Jealousy and Triangular Illusion:

A Reformulation of the theory of Mimetic Desire

Jean-Pierre Dupuy
École Polytechnique, Paris
& Stanford University

Everyone thinks that he alone is condemned to hell,
and this is what makes it hell.

— René Girard

[T]hey were a fragment of another world, of an unknown
and accursed planet, a glimpse of hell.

— Marcel Proust

thus does the jealous imagination confer as much
significance on the possible as on the real.

— Nicolas Grimaldi

[I]f by any chance the reality presented itself
to our eyes, it would be so far outside the
limits of the possible that . . . we should
fall over backwards in a daze.

— Marcel Proust

René Girard's first work, Desire, Deceit, and the Novel, dates from 1961.1 For a first book its brilliance was astonishing, and it probably remains his masterpiece. In it Girard expounded a theory of mimetic desire with reference to "great literature" and, more specifically, the Western tradition of the novel. Rereading this book on the fiftieth anniversary of its publication, one realizes that the themes of the two great works that came after it during the course of the following decade, Violence and the Sacred and Things Hidden Since the Foundation of the World2— which is to say the theory of the joint origin of religion and human society, and the interpretation of Christianity as the religion that brings religion to an end— were already to be found in embryo in this first book. Whereas the later anthropological work was able to draw strength from the controversies it aroused and the criticisms that were directed against it, the theory of mimetic desire— or mimetic theory, as it is more commonly known— did not elicit the same kind of critical elaboration for, paradoxically, it was recognized at once as a major breakthrough in the attempt to identify the wellsprings of human actions and behavior. Over the next fifty years it came to be applied to a great many fields, but no one has really thought to inquire into its conceptual foundations, as though they had been laid once and for all. It is not too late to do this, however and the celebration of the fiftieth anniversary of DDN furnishes us with the perfect occasion. One does no injury to René Girard's genius by taking up such a challenge— quite to the contrary. Girard himself has always claimed that mimetic theory is a scientific theory, even if it is not wholly reducible to one; at all events, like science itself, it is a living thing only insofar as it can be put to the test of facts and reflection. It is exactly this that I propose to do in treating an apparently minor problem that will nonetheless in due course reveal itself to be one of surpassing significance.

    What is the true place of jealousy, as distinct from envy, in mimetic theory? Not long ago I put just this question to a few scholars in the field. Taken by surprise, none of them knew what to say. The truth is that Girard himself did not lay any particular emphasis on jealousy, by comparison with envy, subsuming both of them under the category of mimetic desire. This may have been a mistake.

    My own thinking was shaped by the possibly fortuitous conjunction of three things. First, Nicolas Grimaldi's most recent book, Essay on Jealousy3, which impressed me more forcibly than any work of literary criticism since DDN itself. To be sure, Grimaldi does not share Girard's grand ambition of constructing from literary sources a general anthropological theory of desire. He contents himself, more modestly it would appear, to reading the work of Proust through the prism of jealousy. And unlike Girard, again, he sees Proust's work an exception to something like a rule of literature. "The experience of jealousy," he says, "is almost as common as that of love. It is therefore surprising to find so few descriptions of it in novels, when almost all of them are devoted to love." [EJ, 7; my emphasis] For Girard, of course, all great literature speaks of only one thing: mimetic desire. Nevertheless the quotations from Proust that Grimaldi gathers and regroups in a suggestive way compose a clinical picture in which one cannot help but recognize a trait no less universal than mimetic desire. Could it be that at bottom they are one and the same thing?

    The second thing that got me to wondering was my reading of another, though very different kind of text. In a programmatic essay aimed at testing the mimetic model,4 Mark Anspach seeks to refute the idea that the propositions of mimetic theory do not lend themselves to empirical confirmation or disconfirmation, and reviews a number of experiments in cognitive and developmental psychology where mimetic theory seems to account for the observational data more satisfactorily than other interpretations. One of these experiments is of the greatest interest, less for what it shows than for the awkward position in which it puts the very people who designed it. It involves a triangle whose three apexes are a baby, its mother, and a doll the same size as the baby to which the mother pretends to pay great attention, provoking the anger of her own child, which is directed toward both the mother and the doll. Curiously, the experimenters seem to have had great trouble in making sense of these events. Anspach, for his part, claims to provide an explanation based on the familiar triangle of mimetic desire comprised of a subject, a mediator, and an object. He cautions, however, that there is an interpretive trap here, for the occupants of the apexes of the triangle are not the ones that first come to mind. I shall say no more of this for the time being. It is enough simply to say that most people— without looking for difficulties where there are none, without having read either Saint Augustine or Jacques Lacan— would spontaneously say that the baby is jealous. If mimetic theory has a hard time recognizing infantile jealousy, it is, I believe, because jealousy poses a serious problem for it.

    Finally, the circumstances of my personal life led me recently to revisit certain past experiences that remained partly hidden from view in the half-light of jealousy. The reader need not worry, I have no interest in emulating Proust in his search for lost time. Nevertheless I was able to recover memories of torments very similar to the ones analyzed by Nicolas Grimaldi. It seemed to me that they must have occurred very early in my life, probably before I had any experience of triangular desire. Does jealousy precede mimetic desire, even to the point of supplying a basis for it, rather than the other way around? Is the hell imagined by Proust populated in reality not only by adults, but also by babies and prepubescent children? In what follows I hope to seek to give a few preliminary answers to these questions.

1. The Absent Triangle of Mimetic Desire

Mimetic theory holds in principle that all human desire has a triangular structure. Desire does not take the form of a straight line connecting a subject with an object, but of a triangle, composed of a subject, an object, and a third term, the model. This model is the mediator of desire in the sense that the subject's desire develops in accordance with it; in other words, his desire for the object is not spontaneous, it is the result of imitating the desire of his model. When possession of the object cannot lend itself to sharing (inevitably one thinks of the exclusive passion of a man for a woman), the model, who is seen as a positive figure, worthy of imitation, is mechanically transformed into a rival, who is now viewed negatively— without his status as a model thereby being altered, and without any conscious motive intervening. Quite to the contrary: the more pronounced the rivalry, the more the model becomes both an object of fascination and an obstacle on the road leading to the object. Model and obstacle reinforce each other while the object constantly increases in value, in keeping with a dynamic to which Girard gives the name (borrowed from the anthropologist and cybernetician Gregory Bateson) "double bind." Here is what Girard has to say on this subject in TH, a more considered statement of the position originally formulated in DDN:

    The object of desire is indeed forbidden. But it is not the "law" that forbids it, as Freud believes— it is the person who designates the object to us as desirable by desiring it himself. The non-legal prohibition brought about through rivalry has the greatest capacity to wound and traumatize. This structure of rivalry is not a static configuration of elements. Instead the elements of the system react upon one another; the prestige of the model, the resistance he puts up, the value of the object, and the strength of the desire it arouses all reinforce each other, setting up a process of positive feedback. Only in this context does it become possible to understand what Freud calls "ambivalence"— a pernicious force that he identified but was unable to explain adequately...
    Once he has entered upon this vicious circle, the subject rapidly begins to credit himself with a radical inadequacy that the model has brought to light, which justifies the model's attitude toward him. The model, being closely identified with the object he jealously keeps for himself, possesses— so it would seem— a self-sufficiency and omniscience that the subject can only dream of acquiring. The object is now more desired than ever. Since the model obstinately bars access to it, the possession of this object must make all the difference between the self- sufficiency of the model and the imitator's lack of sufficiency, the model's fullness of being and the imitator's nothingness.
    This process of transfiguration does not correspond to anything real, and yet it transforms the object into something that appears superabundantly real. Thus it could be described as metaphysical in character. We might well decide to use the word "desire" only in circumstances where the misunderstood mechanism of mimetic rivalry has imbued what was previously just an appetite or a need with this metaphysical dimension...
    The "metaphysical" threshold or, if we put it a different way, the point at which we reach desire properly speaking, is the threshold of the unreal. [TH, 295-297; my emphasis]

    The reader will have noted the phrase that I have placed in italics: "the object that [the model] jealously keeps for himself." It is true that jealousy is referred to here, but it fails to rise to the status of a concept. Nor does Girard say anything elsewhere to alter this state of affairs. Nevertheless it is important to keep in mind that the subject is held to perceive the model's desire for the object as a jealous longing from the moment that triangular desire begins to have effect.

    It is therefore by definition that all desire in Girard is "metaphysical" in the sense that it transfigures both the model and the object. Is this enough to make it triangular? At the risk of offending Girard's disciples, I maintain that the basic figure of the triangle— subject, object, mediator— is not in fact met with in its pure state in any of the multiple and protean forms assumed by mimetic desire; or rather that, if this triangle can be said to appear, it is always as a transfiguration brought about by the subject's mimetic rivalry, and never as an analytical tool that a Girardian theorist could use for the purpose of demystifying desire. Either the triangle turns out to be an arbitrarily isolated part of an expanding rhombus, or else it lacks an apex (unless perhaps two apexes merge to form a single apex). For a mathematician, admittedly, two secant lines can be said to be a hyperbola of a sort, one straight line a parabola, a point a square or a circle, as the case may be (which means, by the way, that there is such a thing as a square circle— a figure that has a radius or side of zero length). Mathematicians apologize for transgressing against our folk geometry, designating such figures as "degenerate" cases of canonical forms. But the famous triangle of mimetic theory is nothing more than a sheer metaphor, and as such it cannot be allowed to abuse intuition. If Girard's purpose was to show how mimetic desire functions, to illustrate its extraordinary powers of transformation and evolution by demystifying it, then the triangle was a most unfortunate choice.

    The only mimetic pattern that one might label triangular if one were hard pressed to, is what Girard in DDN calls external mediation— a notion that subsequently he was to abandon, and this choice is significant. External mediation occurs when the distance— ontological, spiritual, social, cultural, geographic, and so on— between a subject and his model makes it unlikely, if not actually impossible, that the subject, in imitating the model's desire, will poach on his territory. Consider the traditional relationship between father and son, master and disciple, teacher and pupil. Even here the dynamic of the double bind may be set in motion— for example, when the disciple comes to be the equal of his master and threatens soon to surpass his achievements— but the rivalry that comes about in this way is repressed in the mind of each of them, since no other relationship than hierarchy can be publicly acknowledged.

    In external mediation, three apexes are present and distinct. But can they be seen as forming a triangle? In the transfigured perspective of the subject they do, of course, since model and object are connected to each other by a straight line— which amounts to saying that the subject believes that the model, being self-sufficient, escapes the hell of imitation, and therefore that his desires are spontaneous. But this vision cannot be the view of the mimetic theorist, who maintains that no one escapes mimetic desire, neither the model nor his subject, and that the model therefore himself must have a model. Facing the subject, then, the theorist sees, not a straight line, but another triangle joined to the first one, and then a third triangle joined to that one, and so on— in short, a string of triangles. A string of triangles is not a triangle.

