Conditions for a Taxonomy of Transactions

I use ‘transfer’ in the sense of Anthony J. Pickles (2020: 14) as a single movement of an item (thing, service, person, knowledge) from one (set of) person(s) to another, and ‘transaction’ for types composed of one or more transfers. Thus, a sharing transaction consists of minimally one transfer, while an exchange transaction requires at least two transfers in reverse direction. I use ‘taxon’ mostly for the elements of (etic) analytical taxonomies, while I use ‘type’ for local (emic) categories of transactions. This does not preclude that a local type may become the model of an analytical taxon, comparable, for instance, to Marcel Mauss's (2016: 63) attempt to establish potlatch as a taxon beyond the American Northwest Coast.

Like other anthropological taxonomies, the taxonomy of transactions emerges from the tension between the historical baggage of Western-modern self-descriptions and the discovery of alternative types of relationships, often beyond the Western-modern setting (Carrier 1995). Therefore, taxonomies need to be adaptable to specific cases without becoming too imprecise. Their role cannot consist in pigeonholing data into a fixed system of boxes. If they did so, any given example of a transaction type would properly belong to one taxon only. This would immediately lead to anomalies and render the taxonomy useless. The effect is obvious in the debate on gifts and commodities. All too often, these terms are used as ideal types that mysteriously or tragically clash with ‘reality’ (see Bird-David and Darr 2009; Tsing 2013). Grounding a taxonomy in shared analytical terms instead means that each taxon (like gift exchange or commodity trade) is defined by a specific configuration of the same terms (like transaction, value, expectation, alterity, opacity—see below). The defining elements of each taxon can then be reassembled to fit the fine-grained analysis of specific cases, for generalization or comparison, and to create new taxa. A theory of transactions will only do its analytical work if it defines taxa by relating them to each other in shared terms, thereby opening a horizon of possibilities for determining the differences between them.

I will operate mainly with three taxa. Two of them, gifts and commodities, have been central to economic anthropology from its outset, with Mauss's (2016) The Gift as point of departure and further work by other authors (such as Gregory 1982; Robbins 2008; Rus 2008). Other important taxonomies differentiate reciprocity, redistribution, householding, and market (Polanyi 1995), generalized, symmetric, and negative reciprocity (Sahlins 1972), locally specific exchange spheres (Bohannan 1955; Röschenthaler 2010), and modes of exchange (Karatani 2014). Additional taxa, such as barter, gambling, inheritance, and so on, have found less space in comparative schemes. I will reconfigure the classical dyad of gifts and commodities by including sharing, which a number of authors (Descola 2011; Widlok 2013, 2017; Woodburn 1998) identify as a separate taxon with its own, unique features.

I will begin by grounding my argument in theories of gift exchange. This does not imply that sharing and commodity trade really are forms of the gift. Rather, their respective specificity can be accounted for in terms developed by these theories. However, I will elaborate the established terms by a future orientation, that is, expectation.

One reason for the usefulness of terms from gift theories is the fact that Mauss based his inexhaustible essay on a seeming contradiction: the gift is at once deliberate and obligatory (Mauss 2016: 58). This suggests that the dichotomy of egoism and altruism that Western-modern thought has naturalized is historically and culturally specific (see Parry 1986; Sanchez et al. 2017). While many interpretations of the essay stressed one term of this dichotomy over the other, I argue that gift exchange can best be accounted for not by dissolving this contradiction or hierarchizing its terms, but by conceiving it as a productive dynamic. This contradiction then leads to contingencies that break up the static character of much of the current taxonomies without abandoning their analytical value.

Gift exchange occupies a key position within this taxonomy, not least because it unites features of commodity trade and sharing within a single taxon. Aspects of interest, appropriateness, obligation, and contract are shared by both gift and commodity exchange. Sacrifice, generosity, the possibility of unilateral transfer, and asymmetric demands (as in demand sharing) are features of both gift exchange and sharing (Peterson 1993; see also Laidlaw 2000: 628). This does not make gift exchange a mixture or hybrid any more than any other type of transaction. Yet it shows how terms can combine in various ways to form taxa that can still be fruitfully differentiated from each other.

However, placing gift exchange at the core of a taxonomy of transactions only works under certain conditions: First, societies cannot be reduced to a single type of transactions, and, as Hans-Peter Hahn (2015: 14) has argued, this was not Mauss's point either. Rather, different types by necessity complement each other in any given setting, for instance, long cycles of gift-like exchanges and short cycles of commodity-like exchanges (Iteanu 2004; Parry and Bloch 1989; Sprenger et al. 2017). The definition and differentiation of these types is always local and specific. This, however, does not exclude the possibility that one type dominates a specific setting (Karatani 2014: 9). Such hierarchies of local types should be seen as configurations of values. Values and modes of communication that are associated with the dominant type of transaction orientate other types or restrict their operation. For instance, the dominance of interested commodity trade supports the development of its reversal, the entirely disinterested gift (Parry 1986). Therefore, I distinguish transaction types and their settings where they coexist with other types.

A second condition for a differentiated theory of the gift is that its basic premises take the form of seemingly uncontrollable contingencies. Considered as a process, any (local or analytical) kind of transaction appears as a movement between opposing terms, such as determination and indeterminacy, identity and alterity, or opacity and transparency—a movement as necessary as the terms between which it is oscillating. The analysis of empirical types of transactions proceeds within a matrix of analytical terms derived from these contingencies, as exemplified at the end of this piece.

I identify three contingencies that dynamize the analysis of gift exchange. The clue to all of them is their future orientation: the fact that actors expect something to happen. Expectations cover a range of possible consequences in situations of contingency. The first contingency has to do with the fact that gifts enable return gifts, but cannot enforce them. If return gifts can be enforced, it is appropriate to speak of ‘obligations,’ but as long as this future remains contingent, gifts only come with ‘expectation.’

The second contingency addresses the question of whether gift exchange produces a social whole—a ‘society,’ for instance. On the one hand, exchange demands a minimum of shared values, so that expectations are not disappointed to a degree that precludes future exchanges. This suggests defining such conceptual social wholes by the validity of a value system, that is, the range of relationships to which these values apply. On the other hand, the boundaries of these wholes seem to shift constantly. If the whole to which the transaction refers in order to be valid remains unfixed, the underlying value system is at once contingent and mandatory.

The third contingency relates to the persons connected by transactions. In many societies, full personhood is defined by the capability to engage in transactions. The contingency emerges from the fact that the transactions constituting a person or demonstrating its personhood are often highly visible or public; however, the person engaged in them is opaque.

Although I will not demonstrate my approach with one specific example, my inspiration is drawn from my own and others’ work in upland Laos and Southeast Asia more generally, as well as from Melanesian and African ethnography.

Contingency 1: Every gift creates futures that remain indeterminate.

We are among us, in society, in order to expect such and such an outcome among us; that is the essential form of community. . . . I know of no other concept that produces law or economy: ‘I expect’ is the definition of any collective action at all. (Mauss et al. 1934: 60–61, my translation)

In The Gift Mauss established the three obligations of giving, taking, and returning and developed a theory of the origin of the contract. Some authors following this argument have based their understanding of society on the logic of the binding contract, among them Claude Lévi-Strauss (1993). The cycles of marriage that he conceived as the basis of society only work under conditions of obligation. In symmetric exchange, Group A give a daughter in marriage to Group B, knowing that Group B will return a bride. In asymmetric exchange, both the future and the moral horizon of the exchange expand, but the contract is no less binding. Group A give a daughter to B, who themselves pass their daughters to C. Group C, in turn, give their daughters to A, such that A can trust ‘society’—a whole constituted by the exchange cycle—to provide brides in a delayed, extended return.¹

However, when Mauss argued for the importance of the gift for his own society, he did not highlight obligation, but generosity (Mauss 2016: 177–184). A binding contract has no need for generosity—it serves both parties anyway. It requires neither giving in excess of its fixed terms nor exchanging pleasantries. For Lévi-Strauss, there would be no difference between a gesture of generosity and forceful abduction, as long as the transfers are balanced. But if the gift is nothing but contract and obligation, why would there be generosity, “even when there is only a fiction, a formality, a social falsehood in the gesture” (ibid.: 58, my emphasis)?

This question can be answered by taking the quotation opening this section into account: in a conversation with François Simiand, Mauss stated that expectation, really, is the foundation of community and collective action. Such expectation, clearly, goes beyond obligation. Expectation is not a subjective assessment, a cognitive stance, but something that people have “in society.” Neglecting an obligation leads to sanctions—expectations, however, can be disappointed. They do not by necessity imply punishment, although they may. Expectations can also be the subject of detailed ethnographic enquiry—they motivate action in an often reflective, articulated way.

Among the various orientations toward the future, according to Bryant and Knight (2019: 50), expectation is distinctive for its reliance on a dense web of existing relationships. Relationships between people or circumstances allow for fairly precise assessments of what is going to happen next. While expectations do not necessarily suggest that past experiences are merely repeated, they are selective extrapolations of these experiences. A future-oriented definition of ‘culture’ or ‘structure’ emerges from this notion of expectation (Luhmann 1984: 387–394). Culture, then, is not a bundle of transmitted traits, but rather the condition for having certain expectations, as reductions of a myriad of contingent possibilities.

However, expectations do not simply concern a sum of equally possible futures, but also their variability and likelihood to occur. While some situations produce a narrow range of expectations, in others, a wider range of outcomes may be imagined. In an election, for instance, the more candidates there are and the more even their chances of winning appear to be, the wider the range of expectations of future policies.

Described in these terms, a gift generates the social expectation of a return, but does not necessarily enforce it. In fact, numerous researchers have stressed that gifts are sometimes not reciprocated (Drucker and Heizer 1967: 57; Hunt 2012; Laidlaw 2000). Especially in systems of gifting, unreciprocated gifts may even strengthen the very relationships that they travel along, thereby stabilizing the asymmetries that these systems are based upon (Yan 2002). Stavroula Pipyrou (2014: 413) describes unreturned gifts from Italian mafiosi as “dormant potentiality,” and Yosuke Shimazono (2008) reports that Philippine organ receivers feel obliged to care for their received organ as if it was not their possession but just in their custody (see also Strathern 2012). Thus, instead of seeing unreciprocated gifts as social failure or evidence against the concept of gift exchange, I argue that the occasional lack of a return is among its defining features. As Bourdieu (1993: 181) has argued, the gift's indeterminacy makes it liable to the contingencies of practice, with its subtle variations of accumulating various types of capital. The futures that the gift brings into being contain a variety of—often almost equally—likely possibilities, including nonreturn.

This implies that the value of the gift ultimately depends on the way it is received and possibly returned. A sense of a lingering bond would then corroborate the gift's value. This is where generosity becomes important. Gestures of generosity work as safety mechanisms that retain the value of the gift even when no return occurs. They signal that the receiver is free to return the gift. If they fail to do so, the expectation of a return may be disappointed, but the original gift is not therefore considered a failure—the giver may still be recognized as a generous, exemplary person who may even expect a return from a transcendent third party, like God (see Strathern 2012: 402–403). A reversal of the same idea are forms of theft that enjoy a degree of social acclaim, for instance, when interpreted as correcting stinginess. Again, a transfer without return is valorized positively (Bathurst 2009; Retsikas 2016).

This would explain why Mauss stressed the importance of both interest and deliberate giving, of freedom and obligation. He implied a model of society that is not so much inspired by the state that gives and enforces laws, but by the question of how communal life is regulated without centralized power. The notion of expectation provides an answer. Instead of determining effect by obligation, the gift opens up future possibilities. Expectations can be fulfilled, disappointed, or surpassed. Reactions on these outcomes may vary, with sanctions, punishment, or ostracism—if the expectation was considered an obligation—being just some among them.

Therefore, I suggest a graded analysis. Different types of transactions engender stronger and weaker expectations. I define weak expectations as consisting of a relatively great variety of future developments, while strong expectations are comparatively restricted and narrow.

Insofar as the gift is associated with generosity, it allows for a wide range of different expectations, which generates the term's analytical value. In contrast, if the gift was tied to a strong expectation of return, social life would often appear as continuous failure, a chain of disappointments and miscalculation, both from the anthropologist's and the locals’ point of view (Pickles 2020: 17).

