Andrew Strathern & Pamela J. Stewart, University of Pittsburgh
During the 1960s and 1970s a major theme in the anthropology of New Guinea Highlands societies centred on issues of affiliation, seen variously in terms of recruitment of groups, the claiming of identities by persons, flexibilities and shifts in such claims, and the constraints on choice deriving from demography, politics, or cultural ideas. One consensus emergent from such studies was that it is worthwhile to distinguish between ideas of descent as used ideologically and pragmatically to justify action or organize support for action, and ideas of descent used to assign or claim identity as a member of a category or group. This distinction enabled ethnographers to accommodate flexibility and choice at the level of ¡¥recruitment¡¦ with fixity of rhetoric at the level of politics or ¡¥social structure¡¦. Flexibility was therefore written in as an important characteristic of practical social life and the constitution of social aggregates in this part of the world.
Given this, it is particularly interesting to ask what has happened to these societies, so intensively studied with one set of anthropological concepts in the 1960s and 1970s, in the sphere of affiliation during more recent times. Since flexibility and choice were identified as significant factors in the earlier literature, we might expect these features to appear also today. At the same c Royal Anthropological Institute 2000.
J. Roy. anthrop. Inst. (N.S.) 6, 1¡V15 time, with cash cropping and increasing effective population density, we might expect to find a contrary process, the tightening of social practices and the formulation of restrictive ¡¥rules¡¦ governing such practices. Both practices can, in fact, be at work: individual persons may attempt to maximize their family followings, while the collectivity seeks to restrict access to resources with rules against ¡¥outsiders¡¦.
We are concerned with the analysis of a case from mid-1997 among the Kawelka people of Mount Hagen in the Western Highlands Province of Papua New Guinea.They have been the subject of intensive ethnographic work since the 1960s, and Hagen society in general was studied in the 1930s, soon after initial explorations in the Highlands, by the Lutheran missionaries Vicedom and Strauss (Strauss & Tischner 1962; Vicedom & Tischner 1943¡V8). Thus, we have by now quite a considerable depth of history in terms of which to consider the ethnography of this area. Here we show how ideas, both old and new, are brought into play or are created in novel situations, illustrating the continuing ingenuity of people in the pursuit of their aims. The overall approach we take to these materials is influenced by practice theory, as initiated by Bourdieu (1977) and reviewed subsequently by Ortner (1984), among others.
Bourdieu¡¦s approach was designed to avoid the mistakes of ¡¥determinism¡¦ on the one hand and the ¡¥occasionalist illusion¡¦ on the other. He saw social action neither as simply governed by rules nor as simply emerging out of the events of interaction. He interposed between rules and events his concept of the habitus, allowing for both the influence of culture and the play of improvisation and strategizing in social action; and he implied also the influence of political and economic considerations in the creation of social processes.The study of practice, therefore, requires attention to case histories of what people do, as well as how they explain and justify what they do. Practice theory argues for a concentration on personal agency, on lived experience and on the individual as a mediating category between structure and event. It also attempts to resolve the tension between materialist and symbolist explanations of social processes by showing how these may both apply in concrete cases (Knauft 1996: 105¡V39). Neither cultural logic nor political economy in itself accounts for social action, since personal agency is also at work, and this includes the gendered agencies of both men and women.
Many writers have recognized, explicitly or implicitly, the importance of this perspective for understanding New Guinea Highlands materials, as well as materials from elsewhere in Melanesia and further afield (e.g. Carrier & Carrier 1989; 1991: 8¡V26; Feil 1984a; 1987; de Lepervanche 1967¡V8; Scheffler 1985).The recognition followed from the realization that principles of descent were insufficient to account for the factualities of local group composition and that all aspects of life were deeply affected by gendered practices of exchange (e.g. Brown 1978; Feil 1984b; Josephides 1983; Lederman 1986;
LiPuma 1988;A. Strathern 1969; 1979; M. Strathern 1972; 1988;Weiner 1982). Further, in attempting to understand exchange practices it is important to locate the aims of the transactors as well as the overall cultural meanings of the transactions: to consider how people enter into exchanges in order to obtain certain results and in so doing may operate both within and between cultural frameworks of meaning, especially in situations of change. Indeed, 2 ANDREW STRATHERN & PAMELA J. STEWART historical change is an important context in terms of which to evaluate the inter-relations of structure and process.