    Besides, external mediation inevitably gives way to internal mediation when the mediator is drawn near to the subject, to the point of becoming indistinguishable from him. Then the double bind turns into a rivalry that tethers them to each other, so that each one sees the other as a mesmerizing reflection of himself. The rivals' perception of the dynamic that propels them onward to ever more imitation and ever more violence now reaches a tipping point: henceforth they see only rivalry, while fiercely denying any imitation, with the result that each one becomes so important for the other that the two of them lose sight of the object of their desire and the point of their quarrel. Let us reread what Girard said in DDN about this destructive process:

The subject is torn between two opposite feelings toward his model— the most submissive reverence and the most intense malice. This is the passion we call hatred.
    Only someone who prevents us from satisfying a desire which he himself has inspired in us is truly an object of hatred. The person who hates first hates himself for the secret admiration concealed by his hatred. In an effort to hide this desperate admiration from others, and from himself, he no longer wants to see in his mediator anything but an obstacle. the secondary role of the mediator thus becomes primary, concealing his original function of a model scrupulously imitated. [DDN, 10-11]

I spoke a moment ago of the disappearance of the object. One should refer instead to the disappearance of interest in the object, interest and attention being exclusively transferred to the model-obstacle as the double bind of mimetic desire grows more intense. This is a very important theme in Girard, and one that many people on first encountering his work have trouble understanding and assimilating. Girard himself jealously insists that mimetic theory completely revises everything that until now had been thought about imitation, so that imitation is understood to bear not only on ways of being and of appearing, of behaving and of feeling, but also on the desire to possess an object; indeed, he goes so far as to coin a term for this urge: acquisitive mimesis. How can it be, then, that the stronger the rivalry the more desirable the object becomes, to the point of being transformed into a metaphysical entity, into the hypostatized nothingness that we call prestige— and yet the object suddenly disappears, as though a magic wand had been waved over it?

    The truth of the matter is that the object, in internal mediation, does not preexist the desire. The basic triangle lacks one apex. What each of the rivals sees, from his transfigured perspective, is a single indissoluble bond of jealousy that, in directly linking the rival to the object, makes the rival a self-sufficient being. This bond may be real, at least to some very partial extent; in most cases, however, it is wholly imaginary. This is true of what Girard calls double (or reciprocal) mediation. One of the finest illustrations he gives of this concept has to do with the making of Julien Sorel in Stendhal's The Red and the Black into a desirable object, and therefore into a subject. The village mayor, M. de Rênal, has the idea of engaging Julien, a young man without any notable qualities apart from an obsessive attachment to the image of Napoleon, as the private tutor of his children. Stendhal is careful to make sure that we understand the true reason for his initiative: M. de Rênal believes (wrongly, as it happens) that Valenod has the same thing in mind. Valenod, a wealthy liberal bourgeois who wields less influence in local affairs than M. de Rênal, a conservative aristocrat, is obsessed with the mayor; M. de Rênal, in his turn, is obsessed with Valenod. In theory the aristocracy has lost its privileges, but the bourgeois are more fascinated with it than ever. The aristocrats, for their part, are fascinated by the bourgeoisie. Later on Stendhal shows us Valenod approaching Julien's father about the possibility of taking on his son as a tutor. "Is Stendhal confusing the Valenod of M. de Rênal's imagination with the real Valenod who [until now] hasn't given a thought to Julien?" Girard asks. [DDN, 101] The author does not commit this error: this passage is a story of the copied copier— a copier who is copied in his turn. Valenod is M. de Rênal's mediator. And yet, little though he suspects it, Valenod himself imitates the desires of his disciple-model. Two persons who imitate each other in desiring an object lack a crucial piece of information: the identity of the object. But this is no obstacle, they imagine it. More precisely, one of them first imagined it about the other. M. de Rênal, anxiously wondering what his model-disciple really desires, invents the answer and moves at once to preempt him— something that Stendhal reveals to us later in the novel. By inventing the answer, the mayor signals to his alter ego what is at stake in their rivalry, so that when Valenod manifests his own desire, which is an imitation of M. de Rênal's, the initial illusion has become reality. M. de Rênal, the first to imagine that such an object of desire existed, now has proof that he was not dreaming. "Reality springs from the illusion," Girard observes, "and provides it with a misleading guarantee. It is by an analogous process that peoples and politicians blame each other, with the greatest possible sincerity, for the conflicts between them." [DDN, 101]

    Mimetic struggle is waged in order to decide which of two competitors will be the exclusive focus of desire. It is as though there were an imaginary object— I shall call it a "quasi-object"— that the rivals seek to gain possession of, since triumph in this contest channels the admiring regard of others toward the victor and further increases the chance that he will continue to possess the quasi-object at the next moment. The convergence of desire is a self-reinforcing mechanism— what Girard would call a positive feedback. Every culture seems to have interpreted this mechanism in objective terms by reifying a relational dynamic. What is more, every culture has given it a name: the Greeks spoke of kudos, the Polynesians of mana, the Maoris (studied by Marcel Mauss) of hau, and so forth. We ourselves use words such as "prestige," whose etymologies remind us that they refer to an illusion. The quasi-object is an object of desire. However— and this is the paradox of mimetic desire— one must disavow any claim to it oneself in order that the mediator may possess it— and this, not through either generosity or some sort of masochism, but because the quasi-object has value, even existence, only if it is possessed by a model. the double bind of mimetic desire therefore assumes the following form:

        1. I wish to possess the quasi-object because it has value;
        2. For it to have value, I must in no way possess it.

    If the mediator possesses the quasi-object, I want to have it; but if I gain possession of it, it vanishes into thin air. For me to be able to desire it and gain possession of it, it must exist, and therefore the mediator must possess it; but in conquering it, I lose it. No one expressed this point better than the heroines of Corneille's comedies. "When I want to be rid of him," Célidée reflects in the La Galerie du Palais, "he is the perfect lover/When I want to be loved by him, he no longer cares"; or, still more succinctly, Angélique in Place Royale: "If I love I am betrayed; I betray if I am loved." Angélique's lover, Alidor, having quit her because her love, he says, keeps him in chains and he is eager to regain his liberty, and having "given" her to his best friend Cléandre (who, as it happens, is secretly in love with her), now changes his mind and does his utmost to win her back. La Place Royale portrays three reversals of this type.5

    Each Molière's Don Juan and Da Ponte's Don Giovanni tries to evade this paradox by throwing himself into an impossible task, the arithmetic exhaustion of reality— mille e tre, and so on. Of course, the properties of the quasi-object seem paradoxical only insofar as one insists in regarding as an object what is really a relation of desire. No object has the properties of a quasi-object. The basic mimetic triangle has lost one of its apexes.

    In place of a triangle, one finds a form whose features now begin to emerge more clearly: on the one hand, a subject; on the other, in the subject's imagination, an exclusive relation between another subject and a quasi-object. The key word here is "exclusive," for the relation that the subject constructs in his mind excludes him, and it is because it excludes him that he feels cruelly deprived of the very property he ascribes to the other— self-sufficiency. Note that we are not dealing here with a triangle, even if it were to be saddled with the prefix "pseudo-" or "quasi-," because the subject is not faced with two elements, another subject and an object; instead he sees the relation between this other subject and the quasi-object, which for him constitutes an indissociable and unanalyzable whole. The plausibility of this reading is further strengthened if we consider another stock figure of mimetic theory: pseudonarcissism.

    In the hell of imitative rivalry imagined by Girard, the secret of success is the ability to dissemble. Making a public display of one's desires is a recipe for failure. Thus Girard speaks of "askesis for the sake of desire." [DDN, 155] this form of self-discipline, like that of the mystic, of which it is the perfect and reversed image, is the key to wisdom. Indifference, withdrawal into oneself, is what moves others and ensures their submission. The subject is forever searching in others for the self-sufficiency that he is acutely conscious of lacking himself, and thinks it is to be found in a person whose aloofness appears to mean that he has no need of anyone, someone who relies on his own devices, who does not suffer from the sense of inadequacy that causes the subject such pain. Is this show of indifference authentic, or is it faked? In mimetic competition, the best faker is someone who does not know that he is faking, or, better still, someone who is not faking. A naturally cold and insensitive temperament can be an asset, no less than stupidity. Girard quotes a remark by French philosopher Alain that would be condemned by the politically correct of our own day: "the amorous person desires the soul, that is why the coquette's stupidity looks like cunning." [DDN, 284] There is no greater source of support, though, than the admiring regard of others. "Nothing is more apt to divert us from others and turn us back on ourselves, reassuring us about ourselves," Girard says, "than the spectacle of others taking us for their object of desire, thus conferring upon us the blessed self-sufficiency of which they deprive themselves." [TH, 371] Conversely, the feeling of mastery that results from it attracts new and ever more intense desires. Success in one's own eyes and success in the eyes of others are not two independent poles, or things that vary inversely with respect to each other; like everything that is subject to mimetic desire, they participate in a relation of circular causality, influencing each other in a positive feedback loop. Such a dynamic is eminently unstable: one spark is enough to cause it to converge in one direction; another spark will cause it to converge in the other. Whether one is welcomed or turned away in even the most commonplace encounters of daily life thereby assumes considerable importance. Anyone who wishes to succeed has only a single rule to follow, namely, to persuade others that success has already been achieved.

    These considerations lead Girard forcefully and mercilessly to deconstruct the Freudian theory of narcissism. Freud really believed in the existence of narcissistic personalities, people who are in love with themselves, entirely absorbed by themselves— like those fashionable women promenading on the Graben in Vienna who could not be bothered even to cast a glance in his direction! Taking up again the example of the coquette mentioned by Alain, Girard writes:

The coquette knows a lot more about desire than Freud does. She knows very well that desire attracts desire. So, in order to be desired, one must convince others that one desires oneself. That is how Freud defines narcissistic desire, as a desire of the self for the self. If the narcissistic woman excites desire, this is because, when she pretends to desire herself and suggests to Freud a kind of circular desire that never gets outside itself, she offers an irresistible temptation to the mimetic desire of others. Freud misinterprets as an objective description the trap into which he has fallen. What he calls the self-sufficiency of the coquette, her blessed psychological state and her impregnable libidinal position, is in [fact] the metaphysical transformation of the condition of the model and the rival...
    The coquette seeks to be desired because she needs masculine desires, directed at her, to feed her coquetry and enable her to play her role as a coquette. She has no more self-sufficiency than the man who desires her, but the success of her strategy allows her to keep up the appearance of it, since it offers her a form of desire she can copy. If the desire directed toward her is precious to her, this is because it nourishes her self-sufficiency, which would fall to pieces if she were wholly deprived of admiration. To sum up: in just the same way as the admirer caught in the trap of coquetry imitates the desire that he really believes to be narcissistic, so the flame of coquetry can only burn on the combustible material provided by the desires of others. [TH, 370-71]

Pseudonarcissism is in fact a particular case of double mediation, in which model and object occupy the same apex. Two concurrent desires imitate and excite each other: the lover imitates the desire of the coquette by desiring her; the coquette imitates the desire of her lover in desiring herself. This configuration has nothing to do with a triangle. It illustrates instead the very abstract form I have begun to sketch, of a subject who is faced with what he takes to be a self-sufficient totality, closed upon itself, that denies him access to its own self-sufficiency.

    I shall now consider a final mimetic pattern, one of the most extreme forms of mental pathology and of the suffering that accompanies it. For a long time I believed that jealousy could be analyzed in terms of this configuration. I now believe that the truth is both simpler and more complex. It has to do with pseudomasochism, a notion Girard uses to deconstruct Freud's concept of a death instinct, invented in the first place to account for self-defeating behaviors in which a person overcome by a compulsion to repeat (Wiederholungszwang) seems deliberately to put himself in situations that he knows will lead to his defeat or humiliation. Girard's use of the term "pseudomasochism" makes it clear that he does not for a moment believe that the subject (much less an "id" that might be supposed to act in his name) in any way desires failure or suffering. The subject has fallen victim instead to an especially acute form of pride, an exaggerated amour-propre that perversely takes on the features of self-hatred.

