An overemphasis on equivalences in exchange procedures forces us to continue to depend upon theories and models based on ’norms of reciprocity.’ This ’reciprocity approach’ disregards the significance of time and analyzes reciprocal equivalences for what, in fact, may be a process of replacement. Significant wealth objects, originally given in exchange, may be reclaimed years or generations later and in this way the replacement process gives an exchange system a dimension of reproductive potential. From this ’reproductive perspective,’ death rather than marriage operates as a regulatory force, allowing for the replacement of wealth, and the long-term regeneration of social relations.
Publié in American Ethnologist 7 (1), February 1980
This paper represents an addition to a series of recent essays (Weiner 1978a, 1978b, 1979, in press) in which I introduce a perspective based on what I call a model of reproduction. Using a similar point of view here, I illustrate, first ethnographically and then theoretically, the necessity to move beyond traditional approaches that treat reciprocity and generosity as central analytical features or structures in exchange systems. I argue that in focusing on ’norms of reciprocity’ as a core principle, we continue to employ a Western construct of linear sequences basically concerned with discrete acts of giving and receiving. While it may be argued by others that the ’norm of reciprocity’ must be the primary focus having logical priority, in that without the obligation to give and to receive there would be no system of relationships, my premise is that ’norms of reciprocity’ must be analyzed as part of a larger system — a reproductive system — in which the reproduction and regeneration of persons, objects, and relationships are integrated and encapsulated.
In formulating a model of reproduction, my basic premise, which I condense from the above essays, is that any society must reproduce and regenerate certain elements of value in order for the society to continue. (Although the meanings of reproduction and regeneration have areas of overlap, I try to emphasize two distinctions, In using ’reproduction’ I am concerned with the cultural attention and meaning given to acts of forming, producing, or creating something new. ’Regeneration’ refers to the cultural attention and meaning given to the renewal, revival, rebirth, or re-creation of entities previously reproduced.) These elements of value include human beings, social relations, cosmological phenomena such as ancestors, and resources such as land, material objects, names, and body decorations. In this paper I am particularly concerned with one way in which these latter elements of value — resources — are regenerated through time. This process of regeneration, as I discuss shortly, I call replacement — a process I take to be analytically central to exchange theory.
What makes these elements of value of such important concern is that the death, decay, or loss of any of these elements is not left to chance or to nature. The physical processes of rotting and decay, as well as the processes of birth and growth, are culturally elaborated to the degree that enormous amounts of energy, time, and attention are expended in efforts to avert and/or transform the effects of deterioration and to facilitate and foster the effects of birth and growth, as well as the production of material resources. Whatever the cultural manifestations of these reproductive/regenerative concerns may be, the process is not merely automatic or habitual.
For example, some Melanesian societies marshal enormous amounts of energy in an effort to regenerate relationships shattered at death. 1n other cases, the system may not allow for long-term continuity of relationships. ln the Trobriand Islands, all of one’s wealth and even the bones of the deceased are recirculated in order to regenerate the most important matrilineal and patrilateral relationships disconnected at a death (Weiner 1976). Conversely, among the Melpa of the New Guinea Highlands, mortuary exchanges appear to be minimal, and there seems to be much more disconnectedness in affinal and patrilineal kinship relationships (A. Strathern 1972 ; M. Strathern 1972 ; and Weiner in press on an expansion of this comparison). ln the Sepik River area the activities of men’s cult houses with extensive rituals, exchanges, and presentations of masks and ancestral figures (e.g., Bateson 1958 ; Forge 1966) seem to be the point of reproductive interaction for men where both patrilineal and matrilateral relationships are regenerated through the propitiation of the deceased, now reified as ancestor.
ln each of these societies, the reproductive/regenerative sectors are organized differently, with varying degrees and attention to the running down and the building up of each kind of valued element. From this perspective we may find that disproportion exists in terms of female/male attention to these processes. Therefore, inequalities in female/male relationships (as occurs in Highlands New Guinea) may cause a system to be minimally integrated in these processes, and men alone may have to labor over and give enormous attention to reproductive areas, as evidenced through the seclusion of women, the polluting women, and male rituals from which women are excluded. ln other societies, however (the Trobriands being a good example), the complexity of the social system in terms of rank and hierarchical control may be achieved through more prominent complementarity between female and male reproductive domains (see, e.g., Weiner 1976, in press).
ln this way, societies may be seen to differ in their cultural/symbolic attention and interest in reproductive processes. But what is basic to all such systems (at present, 1 am applying this perspective only to Melanesian systems) is a cyclical world view in which the processes of reproduction and regeneration are perceived as essential cultural concerns. From this perspective, exchange interaction is reflective of the kinds of symbolic and material values a society accords its reproductive and regenerative flow. My hypothesis, as described elsewhere (see especially Weiner 1978a) is that the flow must be ’fed’ or the system (or part of it) begins to collapse. The modus operandi of this ’feeding’ is exchange.
Of special interest here will be the interplay between human life cycles and the life trajectories of material and immaterial resources. 1 am concerned with the processes of regeneration that take place in the face of contradictory processes of loss and decay. 1 emphasize that each element moves through its life cycle or trajectory organized around the cultural meaning of loss (e.g., the deliberate rotting of food, the death of a person through sorcery). Such movement is either in complement, or in opposition, to the natural length of time in which food rots and the time it takes for objects to break and disintegrate, for memory to fade so that ancestors are forgotten, and for individuals to die of old age. Thus, given the cultural problems surrounding natural loss and the equally perverse dangers caused by human intervention (such as witchcraft or theft), the processes of regeneration constitute an unrelenting attempt to counteract the constant threat of deterioration, degeneration, infertility, and eventual or immediate loss. ln this way, the system reflects the limitations imposed by biology, nature, and culture.
