What is a gift?

This work is licensed under the Creative Commons | © Alain Testart. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 DOI: http://dx.doi.org/10.14318/hau3.1.026


What is a gift?

Alain TESTART, Centre national de la recherche scientifique

Translated from the French by
Susan Emanuel and Lorraine Perlman

This article introduces the subject of defining a gift in an ironic way, imagining Martian anthropologists conducting fieldwork among terrestrial human beings, hearing so many times the verb “to give” (for instance “giving” money to pay in a shop, or to pay taxes) and concluding to the importance of “the gift” in modern human society. Such a fiction serves to show that “to give” is not the same as “to make a gift.” It can also be shown from Latin etymology. Finally, the article ends by way of a precise definition of what a gift is, and insists on the notion of requirement to pay. What marks an exchange is the requirement to repay with an appropriate counterpart of each of the transfers. With a gift, on the contrary, the counterpart (counter-gift) may be given, but cannot be required.

Keywords: gift, counter-gift, exchange, Mauss, obligation

On the difference between “give” and “gift”

In English, does “to give” mean the same thing as “to make a gift”? In French, donner and faire un don do not mean the same thing. Let’s consider the followingexample drawn from a common experience of daily life in France.

I went to the butcher but did not know what to buy, and I ended up saying to him: “Give me a steak!” And he answered: “I'm going to give you some sirloin, it’s very good today!” When I went to pay, I had no change and I told him: “Can I give you a 50 Euro bill?” Then we went on to talk about taxes, a favorite subject of all small businessmen who always consider them excessive, and the butcher summed things up by saying: “Just think of how much we give them!”

This small example shows that the verb “give” was used four times, but not once in the context of a gift. Of course I never thought that my butcher was offering me a gift of a steak, nor did he think he was giving away his sirloin, any more than I was giving him a gift when I held out my banknote to pay, or that it was a matter of gifts when we talked about taxes.

“Giving” therefore is not making a gift; “to give” is not to make a donation. This is our first observation, and it is indisputable. However, there is also no doubt that if we confuse these two meanings, we create serious misunderstandings. We will return to this observation.

First of all, that such a great difference exists between the two terms might seem strange, since in Romance languages it seems they might have the same root, that of don-. But this linguistic inference is incorrect, since, our verb “donner” actually derives from a confusion between two distinct Latin verbs. The first, donare, signifies to make a gift, to make a donation. The second, dare, has a much more general meaning: it signifies to put into someone’s hands, to bestow, grant, concede. Dare actually applies to any movement of goods, any change of hands, any translation or transfer (the latter is the general term we are going to employ) and is just as suitable to use for a gift as for any other transfer.

The verb “to give” therefore can be applied to any kind of transfer (exchange, taxation, transmission through inheritance, etc.), not just to a gift. The noun “gift” (don), on the contrary, designates a particular mode of transfer, a mode of transfer that possesses a particular quality that we are trying to characterize. An initial approximation might be reached through the fact that a gift is a matter of a free or gratuitous act. It is obvious that receiving something as a gift (when it is a gratuitous act on the part of the person who is providing it) is the opposite of having to pay for receiving it (when, from the perspective of the person who furnishes it, it is an act that must be paid for); we may now summarize these first reflections in a little diagram:

It remains to be determined which term should fill the position we have left blank (namely, how we should label “the opposite of gift”), but for the time being, let’s explore some of the simple ideas we have mastered.

Mistakes by Martian anthropologists

Let us imagine that anthropologists from another planet—let us say Mars—were present at the butcher’s shop and witnessed our conversation. Imagine, too, that these anthropologists, like human anthropologists from earth, have little taste for law and economics or for studying arid treatises, and prefer the direct and living observation of what they call “a slice of life.” Imagine that they carefully note down everything said around them. The words I exchanged with my butcher were quite banal, and they could easily have heard similar exchanges with many other customers. It is likely they would have noted the recurrent use of the word “give.” Ignorant of its Latin etymology and therefore assimilating “give” and “make a gift,” the Martian anthropologists might certainly conclude that “the gift” is very import-ant in modern French society. “You can see,” they would say, “commercial prin-ciples are unknown in this society, but gift-giving practices are widespread: the butcher gives his cuts of meat and the customers make counter-gifts. Even the principle of taxation is unknown, since taxpayers don't mind giving to the revenue authorities, etc.”