A graded analysis also allows for further differentiation. I demonstrate this with an important local type that is often taken as a prime example of the taxon gift exchange. The Kula on Dobu—and elsewhere in the Massim—sets up two contrasting, but equally necessary sets of expectations. In what Bourdieu (e.g., 1979: 74–75) may have called the “official” version, a gift of a necklace is reciprocated by an armshell of matching rank and value. However, if this rule was followed, no one would ever get “bigger” and grow in fame, which is the whole point of Kula. Therefore, the expectation of a return gift of the same value is complemented by the expectation of cheating, for which Kühling (2005: 212) reports a specific Kula term, butubutu. There are several ways of cheating—giving a gift of lower value, giving no return at all (especially when the original giver is inexperienced), or making promises to give a valuable shell to several partners, of which all but one will be frustrated (ibid.: 212–214). The latter option is risky, but nevertheless part of a field of weak expectations. The giver of a Kula gift has a strong expectation that they will receive a certain class of valuable as a return gift, strong enough to call it an obligation. But shells are individualized and ranked, and expectations are weak in respect to the particular item they will end up with. Even a strong desire for a specific shell and careful planning to elicit it from a partner do not guarantee that desired outcome (Fortune 1963: 217). This tension may also account for an apparent contradiction in behavior: When negotiating Kula deals in private, Kula traders treat each other as trusted friends and equals, but the public transfer of the shell is strongly asymmetric, with the giver assertively announcing the return gift they expect—and of course, this elicitation may fail (Kühling 2005: 215). Speculations on what someone will give to whom keep communication going in gift exchange systems. The tension between strong expectations of some kind of return and the much weaker expectations regarding the specific item may produce public discourses that mix admiration of and outrage about those who play the system well (see Liep 2009: 306, 310; Kühling 2005: 216). For the Trobriands, Annette Weiner (1988: 155) states the relation between these two types of expectation this way: “Reciprocal returns [the exact match in value of two valuables] produce the illusion of stability, generating the possibility for gain, while gain itself creates a rupture in continued reciprocity.”

Therefore, analyses of gift exchange that account for the alleged failure of return by a dichotomy of ideal and practice take Mauss's obligations too literally. Instead, seeing expectation as its driving force allows for capturing the range of potential future events more precisely, without running up against a kind of normative reciprocity that may only exist in theory. The framework of expectations thus allows for an ethnographically fine-grained comparative analysis of transactions. Thus, the apparent contradiction between generosity and obligation resolves into a dynamic that indeed enables one to expect returns but with weak expectations regarding certain of their aspects—the very item itself, its timing, the person presenting it, and sometimes the return as such (Bryant and Knight 2019: 131; Callari 2002: 254; Bourdieu 1993: 183).

The gift—exemplary for other transactions—thus connects future and past. It is a ‘project’ in the original sense of ‘to be thrown ahead.’² It creates society by working on a shared future—a future, however, that emerges from the past, insofar as expectations are generated by experiences. Any gift, any current transfer indeed, can only appear as some kind of follow-up to an earlier transfer. Therefore, the inception of any system of gift exchange is so often lodged in myth.

Gifts make counter-gifts expectable, but equally so disappointments, the recognition of generosity, and the emergence of sociality. Indeed, a gift transferred not only elicits some kind of future. Its reception reveals a relationship that implies a larger social whole that allows this relationship to exist.

Contingency 2: Every gift presupposes a social whole whose boundaries are unclear.

Mauss's identification of the gift as a ‘total social fact’ suggests that systems of gift exchange create society as a whole (Hart 2007). The condition for this is a shared system of values that are at the same time culturally specific ideas (Graeber 2001, 2013; Godelier 1999). Values are realized in relationships between givers and receivers just as in relationships between the gift and its responses. The value of a gift depends on the entirety of responses—not just returns—to it that are culturally expectable; the same applies to commodity exchange and sharing. As some responses are considered better than others, value consists of the ranked possible futures following a transfer. This system of values underlies the structure of the social, structure being the expectation of predictable relations. Therefore, the validity of the value system should correspond with the extent of the social configuration that is relevant for the transaction in question (Godelier 1999: 150).

A synthesis of these two definitions of values, as ideas and as the background of transactions, was developed by scholars inspired by Louis Dumont (Barraud et al. 1994; Platenkamp 1990, 2004). This approach is based on the constitutive hierarchy of parts and wholes (Dumont 2013; Parkin 2003). If society is a whole, its parts are locally defined types of relationships, each of whom represents—makes present anew (Coppet 1992: 65)—a specific value that occupies a position within a system of values. Relationships are typified by the transfers they consist of. Therefore, a type of relationship is also a value, and the whole of relationships constitutes society as a value system.

The value of a social relationship relates to asymmetric couplings of the values of gifts and counter-gifts in ritual exchanges (Foster 1990; Strathern 1992). These exchanges establish the values represented by the two gifts as necessary complements, and as complements, the values form a whole (Godelier 1999: 170). In each exchange, one of these values supposedly represents the whole of society and is thus superior to the value of the complementary gift. This determines the value of the relationship as such and therefore its position in the society-as-value system. This way, all ritual exchanges are oriented toward the highest value of the given society and thus make it possible to consider society as a whole. The part–whole hierarchy also determines the hierarchy of givers and takers, at least in the context of the exchange.

One may criticize this Dumontian approach as a return to functionalism that postulates the closure of society as a system of institutions and groups of people (Graeber 2001: 20). However, the assumption that the concept of society as a whole and the sharing of values complement each other has much wider, and more dynamic, implications (Kapferer 2010; Robbins 2015). Thus, David Graeber (2013), despite being skeptical about the notion of society as a totality, stresses that the imagination of a social whole is necessary for sharing values and entering durable relationships. Imagination is not illusion or individual fancy, but, I suggest, the kind of expectation that Mauss, in the quotation above, sees emerging in society regarding the consequences of action.

Similarly, Knut Rio and Olaf Smedal (2009), in their rereading of Dumont and Mauss, go beyond the total social fact and stress totalization instead. The total social fact—of which gift exchange is just one instantiation—is that institution that enables considering a plethora of relationships and values together as a whole. Any action that this institution consists of contributes to the totalizing of society into the image of a whole. Wholeness is neither a given nor a normative ultimate state of society, but a feature of a movement that unfolds in certain contexts and settings. Rio (2014) shows how the image of Ambrym society as a whole appears in a brief period during a large exchange ritual, when all gifts are displayed, only to be disassembled afterward. An analysis of transaction types would then be the royal road to grasping the respective ‘total social fact,’ as it is realized through transactions.

An example from my ethnography of the Rmeet, uplanders in Laos, will demonstrate this. Rmeet bridewealth exchanges serve to produce life and thereby represent a contingent image of the whole of society. From the perspective of these exchanges, the bride is an item of transfer that cannot be returned in kind, as Rmeet marriage rules allow women to marry only into previous wife-taking households of their family or to strangers. There is thus a permanent asymmetry built into the exchanges (Lévi-Strauss 1993, see above).

The most important aspect of women in this system is their ability to create a new ‘house’ through their children. The bridewealth that their husband's family transfers to their own consists of a buffalo and colonial silver coins. Buffaloes are, if not sold, sacrificed to ancestral spirits, while silver coins serve various ritual functions but figure prominently in mortuary rituals. Thus, there is no equivalence of bride and bridewealth. Rather, relationships with future generations, embodied by the bride and her prospective children, are exchanged for relations with past generations, embodied by the future gifts to the ancestors. Relations with future generations are valued higher than those with the past—not only are wife-givers superior to their wife-takers (independent of their social standing), but people and houses are teknonymically named after their children. Each marriage exchange thus instantiates an image of society as a whole that is composed of the complementary values of relationships to past and future generations, with the past servicing the future (Sprenger 2006). Other exchanges, for instance with village or forest spirits, create different imaginations of society as a whole.

The dynamic and contingent character of imagination and totalization implies that each social setting may contain various institutions realizing different types of transactions and thereby different totalities. Such totalities that coexist with each other as imaginations or projects of transactions I call “moral horizons” (Sprenger 2022). Networks of transactions based on particular values thus appear as closed in the sense of coherence, without necessarily being exclusive. For instance, local communities can realize totality by sharing game meat or celebrating village festivals; ethnic groups may realize a different totality by ethnic endogamy, and the state does so in the form of taxes, elections, or compulsory military service; on the level of humanity at large, institutions like Fair Trade or the Declaration of Human Rights create totality as an imagined reality. Every transaction projects a totality of which it is but one instantiation, a moral horizon of the validity of its values. Identifying exchange types and value systems does not require bounded societies. Totality or systemic closure does not imply isolation or exclusivity.

This raises the question of how shifts between such coexisting wholes are accomplished. For this, we need to direct attention away from the sharing of values and toward the other aspect of gift exchange that I have introduced above, the question of why items can be exchanged in the first place. Each gift exchange is based on an asymmetry of expectations—not on quantifiable equivalence but on qualitative difference. This not only refers to the items of exchange, as in my account of Rmeet marriage, but also to the relationships produced or maintained by them. Each transfer contains a constitutive difference, the one between giver and taker. This difference is, like the complementarity of exchange items, a necessary alterity complementing the identity that value sharing suggests. As Marilyn Strathern (1988: 14) argues, this duality of identity and alterity pervades all levels of exchange, connecting the difference between giver and taker with that between kin groups, social strata, or ethnic categories. This difference is not a contradiction of totality, but its condition—the (imagination of a) complementation of two values into a whole. Therefore, imaginations of social wholes as enacted in exchange do not stop at ethnic or other boundaries, but may expand beyond them. Wholes may be built from any difference, as long as it is imagined as complementary.

Iris Därmann (2010) radicalizes this by considering the gift as an essential experience of otherness. Building on Mauss's insight that one always gives away part of oneself, she suggests that the difference between giver and taker corresponds to the difference between societies that are related by exchange. Gift exchange effectuates the integration of the other into the self, without denying its otherness.

This implies a critique of authors like John Furnivall (1956: 311–312), according to whom interethnic relationships are best achieved by markets. In fact, the compatibility of gift exchange with relations between strangers, I argue, may go some way to explaining the continued presence of features of the gift in commercial transactions (Bird-David and Darr 2009; Rus 2008; Sprenger et al. 2017). The relational emergence of cultural identity at least partially articulates through asymmetric transactions: you are what you are because you receive certain items from certain others or give them to them. This allows for connecting different value systems via gift exchange-like structures (Platenkamp 2013). Kula is a major example of this. The islands of the Kula ring contribute specific kinds of firewood to shared rituals that constitute the ring as one internally differentiated whole (Malinowski 1964; Damon 2000: 52). Groups that consider each other as different can thus still imagine forming a larger entity through these differences.

This dynamic complements the contingency dealt with in this section. The tension between identity and alterity is a condition of the whole. Each whole by necessity contains givers and takers that are posited as different from each other within the context of the sharing of values. Pairings of identity and alterity are thus at the center of social reproduction via gift exchange. As this dyad may appear on various levels of social organization, it is difficult to maintain the equation of ‘sharing with one's own kind, exchanging with outsiders’ that has been suggested by Sahlins (1972) or Woodburn (1998), at least as a universal pattern. Rather, I propose terms for comparing types of transactions that allow for differentiation without exclusivity. The question for each case then is: How are identity and alterity conceptualized and how are they weighted against each other?

This approach also allows for analyzing the matches that the symbolism attached to the parties of exchange often states between such levels. A common example is the association of affines with the outside. Potential affines appear as enemies and animals in Amazonia (Viveiros de Castro 2001), while in eastern Indonesia, brides appear as daughters of the soil and bridegrooms as “flotsam” from across the sea (Barraud 1990: 206; see also Sahlins 2008). The “perpetual disequilibrium” (Viveiros de Castro 2001: 29) of the relationship allows for the shifting of scales and the transposition of heterogeneous models of the whole. Shared local concepts of marriage are equaled with an encompassing alterity of the internal and the external. Strangers are cast in the role of exchange's others in the moral horizons created by the gift.³

The whole thus emerges from the alteration between these two principles: integration via shared values versus differentiation of givers and takers of complementary items. This accounts for the dynamics of the second contingency. Sharing values is crucial, but the whole remains contextual. Each exchange is a ‘project’ that evokes the imagination of a particular totality that is never ‘just there’ and ready to respond. Each act of giving risks that the whole it addresses, the moral horizon it implies, refuses to emerge. The expectation of a return thus equals the expectation that a particular whole reveals itself. This includes the possibility that unreciprocated gifts evoke the whole on a cosmic scale—transcendental blessings or the promise of Heaven for the generous. The constitutive difference between givers and takers may be used to model and thus integrate other differences that define specific levels or contexts of the social and are equally asymmetric—like autochthonous versus immigrant, self versus stranger, human versus animal, powerful versus powerless, immanent versus transcendent, and so on. The difference between givers and takers finally raises the question of how the persons involved in transaction are conceived.

