French Department, UCLA
Los Angeles CA 90095-1550

Generative anthropology, like all projects of fundamental reflection on the human, whether they be called myth, theology, or social science, is a bootstrapping operation that seeks to explain in human language the origin of human language. This apparent aporia has traditionally been masked by transcendental figures that we may now understand as projections of the originary scene of language. The historical “death” of these figures in the modern era has been taken by some as revealing the inconceivability of the operation itself. What is indeed revealed is its paradoxicality. But the paradoxical is not the unthinkable; on the contrary, without paradox, thinking would be impossible.

Paradox is the privileged road to understanding the human, because paradox reveals the seam–the umbilical hole–in the hierarchy of sign and referent that is the essence of human language. The foundational modern definitions of the sign fail to grasp its double essence as a relation both real and ideal, dualist and monist, “vertical” and “horizontal.”

For Charles S. Peirce, the sign is defined as “determined by something else,” that is, it stands in a horizontal relation to its referent.[1] The inadequacy of this relation is then supplemented by a hypothetical third term or “interpretant,” along the lines of the “third man” of Greek philosophy who furnishes the ground of resemblance between a real man and the idea of a man. The sign-relation is explained through a movement of infinite regress, thereby deferring the horizontal encounter between sign and referent at the cost of the definitional rigor of the system. In distinguishing families of signs by their type of motivation, Peirce can make no place for the arbitraire du signifiant that distinguishes human language; the arbitraire is not a zero degree of motivation but a formal absolute–one that, like all absolutes, is not immune to deconstruction.

In contrast to Peirce, Saussure sees in the sign nothing but verticality. In giving to the bar that separates signifier from signified the–in his perspective primordial–anthropological function of paternal interdiction, Lacan implies the necessity of a generative-anthropological explanation for the emergence of the formal-vertical from the horizontal. The bar is a mystery; if the sign and what it refers to are identical areas on either side of a sheet of paper, one wonders not only what function can be served by turning the paper over, but how we ever got from the real world to the paper in the first place. By bracketing the referent of the sign and substituting its signified or concept, Saussure only defers the understanding of the horizontal relationship between sign and referent as two worldly things.


My solution to this aporia was published well over a decade ago in The Origin of Language. The terms in which it was expressed, as well as those in which it has been repeated and refined in my more recent books, have sometimes been misunderstood as formulating a new “myth of origin.” No doubt I grasped the essence of the problem better than the strategy for articulating it. I offered a minimal hypothesis for the origin of language, of the human–of “man,” as we said at the time. But depicting a scene of origin of language, as opposed to merely affirming language’s essential scenicity, could not fail to give the appearance of an excess rather than a minimum of content. Which is all the more the case when the rival hypothesis is that no hypothesis is conceivable.

The “triangular” version of the originary hypothesis that I present here differs little in substance from that of The Origin of Language, but that difference makes it henceforth impossible to tax the hypothesis with naturalistic naivete. Our fundamental anthropological intuition is far more sensitive to the mode of narrative presentation of the hypothesis than to its real content. The description of a collective scene of origin goes against the grain of a postmodern intellectual climate suspicious of centers of mimetic attraction. My early formulations of the originary hypothesis defied this fact of our intellectual life, as was no doubt necessary to permit something new to emerge. Our intuition of minimality is infallible, but only in the long run; in the immediate, one adds to the imaginative burden in order ultimately to subtract from it. Now that the subtraction has been made, the reader should find it easier to grasp the minimizing principle at work both here and in my earlier narrative reconstructions of the originary scene. The crux of the origin of language is the emergence of the vertical sign-relation from the horizontal one of animal interaction. The originary hypothesis claims that this emergence is conceivable only as an event because the communication of the new sign-relation to its users gives them a conscious, directly manipulable access to the sign as a transcendent form of representation. One cannot be given access to the sign without knowing it, which does not mean knowing what this access is–what language is–in our terms.