    If the subject despises himself for what he perceives to be his utter inadequacy, it is because he believes that self-sufficiency, which is to say the way out from the hell of mimetic desire, can be achieved— at least by people other than himself. It is the desperate hope to attain this state himself that irresistibly attracts him to such people, as though he sought to absorb the essence of their being, and causes him to stumble upon the obstacle of the model-rival. The double bind that intensifies the subject's fascination for his model also confirms his own sense of inferiority. All the more, then, is he encouraged to look outside himself for a way to remedy the dreadful inadequacy from which he suffers. The circle now closes on itself, as the desire to relate everything to oneself and the flight away from oneself to another person feed on each other.

    In this game, every victory is fated to be tranformed into failure. Only the metaphysical distance that separates the subject from the model-object pair— and we know that this distance varies inversely with social and spiritual distance— is capable of imparting value to each element of the pair. Victory, however, consists in abolishing the distance between model and object: the subject succeeds in possessing the object, whether it is a person or a thing, and the rival, fascinated, submits. At this juncture the gaze of others, which converge on the subject and designate him as a model or desirable object immediately lose their power to channel desire in the eyes of the subject himself. Paradoxically, this victory deprives him of the orientation that he found in the others' gaze. there is only one thing of which he is sure: his own worthlessness. He hates himself too much not to despise anyone who admires him or who simply wishes him well. It is this "underground psychology" that Dostoevsky so ruthlessly dissected in Notes from Underground (1864).

    The subject experiences desire as something that converts every model into an obstacle and causes every obstacle that is overcome to give way to failure. Constant frustration leads him to change object, then model— but in vain. With implacable yet demented logic, he infers from his repeated failures that he will be able to find happiness at last by measuring himself against an invincible opponent. Mimetic desire therefore begins by turning models into obstacles, and ends up choosing as models only those who confront the subject with insuperable obstacles.

    Does the subject desire failure? What he desires is the proof and sign of the mediator's divine self-sufficiency. He therefore desires the mediator's success— not because this success signals his own failure, but, on the contrary, because he hopes to have it for himself. Self-disgust is inextricably mixed with an inordinate pride, the subject being hopelessly torn between I and the Other. In order to escape from this double bind, he would have to be the Other. Failing that, the subject's search for transcendence ends up in the quest for the mediator's contempt, or even in the obstacle posed by the mechanical and the inanimate. Self-deification is no longer manifested, as in ordinary individualism (that lie born of pseudonarcissism), by an unquenchable thirst for the admiration of others and self-veneration; it is manifested by a longing for humiliation and self-destruction. The symmetry is perfect, the progression relentless: the search for life leads straight to death. Even if this is not everyone's fate, Girard insists that everything that is identified with a taste for risk, a hunger for the infinite, a desire to go beyond oneself, to press on ever further, to the last frontier, to the ultimate limit of desire— all these things are the expression, in various socially acceptable forms, of a single doomed quest.

    In the light of pseudomasochism, the mimetic triangle, if it ever had any analytic justification at all, can be seen at last to have evaporated. No longer can the position of either the object or the mediator (if by this latter term is meant an actual person) be identified. The reef of fascination on which the subject is about to wreck himself can be anything at all, not merely a relation between another subject and the object of his desire. Even in that case, as I have insisted, it is the relation between the two terms that constitutes the obstacle facing the subject, and not either one of these terms themselves. But the obstacle may also be a milieu or social circle, or any such closed totality that excludes the subject— a "world," to use Proust's term. In an admirable chapter of DDN called "the Worlds of Proust," Girard analyzes Mme. Verdurin's salon by comparing it to another closed world, albeit an immeasurably more likeable one, that of the town of Combray: "the Verdurin salon is not simply a meeting-place, it is a way of seeing, feeling, judging. The salon is also a 'closed culture.' Thus the salon will reject anything which threatens its spiritual unity. It possesses an 'eliminative function' similar to that of Combray." [DDN, 198]

    Swann comes to learn what it means to be a victim of this "eliminative function." But what looks like victory— finally being admitted to a circle that one revered all the more as access to it was denied— leads on to bitter disappointment: "the narrator suffers a terrible let-down when he eventually gains admittance to [Mme. de] Guermantes' [salon]! He discovers that the conversation and thought in [her] salon does not differ from that to which he is accustomed. The essence of the Faubourg [Saint-Germain] seems to vanish. The Guermantes salon loses its individuality and blends into the vague grey of already known milieux." [DDN, 217] The question therefore arises how the contours of these closed worlds are defined, how the boundaries that close them upon themselves by cutting them off from the outside world have appeared. Can the social sciences here be of any help? The answer is perfectly consistent with the principles of mimetic theory, namely, that these contours and these boundaries are a pure creation of mimetic desire:

The sociologist interested in the Faubourg Saint-Germain should not turn to [In Search of Lost Time]. This novel is not only useless, it can be dangerous. The sociologist thinks he has hold of the object of his research and suddenly he finds it slipping between his fingers. The Faubourg is neither a class, nor a group, nor a milieu; none of the categories currently used by sociologists is applicable to it. . . . [T]he Faubourg vanishes when scientific instruments are brought to bear on it. This object cannot be isolated. The Faubourg ceased to exist a hundred years ago. And yet it exists because it excites the most violent desires. Where does the Faubourg begin, and where does it end? We do not know. But the snob knows; he never hesitates. It is as if the snob possessed a sixth sense which determined the exact social standing of a salon. [DDN, 218; my emphasis]

As in the case of individual mediation— Girard speaks here of "collective internal mediation" [DDN, 202]— the object that attracts desire is a creation of desire.

    As we have already seen, the mediator-obstacle may also consist in the apparent closing upon itself of a person who is utterly lacking in imagination. This may be someone whose cultural, spiritual, or intellectual inferiority fascinates the subject (Odette for Swann, Morel for Charlus— Proust's novel abounds in examples of such "'descending' snobbism." [DDN, 209] But it may just as easily be an automaton, like the "statue that walks and talks" that so frightens Sganarelle while leaving Don Juan unshaken; or E.T.A. Hoffmann's Olympia, the mechanical doll whose uncanniness plunges the student Nathaniel into madness6; or Faustine in The Invention of Morel by Adolfo Bioy Casares, a mere image in a three-dimensional film that plays over and over again in an endless loop; or Madeleine, her double in Hitchcock's Vertigo, who is so devoid of being that she is only a fiction within a fiction. Or this being may actually be a thing— a stone, a reef. It may be the totality of all that is. It may be death.

    As I look back over what I have written so far, I am struck by how many times locutions of the type "as we have already seen" or "as I have already insisted" occur— not a sign of the best literary style, I admit. But such phrases are evidence that every figure of mimetic desire can be characterized in the following manner: a subject faces a closed world whose definition and closure are the effect and cause of his desire. Both the world and the subject are constituted and sustained by the subject's painful sense of being excluded from this world. One might say that the subject is the supplement of a closed and apparently self-sufficient totality.

    Is there a name for this structure? It is tempting to call it pseudomasochism, since it is in analyzing this figure that we have come this far, but evidently it would be an unsatisfactory choice. Girard, as we know, has given it the name of metaphysical desire. But I believe there is an even better term: jealousy. For all the figures of mimetic or metaphysical desire are kinds of jealousy. The reader will recall that I began by wondering what place should be assigned to jealousy within mimetic theory. It has always seemed obvious to me that envy, which is a relation between a subject and a mediator, should be understood within mimetic theory in terms of the double bind in which they find themselves. A number of my articles and books amount to variations on this very theme.7 But jealousy? The answer that I am now inclined to give takes the form of a paradox in the manner of Agatha Christie or Edgar Allan Poe: it is impossible to assign jealousy a place within mimetic theory, because it is found everywhere, hidden behind each of the figures of mimetic desire, from external mediation to pseudonarcissism, and from pseudonarcissism to pseudomasochism. What prevented me from seeing this until now was the obstacle thrown up by the misleading image of triangular desire.

    The reformulation of mimetic theory that I am proposing seems to me to have one enormous advantage, which I cannot elaborate on here, except to say that the many forms assumed by mimetic desire exhibit the same logical structure as the sacrificial expulsion that Girard, in Violence and the Sacred, identifies as the source of the universal presence of religion in human societies. The surrogate victim occupies the place of the subject— in the etymological sense of the subject of the king, the one who lies beneath (sub-jacere)— and the world is the undifferentiated totality of the crowd that drives him out. At last the two great pillars of Girard's work, the theory of mimetic desire and the anthropology of violence and the sacred, can be seen to be linked by a formal analogy that until now has been unsuspected.

    Why should we choose to call this universal form by the name of jealousy? Nicolas Grimaldi's book gave me the reason. Before going further, and by way of introduction to this next step in the argument, I would like to return to a point that earlier I mentioned only in passing. At first I was tempted to locate jealousy solely in pseudomasochism, one of the most advanced stages on the road that leads from mimetic desire to death. It was, as I say, my analysis of Molière's Don Juan that suggested this idea.8

    In Act I, scene 2, Don Juan confides to Sganarelle:

Come now, let's give no thought to the bad things that might happen to us; let's only think of what might bring us pleasure. The young beauty I mentioned, an utterly delectable creature, is engaged to be married, and she's been brought here to this town by her fiancé. Quite by chance I caught sight of these lovers elsewhere, three or four days before their coming here. Never had I seen two people so enchanted by each other, so radiantly in love. Their open tenderness and mutual delight moved me deeply; it pierced me to the heart, and aroused in me a love that was rooted in jealousy. Yes, from the moment I saw them I found their shared happiness intolerable; [spite]9 sharpened my desires, and with keenest pleasure I began to consider how I would mar their felicity, and disrupt a union which it pained my heart to behold. But thus far all my efforts have been in vain, and so I must resort to extreme measures.10

Don Juan is moved, then offended, by the sight of mutual love. It hurts him. He suffers, his stomach tightens. Here Molière lets us see the thing that holds sway over him, a man who holds such sway over others. What makes him suffer so? Being excluded from an autonomous totality. For nothing better conveys the idea of autonomy and of self-sufficiency than a love that is shared.

    Don Juan is attracted by women who have committed themselves to a certain type of relationship: women who are engaged to marry (Charlotte and the young woman just referred to), women who have taken a vow to serve God (nuns such as Elvira), and women who have already married. It will be granted, then, that what Don Juan chases after— the "quasi-object" that exerts a peculiar fascination over him— is a closed and autonomous totality whose invincible appeal arises from just this fact, that it excludes him. Does this amount to masochism, or a death instinct, as a whole tradition of commentary maintains? To be sure, Don Juan will find rest only once he has reached the end of his quest, which will ultimately land him in hell. But obviously it is not exclusion that Don Juan seeks in repeatedly confronting closed and autonomous totalities that exclude him. It is because they exclude him that they, and they alone, appear to him to be worthy of his desire. His misfortune, the cause of his suffering, is that he never manages to possess any woman, in the sense of forming with her the kind of union that he covets in others. He therefore strives to possess all of them, confusing the bad infinity of an unending listing of conquests with the fullness of being. This tragic error, which Pascal calls "concupiscence," leads inevitably to death— the most undifferentiated and ungraspable totality of all, a realm that one may be sure of never mastering. It seems to me of great significance that Don Juan should choose as his guide in this ultimate quest an animated statue, one that "walks and talks," which is to say an automaton— the most pathetic form of autonomy imaginable, the one that is best suited to a rationalist age. (One thinks of those tales of our own time in which the hero, tragic and pitiable, perishes from a consuming passion for an inflatable doll.)