Contingency 3: Gifts visibly constitute persons that are opaque and unpredictable.

The contingency I want to address here is most clearly present in Melanesian ethnography, but not exclusively so. I understand ‘person’ as a being locally identified by the qualities that enable active participation in a given system of transactions—for instance, to be able to communicate, have intentions, express needs, or take up responsibilities (Harris 1989). As these features vary between societies and local transaction types, the definition of person and its access to transfers vary accordingly. Often, these qualities only show in the practice of relationships or are even constituted by them. Marilyn Strathern (1988) and Roy Wagner (1991) have argued that Melanesian persons emerge from exchange. The dichotomy of individual and society is replaced by a network of relationships that constitutes each single person as a unique node of exchanges; persons are thus partible or dividual (Strathern 2018).

The transactions making up persons are social, often public. One would assume that persons thus constituted were equally transparent and legible. However, the very opposite seems to be the case. Joel Robbins and Alan Rumsey (2008) or Jürg Wassmann and Joachim Funke (2013) show that the opacity of the person is a central idea in Melanesian discourses about personhood. Melanesians routinely assert that it is impossible to know what is going on inside another person and even inappropriate to ask in the first place.

This, I suggest, is but one form that a more general problem may take, the unpredictability of people engaged in exchanges. There is a relationship between the way that persons and their (un)predictability are conceived and the expectations that structure exchange systems. In this respect, Melanesians not only articulate a local idiosyncrasy but demonstrate a heightened awareness of what Niklas Luhmann (1984: 156), following Talcott Parsons, calls double contingency. One's own communicating depends on how the other could possibly react, but as we know that the other operates under the same premise, successful communication becomes utterly improbable. We are thus constantly communicating into uncertain futures.

Under conditions of relational personhood, the opacity of the other is a specific form of the unpredictability of the future. Saying: “I know what she is thinking” ultimately means to say: “I know what I have to expect from her.” But in systems of gift exchange, the fixing of the future, the narrowing of expectations is ultimately not desirable. Because exchanges are public, legibility of minds and predictability of behavior would infringe on the play of matching returns and cheating (see above; Buitron and Steinmüller 2021: 43–44).

Robbins (2007: 308) describes as much for the Urapmin: Everybody is obliged to exchange, but has to consider carefully what to give to whom. Receiving a specific gift does not determine a specific return. Expectations thus need to remain weak in a system of gift exchange. Consequently, in societies where explicit speculation on other people's thought is disapproved of, empathy is expressed not by verbal promises but by gifts and support—that is, invitations to exchange that establish possible futures (Wassmann and Funke 2013: 240). When persons are defined by gift exchanges and gifts are obligatory but not fixed in their detail, it is plausible to express the contingency of future possibilities in terms of the opacity of the person. When persons ontologically equal relationships, it is not an individual that is opaque, but its relationships. Therefore, it is plausible that the opacity of mind becomes an explicit value in societies that are equally explicit about their self-definition through gifts as the dominant form of transaction. However, the basic idea is found in most other societies, even in the commonplace notion that you can't look into someone else's head. One may even assume, along with Philippe Descola (2011), that the distinction between the observable effects of physicality and the impenetrable workings of interiority is a cognitive universal.

There is a certain tension here between what psychologists call the theory of mind and the various (possibly even absent) cultural concepts of interiority, mind, agency, and so on. Psychologists argue convincingly that the theory of mind is a human universal—the recognition that other people have a mind and perspective on the world that differs from one's own (Wassmann and Funke 2013). This, however, does not imply that all people in the world share the same concepts of interiority or draw the same links between interiority and outward communication and action. Studies of the opacity of mind, such as those quoted above, are primarily concerned with culturally specific concepts of interiority. They are thus restricted to conceptual systems that assume that minds or intentions are necessary for action.

The notion of personhood, however, is wider than that. As I have argued elsewhere (Sprenger 2017), the recognition of personhood is a matter of identifying events as communication and then ascribing these communications to a person. Personhood is thus first of all the observation of communication, which, in turn, is structured by expectations (Luhmann 1984). Specific concepts such as ‘mind,’ ‘intention,’ ‘heart,’ and so on are important conjectures about interiority that serve to interpret observable action and communication.

Identifying a being as a person thus entails certain expectations regarding its communication and action. This includes a certain degree of unpredictability—the expectation that expectations will not always be met. In a sense, personhood can be understood as the acknowledgment that certain events are communication and the expectation that communication is unpredictable in a specific way.

I thus suggest differentiating between the opacity of mind and the opacity of personhood, with the latter encompassing the former. While the opacity of mind concerns certain local theories of interiority, the opacity of personhood concerns the unpredictability of communication that is assigned to persons. Certainly, the latter goes along with concepts of interiority and assumptions about their relationship with outward action, but there may be cases in which this link is not very pronounced. What is crucial for expectations of the gift, though, is that the public and communicative character of gift giving connects to the unpredictability of persons. The range of weak and strong expectations concerning the future and the social wholes thus also applies to personhood, all articulated through gift exchange.

Expectations of Sharing

The three contingencies of the projecting of uncertain futures, the creation of contextual wholes, and the constitution of opaque persons through visible exchanges provide a set of analytical terms that accommodate a continuum of forms of transaction. Analytical taxa like commodity trade, gift exchange, and sharing are incisions in this field, but merely heuristic options for comparison. A terminology of expectations, value, moral horizons, alterity, and opacity of personhood helps with deciding if these taxa are useful in any given case or if additional taxa—for instance inspired by local conceptual systems—are more appropriate. In any case, as various types of transactions coexist in any given setting, the use of analytical taxa always contains a comparative dimension.

Let me begin with sharing. My models are egalitarian hunter-gatherers like the !Kung of Namibia or the Hadza of Tanzania as described by James Woodburn (1982). They are organized in bands in which any big game is shared among all members, irrespective of the question of whether receivers will contribute anything later. Hunting success is hard to predict, and some people are better hunters than others. Still, differences in their contributions neither translate into privilege, debt, or dependency, nor do they alter the right of any group member to their share (Woodburn 1998). Even strangers from whom no return can be expected are allowed to participate (Widlok 2013: 16).

Therefore, sharing is unlike an unreciprocated gift in a gift exchange system. As shown above, even unreciprocated gifts may engender an expectation of return or a sense of indebtedness. In contrast, as Woodburn (1998) and Thomas Widlok (2019) point out, sharing does not create debt. They thus argue against attempts to identify sharing as ‘generalized reciprocity’ (Sahlins 1972) or ‘exchange’ (Winterhalder 1997).

The expectation that everybody will contribute equally to the group's meat provision is thus very weak. Among many future possibilities, it is by far not the most likely one. There is, however, a strong expectation that if meat is available, it will be shared. This differentiation of expectations thus contrasts with the Kula ring, where expectations of giving, receiving, and returning are equally strong, but those regarding the specifics are not. Kula expectations are also strong in respect to who will return gifts that are clearly marked as such, while this is absent in hunter-gatherer sharing.

This brings us to the question of moral horizons and the range of shared values. The inclusion of strangers, highlighted by Widlok (2017), relates to the fact that sharing occurs among people present to each other and the construction of value. According to him, sharing draws attention to the intrinsic value of items, while gift exchange emphasizes the value of the relationship thus maintained. The gift is subject to a Saussurean arbitrariness of the sign: What is actually circulating is less important than its trajectories. Thus, while Maussian gifts are mostly luxury items, everyday necessities like food are subject to sharing (Widlok 2013). Sharing then stresses the independence of the value of items from the social relationships they travel along.

This argument does not require a distinction of universally valued items of sharing from culturally specific valuables of gift exchange. As Marriott's (1976: 110) account of transactions between Hindu castes shows, a system that creates lasting and asymmetric relationships also accommodates food as a “substance-code.” Therefore, even when people stress the intrinsic value of an item, the transaction presupposes a shared understanding of this value.

The difference between the valorization of the relationship in gift exchange and the valorization of the item in sharing is however decisive for the imagination of moral horizons. Sharing a valorized item like meat creates a community of people sharing a value. When strangers are included, sharing projects the possibility of a community of value that extends beyond the people commonly participating in current occasions of sharing. Sharing thus problematizes the group as a bounded community of value. Expectations of sharing support the coherence of the given group and at the same time suggest that this group only exists because there are other such groups. The hunter-gatherer groups described by Woodburn (1982: 435) are rarely stable over time. People often change bands, and bands split or merge quite regularly. Therefore, one's own group is just one level of a society that encompasses several groups sharing the same values and expectations. The integration of outsiders is proof of the validity of the value system beyond the confines of one's group. Even though sharing occurs in the presence of people, the intrinsic value of the thing shared suggests that matters are similar in other groups. The imagined whole extends beyond the forager group.

Therefore, the actual boundaries of the value system are an open question. The hunter-gatherers on whom Woodburn bases his argument have been living in close vicinity to herders and farmers with different value systems for a long time (Woodburn 1982: 447). They were thus aware that their way of life was not universal. It is possible that hunter-gatherers perceive their value system to be expansive, not least because of its two levels, the forager group and the ‘ethnic’ category, the sum of the groups that people can move between. Sharing with strangers tests the validity of the value system. It is an experiment concerning their ability to share ideas of what is valued intrinsically. The expansibility of the validity of the value system corresponds with the expansibility of its moral horizon. This relates to the unpredictability of the future. Just as possibilities are projected onto the social time of an indeterminate future, they are projected into the social space of the foreign. Even when dealing with strangers, a transaction may work as the source of future possibilities.

This suggests that, given the differentiation of identity and alterity on any level of transaction, transaction systems generally have the potential to expand or contract. Like the matching of affines with outsiders, the differentiation of scales—from group to ethnicity to strangers—allows for experimental extension.

The emphasis on intrinsic value helps abstract the value system from its respective social bonds. This also occurs in different ethnographic settings in strategies of objectification or fetishization. The items of transfer transform the social into matter that potentially carries values across the boundaries of groups, value systems, and time (Godelier 1999; Graeber 2005). This way, the objectification of the social in the value of what is shared addresses the problem of the coupling of identity and alterity. Egalitarian hunter-gatherers stress the principle of shared values in the moment of sharing with everybody present. The alterity of successful hunters and meat receivers, members and strangers is subdued—except when hunters claim special shares of the meat or are excluded from eating their own game (Woodburn 1998). Alterity is thus subordinated to identity.

This brings me to the third dimension of expectation, opaque persons. Richard Lee (2003) and others have observed the denial of hunting success among hunter-gatherers: a hunter returning to camp would silently lay down his weapons and thus signal that large game is ready to be brought in. However, he would not state what exactly he has caught. This has been interpreted in terms of egalitarianism, the avoidance of stressing one's importance for the group. In light of the above discussion, it also appears to be a kind of opacity. When a !Kung hunter denies having any good meat to offer to his guests, he is not only indicating that food will be excellent (ibid.), he also makes himself open to interpretation. He communicates that you should not know exactly what to expect. He thus stresses unpredictability in transfers. Both the opacity of exchangers under conditions of asymmetry in Melanesia and the obliteration of hunting success under conditions of egalitarianism are forms of acknowledging the unpredictability of the future. The opacity of others that is, the unpredictability of their communication, appears as a condition of the expectation of giving in sharing systems.

Obligations of Commodity Trade

Let me now attempt to demonstrate similar problematics with different emphases for commodities. As above, I restrict consideration to a particular field, modern-Western capitalist transactions, by itself a vast field that allows only a sketchy discussion. In addition, capitalism is accompanied by a prominent social science, economics, that generalizes its models of capitalist markets and is itself future-oriented.⁴ Still, I believe that the approach I suggest here puts some features of capitalist commodity trade into a different perspective.

Systems of contractual commodity trade are regimes of strong expectations. Each transaction fixes the exchange items in relation to a precisely calculated amount of money. They narrow down possible future developments to a few, preferably just one single option. This may be hardly perceptible when we take groceries off a rack and carry them to the cashier, expecting that the clerk will accept a known sum of our money for them. Expectations become more palpable while we are waiting for an item ordered on the internet. When these expectations are not precisely met, the exchange can normatively be undone.

