The three Graces, or the allegory of the gift

This work is licensed under the Creative Commons | © Denis Vidal. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau4.2.024


The three Graces, or the allegory of the gift

A contribution to the history of an idea in anthropology

Denis VIDAL, Université Paris Diderot

Translated by Eléonore Rimbault

Marcel Mauss’ The gift is one of the most revered texts of social anthropology. It is also one of the most debated. But, paradoxically enough, these debates have not focused on the main cultural tradition to which the famous essay may be attached. In this article, I attempt to show that Mauss’ anthropological theorization of the gift perpetuates and slightly modifies a very ancient tradition of reflection, fundamentally based on a few concepts—charis, gratia, and grace— all of which played a crucial role in European cultural history. This article also reveals the specific function played in this context by the allegory and iconography of the three Graces.

Keywords: Three Graces, allegory of the gift, Mauss, gratia, charis, Seneca, Chrysippus, Lévi-Strauss

Why are there three Graces and why are they sisters? Why do they hold hands?
Why are they smiling, youthful, virginal, wearing a loose and transparent dress?

—Seneca, On benefits1

As we learn from eminent scholars, art historians, students of antiquity, archaeologists, and philologists, the signification of the allegory of the three Graces has been known since Seneca. Seneca had found its meaning in Chrysippus, who certainly knew it from Hecaton. And it is not impossible that Chrysippus had himself taken this allegory from Epicurus. It is, in any case, a plausible hypothesis according to Edgar Wind, knowing, as Diogenes Laertius reports, that Carneades did not hesitate to call Chrysippus a “literary parasite of Epicure,” because he wrote on every topic that the latter had dealt with before him (Wind 1958: 34–35). Anyone who has, even absent-mindedly, read Marcel Mauss’ The gift will, by reading the persisting interpretation of the allegory of the three Graces, feel a strange sense of familiarity—a sense of déjà lu not necessarily stemming from a flawless knowledge of classical authors. Allow me to quote in full the answers Seneca gives to the series of questions I have just reported:

Some people advance the view that one of them stands for giving a benefit, one for receiving it, and one for returning it. Others hold that they represent three kinds of benefactors: those who confer benefits, those who return them, and those who accept “benefits” and return them at the same time. But no matter which of these interpretations you decide is true, what good does this specialized knowledge do for us? And what about the fact that the group dances in a circle with intertwined hands? It is because a gift (or “benefit”) goes through an orderly sequence, passing from hand to hand and yet returning to the giver, and loses its integral character if the sequence is at any point broken, being most beautiful if the continuity is maintained? In the dance, though, the older sister has a greater value, like those who confer “benefits.” The Graces have joyful expressions, just as those who give and receive benefits generally do. They are youthful because the remembrance of “benefits” should not grow old. They are virginal because benefits are unspoiled, pure, and revered by all. Benefits should not be constrained or obligated—that is why the Graces wear loose robes. And the robes are translucent because benefits want to be in full view. (Seneca 2011: lines 20–21)

It is difficult to find a better illustration and a clearer explanation of Mauss’ theories of the gift. But that is also why it is surprising to note the absence of any reference to Seneca in general (and to this text in particular)—not only in The gift, but also, it seems, in all his published work.2 It is especially startling if one considers the imposing collection of references, particularly those in Greek and Latin, that Mauss used to support his work, as well as the deep familiarity many of his closest collaborators had with classical literature (e.g., Huvelin,3 Davy,4 and many others).

This is all the more curious considering that in the numerous commentaries that followed The gift, scholars agreed almost unanimously on one point: the discrepancy between the quality of the intuitions expressed in the text and the less convincing character of some the demonstrations (based on ethnographic facts) included in it. Students of Oceania, such as Raymond Firth and Marshall Sahlins, raised this criticism concerning Mauss’ interpretation of the hau, a Polynesian category quite central to his argument on the gift (Firth [1901] 1965; Sahlins [1972] 2004; Boyer 1986). The perception of this discrepancy has led some of Mauss’ commentators to criticize him, others to see in it further evidence of his perspicacity, but it seems that very few of them questioned the origins of his intuitions. One may reconsider, from this perspective, the meaning of Lévi-Strauss’ famous observation about The gift in his Introduction to the work of Marcel Mauss ([1950] 1987: 47): “Are we not dealing with a mystification, an effect quite often produced in the minds of ethnographers by indigenous people? Not, of course, by ‘indigenous people’ in general, since no such beings exist, but by a given indigenous group, about whom specialists gave already pondered problems, asked questions and attempted answers.” Lévi-Strauss continues, explaining that

we may infer that Mauss is seized by hesitation and scruples at the most crucial moment. He is no longer quite sure whether he must draw a picture of indigenous theory, or construct a theory of indigenous reality. He is very largely right to be unsure, for indigenous theory is much more directly related to indigenous reality than a theory developed from our own categories or problems would be. So it was a very great progress, at the time when Mauss was writing, to approach an ethnographic problem from the starting point of his New Zealand or Melanesian theory, rather than to call upon Western notions such as animism, myth or participation. But indigenous or Western, theory is only ever a theory. At best, it offers us a path of access, for, whether they be Fuegians or Australian Aboriginals, the interested parties’ beliefs are always far removed from what they actually think or do. Once the indigenous conception has been isolated, it must be reduced by an objective critique so as to reach the underlying reality. We have very little chance of finding that reality in conscious formulations, a better chance, in unconscious mental structures to which institutions give us access, but a better chance yet, in language. (Lévi-Strauss [1950] 1987: 48–49)

One may suspect, however, that Lévi-Strauss, too, was seized by hesitation and scruples at the most crucial moment while analyzing the work of his predecessor. Is he not affirming, for example, in a most magisterial manner that “indigenous or Western, theory is only ever a theory”? But, in the particular case of Marcel Mauss, as I will attempt to demonstrate here, this sentence should be reformulated more accurately in the following way: “Western or Western, theory is only ever a theory.” The progress of logic could have given rise to the suspicion that the issue deserved further consideration. Unfortunately, this is where Lévi-Strauss, in his exegesis of Mauss’ work, seems to abandon the application of his own analytical principles and neglect their implications. Once Mauss’ conception had been defined, like that of any “indigenous people,” “it [had to] be reduced by an objective critique so as to reach the underlying reality. We have very little chance of finding that reality in conscious formulations, a better chance, in unconscious mental structures to which institutions give us access, but a better chance yet, in language.” (ibid.) Lévi-Strauss was probably right when thinking that Mauss had been “mystified” by indigenous people. But by identifying the “indigenous” people as Māori, he ran the risk of being mystified at once by a native and as a native. In line with Mauss and Lévi-Strauss, let us see now if we can catch sight of the “many moons that are dead, or pale, or obscure” (ibid.: 66).


The short passage in Seneca, which we previously cited, combines several elements that can be briefly enumerated. It primarily discusses two notions: one is originally Greek, charis, and the other one is Latin, gratia, and was to be considered as the translation of the former. It also involves a group of deities, also originally Greek: the Charites, who became the three Graces, and whose iconography can be traced over more than twenty-five centuries. Finally, it speaks of the interpretation of this allegory, which involves a reflection on exchange as a concept.

Taking these various associations as a point of departure, we shall wonder what kind of continuity can be detected between these notions, deities, this iconography, and some of the contemporary anthropological developments regarding forms of exchange. In particular, we will try to elucidate whether the consistency between these different elements is due to anecdotal similarities or if it is possible to detect the existence of a genuine common ground, firmly rooted in our traditions of thought, which is still reflected in our way of thinking, analytically or not, about exchange practices. This is indeed the hypothesis that will be developed here, and which summarizes the perspective adopted in the comments that follow.