The emergence of the sign is the product of the becoming-paradoxical of mimesis in a situation of “mimetic crisis,” of imminent conflict among beings that animal means of differentiation can no longer protect against dedifferentiation. Our model of this transformation requires only one presupposition; that mimesis, having reached a certain level of intensity, becomes incompatible with prehuman forms of differentiation. There are of course many kinds of such differentiation; among the higher apes there is no question of hard-wired divisions of labor as among insects. Differentiation serves the purpose of maintaining a social order, of avoiding or restraining conflict. Pecking-order hierarchies limit conflict to one-on-one attempts to rise in the hierarchy. But it is only as the result of viewing animal hierarchies from the standpoint of human equalitarianism that we understand their essence as limitation. In terms of the evolution of animal social organization, animal hierarchies introduce new degrees of freedom by rechanneling the mimetic energy of intraspecific rivalry.


It suffices then to hypothesize that the indifferentiation of mimesis overcomes at some point the differentiating force of animal hierarchy. Since this hierarchy did exist, there was mimetic conflict to be controlled; since the hierarchy has ceased to function, the conflict can no longer be controlled by it. Hence a new system of control is necessary, one that can operate under the condition of a collective dedifferentiation. This system is language.

The linguistic sign as an aborted gesture of appropriation is detemporalized, cut off from the practical domain in which imitative action slips unnoticeably into violent rivalry. The sign points before it imitates; its horizontal, metonymic relation to its referent turns back on itself as verticality, metaphor.[2] As the object of representation, the central figure takes over the negative role of the mimetic obstacle. The goal of the imitated worldly activity has become its other-worldly model.


Imitation leaves its ontology unthematized; it knows only that since you are like me, I can do as you do. Mimesis thematizes its ontology. This great misunderstood concept of the metaphysical tradition was confined by Aristotle’s Poetics to the esthetic domain for over two millennia until Girard gave it its due by revealing that human desire, and the human as such, obey the paradoxical structure of mimesis.

Imitation of behavior among similar creatures is generally unproblematic. More precisely, I can imitate your actions unproblematically so long as they do not involve the appropriation of a scarce object that we both desire to possess. But the search for such objects is precisely the kind of behavior that makes imitation advantageous. The evolution of higher animals has been driven by the difficulty of obtaining appetitive satisfaction, particularly food. If I serve as your model in the hunt, all will go well until your imitation reaches the point of reproducing my appropriative gesture toward the same object. At this point imitation provokes rivalry; the mimetic model becomes an obstacle.

The becoming-obstacle of the model is not in itself uniquely human. At the most elementary level of imitation, when a swarm of animals gather around a source of nourishment, each one becomes sooner or later an obstacle for the others. But the energy and attention of members of the group are directed to the prey, not to one another. If they do enter into conflict, or even begin to devour each other, this remains incidental to the appropriative operation that ultimately benefits the swarm and the species to which it belongs. The mimetic obstacle is there, but it remains epiphenomenal with respect to the benefit conferred by imitation.


In animal imitation, the becoming-obstacle of the model remains an unpleasant side-effect that must be countered by the very process of mimetic evolution that serves to increase it. Whereas the less fit among the multiple members of lower species can easily be sacrificed, higher animals, of greater individual value to their species, are worth preserving within a hierarchical order that prevents mimetic conflict, or limits it to one-on-one battles for supremacy. The “alpha” animal is the product of a higher level of mimetic tension than can exist within the leaderless swarm. His maintenance of order implies a degree of rivalry with his fellows. But this order is not threatened by the reinforcement of collective mimesis. Animal imitation is naturally self-controlling; the tensions of rivalry reach a semi-stable equilibrium in the dominance of the most robust individuals. The mimetic model stabilizes at higher evolutionary levels in greater order rather than less.

But the emergence of humanity demonstrates the limits of this correlation between imitation and order. The mutual reinforcement of collective imitation leads to the overthrow of the one-on-one mastery of the alpha animal and the formation of the equalitarian human community. Once the conflictive structure of mimesis reaches the point of overwhelming the constraints of animal hierarchy, it can only be controlled, or more precisely, deferred, by the formal hierarchy that is human language. The origin of language is describable as the establishment of this new hierarchy, through which the linguistic sign acquires its uniquely human verticality. The originary hypothesis addresses the mystery of the generation of the vertical from the horizontal, form from content, or, in the old dialectical vocabulary, “quality” from “quantity.”

In our elucidation of this mystery, we must concentrate our attention on the inherent conflictuality or proto-paradoxicality of prehuman mimesis, where we will find the horizontal correlative to the verticality of the sign. This in turn obliges us to analyze more closely than previously the central idea of the originary hypothesis: that the originary sign is an “aborted gesture of appropriation” transformed into a gesture of representation out of fear of the mimetic rivalry of the others.