However, although commodity transactions are morally less charged than gift exchange (Parry and Bloch 1989), many of them are embedded in long-term relationships (Granovetter 1985). Loans, insurances, mortgages, and the like are forms of commodity exchange with long-term expectations, and a significant amount of wealth is currently made by transferring them. Trade in, for instance, futures is a transaction of commodities whose value has yet to materialize but has to be calculated quite exactly for transfer in the present.

These arrangements reveal certain tensions of the commodity. The precision of calculation in each current transaction contrasts with the volatility of prices in the longer run. This contradiction between strong expectations in the short term and weak expectations in the long run produces a problematic labeled ‘uncertainty’ that marks the variability of future outcomes. As determination is impossible, but exchangers desire a reduction of possibilities in long-term calculations, an “ethics of probability” (Appadurai 2013: 295) emerges that tries to make the likelihood of outcomes at least calculable, thereby providing them with features of prices while reducing and hierarchizing the range of expectations.

The tension is evident in studies of financial traders. While the products they trade with are supposed to yield money only in the comparably long term, traders themselves aim at short-term profits (Garsten and Hasselström 2003; Ho 2009). Their options for dealing with this contingency include shifting from rational calculation of the future to intuition (Carrier 1997: 12) or embracing ‘risk’ and ‘uncertainty’ as values (Zaloom 2004). However, the same tension seen from an external perspective may lead to pathologizing traders as gamblers (Núñez 2017). In any case, the strengthening of expectations is still confronted with unpredictability. Under these conditions, however, even thoroughly calculated forms of exchange with fixed expectations, like labor contracts (Akerlof 1982), may assume features of the gift, such as generosity and asymmetry (as opposed to equivalence). The tension between the fixation of conditions for single transactions and the flexibility of the system as a whole, however, seems to be an inevitable result of the attempt to strengthen expectations.

Partially because the market orientation of Western societies is a form of self-idealization (Carrier 1997: 31), the market provides a ‘model for’ the way transactions should be understood more generally. In Western-modern social formations, the prediction of the future becomes a value in itself and proliferates in numerous forms, some serious, others more playful: scientific calculations, science fiction, even divination and soothsaying. Prediction spills over from the system of commodity trade and permeates all walks of life, indicating the hegemony of this specific type of transaction (Pels 2018).

For the problem of the social whole, modern capitalism has developed an approach that is at once similar to and different from sharing. Similar to sharing, modern markets stress the intrinsic value of items, at the expense of the social relationships by which they come about. However, the expansion of the value system that this makes possible is not so much experimental, as I surmise for sharing, but universalizing. Capitalist sociality requires the assumption that everyone is potentially interested in buying a car or a smartphone, irrespective of the community or value system they identify with. While the presence of actors elicits the experimental application of the value system in sharing, capitalism posits an imagination of ‘humanity.’ Along with it, a notion of ‘human nature’ emerges that underlies all cultural difference and is defined by ‘unlimited needs’ (Carrier 1997: 18; Sahlins 1996).⁵ This projection of a moral horizon only bounded by the human kind augments the intrinsic value of the transfer items. Modern market transactions famously ignore cultural boundaries, while positing the universality of their models (Dumont 1977). Buying in your neighborhood store or in a faraway country is not much different when the price is fixed, that is, isolated from who you are and how you relate to the seller. Engaging in gift exchange in an unfamiliar social setting, in contrast, can be excruciating.

The universalizing of the moral horizon and the fixation of value to the items comes with a double effect: a strict separation of human persons from objects and other species, all of whom are denied agency (Callon 1999; Sprenger 2016). The separation of humans from other species circumscribes the moral horizon of shared values, while the separation of humans from objects allows universal value sharing by isolating value in the objects. Both support the constitution of the person as an indivisible, autonomous ‘individual,’ in the sense of individualism as a value system (Dumont 1991). This in turn has made the contract between two individuals the hegemonic model for transactions in capitalism.

Individualism and the universalized sharing of values together obliterate the hierarchical alterity of superior givers of money and inferior givers of commodities, with the latter depending on the former to actually turn objects into commodities (Karatani 2014: 14). The concealment succeeds because, in individualism, asymmetries are not phrased in terms of hierarchical complementation, as in gift exchange, but criticized in terms of power. The hierarchies immanent to asymmetric gifts can potentially be reversed but never eschewed. Power (in Weber's sense), however, is thought to be something that can be ameliorated, diminished, even entirely subtracted from individualist relationships—the optimal relationship being free of power (Iteanu 2009). In contrast, trying to make actors equal in gift exchange misses the point, as their relationship is based on valorized asymmetry.

This has consequences for the question of the opacity of the person. Opaque personhood poses a problem for capitalism. Along with the narrowing of expectations, capitalism fosters either an obliteration of personhood or a growing desire for transparency. Obliteration of personhood, which is related to the classic Marxist notion of alienation, occurs when economic actors are removed to faraway places, such as laborers in poor countries producing for rich ones, or when actors are objectified, as in discourses on ‘human resources.’

The opposite effect, desire for transparency, shows when people with stakes in market contracts, mortgages, insurances, and the like wish to know with whom they are dealing. This demands a reduction of opacity, and the idea that persons could possibly be fully transparent becomes as irrefutable as nontransparency in societies describing themselves in terms of gift exchange. This produces a contradiction between the privacy of the individual and the need to calculate its future.

One field in which this articulates is the popularized image of neuroscience that makes even free will calculable. Here, models of neurological predictability clash with traditional ideas of individual self-determination (Gardner et al. 2018; O'Connor and Joffe 2013: 258–261). While capitalism places great value on the freedom of contract, the ontological freedom of those entering contracts is increasingly cast in doubt. Everybody may enter a myriad contracts—in this respect, unpredictability is valued. However, once a contract is concluded, everybody is supposed to be calculable, to make sure obligations are met. In this respect, capitalist modernity exemplifies a reversal of Melanesian dynamics. The tension between the obliteration and invisibilization of personhood and the demand to render it transparent creates a continuum of moral debates with subjects as diverse as data privacy and human rights in supply chains.

The table below summarizes the relative qualities of the taxa I propose. The greater-than sign indicates that both ideas are present but hierarchized.

Conclusion

Transactions produce expectations of a future that remains unpredictable. However, they deal with unpredictability in rather diverse ways. Contracts in commodity regimes aim at a closure of expectations, narrowing them to a few possibilities: what is given and what is returned, as well as the persons involved, are fixed in detail. In many forms of gift exchange, the contingency of the transaction is much more apparent and expectations of returns are weaker. In sharing, unpredictability regarding returns and the persons involved becomes predominant.

This shows how a taxonomy of transactions can be derived from combining a variety of continuous parameters. Taxa like sharing, gift exchange, and commodity trade are tentative, but demonstrate how to make incisions in the vast continuum of possible types. Such taxa are thus of comparative use. A specific example of transaction is not to be identified as gift exchange because a fixed range of possible outcomes is expected, but because expectations are weaker when compared with commodity-like transactions, either within the same setting or across settings. Comparisons could be based on the following analytical terms:

1. •the variety and strength of expectations;
2. •the validity or range of the moral horizons presupposed by the transactions in question, that is, the projected social whole;
3. •the respective emphasis on identity (shared values) vis-à-vis alterity (minimally the differentiation of givers and takers);
4. •the transparency and opacity of the person.

All these aspects of comparison derive from the principle of an unpredictable future. Expectations of returns or other follow-up events as social imaginations emerge within the relative validity of a value system, that is, the moral horizon it projects, and the transparency or opacity of transaction partners. This avoids a rigid separation of gift exchange, commodity trade, and sharing as mutually exclusive taxa, albeit without making them obsolete. Each society contains various types of transaction, and in order to transpose these local differences into a shared analytical language, keeping the established terms may be worthwhile. However, this requires the taxa to be relational. Identifying sharing or commodity trade is less about checking the boxes of a list of features, but rather about asking if the difference between two empirical types of transaction can be better understood in terms of the difference between the taxa. This way, we can relate our analytical schemes immediately to questions that we may ask in the field: If you give this, what do you expect? Are there different outcomes and how do you assess them? Whom does this affect? What do you call this kind of transaction, and how is it different from other types of transactions?

Therefore, the proposed grid of analytical terms invites the identification of new taxa alongside the established ones. It draws attention to features often overlooked in the analysis of transactions: by extrapolating from known patterns of relationships, transactions constitute society not as tradition but as a shared project of expecting the future.

Acknowledgments

An earlier version of this article, significantly shorter and different in emphasis, was published in German (Sprenger 2019). I thank Thomas Widlok, Hans Peter Hahn, Jos Platenkamp, Mareile Flitsch, and Hans Steinmüller for important discussions, as well as the five anonymous reviewers for Social Analysis.

Comments on “Expectations of the Gift,”

by Guido Sprenger

Anthony J. Pickles

University of Birmingham

This was a sparklingly original and thought-provoking article. In a field where complicating particularism (‘but here things are more muddled’) and conceptual neophilia (‘try these novelty glasses’) are the norm, Sprenger's broad, sophisticated comparativism is a cool drink on a hot day.

The argument springs from two postulates: (1) Transactional categories should not begin from archetypes, they should be derived from analytical features that emerge out of ethnographic findings and comparison. (2) The prime referent for transactional categorization should be the way each transaction shapes expectations of future transactions. I question whether Sprenger really achieves the former, and query whether the latter is really necessary, but Sprenger has built an elegant algebra of transactional ‘taxa’ from both in combination. The resulting analytical techniques have broad potential to refine our understanding of economic activity.

Postulate 1: As Sprenger noted, in 2020 I argued the case for a more inductive approach to transactional categorization and outlined how to build descriptions using a less theoretically burdensome lexicon based around transfers. Such a minimalist lexicon would establish a baseline from which we can make inductive descriptions of actual transactions. If this is too much to hope for, at least we might reconcile our observations. I hoped that my effort would become a staging post in an imaginative project that takes our understanding forward. One that breaks from the contemporary monotony of anthropologists revealing that the world complicates our transactional categories. Sprenger shares my concern, but his approach is a little different. While I advocate for minimalist and theoretically unladen terms to construct complex ethnographically determined descriptions, Sprenger brings in a second raft of concepts that are also deductively generated and strictly defined (value, expectation, alterity, opacity, among others). The arrangement of these concepts can be reconfigured by the anthropologist to fit specific cases, or so it is argued by Sprenger. Noting how these concepts are transformed while they are adapted to different ethnographic examples unlocks an open-ended theoretical language for comparison. Sprenger demonstrates this using a range of gifts, commodity trades, and instances of sharing as his examples. Presumably in order to make space for the broad logical connections between transactional taxa, the ethnographic description becomes thin and reductive in places. The level of generality of the ethnographic descriptions and the apparent homogeneity of their meanings in Sprenger's writing makes for compelling logical connections between transactions across contexts, but there seemed little room for local variations in interpretation within a given context and across different contexts where similar transactions occur. This may be a price worth paying for conceptual clarity, but there is a bargain here and there are costs. A longer piece might achieve insightful comparison while sacrificing minimal descriptive fidelity, but more often clarity falters under the weight of specificity.

Sprenger improves my terminology by using ‘taxa’ to denote the ideal-typical transaction types coined by anthropologists. He also uses ‘type’ for local or emic categorizations. A quibble: I personally think ‘type’ is too similar a word to ‘taxon’ in an already jargonistic field of inquiry; a word like ‘strain’ instead of ‘type’ would make the distinction between emic and etic categorizations more obvious, and the word ‘strain’ conveys the sense that emic categorizations may be fuzzy. As an example, the common-use English-language word ‘present’ is strained from the soup of overlapping words (‘gift,’ ‘loan,’ ‘donation,’ ‘handout,’ ‘benefaction,’ ‘gratuity,’ ‘largesse’) that may have been chosen for the transaction depending on the perspective of the actors involved. A word like ‘type’ implies unambiguous boundaries that should not be assumed. The boundaries of emic categories are not always fuzzy, but the language should account for the fact that they might be.

