2. How should we define our terms?

To be understood and to be persuasive scholars must be clear about what words mean when we use them (Hodgson, Reference Hodgson2019: 1). While we can mostly rely on a shared linguistic background with our readers, part of the task of the theorist is to expand our understandings beyond, and sometimes in opposition to, lay beliefs. Sometimes this entails using words in ways that may not match the linguistic backgrounds of most readers. This need not create difficulty: all specialisms employ their own vocabularies, the terms of which may be quite clearly understood by their users. The social sciences, however, face a more challenging problem because our terms relate to everyday social phenomena that are both described by terms in common use and influenced by our use of those terms; indeed institutionalised practices such as gift giving are governed through the words we use to delineate and endorse them. If we are to speak of these phenomena it is natural to feel that we should do so ‘as much as possible in familiar everyday language … that is widely accessible to non-specialists’ (Riggs, Reference Riggs, Sonneveld and Loening1993: 195),Footnote ¹ and yet this raises the thorny question of the relation between the everyday and the theoretical meanings of the same term.

Consider, for example, the following definition of the gift: ‘Any exchange of goods and services with no guarantee of recompense in order to create, nourish, or recreate social bonds between people’ (Godbout and Caillé, Reference Godbout and Caillé1998: 20). Here, the role of the phrase ‘in order to create, nourish, or recreate social bonds between people’ is to build an element of theory into the definition, with the effect of excluding those transfers of goods and services that do not nourish social bonds. But as Hodgson has argued, building theory into definitions can generate disagreement and impair understanding (Hodgson, Reference Hodgson2019: 9, 13), in this case because there is no such exclusion from the everyday concept of the gift: an anonymous gift to an animal welfare charity, for example, is still a gift according to common English usage but it may make no significant contribution to nourishing social bonds, and the donor may have no intention that it should (cf. Laidlaw, Reference Laidlaw and Osteen2002: 49).

We can understand this definitional issue in terms of the relation between stipulative, descriptive (or lexical), and explicative definitions. A stipulative definition ‘imparts a meaning to the defined term, and involves no commitment that the assigned meaning agrees with prior uses (if any) of the term’ (Gupta, Reference Gupta2014: section 1.3); a descriptive or lexical definition, by contrast, aims to describe the existing usage of a term; and an explicative definition ‘aims to respect some central uses of a term but is stipulative on others’ (Gupta, Reference Gupta2014: section 1.5). Godbout and Caillé’s definition is an explicative one: it respects the basic everyday sense of the gift as something that is transferred ‘with no guarantee of recompense’ but also seeks to be stipulative about a different aspect of the gift: that it must nourish social bonds.

Such interference with generally accepted understandings of a term is risky. An author might make some claim about gifts, for example, meaning it only to apply to that subset of gifts stipulated in her explicative definition, but be read as meaning it to apply to all gifts, as commonly understood, generating unnecessary criticism and conflict. As Riggs puts it, ‘the author's stipulated meaning may [easily be] lost while readers struggle vainly to make sense out of a text in which well-known words do not carry their normal connotations’ (Riggs, Reference Riggs, Sonneveld and Loening1993: 196; and see Gerring, Reference Gerring1999: 361; Hodgson, Reference Hodgson2019: 5).

Although this is a good argument for respecting common usage in our definitions, uncritically accepting descriptive definitions of terms that are to be used for theoretical purposes can also cause problems. One set of problems arises from the widespread polysemia of everyday words (Gerring and Barresi, Reference Gerring, Barresi, Collier and Gerring2009: 243). The same word, that is, may have multiple meanings. If, for example, we look up the verb ‘give’ in the Oxford English Dictionary, we find no less than 23 different meanings, some of which have up to four variations (Oxford University Press, Reference Thompson1996: 417). The social study of giving is concerned with one of these senses or perhaps one group of them, in which, to put it simply ‘To give, is to hand over something to somebody free of charge’ (Testart, Reference Testart, James and Allen1998: 97). Yet some of the dictionary meanings are quite unrelated to this, such as ‘utter (gave a shriek)’, or even diametrically opposed, notably ‘make over in exchange or payment (gave him £30 for the bicycle)’ (Oxford University Press, Reference Thompson1996: 417). While we may use the same word to express these various meanings, these are not all the same concept, and we must clearly exclude them from the study of giving in the sense of handing things over free of charge. For these purposes at least, a fully descriptive definition of giving is thoroughly inappropriate.Footnote ² As Hodgson puts it ‘in these circumstances, reliance on lexical appeals is problematic and some stipulation is unavoidable’ (Hodgson, Reference Hodgson2019: 5). In the case of giving, then, we will need an explicative definition, one that excludes all these other senses so that we can focus on one concept, referring to a specific set of social practices or institutions, those in which things are handed over free of charge.