    Don Juan himself calls the illness from which he suffers jealousy— a much more astute and accurate diagnosis, it seems to me, than the improbable suggestion of a death instinct. that is why I thought to liken the role of jealousy in mimetic theory to that of pseudomasochism, the concept that Girard uses to demystify Freud's analysis. Evidently this is much too restrictive a classification, however, since every metaphysical desire is a jealous desire, and vice versa.

2. A Brief Transit through Proustian Hell

Nicolas Grimaldi prefaces his commentary with this observation:

Jealousy almost always resembles the shadow of love. It keeps pace with [love], accompanies it, follows it. Thus we find it natural to believe that there can be no jealousy without love. Proust alone has reversed this relationship, and made love the double of jealousy. In his writings love does not precede jealousy, it follows it. Love is therefore characterized less by the pleasure that the presence of a person brings us than by the pain to which her absence gives rise. It is this pain that then turns her presence into a need, the only analgesic that can soothe the anxiety of not knowing what she is doing. If the violence of love can be felt only in feeling oneself crushed by that of jealousy, it goes without saying that one will therefore have ceased to love in ceasing to be jealous. The jealousy that arouses love but that no love has brought into existence is a psychopathology of the imagination. [EJ, 8; my emphasis]

    Grimaldi is mistaken on one point. Proust was neither the first nor the only one to reverse the relationship between love and jealousy. Molière, as we have just seen, preceded him by two and a half centuries. Let us listen once more to Don Juan: "Their open tenderness and mutual delight moved me deeply; it pierced me to the heart, and aroused in me a love that was rooted in jealousy." In reflecting upon the logic of his desire, Don Juan prefigures Proust's analysis— with nonetheless one manifest difference, not in the structure of the emotion, but in the behavior that it engenders. In both cases, the Girardian theorist is liable to fall into a trap that my method of interpretation sidesteps. For however much the form of jealousy may seem sometimes to coincide with that of the mimetic triangle, this is an illusion.

    Grimaldi wonders, for example, how Swann came to fall in love with a woman who not only was not his type, but whom he strongly disliked; in short, a woman whom at the outset he did not love, in the ordinary sense of the word. Here is the response given by the narrator of the Recherche: "It was true that Swann had often reflected that Odette was in no way a remarkable woman...; but since he had observed that to many other men besides himself Odette seemed a fascinating and desirable woman, the attraction which her body held for them had aroused in him a painful longing to secure the absolute mastery of even the tiniest particles of her heart. And he had begun to attach an incalculable value to those moments spent in her house in the evenings, when he held her upon his knee..."11 The mimetic triangle seems clearly established— the mediators of the desire Swann feels for Odette are these men who orbit around her. And yet, a few lines later, the narrator adds this crucial detail, which ruins the interpretation in terms of triangular desire: "[T]he supreme pleasure [which it was in her power to bestow upon him was] to guarantee him immunity... from the assaults of jealousy." [SW, 385; my emphasis]

    This passage becomes clearer if we compare it to what Grimaldi has to say about jealous desire, which in Proust is identical with love: "that we should ceaselessly need to be near someone, not on account of the charm she radiates or for the pleasure that she brings us, but solely to avoid the anxiety that comes of not knowing what she is doing, is the whole of what explains love in Proust. It is not so much that love binds us to a person than that jealousy makes her indispensable to us. Our desire for the pleasure she gives us is much less than the suffering we feel from imagining [the pleasure] that others might give to her." [EJ, 23] Swann's desire for Odette is not an imitated desire that feeds on the desires other men have for her and that is exacerbated by the rivalry between him and them. It is a desire for peace, calm, and tranquility. Swann wants Odette for himself simply so that no one else will possess her.

    Proust writes: "And yet this Odette from whom all this evil sprang was no less dear to him, was, on the contrary, more precious, as if, in proportion as his sufferings increased, the price of the sedative, of the antidote which this woman alone possessed, increased at the same time." [SW, 517; quoted in EJ, 103] Grimaldi comments:

We have just seen Odette paradoxically become more dear and more precious to Swann, at the very moment when the crushing avalanche of her revelations seemed bound to turn him against her for good. Yet the opposite occurs. For in learning of her wild behavior, Swann also discovers that no one is more unknown, more secret, more unpredictable. And so it is at the moment of his greatest pain that he experiences something like an amorous passion, a still more violent need to attach himself to her, to moor his existence still more securely to hers.
    In just the same way, we have seen the narrator's love for Albertine ebb and flow with the tide of his suffering. In confiding this [to the reader], the most remarkable thing is the explanation he gives of it. "I no longer loved Albertine, for I no longer felt anything of the pain I had felt..." [TC, 17] But he had only to suffer again in order to begin to love again. [EJ, 104-5]

    What is the mimetic theorist to make of this? Should he try to explain it in terms of pseudonarcissism? Odette does not display indifference toward her lover as a self-sufficient person would do, aiming to attract mimetically the desire of someone who is ready to sell his soul in order to conquer this self-sufficiency. Quite to the contrary, it is when she shows herself to be, like Swann himself, a creature of desire (for others), and therefore an incomplete and needy person, that Swann's desire (for her) is rekindled. Pseudomasochism then? Swann is not searching for a mediator so invincible that confrontation can only leave him hurt and bruised. Again, the truth of the matter is quite opposite. If he burns with desire to (re)conquer Odette, it is simply in order to put an end to his suffering. Here is what Grimaldi has to say, referring at the end to the narrator's feeling for Albertine:

For a person whom we have never loved suddenly to become necessary to us, anguish is enough. In making us experience her absence as a [kind of] misery, she makes her presence a need. Love is another name for this need. Thus no sort of joy is expected from it in Proust, only a sort of anesthesia. Far from being identified with the happiness of a presence, [love] is limited to making us desire the end of an absence as [the end] of a torment.
    According to Proust, then, and contrary to what is ordinarily believed, it is not because we are in love that we are jealous. It is jealousy that makes us be in love, just by making us desire someone's presence as an end to anguish. Not because her presence enchants us, or delights us, or overwhelms us. Yet she is the only one who gives us this opiated feeling of calm— a point that the narrator constantly repeats: "the calm that my friend brought me was an appeasement rather than joy."12 [EJ, 46-47; my emphasis]

Attentive readers of Girard's work will not have failed to notice that sexual pleasure plays no notable role in it. No sooner is it mentioned as a natural "need," on the same order as the necessity to feed and clothe oneself, than it finds itself absorbed by mimetic desire, losing its identity in it. Nothing of the sort occurs in Proust, who to the contrary sees sensual pleasure as the means, at least temporarily, to be sufficient unto oneself, to coincide at long last with oneself. And it is here that a new source of anguish arises for the jealous person. As Grimaldi observes:

One of the most constant themes of jealousy in Proust is that even when we hold someone in our arms she still eludes us. Her body provides us only with the illusion of her presence. For she does not inhabit her body. She has fled toward what she desires or what she expects. And so no one possesses another so long as sensual pleasure has not reunited her with herself by fulfilling her expectations and deadening her desire. Only the lover who has accompanied his mistress in sensual pleasure to the point of making her coincide utterly with herself can be said to possess her entirely. [EJ, 29]

    For mimetic theory, personal completeness and the accompanying deadening of desire can exist only as part of a strategy of deception. In the case of the coquette, it is because her desire imitates the desire of another, who desires her, that she is able to give the illusion of self-sufficiency; and it is because she creates this illusion that she is able to attract the desire of another— indeed, of many others. Girardian analysis, in regarding the coquette as the point at which all desires (including her own) converge, fails to distinguish between the body and desire, between nature and the mind. It is therefore incapable of explaining the kind of sexual jealousy that well knows physical possession is not enough, for the mind can still elude one's grasp. Like Honoré in Proust's early collection Les Plaisirs et les Jours (1896), the lover can exclaim: "I can accept that someone gives her happiness, or gives her love, but I don't want anyone to give her pleasure. I don't want her senses to be aroused, for her to be given more pleasure than I have given her... I am jealous only of pleasure, it is my body that is jealous."

    It should be clear, then, that Proustian jealousy obeys another dynamic than the one associated with the forms of mimetic desire analyzed by Girard. Nowhere do the categories of mediator and imitation figure in it. And yet it is legitimate to subsume both Proustian jealousy and the forms of mimetic desire under the general category of "jealousy", as I have defined it. The basic structure of this category is invariant: a subject finds himself shut out from a totality closed on itself; what varies are the subject's reactions.

    In Proustian jealousy, the subject wishes to reserve to himself the exclusive right to possess an object of desire in order to forestall the suffering that would follow if the object were to be exclusively, or even partially,13 possessed by someone else. It is often in this sense that the word "jealousy" is employed. But there are many other usages. Recall the sense in which it is used by Molière's Don Juan— and by his Italian counterpart, Da Ponte's Don Giovanni. There, too, triangles appear. Don Giovanni openly flirts with Zerlina, for example, in the presence her fiancé, Masetto. But no one would suppose for a moment that Masetto is the mediator of the desire that Don Giovanni feels for Zerlina!

Don Giovanni: And we're alone, Zerlina, for the first time alone without that idiot...
was it not very neat, how I dispatched him?
Zerlina: But, sir, we're getting married!
Don Giovanni: You! To him! What can he know of beauty? And how can I,
who live by fine perceptions, seeing your grace and sensing your refinement,
how can I see you wasted in servitude to that unworthy bumpkin?
Zerlina: But, my lord, I can hardly go back on my promise!
Don Giovanni: It would be hardly right to keep it.14

Don Juan's jealousy, on the other hand, does not drive him to try to make a particular woman his prisoner, as the narrator of the Recherche does with Albertine, in the hope that one day they will be able to doze off together in the comforts of domestic bliss. He says it magnificently in scene 2 of the first act of Molière's play:

But once one is the master, there's nothing more to say or wish for: the joy of passionate pursuit is over, and all that remains is the boredom of a placid affection— until some new beauty appears and revives one's desires, enchanting the heart with the prospect of a new conquest. To be brief, then, there's nothing sweeter than overcoming the resistance of an attractive woman, and I bring to that enterprise the ambition of a conquering general, who moves on forever from victory to victory, and will set no limit to his longings. Nothing can withstand the impetuousness of my desires: I feel my heart capable of loving all the earth; and, like Alexander, I wish there were still more worlds in which to wage my amorous campaigns."15

    In Don Juan we see the irresistible need to break through the closed circle that excludes us, to smash open the self-contained world that ignores us and that it appears futile to wish to possess. With Meursault, the eponymous stranger of Camus's L'étranger, we climb yet another rung on the ladder of evil. Now the feeling of being excluded gives such pleasure that one readily does what is needed (in this case, killing a man for no particular reason, unless on account of "the sun") to be excluded— the opposite of Max Weber's Calvinists, who assured their own divine election. In moving from self-election to self-exclusion, and so in drawing nearer to nothingness, the paradox of pride is deepened.16

    Girard himself, as we saw at the outset, uses the language of jealousy in relation to the subject's perception of the mediator's exclusive attachment to the object: "the object [that the model] jealously keeps for himself." this is obviously an expression of the subject's own jealousy, since he too would like to keep the object for himself. Indeed, there are as many uses of the word "jealousy" as there are attitudes and reactions to a range of situations whose structure remains fundamentally the same.