The Triangular Hypothesis

In a plausible historical reconstruction of the originary scene, the mimetic “other” or model would be a plurality of individual others. The novelty of the present exposition of the hypothesis reflects the hitherto unexploited fact that the original Girardian triangle is already a sufficient, and more truly minimal, model of this scene. At the moment in which the appropriative gesture is initiated, the prehuman mimetic relation is simply imitative: the other is my model, but not yet my rival. I reach for the object on the model of my model’s gesture. But at that moment, I realize the incompatibility of my gesture with his; two movements toward the same object cannot both be fulfilled. The other-model leads me to intend an act that would make him an other-rival.


Here we must stop to consider precisely what makes this situation intolerable. It is easy enough, in the collective version of the hypothesis, to introduce two naturalistic explanatory elements. One is the obvious one of a plurality of others; in a collective situation, a single individual wishing to appropriate the object of common interest risks provoking the hostility of the group. The second, less obvious, element is the preexistence of the animal hierarchy spoken of above. Freud’s scenario of the murder of the father in Totem and Taboo may be rewritten in ethological terms: the alpha animal attempts to exercise his normal privilege in appropriating the object of common desire, but because of the increased level of mimesis and consequent dedifferentiation, the others no longer defer to him but imitate his appropriative gesture “out of turn,” with the consequence that he must abort his own action, the others then following suit.

There is no reason to doubt the plausibility of this scenario; but its naturalism is incompatible with minimalist rigor. This is more obvious in the case of animal hierarchy, but it applies to the plurality of others as well–and this despite the necessarily public nature of the event of origin. For the “public” can be modeled as easily by two people as by two hundred.

The problem is not that an alpha animal may not have existed, but that its empirical existence, even were we certain of it, cannot be substituted for an explanation of the breakdown of its dominant role. To account for the end of hierarchy by its inherent instability is merely to beg the question. Thus the alpha’s existence does not serve as a true explanatory element, but as a relay or intermediary stage that avoids the crucial question of fixing the degree of freedom inherent in the specific operation of human mimesis.

Animal hierarchy arises in order to avert the conflict implicit in mimesis. But this intermediate stage between lower life-forms and the human only interests us insofar as it determines the minimal conditions of emergence of the latter. That reliance on animal hierarchy is inherently misleading is already apparent from the examination of Freud’s model, despite its lack of ethological references. Freud envisioned the prehuman horde as a hierarchically organized group, the liberation or dehierarchization of which corresponded to the appearance of man. But we cannot understand the mimetically dedifferentiated state in which the originary scene takes place simply as a product of the unexplained dissolution of a previous hierarchy. It must be explained from within as the subject’s state of undecidability between the mimetic other-mediator’s two roles of model and rival. This statement of the problem makes clear that it is unnecessary to postulate a protohuman animal hierarchy; mimesis itself defines a hierarchy, however unstable, between subject-self and other-model, and this hierarchy is the basis upon which all others are founded.


Similarly, the triangular formulation of the hypothesis eliminates the need for the independent postulation of the plurality of others. The determining factor in the conversion of the appropriative gesture into a sign is not fear of the violence of the other(s), but the incompatibility of the two roles of subject and other in the mimetic process. This correction should not be taken as a sanitization of our bloody past. “Fear” and “violence” are not the clear-cut categories they appear to be. If fear of the violence of my mimetic model(s) is supposed to explain my failure to carry my appropriative gesture to completion, it hardly explains why I continue to perform the gesture under a new intention, or why I remain within the mimetic configuration rather than seeking to escape from it. The subject’s attachment to the scene, whatever its dangers, demonstrates that mimesis rather than fear is the explanatory element; but in that case, the most economical explanation is the one that presupposes nothing beyond the triangular mimetic configuration. Reduced to the mimetic triangle purged of all naturalistic elements, the originary hypothesis may be formulated as follows: the sign originates as the solution to the “paradoxical state” or “pragmatic paradox” engendered when the mimetic relation to the other-mediator requires the impossible task of maintaining the latter as model while imitating his appropriative action toward a unique object. Put in geometric terms, the parallel lines of imitation must converge toward a single point. The mimetic model is both model and (potential) obstacle; it is at the moment when this contradiction prevents action that the human linguistic sign appears.