Perhaps this is one reason why scholars have often spoken not of giving but of the gift. Superficially, these two words appear equivalent in the sense that they are merely the verb and noun forms used to speak about the same phenomenon: when one gives, at least in the sense of handing things over free of charge, what one gives is a gift, and a gift is something that is given. Perhaps the two words were once related only in this way, but today they are not. When one gives a shriek, what is given is not a gift. When one gives £30 in payment for a bicycle, what is given is not a gift.

One might therefore try to avoid the polysemic meanings of giving by adopting gift as the theoretical term instead. However, although the range of meanings attached to gift is smaller, it still does not converge precisely on the practices that interest social theorists of the gift. The most obvious exception is the sense of gift as referring to ‘a natural ability or talent’ (Oxford University Press, Reference Thompson1996: 416). Once it was widely believed that natural abilities were given by God. Few of us now accept this, yet we continue to refer to such abilities as gifts. Talent is not given, yet we call it a gift. So even if we were to adopt gift as a term instead of giving, we would still need to define it explicatively, excluding the case of natural talents.Footnote ³

There seems no viable alternative, then, but to use explicative definitions for the central theoretical terms in the study of giving. This does not mean that we can never adopt terms that refer to more restricted ranges of meaning: my objection is to using everyday terms in this way. There is a simple technique we may use to generate more restricted concepts: the qualified term. Thus, for example, Godbout and Caillé may wish to work with a concept that refers only to those gifts that nourish social bonds. Rather than redefining gift to mean this, they could stipulatively define a qualified term, such as bonding gift, to represent this concept. This has the merit of engaging the reader's understanding of the common meaning of gift while signposting that we are using a restricted version of the concept and indicating the nature of that restriction. This technique is also useful for distinguishing between different kinds of giving. Any plausible definition of the gift must respect what Silber has called ‘the astonishing variety and multiplicity of forms and functions of gift-giving’ (Silber, Reference Silber, James and Allen1998: 146). We will therefore need terms that distinguish different kinds of gifts within the larger class of gifts-in-general, even when that larger class has already been explicatively bounded.

3. Maussian definitions of the gift

As Mirowski has said, ‘all roads to the gift lead back to Marcel Mauss’ (Mirowski, Reference Mirowski, Cullenberg, Amariglio and Ruccio2001: 438). Mauss's analysis of the gift, based on anthropological evidence, was intended as a critique of utilitarianism, showing how societies could be integrated and coordinated through non-market forms of economy. While some of his followers have focused on developing this critique of utilitarianism, the Maussian tradition is diverse, and others have sought to apply his analysis of giving practices to a wider or even universal range of societies. Mauss himself never defines the gift (Testart, Reference Testart, James and Allen1998: 97), but he constantly ties it to three features: to reciprocation, to an obligation to give and to the maintenance of social relations. At least the first two of these refer to ‘systems of established and prevalent social rules that structure social interactions’, and thus to institutions (Hodgson, Reference Hodgson2006: 2). Although his own argument was founded on anthropological examples, some of his interpreters have concluded that these theoretical claims are definitional of gifts in general – including contemporary gifts – and have thus failed to recognise that institutions are particular to specific communities (Hodgson, Reference Hodgson2006: 3).Footnote ⁴ While giving persists in the contemporary world, the forms of its institutions are not the same as those discussed by Mauss. Let us consider the three features highlighted by Mauss in turn.