    Grimaldi devotes a whole chapter of his book to a particularly devious form of jealousy, what Proust himself calls "retrospective" or "delayed-action" jealousy [TC, 106; quoted in EJ, 119]. More completely than the forms we have considered so far, it illustrates the pointlessness of trying to discern a triangular structure in jealous desire. A story published almost sixty years ago now by Jean-Baptiste Lamothe explores the suffering caused by this particular sickness.17 The protagonist is a Frenchman who is struggling to get over the breakup of a love affair. He takes refuge in a boarding house in the English countryside, where he comes under the spell of his neighbor, a melancholy young woman whose beauty is somewhat dulled by a constant air of sadness. He learns that the previous year she had had a stormy relationship with a man who occupied his room, and who has since gone back to his native Northumberland. The Frenchman falls madly in love with his neighbor. Nevertheless he spends most of his time thinking about her former lover. Obsessively he gathers information about him and about Northumberland, of which until recently he knew nothing. He wants to know everything there is to know about its landscape, the local weather, the people who live there, and so on. This fixation ends up driving him mad.

    Here again it would be pointlessly reductive to see the former lover as the mediator of the Frenchman's desire for his neighbor, whether through external mediation or anything else. The protagonist of Lamothe's story neither imitates the desire of his predecessor nor enters into rivalry with him. His love has its origin instead in jealousy. He cannot accept having been excluded by the force of circumstances (in this case, the irreversibility of time) from the very powerful and intimate bond that united the Englishman with the young woman next door. If his every waking moment is dominated by thoughts of the former lover, if he all but deifies his predecessor, it is not because the former lover is an obstacle to his present passion; quite obviously there is no such obstacle. It is because the former lover was, and therefore remains, a part of a closed totality that excludes the protagonist. Proust writes:

Jealousy is thus endless, for even if the beloved, by dying for instance, can no longer provoke it by her actions, it may happen that memories subsequent to any event suddenly materialize and behave in our minds as though they too were events, memories which hitherto we had never explored, which had seemed to us unimportant, and to which our own reflexion upon them is sufficient, without any external factors, to give a new and terrible meaning. There is no need for there to be two of you, it is enough to be alone in your room, thinking, for fresh betrayals by your mistress to come to light, even though she is dead.
[TC, 107; quoted in EJ, 75]

Grimaldi comments:

Everything jealousy imagines it experiences in the present. It has freed itself from any connection to [past] time. If it were aroused by actual events, [if it were] founded on reality, it would be subject to the same temporal causes as reality. But— jealousy develops solely through the spontaneous activity of the imagination, independently of all reason as of all reality. In imagining what it suspects, it experiences it. In experiencing it, it makes it present. And so it is that in retrospectively suspecting betrayals long ago in the past, jealousy experiences them with as much pain and intensity as if it [had just] discovered them, since [as Proust observes (TF, 662)], "for jealousy there can be neither past nor future, and what it imagines is invariably the present." [EJ, 120-21]

3. Proust in the Nursery

I will have made myself very poorly understood if anyone believes that a minor change in terminology will be enough to solve the problem I have pointed out. Why should we not simply call "jealousy" what Girard describes as "metaphysical desire," and be done with it? Because ultimately what is at issue here is nothing less than the role that imitation plays in desire, and therefore the very concept of mimetic desire.

    We must be clear about what the notion of imitation involves. Two people who do the same thing at the same time are not necessarily imitating each other. People who go out for a stroll on Sunday afternoons in Paris, for example, walking up the Champs Élysées from the Grand Palais to the Arc de Triomphe, all behave in much the same way. They pause before the same shop windows, stop at the same sidewalk cafés, withdraw money from the same cash machines, and so on. None of this allows us to say that they are imitating one another. When it comes to behavior, one can speak of imitation only if it can be shown that there exists a link among a certain set of actions, either a blindly causal relation (as in the case of a baby who sticks out her tongue in response to the same gesture made by her mother) or one that depends on rational calculation, even if only at the unconscious level (thus, for example, individuals caught up in a collective panic are apt to suppose that others may have information that they do not, and that in following them they will indirectly benefit from it).

    With regard to the fundamental triangle of mimetic theory, two hypotheses present themselves as explanations of the behavior of a subject who desires the very object that another desires or possesses: the mimetic hypothesis, which asserts that the subject imitates the desire of this other person, who is then a model; and the hypothesis of jealousy that I have advanced, according to which the subject wishes to have exclusive possession of the object in order to spare himself the suffering that possession of the object by another would cause. It would certainly be odd to say in this case that the subject imitates the desire of a model or mediator, for this desire is, or can be, counterfactual or virtual in nature, no less than the identity of the mediator. Thus, for example, if I do not claim full possession of an object as my own, then someone else might appropriate it for himself, at least to one degree or another. I have already indicated how long it took me to realize that when Girard formulated the mimetic hypothesis in Things Hidden Since the Foundation of the World, he did so using the vocabulary of jealousy. But this, it seems to me, involves him in an inconsistency.

    Another story collected in the volume I mentioned a moment ago by Jean-Baptiste Lamothe— like André Maurois before him, a sensitive and droll observer of cultural misunderstandings between the French and the English— tells the tale of a thirteen-year-old boy from Paris named Philippe who has been sent by his parents to spend the month of July with a family on the Isle of Wight.18 The father, Mr. Rolfe, teaches French in a secondary school outside London. The mother, about thirty years old, is a housewife. Their son, John, is the same age as Philippe and the excuse for his stay with the family, which it is hoped will improve his still very rudimentary English. In the event, John's company bores Philippe. An only child long overprotected by his mother, who had only reluctantly agreed to let him go away for the sake of his studies, Philippe ties himself to Mrs. Rolfe's apron strings, refusing to leave her side. No one is there to stop him, for Mr. Rolfe spends the entire week in London, and it is only on the weekend— after a tedious trip that takes him first by train to Southampton, then across the strait from Southampton to Newport by boat, and finally by bus from Newport to the family's home in Shanklin— that he is able to rejoin his family. The year before, during his first stay with an English family, at Ramsgate, a seaside resort on the southeast coast of England, Philippe had fallen madly in love with the young daughter, a very pretty girl with red hair named Sonia who constantly teased him. Philippe did not understand why, for everything was new to him, the language, the food, the customs, the style of dress. The title of the story, "The Boy in the Plus Fours," refers to the ridiculous short trousers that Philippe's mother had forced him to wear, making him look like Tintin out for a round of golf— to Sonia's great mirth, since everyone in England then wore blue jeans. Naturally Philippe is more at ease with Mrs. Rolfe, who inspires a tender and profound feeling in him. It is a feeling he has never felt before.

    Disaster strikes the weekend before Philippe's scheduled return to France. On arriving home from London, Mr. Rolfe on once takes his wife into the bedroom. Philippe follows, clinging to her dress. The master of the house abruptly slams the door in the young boy's face. Wild with rage and despair, Philippe flies into a frenzy and sets about smashing everything in the living room he can lay his hands on. The story ends abruptly with a family conference in Paris. Philippe's father has just received a letter from Mr. Rolfe and reads it aloud to his son, his wife, and to his wife's father, Philippe's beloved grandfather. Written in proper French, the letter says more or less this: "Dear Monsieur Puech, I regret to inform you that, contrary to our prior understanding, my wife and I shall not be able to welcome your son next summer. Although he is possessed of great intelligence and great sensitivity, or perhaps for this very reason, Philippe has behaved inappropriately during his stay with us. We do not wish to undergo this painful experience again. Yours sincerely etc." The young boy's humiliation is complete.

    By the end of the story the reader is left to wonder not only whether Philippe will ever learn to speak English, but also what sort of emotional and sexual life awaits him. Will the trauma of a distant summer in England have permanently stunted his growth, so that like the unhappy hero of L. P. Hartley's novel The Go-Between (1953), superbly adapted for the cinema by Joseph Losey almost two decades later, he remains incapable of confiding in another person, incapable of love? Or will he always carry deep inside him, and cherish in his heart, the sweet memory of a boyish infatuation that, if he had read Hartley's tale, would often remind him of its opening sentence— "The past is a foreign country: they do things differently there"?

    When Lamothe's story was first published, in the early 1950s, Freudianism in its Lacanian-structuralist interpretation was all the rage in France. One critic suggested that the author had redefined what Freud called the primal scene: no longer was it the trauma of a child seeing his parents make love (something that is seldom likely to occur under modern living arrangements); now it was the door that closes off the parents' bedroom. Behind this door lies a hermetically sealed world that for young Philippe is no less unknown, frightening, and fascinating than the singular customs of perfidious Albion. Even granting that a slender literary fiction may not by itself constitute the basis for an entire science of mankind, we are nonetheless justified in wondering if the experience of jealousy does not in fact precede that of the primordial triangle, no matter whether this relationship is interpreted in the manner of Freud, Lacan, or Girard. What happens behind the door irrevocably joins together father and mother, man and woman— since the subject, the child, cannot distinguish their respective roles. If, as seems obvious, the subject is faced with a closed world from which he is excluded, and not with two distinct beings, one of whom is the model and the other the object, then we are dealing here with jealousy and not with a mimetic triangle.

    I referred at the outset to an experiment in developmental psychology to which Mark Anspach, one of the most subtle and inventive exponents of mimetic theory, has called attention. Adding to its interest is the great difficulty Anspach has had in accounting for what happens, as he puts it, when a doll comes between mother and child. This is the title of the relevant section of his paper on the subject. The subtitle, significantly, is "Anatomy of a Mimetic Triangle."19

    The experiment examines how twelve-month-old babies react in the presence of a third party who monopolizes the attention of their mother.20 In the middle of a room a mother plays with her child. Then an unfamiliar person intrudes upon the scene— a woman holding in her arms either a plastic baby doll roughly the same size of the baby or an illustrated cookbook. The mother gets up and joins the newcomer, sitting next to her in a corner of the room and outwardly pretending no longer to be interested in her child. The two women talk about the object, doll or book, and take turns putting it on their knees. A hidden camera records all signs of interest, distress, or protest on the part of the neglected infant.

    The remarkable thing is that even though the results of the experiment are entirely predictable, they come as a surprise to its authors. The babies are much more upset when the object is a doll, and even more so when the mother holds the doll rather than the visitor. When the object is a book, no matter which woman holds it, the attention that the baby gives the object is the same. The authors conclude that the doll is seen by the baby as a rival, whereas the book is not. Yet they admit to being at a loss to explain why the rivalry between baby and doll should be stronger when the mother, rather than the visitor, holds it on her knees.