Situating ’norms of reciprocity’ in this larger reproductive system immediately forces a shift in perspective, making an analysis include a range of phenornena that is usually excluded from exchange theories and models, and often overlooked in ethnographic descriptions of exchange. The traditional priority given the obligation to give and to receive will simply not account for the complexity of the range of transactions within a society when it becomes apparent that although transactions may operate in totally different contexts, they still remain linked to each other through specifically defined modes of reproduction. lt also becomes necessary to account for long-range cycles of exchange interaction (see also Bourdieu 1977 on the importance of temporal aspects in exchange processes), which may involve relationships that continue for the living, beyond the lifetime of any one individual. ln addition, the nature of what is being exchanged must also be expanded to include as significant objects of exchange, v,1lued elements such as blood (see, e.g., Schneider 1968 ; Silverman 1971), social identity and yams (e.g., Weiner 1976, 1978a), semen (Kelly 1976), and even an element as seemingly insignificant as sago grubs which Ernst (1978) recently described as constituting an essential affinal exchange for the Onabasulu. Further, cosmological phenomena such as ancestors must be directly incorporated (where applicable) into the analytic framework. As I illustrate elsewhere for the Trobriands (Weiner 1976, 1977), analyses of exchange and kinship must recognize and integrate the resources and controls in the cosmological domain with those within the social domain. 1
Therefore, from a reproductive prospective, concepts associated with traditional theories and models of exchange such as ’balanced reciprocity,’ ’pure gifts,’ and ’generosity’ completely oversimplify the much deeper complexities inherent in the way transactions allow for the buildup and the embeddedness of wealth and value in others. What is given from one person to another may initially and superficially be labeled ’pure gifts,’ or ’generosity,’ but in fact, what is being given may constitute a long-range set of obligations that switch back and forth between giver and receiver through time. Further, what is given may be reclaimed (a point I return to below) at a much later time, even after both the giver and receiver have died. ln this way, models based on reciprocity ignore the significance of a reproductive frame of reference in which death, rather than marriage, operates as a regulatory force in constituting regenerative cycles of relationships. The strength, weakness, integrity, or viability of descent, kinship, and affinal relations may be more clearly analyzable in a comparative framework from the perspective of mortuary transactions rather than gifts between men at marriage.
But in the classic models of exchange, most of which emerged from Malinowski’s Trobriand studies, an exchange system is described and analyzed around the anthropologist’s observations that A gives to B, B is obligated to return something of equivalent value to A, and A must eventually make a counterpayment to B. Or, in complex systems, as defined by Lévi-Strauss, A gives to B, B gives to C, and so on until A receives a return. The linearity, even in the seemingly circular exchanges of the second example, indicates that the system maintains continuity through the flow and counter-flow of objects causing oscillation between disequilibrium and equilibrium. A particular time-space orientation is set by the emphasis placed on the return or loss of objects between individuals or groups, thereby still giving extreme weight to the obligation or return side of the exchange event (cf. Lévi-Strauss 1966). From my perspective, it seems appropriate to raise the question of whether or not the enormous energy, preoccupation, and ritual extravagance found, for example, in Melanesian societies, can be reduced to mere sets of linear ’gift’ and ’counter-gift’ situations. Do elaborate and well-articulated exchange systems function only to insure, for example, that a man replaces his sister with another man’s sister, or that the tradition of the exchange of an armshell for a necklace previously given is perpetuated ?
I refer to some very sensitive insights by A. Irving Hallowell (1955:86-89), who argued that, although our contemporary classifications of phenomena to which we give scientific attention - models based on our ’mechanical laws’ - may be equally relevant to some aspects of human behavior in societies other than our own, they may be much less salient for the individuals we study. Among the reasons for our biases, Hallowell cautioned, is our lack of historical perspective. We tend to forget that the conception of the world as ’mechanism - a world machine, no longer animate, but mechanically responsive to the ’laws of Nature,’’ only developed during the 17th century. Prior to the rise of modern science the world was conceived as something animate - alive and flourishing, yet susceptible to decay and death, mirroring the life cycle of individuals (Hallowell 1955:86-87). Hallowell’s view expresses several points germane to my approach. One of these is that we need to think about an animate world view - one in which both the life cycles of human beings, including the cosmological domain, and the life trajectories of material objects of exchange are given priority as organizing principles. Another point is that we need to give priority to a space-time framework that is structured around the reproductive and regenerative cycles culturally/symbolically defined in the societies we study. For in focusing almost total attention on the nature of reciprocity within the construct of ’gift’ and ’counter gift,’ we codify properties that seem significant to us, but in fact, as Hallowell argued, these properties may be much less salient to the individuals we study.
ln order to illustrate these points ethnographically, 1 now discuss the importance of the concept replacement, a process that has not seemed especially significant to us. ln my own experience, what seems curious about the notion of replacement is that the process was explained to me by my informants, and yet, because of Malinowski’s original description of mapula as ’equivalent, repayment,’ and the way in which mapula has been used as central to other theories of exchange, 1 almost missed what my informants were saying. Yet, the concept of replacement gives full recognition to the ethnographic reality of long-term complexes of giving, receiving, and most importantly, the act of reclaiming.
ln Chapter 6 of Argonauts of the Western Pacific, Malinowski (1922:166-194) set out the kinds of internal exchanges that occur in Kiriwina, and the relationship of these exchanges to each other and to a range of different kinds of social relationships. With the exception of kula itself, no other body of ethnographic data had such an impact on subsequent theoretical formulations of ’primitive exchange’ (see, e.g., the work of Marcel Mauss, Claude Lévi-Strauss, Raymond Firth, Karl Polanyi, Marshall Sahlins). Basic Kiriwina terms such as mapula and gimwali became common currency as key terms in various theoretical models. ln Chapter 6, Malinowski was concerned to understand Trobrianders’ attitudes toward ’wealth and value.’ 1 now proceed from the same concern and, therefore, first reevaluate the legacy of ’pure gifts’ and ’balanced reciprocity’ developed in this chapter.