First, Mauss constantly stresses the ‘obligation to reciprocate’, indeed the phrase appears in the title of the first chapter of his book (Mauss, Reference Mauss and Halls2002: 10). In Mauss, however, this is a phrase that is always connected to particular societies, to particular practices (e.g. the potlatch: Mauss, Reference Mauss and Halls2002: 11, 53), or at most to groups of societies sharing a similar institutional context (Mauss, Reference Mauss and Halls2002: 23). Though he extends his analysis rather vaguely to contemporary society in his closing chapter, he never says explicitly that all gifts everywhere necessarily involve an obligation to reciprocate. Yet, influenced by Mauss, Gregory, for example, has argued that ‘reciprocity, along with the inalienation and the obligation to return it implies, is the defining characteristic of a gift’ (Gregory, Reference Gregory1997: 65). Though various thinkers have sought to make this true by redefining reciprocity to cover an increasingly wide range of referents, in doing so they have taken this term far beyond the corresponding common usage. As Graeber says, ‘if one has already decided that all relations are based on reciprocity, one can always define the term so broadly as to make it true’ (Graeber, Reference Graeber2011: 405). But in doing so writers not only take the concept of reciprocity beyond the range of ordinary understandings, they also deprive Mauss's argument of its original meaning. Mauss's point is that gifts result in an equivalent or superior return to the original giver (though sometimes after a cycle of intermediate transfers). But to preserve the idea that gifts must be reciprocated, reciprocity has been redefined to include, for example, what Sahlins calls ‘generalized reciprocity’ (Sahlins, Reference Sahlins1974: 199): an institution in which a gift from A to B must lead to a gift from B to C, but without any necessity that the circuit will continue until A receives a return (cf. Moody, Reference Moody2008). Once reciprocity is redefined in this way the claim that gifts must be reciprocated no longer corresponds with Mauss's original argument.

Furthermore Mauss's original argument, if it is to be read as meaning that all gifts must be reciprocated in something like the normal everyday sense (as it is, for example, in Douglas, Reference Douglas and Mauss2002), is ‘manifestly false’ (Testart, Reference Testart, James and Allen1998: 98) (also see Elder-Vass, Reference Elder-Vass2015). The examples of unreciprocated gift institutions can be multiplied endlessly: presents to children (Cheal, Reference Cheal1988: 8, 57; Godbout and Caillé, Reference Godbout and Caillé1998: 24, 41), blood donation (Silber, Reference Silber, James and Allen1998: 138; Titmuss, Reference Titmuss1997), organ donation (Healy, Reference Healy2006), gifts to beggars (Testart, Reference Testart, James and Allen1998: 98), gifts to charities and political causes (Lainer-Vos, Reference Lainer-Vos2013: 30), and the various forms of the Indian institution of dan (Laidlaw, Reference Laidlaw and Osteen2002; Parry, Reference Parry1986), for example.Footnote ⁵ Even when return gifts actually do occur, this does not necessarily mean that it makes sense to understand them as obligatory reciprocation, as indeed other Maussians have pointed out (Godbout, Reference Godbout and Vandevelde2000: 30, 40; Godbout and Caillé, Reference Godbout and Caillé1998: 95).

Mauss also stresses the obligation to give, again in the context of specific types of society (Mauss, Reference Mauss and Halls2002: 16–17). But Carrier, for example, generalises this into a fundamental distinction between gifts and commodities: ‘From the Maussian perspective … gift transactions are obligatory: the parties are obligated to give, receive and repay. On the other hand, commodity transactions are voluntary’ (Carrier, Reference Carrier1995: 11). When Carrier goes on to define gifts, again invoking Mauss, he insists that ‘gift exchange is (1) the obligatory transfer of (2) inalienable objects or services between (3) mutually obligated and related transactors’ (Carrier, Reference Carrier1995: 21). All three elements of Carrier's definition are problematic, but here I focus only on the first: What could it mean to say that a transfer is obligatory?