The mistakes of the Martians are clear, as are the misunderstandings they might infer/promote about the nature of our society. Now, I believe that social anthro-pology, from Marcel Mauss in the 1920s to our day, has made the same mistakes relating to preliterate societies and has correspondingly misunderstood the nature of these societies.

More specifically, I maintain that this anthropology:

  1. has always confused gift and giving
  2. because it has never had a clear definition of what a gift was,
  3. and consequently has tended to overestimate the importance of the gift in preliterate societies.

As an example of the first error, we can cite the recurrent usage in French and English anthropology of the terms “givers” and “takers” of women in generalized exchange,1 and the concomitant idea (often implicit, occasionally explicit) that this involves a “gift” of women. This is curious because most examples of generalized exchange from Burma to eastern Indonesia—for instance, the Kachin—are associa-ted with particularly large marriage payments. In order to marry, considerable goods must be furnished, such as several buffalo; a man may even commit himself to be a “buffalo” if he does not have one (meaning he becomes a beast of burden), he can be prosecuted for debts his grandfather contracted on his wedding day, or else be reduced to slavery for not being able to pay.2 These practices seem scarcely compatible with an ambiance that (usually) involves gifts and presents.

The opposite of the gift

In order to fill the empty box in our last diagram, let us take a simple social example. Children, after school, still play marbles. We adults, even if we still remember the charm of the iridescent sparkle of glass balls, attach no great importance to marbles. But once upon a time, we eagerly collected them, played games to win them, and even tried to acquire them by other means. And as the children that we all were, we already made a clear difference between a gift and an exchange. “I'm giving you that one!” meant—with no ambiguity possible—that the child engaged in this generous gesture was giving up his marble without wanting to recuperate one in exchange. We knew very well that giving (making a gift) was the opposite of exchange.

To be precise: “I am giving it to you” means that the person who gives (i.e., the donor) gives it without any need for me to give him back anything at all. In contrast, the expression “I am swapping it for that one” (identical in form to “I am giving it to you for that one”) means that the one who gives (and who is definitely not a donor) only gives his marble if I give him mine. The reader can see that in these two different situations the word “give” has the different meanings of donare and dare. This we already know; and here we are making another point: in the exchange, my marble that I should “give” in order to obtain the one offered to me plays exactly the same role as the money I “give” to the butcher to obtain the sirloin he offered me. Handing over my marble is the payment necessary to obtain the coveted marble. The exchange is in all respects an act that is paid for. The word “pay” by no means supposes the existence of money: one may pay in kind, as once was done for taxes. And in the exchange of marbles, the marble that is relinquished constitutes the payment for the marble obtained. The same is true in any barter economy: the ceding of one good to obtain another represents the payment for this other good.

We may now complete our diagram by writing:

Thus, exchange is the opposite of gift, just as paying for something is the opposite of getting something for nothing. We can now see that if Martian anthropology confuses the two meanings of “give,” it likewise confuses gift and exchange. As does social anthropology.

Marcel Mauss was guilty of this in his excessively famous “Essay on the gift” (and, actually, should have been called “Essay on giving,” which would have eliminated the ambiguity). In numerous passages, he maintains that certain “primitive” transactions relate to both gift and exchange, and he introduces the expression “gift-exchange.”3 But this mixes up two things as one: the confusion is either in the heads of the so-called primitive peoples or else it is in the heads of the anthropologists. Mauss’ thesis—and it is explicit—is that the confusion is in the heads of the primitive peoples, and that they are, in this respect, more primitive than children who easily make a distinction between exchange and gift. We may trace this thesis directly back to the old assimilation (common to both anthropology and psychoanalysis): primitive = childish = pathological. It also conforms exactly to Lucien Lévy-Bruhl’s view that “a primitive mentality” is characterized by a confusion of ideas.4 Such notions were widely shared at the time but today are strongly rejected as ethnocentric, if not racist. But when Lévy-Bruhl indignantly described “the primitive” and Mauss became incensed about it, we forget the intellectual atmosphere in which they worked. This atmosphere was governed by two great ideas that oriented all that era’s anthropology (as it does a part of anthropology today): primitive societies are supposedly simple, much simpler than ours are, and as a consequence, their intellectual categories (legal, economic, social, etc.) seem to confuse things that we can distinguish/analyze.