The Kiriwina term mapula has played a complex role in the development of classifications and theories of ’primitive’ exchange. Malinowski described mapula as ’repayment, equivalent.’ He quoted an informant as saying, ’ ’This is the mapula for what he has done’ ’ (Malinowski 1922:182). ln another passage, mapula is defined as ’the general term for return gifts, and retributions, economic as well as otherwise’ (1922:178). An example of mapula that initially caused Malinowski confusion was the use of mapula for the small ’gifts’ (e.g., tobacco, betel) a man gives his wife and, equally, ’gifts’ (e.g., fish, rice, betel) he gives his children. From Malinowski’s observations, these appeared to be ’pure gifts’ - ’an act, in which an individual gives an object or renders a service without expecting or getting any return’ (1922:177). The problem was why a man should be giving ’repayment’ to his children to whom he was merely a ’stranger’ (cf. Weiner 1976:124-130),and why mapula should begiven to a woman for ’sexual favors.’ Malinowski (1922:179) assumed that such ’gifts’ ran counter to ’the matrilineal principle as well as to the general rule that all gifts require repayment and so these gifts were explained away by the natives in a manner that agrees with these rules.’ So, although Trobrianders called these ’gifts’ mapula, Malinowski (1922:176-177) categorized them as ’pure gifts’- a classification lacking a corresponding Kiriwina term.
The problem of mapula did not end there. Marcel Mauss (1954:71) regarded ’the mapula, the sequences of payments by husband to his wife as a kind of salary for sexual services ... [as] ... one of the most important acts noted by [Malinowski).’ What Mauss wanted to tease out of ’gifts’ such as mapula was the important idea that such ’gifts’ were not merely used as the means for exchanges of goods and services, but became the very underpinnings of alliance relationships. Mauss further questioned Malinowski’s general category of ’pure gifts’ (see also Firth 1951:123-131), sensitively perceiving that it seemed unlikely that any ’gift’ would be given without some expected return. Attending to the problem of ’pure gifts’ in a later publication, Malinowski qualified his definition, explaining that he had been guilty of ’tearing the act out of its context,’ and therefore has failed to take ’a sufficiently long view of the chain of transactions’ (Malinowski 1962:40). Malinowski went even further and amplified this position, especially as it related to what admittedly he wrongly had called the ’free gifts’ between a man and his sons and a man and his wife :
But the really correct account of the conditions - correct both from the legal and from the economic point of view - would have been to embrace the whole system of gifts, duties, and mutual benefits exchanged ... the system is based on a very complex give and take, and ... in the long run the mutual services balance (Malinowski 1962:41, emphasis mine).
What Malinowski suggested in his reevaluation was the importance of analyzing individual transactions as a system operating between participants over a long period of time (a view of time also astutely proposed by Mauss). Although concerns with balances and equivalences remained a fundamental part of the organization of such a system, Malinowski’s qualifications might have been considered a red flag of caution to those who later would pursue ideas of ’free gifts,’ or who would attach too much meaning to the mapula ’gifts’ from a Trobriand man to his wife and children. But despite Malinowski’s rejection of his earlier ideas, the original legacy of 1922 continued to play a strong role in two other theoretical directions : (1) in Claude Lévi-Strauss’ analysis of marriage exchanges, and (2) in Marshall Sahlins’ model of ’primitive’ reciprocity.
Lévi-Strauss enlarged upon Mauss’ implications of mapula to illustrate that women are one among the many objects exchanged in societies. Using the Trobriand example of mapula (also the term buwana for ’gifts’ from a Trobriand man to his female sleeping partner of the moment), Lévi-Strauss chided Malinowski for trying to formulate a functional answer where none was needed :
The lack of reciprocity which seems to characterize these services [sexual or domestic] in the Trobriand Islands, as in most human societies, is the mere counterpart of a universal fact, that the relationship of reciprocity which is the basis of marriage is not established between men and women, but between men by means of women, who are merely the occasion of this relationship (Lévi-Strauss 1969:116).
Elsewhere (Weiner 1976:195-210, 227-236 ; and especially Weiner 1979) 1 present opposing views to Lévi-Strauss’ conception of women defined as objects. My main purpose here is to show that in Lévi-Strauss’ analysis, mapula became exaggerated as a symbol of the insignificance of women and the relationships between men. 1 return to this point below.
ln developing a universal mode( of ’primitive’ exchange, Sahlins reified Malinowski’s 1922 classification of Trobriand exchange events. ln Malinowski’s (1922:176) formulation, mapula figured significantly as the center point along a continuum which was bounded at one end by ’pure gifts’ and at the opposite end by ’real barter,’ gimwali. Stressing Malinowski’s concern with the ’directness of returns,’ Sahlins expanded Malinowski’s principles by developing a mode ! applicable to ail ’primitive’ societies.
lt seems possible to lay out in abstract fashion a continuum of reciprocities based on the ’vice-versa’ nature of exchanges, along which empirical instances encountered in the particular ethnographic case can be placed. The stipulation of material returns, less elegantly, the ’sideness’ of exchange, would be the critical thing (Sahlins 1965:146).
Placing the ’sideness’ of exchanges along a continuum, as Malinowski had done, Sahlins (1965:147-149) defined ’balanced reciprocity’ (Malinowski’s mapula) as the midpoint be tween ’generalized reciprocity’ - ’the solidary extreme’ (Malinowski’s ’pure gifts’), and ’negative reciprocity’ - ’the unsociable extreme’ (Malinowski’s gimwali [barter]). Again, following Malinowski’s (1922:191-194) association between the interest and disinterest in exchanges and types of social relations, Sahlins correlated each sector of return with a particular degree of kinship distance. Thus, ’lt follows that close kin tend to share, to enter into generalized exchanges, and distant and nankin to deal in equivalent or in guile’ (Sahlins 1965:149).
But, just as Malinowski experienced problems in coming to terms with the ethnographie details of mapula, so too, Sahlins found the notion of ’balanced reciprocity’ somehow in adequate. . . in the main run of primitive societies, taking into account directly utilitarian as well as instrumental transactions, balanced reciprocity is not the prevalent form of exchange .... May we conclude that balanced reciprocity is inherently unstable ? Or perhaps that it requires special conditions for continuity ? (Sahlins 1965:178).