Testart, in a brilliant analysis of concepts of obligation employed in the Maussian tradition, argues that it covers a broad range of degrees of obligatoriness. At the weakest end of the scale, there are gifts that involve no obligation whatsoever, either to give or to reciprocate, as for example when one gives a coin to a beggar (Testart, Reference Testart, James and Allen1998: 98). If someone invites me to dinner, I may feel obliged to return the invitation at some point, but ‘there is no sanction attached to this “obligation to reciprocate”, which is merely a feeling’ (Testart, Reference Testart, James and Allen1998: 99). The sense of obligation to return the Native American potlatch was rather stronger (although only for some of its recipients) in that failure could lead to some loss of prestige, but there was no sanction exercisable against such failure (Testart, Reference Testart, James and Allen1998: 99–101). The Trobriand Islands institution of kula by contrast explicitly allowed someone who was entitled to a certain type of object from a certain person to seize it if it was not given, and in such cases we can say that reciprocation is truly obligatory, although there was no sanction against a defaulter who did not possess the required object, and Testart questions whether gift is really an appropriate term here: perhaps this is more a loan than a gift (Testart, Reference Testart, James and Allen1998: 101–102). There are even stronger senses of obligation attached to modern debts, in that bailiffs can break into our houses and take anything they can find to the value of the debt, but enforceable modern debts are neither gifts nor required reciprocations of gifts (Testart, Reference Testart, James and Allen1998: 102). Clearly, obligations to give and reciprocate vary considerably between institutional contexts; they may be rather weaker than many other social obligations, and sometimes do not exist at all. Tying our definition of the gift to obligation thus excludes many good examples of gifts but it also draws in cases that common usage might exclude.

The Maussian theoretical claim that has been most widely adopted into definitions of the gift, however, is the idea that gifts are necessarily moves in developing or sustaining social relationships. We have seen this already in Godbout and Caillé’s definition: ‘Any exchange of goods and services with no guarantee of recompense in order to create, nourish, or recreate social bonds between people is a gift’ (Godbout and Caillé, Reference Godbout and Caillé1998: 20). Similarly Douglas argues that ‘A gift that does nothing to enhance solidarity is a contradiction’ (Douglas, Reference Douglas and Mauss2002: x). And Caillé (Reference Caillé and Vandevelde2000: 47) gives an interesting double definition:

1. (1) A sociological definition: every allowance of goods and services made without a guarantee of return, with a view to creating, maintaining or regenerating the social bond. In the relationship of the gift, the bond is more important than the good.

2. (2) A general definition: every action or allowance made without expectation, guarantee or certainty of return, and which, because of this fact, has a dimension of ‘gratuity’.

Caillé’s sociological definition reiterates the definitional link to social relations. As we have seen, this link cannot stand if our concept of the gift is to correspond to any sense of the term in common usage, since there are significant cases of the gift that do not make any direct contributions to nourishing social bonds, such as certain gifts to charity, and some of the many gifts that occur over the Internet (Elder-Vass, Reference Elder-Vass, Garnett, Lewis and Ealy2014a). If I share the MP3 file for a song over a peer-to-peer service, for example, it may be anonymously downloaded by people who neither know nor care that I contributed part of their download. While they may feel a general sense of community with other users there is no one-to-one relationship building (Giesler, Reference Giesler2006). Caillé’s second definition is far closer to the core common usage but using the same term for both seems destined to create confusion: when Caillé goes on to talk of the gift later in the paper, how are we to know which of these definitions is intended? Nor is it clear whether Caillé recognises that these two definitions of the gift have different extensions (i.e. they describe different though overlapping sets of phenomena), or whether he believes that he can tie the extension of the general definition to the theoretical claims in his sociological definition. It would be far clearer to use his two definitions to define two separate terms, such as bonding gift and gift. Stripped of this definitional confusion, we could then make use of these to address the interesting and important question of when, how, and how widely gifts play the potentially crucial social role of nourishing bonds between people: the question, in other words, of which gifts are bonding gifts (cf. Gudeman, Reference Gudeman, Cullenberg, Amariglio and Ruccio2001: 460–461).