My own position, by contrast, is that preliterate societies are extremely complex (and, on many points, more complex than ours) and that their conceptual and linguistic structures are highly subtle (often more so than ours). There would be no argument about this if we examined the great complexity of these societies’ vocabularies relating to the gift and to exchange.5 In other words, I think that the confusion lies not in primitive structures (intellectual or social), but in the heads of anthropologists.

We should now progress to the definition of gift, a concept that we were beginning to discern without having yet defined it. Our example of marble games leads us to think that the question of counterpart (what one needs or doesn't need to offer in order to obtain something) may be central.

First considerations on the counterpart

By counterpart, we usually refer to what comes back in return for the first transfer. We call “counter-transfer” the movement (in the sense we spoke of the “move-ments of goods”) that delivers this counterpart.

There is no exchange without counterpart: each of the goods exchanged is found to be the counterpart of the other. But counterpart may also exist in the activity of making a gift. The child who received a marble from a playmate may, some time later, make him a present of one of his own marbles: this is a way of thanking and of maintaining a friendly relationship. This, then, is an instance of a counter-gift. There may be a counterpart (counter-transfer) in a gift as in an exchange, and therefore it is not the existence or absence of a counterpart that differentiates the two.

Nor is it that gift and counter-gift may be deferred in time. In our example this is the case, but the recipient might wish to make a counter-gift on the spot, which does not change the fact that he has indeed received a gift and offered a counter-gift. Similarly, an exchange may be immediate or deferred.

We might pursue this parallel by saying that the regularity that seems to charact-erize exchanges (in capitalism a regularity interrupted only by crises and serial bankruptcies) may also arise in gift-giving. Very well brought-up people may unfailingly respond to a received gift with a gift offered, or to an invitation received with an invitation offered. But, I think we see well enough that it is not the kinetic aspect—an expression referring to everything concerning the movement of goods, the possibility of a counterpart, its regularity, etc.—which enables us to make a distinction between the different modes of transfer. It must be something else. What is it?

To answer this question, we need to explore “exchange.”

The three meanings of the term “exchange”

Dictionaries (Le petit Robert and the OED) tell us that the noun “exchange” possesses various meanings, the first two of which correspond to two meanings of the verb “to exchange.” The first meaning is economic: exchange is an exchange of goods, and to exchange is “to cede in return for a counterpart.” The second meaning (said to be by analogy, and in existence in both French and English in the seventeenth century) concerns any reciprocal communication: to exchange is to address and receive reciprocally. Thus we speak of exchanges of smiles, courtesies, blows—the verb applies to all these situations. The third meaning concerns only the noun and in French was originally (1865) biological: “the passage (in both directions) and circulation of substances between the cell and the external milieu.” This meaning easily extends to physics when we speak of heat or fluid exchange. But we see it could be applied to all other domains, for example if speaking of the exchange of cars between the city and the country. Of course, I am not saying these meanings are different because they arise in different domains (economy, linguistics, biology) but that they differ in a more profound way because they have different formal properties. Let’s start with the least rich, the third.

This is a purely mechanical (or kinetic) exchange with no link between cause and effect. It refers only to a movement in both directions through a permeable partition that nevertheless permits distinguishing between an interior and an exterior. This is why the term can be applied equally well to the cell as to an inanimate body, to an organic body as to a social ensemble like a city. It refers to a simple action in two directions, a reciprocal action or interchange (in the very general sense that one speaks of a reciprocal action or force, without intentionality.)

In contrast, the second meaning is inseparable from the notion of intentionality. The dictionary speaks of “address.” In an exchange of smiles, each person addresses a smile to the other. Without this address, there would be only two smiling humans; but without one to the other; there would not be an exchange of smiles. Addressing signifies something, aiming at a certain goal, or hoping to do so. If I address a smile to someone, it is because I hope for a response that may take the form of a smile or something else. The address presupposes the hope of a response. Of course, in an exchange of blows, I do not give one in the hope of receiving one in return, but the other person gives me one in response to the one I gave him. Indeed, the common denominator of these exchanges (smiles, blows, courtesies) lies in the idea of response. An act (a smile, a blow, a courtesy) responds to a precedent. There is no exchange except to the extent that there is a response that conforms, is suitable, and in proportion to the demand implicit in the first address. There has to be a certain balance between the two. And there is an order between the two acts, one following the other, and possibly occasioning a third: the exchange is a series of acts that respond to each other; the preceding one is always the cause of the following one.