Rather than consider that the analytical framework of a continuum based on ’returns’ may be inappropriate in its conception, Sahlins’ solution to the instability of ’balanced reciprocity’ placed additional weight and emphasis on the ’generalized’ sector of the continuum - on the very legacy Malinowski himself finally rejected. ln this way, important kinds of transactions, including marriage exchanges, were thrown back toward ’generalized reciprocity’ - where the counter obligation ’is not stipulated by time, quantity, or quality : the expectation of reciprocity is indefinite’ (Sahlins 1965:147).
I argue that the ’expectation of reciprocity’ is realistically evaluated and calculated. Else where (Weiner 1976:211-215) 1 show that it is the very style of giving (i.e., the quality, quantity, and kind of giving) that indicates the state of a relationship at a particular moment in time. The objects themselves contain the messages of intent because future expectations and strategies should not be publicly expressed in words. What appears to have great meaning for the anthropologist (i.e., the reciprocity for the ’gift’ given) only has significance for the actors when the complex of giving and receiving is evaluated over many years.
For example, the elements of value that a Trobriand man gives his children may be separated for my purposes here into two categories : (1) those things identified with the man’s lineage (dala) and (2) those things that belong individually to the man. An ethnographic note interjected here on the meaning of dala will be useful. Although dala was glossed ’subclan’ by Malinowski and the term was continued by Harry Powell and others who reanalyzed Trobriand material, dala is not a smaller replica of clan (see Weiner 1976:37-60 for a discussion of dala and clan). For this reason I gloss dala ’lineage’ as a convenient reference, but dala is not a corporate unit of individuals who work and live together. Of importance here is that dala is a concept referring both to an internal matrilineal identity (blood) and to plots of land, hamlet sites, and other paraphernalia (e.g., body and house decorations, magic texts, dances, distinctions of rank) that were carried or claimed by the original founders (see Weiner 1977).
Therefore, the two categories above may be classified as (1) dala property - resources such as a persona ! name, certain kinds of shell decorations, the use of land and palm trees ; and (2) persona ! resources such as a new persona ! name achieved through kula participation,2 manufactured objects, kula roads, some kinds of magic texts, betel, raw yams, seed yams, and cooked food. The former group represents resources that may circulate outside dala for many years, but eventually must return to dala. Therefore, a man may give these things to his children who may use them until the property must be replaced and returned to the control of a member of the original dala. But so long as these resources remain outside dala, the user incurs obligations to his or her father and to his or her father’s kin. The latter group represents resources that contribute to the growth of the child and to the growth of social relationships with the child’s father and his kin. But as these resources are unconnected to dala, they do not return to dala. Yet the obligations of the children to members of the father’s dala may last beyond the lifetime of the father (see Weiner 1976:121-130, 141-156 ; 1978a for examples).
Therefore, Sahlins’ description of ’generalized reciprocity’ (at least in the Trobriand casel3 remains an inefficient model for recognizing the strategies and long-term significance of what appears to be a ’pure gift.’ ln general, one act alone, one set of exchanges at marriage, one ’gift’ from a man to his children, does little more than initiate a process that can only be properly assessed from a larger reproductive perspective.
Initially I too fell victim to the legacy of mapula. But while I accepted Malinowski’s definition of mapula, I was discovering many differences between what my informants were telling me and what Malinowski had written about land tenure. For example, hamlet and garden lands (dalaland) are controlled by one or two men (brothers), but the rest of the residential pattern varies somewhat from hamlet to hamlet. The basic organizational pattern, however, is a combination of men living with their wives and children in the same hamlet with older brothers, and married sons with their wives and children living in their father’s hamlets (see Weiner 1976:141-155, especially figure 11, for variations and accommodations of this basic pattern). Therefore, in terms of residence and garden land use, an ego-centered organizational pattern is intricately combined with dala-controlled continuity of land and with the accommodation of patrilateral relation ships. All members of the same dala only come together (and not always as a unified group) at the death and subsequent mortuary distributions when one of their own dala members dies.
It was in analyzing these land tenure arrangements that I first became aware of the inadequacy of using ’gift’ and ’counter-gift’ to describe certain transactions. I observed, for example, that in land tenure arrangements, plots of land could be given by a hamlet manager to someone who was a member of another dala to use. These use rights (as opposed to control rights with rights to determine future transmission) are held by the receiver until his own death. At that time the land must be returned to the controller. The return of land, however, is not automatic, and the return only comes about when the original hamlet manager formally reclaims the land by making payments of valuables to the hamlet manager of the deceased’s dala (see Weiner 1976:154-167, for examples).
These activities I originally characterized as a process of ’loan’ (see especially Weiner 1974). 1 found similar situations in many other contexts, such as when a man gives his newborn child a name from his own dala genealogy. Years later the man’s sisters make payments of women’s wealth (skirts and bundles of banana leaves) to the child and, thus, reclaim the name for reuse in the original dala (Weiner 1976:126-127 ; see Tuzin 1976 for similar processes among the Arapesh in naming procedures). ln each of these cases, and others as well, the general term for the process of reclaiming objects previously given is sagali (to settle accounts ; to reclaim control and rights to future transmission). But the general term I had for the act of giving land use or giving a name was mapula. (Each individual kind of transaction has its own separate specialized term as well.)
The problem with mapula was that, a priori, I accepted its original Malinowski definition, and I then proceeded to take its meaning for granted. The same situation might have occurred with sagali (which Malinowski glossed ’distribution’) had I not pursued the activities and exchanges in mortuary ceremonies. But if mapula meant ’repayment, equivalent,’ then how could I use it to define a concept such as ’loan’ ? Without a general Kiriwina term indicating that ’loan’ (or some variation) was a Kiriwina concept, 1 was missing a way to formalize the process as a structural feature of Kiriwina exchange.
ln retrospect, mapula was so much a part of my own exchange vocabulary that I remained deaf to what my informants were really saying to me. Although I recognized that the meaning had to be expanded to include an emphasis on future benefits between the participants (Weiner 1976:256), 1 did not explore these ideas in depth. Only when I returned to Kiriwina in 1976 did 1 begin to ’hear’ the more subtle implications associated with the word. The following is the way an informant explained mapula tome :
If my father gives me (mapula) a coconut palm and several years later a strong wind cornes and knocks down the palm, my father will give me another one. If I go to the trade store and buy (gimwali) a kerosene lamp and later the lamp breaks, do you think Mr. Holland will give me back my money ? Mapula is not the same as gimwali. If anything ever happens to that coconut palm, my father will always replace it (mapula). When my father dies, his brothers will come and give me (sagali) money [traditionally a small stone or shell valuable], and they will take the palm back. If they do not do sagali (give me a payment), 1 continue to use the palm until I die. Then someone from my father’s dala must come and do sagali (make a payment) for the palm tree. If no one comes, the palm is lost to them and my own relatives [members of my dala or my children] will get the palm.