The use of the term charis is longstanding. We find it in Homer. The two meanings of the word “grace” (as beauty and as favor) reflect the two chief meanings of charis, which seemingly coexisted from the outset. Charis also refers to the two ideas of pleasure and granted favor (Chantraine 1980). More importantly, the set of notions associated with this term in ancient Greece are consonant with a stream of thought on gift and exchange, as the works of Jean-Pierre Vernant ([1965] 2006) and Marcel Detienne ([1972] 1994) show.

Vernant defines the Greek charis as “the divine power manifest in all aspects of gift giving and reciprocity (the round of generous liberality, the cordial exchange of gifts), which, in spite of all divisions, spins a web of reciprocal obligations” (Vernant [1965] 2006: 163). The author adds that “one of the oldest of all functions of charis is a woman’s giving herself to a man.” Detienne’s studies on mythology, in particular those related to the myth of Ixion, corroborate this general analysis of the charis, showing what happens a contrario in its absence:

In archaic Greece, as in other societies of the same type, the circulation of women cannot be dissociated from the exchange and the circulation of goods. At every level of social life the binding rule of gift and counter-gift applies, not only between men and gods, and between men and nature, but also between different groups of men. To thwart charis at any of these levels is to upset the entire system of exchange and presentation and to carry corruption to the very heart of the social order. (Detienne [1972] 1994: 88)

In order to draw these reflections closer to the analyses of Mauss, especially those on the Polynesian concept of hau in The gift, several other aspects of the concept of charis must be noted, although it is not possible to develop them fully here. First, there is a double lack of differentiation associated with it. It appears, on the one hand, in the absence of a distinction between charis seen as a quality that would have been deified and the Charites (i.e., the three Graces) as a group of deities. Moreover, at this stage it seems that nothing radically distinguishes each of the deities’ functions, even if, in some cases, they appear as a collective group of goddesses in which none are individualized, while in other cases they are singularized, and each one has a different name (Vernant [1965] 2006: 360). There is also the association, ancient but never lost, of the Charites with the notions of growth and fertility (ibid.: 277). Finally and above all, there is this specific quality which the charis grants to everything from which it emanates. Let us cite Vernant once more:

To the Greek, the charis does not only emanate from a woman’s body, or from any human being who “shines” with the beauty of youth, with a sparkle (often found in the eyes) that inspires love; it also emanates from finely chiseled jewelry, carefully carved jewels, and from certain precious fabrics; from the scintillation of metal, the bright reflection of a precious stone’s water, the polychrome quality of a weaving, and the vivid colors of the depiction of an animal or a vegetal setting with an intense liveliness. The silversmith’s and the weaver’s works also shine splendidly and render the gleam and light of flesh. (Vernant 1965: 261 n. 31)5

There is no doubt that this specific quality of the beings and objects that circulate in an exchange (which Vernant identified) very much resembles the mysterious quality that Mauss tirelessly tracks in every monograph, and over the continents, which he designates metaphorically as the mysterious form of power manifesting itself in the practice of exchange. Hence we may affirm that everything called for the Greek concept of charis to figure as a prominent example in Mauss’ work. Its almost complete absence should only be the first among the reasons for our astonishment.


In spite of their lack of etymological kinship, the term gratia came to be considered as a legitimate translation of charis given that it was precisely and constantly used

Figure 01
Figure 1: The Three Graces, fresco from Pompeii, Naples Museum of Archaeology.

by Latin authors as its equivalent. Émile Benveniste ([1969] 1973: 160) describes as follows the social process in which the word gratia and its linguistic family are involved:

The connection with Latin words shows that the process at the beginning consisted of giving service for nothing, without reward; and this service, which was literally “gratuitous,” provokes in return the manifestation of what we call “gratefulness.” The notion of service that does not demand a counter-service is at the root of the notion, which for us moderns is twofold, “favor” and “gratefulness,” a sentiment which is felt by the one who gives and by the one who receives. They are reciprocal notions: the act conditions the sentiment; the sentiment inspires a certain form of behavior.

Once again, we are definitely in the realm of the gift. But with this Latin equivalent to the notion of charis, the “sentiments” inspiring the actors, as well as the way in which these feelings are called upon in all aspects of social life, are paramount. This is consistent with what Claude Moussy has shown in a work entirely dedicated to the analysis of gratia and its linguistic family:

All the values of gratia common in the Classical period can easily be explained from its fundamental meanings, both abstract: gratia as “recognition”; and concrete: gratia as “payment in return” or “benefit.” They all bear a link to the practice of benefaction, and this extension of gratia, stretching out from the vocabulary of recognition toward various domains, in particular to that of politics, is a good illustration of Seneca’s thesis, which envisions the entire edifice of social life resting on the pair “benefit–recognition.” (Moussy 1966: 410)

However, the common definition of gratia as “charm, grace, pleasure,” is incommensurable with the previous meanings, and the intervention of charis in this new development of the Latin word seems quite plausible. As to the Christian sense of gratia, it derives from charis, usually translated as gratia in the Latin versions of the first works of Christian literature.

Hence, from a comparison of the Latin notion and the Greek one, we can draw several observations. With gratia, the relational structure fostering cohesion between actors emerges, along with the definition and the analysis of sentiments that inspire or should inspire the practice of gifting. Thus is it not surprising that Seneca’s On benefits, from which I excerpted the interpretation of the allegory of the three Graces, is conceived as a practical and pedagogical treatise destined for those involved in gift giving. But it is precisely this pedagogy that presents a difficulty.

When speaking of practices associated with beneficence, obligations to give, to receive, and to return, Seneca affirms that the challenge is to “organize the topic which more than any other binds together human society” (Seneca 2011: 22). But this affirmation comes here to support a plea in favor of practices that he judges insufficiently present in collective life. This is how one sees, in Seneca, the awareness of a dissociation between two exchange circuits—which the terminological usages account for more generally, as Benveniste ([1969] 1973: 161–62) points out:

It would be a serious error to believe that economic notions originate in needs of a material order which have to be satisfied, and that the terms which express these notions have merely a material sense. Everything relating to economic notions is bound up with a far wider range of ideas that concern the whole field of relationships between men and the relations of men with the gods. These are complex and difficult relations in which both parties are always implicated.

Yet the reciprocal process of supply and payment can be interrupted voluntarily: thus we have services without return, offerings “by grace and favor,” pure acts of “grace,” which are the starting points of a new kind of reciprocity. Above the normal circuit of exchange—where one gives in order to obtain—there is a second circuit, that of benefice and gratefulness, of what is given without thought of return, of what is offered in “thankfulness.”

Thus, contrary to charis, which according to Hellenic scholars was used in ancient Greece to signify a quality inherent in any authentic social exchange, gratia seems to have been dedicated to a much more specific exchange form, with specific connotations. These connotations, more and more present in the use of gratia, also mark the modern notion of “gratuitousness.”

Moreover, in the passage from charis to gratia we see the survival of other connotations associated with this term in ancient Greece. The concepts of “charm” and “pleasure” are still present, but there is more. When describing an old usage of charis, Vernant, as we have seen, spoke about “a woman’s giving herself to a man.” Let us refer now to the Christian signification, especially in St. Paul, “who seemingly introduced the word charis into the vocabulary of Christianity; who, in any case, was first to systematize the theology of Grace. Grace is above all a gift from God: it is God giving Himself gratuitously to mankind in the person of Jesus Christ” (Moussy 1966: 451). It is, in fact, that same charis that Christian baptism bestows on human beings.