The cessation of action in the situation of mimetic crisis is more radical than in that of hierarchical submission, where the non-alpha animal acts out its submission by its very stasis–and where it would normally expect to take its turn after its superiors. We need a more general word than “action” or even “behavior” to describe what is prevented in a truly paradoxical situation–habitus, perhaps–a term that designates simply a coherent mode of being. The psychological correlate of the paradoxical state of mimesis is anxiety, as was the case with Pavlov’s dogs. The situation is obviously similar, but here the feedback loop is minimized; it is not determined by the interference of two conditioning factors that drive the subject to two incompatible actions at the same time, but by an internal contradiction in the (mimetic) mode of behavior itself. When mimetic attraction has reached a sufficient intensity, behavior as such becomes impossible.


What is done in this circumstance is no longer to “behave,” but to produce a sign. The triangular model of the hypothesis permits a more rigorous analysis of how the function and character of this designating sign differ from those of the original appropriative gesture. In the naturalistic model, the clearest function of the sign is apotropaic; it averts potential violence from the group by demonstrating to it that the emitter no longer intends to appropriate the object. To designate is to renounce, to defer possession through representation. This still leaves unclear the nature of the link between renunciation of appropriation on the one hand and imaginary possession through representation on the other. The very fact that we talk about a “link” between what appear to be two wholly distinguishable operations is proof enough that this is not yet a minimal exposition of the hypothesis.

This is the appropriate point at which to recall the Aristotelian notion of mimesis as (theatrical) representation. This notion leaves the potentially conflictive horizontal imitation of others to the subject-matter on stage and retains as its formal definition only the conflict-free vertical representation of reality. In the originary scene, the Aristotelian concept of mimesis applies, not to the original appropriative gesture, which depends on the other as its mimetic model, but to the new designative-representative gesture of the sign.

The sign emerges as a turning away from the other as model to the object of desire as model. In the transformation of the mimetic relationship wrought by language, the subject displaces the intention of his gesture from (unconsciously) imitating the other to (thematically) imitating the object. Signing remains a mimetic operation in the older sense; the sign in its otherworldliness, its “arbitrariness,” can be learned only from others. But unlike even the most stylized of animal behaviors, the sign is intended to make-present a referent other than itself.

How does this doubling of mimetic models resolve the “paradoxical state” of mimesis, the originary mimetic crisis? Prehuman imitation, whether one-on-one or in a herd, has a two-place structure of actor and model; the only difference in the latter case is that the model has many bodies. Animal imitation, lacking in triangularity, can only be the basis of a dualistic order, a “pecking” order of one-on-one relationships. Unlike the triangle of human mimesis, it cannot be expanded into a communal circle in equilibrium around its center. If a central object is the actors’ common goal, they go to it in accordance with the pecking order; the appropriative gestures of the non-alphas are not converted into signs but merely postponed. Hence the circular structure of the scene of representation, fundamental to human cultural phenomena, is unknown among animals, except as it may have evolved in hard-wired form for a single purpose–as in the famous waggle-dance of the honeybee.


The formation of the triangle of human mimesis, the minimal structure in equilibrium, resolves the crisis by permitting the resumption of mimetic activity. The sign begins as the same physical action as the aborted gesture of appropriation, but the intended deferral of horizontal interaction with its object allows it vertically to “intend” this object in the phenomenological sense, to take it as its theme. What unblocks the mimetic process has its source within mimesis itself. As the appropriative intention of the original gesture makes its imitation impossible within the framework of animal relations, it increasingly focuses attention on the object to be appropriated. When I imitate the other in his appropriation of an object, my attention focuses on him if I have my own counterpart to his object, but on the object if it is the same for us both. The intensification of mimesis, by putting into question the equivalence between my object and the one that partakes of the aura of the mediator, makes it increasingly less satisfactory for me to choose an object different from his. The normal child chooses a mate in imitation of his father’s choice; Oedipus, the mimetic archetype, can take no wife other than his father’s.