Clearly, from this explanation, mapula represents one set in a complex series of transactions denoting more than Malinowski’s gloss of’ repayment, equivalence,’ and more than my original notion of ’loan.’ According to my informants, from the time of European contact, the term mapula included the cost of items in the trade store (but not the actual process of buying and selling [gimwali]) ; the price of a woodcarving, the amount of a salary. ln these contexts, mapula means ’payment,’ ’cost,’ and the relationship between the participants in a transaction may or may not be of a continuous nature. ln traditional contexts, however, mapula carries with it a similar notion of ’payment’ for something previously given, but it also carries with it the notion of ’replacement,’ which itself includes possibilities and expectations for future benefits.
The concept ’replacement’ in the explanation above describing the coconut palm involves first the actual replacement of the palm and, second, the replacement of the palm with another abject of value. For example, at the death of the man who gave his son a coconut palm, the palm is reclaimed (sagali) by the dead man’s successor. But it must be replaced with male valuables (e.g., stone axe blades, clay pots, cash, or shell ornaments). Other kinds of elements are replaced with female valuables (bundles of banana leaves and fibrous skirts) when a death occurs, especially the death of an important hamlet manager, where large amounts of valuables are needed in order to reclaim (sagali) each object in circulation. Land constitutes the most costly and complex resource in the reclaiming process. (For land the entire process is more complex than the above coconut palm example, see Weiner 1976:137-168.) But, fundamentally, the process remains the same - what is given must be reclaimed or it is lost. Again, a reversal in the process may occur. ln order to keep what was once given (e.g., land use, palms), a man may present valuables and yams to the halmet manager of the dala of the deceased. For example, by contributing to mortuary distributions when his father dies, a man retains rights to continue to live in his father’s hamlet and to use garden land. ln switching roles from receiver to giver within the same frame of mortuary distributions, a man ’takes his deceased father’s place.’ (These replacement payments also occur within a dala as a younger brother must contribute valuables [for sagali] to his older brother who has taken over the control of the hamlet.)
lnherent in the replacement process is that payment received has long-range expanding effects. What one individual gives to another not only creates obligations between giver and receiver, but eventually involves other kin who must use their own resources to reclaim what has been given to a nonkin - to ’others’ who are members of different dala - by someone else. Therefore, mapula transactions necessitate reclaiming actions that may impinge on the resources of individuals who had nothing to do with the initial transaction, but who must bear the burden of reclamation. The replacement process, on one hand, puts a drain on resources, but on the other hand, the process allows for long-term regeneration of intergenerational (and intragenerational) social relations, thereby giving the system a dimension of reproductive potential.
Objects are not just creating reciprocal obligations between two individuals (or groups). Rather, objects -those elements of value conceived as dala property - may circulate for many years. ln the course of circulation, social relations are reproduced, nurtured, and regenerated until finally male valuables and or female valuables secure the return. Through time some elements are lost as a resource of the original dala (via default on reclamation, theft, or compensation payments). But in general, the system enables men to control resources and to maintain some measure of power over other individuals through the circulation of dala property without depleting the property as a potential future resource.
Further, the concept of replacement applies to individuals, as well as to material objects, names, decorations, etc. For example, mapula carries significant meaning in situations when a man ’takes the place’ (kemapula) of his deceased father as he continues to live on land belonging in name to the dala of his father. Similarly, a woman ’takes her father’s place’ (kemapula) when she gives away her wealth with members of her father’s dala in mortuary distributions (see Weiner 1976:62-63, 141-153, 196-201 for details). But in each of these situations, ’taking the place’ of someone is not an automatic procedure. Each act has been preceded by explicitly defined transactions, including mapula and exchanges involving yams and women’s wealth (Weiner 1978a) which extend over decades, and switch back and forth between giver and receiver.
Even among members of the same dala succession to a chief’s position (kemapula, ’taking the place’ of the chief) or succession to land control of a hamlet is not automatic. Each hamlet manager has the right to choose his own successor from among his brothers and sister’s sons, unless only one possible heir is living. Therefore, younger men expend much effort in working for older men in order to influence the choice (see Weiner 1976:137 ff., for details). ln the case of ’taking the place’ of someone within one’s own dala, a man who becomes a chief or a hamlet manager assumes full control over all future reclaiming procedures each time a death of a member of his dala occurs. I expand on these circumstances shortly.
ln the case of the relations between a man and his children, when an individual stands in place of another, that individual then becomes part of the reclamation procedures of another dala - in the name of the dala persan he or she replaces. For example, a man ’takes his father’s place’ and he then must contribute yams and valuables to the mortuary distributions of his deceased father. From that time on, ’taking his father’s place,’ he must also contribute to any other mortuary distributions organized by the hamlet manager of his father’s dala land. ln this way, the small transactions of mapula given by a man to his children begin the process of securing for the man, and much later for his successor, additional work power and resources.
