspiration for an article that we devoted to the ritual origins of democracy. [40] In this case, a ritual hunt was resorted to in order to resolve a futile dispute that threatened to degenerate into a violent confrontation between two hardline factions. At the end of the hunt, the dispute was resolved in favor of the camp that brought back most game, and culminated in a feast which reconciled the whole community. We argued that our democratic process – which consists in counting votes rather than the spoils of a hunt in order to settle political debates, sometimes of enormous consequence – is no more nor less rational than this customary practice. What is important is rather that there is unanimous agreement over the process to be adopted and the largest possible number of participants in it. We have renounced the ritual hunt and lost its cathartic effect, but we are left with the hunt for votes and the low blows and mud-slinging of our electoral campaigns. Judging by its history and lexicon, voting itself bears the traces of its violent origins. Votes were initially cast with stones, shards of broken pottery (the original meaning of “suffrage,” for the Latin suffragium) or shouts. The peaceful process with which we are now familiar was born amidst this sound and fury. In a myth related by Walter Burkert, Hermes is stoned with voting pebbles. This myth seems to reverse the real course of events or to describe a regression: in the beginning, we stoned individuals, then we put these stones to a new use in voting – first in courts, and then in other assemblies. [41]

38As Hobbes knew, the first task of politics is to ensure a state of peace, if not of harmony. But his solution to the question of how peace might be attained is incompatible with his own theories, just as it is with those of Girard. If every indivisible object is ipso facto an object of potential rivalry, and if sovereignty is by definition indivisible, it is difficult to see how human beings, however strong their desire to escape the war of all against all may be, could come to choose the same sovereign and accept his authority. They will rather fight over who should exercise sovereignty.

39Furthermore, history tells us that the majority of those who desire power desire it as an end in itself rather than as a means to an end. The more individuals are of this tendency and are prepared to eliminate rivals in their own camps as much as, and even more than, their adversaries in opposing camps, the more likely they are to gain and wield power. This is hardly surprising. It is a political illustration of the Girardian theme par excellence: the violent enmity of twin brothers. In such a state of affairs, the more institutions tend to concentrate power and render it unassailable, the more they provoke frustration and resentment. On the contrary, it is by dividing up power in order to weaken it, and by rendering it precarious in order to ritualize the process of its conquest, that these institutions might exercise some control over the outbreak of violence. In other words, the best hope for the preservation of people and property does not lie in the attempt to use politics to eliminate all conflict, but in making political power itself the objective of most conflicts.

40This is exactly what occurs among the Nambikwara, where competition to become a chief, and thereby to gain the right to polygamy, staves off a competition between the men for possession of the women, which would prevent everyone from marrying and thus put the community in danger. [42] The unity of the community is maintained by substituting an institution for a group of individuals, and by making it the focus of the principle conflicts that might break out between the members of the community.

41But this model is also applicable to large modern societies. It allows us to better comprehend the nature of Hobbes’ theory, which sees civil war as inevitable in the absence of an indivisible power, or that of Rousseau, which rejects the party political system. Both are determined to avoid the confrontation of particular wills to the detriment of the general will. Montesquieu showed that it is necessary to divide up power in order to neutralize the threat of violence, while Schumpeter has shown that democracy is inseparable from a state of competition between several political parties. [43] Such pluralism is not only necessary for the protection of freedom, but also in order to quell a lust for power and to ritualize conflicts.

42Girard’s work could be thought of as a clarificatory supplement to this last point, even if his predecessors had already laid the essential groundwork for it, albeit in an incidental and unsystematic manner. For as the history of ethnology has taught us, wherever relatively stable human societies are concerned, we are never confronted with an undifferentiated mass, but always a segmented body containing divisions within itself and ritualized competition between the groups that compose it. Only the configurations and the denominations of these groups change. In this regard, Murdock rightly draws a connection between the bipartisan nature of modern democratic societies and the many “dual organizations” present to some degree throughout the world – structures composed of two opposed but complementary moieties which are connected via the services they render each other and the ritual confrontations between them, and which are associated with, respectively, Elders or Youths, Heaven or Earth, Right and Left, Lightness and Darkness, and so on. As he notes, the jousting matches that take place within these structures help them not only to control outbreaks of endogenous violence, but also to prevent this violence from accumulating and overwhelming other communities. [44]