It is as if the comparison of charis and gratia reflected the gradual dissociation of a “total phenomenon.” In ancient Greece, the notion of charis—perhaps a genuine equivalent of the Polynesian hau—is commonly interpreted by scholars as the cement of any social exchange, and a mysterious quality present in both things and people; the notion of gratia, however, disseminated and preserved the Greek connotations of charis, but also altered them quite significantly. It came to denote an increasingly particular type of exchange (“gracious,” “gratuitous”) and specific qualities, be they sensible—the beauty of a woman—or, on the contrary, immaterial—Divine Grace. What has been lost apparently is the sense of an immanent and general connection between all the different connotations once associated with the term charis and which referred to all exchanges that bind society. But is it really from such a perspective that we must interpret the allegory of the three Graces?

The Charites

In order to progress further in our analysis, we must now turn to the Charites, minor deities of ancient Greece, whose memory was preserved thanks mainly to their iconography. We know little about these deities and their worship in ancient Greece. The references found in ancient authors, although quite frequent, are extremely laconic.6 The cult of the Charites seems to have been primarily associated with the “fertility” of nature, the “pleasure” and “joy” in love and human relations, and, finally, with gratitude and “benefits.” Each of these associations, of course, must be understood and resituated in the cultural and psychological context of ancient Greece. Suffice it to say here that the three Graces each bore names and more importantly had a filiation and alliances that varied greatly according to the poets who mentioned them (Homer, Hesiod, Pindar, Pausanias, etc.). Sometimes considered as three sisters, the daughters of Zeus, they were not always called by the names which Hesiod reports: Aglaea, Thalia, and Euphrosyne. As minor deities, the Charites were also relatively undifferentiated from nymphs, Muses, and Hours. Furthermore, the worship of these various entities was frequently conflated. Their function can also be contrasted with other groups of minor deities like the Erynies (Vernant [1965] 2006: 412 n. 43). Their cult was sometimes brought to the forefront, as it was at Orchomenos in Boeotia; and it could occasion specific celebrations (the charitesia). But in general, their cult was associated with and subordinated to that of other deities such as Aphrodite or Apollo.

Figure 02
Figure 2: The Three Graces, Acropolis Museum, Athens.

In the fragments of scattered information available about the Charites in ancient Greece, one can thus discern the different connotations that the Greeks associate with the notion of charis. But it was not until the Hellenistic period, with the first Stoics and their Roman successors, especially Seneca—and to a lesser extent, Servius—that the Charites came to be linked closely with the logic of the gift and of “benefits,” to the extent of becoming their privileged allegory. As we have seen above, Seneca describes the gift in such a way that, regardless of the temporal logic, one feels like one is reading a pastiche of Mauss! One can also appreciate the rhetorical artifice with which Seneca equips himself to report the allegory of the three Graces: a very specific blend of false humility and genuine self-importance, authentic admiration and easy denigration—in short, an overall tone and commentary style where one easily identifies a stranger’s viewpoint, almost, even, an ethnographer’s.

How can we understand, Seneca wonders, why “Chrysippus, who is famous for his sophisticated intellectual analysis that gets to the heart of the truth, and who only says what is needed to get the job done and never uses more words than he needs in order to be understood—Chrysippus, too, filled his entire book with this nonsense, leaving himself only a little bit of room to discuss the actual process of giving, receiving, and returning benefits” (Seneca 2011: 21)? That is a question familiar to anthropologists, and not only concerning the Greeks. And the following remark also has not always been without an equivalent, in the ethnographic literature on other cultures or different scholarly traditions: “Just you look out for me, if anyone takes me to task for knocking Chrysippus off his pedestal—he is a great man, of course, but still he is a Greek and his overly subtle sharpness gets blunted and even turned against him” (ibid.: 22).

One problem with which anthropologists are also familiar is the impossibility of knowing—given how the allegory of the three Graces is reported to us—what the respective contributions were of Seneca, Hecaton, or Chrysippus, and what the sources of the latter author were. Hence, by a coincidence that is perhaps not meaningless, this astonishing description of the gift has an epistemological status reminiscent of our modern ethnographic descriptions (Sperber 1985). First, this wavering is due to the difficulty of associating the interpretation of the three Graces with an author or even with a precise culture—Greek or Latin. This comment on the gift was ultimately preserved thanks to the confrontation between two cultures. Moreover, it is because Seneca’s text shows us clearly that during the Hellenistic period, the logic of gift giving, this illusion of a “total phenomenon,” is, in fact, already merely a dreamed universe of sociability. Only the artifice of an allegory can introduce us to it, and this arouses Seneca’s suspicions as it is not good to believe “that frivolous fictions and arguments fit for old women might be able to prevent the most destructive possible turn of events: a universal cancellation of benefits” (Seneca 2011: 23).

It would therefore be an error of interpretation to mistake what is presented unambiguously as an allegory for what could be taken, outside the context of its utterance, as a representative description of a state of society; a society that would, moreover, remain in need of precise identification. By describing this allegory, Seneca does not pretend to do more than to submit to the rules of art and to the intellectual trends of his time, although not without a certain disdain, as some of his reflections, like this supposition, indicate: “But suppose that someone is so dedicated to the Greeks that he thinks these questions are vital.” (Seneca 2011: 21). However, Seneca underestimates the possibility that Chrysippus had used this allegory in the same way as he does, that is, precisely as an allegory and for purely allegorical purposes. So what is transmitted from one thinker to another, and Seneca is an integral part of this chain, is probably neither a belief nor a superstition. It is a particular way of thinking about exchange in society through select images that illustrate this conception, in order to have it understood and to make it more desirable. To think that scholars like Mauss may have thought, many centuries later, that one could find in ancient or exotic societies the same ideal of sociability because the desire for it was maintained while the allegorical character of this ideal had been lost, is not only an attractive hypothesis but also one that may have the advantage of being accurate, as I will try to show. But let us still continue our investigation by focusing this time, more precisely, on the iconography of the three Graces as such.

The iconography of the three Graces

Figure 03
Figure 3: The Three Graces, marble statue, Metropolitan Museum of Art, Roman copy of a Greek original from the second century BC.

It is convenient to distinguish two periods in the iconography of the three Graces, one “ancient” and the other “modern.”

The ancient period


One may find a few pieces of information on the first visual representations of the three Graces in the same way as one can find rare and scattered references to their cult and mythology.7 At the time of their debut, the three Graces had not yet started the round dance that would immortalize them. In some temples, in Orchomenos, for example, they were represented by rocks and at Elis by wooden statues in golden clothing, whose faces were made out of marble. But most importantly, in the basreliefs of the time, it seems the Charites were conventionally shown in profile or in a three-quarter view in three-quarters, holding hands and in a single file. Moreover, they were always represented draped in cloth.


It is during the Hellenistic period, first in Greece and then in Rome, that the three Graces acquired the iconography that would serve as their canonical representation in the modern era (Deonna 1930). From that moment on, the representations of the three Graces showed them dancing in a circle: the one in the center is represented from the back while the others are more or less facing the viewer, while also in a three-quarters orientation. As in the previous period, the three Graces hold each other by the arms or shoulders, and they often have symbolic attributes (fruits, vegetables) in hand that they circulate among themselves. Finally, following an aesthetic trend that certainly began in Aphrodite’s representations, the three Graces were increasingly shown naked. If one believes Pausanias, this new representation of the three Graces spread from one particular artistic work. There was, indeed, in the new configuration of these deities, an element of novelty that did not seem to be only related to their symbolism; it may have also been related to the masterly way their dance was depicted from an aesthetic point of view, thanks to an ingenious spatial arrangement. Moreover, considering that the representation of the three Graces required the artist to depict the female body from all angles in one single composition, we may sense there a reason that, regardless of any symbolic association, can explain the particular appeal of this divine group.