Affiliations in general depend on a host of practical factors, and in particular can be handled through the medium of exchange payments, and by presentations of goods designed to influence, modify, strengthen or subvert rules derived from the transmission of substance. Ethnographies exploring these dimensions have led to a widespread recognition of degrees of flexibility in social arrangements, often arising out of warfare, shifting patterns of co-operation, the wishes of individuals of either sex, the attachment of persons to localities, and competition between leaders (e.g. Kaberry 1967; A. Strathern 1972).
Although the anthropological debates of the 1960s on kinship and descent no longer occupy the front stage of theoretical discussion in Melanesianist ethnography, conflicts about kin relations have moved into quite a central position in the contemporary lives of many peoples of Papua New Guinea, who have been exposed to massive concentrations of change since Independence in 1975, through urbanization, the growth of violence, the entry of the Christian churches, wage labour, migration, monetization and the commoditized consumer economy, as well as the effects of new political and legal frameworks. Our case study from Mt Hagen in 1997 is set against the backdrop of these changes and shows us principles of conflict in action, centring on the ambiguous affiliation of a child and attempts to resolve this.
Buying¡¦ the grandson: an attempted affiliation
R. is a leader of the Kawelka group in Hagen. His only grown-up son, H., lived and worked in Port Moresby in 1997, and had involved himself in sexual relationships with two women. One of these had a Chimbu (Highlands) mother and a Kerema (Papuan Gulf) father. The other woman was from Mendi in the Southern Highlands Province. Both women were living and working in Port Moresby. H. did not have a customary or legal marriage with either of these women. His relationships therefore reflected the fluid, transitory, inchoate character of contemporary urban liaisons (Zimmer-Tamakoshi 1995).
From the relationship with the woman whose mother was from Chimbu, a male child was born. As might be expected from this triad of relationships, there arose jealousy, violence and tensions between the individuals involved, each of whom had different aims that they sought to reach in the relationship( s) as well as conflicting life ¡¥projects¡¦. The child became a tool that the Chimbu woman could use in her attempts to gain more from H. on behalf of her family, who knew that no brideprice had been paid. When R. discovered that this male grandchild had been born he was eager to secure the affiliation of the boy to the Kawelka group and was concerned that the mother might take her son back to her Chimbu kin. His son had told him about the ambiguous relationship he had with the boy¡¦s mother and with his other partner, the Mendi woman, who was a co-worker with him at his place of employment and who was valued because of the income that she was able to bring to the relationship.
R. dispatched his lineage brother W. to go to Port Moresby and negotiate with the Chimbu woman to come to Hagen and bring the child with her. This was an effort to show her that the Kawelka kin wanted to strengthen their tie with the male child and by extension with the mother and her kin. Owing to conflict with the Mendi woman and some of her kin, who were said to be prone to violence and had come to break up the furniture in his flat, the Chimbu woman had threatened to leave H. and take her child back to Chimbu to be with her kin there. She had already been living separately from H. with her relatives in Erema, a suburban village area. Indeed, they had never set up a household together and H. had lived in a flat next to the Mendi woman. This posed a problem for R. and his kin group, who did not want the boy to learn the Chimbu language as his first language and therefore be more closely affiliated with his Chimbu kin than with the Kawelka. (The fear here was related to the indigenous Hagen theory of personhood, since children begin to acquire noman, in the sense of ¡¥social consciousness¡¦, when they first start to speak in words; Strathern & Stewart 1998a.)
In an effort to secure the affiliation of the boy with the Kawelka, a feast (mumu) was prepared and the mother, the boy and some of the Chimbu woman¡¦s female kin were brought to Hagen to be honoured and to receive money. These acts were to show them the good intent of R.¡¦s family group towards them.