    This is quite amazing. The authors' blindness to what is going on is all the more incomprehensible as they use the word "jealousy" from time to time in commenting on the experiment, although they seem to regard it not as a concept, but simply as a manner of speaking. For this very reason they seem not to understand the import of their own quotation from Darwin, who remarked about his son at the age of fifteen months: "[J]ealousy was plainly exhibited when I fondled a large doll."21 But the confusion only gets worse. The authors find it intriguing that the babies should scream and cry when their mother coos and fusses over the doll. Normally, they remark, using the jargon of experimental psychology, a "positive vocal affect" on the mother's part arouses the same sort of affect in her child. The discovery of an anomaly prevents them from openly characterizing the baby's behavior in terms of imitation alone, but plainly the idea has not gone away, and Anspach has no hesitation in pronouncing the forbidden word for them. The baby should imitate her mother— but here she does the opposite. What a surprise! the real question is how anyone could possibly be so obtuse. As if babies are immune to destructive passions. As if babies are incapable of suffering horribly when their mother neglects them in favor of another, even if it is only a facsimile of a child. As if babies could not be jealous!

    And yet Anspach attempts to dispel the authors' perplexity by appealing to the concepts of mimetic theory, and in particular to the notion of triangular desire. But exactly which triangle does he have in mind? Who plays which role in this infantile comedy? One might be tempted to cast the mother in the role of the mediator and the doll in that of the object. The mother is certainly a more attractive model than the unknown visitor, which would explain why the baby is more interested in the doll when the mother herself shows an interest in it. But if this were a satisfactory explanation, as Anspach rightly observes, it should also apply to the illustrated cookbook. Yet this is not the case: whether the book is held by the visitor or by the mother, the effect on the baby is the same.

    Anspach therefore searches for another explanation. He finds it by contorting the theory of triangular desire in such a way that the mother is now seen as the object of the rivalry and the doll as the mediator! After all, in mimetic theory a double bind transforms the rival into a model and the model into a rival. True enough— and yet it could not be more plain in this case that the doll is the rival of the baby and the mother the object of the rivalry. The baby would like to be in the position of the doll; indeed, the baby would like to be the doll. Anspach seems to sense that he has pushed the theory of triangular desire as far as it will go or even beyond its limits. Couldn't he be right, though? Since the mediator in the sort of world Proust imagines may be anyone at all— even a person who is in some way inferior to the subject— why should the mediator not be a thing? Is not Don Juan guided along the road to hell by a statue? This gambit immediately runs up against another obstacle, however, and a very considerable one at that, for in the experiment Anspach considers he is forced to assume the existence of a mediator who, constitutionally one might say, has no desire; a mediator whose desire cannot be imitated by the subject because there is no desire in this case. To say that the doll is the mediator of the baby's desire is no less baroque, I fear, than to regard the unfortunate Masetto as the mediator of the desire that Don Giovanni feels for Zerlina.

    But Anspach is not to be deterred. Why not simplify and at the same time generalize the theory of mimetic desire, he asks, by accepting that there can in fact be mediators without desire— as in the case of advertising?22 Think of the gigantic billboards that show George Clooney delighting in the aroma of a cup of instant coffee. Does anyone really believe that Clooney is doing anything other than his job, which is to mimic, in exchange for fabulous sums of money, desires that he does not himself have? No— and yet people are eager to buy the product that he endorses. From this Anspach concludes that imitation is not in this case the imitation of a desire; instead it is the reflexive imitation of an acquisitive behavior. Later we shall see that the case of advertising turns out to be rather more ambiguous than one may be inclined to suppose at first sight.

    Anspach's efforts to save the theory of triangular desire call to mind the futile attempts to rescue the geocentric conception of the solar system by resort to the abstruse complications of the theory of epicycles. Eventually it became clear that there was no choice but to embrace a new paradigm that replaced the earth at the center of the system by the sun. In the present case the sun is a black sun— jealousy. Perhaps the most remarkable aspect of the matter is that Anspach himself shows beyond any doubt that this is so, without, however, recognizing the obviousness of it. Against whom is the baby's rage and fury directed when it sees the mother show her affection for the doll? Anspach acknowledges that the experimental result is surprising. And yet it is perfectly consistent with what we know about jealousy. The jealous subject does not find himself confronted with two apexes— model and object— of a triangle of which he is the third apex. He sees a relation that is closed on itself, and therefore unanalyzable. And indeed the experiment confirms that the target of the baby's anger is indeterminate: the mother and/or the doll. The reason for this, of course, is that it is a relationship from which it is excluded that causes the baby's distress, and not specifically the behavior of any of its members. "Jealousy is an equation involving three permutable (undecidable) terms," Roland Barthes wrote in Fragments d'un discours amoureux (1977). "One is always jealous of two persons at once: I am jealous of the one I love and of the one who loves the one I love."23

    It is clear, then, that patterns of desire that are apparently triangular can be subsumed under the general category of jealousy. On the other hand there are certain jealous dynamics that it would be altogether inappropriate to treat as instances of triangular desire. Infantile jealousy is one such case. And yet doing away with the triangle seems also to do away with imitation, and therefore with mimetic desire. Is jealousy therefore more primitive, more fundamental than mimetic desire? the question is staggering in its implications. Let me conclude with a few remarks that may perhaps form the basis of a program for further research.

4. Implications for Mimetic theory

In one sense, there is nothing I have said until now that does not already figure in the theory that Girard has constructed. If the perspective afforded by jealousy changes anything, it is our perception of the topology of the manifestations of desire.

    I now come back to the example that is at the very heart of my proposal: the disappearance of the apex occupied by the mediator (or model) in what is supposed to be a mimetic triangle. Earlier I recalled Girard's own analysis regarding the disappearance of the apex occupied by the object. This analysis is crucial because it is meant to explain the transition from desire to violence: a world in which human beings are no longer separated from one another (and yet at the same time are joined together) by a common world of objects is a world in which violence can spread like a virus through a population that has no immune system to defend it against infection. Now imagine what happens when the apex occupied by the mediator also disappears. It is hard to dispute that a triangle lacking two apexes has nothing triangular about it apart from the name. To this it must be added that the simple existence of a relation between three persons— let us say, two men and a woman— does not justify us in speaking of a mimetic triangle. Take the trio of Don Giovanni, Masetto, and Zerlina. It can hardly be said to be mimetic, since Masetto is in no sense the mediator of Don Giovanni's desire.

    As it happens, the case of a triangle lacking two apexes has explicitly been considered by René Girard. He speaks of it in connection with pseudomasochism in Dostoevsky. I reproduce a long quotation here, for it reveals at once Girard's analytical genius and the utter inadequacy of the triangular metaphor:

    At the stage illustrated by Dostoevsky, the object and the model are both necessary, but they only have value in terms of their mutual relationship. In fact, it is neither the woman nor the rival that the subject desires, but [in some sense] the couple as such. This alone seems to be capable of realizing the autonomy the subject dreams of — a kind of blessed narcissism for two, from which the subject feels himself to be excluded. Likewise in Racine's Phèdre the heroine's desire becomes aggravated when she learns that her beloved has a beloved and the two young people seem to desire one another. Of course, this is also the theme of Rousseau's Nouvelle Héloise.
    So it can easily be explained why, in a number of Dostoevsky's early works, the subject actively tries to bring the woman he loves and his rival together; he hopes that the couple will [be grateful] and will find a little place for him as a third party in their paradise of intimacy.
    This theme, as it occurs in Dostoevsky, reveals the undecidable element in the situation created by mimetic rivalry. The subject has no wish to triumph completely over the rival; he has no wish for the rival to triumph completely over him. In the first event, the object would fall to him, but it would have lost all value. In the second event, the object would attain an infinite value, but it would be forever outside his reach.
    However painful it may be, the triangular relationship is less painful than a decision that would end it in one way or another. That is precisely why [the triangle] has a tendency to perpetuate itself and to reproduce itself if [by chance] it has collapsed, [henceforth by imitating a desire that has always already been a rival, before it ever was a model. It is on the basis of the model, however, and of imitation, that the whole sequence must be conceived.] Rivalry is intolerable, but the absence of rivalry is even more intolerable. It brings the subject up against nothingness. That is why the subject makes every effort to persevere or to begin again, often relying on the undisclosed complicity of partners who are aiming for similar goals.24 [TH, 360-61]

This complex passage needs to be read carefully in the light of the argument I have developed until this point, so that we can have a better idea of the distance, modest or considerable depending on one's point of view, that separates the reformulation of mimetic theory that I recommend from the original version given it by Girard. The following six points will summarize our convergences and differences.

    1. The figure that I call "jealousy" and that Girard calls "metaphysical desire" is perfectly described at the outset, and the fatal inadequacy of the fundamental triangle in principle clearly indicated: the object of the subject's desire is "in some sense the couple as such," and not the woman or the rival, for the couple "alone seems to be capable of realizing the autonomy the subject dreams of— a kind of blessed narcissism for two, from which the subject feels himself to be excluded." This relation, which Girard identifies with pseudomasochism, ultimately reduces to pseudonarcissism: the subject is attracted by a being— here a collective being, the couple— who seems to him to possess the autonomy to which he aspires. The use of phrases such as "in some sense," "seems," and "feels" betrays Girard's conviction that in a world transfigured by desire things are not as they appear: it is not really true that two young persons can desire each other, or that the subject is really excluded from the couple they form, for the exclusion that constitutes the subject and the closure that separates the two lovers from the rest of the world feed on each other, no less than the pseudonarcissism of the coquette and the convergence of male desires upon her.

    2. Yet this disappearance of the separate positions of object and mediator is held to correspond to the "stage [of mimetic desire] illustrated by Dostoevsky," which is to say to one of its most pathological manifestations— "underground" psychology, or what Girard calls pseudomasochism. At this stage, as we have seen, priority is given to the rival over the model. The subject, in other words, chooses as models persons who he knows will have the upper hand over him when, as one day he must, he enters into rivalry with them. His desire is therefore the result of "imitating a desire that has always already been a rival, before it ever was a model." I maintain to the contrary— and this, I hasten to add, on the very basis of my reading of Girard's major works no less than on my own personal experience and reflection— that all the stages of metaphysical desire, the least pathological and the most innocent alike, exhibit the same structure. For anyone who has difficulty accepting that pseudomasochism (or a death instinct) is found in babies, there remains the possibility of admitting that babies may, quite simply, be capable of jealousy.

    3. Girard's remark on the nature of the object in Dostoevsky makes it clear once again that an object of desire is not independent of the desire, that it does not preexist it, but that, on the contrary, it is created by it. In other words, the object is a quasi-object; it has value only on condition that one does not possess it.

    4. No matter that the triangle has been sacrificed, having lost two of its three apexes (object and model), it reappears at the end wholly intact, like a Phoenix reborn from its ashes: "that is precisely why the triangle has a tendency to perpetuate itself and to reproduce itself if by chance it has collapsed... It is on the basis of the model, however, and of imitation, that the whole sequence must be conceived." The "however" that I have emphasized here, like the "in some sense" at the beginning of the extract, betrays the awkward position in which Girard finds himself, torn between his realization that the triangle has disappeared (actually, was never there in the first place) and his determination to save it. Something quite considerable is at stake: "It is on the basis of the model, however, and of imitation, that the whole sequence must be conceived." Without a model, without a triangle, it is the mimetic character of desire that is in peril.

    5. And yet there are cases in which no triangle can plausibly enter the stage of desire. They occur when the roles of subject, object, and, especially, mediator cannot be embodied by three persons made of flesh and blood; and also when the actor to whom one may be tempted to assign the role of mediator exists, but either he is incapable of playing the role or he is absent just when he is needed.