The modern period


The memory of the three Graces was never completely lost in the Middle Ages (Wind 1958: 44 n. 35). By contrast, it is interesting to notice how, concerning its iconography more specifically, contact was lost for several centuries with the depictions that were prominent during the Hellenistic period, and which only resurfaced during the Renaissance. Medieval artists, who did not draw inspiration from the models of the past, used their own traditions of representation to invent an image of the deities matching the textual descriptions they built upon in a rather literal manner. Jean Seznec, who studied in detail how images of ancient gods were preserved, modified, or transformed up until the Renaissance, provides an example regarding Remi’s commentary on Martianus, dating from the year 1100, which illustrates well such a process:

Yet as we examine them more carefully, we see that the artist has not indulged in pure fancy; on the contrary, he has taken pains to follow as carefully as possible the directions of a certain text. . . . The text tells him, for instance, that Apollo carries the three Graces in his hand. Remi has taken this detail from Macrobius, who had it from Pausanias: “Apollinis simulacra manu dextera Gratias gestant” (Saturnalia I, 17). What is called for here is thus a small replica of the group of the three Graces. Our draftsman, however, who has never seen anything of the sort, naïvely pictures a kind of bouquet out of which emerge three female busts. (Seznec [1940] 1961: 168)

There are other “deviant” illustrations of the three Graces, related to the same period in their iconography, for instance in the moralized manuscripts of Ovid, or in the Tarocchi (Tarot card games) (ibid.: 199).


Figure 04
Figure 4: A detail of Primavera, Sandro Boticelli, 1478/1482, Uffizi, Firenze.

The renewal of the Hellenistic representation occurred in the second half of the fifteenth century, and from then on it quickly became dominant. The discovery of an ancient sculpture depicting the three Graces in Siena played a crucial role, as it allowed artists to become acquainted with the old way of portraying them.

Figure 05
Figure 5: The Three Graces, Raphael, 1534/1535, Musée de Chantilly.

This discovery is known to us thanks to Raphael, who drew numerous sketches of the sculpture.

Subsequently, a host of great artists offered an interpretation of the allegory: among them were painters, printmakers, and sculptors of the Renaissance. Raphael, Raimondi, Correggio, Francesco del Cossa, Vasari, Dürer, Botticelli, Clodion, Germain Pilon, portrayed the Graces strictly following the ancient models, more or less, while Titian or Tintoretto gave freer interpretations. From then on, artists, like Rubens in the seventeenth century or Boucher in the eighteenth century, continually drew inspiration from the iconography of the three Graces, even as the meaning of the allegory was progressively lost.

Figure 06
Figure 6: The Three Graces, Lucas Cranach the Elder, 1531, Musée de Louvre, Paris.
Figure 07
Figure 7: The Three Graces, Rubens, 1639, Prado Museum, Madrid.

Perspectives on the gift

Representations of the Charites existed in ancient Greece. Similarly, the notion of charis had always been associated with the system of presentations typical of the gift. But it was really during the Hellenistic period, in Athens and in Rome, that an effort of conceptualization regarding the social implications of the gift took as its starting point the theme of the three Graces. As we have already seen briefly, it was based on a constant renewal of the iconography of these deities.

Seneca’s perspective

In Seneca’s commentary, which draws inspiration from Chrysippus, several points are noteworthy. Seneca refuses, for instance, to individualize each of the Charites. With a biting irony, he ridicules the way Greek poets and thinkers individually assigned them a name and a genealogy: “Each authority twists the interpretation of these names as it suits him, trying to reduce them to some orderly plan; in fact, though, Hesiod just assigned to the girls the names that he felt like giving them” (Seneca 2011: 21). By ignoring completely the specific identity of each of the Charites, Seneca does not purport to innovate, but he intends to give a more universal scope to this allegory. It is also why he does not seek to define what each of the Charites precisely stands for. Let us cite again this passage:

Some people advance the view that one of them stands for giving a benefit, one for receiving it, and one for returning it. Others hold that they represent three kinds of benefactors: those who confer benefits, those who return them, and those who accept benefits and return them at the same time. But no matter which of these interpretations you decide is true, what good does this specialized knowledge do for us? (Seneca 2011: 21).

Each of the Charites stands for a significant gesture of the gift relationship. But just like Lévi-Strauss when he analyzes the structure of exchange, Seneca grants less importance to the precise definition of each of these gestures than to the “orderly sequence” that constitutes itself through them. He insists that it is not enough to find a willingness to give. There must be simultaneously a willingness to receive and a willingness to return. Closer in this regard to Mauss than to Lévi-Strauss, for Seneca, taking this simultaneity into account does not imply any form of automatic complementarity among these three archetypical gestures. What really is needed, he suggests, is to find out the necessary conditions so that such synthesis may effectively take place: “Our job is to discuss benefits and to organize the topic which more than any other binds together human society” (ibid.: 22). For he has assigned to himself the duty to help “prevent the most destructive possible turn of events: a universal cancellation of benefits” (ibid.: 23). From the same perspective, the remarkable advantage of the allegory of the three Graces is to allow the representation and thereby the comprehension, though a series of symbols, of the necessary complementarity of the gestures involved in gift giving and the no less indispensable role succession implied by these gestures.

There was indeed a difficulty, both logical and iconographic, in attempting to represent, in a unified iconographic space, gestures that are otherwise defined by a temporal succession. It is the dynamic intuitively associated with dancing, and to round dances in particular in this case, that allowed the paradox to be resolved. For Chrysippus or Seneca, it was necessary that there be three Graces because it is the minimal number needed to represent the three fundamental gestures characteristic of the gift in one single moment, and thereby in a single representation. Hence, if one of the Graces is presented to us from the back and the two others in profile or facing us, this is only due to the constraint of solid geometry. However, space is not oriented here from the viewer’s position. The fact that the deities are facing him or her does not have a particular signification.8 This is a fundamental difference between the analysis offered by Seneca and the one proposed by Servius, which was also abundantly cited and analyzed by commentators after him.

Servius’ perspective

Servius wrote a famous gloss on the three Graces in the fourth century, that is, three centuries after Seneca. Contrarily to the latter, who still described the Graces wearing transparent dresses, Servius accounted for their nudity. But above all, it is the interpretation of the positions assigned to each of the Graces that changed with this author: “That one of them is pictured from the back while the two others face us is because for one benefit issuing from us two others are supposed to return” (Servius 2004: 1.720; English translation found in Wind 1958: 28).