From this perspective, when an individual replaces (kemapula) another individual, the former has already embedded many elements of value in the latter. For example, the older person has given magic texts, knowledge of land boundaries and origins, raw yams, and many smaller elements such as food and betel nut. The process enables the older man to benefit from objects and tabor returned by younger individuals. But also, the process of embedding wealth in another contributes to the eventual replacement of one’s own self. Simultaneously, the process allows for a flow of material resources for the future recovery of dala property by one’s heirs, and for the regeneration of social relations beyond one’s own lifetime.
ln the case of the relations between a man and his wife, the mapula given, which Malinowski first described, must be understood as part of long-term exchange interactions also involving the production of raw yams and women’s wealth (Weiner 1978a). But neither women nor men replace each other as husband and wife in the way I described for individuals of the same dala, or for children and their fathers. When a woman, however, divorces her husband and marries another man, her new husband must present valuables, mapula veguwa, to her former husband. The transaction signifies the ’replacement’ of the man’s children. But the former husband has the right to refuse, and instead may take one of his children as ’mapula veguwa.’ Further, if in divorce a man takes his son to live with him as mapula veguwa, when the son begins to work a yam garden for his father (Weiner 1976:146-151), his father may take the harvest to his own sister and present it to her as mapula veguwa. According to my informants, these two examples are the only instances when an individual or a yam harvest is called mapula veguwa.
Therefore, the small things a man gives his wife should not be reduced to reciprocity for sexual or domestic services as Lévi-Strauss would have it. Evidence that such mapula is perceived by Trobrianders to extend throughout a life cycle and to be specifically ’reciprocated’ at some point in time can be defined during mortuary distributions. A major share of the distributions (both female and male valuables) goes to the spouse and to the father of the deceased for the ’care’ (i.e., the things given as mapula) presented by the father and the spouse throughout the deceased’s lite (Weiner 1976:61-120 for illustrations of these transactions). Further, the mapula should not only be thought of in terms of individual (or group) relationships, but the transaction also must be understood as an aspect of the larger reproductive system in which the accumulation, circulation, and replacement of elements of value, along with the buildup and replacement of individuals, only occurs through constant attention, nurturance, and ’feeding.’
The problem in Malinowski’ s early legacy was his lack of understanding of the specific kinds of obligations a man has to his wife and children (and the reversal of these obligations) as they are played out in many different contexts. Unfortunately, Malinowski’s own 19th-century notion of how a matrilineal society should function severely limited the organization of his data in many respects (Leach 1957 ; Fortes 1957 ; Sahlins 1976 ; Weiner ·1976). Yet his 1926 correction of his categorization of obligations from father to children and husband to wife as ’pure gifts’ (Malinowski 1962:40-41), certainly suggests that he was aware of the complexity of these relationships.
There exists, however, a deeper methodological problem. Malinowski’s preconceived notions about matrilinearity (see especially Weiner 1979) are not the only source of difficulty. ln all the criticisms and reconstructions of Malinowski’s data (e.g., Firth 1957 ; Fortes 1957 ; Harris 1968 ; Leach 1957, 1958), no one ever suggested the necessity of understanding mortuary transactions. Malinowski has been criticized for never producing his promised book on kinship ; no one ever discussed the absence of his book on mortuary ceremonies and exchanges (which he himself said would be larger than The Sexual Life of Savages [Malinowski 1929)). Unfortunately, by ignoring a most significant part of the life cycle, our Western notions and assumptions may completely bypass the saliency of the ’native’s point of view.’
ln Kiriwina, the most significant long-term relationships are located in the sector Sahlins (1965) would define as ’close kin.’ But the transactions that occur between ’close kin’ should be called neither ’gifts’ nor objects ’freely given’ in the sense of the nonexpectation of a return. The rules of reciprocity can be characterized as neither ’generalized reciprocity’ nor ’balanced reciprocity’ for these rules are based on mechanical models that take no account of the passage of time, the strategies played out, and the future expectations of reclaiming and replacement. ln Search for a Method, Sartre (1963:96) criticized mechanistic models because
... they never conceive of cultural behavior and basic attitudes (or roles, etc.) within the true, living perspective, which is temporal, but rather conceive of them as past determinations ruling men in the way that a cause rules its effects. Everything changes if one considers that society is presented to each man as a perspective of the future and that this future penetrates to the heart of each one as a real motivation for his behavior (emphasis in original).
The lived realities of social interaction necessitate an awareness and understanding of future perspectives -as much for Kiriwina women and men as for Sartre’s factory workers. The future is not perceived as a hazy, undefinable state. ln Kiriwina the future is marked by the events of death. Each death exposes the connections of individuals (and human life cycles) to the circulation of objects (and the life trajectories of objects). The deep interconnectedness of relationships reproduced and regenerated through the processes of replacement - giving, replacing, and reclaiming - remain rigorously held together through the perspective of death.
For example, when a death occurs, and what has been given away must be reclaimed, individual relationships such as I have been describing become a ’dala affair.’ The reclaiming process (for a range of elements) is enacted more than a hundred times over throughout the series of mortuary distributions (sagali) that periodically take place following a death. Death is the beginning of the enormously heavy (material) burden for the members of the deceased’s dala (the ’owners’ of the sagali) who now must repay the father, spouse, and all others for the care of the deceased during his or her life. The owners must also provide payments to others in order to reclaim objects originally circulated by the deceased (Weiner 1976:61-120).
A death must be understood as the moment when huge amounts of re-sorting occur : when all retrievable elements (including the deceased’s body and bones) are reordered and redefined by the dala owners. ln this respect, death triggers the return of the deceased’s body and property to the members of its own dala, followed by the concern to make the loss of the deceased (and the loss of social relations) into a positive resource for the regeneration of equivalent social relations. As I detailed elsewhere (Weiner 1978a, 1979), the basis of dala (following conception) allows for the accumulation of resources from ’others’ (first a father, and then a spouse, and extended relationships), who are members of different dala. From this reproductive perspective, by breaking the autonomy of dala and by allowing for the absorption of new contributions - resources gained from ’others’ who possess access to material and symbolic elements of value in addition to, or as an alternative to, the resources of one’s natal dala - the Kiriwina system achieves some measure of reproductive expansiveness exemplified by rank, hierarchical control, and the strength of paternal kin relations.