The movement toward the object–and concomitantly away from the model–is inherent in mimesis as such. The appropriative gesture is so to speak already “predisposed” to re-present the object even as it performs its practical function.[3] What remains for the originary scene to accomplish–but it is the accomplishment that makes all the difference–is the thematization of the intention to represent and defer appropriation. Once attention to the object and its interdiction by the other have increased to the point of rendering appropriative action impossible, the mimetic shift to the object formalizes–in effect, brings to (human) consciousness–this already-existing tendency.

This analysis moves in the opposite direction from Girard’s original exposition of mimetic desire in Mensonge romantique, which consists in bringing to light the mediating third element behind the “romantic lie” that conceives of desire as a dual object-relation. Here it is the object of desire rather than the mediator that is exposed as central to what had appeared to be a one-on-one relationship of behavioral imitation. This counter-intuitive result requires explanation.

Appetitive behavior normally directs itself to objects, and it is not stretching the analogy too far to say that it “intends” these objects. A cat hunting a mouse knows what object it is looking for as much as a human hunter stalking a deer. But the cat’s behavior, unlike the hunter’s, is an unlearned routine that includes its object categorically within it–not hunting behavior that happens to alight upon a mouse, but mouse-hunting behavior. When imitative learning does take place with respect to such behavior, the object, as part of the behavior itself, does not fall under the spell of the imitation; if one cat learns from another a new mouse-hunting technique, no particular mouse receives thereby a supplementary value. The supplement that comes from re-presentation of the object can only arise when it is not already present as an element of the learned activity.


Desire is always mediated desire. The movement of appetite toward desire is that of an intensified mimesis that discovers not only behavior but the goal of behavior in the other. What causes the late emergence of the object into the mimetic equation is not indifference to it but, on the contrary, the practical object-orientation of animal behavior. Because the object as source of food, shelter, sexual release, and so forth is less freely chosen (more “scarce”) than the behaviors of the subject by which it may be appropriated, the techniques of appropriation are subject to mimetic learning before there is any need to “learn” the object of appropriation. It is only at an advanced stage of mimesis that not merely the action itself but its goal falls under the influence of the mediating other.

Why should the intensification of mimesis lead the subject away from the other’s behavior toward the object to which it is directed? This movement reflects an internalization of the model’s motivation, the self’s closer assimilation to the other’s own reality. The more closely I imitate my model’s goal-directed action, the more I share the goal of this action, which is not located in the action itself but precisely in its external object. (This analysis applies as well to self-directed actions; a higher level of mimesis will lead me, for example, to imitate the other animal’s “narcissism” and groom it rather than myself.) Whence the apparent paradox that as imitation becomes more intense, it prefigures the triangular structure of human representation, focusing less on the model’s behavior and more on the object to which it is directed.

The paradox of the sign

The sign, as we have noted, is the conversion of a gesture begun in imitation of the model’s appropriative gesture into the “imitation” of the object that was the aim of this gesture. In performing the sign, I abandon my imitation of the other’s original intention of appropriating the object; I turn back from the object we desire in common. In consequence, my situation in performing this gesture is once again compatible with that of the other, whose action can take place simultaneously with mine without any danger of convergence on the object. The two gestures are not parallel as before, they are both directed toward the object, but they no longer seek to remove it from its central position. The object has now become the center of a scene.

The emission of the sign creates verticality out of what was previously a horizontal relationship of appetite and appropriation by combining the roles of model and object in a single behavior. I imitate the other in my énonciation and the object in my énoncé. Instead of my action being a simple means of self-expansion into the world through the incorporation and obliteration of external objects, it becomes a means to preserve these objects by reproducing them within myself. I can now continue to imitate the gesture of my model despite the presence of an obstacle to appropriative action. Because the model does not disturb my signing behavior, it is the object that is perceived as the obstacle to its own appropriation; this is what we call its sacrality.


This model of the first emission of the sign is that of its “early” or “thoughtful” emission. In the sense that I emit the sign as the result of my own abortion of the gesture of appropriation rather than in imitation of my model, the decision to emit the sign is the originary example of thinking. This may be contrasted with the emission of the sign under the mimetic or “rhetorical” influence of the model by the “late” participant still seeking to appropriate the object.