Servius’ interpretation involves a new element. As we have seen, in ancient Greece the notion of charis was associated with growth and fertility. A symbol of this association was in fact preserved through the vegetal attributes often placed in the hands of the deities. But this reference disappears completely with Chrysippus and Seneca. In their conception of the “orderly sequence” of benefits, it is the gesture that matters, as well as the human relations that are established and affirmed with gift giving. But as Seneca repeatedly and precisely indicates, its signification is independent from the material reality of the objects that circulate from hand to hand by way of the gift. By contrast, for Servius and other authors after him like Boccaccio, the benefactor can legitimately expect the “benefit” associated with his gesture. The apparent gratuitousness of the gift does not show the disinterest of the giver since, on the contrary, a benefit in return is promised to him—a benefit delayed, but that will amount to twice what was given. It is also interesting to note that such a perspective, relying on the hope of future gains, is conveyed by a new interpretation of the representation of the three Graces. As we have seen before, in ancient Greece, the three deities were shown in profile and were advancing in line, parallel to the viewer. Then, during the Hellenistic period, they started dancing in a circle, and the viewer found him- or herself in a decentered position vis-à-vis them. But at the end of the Hellenistic period, the perspective changed again. The depiction of the three Graces became still. This time, the axis that extends from the viewer to the deities was explicitly taken into account. The giver’s function was identified with the central deity, who turns her back to the viewer. The two other Graces, situated on her two sides and facing, represent the recipients of the gift. In Chrysippus’ interpretation, the one relayed by Seneca, the logic of the gift was defined by three fundamental gestures—to give, to receive, to return—each embodied by one of the three deities. With Servius, however, the perspective changed. There are only two fundamental gestures left—to give and to return—and if one of them is symbolized by one deity and the second by two of them, it is only, we are told, to express the quantitative difference between what passes from hand to hand on the occasion of the gift and what compensates it in a second movement. Servius’ interpretation, however, although it is frequently cited, more recent and detailed, did not supplant Seneca’s, and these two perspectives were in fact maintained up until the Renaissance.

The perspective of the Renaissance

The theme of the three Graces was at its peak during the Renaissance, in particular among humanist authors like Marsilio Ficino or Picco dela Mirandola. The three deities were assigned a central position among allegorical figures in their new versions of Neoplatonism. Ficino, for instance, gave the title Charitum Ager (the Graces’ place) to his villa in Carregi, where he hosted a Neoplatonic academy. Picco dela Mirandola also took them as his emblem and engraved them on the obverse side of his personal medal. Three words accompany their representation on the medal—pulchirtudo, amor, voluptas—each word supposing to correspond to one of the Graces. This citation, taken from Wind (1958: 36), gives a general idea of the symbolism of the three Graces in the Renaissance:

While the triad of the Graces signified liberality to the Stoics, for the Neoplatonists it was a symbol of love, inviting celestial meditations. Since the Graces were described and pictured as attendants of Venus, it seemed reasonable to infer that they unfold her attributes: for it was the axiom of Platonic Theology that every god exerts his power in a triadic rhythm.

This new signification of the symbolism of the three Graces was associated, subsequently, with a renewed interpretation of their iconography. For Servius, the most important element was the confrontation between the central deity, identified with the act of gift giving, and the two others, symbolizing the return of the gift. During the Renaissance, however, the scene came to be interpreted differently. The central deity’s volte-face in relation to the viewer was now enhanced. It was understood as a way of turning away from earthly considerations to contemplate the spiritual realm.

This signification was reinforced by the way the two other deities were consequently reinterpreted. The accent was no longer placed on their symmetry in relation to the central deity, but, on the contrary, on the asymmetry that appears as soon as one neglects the overall position of the bodies to pay attention to the orientation of their faces and gaze. The reading of the scene was no longer static: the new interpretation had to be based on the dynamic that animates the scene. It was no longer about dance, as in the Stoic tradition, but about the reinvention of a particular genre:

All that we have to remember is that the bounty bestowed by the gods upon lower beings was conceived by the Neoplatonists as a kind of overflowing (emanatio), which produced a vivifying rapture or conversion (called by Ficino conversio, raptio or vivificatio) whereby the lower beings were drawn back to heaven and rejoined the gods (remeatio). The munificence of the gods having thus been unfolded in the triple rhythm of emanatio, raptio, and remeatio, it was possible to recognize in this sequence the divine model of what Seneca had defined as the circle of Grace: giving, accepting and returning. (Wind 1958: 38)

While Servius’ commentaries on the gift had privileged a rather materialistic interpretation of gratia and had essentially dealt with the nature of human transactions (their social and material quality), the Humanist reading of Plato instead placed the emphasis on a very specific and largely metaphysical notion of love—amor— which took preeminence in their interpretations over the previous connotations associated with charis and gratia:

If we further consider that all communion between mortals and gods was established, according to Plato, through the mediation of Love, it becomes clear why in Ficino’s and Pico’s system the entire Greek pantheon began to revolve around Venus and Amor. All the parts of the splendid machine (machinae membra), Ficino wrote, “are fastened to each other by a kind of mutual charity, so that it may justly be said that love is the perpetual knot and link of the universe: amor nodus perpetuus et copula mundi.” Although Venus remained one deity among others, and as such the bestower only of particular gifts, she defined, as it were, the universal system of exchange by which divine gifts are graciously circulated. The image of the Graces, linked by the knot of mutual charity (segnesque nodum solvere Gratiae), supplied a perfect figure to illustrate the dialectical rhythm of Ficino’s universe. (Wind 1958: 38–39)

The search for spirituality in this new interpretation of the three Graces found a superior aesthetic translation in the painting of Botticelli, Primavera, currently on display in Florence’s Uffizi gallery (see above, Figure 4). In this picture, the representation of the three Graces, wearing transparent dresses and holding hands, seems to be directly inspired by Seneca’s description. Indeed Botticelli was a keen illustrator of these kinds of allegories, inspired by the Neoplatonists, and which had become popular in the Medici circle.

However, this new proximity between the image and its ancient gloss must not delude us. There is some paradox in closely associating a spiritual quest with the rather carnal theme of the three Graces; and such a paradox was ever apparent, as Wind suggests, in the ancient use of this theme by the Stoic masters of antiquity. This also led Wind to postulate the possibility of an Epicurean origin for the allegoric use of the three Graces. But whether it is the case or not, there is no doubt that, during the Renaissance, not all artists felt obliged or inclined to follow the scholastic interpretation of the most fashionable allegories of the time, and to attenuate the carnal quality of the representation of the three Graces in the manner of Raphael (see above, Figure 5) or Botticelli. Correggio, as well as many painters of the Fontainebleau school, for example, did not hesitate to emphasize the theme’s intense sensuality.

This also explains what happened by the end of the Renaissance, when there was a progressive breaking of the link between the scholarly interpretation of philosophical origin and an iconography, largely inspired by it, but also loaded with sensuality. Then, the theme of the three Graces survived thanks to its iconographic appeal, but it also became more and more remote from any association—scholarly or not—with the idea of gift giving. The three Graces were still frequently represented dancing, but not necessarily in a circle, and regardless of the viewer’s position. However, they remained associated with Love, most of the time depicted conventionally as the flying baby (putto) escorting the Graces. One may then follow from the sixteenth to the eighteenth centuries—with Tintoretto, Rubens, Boucher, or Watteau—how this visual allegory progressively lost its conventional sense and also became more closely associated—Watteau is a good example of this transformation—with the courteous pleasures of Love.

An evolutionist temptation

From what has been seen above, it is possible to trace the guiding lines of an evolution of the theme of the three Graces. And one could argue that the history of such a theme reflects, beyond this specific example, an evolution of behaviors and attitudes that bears witness to a progressive change in civilization. This is what I would like to show briefly. But on this occasion, it is better to summarize first the various elements of our investigation.

In the beginning, we have the Greek notion of charis. This notion, as is often the case in ancient Greece, was represented by a group of minor deities—in this particular case, the Charites. Poets like Homer, Hesiod, or Pindar progressively assigned names to these deities, as well as genealogies and sometimes mythologies. There was also some worship of them, often in the context of the cult of more important divinities of the Pantheon: Zeus, Apollo, Hermes, Aphrodite. The Charites were often represented with the latter deities or placed in the entrance hall of their sanctuaries. It is possible to learn all this thanks to the compilations of archeologists, philologists, and historians of ancient Greece; and our sources are altogether numerous, diverse, and reliable.