43If we look closely, we can see that democracy in no way abolishes the divide between people and power. It is not, nor could it ever be, the government of the people by the people. In reality, the people never govern, and as Schumpeter notes, we can only make it govern by definition. The people can choose who will govern them, and democracy consists in this and nothing else. This choice, however, is only valid on two conditions: one, that a number of candidates are in the running; and two, that those who are elected can be discharged from their offices. The capacity of an unhappy electorate to dismiss its governors is the touchstone of the democratic system. A prime minister, as Schumpeter explicitly states, is a scapegoat. The official function of a government minister is to ensure that the will of the people prevails in her electoral district. Her main function is to serve, when necessary, as a safety valve. [45] During a heat wave, we no longer sentence the rainmaker to death, but we do get rid of the health minister who was unable to ensure that the most vulnerable had enough to drink.

44There is no need to go into further detail in order to see that the victimage mechanism has more than one trick up its sleeve, and that it underpins modern societies just as much as traditional societies. If Schumpeter is right, it even constitutes one of the driving forces of democracy. Let us just note here that this has unpleasant consequences for the “demi-habile” slumbering in each of us. It is, for example, pointless to lament or make a mockery of the excessive number of candidates for the presidency of the French Republic, on the pretext that certain candidates have no chance of being elected, or to maintain that political rallies, which reduce complex problems to simplistic slogans with floods of rhetorical eloquence, are a waste of time and energy, or to become indignant at the barbs traded by the various candidates, etc. It is inevitable that in the tumult of election season far greater focus will be put on the clash of personalities than of political programs. Still, in the agora just as in the courtroom, it is surely healthy for the views and emotions of the different groups to be expressed and challenged with some vehemence. This theatrical dimension of politics is a source of sadness for independent spirits, but the shared ritual roots and cathartic processes of politics and theatre may well make the two forever inseparable. Inverting Clausewitz’s expression, we might even say that politics is the continuation of war by other means. In modern democratic societies, this takes the form of a covert war which spares the population the horror of riots and massacres.

45Of course, this conception of politics might be deemed too reductive. It will be pointed out that power is also a means of modifying the norms at work in a society and the range of possibilities available to individuals and groups, [46] and that political battles are not only about rival ambitions but also about values. This is quite true, and if such values come to be opposed in the form of Good versus Evil, the ensuing conflict can turn into a civil war and degenerate into the massacre of one camp by the other. What generally prevents such a disastrous outcome is the fact, noted and analyzed by Tocqueville and Pareto, [47] that the majority of conflicts do not involve masses but resolute minorities united by common values – elites, in Pareto’s sense. A self-confident and united elite is almost impossible to remove from power. It is internal doubt and division that leads to its decline, allowing it to be ousted by a new, more determined elite. This principle prevails under almost all regimes, albeit according to different modalities. A revolution is never brought about by the people, but by minorities transforming the vague aspirations of this or that element of the people into a political will and a political force. As in Westerns and in traditional societies, it is small groups that square up to each other, with the people rallying to the camp of the victor. The latter’s victory is the proof of its legitimacy. [48] In all domains, whether political, moral, aesthetic, or otherwise, it is the elites who set the agenda, who decide whether something is either no longer possible or, on the contrary, has become indispensable. We are probably wrong, then, to speak of a “democracy of public opinion.” In democracies too, the elite passes judgment and the people follow; the latter are then consulted and a poll or an election “confirms” that the elite really did express, and thereby sanctified, the opinion of the people. Each individual prides himself on having his own opinion and rejecting what his fellow men regard as a superior authority; but still he bends to collective opinion or feels guilty about straying from it. It is under the influence of the external mediation from which we set out that shifts in values take place or that a “tyranny of the majority” comes to be established. [49]

46As concise as they are, these reflections show that the Girardian hypotheses that seemed to us most plausible in their original domain might prove equally relevant to the consideration of political institutions and activities. If it does not yet prove this, this short exercise has at least given an indication of the robustness of these hypotheses.