    Masetto in Don Giovanni and the former lover in Lamothe's story illustrate these two latter cases, respectively. I shall not pause to consider them further here, except to note that whereas jealous desire gives the former lover, in his capacity as rival, a fantastic prestige, without thereby setting him up as a model (since the object has already lost her importance), the desire manifested by Don Giovanni for Zerlina cannot possibly strip Masetto of the complete and total insignificance that Don Giovanni has conferred upon him.26

    One case where a triangle is physically impossible, because there are no human actors available to occupy the apexes of object and model, is the direct encounter between a subject and a world, in the sense that Proust gives this word— a world that the subject wishes to conquer because until now its gates have been closed to him. Charlie Chaplin, on arriving in New York, passed the Statue of Liberty and exclaimed, "America, you will be mine!" Many people have set similar challenges for themselves and gone mad. This was not to be the fate of Chaplin, for the good and sufficient reason that he won his bet. Few have his genius or his luck.

    Yet we are dealing here once again with pseudomasochism, an advanced stage in the evolution of desire which causes such difficulties for standard mimetic theory that it is forced, at least implicitly, to jettison one of its basic concepts. It will therefore be instructive to consider less extreme cases in which the impossibility of a triangle is equally patent.

    Capitalism has so inured us to advertising that we do not see that an apparently perfect illustration of the standard theory of mimetic desire in fact constitutes a serious challenge to it. In this case the person who ostensibly occupies the position of mediator is a movie actor who is playing a part, and everybody knows it. Thus, for example, a very handsome man, known to millions, pretends on a giant billboard or in a television commercial to be so captivated by the aroma wafting up from a cup of instant coffee that he is in danger of fainting. No one is fooled by his apparent indifference to everything else around him: he plays the part very well, both as an actor and as a character. Must we conclude that the viewer who stops on his way home to buy the same brand of coffee will have succumbed to the force of mimetic desire? This would overlook the fact that the mimetic mechanism operates on the assumption that the subject believes— whether wrongly or not does not concern us here27— in the reality of the model's desire. Would jealousy provide a more satisfactory explanation? In a world in which the jaded consumer prides himself on not being taken in by the claims of advertising, this seems unlikely. But it is to exactly such an audience that advertising seeks to appeal, and it is willing to take matters a step further: an attractive young woman approaches the movie star, who seems now to be savoring the prospect of his next romantic conquest. But then it becomes clear that the young woman has not the least interest in him; she is only attracted by the coffee's aroma. With the sudden confusion and humiliation of an ideal candidate for the role of mediator, this role has brilliantly been reduced to nothing. The ad is meant to tell us— and this is how it fools us— that the product's intrinsic qualities alone are responsible for consumer loyalty, without there being any need for mediation in the form of a celebrity sponsor. The young woman is Jane Doe— just anybody, anybody at all, you or me.28 And it is just such a person, which is to say the undifferentiated crowd, that turns out to be the mediator of desire.

    But does a group of people who flock to a particular object indicate it mimetically to others who have not yet become part of the group? If we leave seven small children in a room with seven perfectly identical teddy bears and then come back a quarter hour later, there is good reason to bet that we will find all of them fighting over a single teddy bear that one of the children happened to claim for himself. But one may be forgiven for wondering whether this pet example of mimetic theory really decides anything.

    In the case of any crowd that forms around an object— whether a rocking horse, a bonfire, or the dying victim of a traffic accident— the crowd and what lies at the center of it constitute a closed world sealed off from the outside. Who among us, child or adult, can resist the temptation to penetrate its secret? Evidently this is a consequence of jealousy, not of mimeticism.

    6. Whereas every figure that comes under the head of what Girard calls metaphysical desire in the standard version of mimetic theory may be interpreted in terms of jealousy, the converse is not true. The baby, in Anspach's revised version of the theory, takes as its model a doll that her mother dotes on, and desires to be this doll; in my account, the baby is simply jealous. Don Giovanni is smitten with Zerlina because she is already desired by another, according to Girard's theory; according to my version of it, Don Giovanni destroys the couple because it excludes him by virtue of being closed on itself. Mimetic theorists say that the seventh child contends with the other children for possession of the teddy bear that they have "chosen"; I say that it joins the crowd in order to avoid being excluded from the totality that it forms around a central object. Girard's followers claim that romantic attraction is mediated by a rival who points out to me the desirability of the woman that he himself desires; the jealousy model claims that such rivalry arises from the simple fact that the woman I love prefers someone else to me. In the latter case, the rival's prestige is a consequence of desire, and not the cause of it.

    Not only does the interest in saving the mimetic triangle have the disadvantage of making it necessary to devise expedients having little or no psychological plausibility; the argument from jealousy that I advance has the great advantage of parsimoniously explaining a great variety of types of behavior with reference to a single generative fact— the suffering felt by a person who feels himself to be excluded from the "sacred center"29 of an apparently autonomous totality. This suffering is universal, yet the number of distinct reactions it can generate is very large. I take up here only the ones that I have already mentioned.

a. The subject seeks to be admitted to the sacred center, as in the case of the narrator of the Recherche, the seventh child, and so on.
b. If the center is constituted by an exclusive relation between another subject and an object (for example, a woman and a man whom she loves), the subject (in this case a woman) desires the object. Here we have a straightforward case of mediation.
c. If the center is constituted by an exclusive relation between another subject and an object (for example, a woman and a man whom she used to love), the subject (a man in this case) is fascinated by the former lover. Here we are dealing with retrospective jealousy.
d. The subject cannot resist the temptation to destroy a commitment that unites one person in love with another, as in the case of Don Juan and the couples he breaks up.
e. A particularly pathological case is the figure of the loner, famously incarnated in the character of Meursault, Camus's "stranger." Meursault finds himself at the margin of society, not because he is excluded from it but simply because there is no center in a society of anonymous and undifferentiated individuals. In order to be at the center, a center must first be created. Meursault's solution— doing something in order to be excluded— is paradoxical but nonetheless effective: paradoxical, for, normally, where there is a center, it is the exclusion from it that causes the suffering of jealousy— and yet, in Meursault's case, the position of the excluded is the center!; effective, since in committing a murder and in ending up on the scaffold, he has his moment of glory— albeit his last moment ("For everything to be consummated, for me to feel less alone, I had only to wish that there be a large crowd of spectators the day of my execution, and that they greet me with cries of hate").30
f. The case of the loner, the outsider who manages to makes others finally take an interest in him by expelling him, is superseded in its turn by that of jealousy itself, at least if one accepts Deleuze and Guattari's interpretation of the Recherche, according to which jealousy is the end of the road traveled by desire, the place that one hopes to reach without even knowing it: "When Proust seems to be describing jealousy in such minute detail, he is inventing an affect, because he constantly reverses the order in affections presupposed by opinion, according to which jealousy would be an unhappy consequence of love: for him, on the contrary, jealousy is finality, destination; and if we must love, it is so that we can be jealous..."31

    For mimetic theory, all these figures are incarnations of metaphysical desire. But in order to make them fit the mold of triangular desire it is necessary, as I say, to resort to unusually strenuous intellectual acrobatics. René Girard's genius is to have understood that the manifestations of desire obey a single scheme. His mistake, it seems to me, was to picture this scheme as a mimetic triangle.

    One last, but essential point. In dispensing with the triangle, one loses the mediator and imitation both. Does this matter? If so, how serious a problem does it pose? These questions may strike the reader as naïve, since for fifty years now Girard's theory has been known as the theory of mimetic desire, or simply mimetic theory. Surely, it will be said, removing imitation would deprive it of its identity. But is this really true? In that case the notion of imitation would have to be well defined. But it is not.

    Let us consider the case of violence, by contrast with that of desire, and accept the point of departure assumed by mimetic theory, namely, that violence is contagious, that it spreads like a summer fire in a dry forest. But Girard's anthropology of violence very clearly distinguishes between two forms that the imitation of others' violence may take, which I shall call mutual and convergent, respectively. The classic instance of the mutual form is the duel, characterized by an exchange of blows. A hates B in reaction against the hatred that and convergent, respectively. The classic instance of the mutual form is the duel, characterized by an exchange of blows. A hates B bears for him, and vice versa. The convergent form involves the unanimous hatred of two (or more) persons against some other individual: A hates X because and convergent, respectively. The classic instance of the mutual form is the duel, characterized by an exchange of blows. A hates X because B hates him, and vice versa. The transition from the mutual form (the war of all against all) to the convergent form (the collective murder of a victim) lies at the heart of Girard's theory of the sacred. On this view, the sacred arises from violence by means of self-transcendence: in moving from the mutual form to the convergent form, violence distances itself from itself, so to speak, by erecting a religious institution upon the corpse of the victim, and by virtue of this act is able to contain itself.

    The nature of imitation in the theory of mimetic desire, on the other hand, is much less clear. One would like to say that desire in this case assumes the convergent rather than the mutual form, as the following anecdote suggests. The poet Jean Cocteau received a visit from a young admirer, himself a poet. Stammering from shyness, the visitor said to Cocteau, "Master, I cannot tell you how much I admire your work!" "But I too," Cocteau replied, "I too, my young friend, be assured that I greatly admire— my work." The rules of politeness, though not those of desire, would have dictated reciprocity, with the master affecting to value the work of his disciple. The convergent form adopted by the master is that of pseudonarcissism, according to which the love one has for oneself feeds on the love that others have for you, and vice versa.

    The problem is that mimetic desire may also sometimes assume the mutual form— and yet one continues to speak of imitation. Two subcases of reciprocity must be distinguished. There is, on the one hand, the situation where A desires A because B desires B, and vice versa. Each plays the game of pseudonarcissism, pretending not to be interested in the other, who plays the same game. Alfred Hitchcock brilliantly portrayed an amorous conflict of this type between Cary Grant and Ingrid Bergman in Notorious (1946).

    But there is also the subcase of genuinely mutual love, a situation that the theory of mimetic desire does not recognize, holding that it is doomed to vanish as soon as it appears. Astonishingly, there appears to be no place in mimetic theory for an essential human experience, one that it must be hoped everyone each one of us will have known, at least once in his or her life. Why should it be necessary to completely wall off love from desire? Even if one reserves the category of love for Christian agape, interpreting conjugal love as love for God, can it really be said to have no erotic dimension whatever? Surely we must suppose that before Don Giovanni came along to break up their relationship, Zerlina truly loved Masetto, and Masetto Zerlina. If not, what harm could the great nobleman have done?