Although Mauss himself does not explicitly draw reciprocity, obligation and social bonding into a definition of the gift, his work has contributed to the belief that the social sciences should operate with a concept of the gift that is disconnected from common contemporary understandings. Mauss constructs his argument by lifting the concept of the gift from a modern context and applying it to a range of practices in quite different societies. He achieves an impact on his readers by playing on their supposedly naïve assumptions about the nature of the gift and counterposing the reality of non-modern ‘gift’ practices to those assumptions (Laidlaw, Reference Laidlaw and Osteen2002: 57). The gift, we are to understand, is not voluntary but obligatory, not altruistic but designed to generate a return, and not an individualistic act but a compulsory act of constructing social solidarity. A strange inversion has been performed here: the term gift has been uprooted from a modern context, applied to a number of practices that in at least some cases do not conform with the mainstream modern concept of the gift, then Mauss and his followers seek to generalise from this particular mix of practices to a universal theory of the gift (Danby, Reference Danby2002: 21; Silber, Reference Silber, James and Allen1998: 137), while ignoring the modern giving practices from which the term was first drawn and criticising views of the gift that reflect those practices. Some of them, as we have seen, even go so far as to define the gift in ways that are derived from non-modern practices while neglecting modern ones.

Mauss himself writes perceptively about a range of important practices, he universalises less, and more tentatively, than some of those who have been influenced by his work, and he performs the enormous service of showing us that economies can operate on models other than the market one. But to insist, as some other scholars have done, on defining the contemporary gift in the terms he employs in his anthropological analyses excludes a vast range of actual gifts, of things handed over free of charge, from the scope of theories of giving.

4. A new definition of the gift

How, then, should we define the gift? Every significant word in a good definition performs a boundary function: it includes some cases within the extension of the term being defined, and excludes others. I suggest that a good definition should tend to include cases understood as giving in common usage, and to exclude cases that are not, though consistency of meaning is also important. There will, however, be difficult cases that remind us that the boundaries of reference of a term are often, and perhaps inevitably, vague (Hodgson, Reference Hodgson2019: 2).

I propose, then, to define a gift as a voluntary transfer of goods or services from one party (the donor) to another (the recipient) that does not require a compulsory transfer in return, and giving as the process of voluntarily transferring goods or services to another party without any requirement for a compulsory transfer in return. Let us consider each of the significant terms in turn.

First, the gift is voluntary. It is a transfer that is not compelled by law, threat or expected sanction. This contradicts the Maussian view of gifts criticised in the previous section. Gifts may be expected in certain circumstances: I am normatively expected, for example, to give a card and a present to various members of my family on their birthdays, but there is no compulsion for me to do so.

Still, there is a grey area at this boundary: institutions are generated and sustained by a mix of endorsement and enforcement that varies across a range of degrees of sanction. Some may be so strongly enforced that they border on being compulsory. In a society, for example, where women are expected to do all domestic and caring work for their partners and children it may sometimes be problematic whether this is a voluntary act and thus a gift. Perhaps some women enter such relationships voluntarily, and are happy to perform this role. Others may be under stronger compulsion, in the most extreme cases from domestic violence, and in such cases their work in the house must be regarded as forced labour rather than a gift.

One implication is that the payment of taxes is not generally a gift. Paying taxes is compelled by law and thus, as Godbout puts it, taxes are ‘the exact opposite of a gift – a gift that's imposed is not a gift’ (Godbout and Caillé, Reference Godbout and Caillé1998: 60). Similarly, the benefits conferred by the welfare state are not a gift, for a similar reason: once the processes of political debate, power struggle and legislation have led to the creation of a given form of benefit, the state is legally obliged to deliver it to those who qualify for it. As Zelizer puts it, these benefits are not gifts but entitlements (Reference Zelizer1996). One implication is that gifts are only one variety of what Boulding calls grants – there are types of one-way transfer that are not gifts (Boulding, Reference Boulding1973).

The element of voluntariness also implies that whether something is a gift depends on the intention of the donor. Mirowski illustrates the issue nicely with the case of the dropped banknote: if I unknowingly drop a banknote and someone picks it up, that is not a gift: ‘the gift must have an intentional aspect to be a gift’ (Mirowski, Reference Mirowski, Cullenberg, Amariglio and Ruccio2001: 447). This intentional aspect may however be rather minimal: the donor must intend that the good or service be transferred to some recipient without a return being required, but need neither intend that it be seen as a gift, nor intend for example that a particular party be the recipient. Thus, for example, the MP3 file posted by the BitTorrent user is a gift, but a gift to unknown recipients.