Then, during the Hellenistic period, two singular events occurred. A new way of visualizing the three Graces replaced the ancient iconography. There is no absolute proof that this developed from an original model that would have served as a reference, but it is the most likely hypothesis. Moreover, at the same period, philosophers—in particular, Chrysippus, a master of ancient Stoicism—used the image of the three Graces as a central allegory in their conception of the gift, most certainly inspired by the wealth of connotations that had long been associated with the notion of charis and with the Charites. It is, then, plausible that the new iconography of the three Graces was born that way, under the influence of the gloss of these thinkers. But unfortunately there are no contemporary accounts of what happened at that time in Greece. What we know we owe to later sources of Roman origin. This is true both regarding representations of the three Graces—paintings and sculptures—and concerning the interpretation of this allegory proposed by ancient Stoics, which we are familiar with almost exclusively thanks to Seneca. Although we know that numerous intermediaries spread the teachings of ancient Stoicism in Rome, we have invariably lost track of them, as it is the case, for instance, for Hecate’s works, even though it is certain that they influenced Seneca.

Seneca did not conceal that his perspective on ancient Greek authors was both critical and selective. If he transmitted the allegory of the three Graces in On benefits, he said, priding himself on his offhandedness, it was without granting it much importance. But although he did not trust the immediacy and the validity of the teaching that could be drawn from an image or a myth, he paradoxically reinforced the significance of this allegory, somewhat in spite of himself. He did so, first, by endorsing the role of agent in this transmission: it seems that Seneca is the sole source accounting for its importance among the first Stoics. Second, and most importantly, he did so because he systematized its interpretation in accordance with his own conceptions regarding the practice of gifting in society. However, the unexpected consequence of his intervention in this domain was less to abolish the need to resort to an allegory than to guarantee the allegory’s subsequent success. Thus, the Charites became known in Rome not so much as other Greek gods and goddesses whose cult was absorbed by the Romans during the Hellenistic period, but, rather, from the beginning, as purely allegorical figures, allowing their use as such by artists, poets, and thinkers.

However, aside from this fragile thread by which the allegory of the three Graces was passed on and systematized from Athens to Rome, another form of acculturation, more diffuse but more significant, also occurred: it was the slow assimilation of the connotations associated with charis and the acclimation of the notion—rather than a mere translation—in Latin, through the term of gratia. The effect was all the more interesting as three streams of signification soon coexisted in the notion of gratia. The first involved the old connotations attached to charis, which can be found in the close association of gifting gestures, charm, and the interior beauty that the gesture confers. The second stream embraced connotations that were more specifically Roman, referring precisely to the procedures of gift giving; they characterized this practice by associating it with notions such as beneficence and gratuity. Finally, a third stream assembled the new Christian connotations associated at first with charis, then with gratia, and which eventually came to characterize the notion of Divine Grace.

This probably explains the success of the allegory of the three Graces in the Renaissance. Indeed, it was a period during which intellectual milieux took pleasure in rediscovering and deftly handling the different—and often paradoxical—connotations and associations associated with gratia or embodied by the three Graces. Offering an opportunity to reinterpret Seneca in a Neoplatonic sense, but also in a rather Epicurean manner in the way it illustrated a notion then at the heart of Christianity, the allegory of the three Graces offered Renaissance artists and humanists a perfect occasion to celebrate creatively the marriage of spirituality and sensuality. They all leaped at the opportunity.

After the Renaissance, it became more and more dubious to present the image of the naked and dancing Graces as a representation, even if only allegorical, of Divine Grace. However, in the seventeenth century, their representation was still judged sufficiently edifying for the Graces to preside, in Rubens’ painting, over the education of Marie de Medici (see above, Figure 7). But as the paintings of Boucher and Watteau can testify, from the eighteenth century onward, they mainly symbolized the charms of love. From this we can summarize how the fate of the three Graces can be interpreted over time:

  • A first period, associated with ancient Greece. The use of the notion of charis at that time is well reflected by the analyses of Gernet or Vernant, who interpret it as a key notion of the exchange, similar to the way Mauss used the Māori hau for his analysis of the sociocultural systems of Polynesia.
  • A second period, the Hellenistic era. Both in the evolution of the notion of gratia and in Seneca’s analysis, it becomes clear that the logic of the gift can no longer be interpreted as if it were involving the society as a whole. As Benveniste shows, the practice of gift giving is now identified to a “second circuit,” interrupting the normal circuit of exchange.
  • A third and last period, corresponding to modern times from the Renaissance to the present. The notion of gratia has been divided between a purely spiritual dimension, to which the Neoplatonic Humanists of the Renaissance gave a new philosophical dimension, and a more sensual dimension still present today, in particular in the iconographic representation of the three Graces. Here we have come full circle and the three Graces, after a centuries-long peripetias, have become a fully secularized iconographic theme, with no obvious connection henceforth with the gift.

As I will now show, there is little doubt that someone like Mauss would have interpreted such an evolution in the interpretation of the three Graces as a simple illustration of a more significant evolution, that is, the progressive marginalization of gift practices at the core of Western culture and society. The question, however, is to know if one is to take this sort of analysis at face value: is there not another way of interpreting what has been quickly summarized here about this allegory of the gift?

An opening to the world

As a conclusion to The gift, Marcel Mauss clearly advocates what he sees as a return to the archaic and ancient values that characterize a society based on the gift:

First of all, we return, as return we must, to habits of “aristocratic extravagance.” As is happening in English-speaking countries and so many other contemporary societies, whether made up of savages or the highly civilized, the rich must come back to considering themselves— freely and also by obligation—as the financial guardians of their fellow citizens. Among ancient civilizations, from which ours has sprung, some had a (debtors’) jubilee, others liturgies (of duty) such as choregies and trierarchies, and syussitia (meals in common), and the obligatory expenditure by the aedile and the consular dignitaries. We should return to laws of this kind. . . . Thus we can and must return to archaic society and to elements in it. We shall find in this reasons for life and action that are still prevalent in certain societies and numerous social classes: the joy of public giving; the pleasure in generous expenditure on the arts, in hospitality, and in the private and public festival. (Mauss [1925] 1990: 88–89)

One may notice once more that such a plea in favor of evergetism seems to echo some of Seneca’s reflections in On benefits.9 May one then conclude that Mauss’ text should be understood as a recent milestone in the line of commentaries about the practices of the gift, which has been ongoing in Western culture from Chrysippus to Hecaton, Seneca to Servius, Marsilio Ficino to Picco dela Mirandola? Some passages written by Mauss in The gift seem to confirm that one may effectively interpret the essay in this perspective. Does he not explain himself, for example, that “the themes of the gift, of the freedom and the obligation inherent in the gift, of generosity and self-interest that are linked in giving, are reappearing in French society, as a dominant motif too long forgotten”? (ibid.: 87). And in another passage, he explains as well that he is “[posing] once more, in different forms, questions that are old but ever new” (ibid.: 5).