Once again, the specific difference between these two moments of the emission of the sign is clarified by the use of a minimal rather than a naturalistic context. The originary sign is the first instance of the free, conscious, intentional thematization of an object. Our analysis cannot be content with showing that the sign is freely performed, but must show how freedom is born with the sign. Like the birth of verticality from horizontality, the birth of freedom from necessity is another statement of the paradox of originary signification. Its explanation can never be complete; as the birth of a new level of complexity, it is irreducible to any earlier configuration. But rather than lament the futility of intellectual bootstrapping, we should take such paradoxes–and all paradox reduces to this one, the paradox of the human-as-such–as guarantees of the inexhaustibility of originary thinking.

In a brief discussion of the question of freedom in the Introduction to Originary Thinking, I used Kant’s formulation of the esthetic judgment (“without a concept”) as my model for the freedom of the signifying intention: the subject was influenced by the “beauty” of his gesture, that is, by its ability to re-present the all-desirable central object. This is a suggestive formulation; but it complicates the matter by introducing the category of the esthetic, in which the subject’s attention oscillates between sign and referent. The esthetic is dependent on the sign; it perpetuates our paradoxical experience of the sign’s thematization of its referent as already significant. To grasp the originary freedom of the sign prior to the reinforcement provided by the esthetic, we must attempt instead to define “freedom within the limits of mimesis”: to understand how a mimetic act can free itself from “instinctive” or non-reflective dependency on its model.

The transformation of the aborted gesture into a sign is a movement from the imitation of a human model to the “imitation” of an appetitive object. In the mimesis of the object, the subject is not copying another’s gesture, but representing the object itself.

Let us consider for a moment the subject-object relation. I appropriate an object in order to fulfill an appetitive need. Whether or not I am imitating a human mediator is not critical so long as I indeed have this need. Up to this point, mimesis is merely a beneficial way of learning the technique of such necessary appropriative gestures.


We might then be tempted to call even the appropriative gesture “free” when it arises in a non-mimetic context. I may have learned it from another, but my performance of it is dictated by my own needs, and in higher animals these needs themselves need not obey a strict physiochemical calculus of stimulus-response. For example, animals engage in various kinds of play. Here we come up against the traditional question of “free will,” which is, along with the existence of God, one of Kant’s antinomies of pure reason.

Generative anthropology offers a new understanding of the concept of freedom as well as of that of God. Ours is a strictly anthropological explanation; it makes no cosmological claims. The question of freedom vs determinism, like that of the existence of God, is really a purely anthropological question. One can no longer take seriously the nineteenth-century “science of religion” that wanted to derive the concept of God from our awe of the cosmos. The reality is just the opposite: having sufficiently deferred human violence by means of the concept of God, we become interested in the relatively dangerous cosmos on the model of extremely dangerous humanity. Religion tends to apply to the cosmos a model of divine power that is indeed of value in anthropological situations but has little functionality in cosmological situations. In times of crisis, cosmic or otherwise, we appeal to God because in our fundamental, originary model of crisis, the sign as name-of-God provides the solution.[4]

The problem of freedom vs determinism is equally anthropological rather than cosmic, “cultural” rather than “natural.” To say that the future movement of a particle is “determined” is to conceive a mind potentially aware of this determination. The simple anthropological test of determinism is the following; if after calculating the future state of a system, I can inform the system of my calculations without leading it to deviate from them, that system may be called determined. If, on the contrary, I must hide my calculations to avoid such deviation, then the system is free; for someone within the system could eventually perform the same calculations as I have.

The obvious objection to this definition is that it is biased in favor of language-users; how could I convey the results of my calculations without language? But the burden of proof should lie not with the definition but with its critics. It is for them to show why, if human language is just one among many means of communication that makes no real difference to the matter of free will, it has such an effect on the system that contains it, why the animals whose languages they study so intensively are incapable of such feats. Nor should the question be deflected by bringing in artificial intelligence. Jusqu’à nouvel ordre, computers have been constructed by human beings for their own benefit.


The freedom of signing as an act of representation distinguishes it from imitation as a new, human variety of mimesis. To imitate is not to represent. I imitate you because we are analogous beings; I need make no conscious effort to follow your gestures, to thematize them as objects of representation.[5] The mimetic crisis leads to stasis precisely because prehuman imitation is non-reflexive; the subject has no knowledge of itself as a self imitating another. In contrast, my representation of the object is a conscious thematization. I am not like the object; I cannot follow it by analogy. When I imitate you, I imitate your action, make movements analogous to yours; but when I represent an object, I designate it, not a particular action of it. My intention of the object is an intention to recall it into being, to double it using only my own resources. I cannot perform the sign, as opposed to the gesture of appropriation, without thematizing the purpose of the sign to represent its object.