Postulate 2: While Sprenger strenuously denies that one anthropologist-defined taxon (e.g., commodity exchange) is a derivative of another (e.g., gifting), he claims that it would be helpful if one taxon could be described adequately in the terms of another for the purposes of cross-contextual comparison. Sprenger achieves this by bringing in a modifier: future-orientation. Seen through the lens of future-orientation, Sprenger thinks different taxa can appear as algebraic (my word, not his) reconfigurations of each other.

Sprenger is, of course, right to highlight that transactions are temporal phenomena as much as they are movements of value, with dynamic foregrounding and backgrounding of states and statuses that occur through transactional sequences. While skillfully presented, this is not especially revelatory. Iconic contributors to theorizations of transaction all make movement over time a pillar in their models (e.g., Levi-Strauss 1966; Bourdieu 1977; Woodburn 1982; Bloch and Parry 1989; Strathern 1988). Much of the continued potency of these contributions comes down to the clarity and conceptual development of the dynamic in their models (wife-givers and wife-takers, habitus, dynamic exchange hierarchies, the careful recomposition of gendered imbalance). What Sprenger does particularly well is capture a crucial difference between the emic and etic perspectives on timing. Sprenger's formulation evoked the image of raindrops falling into a pond, first seen from above and then from the surface of the water; the effects of individual transactions—the raindrops—cause ripples on the surface of the pond that is social life. This is clear from above the pond, but equally important for Sprenger is the perspective of the pond-skater (those people for whom the transaction has some direct importance), for whom the overall pattern of the ripples can be hidden by the undulations of the water and may be disrupted by other raindrops. If my flighty analogy holds then the turbulence caused by the rain appears to the pond-skater as uncertainty about the future: what is behind the next wave? From up above the pond, Sprenger sees uncertainties as ‘contingencies’ that follow a pattern that can explain the movements, while paying attention to how pond-skaters react to the changeable conditions.

Sprenger concentrates upon the transactions that anthropologists label ‘gifts,’ and three dynamizing contingencies that shape the ripples they produce:

1. 1.Gifts enable but don't ensure returns, and therefore are best understood in terms of expectation rather than obligation.
2. 2.The gift projects a social whole that shares the values the gift embodies and is constituted by a complementary asymmetry between givers and takers. This social whole requires further gifting to consummate and affirm its existence.
3. 3.The gifts are clear when they are gifted, but the person involved and their intentions are hidden in the act, creating the conditions for further gift-elucidations.

The result is a set of attributes that can vary in their expression: the expectation of giving and receiving; the expectation of return; the expectation of specific return; the moral horizon; whether personhood is opaque or transparent; are there shared values?; is the relation between exchangers one of alterity? Think of these as the properties of the ripples. These gift-derived contingencies are then applied to the taxa commodity exchange and sharing, noting the conceptual adaptation that was necessary to make the descriptions operate effectively. Are the ripples denser, are they steeper, do they dissipate quicker?

Not wanting to regurgitate the argument any longer, I found the way that different transactional taxa appeared as specific configurations of Sprenger's chosen attributes very thought-provoking. I intend to try applying this method in future work. I was particularly taken with how the algebra accounted for the ways that gift exchanges produce a valorized and complementary hierarchy and asymmetry while commodity exchanges are criticized in terms of their power imbalance. If anything, I thought more space could have been given to exploring what the different configurations of attributes for different taxa told us about how transactions stand apart from each other. I hope Sprenger has more to say on this in the future.

Sprenger emerged into this conceptual clearing by thinking in terms of future-orientation. It helped him to build his algebra, but I am less sure that the scaffolding need remain once the system is erected. Just as the deductive concepts were well chosen for their purpose but are not immutable necessities for an effective analysis, I don't think that all “these aspects of comparison [necessarily] derive from the principle of an unpredictable future.” They might equally derive from social memory and the imprint of memory on a rolling present. How else should we account for those with unshakeable faith? Or those who are driven by grief? Or slaves to precedent? What about people whose future is determined for them by others? What of an endowment? What about the gift of Christ's sacrifice in Christianity, which can never be returned or matched? Each of these complicates the idea that the future must seem unpredictable for Sprenger's analysis to generate insight. All these examples might be explainable in terms of an unpredictable future, but that doesn't mean they must or even should be. As far as I am concerned, it is the motion that matters rather than the temporal conceptualization. It would be better to include a variety of approaches to time than assume an unpredictable future.

In fact, in crucial passages, temporality (that is, orientations to time past, present, and future) might have served the argument better than the singular focus on the future (and expectations of it). For instance, Sprenger writes that because gifts do not determine a return but rather create an expectation, the gift “opens up future possibilities.” This is not so; expectation narrows the field of possibility, even if it doesn't do so to quite the same extent as obligation would. Elsewhere, Sprenger claims that gifts can only appear as follow-ups to earlier transfers. Not so; initiatory gifts are common, and while these are also commonly described as precipitated by some characteristic of the recipient (beauty, dignity, needfulness) this is not always the case. Reinforcing my previous point, these are times when the text strains to apply a future-orientation framework based on expectations when there isn't really a call for it.

Sprenger overaccentuates the potential of thinking in terms of expectations, and this leads him to commit two major errors: he is wrong when he writes that all gifts are about expectations in general. He is also wrong when he claims that specific transfers necessarily raise specific expectations for social wholes or wholesome persons. But even so, his algebra is compelling and captivating, and it makes an important contribution. The core of this contribution is Sprenger's tripartite formulation: transaction types (contingent, ethnographically determined, and rigorously described using neutral terminology), contingencies (that which is undetermined at any given moment), and attribute configurations (the tightly described concepts that change in describable, comparable ways when translocated to different transactions). Employing these three together enables complex but meaningful comparisons across contexts and transactional forms. Sprenger's algebra is thereby able to reach a level of insight that is extraordinary and extensible. This is a fascinating and intriguing reimagination of transactional categories on noncategorical foundations and I hope that others recognize it as I do. It is not the general algebra of all transfers, however, just one possible mathematics.

Ilana Gershon

Rice University

Contemplating the same theoretical concepts that I often do, Guido Sprenger zigs where I zag, bringing me analytical delight. I sputtered happily and frequently as I read, “but you can't just assume that!” or “why would you make this move?” Reading an article that is intellectually uncanny—using such familiar touchstones and yet to other ends—offers me the much-relished labor of exploring what my zags allow me to do, and what Sprenger's zigs enable. Sprenger's zigs ultimately offer the building blocks for a taxonomy that can yield comparison across a wide range of economic practices. My zags have led to a set of ethnographically grounded questions first laid out systematically in my 2019 article “Porous Social Orders.” For me, it is a matter of taste and ambitions whether to zig or zag, or perhaps engage in an entirely different locomotion through intellectual thickets.

Where does my taste differ from Sprenger's? I will first discuss the role of anthropologists in analyzing what is presupposed and entailed in any meaningful interaction, as well as what might be opaque or transparent. I will then address why I eschew totalities, in contrast to Sprenger's enthusiastic adoption of this inherited take on social orders. I conclude by briefly discussing how some pressing contemporary political concerns might be underlying Sprenger's focus.

Sprenger opens by astutely observing what linguistic anthropologists influenced by Mikhail Bakhtin and Michael Silverstein take to be a starting point of any semiotic analysis. To say that the world is socially constructed is not enough. Silverstein and others view every semiotic utterance as part and parcel of world-making, both presupposing and calling a world into being. In some contexts, the weight of the utterance is best interpreted in terms of what it presupposes, in other contexts, what it entails; although with every utterance, both presupposing and entailing are at play (Silverstein 1993). So far, so good—Sprenger and I are in agreement. The next step is where we differ as we explore: how is it determined whether an utterance predominantly presupposes or entails, and can almost every instantiation of a type of interaction stably predominantly entail? Sprenger's argument hinges on the fact that it is the analyst who determines presupposition and entailment, and that indeed one type of meaning-making action mainly entails across all contexts. I disagree on both accounts. More specifically, Sprenger argues that gift exchange is fundamentally and consistently entailing across all contexts, largely because people are so unpredictable that gifts are always oriented toward attempting to predict or control an uncertain future.

Yet from my perspective, it is not the role of the analyst to decide, irrespective of context or one's fieldwork interlocutors’ own analysis, whether a meaningful interaction is entailing or spontaneous (calling forth forms of interaction into the world experienced as new) or presupposing or reactive (reproducing already established patterns) (see also Wagner 1995). Sometimes, and for some of the people participating, the most important aspect of a gift exchange will be the uncertainty involved. Yet just as often, for some participants, what matters is how gift exchanges reassert already established practices, strengthening relationships and beliefs through repetition. In any given exchange, participants may have a range of different interpretations about whether what is happening is past- or future-oriented. Part of the analytical work of an anthropologist is to understand reasons for this variation, to grasp the enunciative space well enough to know when and why some actions are primarily viewed as spontaneous and others as reactive. By insisting primarily on entailment and future orientation all of the time, Sprenger, from my perspective, sidesteps what it means to analyze social interactions alongside others who are also social analysts in their own right, albeit practitioners and not theorists (see Gershon 2009). In short, he creates an axis for comparison across all human exchanges of objects and labor at the expense of establishing the grounds for understanding our fieldwork interlocutors’ richly textured jockeying in a given context.

I have similar reservations around Sprenger's discussion of opacity and transparency. It is not for anthropologists to know opacity when they see it; anthropologists should ask their fieldwork interlocutors to decide this, not presume a quality a priori. Sprenger suggests that people are always opaque. I tend to analyze otherwise, viewing everyone as potentially on a continuum between opacity and transparency, even when discussing Melanesians—fieldwork interlocutors so many anthropologists take to be committed in general to each other's opacity. Even for Melanesians, opacity doesn't always seem to reign. From a Melanesian perspective, it is in the very moment of gift exchange that the veil of others’ opacity is lifted, offering a brief glimpse into how people evaluate their relations with others. What others’ opacity or transparency consists of depends on the cultural context. One can begin by asking how a group of people tend to respond to others’ fundamental unpredictability, just as one can begin by asking how a group of people respond to mortality, but the answers are so quickly entwined with specific cultural assumptions that asking about opacity as a universal makes little sense, from my analytical perspective.

While Sprenger views what is entailed and what is opaque as analytically similar enough to serve equally as a springboard for comparison, I would argue otherwise. What is presupposed and entailed is analytically integral to utterances, whether or not those communicating are aware of this. Opacity and transparency, by contrast, are qualities (or, more specifically, qualia—see Chumley and Harkness 2013; Harkness 2015; Munn 1986)—and as qualia, are too contextually specific to function as a basis for the types of comparison Sprenger proposes. Not every concept of personhood stresses that there is a continuum between opacity and transparency at every historical moment—objects and people have many affordances or qualities, and only some rise to the level of analytical significance for anthropologists’ fieldwork interlocutors. I learned this from Susan Gal's (2017) compelling article in which she wrote about how the qualities ascribed to and valued about porcelain have shifted over the centuries, addressing in detail the way that while an object (or human being) might be materially similar enough over time, what large swathes of people will view as the relevant qualities of the object will change significantly. Thus to have the opacity versus transparency dichotomy be equivalent to the presumed and the projected, as it is in Sprenger's account, conflates two different kinds of analytical categories. And to decide that one pole dominates consistently, which he does for both dichotomies, is to offer an answer a priori that should instead be ethnographically explored. We have differing approaches as to how and when to locate determining which analytical categories are salient—with the anthropologist; as a shared analytical category with fieldwork interlocutors; or as an analytical concept that fieldwork interlocutors hold dear and anthropologists take as an object of analysis. As an example on the last point, I would prefer not, as an analyst, to commit to the existence of totalities, but to leave this instead to my fieldwork interlocutors. As a corollary, I do not believe everyone is engaging with totalities. One of the reasons that this distinction around where the analytical concept resides is so significant is that our different approaches determine what kind of comparison is possible.

Sprenger's article is evidence that anthropologists are still trafficking in totalities, especially totalities such as Society, even in the wake of Marilyn Strathern's (1988) critique. Sprenger zigs consistently throughout this article toward totalities. My understanding is that asserting totalities’ existence enables the comparison that Sprenger aims to provide, admittedly totalities of unrestrained variety and without clearly demarcated boundaries (because for Sprenger, some totalities are small enough totalities to have fuzzy boundaries—raising questions about the totalness of Sprenger's totalities).