What distinguishes Mauss, however, from his predecessors is the nature of the argumentation he uses in his essay in order to promote a return to a society based on the gift (an ambition that he quickly reduced, like his predecessors, to a modest plea in favor of euergetism). Until then, none of the authors that we have referred to had ever presented a society exclusively based on gift giving as something that may have actually existed. This is also why the state of “grace” that the existence of such a system would imply was usually presented either as a miraculous outcome that could only result from the interaction between men and deities, or as an allegory illustrated by poets and artists, from Goethe to Raphael. In the work of Mauss, however, one does not find a single reference to Chrysippus or Seneca, nor any allusion to the three Graces, to charis, gratia, to amor, to the notions of charm, grace, gratuity—all notions which have been traditionally at the core of such a problematic in Western culture. To argue in favor of a system of human relations founded on the practice of the gift, authors since antiquity had strived to show in what ways such a system was effectively desirable and they had presented it as a charmed ideal. With Mauss, however, the perspective changed completely, because he believed that ethnographic works brought the historical and sociological confirmation that the logic of gift giving actually constituted the milestone of all societies. Hence, what had generally been presented as an ideal altogether desirable and yet in many ways unrealistic by a long succession of philosophers, religious minds, skilled politicians, disillusioned preceptors, and idle courtiers finally acquired the remarkable dignity of being now considered as a “total phenomenon.”

One should not, however, consider, as a result, that the underlying logic of the problematic of the gift had completely changed with Mauss: in fact, one of his merits is that he offered a wider insight into the problem posed with the paradox situated in the “voluntary character of these total services, apparently free and disinterested but nevertheless constrained and self-interested” (Mauss [1925] 1990: 4). What fundamentally changed was the nature of the arguments used. While authors in the past only used rhetorical images and ancient literature in order to illustrate their opinion and defend their point of view, Mauss aimed to show that, as a prudent thinker, it was from social facts, past and present, and only using these, that he built his analyses. This is actually why there is an obvious complementarity between, and, we may say, an underlying logic explaining, Mauss’ absence of interest in the intellectual history which he unknowingly prolonged, and the way he used reconstitutions of sociological realities, even if they were at times fragile, to “demonstrate” that his theses were fully justified. Lévi-Strauss’ conception of the gift, by contrast, breaks away from this underlying logic.

From gift to exchange

The way in which Lévi-Strauss interpreted The gift in his famous introduction dated from 1950, is quite disconcerting. It is easy to perceive all that Mauss’ reflection owed to a scholarly tradition of which his colleagues and he were the direct heirs, even if they never acknowledged it. But, speaking about Mauss, it is more surprising, for instance, to affirm that “empirical observation finds not exchange, but only, as Mauss himself says, ‘three obligations: giving, receiving, returning’” (Lévi-Strauss [1950] 1987: 46). We have seen, on the contrary, that it was precisely his own cultural tradition that Mauss was following—perhaps unknowingly—when he analyzed the gift, after so many others, by way of these three obligations. Hence it is slightly paradoxical to reproach him, as Lévi-Strauss did, for not having followed completely his own principles in favor of a “New Zealand theory” when he did precisely the opposite.

Similarly, when Lévi-Strauss wrote a few pages later that “what happened in that essay, for the first time in the history of ethnological thinking, was that an effort was made to transcend empirical observation and to reach deeper realities” (ibid.: 38), we saw, quite to the contrary, that Mauss’ intellectual originality was precisely of an “empirical” order: it consisted in renewing an originally Western thematic by seeking its validation in new facts, gathered, in particular, from other societies.

For Mauss just as much as for the authors of antiquity, the crucial challenge was to find how to instigate a more desirable form of sociability, seemingly embodied in the practice of the gift. But the difference between Mauss and his predecessors is that instead of seeking to convince the reader of the importance of gift giving by emphasizing its moral, desirable, necessary, or rewarding character, he strived to show that such a set of attitudes constituted, in fact, an ineluctable and universal given in every society. Its apparent absence in our society was, as a matter of fact, merely a temporary aberration, and most certainly transitory. This means that just like Chrysippus, Seneca, Servius, or Marsilio Fino, Marcel Mauss in his time sought to convince that there was nothing worse for a society than the “universal cancellation of benefits.”

Nevertheless, the ambiguity in Mauss’ approach was to partly conflate the system of specific attitudes characterizing the gift and the very different reality of the exchanges that exist in various domains of social life. This confusion was not total since he considered the practice of gift giving and the attitudes it involved as “archaic forms of exchange,” in opposition to what he called “self-interested exchange,” which he viewed as characteristic of modern society. If we are to trust the meaning of words, we can only qualify a gift as a “gift” if a benefit in return is not systematically guaranteed. Of course, this does not imply that the gift must be disinterested. But what may be viewed as the risk or the inherent greatness of the gift is precisely that it is, essentially, a unidirectional gesture. Hence, the gift may, at most, be assimilated to an exchange if it is understood that it will be an exchange of a particular type, whose nature it is to be always problematic and without the guarantee of reciprocity. Ultimately, this was indeed how Mauss envisioned the problem posed by the analysis of the system of presentation of the gift.

Mauss was careful to preserve intact the particular thematic of the gift, while confusing it, in some sequences of his reasoning, with a primitive or fundamental modality of the exchange. Anthropology is not necessarily an exact science. Lévi-Strauss read Mauss literally and refused to see in the gift something other than the objective reality of an exchange. For him, the gift is solely a sophisticated and complex form, whose true nature will generally escape, for this very reason, the indigenous people themselves, as they only have a partial and subjective view of the social universe in which they are immersed.

Confronted with such an argument and without developing it further, it may be useful to note that the reasoning is not in fact that radically new, as it was always invoked by the partisans of the gift: of course, they sometimes brought up the idea that it was merely a disguised, and thereby advantageous, form of exchange. However, they obviously had some difficulties convincing anyone that a society existed in which such a conviction was universally shared. This is why they had to resort to the allegory and to a host of varied arguments to convince those to whom they spoke of the value of their analyses. The real innovation, which was transmitted from Mauss to Lévi-Strauss, was, firstly, to assume that primitive and archaic societies could be founded on systems of total presentations of the kind that were expressed through gift giving. But also, it was to do so while suggesting simultaneously that the functioning of such systems did not create any problem, if only possibly in the mind of rare indigenous theoreticians unable to understand the subtleties of their own societies. It must be noted as well that such a point of view could only be helped by the functionalism that has dominated most ethnographic research.

The problematic of the gift

From Mauss to Lévi-Strauss, therefore, one witnesses the short-lived resurgence of a problematic of the gift, the main interest of which was to open this field of reflection to a diversity of cultures and civilizations. But Lévi-Strauss, because he did not acknowledge, like Mauss, the essentially problematic nature of the gift, struggled on the contrary to assimilate the moments that characterized it as a complex form of exchange. It must be added that, unfortunately building upon what was the most arguable in Mauss’ theses in the framework of his own argumentation, Lévi-Strauss adopted a conception of primitive societies in which the gift, now classified among other categories of exchange, ends up losing its specificity entirely.

It is in such a context that we should consider what was lost when the reflection on the gift was cut off, under cover of scientism, from the ancient tradition to which this debate was related. Thus was ignored the most fundamental fact that there is no automatism in the gift and that there never was any.

The exchange or the gift: Illustration for a reflection

A society based on the gift: it is a dream and a utopia that inspired many reflections and beautiful works of art in the Western tradition. We will give a last illustration, which does not mark a clearly identifiable evolution in our societies but rather reveals the boundless imagination of people (of males in particular, in this case) and of their cultures. We started our investigation by the analysis of the word charis, which led us to refer to Vernant’s analyses concerning this notion:

In this connection, one of the essential aspects of Greek charis should be emphasized: charis is the divine power manifest in all aspects of gift giving and reciprocity (the round of generous liberality, the cordial exchange of gifts), which, in spite of all divisions, spins a web of reciprocal obligations, and one of the oldest of all functions of charis is a woman's giving herself to a man. (Vernant [1965] 2006: 163)

Figure 08
Figure 8: The Three Graces, Robert Delaunay, 1912, private collection.