The regulatory force in the reproductive and regenerative cycles, however, occurs through death. Following each death enormous amounts of persona( wealth are individually drained off in mortuary distributions in the name of dala. But at the same time, property and the deceased are reclaimed in the name of dala. ln this way female and male objects of wealth are used to secure a return - i.e., a replacement of personnel, land, and even the deceased’s body -those elements associated with dala. The procedure is never simple and the transactions remain a long-term effort, in the course of which new positions in the political and social hierarchy become defined in and for the future, and in the process new relationships may eventually be reproduced.
The procedures are organized and controlled by the hamlet manager (and his most powerful sisters), but the legitimization of such control is vested in the name of dala. This is an important political point, for dala as the legitimization of rights to control, serves as access to individual power. (The rights for chiefs come from the same source - dala - as do rights for all other hamlet managers.) The maintenance of that power, however, requires constant individual attention to the demands of reclaiming, which in turn necessitates the equally constant ac cumulation and dispersal of resources. lnternally there is no way out of this system which, on the one hand has enormous reproductive potential, and on the other hand contains equally impressive limitations. Marriage and the birth of a chi Id constitute the reproductive baseline. But the beginning of a marriage and the moment of birth are only indicators of the possibilities for future reproductive capabilities in many areas. The development potential of a marriage and of a child is slow and constantly evaluated and negotiated. One’s self, one’s position, and one’s potential as a reproductive agent depend upon relations with ’others,’ because the reproductive process depends upon the circulation of possessions from ’others.’
Therefore, in order for the system to be reproductive and regenerative through time, the reclaiming actions at a death are vital. Value and wealth are embedded in ’others’ as the primary mode of self-expansion. But such a process of reproducing in ’others’ obligations, identities, and knowledge extracts its own kind of price. Without the drain on resources at a death, some individuals (such as chiefs) might be able to project themselves into a more completely autonomous, hierarchical position. But the need to reclaim one’s position (even when rightfully inherited) and to reclaim one’s control of property (even when rightfully marked as ’dala’), short-circuits any possible fully distinct political separation as we find, for example, in chiefly classes in Polynesia. Ail Kiriwina individuals must respond to the exigencies produced by a death. But the response, given the constraints of the system, becomes a significant social and political factor for the future.
Death then represents an enormous loss, yet death constitutes the major regenerating mechanism within the system. If the drain at a death tends to keep individuals within the boundaries of ’all,’ it also allows individuals great latitude to build on each other. To return for a moment to Malinowski’s (1922:166-194) discussion in the Argonauts, on the importance of understanding Trobriander’s attitudes toward ’wealth and value,’ 1 submit that the significance of such attitudes is to be found in the reproductive and regenerative advantage of embedding ’wealth and value’ in an ’other,’ in order to develop and maintain one’s own present and future value - a value that should remain forceful, operative, and influential for others beyond one’s own lifetime.
Moving beyond ego-centered linkages, however, each individual must be seen as responding to the larger social and cosmological issues of reproduction and regeneration brought into focus at every death. Therefore, individual strategy and the restraints and advantages imposed by the nature of the reproductive system are meshed together into articulating cycles through which the elements of most value (i.e., human, material, and symbolic) are reproduced, regenerated, and/or die, rot, or dissipate (see Adams 1978 on dissipative structures and energy models). Thus, cycles of exchange are regulated by cycles of death (or dissipation). ln this way, to reverse Sartre’s dictum, the future exists as a perspective of the present.
Pierre Bourdieu (1977:3-10) recently argued that in constructing mechanical models based on ’norms of reciprocity,’ the anthropologist substitutes a timeless mode ! for ’a scheme which works itself out only in and through time’ (1977:6). Bourdieu’s focus, like that of Barth (1966) and others (e.g., Kapferer 1976), emphasizes the necessity for a temporal perspective for the purpose of accounting for the working out of strategies and tactics (see also Weiner 1976). Bruce Kapferer (1976:13) points out that one of the major problems in the transactional orientation is that ’it is ego-centered, and tends to see individuals as voluntarily involved in, relatively unrestrained by, their social systems, relationships, and settings.’
Using the outlines of a model of reproduction, rather than the limited features inherent in models based on ’norms of reciprocity,’ an ego-centered perspective remains integrated with a societal perspective in which the restraints and potentialities at each reproductive sector in the system may be charted and measured against each other. ln this way, a transactional orientation becomes incorporated into a societal framework so that the individual and society are not perceived as dualities or as oppositions. The life cycles of individuals (in their biological, social, and cosmological dimensions) in articulation with the life trajectories of objects of exchange (in their natural and cultural/symbolic rates of production and dissipation) establish the temporal dimensions around which a particular societal system is structured. Within and against the restraints of these temporal dimensions, individual interaction, transactions, strategies, and motives are played out. ln this way, the spatiotemporal dimensions of social relations, cosmological determinants, and ecological boundaries operate within the context of the demands and possibilities for the reproduction and regeneration of certain elements of value.
ln summary, a reproductive mode(, unlike a reciprocity model, accounts for a wide range of transactions occurring through cosmological and social time and space. Tracing the life trajectories of objects, as well as the life cycles of individuals, a reproductive mode of analysis recognizes the significance of replacement as a central mechanism for systems with expanding potentialities. The restraints and limitations within each system will vary from society to society. But the variation in the cultural/symbolic reproductive attention and concern should indicate variations in significant sectors such as descent, rank, and hierarchy.
Acknowledgments. This paper is a revised version of the paper titled ’Reciprocity and Reproduction’ presented in the symposium ’Symbolic Aspects of Melanesian Exchange’ at the American Anthropological Association’s 77th Annual Meetings, November 1978, Los Angeles. The research on which this paper is based was made possible by the following support : National lnstitute of Mental Health, Bryn Mawr College, National Endowment for the Humanities, American Council of Learned Societies, and the University Research lnstitute of The University of Texas at Austin. 1 am very grateful for comments by Nancy D. Munn, Richard N. Adams, and Robert M. Palter on an earlier draft of this paper.
1. McKim Marriott (1976:109-142) illustrates the relevancy of integrating persona ! life cycles with the cosmological orientation of Hindu society and he uses this framework for analyzing transactions in Hindu daily life.