    A loves B, in return for the love that B has for her, and B loves A in the same manner. "Love is a contagion," Nicolas Grimaldi writes in another work. "One cannot feel [a love] that is not shared."32 His shrewd insights into the nature of love have enabled me to understand why a theory that does not see that mutual affection is an almost unimaginably stable (though not quite invulnerable) fixed point in the realm of desire is a theory that cannot really understand what it is to suffer from jealousy. There is no better embodiment of the idea of a world completely closed on itself, shut off from everything that lies outside, than two lovers entwined in each other's arms. Just so, there is no worse torment than the one that rips open the heart of a man who imagines his companion in the arms of another, for the sense of exclusion he then feels is total and irremediable. The attempts that have been made in recent decades to transform love into something that can be shared by more than two people have, for exactly this reason, failed lamentably. Grimaldi makes my point for me:

What, after all, is happiness if not the wondrous feeling that one's existence has been granted a reprieve from everything that makes it wither and decay: solitude, fear, anxiety, and torment? To be delivered from these things, it is enough that [lovers] be together: everything has become so simple for them that the simplest thing would be for [their love] to last forever. This is what makes love a sort of rediscovered innocence. It knows no other evil than separation, and scarcely any other good than being together. Freed from all constraint, irreducible to any law, unconcerned with anything that lies outside of itself, love tends toward isolation and forms a world unto itself...
    Love is a desire for the incandescent vitality that another desire, in responding to our own, seems to promise...
    Whereas in order to see something I must be at a distance, and for this reason separated [from it], the privilege of touching, by contrast, annuls all distance between a subject and an object. Nothing separates me from what I touch. But if it is a body rather than an object that I touch, at once it feels from me what I feel from it. No longer do we find ourselves shut up inside ourselves, without being able either to communicate or to share what we feel. The sensation [of touch] is not only mine, it is ours. Not only is it common to us, but there is no one but us who can feel it. Touch therefore annuls the primordial [state of] separation that sight perpetuates. Contrary to what Apollinaire suggested, the simplest way of not having to go on looking at each other, face to face, is to reach out, each of us placing his hands in the other's...
    For love does not consist only in the fact of "being two," or of "being together," but also in that intimacy that puts the other more inside us than we are inside ourselves. . . .
    Love is therefore a state of innocence, a sort of extraterritoriality. Without anyone knowing, it cuts off lovers from the rest of the world, and creates for them a singular world that no one else can enter, a world in which only they exist, in which each lives only for the other. And so the peculiar thing about love is that two imaginations become one and conspire to construct a universe that they alone inhabit.33

It is from such a love-world that a Don Giovanni finds himself utterly excluded. The tragic minor chords of the overture are meant to alert us to the nocturnal darkness of his suffering.

    Love itself is therefore imitated as well, no less than desire, but the form of imitation associated with it is the mutual, and not the convergent, form. I shall say no more about these two figures, for we have already seen enough to be persuaded that to say that desire is modeled on the desire of another tells us nothing about which form it will assume. This underdetermination, Girard would say, is peculiar to desire itself. I quite agree— but the theory of desire does not itself have to suffer from the same underdetermination, any more than the theory of the triangle has itself to be triangular. The advantage of the jealousy-based model is that by holding the basic structure— reciprocal interaction between an autonomous totality and a subject who stands outside of it— constant, it associates mimetic underdetermination with the jealous person's freedom to react to this structure in a great many different ways.

    A final word about the extreme pessimism that accompanies, and perhaps justifies, the extension of the argument from jealousy to human desire in all its manifestations. Jealousy is pure and undiluted suffering, unlike envy, which has a positive, sublimated side to it— the desire to go beyond oneself, the will to push onward, and so on. One speaks in that case of emulation.34 Jealousy, on the other hand, is wholly negative: it corrodes the heart and ruins the brain of whomever it attacks; it is a nightmare from which one has lost all hope of awakening. To be honest, I do not know if the meager resources available to us— reason, self-knowledge, and whatever aid can be drawn from non-pathological ties to others, which is to say relations that are not contaminated by desire— are enough to help us throw off the awful burden of jealousy. Something like divine grace would be required, I fear. My pessimism on this point is the same one to which a lifetime's reflection finally led René Girard.35

Translated by M. B. DeBevoise

1 René Girard, Mensonge romantique et vérité romanesque (Paris: Grasset, 1961). All quotations from this book are taken from the English edition, Desire, Deceit, and the Novel: Self and Other in Literary Structure, trans. Yvonne Freccero (Baltimore: Johns Hopkins University Press, 1965); hereafter referred to as DDN.
2 René Girard, La violence et le sacré (Paris: Grasset, 1972) and Des choses cachées depuis la foundation du monde (Paris: Grasset, 1978). I refer hereafter to the latter work as TH, from the title of the English edition, Things Hidden Since the Foundation of the World, trans. Stephen Bann and Michael Metteer (Stanford: Stanford University Press, 1987), from which all quotations are taken.
3 Nicolas Grimaldi, Essai sur la jalousie: L'enfer proustien (Paris: Presses Universitaires de France, 2010). Hereafter EJ.
4 See Mark Anspach, "Imitation and Violence: Empirical Evidence and the Mimetic Model," in Scott R. Garrels, ed., Mimesis and Science: Empirical Research on Imitation and the Mimetic theory of Culture and Religion (East Lansing: Michigan State University Press, 2011). This forthcoming volume grows out of a seminar organized by Garrels for the Imitatio Foundation with the purpose of initiating a dialogue between mimetic theory and research in the neurosciences and developmental psychology.
5 In an earlier essay I showed that it was not necessary to wait until the advent of twentieth-century literature in the manner of Gombrowicz to encounter the theme of sexual generosity, since French theater in the seventeenth century had already treated it, nowhere more strikingly than in Corneille. See Jean-Pierre Dupuy, "Quasi-object and symbolic exchange: From Corneille's Alidor to Molière's Don Juan," MLN [Modern Language Notes] 104 (September 1989): 757-86.
6 I thank Olivier Rey for this reference.
7 See, for example, my essay "Le signe et l'envie," in Paul Dumouchel and Jean-Pierre Dupuy, eds., L'enfer des choses: René Girard et la logique de l'économie (Paris: Seuil, 1979); my book Le sacrifice et l'envie: Le libéralisme aux prises avec la justice sociale (Paris: Calmann-Lévy, 1992); and a more recent article, "Invidious sympathy in Adam Smith's Theory of Moral Sentiments," The Adam Smith Review 2 (2006): 96-121.
8 See Dupuy, "Quasi-object and symbolic exchange."
9 The translation reads "envy", the French original being "dépit". This is a mistake. The best translation might be "resentment".
10 Molière, Don Juan: comedy in five acts, 1665, trans. Richard Wilbur (San Diego: Harcourt, 2001), 19.
11 Marcel Proust, Du côté de chez Swann (1913), vol. 1 of À la recherche du temps perdu. All quotations are taken from the English translation by C. K. Scott Moncrieff and Terence Kilmartin, revised by D. J. Enright following the 1989 Bibliothèque de la Pléiade edition, In Search of Lost Time (New York: Random House, 1992); this passage occurs on page 385 of Swann's Way (henceforth SW), quoted in EJ, 50. Other volumes cited below are The Guermantes Way (GW), The Captive (TC), and The Fugitive (TF).
12 Here Grimaldi is paraphrasing Proust, who has his narrator conclude a long account of the feelings that Albertine produces in him with the words: "I can never repeat it often enough: [carnal love] was more than anything else an appeasement." [TC, 94] Similar phrases occur at various places in the text leading up to this summation.
13 I thank Olivier Rey for this not unimportant qualification. W. A. Mozart and Lorenzo Da Ponte, Don Giovanni: Opera in Two Acts, Act I, scene 6, recitative, trans. W. H. Auden and Chester Kallman (New York: G. Schirmer, 1961), 61-62.
14W. A. Mozart and Lorenzo Da Ponte, Don Giovanni: Opera in Two Acts, Act I, scene 6, recitative, trans. W. H. Auden and Chester Kallman (New York: G. Schirmer, 1961), 61-62.
15 Molière, Don Juan, 15.
16 See one of Girard's finest essays, "Pour un nouveau process de L'Étranger," in Critique dans un souterrain (Lausanne: L'åge d'Homme, 1976), xx-yy; also Jean-Pierre Dupuy, "Mauvaise foi et 'Self-Deception,'" Raison présente 117 (1996): 59-86.
17 Jean-Baptiste Lamothe, "La pension de famille," in La valise rouge et autres récits (Paris: Éditions de la Table Ronde, 1952), 125-57.
18 Jean-Baptiste Lamothe, "Le garçon en culotte de golf," in La valise rouge et autres récits, 45-99
19 Anspach, "Imitation and Violence."
20 See Sybil Hart et al., "Infants Protest their Mother's Attending to an Infant-Size Doll," Social Development 7 (1998): 54-61.
21 Charles Darwin, "A biographical sketch of an infant," Mind 2 (1877): 285-94.
22 Anspach made this point to me in private correspondence.
23 Roland Barthes, A Lover's Discourse: Fragments, trans. Richard Howard with a correction (New York: Hill and Wang, 1978), 66. The emphasis is mine.
24 The bracketed phrases restore text found in the original French edition but omitted in the English translation; in every instance except one, where Girard himself has added italics ("narcissism for two"), the emphasis is mine.
25 Freud did not have this difficulty, by the way. Recall that the work in which the concepts of the death instinct (Todestrieb) and repetition compulsion (Wiederholungszwang) first appear, Jenseits des Lustprinzips (1920), opens with a description of the game that his grandson invented in response to the comings and goings of his mother, which he reproduced by repeatedly making a toy disappear— "fort!" ("gone!")— and then reappear— "da!" ("there!"). Freud thought that he could explain what caused the child to continually recreate a traumatic experience in this way by reference to a principle located "beyond the pleasure principle." Note, too, that the game played by Freud's grandson has the same structure as the little game that Alidor plays in Corneille's Place Royale: fort!— I give Angélique to my friend Cléandre; da!— I take her back. Angélique is for Alidor what the toy is for the child: a quasi-object.
26 But not Mozart, who reserves him a magnificent aria (Act I, scene 6), "Ho capito, signor, si!" and a no less memorable duet with Zerlina (Act I, scene 5), "Presto, presto, pria ch'ei venga."
27 Here I have in mind the mechanism of double mediation, analyzed above in relation to the case of Valenod and Rênal in Le rouge et le noir.
28 Note that George Clooney's appeal is exerted over men no less than women. I thank Mark Anspach for having challenged me to criticize the standard theory of mimetic desire on the basis of this example.
29 This expression is due to Eric Gans, to whom I am grateful for having reminded me that he had isolated this model some years ago, under the name of resentment, and assigned it a central role in his theory of generative anthropology. Among his many works see The Scenic Imagination: Originary thinking from Hobbes to the Present Day (Stanford: Stanford University Press, 2007); also a recent essay, "Resentment, Jealousy, and Mimetic Desire," posted in the 29 January 2011 edition of his blog Chronicles of Love & Resentment,
30 The final sentence of Camus's book: see The Stranger, trans. Matthew Wald (New York: Knopf, 1988), 123.
31 Gilles Deleuze and Félix Guattari, What Is Philosophy?, trans. Hugh Tomlinson and Graham Burchell (New York: Columbia University Press, 1994), 175. I thank Frédéric Lorton for having brought this passage to my attention.
32 Nicolas Grimaldi, Métamorphoses de l'amour (Paris: Grasset, 2011), 29. My emphasis.
33 Ibid., 49-50, 56, 101-02, 105, 106.
34 "Jealousy and emulation have the same object," La Bruyère says, "which is the prosperity or merit of another, but with this difference, that the latter is a voluntary sentiment, as courageous as sincere, which fertilizes the mind and induces it to take advantage of great examples, so that it not seldom excels what it admires; whilst the first, on the contrary, is violent in its action, and, as it were, a forced acknowledgment of a merit it does not possess . . . ." Jean de La Bruyère, Characters, 11.85, trans. Henri van Laun (New York: Scribner and Welford, 1885), 301.
35 See René Girard, Battling to the End: Conversations with Benoît Chantre, trans. Mary Baker (East Lansing: Michigan State University Press, 2010).

Formatted in HTML by Peter Y. Chou,