Thus there is much more in exchange in the second sense than in the third, with its mechanical or biological sense. Everything in the third is also in the second: a similar reciprocal displacement, we might say, means that something from each goes toward the other. But the second sense of the word “exchange” has more intentionality, signification, address, an idea of response, and causality.

Now look at the first meaning in the dictionary’s listing, an economic sense, an exchange of goods. This is also the fundamental meaning, since all the other definitions are impoverishments of this rich and strong sense, and this is also the strict sense, since it is more limited than the other two. Everything in the third is found here: reciprocal displacement of goods between two actors. And everything in the second is also there: intention, obviously, since I cede my good only to signify to the other that I expect to acquire his; equally present are address and response; and finally, causality, since each of these transfers elicits the other. There is much more as well. As the dictionary says, to exchange is “to cede in return for a counterpart.” In the exchange of goods, one does not cede one’s good unless the other cedes his. One does not cede it except on condition that the other does likewise. One does not cede it except by reason of the engagement of the partner to cede his. And this is what differs completely from the second meaning, since in an “exchange” of words, I do not address words to someone only on condition that he speaks to me, any more than I address a smile to someone solely on condition that he smiles back to me.

To summarize: the third meaning has little relevance for us since it designates simple reciprocity and says nothing more: a thing leaves A and goes to B, while from B something goes in the opposite direction to A. This is simple displacement in two directions, in a purely kinetic sense, a simple reciprocity of actions taking place in space. The meaning is purely kinetic. The second meaning is reciprocity plus the idea of response, which corresponds to reciprocate: this reciprocity is intentional, desired, signifying. But the idea of reciprocity in itself does not take us much farther. The first meaning is quite different: it designates an insitution that is human, complex, and perfectly specific.

It is human because, as Adam Smith once said, we have never seen animals proposing an exchange of goods with each other. They do exchange in the second sense, but not the first. It is complex because there is a prior engagement or understanding, an “agreement of wills,” as jurists say about contracts, which must precede the acts of transfer. And it is perfectly specific, finally, because the characteristics that we have just identified relative to exchange in the first sense distinguish it from other gift transfers; one does not give something on condition that the other promises to give you back an adequate counter-gift.

We are making progress in understanding exchange, and gift as well. (It is always useful—even necessary—to understand non-A in order to understand A). Let us pause for a moment and return to our Martians.

More Martian mistakes

The Martian anthropologists who are still listening to our conversations and recording them in their ethnographic notes, have not failed to observe how frequently the terms “exchange” and “to exchange” appear in Western languages. They most likely conclude that our societies are based on exchange, as most other human societies probably are. Failing to notice that exchange in the third sense was often found simultaneously in biology, in ecology, as well as in physics, they will think they have discovered a great truth of social science by disclosing that human societies are founded on exchange. But their statement is a flat and banal one, as it is true in any sphere—physics, biology, and the social world—and does not define any one of them (reciprocal action is one of the fundamental principles of mechanics). Perhaps they can claim that the differences in political economy, linguistics, and social anthropology are based solely on the fact that the first deals with the exchange of goods, the second with the exchange of words, and the last with the exchange of women among men.

Since they do not make a distinction between the different meanings of the world “exchange,” they see no essential difference separating the exchange of words from the exchange of goods. And after noting the expression “exchange of gifts,” they would probably also say that there is no great difference between exchange and gift—for the simple reason that they would not have seen that the term “exchange” in this expression implies the second sense of the word (we exchange presents like we exchange courtesies) whereas in the expression “exchange of goods” the first sense is implied. Finally, their failure to differentiate between exchange and gift allows the Martians to assert that a statement about the importance of the gift in Western society (their first proposition, remember), is equivalent to this one about the importance of exchange. And thus a complete confusion in Martian anthropology will be established.

This confusion has reigned over structuralist anthropology for the last fifty years. And it will reign over any anthropology and any discipline that employs words from everyday language, with all the polysemy and ambiguity that characterize them, with no concern for defining scientific concepts—and defining them rigor-ously—in order to be scientific.

The result of all this is that if we are interested in the transfer of goods, we will use the term “exchange” only in the first sense, the restricted and economic sense, and the only one that is specific to institutions and the social sciences. The distinctive feature implied in this exchange is the necessity for a counterpart, and this must be understood as simultaneously the condition, cause, and purpose of the exchange. This counterpart is also obligatory. What does that mean?