Second, the gift is a transfer, but what is transferred when we make a gift is not so much a thing as a right to a benefit or set of benefits. I may pass you my mobile phone to show you a message I have received, but in doing so I have not given you my mobile phone, because I have not transferred any rights: as the legal property owner I have the right to control who has physical possession of my phone and the right to control who sees my messages and I have simply exercised those rights in authorising you to hold my phone and look at a message (Hodgson, Reference Hodgson2015). If I were to give you a mobile phone, I would be transferring a bundle of rights over that phone to you and relinquishing a bundle of rights of my own in the process. One way of putting this point is to say that ‘To be given as a gift an object must be alienable, in the dual sense that the donor has the right to renounce ownership of it and that the recipient has the right to possess it as his or her own property’ (Cheal, Reference Cheal1988: 10).

Gifts of services are somewhat different. If someone, for example, mows the lawn of an elderly neighbour without any expectation of reward, they have given the neighbour a service. What is transferred is not an alienable object, but rather the product of the donor's labour: in this case, the transformation of the lawn to a tidier state. The donor has not transferred rights over something but rather improved something over which the recipient already held rights. Gifts of labour also raise the question of power: when young people take unpaid internships, for example, they give a gift of labour to an employer whose control over access to work creates the context in which internships become necessary to those pursuing certain sorts of careers.

Third, what is transferred is goods or services. The previous paragraphs explained both, but what of the exclusions implicit in this formulation? Can things be given that are neither goods nor services? Adloff and Mau, for example, argue that ‘people are permanently giving each other non-tangible ‘things’ in everyday face-to-face situations’ (Adloff and Mau, Reference Adloff and Mau2006: 110): interactional ‘gifts’ of attention, interest, concern, support, responsiveness, and the like (cf. Martelaere, Reference Martelaere and Vandevelde2000: 214). Godbout and Caillé even speak of the words we use in conversation as ‘given and returned’, and if we define the gift this widely, it becomes ‘the basis for a system, a system that is nothing less than the social system as a whole’ (Godbout and Caillé, Reference Godbout and Caillé1998: 13). Donati goes even further when he says that every relationship depends upon a gift, our willingness to ‘recognize the Other as having dignity, give your trust and put yourself in his/her place’ (Donati, Reference Donati2003: 269). Trusting or respecting others seems to count as a gift for Donati, even before it leads to communicative interaction. These are somewhat imperialistic understandings of the gift, which risk turning everything we do into a gift and thus rob the concept of its specificity. I doubt that we should count interactions as gifts, though this raises a challenging question. Perhaps some of the things that are sold as services provide primarily interactional benefits, and if such services can be sold, could they also be given? But in common usage, smiles and supportive comments are not generally considered gifts, and although one might say that A gave B a supportive comment, I take it that here give is being used in one of its other polysemic senses rather than the core sense we are concerned with here. Perhaps the argument that a gift is a transfer of a right to a benefit is enough to distinguish between interactions and services that can reasonably be considered gifts, but this issue requires more attention than I can give it here.

Fourth, the transfer occurs from one party to another. I assume that either of the parties may be an individual, a group of individuals (e.g. a household) or a corporate body. We may therefore regard, for example, a transfer from an individual to a charity as a gift, and a transfer from a charity to an individual. By contrast, Caillé’s general definition leaves out parties as well as transfers. It thus leaves out something essential to our understanding of the gift: on his definition, even cycling down the road would count as a gift, irrespective of whether some other party benefitted from the action, although I assume that this is an inadvertent slip on Caillé’s part.