The key to the freedom of the sign lies in the detemporalizing/retemporalizing movement discussed in Originary Thinking, Chapter 6, (“Narrativity and Textuality”). Imitation has no inherent form. The practical gesture is “horizontally” contiguous with its object; its lack of end-in-itself is visible in its outward formlessness. Because appropriation ends with the object, not with the act itself, if I imitate you successfully, I have no awareness of the limits of my gesture, which are imposed upon it from without. The non-formal quality of the practical gesture is reflected in the continuity between its temporality and that of the life-world to which it is subordinate; the hunter’s movements must obey the rhythms of the animal he hunts rather than his own. In contrast, the sign is detemporalized, cut off from its natural aim and therefore from the time in which such aims are realized.

In the discussion in Originary Thinking, the sign’s detemporalization of the original appropriative gesture was considered tantamount to its constitution as form, even esthetic form (for example, “In the originary scene itself, in the presence of the sacred, the esthetic contemplation of the sign is the complement of the sign’s desiring prolongation toward the center” [103].) But in the present discussion, we stand at the origin of form, which we must explain without recourse to the notion of the esthetic. This explanation will return us to the question of freedom.

Abortion of the gesture is not in itself detemporalization; it is such only when the aborted gesture becomes an action in its own right–an action of a new kind, devoid of direct worldly aim. We may say that it defers this aim, that its very existence as form is a worldly realization of deferral, by which I refer to the fundamental equivalence, pointed at by Derrida’s seminal term différance, between differentiation as marked by the sign and deferral of the mimetic conflict that the loss of difference risks bringing about. The sign re-presents the object as what may truly be called an object of desire, now that its potential appetitive attractiveness is cut off from practical action. Desire is not first experienced and then “repressed,” as in the psychoanalytic model; its thematization of its object is itself a product of the repression of the possibility of discharge in appetitive satisfaction.


The detemporalized gesture possesses a new, formal temporality. The beginning and end of a form are within the form itself. At the origin of formality is the new aim of the aborted gesture, which is transformed from a practical into a representational act. Within the practical realm, the goal is no longer to appropriate the object in imitation of the human mediator but to imitate the object to the latter’s satisfaction, that is, well enough to make him understand the new sense–which can already be called the “meaning”–of the gesture. This is an aim external to the gesture itself, but one that depends on its formal closure as a representation. This closure is not perceived within the practical world but on the other’s imaginary scene of representation. In practical terms, this imaginary aim mediates the deferral of conflict, averting the potential wrath of the other-mediator toward his disciple-rival.

It is justifiable to follow past practice in calling this originary form of representation “designation” and the utterance that performs it an “ostensive,” with its connotation of pointing. (But it is useless in this context to speculate on the oral vs manual character of the “aborted gesture”; we may just as well assume it to have both.) In principle, any appropriative gesture will “point” toward its object in the sense of serving as a natural sign that draws attention to it, but this within the continuum of worldly action that leads ultimately to its appropriation. Now that the latter course has been foreclosed, the sign does nothing but “point,” not in anticipation of further action, but as re-presentation, calling to attention. The aim of the action now having shifted to communicating a representation of the object to the other, the beginning and end of this action, whatever feedback is available from the other, are ultimately determined by the internal or formal coherence of the gesture itself, since it is this coherence that makes it an object of perception and thereby communicates to the other the intention to re-present the object. The formality of the signifying gesture, however different its form may be from that of the object it designates, is of the same kind, in contrast to the nonformality of the practical gesture–the normal aim of which is rather the de-formation of the object, the destruction of its form for the benefit of one’s own, as is the rule in appetitive operations.

The formality of the gesture is an objective quality analogous to that of the forms of objects in the real world; the gesture as part of a sequence having its principle of coherence outside itself becomes an autonomous object of perception or Gestalt in itself. This alone is sufficient to allow us to speak of the sign as an imitation of its object. At the same time, the creation of a formal object in the sign requires that the criteria for formal closure be imposed by the subject. A feedback loop connects the progress of the sign with the perception of its completion, something not required either in prehuman mimesis or in the performance of life-world routines. This loop anticipates the oscillatory structure of the esthetic, but it determines the specificity of the signing action as the creation of form-in-general rather than beautiful form. The sign is well-formed rather than beautiful; it does not call for “a second look” as does esthetic form. This first loop of formal judgment, in which the relation between sign and object is thematized so that we examine our sign and “see that it is good,” is the minimal structure of human free will.