Obligation and requirement to pay back

If there is one aspect for which Mauss’ formula of “three obligations”6 can be reproached, it is that it appears empty of content since the very idea of obligation is coextensive with social life as a whole. In fact there is no social relation without obligation. Whether it is a kinship relation, an amorous commitment, or the relation of the citizen to the state, obligations are found everywhere. But they are not all of the same nature. Obligation makes sense only if the type of obligation is specified—without this, it is almost tautological. Jurists have written comprehensive treatises on obligations, anthropologists never. They think the notion of obligation is self-evident. It is not.

I cannot claim to fill this void here, or even to tackle the subject in a significant way, but we must at least make the distinction between a moral obligation and a legal obligation. A moral obligation merely pricks your conscience. But a legally recognized obligation gives rights to the one who has this obligation over you; and it gives him the means of action. We might say that the exchange includes a legal obligation to furnish the counterpart, whereas the gift does not carry such an obligation, or, at most, a moral obligation. We might even correlate this with the notion of sanction, in which we similarly distinguish between moral and legal sanctions. This already goes much farther than Mauss (and all of subsequent anthropology) went, but it carries difficulties linked to the controversial question of how to define legality in preliterate societies—a subject to which we return in chapter 2. But the idea of legality itself shows us another path.

What is specific to legal obligation is that payment is required. Payment may be obtained by all legitimate means that exist in a society, including by violence, from the moment it is conducted in forms recognized as legitimate. In a state society, the obligation may be requited by resorting to legal procedures (when agents of the state are involved in the recovery); in a stateless society, payment is obtained through violence of creditors (or right-holders) who may resort to a vendetta, this being the normal way of carrying out one’s own justice. If the notion of compulsion to pay appears clear and generally applicable, we can now see how it sheds light on the question of gift and exchange.

First element in the definition of the gift

Suppose that you have seen my fountain pen and find it a very nice one.

“Monsieur, you have a beautiful fountain pen!” you say.

And I say to you: “I am giving it to you!” You protest, you are too kind, there is no reason to, but finally you accept. You offer effusive thanks and you depart with the fountain pen.


A few days later, we meet and I demand that you give me the hundred euros that represent the value of the pen. My demand astonishes you, but I insist, maintaining that, since you accepted my pen, you owe me this sum and that you must pay me. What would you say? Simply: “But then it wasn't a gift!”

It is clear where the problem is: the fact that I require a counterpart demolishes the idea that I made you a gift.

If several days later, as a gesture of appreciation, you make me a present— valued at one hundred euros, whether more or less doesn't matter—no one would deny that because of this act, I have given you a present. The existence of the counter-gift does not cancel the nature of gift in my original gesture.

If, in fact, I had given you my pen in the hope that you would render me some service, this would still make the gesture a gift. It would have been a gesture with an ulterior motive, a selfish one certainly, but a gift nevertheless. The fact of expecting a counter-gift does not cancel the nature of my gesture as a gift.

If I later asked you to lend me one hundred euros, on the pretext that I was short of money and reminded you of the gift I gave you a few days earlier, you might find it hard to refuse me, but you could not say that my original gesture was not a gift. It certainly was self-interested and my behavior was deplorable, but still nobody would deny that this gesture was a gift. The fact of soliciting a counter-gift does not annul the nature of my act as a gift.

What does annul it is that I claim something in return and that I assert my right to do so. Either it is legitimate for me to claim something and so it is not a gift, or else I have no legitimacy to demand anything at all and it is indeed a gift.

We may conclude that the gift is the transfer of a good that implies the renunciation of any right over this good, as well as of any right that might issue from this transfer, in particular something requiring a counterpart.

The idea of gift contains the notion of abandoning. The donor abandons a good, any right over this good, as well as any right that might emanate from its transfer.

In the exchange, on the contrary, whoever exchanges something has a right to require a counterpart—and it is the right itself that defines the exchange.

We may now say precisely in what way Mauss’ expression “exchange of gifts” is (strictly speaking) contradictory: it is because exchange is founded on the right to demand a counterpart, whereas the gift is a gift only because the right to require it is renounced.