Therefore is it difficult to conceive of a more total reversal in perspective than Lévi-Strauss’ when he wrote in a scandalous conclusion that “the emergence of symbolic thought must have required that women, like words, should be things that were exchanged” (Lévi-Strauss [1949] 1969: 496). By considering the subjects of the gift only as the objects of an exchange, the specificity of the notion of gift was lost entirely. In such a perspective, it is not surprising that the taking into account of notions such as charis, gratia, amor, and their equivalents in other cultures disappeared completely from anthropological considerations on alliance. Incidentally, knowing that Lévi-Strauss was a lover of figurative painting, and considering that his own work unknowingly transformed it into complex sets of abstract kinship schemes, we may wonder what he really thought of his personal contribution to the iconography of the three Graces.


Denis Vidal would like to thank Yves Goudineau, Michel Izard, Gérard Lenclud, Anne de Sales, Carlo Severi, and Alexis Tadié for their comments and their recommendation that this text be published.


Benveniste, Émile. (1969) 1973. Indo-European language and society. Translated by Elizabeth Palmer. London: Faber & Faber.

Boyer, Pascal 1986. “The ‘empty’ concepts of traditional thinking: A semantic and pragmatic description.” Man (N. S.) 21 (1): 50–64

Chantraine, Pierre. 1980. Dictionnaire étymologique de la langue grecque: Histoire des mots. Paris: Klincksieck.

Deonna, Waldemar 1930. “Le groupe des trois grâces et sa descendance.” Revue Archéologique XXXI: 274–332

Detienne, Marcel. (1972) 1994. The gardens of Adonis: Spices in Greek mythology. Princeton, NJ: Princeton University Press.

Firth, Raymond William. (1901) 1965. Primitive Polynesian economy. 2nd Edition. New York: Routledge and K. Paul.

Goldschmidt, Victor 2011. “The Stoic system and the idea of time (1953).” In French philosophy since 1945 : Problems, concepts, inventions, edited by Etienne Balibar, John Rajchman, and Anne Boyman, 153–58 New York: New Press.

Gsell, Stéphane. (1877) 1969. “Gratiae.” In Dictionnaire des antiquités grecques et romaines: D'après les textes et les monuments, edited by Charles Daremberg and Edmond Saglio, 16658–67. Paris: Hachette.

Lévi-Strauss, Claude. (1949) 1969. The elementary structures of kinship. Translated by James Harle Bell, John Richard von Sturmer, and Rodney Needham. Boston: Beacon Press.

———. (1950) 1987. Introduction to the work of Marcel Mauss. Translated by Felicity Baker. London: Taylor & Francis.

Mauss, Marcel. (1925) 1990. The gift: the form and reason for exchange in archaic societies. Translated by W. D. Halls. London: Routledge.

Moussy, Claude. 1966. Gratia et sa famille . . . Vol. 25. Paris: Presses Universitaires de France.

Sahlins, Marshall. (1972) 2004. Stone Age economics. London: Routledge.

Seneca, Lucius Annaeus. 1961. Des beinfaits. Paris: Les Belles Lettres.

———. 2011. On benefits. Translated by Miriam T. Griffin and Brad Inwood. Chicago: University of Chicago Press.

Servius. 2004. Servius’ commentary on Book four of Virgil’s Aeneid: An annotated translation. Translated by Christopher Michael McDonough, Richard E. Prior, and Mark Stansbury. Wauconda, IL: Bolchazy-Carducci Publishers.

Seznec, Jean. (1940) 1961. The survival of the pagan gods: The mythological tradition and its place in Renaissance humanism and art. Translated by Barbara F. Sessions. New York: Harper.

Sperber, Dan. 1985. On anthropological knowledge: Three essays. Cambridge: Cambridge University Press.

Vernant, Jean-Pierre. 1965. Mythe et pensée chez les Grecs: études de psychologie historique. Paris: F. Maspéro.

———. (1965) 2006. Myth and thought among the Greeks. Translated by Janet Lloyd and Jeff Fort. New York: Zone Books.

Veyne, Paul. (1976) 1990. Bread and circuses: Historical sociology and political pluralism. Translated by Oswyn Murray and Brian Pearce. London: Allen Lane/Penguin Press.

Wind, Edgar. 1958. Pagan mysteries in the Renaissance. London: Faber & Faber.

Les trois Grâces ou l'allégorie du don: Contribution à l'histoire d'une idée en anthropologie

Résumé : ‘L’essai sur le don’ de Marcel Mauss est une des textes les plus fameux de toute l'anthropologie ; certainement aussi un de ceux qui ont suscité les débats les plus nombreux et les plus variés dans cette discipline. Paradoxalement, ces débats ont assez peu porté sur la tradition culturelle dans laquelle ce texte s'inscrivait de manière manifeste. Je voudrais montrer ainsi que les réflexions de Marcel Mauss sur le don prolongent, tout en le modifiant, un très ancien courant de réflexion sur le don, basé sur un ensemble de notions apparentées ayant joué un rôle crucial dans la culture européenne : charis, gratia, grâce. Je voudrais montrer aussi, le rôle central joué dans ce même contexte par une allégorie et par un thème iconographique qui s'est perpétué pratiquement sans discontinuer depuis la Grèce archaîque : celui des Trois Grâces.

Denis VIDAL is a social anthropologist whose research is primarily in India. He is a Senior Research Fellow at IRD (URMIS-Paris Diderot) and an Associate Professor at the Museum of the Quai Branly and at EHESS (Paris). His actual research focuses on visual culture and on new ways of approaching technology from an anthropological perspective.

Denis Vidal
École des Haute Études en Sciences Sociales
190–198 avenue de France
75244 Paris cedex 13


Publisher’s note: This article is a translation of Vidal, Denis. 1991. “Les trois Grâces ou l'allégorie du don: Contribution à l'histoire d'une idée en anthropologie,” Gradhiva 9: 30–47.

1. Translated from the French edition of Seneca, Des bienfaits (1961: 7), as cited by Denis Vidal in the original version of this article. Our choice to not use the 2011 English translation for this epigraph results from the rhetorical features of the French translation, which offer “a series of questions” as a translation of the Latin passage. The English translation of the same passage (see Seneca 2011: 20) is not rendered in the same interrogative voice.—Trans.

2. Seneca is not cited once in the exhaustive name index of the complete works of Marcel Mauss published by Les éditions de Minuit in France.

3. Paul Huvelin (1873–1924) was a legal historian and contributor to Émile Durkheim’s L’Année Sociologique. He taught Roman law at the University of Lyon, and notably wrote a famous thesis, “Droit individuel et magie,” exploring the relation between magic and the law.

4. Georges Ambroise Davy (1883–1976) was a French legal sociologist and, like Huvelin, a contributor to L’Année Sociologique. He was a close disciple of Marcel Mauss.

5. This is the translator’s English translation of Vernant’s endnote from the original French edition; it seems that this endnote was lost during the translation of the work into English. —Trans.

6. One may refer to Gsell ([1877] 1969) for further analysis on this matter, and a list of references on the Charites and their cult.

7. For a survey of the iconographic corpus of that period, see Gsell (1969: 1664–65).

8. One should note here the substantial compatibility between Seneca’s analysis in that particular context and the Stoic doctrine regarding space and time, presented in Goldschmidt’s work (2011).

9. To learn more on evergetism in Hellenistic antiquity, both in Greece and Rome, see Veyne ([1976] 1990).