Fifth, there is no requirement for a compulsory transfer in return. Much of the literature on the gift concerns reciprocity, and although I have argued strongly that reciprocity is not a defining feature of the gift, it is clear that some gifts are reciprocated, and that reciprocation is an important feature of some gift practices. If I give you a birthday present, it is entirely possible that you will later give one to me, and the fact that I gave you one may influence you to reciprocate, but the original gift does not create a binding requirement to do so. By the time my birthday comes around, we may have drifted apart, or you may be too poor to buy me a gift, or too ill to buy me a gift, or just too mean to buy me a gift, and there is no sanction that compels you to do so. In Testart's terms, the return is not obligatory.

Gifts can therefore be contrasted with commodity transactions, since in such transactions a transfer from one party to another creates a legal obligation for the second party to make a transfer in the reverse direction. The crucial issue is the obligatory requirement to make a return, and not the question of whether a return actually occurs. When we give a gift, a return may occur, but there is no binding requirement that it must. When we buy a commodity, we immediately generate a legal requirement to pay for it, even if that payment is itself deferred. In this respect, gifts and commodities are diametrically opposed. Once again my argument converges with Testart's: if we obscure this distinction by insisting that the gift incurs an obligation to reciprocate ‘we blur all the difference between gift and exchange’ (Testart, Reference Testart, James and Allen1998: 104).

Such oppositions are often features of definitions, but must be treated carefully: we cannot define all the features of the gift by contrast with the commodity, since they also have some features in common. Both, for example, involve a transfer of goods or services from one party to another. These similarities create the possibility of cases that have some of the characteristics of both gifts and commodities: hybrid borderline cases. Consider the case of buying a lottery ticket. In many European countries one condition of licensing a lottery is that part of the proceeds must be disbursed to charitable causes, so that buying a lottery ticket has something of the character of making a charitable donation. If a fixed percentage of the price paid for a lottery ticket is donated to charity, is the purchaser of a lottery ticket buying a commodity or making a gift? On the one hand, they are engaged in a monetary exchange, in which monetary payment is legally required for the ticket, and so in this sense this is a commodity. On the other hand, they are doing so in the knowledge that a portion of the price is a donation to charity. It is possible to argue that the purchase is both a commodity transaction and (in part) a gift, at one and the same time: a hybrid of two types. We should not define the gift so strictly as to exclude the possibility of such hybrids.

One feature of cases like this is that it may be difficult to decide whether a transfer is a gift or a commodity, or both or indeed neither. Consider the case of New York nightclub owners who give models free entry, free drinks and free meals at their nightclubs (McClain and Mears, Reference McClain and Mears2012). These are gifts, in that the models strictly speaking are required to do nothing in return. But inherent in the act of this acceptance, the models deliver a benefit to the nightclubs: their presence at the venue tends to attract high-paying customers. Does that also impart something of the character of a commodity to these benefits? One could make a case that these transfers are neither gift nor commodity but something in between or perhaps a mix of both.

Finally, if we confine our attention to what is said in a definition, we will miss one last essential feature of all definitions: that which is absent from them, that which is left out. Sometimes distinctions may be left out by accident, sometimes they may be left out because the author has not considered cases that would make clear the need for a certain distinction, but sometimes distinctions may be left out deliberately, and mine includes at least one deliberate omission: it does not say that a gift cannot benefit the donor. Those who give may receive praise, their social status may rise, they may gain pleasure from helping others, and they may even receive material benefits as a consequence of their giving, such as open source programmers who give their labour freely to programming projects but as a result acquire expertise that might open the door to lucrative consulting careers (Lerner and Tirole, Reference Lerner and Tirole2002: 213, 217–220). They may even be motivated by the desire for such benefits, for example unpaid interns who give their labour in order to build their CVs and their career prospects. But none of this prevents the transfers they make from being gifts. It is only when a benefit is a return from the recipient that is required as a consequence of the original transfer that the transfer is not a gift, as I have defined it. I have thus ignored extensive discussion in the giving literature on the question of the disinterestedness of the gift, or the question of altruism. Hence there is no moral element to these definitions: giving may often be generous, but it is not generous by definition. Indeed, gifts may even be given in the pursuit of commercial gain (Barnett, Reference Barnett2018; Elder-Vass, Reference Elder-Vass2016: Chapter 8). A comprehensive study of the gift will surely include the ethical evaluation of giving institutions but this is a different matter than the question of how to define the gift.