The key point of this initial approach—which is not yet a definition—is a matter of right: is it legitimate to claim, to require? This key point resides neither in the movement of goods (as does the matter of the counterpart—what we have called the kinetic aspect of transfers), nor in the actors’ psychology (generous or selfish), nor in their behavior (solicitation of the donor).

Definition completed

This first definition is still inadequate. It allows us to distinguish between gift and exchange, but it does not distinguish between gift and a third type of transfer.

Consider the matter of fines, or whatever we are obliged to pay as damages and interest, plus the reparations growing out of a misdeed or a responsibility we are liable for. What is the nature of such transfers? They are obviously not gifts. And they are hardly like exchanges since, although one sees a certain reciprocity at work in the idea of reparation as in the classic exchange, there is a difference of scale enters between the two.7

What marks the exchange is the requirement to repay each transfer: although the reparation may be obligatory, the damage that it is supposed to compensate for is not. Nor is there, in the reparation, the kind of reciprocal causality that characterizes the exchange when each of the exchanging partners agreeing to cede a good causes the other to agree to cede his own. In reparation, the offense is the cause of the reparation, but the reparation is not reciprocally the cause of the fault. We are always wrong to confuse exchange and reciprocity, for if exchange is indeed a form of reciprocity, not all reciprocal acts are exchanges. Finally, the reciprocity associated with reparation is not the same as with exchange. Reparation, being neither gift nor exchange, is a transfer of a different kind, what we will call T3T, a transfer of the third type.8 I will not list here all its characteristics, but say only that it is a transfer that results from an irrevocable obligation (the wronged party having the right to demand reparation from the person who caused him the wrong) and without counterpart. Taxes are another example of a T3T transfer.

Now this is our problem: the definition which we use to characterize the gift applies equally to the T3T and so does not allow a distinction between gift and T3T, which is absurd. Consequently we have to refine our definition of gift so it cannot be applied to what is obviously not a gift.

Could we say that the gift is voluntary and that no T3T (compensation or tax) is? Could we say that the gift is elective, while taxation is coerced and reparation is compulsory? This would introduce endless and difficult discussions, similar to the ones we wanted to avoid regarding obligation. Now everybody has known situations in which we feel “obliged” to give a gift. For example, we do not feel “totally free” to not make a gift when it is a matter of a tip or a charitable donation. Where does the line of demarcation lie? Could we say that what characterizes the gift is that, even if you are not totally free not to give, you still give “as much as you want,” in the hallowed phrase? But note that the tipping rate is known and customary; and when there is a collection you know that you cannot really give what you want, for there a tacit minimum below which you cannot go beneath. In truth, the notion of freedom covers an infinitely varied range—a range similar to what obligation covers. There are moral pressures that limit freedom and they are quite different from legal constraints, which rely on force.

A short parenthesis is appropriate here in order to dispel a false solution that consists of observing: by paying the tax, I do not renounce any right as a result of this gesture, since by this gesture I now have the right to no longer be subject to the demand to pay the tax. A payment acquits a debt; it either creates a right or cancels the obligation that others held over us. We say the same of an individual who caused someone a wrong—once he pays for it. On this basis, we might claim that our first characterization of the gift suffices to differentiate it from T3T. But, in fact, these considerations are not general enough and are valid only for the taxpayer and the author of a reparable wrong. They are not valid for a serf—taillable et corvéable à merci, a French medieval expression meaning “subject to unlimited exploit-ation.” The serf’s tax (payment in money) is, like the corvée (payment in services), a transfer of the third kind. But by furnishing them, the serf acquires no right. For these transfers, our preceding definition of the gift would apply, and this is contrary to reason.

A simple solution is found by reconsidering the necessity for payment. The tax must be paid. Every tax bill states as much. The compensation payment, the serf’s tax and corvée (labor), are similarly mandatory. A gift is not. No one can require a gift of you. A gift can be expected, solicited, etc., but not required without losing its character as a gift. At this point we return to the discussion we had about the counterpart.

To conclude, we will say that a gift is a transfer of a good that

  1. implies the renunciation of any right over this good as well as of any right that might arise from this transfer, in particular that of requiring anything by way of counterpart; and
  2. that is not itself required.