Sprenger's commitment to totalities is such that he understands a totality's vulnerability to be an integral risk with every exchange. He explains that every exchange calls forth a totality as a fundamental part of what exchanges accomplish in the social world. Every gift exchange also contains a grave threat: the possibility of not having an exchange evaluated as minimally successful by the standards of those participating will threaten that totality. He writes: “Each act of giving risks that the whole it addresses, the moral horizon it implies, refuses to emerge.” Each time I read this sentence, as a frequent giver of gifts myself because I read too much Pacific exchange theory in graduate school, I shudder. Who wants to live in a world where such a risk is around every corner? I, for one, do not want the weight of a totality's emergence to hinge on any of my actions—gift exchange, rite of passage, public speech, and so on. This may be fine to theorize, but hard to live by.

In general, by focusing on the totality as the achieved unit to build upon, Sprenger zigs in his retooling of Niklas Luhmann where I am inclined to zag. When reading Luhmann, I too found it helpful to think that modern life is filled with different social orders, all jostling next to each other. For Sprenger, the proleptic quality of these different totalities is quite theoretically generative. I am more fascinated by how people, objects, ideas, and forms move among and between these social orders (Gershon 2019). Yet I also suspect that I am more committed to seeing the social orders as operating along different enough principles, principles that are socially constructed themselves so that, for our interlocutors, the boundaries need to be made to exist, and also to be porous in the appropriate ways (however people on the ground might define appropriate). Thus what interest me more than any given social order's relationship to Society as a whole are the differences both between and within these social orders.

What does focusing on heterogeneity within putative totalities enable scholars to analyze? Let me begin with an example that courses through Sprenger's article—the Market versus what I understand to be markets instead, which are concatenations of ways to circulate resources. The Market is a handy totality for economists and their ilk, especially because economic practices in the Market are so often seen as derivative of the simplest pricing mechanism—as though all markets are structured based on variations of what one experiences in the supermarket, in which one knows before choosing an item what it will cost. (Admittedly, within economics as a discipline, not all hold this view). Writ large, the Market is the mirror that putatively reveals value in the world, allowing the wisdom of crowds to assign price and thus create spontaneous order. Sprenger relies on this take on the Market, claiming: “Each transaction fixes the exchange items in relation to a precisely calculated amount of money. They narrow down possible future developments to a few, preferably just one single option.” His notion of expectations at the heart of the commodity exchange relies on viewing market mechanisms as fundamentally homogeneous in how they structure risk and expectation. Yet this version of the Market, as Mirowski and Nik-Khan (2017) point out, overlooks how markets are in fact multiple, structure time and risk differently, and are all too often engineered to privilege some to the detriment of many. When the Market is taken to be a unified global economy, this overlooks all the labor of maintenance, repair, and jury-rigging that goes into allowing resources to move across differently structured markets smoothly enough. When ignoring these vulnerabilities, scholars turn away from understanding how economic systems could falter, or be determinedly made to function in other ways. Paying attention to the heterogeneity in how a social order is enacted allows scholars to understand more about how social change occurs (see Silverstein 1997 for a linguistic example). In short, Sprenger allows scholars to compare along a certain axis, but at the cost of an understanding of social change, and the potential comparisons of the types of difference that differently structured heterogeneous configurations make.

I want to conclude by turning away from zigging and zagging, and to make an observation about our contemporary moment—Guido Sprenger is not the only one thinking about how everyone's quotidian relationships with an uncertain future shape their practices and social analysis. We are in a historical moment in which many are aware that Change is afoot, not only that the ways things have always been done is unsustainable, but that many tried (and not so true) logics infusing so many widespread infrastructures have been proven woefully inadequate. Illiberalisms and many social movements have sprung up everywhere, out of long-standing frustrations with how different economic and political systems have navigated their varied mixtures of neoliberal economics and classically liberal bureaucracies. Sprenger is offering a zigging alternative to a group of scholars exploring people's social imagination of the as-if, of the otherwise, of prefigurative forms and practices as part of how one engages with multiple social orders. In general, it is no accident that so many these days are concerned not so much with what is presupposed, but avidly with what is entailed.

Joel Robbins

University of Cambridge

Though previously reputed to be mired in an eternal ethnographic present, anthropology has for a long time now been messing around with the past. But until recently it has not messed much with the future. Yet in the last few decades, notions like hope, aspiration, anticipation, and the future itself have led anthropologists to begin looking ahead (Appadurai 2004; Bryant and Knight 2019; Guyer 2007; Miyazaki 2004; Rollason 2014). This has already been a productive move. But it has not yet produced a widely shared, generative body of theory. I wonder if this might partially follow from the fact that anthropological approaches to the future are complicated by a basic conundrum: is the future truly open, or is it only the modern cultural notion of the future that understands it as a site for novelty, rather than for the pretty predictable kinds of broad reproduction of past patterns that anthropology was once so keen to document? Of course, modern models of the open future are globally widespread these days, so more people than ever might themselves equate the future with the unpredictable, but as a theoretical difficulty, this question still remains (Robbins 2016). Guido Sprenger's major rethinking of gift theory brings anthropology's newfound future orientation, and these complications it carries with it, to bear on the discipline's core topic of exchange.

Like Anthony Pickles's (2020) article “Transfers,” the other recent theoretically ambitious treatment of transactions, Sprenger sets exchange in time, treating the social framing of any instance of exchange, as well as what follows from it, as an open question. For Pickles, social work has to be done beyond the transfer of something to make an interaction into a transaction of a given type, be it a gift, a commodity exchange, an instance of sharing, and so on. Sprenger's key temporal move is to see any transfer as beset by three contingencies that affect its eventual social definition: (1) the contingency that besets any exchange by virtue of the fact that it leaves the future open (e.g., it might not be returned); (2) the contingency that follows from the fact that every transfer projects a whole of shared values the boundaries of which, and perhaps even the existence of which, are not evident until after the transfer is initiated; and (3) the contingency that results from the ways that gifts create persons who are defined as to varying extents transparent or opaque. For Sprenger, the ways that various transfers understand and engage these three contingencies become the substance of a tour de force rethinking of anthropology's basic typology of kinds of exchanges as instances of either a gift, a commodity, or something shared. The many insights that this typological innovation produces are highly original. For example, on Sprenger's analysis, in commodity exchange the projected whole locates value in the objects themselves rather than in the relationships that move them between people, and the whole thus sees itself as universally expansive. The many uses of his new way of parsing types of transfers along these three dimensions should be apparent to any reader.

My two sets of comments and questions for Sprenger are going to be more a matter of quibbling about a few details of his article than one of questioning its core contributions.

The first set turns on the notion of contingency that is the key Sprenger uses to unlock the theoretical door to a fully open future. In a strict sense, nothing happens by way of necessity in the world he posits. For example, gifts might be returned or they might not be, transactors might share an assessment of the value of a commodity or they might not, and those invited to share might seek inclusion in the whole that emerges from the activity of sharing or they might not. Since nothing is determined by the initial attempt to begin a transfer, what follows such initial attempts can take any form (or presumably no recognizable form at all, in the event of a truly novel action). Those involved in transfers do not, on Sprenger's account, have any guarantees about how they will go or where they will lead. What participants do have, and this is the one key term of Sprenger's analysis I have not yet mentioned, is “expectations” about these things. Such expectations can be strong (e.g., my gift of a Kula valuable will be returned by some other valuable) or weak (e.g., my gift of a Kula valuable will be returned by some other valuable, but the value of that return may or may not match or exceed that of the valuable I initially gave), but all transfers set up some expectations and this is precisely why they are by nature future-oriented.

So important are expectations to Sprenger's understanding of the exchange that he suggests that this idea should replace Mauss's core claim that the gift is constituted by the three obligations (to give, receive, and return) with one that holds that the gift (and exchange more generally) is constituted by three expectations (about giving, receiving, and returning). As he puts it in his abstract (and in various ways throughout the article), we should “replace Marcel Mauss's three obligations of gift exchange . . . with the notion of expectations.” And here is the first place that I want to quibble a bit. My problem is not with introducing the notion of expectations, but rather with discarding that of obligations. I do not think these two terms are true disjuncts, that the choice between them is really either/or, as Sprenger implies. And I think it is worth hanging on to obligations even as we adopt Sprenger's suggestion about adding expectations to our models of exchange. It is worth doing this because it helps us develop a more nuanced account of the kinds of futures social actors look toward.

To make a case for retaining obligation, it is important to understand why Sprenger suggests “replacing” it. His reasons come out clearly early in the article when he writes of “[s]hifting analysis from fixed ‘obligations’ to open-ended expectations.” In Sprenger's usage, obligations are the foundations of predictable, even determined, outcomes; it is obligations that shut down the open future. But I do not think of obligations as automatically fulfilled, nor that this is what distinguishes them from other social phenomena (including expectations). Rather than determining future actions, what obligations do is set up responsibilities in relation to future actions. People are perfectly capable of failing to meet their responsibilities, but when they do, others are within their rights to hold them accountable for this (see Laidlaw 2014: Ch. 5 and Keane 2016: 77–79 on responsibility and accountability respectively). Obligations thus give rise to, rather than obviate, expectations, and they also provide a social basis for recourse beyond just feeling disappointed when obligations are not met. This responsibilizing effect made the obligation-creating qualities of gift exchange central to Mauss's concern with the role of gifts in the development of the notion of contract. But it fits just as well with Sprenger's concern with contingency, for it gives us a way to track the social lives of contingencies when they occur, a kind of tracking that can explain why people do not simply ignore contingencies as meaningless, but take them as occasions to push social life forward in various directions. None of this puts issues of value or opacity versus transparency out of play, so I would argue that whatever one thinks of the proposed benefits of retaining obligation alongside expectation in gift theory, the costs of doing so are theoretically quite low.

A second issue I want to raise is itself somewhat contingent, relating as it does as much to something I have been thinking about lately as it does to Sprenger's argument (Robbins 2020: Ch. 5). Christian theologians quite frequently write about the gift these days. While the range of studies of this topic in theology is broad, one of the most interesting streams comes out of Lutheran theology. The core issue in this literature concerns the receipt of the gift, which also happens to be the Maussian obligation that is least discussed in the anthropological literature, fixated as it has been on what drives giving and returning at the expense of the middle term of the triad. For Lutheran theologians, the gift whose receipt is problematic is the one of grace given a person by God. In the Lutheran tradition, humans are by nature understood to be so fallen that they are not capable of themselves doing anything to secure grace. This means that they cannot even “receive” the grace God gives, at least if we count reception as an action they would have to perform themselves. This logic underlies the famous Reformation turn from works to grace as the source of salvation: until people have grace, no works they do can be fully good, so people have to receive grace passively in some sense. Therefore, a lot of theological work in the Lutheran tradition has gone into laying bare the mechanics of the reception of the gift to show how this can be passively accomplished. Cutting a long argument short, and relying mostly on Ingolf Dalferth's (2016) rigorously argued account, gifts make people recipients, rather than obligate them to become ones. He uses the example of birth as a gift that one receives through no action of one's own, and also inheritance, in the way that one can be in receipt of an inherited gift even before one knows that one is in this role. Being cast in the role of recipient does obligate one to respond, and sets up expectations about how this might be done. Thus a gift always does open new possible pathways into the future that were not there before it was given. But I think it is closer to the experience of being made a receiver to say that more than an obligation or an expectation, being put in this role is a fate, an event that befalls one and thereby shapes one's future possibilities. Can this work with Sprenger's theoretically driven typology of exchange? More generally, what would Sprenger make of reception? Here, as in most other work on the gift, reception is the backgrounded member of the famous triad, and it is encompassed by “giving” in the table that concludes the argument.

As noted, my questions are attempts to work within Sprenger's frame, the utility of which is not in question on my reading. Stepping back to the general topic of the future, I wanted to close by raising the question of just how open a future anthropology wants or needs to posit in theory to reap the kinds of gains Sprenger does here for the study of exchange. Does anthropology need to approach all aspects of all exchanges, or even of all social action, as liable to take novel forms at all times, or only some things and only at some times, or very few things and then only occasionally? Where would a future that is completely new at every moment leave core human issues like responsibility and accountability? My queries also ask just how or why that future needs to be understood as open—is it open because of the nature of social life itself, or only because social life exists in time, or only when those living a particular social life see it as existing in temporal terms that include an open future, or because human beings are the way they are (whatever one takes that way to be)? Along with its major contributions to the study of exchange, Sprenger's ambitious account of the social reality of expectations raises these questions in a very productive way.