Paradox is a structure of language; it cannot be conceived without the sign. But neither can the sign be conceived without paradox. The horizontal and the vertical cannot be cleansed of one another. The doubling of reality by the sign-world cannot follow either the Saussurean or the Peircean model. The sign that is in the world represents the world it is in; the sign that stands above the world remains within the world of the sign.

There is more than an analogy between this situation and that of the mimetic subject who finds the doubling of his model’s gesture blocked by the collision of converging trajectories. There are neither two places in the same universe nor two separate universes for the one to find room to mimic the other. The constitution of the sign is the creation-and-deferral–the différance–of paradox. Paradox itself is paradoxical; that is what makes it paradox. It cannot be reduced to “lowest terms,” only deferred. But neither is it ever present before our eyes; it is always in a state of deferral.

The subject sets the process in motion by imitating the object, doubling within his own action the inaccessible goal of his mimetic gesture. This action liberates the gesture from its stasis by separating the components of the mimetic blockage. On the one hand, the signing gesture, unlike the gesture of appropriation, can be imitated without further difficulty, since it has its end in itself and not in the material world. On the other, the object is designated, represented as an obstacle to the very appropriative action its designation incites. Instead of two hands converging on the object, we have two coterminous gestures cut off before they can interfere with one another, but which thematically reject the practical aim that incited them in the first place.

What this transformation generates is a radical redefinition of the “practical,” a program for the stabilization of the human community through signification the validity of which has been demonstrated by our continued use of it for at least 30,000 years. This program creates paradox by deferring it. The referent is no longer a simple object of appropriation but an object of signification; there is no way to maintain a barrier between its “natural” and its “cultural” being. Formally to designate the object through language is inevitably to designate it as an object-of-designation; the object I mean is always already an object-meant. This is not an artifact of mimesis that can be overcome with the benefit of a lucid theory of desire; it is already the case for the first sign, the originary aborted gesture, at the first moment of human thought.



1. The complete definition, given in his letter to Lady Welby of December 23, 1908, is as follows: “I define a Sign as anything which is so determined by something else, called its Object, and so determines an effect upon a person, which effect I call its Interpretant, that the latter [person] is thereby mediately determined by the former [object]. My insertion of ‘upon a person’ is a sop to Cerberus, because I despair of making my own broader conception understood.” (Charles S. Pierce, Values in a Universe of Chance, Garden City, N.Y.: Doubleday Anchor Books, 1958, p. 404.)(back)

2. This suggests an originary analysis of Roman Jakobson’s familiar metonymy-metaphor dichotomy as homologous to the opposition between the horizontal and vertical components of the original mimetic paradox.(back)

3. To this mimetic predisposition corresponds the cognitive evolution toward the formation of prelinguistic “concepts” referred to by Derek Bickerton and others as the necessary preliminary to human language. (See Bickerton’s Roots of Language [Ann Arbor, Mich.: Karoma Publishers, 1981] and his later synthesis Language and Species [University of Chicago Press, 1990], which is even less concerned than the earlier work to hypothesize an originary scene of language.) Imitation is “always already” protolinguistic, not merely in the abstract sense that après coup we can recognize an unthematized version of concentration around the center, but in the very concrete sense that neurons are becoming devoted to differentiating among categories of objects as a result of this concentration. This having been said, because Bickerton’s conception of early linguistic evolution, although a considerable advance over purely linguistically-oriented theories, does not recognize the centrality of mimesis, it fails to take issue with what is after all the fundamental question of the origin of language: the crossing of what Bickerton calls the “Rubicon” of interactive speech.(back)

4. See my “The Unique Source of Religion and Morality,” Anthropoetics I, no. 1 (June, 1995; URL:, and Contagion, forthcoming.(back)

5. The game of “Simple Simon Says” is a practical demonstration that in matters of simple mimesis, language only gets in the way. The ironic point of the game is the difficulty of listening for the words when it is so natural “simply” to repeat the gesture.(back)