2. Men who participate in kula may create a name for their child which has to do with some particular kula transaction. This name will be given to a chi d in addition to names given from its father’s dala and from its own natal dala.
3. But data presented by Jill Nash (1974) on the relationships between a man and his wife and children among the Nagovisi of South Bougainville suggest that this is not just a Trobriand exception.
4. At times women are very important in the reclaiming process for land. 1 recorded cases where women took wealth objects, and as my informants said, ’saved’ the land from being lost.
Adams, Richard N., 1978, “Man, Energy and Anthropology : I Can Feel the Heat, But Where’s the Light ?”, American Anthropologist 80:297-309.
Barth, Fredrik, 1966 Models of Social Organization. Royal Anthropological lnstitute Occasional Paper, No. 23. London : Royal Anthropological lnstitute of Great Britain and lreland.
Bateson, Gregory, 1958 Naven. Stanford : Stanford University Press. Bourdieu, Pierre ; 1977 Outline of a Theory of Practice. Cambridge : Cambridge University Press.
Ernst, T., 1978 “Aspects of Meaning of Exchanges and Exchange Items Among the Onabasulu of the Great Papuan Plateau.” ln Trade and Exchange in Oceania and Australia. J. Specht and J. Peter White, eds. Mankind 11:187-197.
Firth, Raymond T., 1951 Elements of Social Organization. London : Watts.
– 1957 (ed.) Man and Culture : An Evaluation of the Work of Bronislaw Malinowski. London : Routledge and Kegan Paul.
Forge, Anthony, 1966 “Art and Environment in the Sepik”. Proceedings of the Royal Anthropological lnstitute 1965. pp. 23-31. London : Royal Anthropological lnstitute of Great Britain and lreland.
Fortes, Meyer, 1957, “Malinowski and the Study of Kinship.” ln Man and Culture : An Evaluation of the Work of Bronislaw Malinowski. R. Firth, ed. pp. 157-188. London : Routledge and Kegan Paul.
Hallowell, A. I., 1955 Culture and Experience. Philadelphia : University of Pennsylvania Press.
Harris, Marvin, 1968 The Rise of Anthropological Theory. New York : Thomas Y. Crowell. Kapferer, Bruce, ed.
– 1976 Transaction and Meaning : Directions in the Anthropology of Exchange and Symbolic Behavior. Philadelphia : lnstitute for the Study of Human Issues.
Kelly, R. C., 1976 “Witchcraft and Sexual Relations : An Exploration in the Social and Semantic Implications of the Structure of Belief. ln Man and Woman in the New Guinea Highlands.” Special Publication of the American Anthropological Association, No. 8. P. Brown and G. Buchbinder, eds. pp. 13-35. Washington, OC : American Anthropological Association.
Leach, E. R., 1957, “The Epistemological Background to Malinowski’s Empiricism.” ln Man and Culture : An Evaluation of the Work of Bronislaw Malinowski. R. Firth, ed. pp. 119-138. London : Routledge and Kegan Paul.
– 1958 “Concerning Trobriand Clans and the Kinship Category Tabu. ln The Development Cycle” in Domestic Groups. Jack Goody, ed. pp. 120-145. Cambridge : Cambridge University Press.
Lévi-Strauss, Claude, 1966, « Introduction à l’œuvre de Marcel Mauss. » ln Sociologie et Anthropologie. Marcel Mauss. pp. ix-lii. Paris : Presses Universitaires de France.
– 1969 The Elementary Structures of Kinship. Boston : Beacon Press. Malinowski, B. K., 1922 Argonauts of the Western Pacific. New York : E. P. Dutton.
– 1962 [1926] Crime and Custom in Savage Society. Paterson, NJ : Littlefield, Adams.
– 1929 The Sexual Life of Savages in North-Western Melanesia. New York : Harvest Books.
Marriott, McKim, 1976, “Hindu Transactions : Diversity Without Dualism.” ln Transactions and Meaning : Directions in the Anthropology of Exchange and Symbolic Behavior. B. Kapferer, ed. pp. 109-142. Philadelphia : lnstitute for the Study of Human Issues.
Mauss, Marcel, 1954, The Gift. London : Cohen and West.
Nash, Jill, 1974, “Matriliny and Modernization : The Nagovisi of South Bougainville.” New Guinea Research Bulletin, No. 55. Canberra : Australian National University Press.
Sahlins, Marshall, 1965, “On the Sociology of Primitive Exchange”. ln The Relevance of Models for Social Anthropology. M. Banton, ed. pp. 139-236. London : Tavistock.
– 1976 Culture and Practical Reason. Chicago : University of Chicago Press.
Sartre, Jean-Paul, 1963, Search for a Method. H. E. Barnes, transi. New York : Vintage.
Schneider, David M., 1968, American Kinship. Englewood Cliffs, NJ : Prentice-Hall.
Silverman, Martin, 1971, Disconcerting Issue. Chicago : University of Chicago Press.
Strathern, Andrew, 1972 One Father, One Blood : Descent and Croup Structure Among the Melpa People. Canberra : Australian National University Press.
Strathern, M., 1972 Women ln Between : Female Roles in a Male World, Mount Hagen, New Guinea. London : Seminar Press.
Tuzin, Donald, 1976 The llahita Arapesh. Berkeley : University of California Press.
Weiner, A. B., 1974 Women of Value. Ph. dissertation. Anthropology Department, Bryn Mawr College.,
– 1976, Women of Value, Men of Renown : New Perspectives in Trobriand Exchange. Austin : University of Texas Press.
– 1977, “Trobriand Descent : Female/Male Domains.” Ethos 5 : 150-174.
– 1978a, “The Reproductive Model in Trobriand Society.” ln Trade and Exchange in Oceania and Australia. J. SpechtandJ. Peter White, eds. Mankind 11(3):175-186.
1978b, “Sexuality among the Anthropologists, Reproduction among the ’Natives.’” Lecture presented at the Société des Océanistes, Musée de L’homme, Paris. June.