Rebecca Bryant

Utrecht University

In his article, Guido Sprenger proposes to replace obligation with expectation in the study of the gift. In particular, he aims to deflect attention from the sanctions that enforce gift giving and to focus on the role of reciprocity in creating and enforcing moral horizons. As I understand it, Sprenger's intent is to think through the significance of transactions to the reproduction (rather than only the production) of social life. If giving and receiving gifts only happens because of a sense of obligation, that would result in the production of social life, its mechanical repetition. Instead, Sprenger argues that when contingency and the possibility of nonrepetition are thrown into the mix, moral horizons emerge as the promise of future relationships. Transactions more generally promise society's future.

Expectation, then, is the answer that this article proposes to Marcel Mauss's fundamental question: “What power resides in the object given that causes its recipient to pay it back?” (1990: 4). That power, according to Sprenger, lies in the expectations that a gift creates. While I agree with Sprenger in his more ontological intent to see the reproduction of social life as a horizon of expectation, the matter gets more complicated when we move to the specific expectations generated through transactions. Encapsulating them only under the rubric of expectation may be more muddying than clarifying.

What I mean by that is the following. In my book with Daniel Knight, The Anthropology of the Future, we argue that the expectation of expectation is the foundation for what is usually thought of as “normal life.” We expect, for instance, that we can expect our house to keep standing or our partner to return that afternoon. Expectation of this sort is the foundation not only for social life but for our ability to act in the world. Mauss says as much in a quotation that begins one section of Sprenger's article: “‘I expect’ is the definition of any collective action at all.”

Yet expectation is always about that which is beyond our control; we can only expect, and our expectations may be disappointed. For this reason, philosophy has most commonly used the metaphor of the horizon to describe expectation—something visible but not in its details. Expectation always contains the possibility of its undoing, which Sprenger discusses through contingency. Nevertheless, because the horizon is such an important outline and limit of our lives, expectation serves as an anchor that produces what Gaston Bachelard (2016: 12) calls “vague and viscous” senses of how things ought to be. In other words, expectation becomes the ground that makes futures possible but is not something of which we're always aware. It is “the promise of the future tense that becomes a conscious expectation only in the past tense” (Bryant and Knight 2019: 77).

This observation appears encapsulated in Sprenger's concept of the “moral horizon,” which is not the focus of the current article but which appears to serve as a conceptual context for it. Sprenger calls moral horizons “totalities that coexist with each other as imaginations or projects of transactions.” As such, they appear to be horizons in the sense that I discuss above, giving us “vague and viscous” senses of a moral order that are conjured in the context of transactions, in which we can (or want to) expect both ourselves and others to behave in certain ways. It is such a moral horizon that seems to be what Mauss is suggesting when he says that expectation is the grounding for action.

Sprenger, then, is trying to resolve the paradox of reciprocity through expectation. The paradox is that on the one hand anthropology acknowledges reciprocity as foundational for social life; on the other hand, anthropology has tended to explain the force of reciprocity through obligation. This then becomes a chicken-and-egg conundrum: which came first, reciprocity or the obligation that regulates it? Sprenger's answer is that this is the wrong question altogether and that what we should be looking at are instead the kinds of expectations that emerge through transactions and create moral horizons.

This seems like a promising route to take at the level of abstraction where we can speak of reciprocity and moral horizons conceptually. Sprenger's use of expectation in this argument is more specific, however. In comparing commodity transactions, gift exchange, and sharing, he asserts that what all of them have in common is “the fact that actors expect something to happen.” That assertion is problematic on two fronts. Firstly, it suggests that expectations define the transaction and are intrinsic to it. A transaction becomes such because of the type of expectation that enables it. We know, however, that all transactions may come with a wide range of expectations, ranging from the obligatory to the complete lack of expectation. What we expect from any transaction is highly contextual.

Secondly, in this framing, expectations seem to be created by the transaction itself. As we know, however, transactions themselves do not create expectations. This would be the equivalent of saying that the act of someone leaving my house leads to the expectation of that person's return. We know, though, that what creates expectation is not the act itself but rather relationships, context, and history. If the person leaving lives in the house, we may expect that person to return, but there are also many good reasons for not expecting that. The person leaving may be a grown child whose bedroom I've kept intact but who now has a life of their own. It may be a partner with whom I've broken up. In such cases, we may hope for return but not expect it, based on previous experience and what we know from similar situations.

In order to think about these particular transactions as generators of expectations, one must point to the obvious fact that was also Mauss's concern: that they revolve around objects changing hands. And here is where I see the biggest problem for Sprenger's argument: it sidelines rather than addresses the materiality of these transactions, focusing instead on the intentions and opacity of other people. The force of materiality is, after all, Mauss's central insight: that “the thing received is not inactive” (Mauss 1990[1925]: 15). The object has a power that compels the recipient to reciprocate. To that extent, one may say that the gift materializes expectation. In particular, it materializes for the receiver the expectations of relationships, social wholes, and reciprocity that are initiated in the act of the object changing hands.

What is missing, here, then, is a confrontation with the power of objects to conjure the past and project into the future. Elsewhere, I argue that objects are important for such temporal work precisely because of their temporal alterity, having a lifespan or timespan that is opaque to us (Bryant 2014). In my book with Daniel Knight, we refer to the capacity of objects to bring other possible futures into our present as “temporal dynamism.”

Expectation, then, can deflect from but not answer Mauss's original question about the power that resides in the object. This is particularly the case because, as I remark above, expectation always contains the possibility of its undoing. Ultimately, we expect from others because we can only expect. I suggest that both the temporal alterity and the temporal dynamism of objects may be places to start in thinking about how transactions involving material objects generate or manifest social wholes and thereby expectations. The “spirit of the thing” (Mauss 1990[1925]: 13) should then not be abandoned but brought into conversation with the varieties of expectation that Sprenger outlines here.

Marilyn Strathern

University of Cambridge

Anyone tilling an already well-tilled field runs the risk of stumbling over the crisscrossing furrows left by previous investments of labor and thought. However, this sure-footed piece is not a fresh ploughing at all—the old furrows are still there—but rather a fresh approach to cultivation that considers the very soil itself. Mauss's impacted term ‘obligation’ was holding back the regeneration afforded by creating an altogether livelier ground in the concept of ‘expectation.’ Sprenger is able to offer us a reimagination of the gift without obliterating the contours of much that is already known.

Everything is familiar and unfamiliar at the same time. The substitution of a grounding concept generates an illuminating framework for analysis. (‘Obligation’ is not discarded but ceases to be grounding.) One cue to the framework's persuasiveness lies in the ‘solutions’ offered to puzzles along the way: for instance, why so many (Euro-American) commentators approach transactions as though the starting point were an engagement between two individuals or why people are not surprised by unreciprocated gifts. Sprenger's framework takes its illumination from an address to a particular field triangulated by gift exchange, sharing, and commodity trade. From the outset he says that he is bringing “a general perspective of social life,” the preoccupation with outcomes and futures, to bear specifically on “economic anthropology,” here largely in terms of a taxonomy of transactions. Seizing this as an invitation to ask about other fields of analysis to which his observations might also apply, the present exercise pursues the point in reverse:⁶ perspectives suggested by investigating economic relations through his taxonomy may also bear on diverse social interests.⁷ In brief, one way to gauge the effective reach or limits of Sprenger's substitution is to ask about the extent of its generative potential.

I start with his astute intervention into the debate about opacity of minds and make a supposition concerning self-consciousness. Second, there is the issue of coercion and violence in the gift, on which the author's particular choice of contingencies places no emphasis. A third comment turns to his foundational concept of expectation, and what the notion of an unpredictable future may tell us about the past. Differently located as these perspectives are with respect to Sprenger's exposition, the issue in each is what is learned from his argument.

Opacity. Among the terms of analysis, the opacity or transparency of persons seems most ingenious. It is a pivot for a strong comparison between commodity trade on the one hand and gift exchange and sharing on the other. By detaching the opacity of minds from the opacity of persons, of which the former is one aspect (a supposition about interiority), the author adds to the folk model of the impersonal market a figuration of those who populate it: beings as transparently calculable as the numbers with which they deal. An observation concerning gift exchange encourages the thought that what is ‘impersonal’ about the market could as well be described as ‘relation-less’ (ostensibly free of entanglements). For Sprenger observes how often the opacity of persons has been recorded in gift exchange systems, that is, in the context of transactions brought into the open;⁸ translating persons into relations writes this in terms of relationships being called forth and made visible thereby yet remaining opaque. Satisfaction or failure will only appear in the afterlife of those relations.

Stimulated by Sprenger's strong comparison, let me recall another way of treating persons, Wagner's (2012) contrast between impersonation and expersonation, orientations also to be found in social life at large. Impersonation might cover those situations where a subject mimes themselves as an individual, such as a clan reinforcing its identity through embodying life-forces (being one's own ancestor), while expersonation could describe distributions of the self over multiple relations, forming a reflexive arena of ‘self-consciousness.’ After Sprenger, the reflexive arena could well comprise relations whose unpredictability must be forever worked upon. These examples (mine, not Wagner's) come from Melanesian gift exchange. Consider instead imagining self-consciousness as an interior condition, through which an impersonated self becomes recognizable in terms of individual privacy and ‘personal’ relations. Just such has been attributed to the philosophical musings of the Scottish Enlightenment (Silver 1990). So what then is the market, when it is envisaged as an arena of indifferent commerce and anonymous transactions, but the expersonation of the self via a generic calculus of self-interest? The latter's transparency, we can now argue, creates no incidental contradiction: it is ideationally separated from the opacity of the private (‘personal’) person.

Coercion. Gift exchange gives Sprenger cause to talk about cheating and failure, but does not additionally lead him to consider the coercive character of detachment or extraction,⁹ as emerges, for example, from one comparative Melanesian foray (Strathern 1988, esp. 299–305). Perhaps an aspect of obligation overlooked in the substitution of expectation was the former's tenor of compulsion. I do not mean to bring back the quasilegal language of sanctions and enforcement, which Sprenger appropriately downplays, but remark upon a particular entailment of gift exchange that arises—as he notes—from the relation of gift to giver when the giver is perceived to be detaching a part of (them)self. Elicitation of the detachable part becomes the quite deliberate act of turning the giver's attention toward the recipient. Diverse modes of seduction—so well known from Kula magic, for example—promise the hope of a coercive effect.

However, the ‘contingency’ noted in the 1988 Melanesian discussion concerns less a generic form of transaction than a generic form of action as such. Entering into relations requires counter-action, such that each party becomes a cause of the other's acts, yet elicitation is only effected under conditions of uncertainty. From the recipient's perspective, the inscrutable and chronically unpredictable giver must be compelled to give.¹⁰ Suddenly, Sprenger's analysis of asymmetry comes into its own; as he says of opacity under conditions of relational personhood, asymmetry also appears as a specific form of unpredictability. His analysis helps us see that, at the point of transaction, donor and recipient are under not just different but ‘unequal’ compulsions to act. More abstractly, if an act asymmetrically at once summons an agent and the cause of that person's taking action, that is, the source of the act (who elicited it), then uncertainty inheres in all acts: the determination of the cause or source (who is behind someone's doings?) may be as much in doubt as the determination of effect or outcome (what actually happened?). There is a measure of what we might call violence here, whether because a source is known through the agent's own detachment from it or whether because the possibility of furthering a particular relationship requires agents, however momentarily, to extract themselves from all other causes of their acting.

Time. The weight of Sprenger's article is on the productive future of an analytical framework focused on the anticipatory or prospective orientation of certain transactions. Gift exchange in particular creates an indeterminate future where anticipation can be no more than expectation. He modulates areas of stronger or weaker expectation to discriminate among diverse transactional types. But what kind of time is being imagined here?