[In the rest of this book (Testart 2007),] these definitions will enable me to demonstrate that of the two great institutions of the preliterate world that are considered to be founded on the gift—the potlatch and the kula—only the former actually is based on gift-giving; the latter is not (see chapters 3 and 7). These definitions will also permit us to perceive that a form of commercial exchange that has been mistakenly conceived in terms of gift and counter-gift is in fact an exchange—but not a mercantile one (chapter 5). In a general way, these precise definitions will serve as the basis for a critique of the classic theses of social anthropology concerned with the role and importance of the gift in preliterate societies.


Lévy-Bruhl, Lucien. 1923. Primitive mentality. Translated by Lilian Ada Clare. London: G. Allen.

———. 1928. The “soul” of the primitive. Translated by Lilian Ada Clare. London: G. Allen.

Malinowski, Bronislaw. 1963. Les argonautes du Pacifique occidental. Translated by Simonne and André Devyver. Paris: Gallimard.

Mauss, Marcel. 1950. “Essai sur le don: Forme et raison des l’échange dans les sociétés archaïques.” In Sociologie et anthropologie, edited by Marcel Mauss, 145–279. Paris: Presses Universitaires de France.

———. 1990. The gift: The form and reason for exchange in archaic societies. Translated by Wilfred Douglas Halls. London: Routlege.

Smith, Marian W. 1940. The Puyallup-Nisqually. New York: Columbia University Press.

Testart, Alain. 1997. “Les trois modes de transfert.” Gradhiva 21: 39–58.

Testart, Alain, Valérie Lécrivain, and Nicolas Govoroff. 2002. “Le prix de la fiancée: Richesse et dépendance dans les sociétés traditionnelles.” La Recherche 354: 34–40.

Testart, Alain. 2007. Critique du don. Paris: Syllepse.

Qu'est-ce qu'un don?

Résumé : Cet article introduit la question de la définition du don sur le mode humoristique en imaginant des anthropologues venus de la planète Mars et entendant si souvent le verbe « donner » (par exemple « donner » de l'argent pour payer dans une boutique, ou pour désigner tout ce que l'on « donne » en impôts) qu'ils en concluent à l'importance du don dans notre société actuelle. Cette fiction sert à montrer que « donner » n'est pas la même chose que « faire un don ». On peut aussi montrer cela en remontant à l’étymologie latine. L'article se termine par une définition précise du don, insistant sur la notion d'exigibilité. Ce qui caractérise l’échange est que l'on est en droit d'exiger la contrepartie de chacun des transferts. Dans le don, au contraire, la contrepartie (le contre-don) peut être donné, mais pas exigé.

Alain TESTART is Director (emeritus) of Research at the CNRS in Social Anthropology and has published fourteen books on Australian Aborigines, Hunters and Gatherers, religion, slavery, money, and other topics. He is currently working with Prehistorian Archaeologists on the interpretation of deposits, Neo-lithic and Palaeolithic iconography, and the conception of social evolution.

Laboratoire d'anthropologie sociale
CNRS, EHESS and Collège de France
52 rue du cardinal Lemoine
75005 Paris


Publisher’s note: This article is an edited translation of Chapter One of Testart, Alain. 2007. Critique du don. Paris: Syllepse. We have retained all references to other chapters.

1. The expression used by Lévi-Strauss and structuralist anthropology that is inspired by a particular form of matrimonial system.

2. Bibliographic references in Testart, et al. (2002: 34).

3. Elsewhere in the essay, Mauss speaks of a “salary-gift” and of “gifts to the chief” that are “tributes.”

4. What was implicit in Lévy-Bruhl (in Primitive mentality, English translation of 1923, and The “soul” of the primitive, English translation of 1928) is explicit in Mauss’s essay, where he speaks of “the slightly puerile legal language of Trobriand Islanders” and of their vocabulary complicated “by a strange inaptitude to divide and define.”

5. For the Trobriand Islanders, see Malinowski (1963: 238–52); for the Puyallup (Salish south coasts of Puget Sound), see Smith (1940: 146–50); see also the chapter on the kula. Note that it is rare for observers to take the trouble to study in detail the vocabulary of these societies; partial and incomplete notes always give the impression of simplicity, when the fault really lies in deficient ethnographic observation.

6. See the systematic critique in chapter 4.

7. This point will be developed in the following chapter (Three situations compared).

8. The presentation of this type transfer in my 1997 article, “Les trois modes de transfert,” was incomplete because reparation was not envisaged there, and so it was partially erroneous due to the overly exclusive link I made with a dependent relationship (which does not exist in the case of reparation), a topic taken up in the following chapter.