That the future brings uncertainty is often contrasted with a past imagined as somehow fixed (it is only accounts of it that vary). However, I do not think that in old Melanesia time worked quite like that. As Sprenger says, past transactions are meant to create conditions for the future, the gift specifically connects past and future, and in any case—since expectations are generated from experiences—the future emerges from the past. I particularly appreciated how this framework molds his discussion of Rmeet (Laos) bridewealth: relations with future generations embodied by the bride and her progeny are exchanged for relations with past generations embodied by future gifts to be sacrificed to ancestors. This is generalized in terms of complementary but asymmetrical values pinned on relating to past generations or to future ones; creating the future renders wife-givers superior to wife-takers. I am not sure that for parts of Melanesia at certain epochs (including 1960s Hagen) the values could not be reversed: although in lifelong debt to wife-givers, wife-takers (recipients) could be thought of as superior in terms of the power of their wealth to seduce or extract brides. While this shows up the cultural bias of my earlier observations, the author's argument about expectation nonetheless illuminates a contingent point made decades ago (Strathern and Strathern 1969). Hagen pairings of marriages as ‘exchanges’ between clan groups, and accompanying computations of superiority, were largely post hoc evaluations of existing relations. Quite divergent accounts of what had happened (which marriages were paired, what span of grouping was involved) could be drawn on as guides for the future. More generally, where the past contains life-forces for further regeneration, everything depends on them being elicited or extracted, and thus captured for and ‘found’ in the new time to come. Along this axis, Sprenger's analysis suggests that a future orientation was also about the unpredictability of the strength or wit that people were able to summon in turning the past to future effect.

This rather arbitrary trawl of old material suggests that there is much to be learned from Sprenger's substitution. It is not that he is alone in turning toward people's futures (on the Pacific, to add just one case, there is Rollason's (2014) collection), but that he has taken the issues back to what has long been considered a foundational formulation—the imperative of obligation. There will always be reasons for revisiting the past,¹¹ but such visits are not always so generative.

I pick a moment from each brief discussion. Extending his analysis of opacity throws up the fruitfulness of the author's contrast between making visible and making transparent. The issue of coercion started from an apparently neglected contingency, but quite rapidly found reason to engage with his argument about asymmetry. Finally, the ramifications of expectation allow an interesting refinement to some of the temporal aspects of gift exchange. I feel I have benefited a lot from his reformulations. It is nice to see fresh signs of growth.

Response to Comments

Guido Sprenger

Heidelberg University

I am grateful for these inspiring comments from colleagues whose work substantially contributed to the development of my own. I take their critical insights as challenges that will develop my thought further.

Anthony Pickles asks if the notion of an uncertain future is necessary for my scheme and provides a series of apparent counterexamples. It is certainly true that expectations of the future vary greatly in strength, and quite a number of Pickles's examples—precedence, faith—are instances of rather strong expectations, for instance, that the future will resemble the past or someone else is going to make the decisions (a strong expectation regarding the person, weak expectations of the decision itself). But, first of all, I see expectations as institutions, not individual dispositions. Sharing them within a moral horizon, between givers, receivers, and witnesses, makes them work. Secondly, expectations apply to specific transactions and people are always engaged in a variety of them. Thirdly, many transactions need stabilization by additional institutions and transactions in order to meet their expectations. Foundations, for instance, are expected to install asymmetric gifting for an unlimited period of time and thus require institutions that safeguard property rights. The latter point also relates to Pickles's question regarding “what the different configurations of attributes for different taxa told us about how transactions stand apart from each other.” I take this as referring to the structure of transaction types in a given field. If, for instance, gift giving is seen as starkly contrasting with commodity exchange (as Carrier (1995) suggests for Euro-American Christmas), the divergence of their expectations and moral horizons will make them structural complements.

Anthony Pickles and Ilana Gershon both point out the conundrum of comparison that is also its achievement—it comes with a loss of specificity. However, my analytical scheme hopefully allows for scaling complexity on at least two levels of reduction. On the local level, it suggests questions that any ethnographer could operationalize: How do people classify their transactions and how do these types relate to each other? Doing so by asking for people's expectations and the moral horizon implied is going to produce rich ethnography on types of transactions that are important to people, and one may decide to stop there. On a more general level, one could argue that the relationship between Local Type A and Local Type B corresponds to the difference between gift exchange and sharing, for instance, in terms of the relative strength of expectations. Alternatively, one may consider that Type A or B could serve, with some necessary reductions, as analytical types. This opens a door between the two levels of abstraction, an aspect of my design's “generative nature,” which Strathern highlights.

I fully go along with Ilana Gershon's principal thrust that values the knowledge gained through shifting perspectives over the approximation of absolute truth. In fact, it is perhaps for the sake of contrast that she claims I am zagging and she is zigging when I am in fact zigzagging. Thus, she argues that I overestimate future orientations and deny those to the past; I am excluding transparency of personhood for my stress on opacity. However, my analytical tools are axes between opposites—one cannot provide an analysis based on a single term. For sure, the question of whether a transaction is more entailing or more presupposing is a matter of context and what is shared among actors. However, I suggest that the entailing aspect, as much as it is present in the work of Silverstein or, for that matter, the economists, has been underestimated in anthropological exchange theories. Presupposing, even being “a slave to precedent” (Pickles), requires a strong expectation that the future resembles the past. This perspective may change the way we ask questions. Gershon perhaps refers to the ever-frustrating (but nevertheless honest) answer anthropologists often get when they ask: “Why are you doing this?”—“Because we always did it this way.” I elicited more precise and actually future-oriented answers when I asked: “What would happen if you did not do it?” Regarding opacity and transparency, I consider their respective degree or relative strength as decisive. Opacity of the person—as a form of opacity of the transaction—as Strathern's comment suggests, does not just vary across ‘cultures’ but also according to the moral horizon of the given type of transaction. When Gershon says that opacity is lifted in gift exchange, she probably refers to the moment the exchange is accomplished. This is in fact my point: At least in Melanesia, while there is public speculation about the options of an exchange, any debate about the possible motives or thoughts of the givers is discouraged. This is what constructs the actual moment of exchange as revelation, an answer to the open question the gift poses.

Gershon rightly questions the totalness of the totalities I look at. Totalities exclude while being in the process of totalizing, and acting upon them requires repressing the possibility that they do not exist. I concur with Graeber (2013) who argued that totalities need to be imagined, explicitly or implicitly, for social life to proceed. For that reason, the threat that totalities may not emerge as expected may be far less severe than Gershon makes me assume it to be. However, I also admit that if Gershon, or myself, find it hard to live by the constant risk that our everyday idea of sociality may prove to be untrue, the notion of an ‘obligation’ in exchanges is reassuring.

This also addresses Joel Robbins's preference for keeping Mauss's ‘obligation.’ As Strathern observes in her comment, obligations are not opposed to expectations, but rather a category of them. I moved away from obligations perhaps because I've heard the common response to Mauss—“but what if the gift is not returned?”—a few times too often. However, Robbins adds an important caveat. Obligations allow for assigning responsibility, while this is not automatically implied in expectations. If your expectations are not met, you were perhaps uninformed or deluded. If your expectations were of obligations, some social mechanism will jump into action against those who did not meet them. This is an excellent way of differentiating moral horizons or imagined social wholes within which transactions happen, and of hierarchizing various expectations. The ties within this horizon become stronger, its outline more defined when third parties are compelled to change their own actions toward someone who does not meet their obligations. Strathern suggests a comparable argument with the language of compulsion or detachment, terms with a less individualistic ring than the language of ethics that Robbins—and I, partially—employ.

The second point Robbins raises is another one on which I agree. The notion of agency is overrated without proper consideration of passivity. For Mauss, receiving is an obligation and thus an action. But what to make of situations where people have no choice but to receive in passivity, such as the gift of life or of grace from God? This is a limiting case, in the sense that both social life and opacity of personhood meet their limits here. God as an exchange party is hard to top when it comes to opacity, and Koselleck (1989) has shown how the Church has managed the expectation of Judgment Day, when the value of all transactions with Him will be revealed. A first attempt to integrate passivity into my approach is by pointing out once more that the expectations and imaginations of the whole are hardly ever those of an individual, but emerge within sociality. The passivity of humans in receiving grace is something that needs to be acknowledged by a society of witnesses. The result is a communal moral horizon defined by the expectation that each single person is obliged to recognize this state of affairs and act accordingly. The ultimate opacity of God, however, creates a horizon in which expectations of human action are strongly posited as obligations but very weak in terms of return. This makes the Lutheran concept of the paradoxical passivity of humans in the face of God a peculiar invention.

According to Rebecca Bryant and Daniel Knight (2019), expectations are distinguished from other future orientations by indicating a degree of familiarity with a given situation. Expectations are thus strongly informed by knowledge of the past. In contrast to the concept so well detailed by them, mine is more inclusive, as I do not need to differentiate expectations from other future orientations, which is the project of their volume. My concept includes outcomes that are unlikely and less shaped by experience, but still imaginable. This is what my notion of ‘weak expectations’ tries to achieve.

However, Rebecca Bryant clarifies an important point that may be misunderstood in my argument. Types of transactions indeed do not create the content of expectations by themselves, they just provide conditions. However, without expectations people would not engage in them. Bryant points out that context is decisive, to which I respond by asking: What is structuring the context? When we cannot know the future, where do our expectations come from? Personal or collective experiences belong to it, traditions and notions of time and causality. My point is that when both observers and actors compare types or taxa of transactions, the form and strength of the expectations suggested by them is a decisive variable. For comparative and analytical purposes—purposes of abstraction, if you will—the content of expectations is less decisive than their comparative range. Or, in order to attenuate the language of agency and decision, people are drawn into the field of gravity that surrounds certain conceptual types that elicit actions from them. It may help to add that expectations are usually ranked. This ranking or hierarchizing may be structured in a variety of ways, such as desirability, appropriateness, efficiency, or probability.

I am unsure if I understand Bryant's point about the materiality of the item of exchange. This requires a rather wide notion of materiality that includes services, knowledge, thankfulness, and, not least, Mauss's “spirit of the gift” (my emphasis). If materiality includes everything that endures—and therefore carries relationships with the past—her argument works. Still this assumes a kind of fixity of the past that Strathern's comment suggests is not without alternatives. Seeing the past encapsulated in materiality is at least partially dependent on the future. Relationships with the past are often selectively invoked or suppressed when items are transferred. When jewels worth millions were stolen from the Green Vault in Dresden in 2019, one major worry was that the thieves would recut the diamonds in order to undo their eighteenth-century cutting and reduce them to their material value. Materiality and the value of heritage were played off against each other—expecting that diamonds would sell better on the illegal market with their history detached. (This, of course, involves “negative reciprocity” in Sahlins's (1972) sense.)

Bryant also suggests that moral horizons compare to horizons of expectations and that the latter are often vague. Formally, moral horizons are like horizons of expectations, as she suggests, but need to be differentiated from them. Their qualities are primarily social. For instance, they address wider or narrower social arenas. Gift giving for Christmas may be experienced as a moral horizon that encompasses only the members of a family. Buying chocolate with a Fair Trade label promising benefits for smallholders in Africa expands the moral horizon beyond one's own nation. There are more ways to analytically differentiate moral horizons, such as duration versus volatility of relationships, which I expect to explore in future publications. Moral horizons are often implicit, vague, and taken for granted by actors, but in analysis, they should be explicated as much as possible, as the context of action and communication.

Moral horizons, totalities, expectations, materiality, and the relation between expectation and obligation all open up frameworks of interrogation, due to the tensions inherent in the continua they imply. Marilyn Strathern considers the opacity of personhood in this light. She elegantly delves into some of the fissures of my argument and, to stay with her agricultural metaphor, sows unexpected insights into them. Regarding the modern concept of personhood, she rightly points out that it is not all transparent, as my argument may wrongly suggest, but only so in its relationship to the market. Its complementation, the private person, remains opaque. This is due to the person becoming an aspect of transparent, calculable market relationships or, as she suggests, the other way around, market relations becoming an expersonation of the person. This suggests a more contingent relationship between person and the imagined social wholes of my argument. I was implying that the social wholes that transactions suggest are complex enough to sustain a concept of personhood that applies to contexts even beyond the specific (sphere of) transaction. However, if concepts of personhood show different aspects in different transactional contexts, this requires some rethinking on my part as to how closely such concepts of personhood are aligned with specific types of transactions.

Finally, responding to a range of comments, my approach does not posit that the future ontologically is always entirely unpredictable. A lot of our daily transactions work because they appear to us as repetitions of tried and tested patterns. Axiomatically positing the future as uncertain just helps with theory building by making a clean sweep. It allows us to recognize a diversity of expectations in ethnographic situations without preemptively opposing certainty to uncertainty. Instead of reproducing some version of the modernist dualism of tradition versus innovation, I seek a terminology that treats both as possibilities within a continuum. The common question that has bothered so many anthropologists—this looks new, but is it really new or just another way of pursuing tradition?—would make way for a more nuanced description of the possibilities ingrained in the range of future expectations.