Frozen beef for the feast was purchased in the town, along with greens and other vegetables and fruits.The money for these had been collected from the concerned parties. A pig was given for cooking by W. (R.¡¦s lineage brother), who had fetched the child from Moresby. This pig was to be given to the Chimbu woman and her kin while the beef was distributed to the others at the feast. The cooking pit was prepared in the style of an area to the south of Hagen,Tambul, rather than in the ordinary Hagen way because immigrants from that region were assisting with the occasion as a part of R.¡¦s support group. Many people gathered for the feast, bringing foods to be cooked and taking the opportunity to get to know the Chimbu boy and his mother.The occasion accordingly took on the character of both a community event and an experiment in inter-ethnic relations.
After the food had been cooked but before the distribution and formal consumption, speeches were made to those gathered while the Chimbu woman, her son and the female kin who had accompanied her sat on a square of fabric in the centre of the encircling crowd. The speeches were made in the local language, Melpa, and in Tok Pisin for the benefit of the non-Melpa speaking Chimbu visitors. It was notable that when speeches were made in Melpa it was said that the boy¡¦s affiliation was being ¡¥bought¡¦, while in Tok Pisin the event was not spoken of as ¡¥buying the boy¡¦. (We analyze this further below.) All who were gathered were asked to give something to the boy¡¦s mother and her kin to show that they recognized the care they had been taking of the boy and that the Kawelka also wanted to show their care for him. The largest sum of money was given by R., the grandfather of the child. The second largest sum was given by E., the full sister of H. (the father of the child). These larger sums marked the special relationship that these kin felt they had with the boy and the hope that they would 4 ANDREW STRATHERN & PAMELA J. STEWART have a lasting tie with the child as a member of their group. H., the boy¡¦s father, was not himself present and it was not clear what his feelings were about these events. (His full-time employment in Moresby may have kept him from attending. Also, because of his continued relationship with the Mendi woman he may have decided to avoid the event. His presence might have complicated the occasion by producing a tension that could have escalated into violence. By mid-1998 he was back in Hagen after some problems in his work situation, but was still not available for interviewing on the issue.)
Speakers (W. and R. mainly) said that the money was to repay the mother and her kin for looking after the child and was to indicate that the child and the mother could visit and stay for periods of time in Hagen in the future without having to worry about the bus fares to travel there. The money that was given also indicated that the Kawelka kinsfolk could go to Chimbu for the purpose of visiting the boy.
At the time of this feast no brideprice had been given, but there was a promise to pay an unspecified amount at some time in the future. It was important for the mother of the boy to hear this promise. R. specifically scheduled the question of brideprice for a later time, reversing the ordinary sequence of events, in part because a brideprice would be a much more expensive event, in part because R. could not know for sure if his son H. wanted to continue the relationship with the mother. Also, because of an inflation in brideprice payments, it would have been difficult to raise the sum needed at that time (P. Stewart & A. Strathern 1998). Many people were short of cash because a great deal of money had been recently spent to support a Kawelka candidate in a national election, who unfortunately lost. In addition, the Kawelka had also recently raised and paid out a large compensation payment (26,000 Kina and 130 pigs) to the neighbouring Mokei people for the murder of a young man in a bar-room brawl (Strathern & Stewart 1998b).
Analysis: creating difference
R.¡¦s actions in instigating this boy-affiliation occasion related to his own individual circumstances, to the position of his small lineage within its wider clan and to the political situation of his whole group, the Kawelka, within their territorial niche in 1997. Our analysis here is based on several conversations with him during our field visit of 1997 and background knowledge based on previous conversations in 1994 and 1995.
R. is a polygynist, who has many daughters and through them grandchildren, especially by his eldest wife. But he has only two sons. One is H., the absent focus of our case-history. Another son, by a different wife, died earlier while young. A third, by his last wife, is still young (of primary school age) and had to be nursed through repeated bouts of illness as a baby. In 1983¡V4 R. sponsored a major performance of the Female Spirit cult in his locality, one of the aims of which was to promote the agnatic continuity of the clan through the production of sons, so his concern over this issue is patent, and was shown when after the cult performance his youngest son was born. At that time he spent many months intensively helping to keep the baby alive.
In 1997 his concern for his only agnatic grandson was heightened further by the fact of the death of his own father earlier in the year, leaving him to contemplate the fact (as he explained to us) that now he must be the one to ensure the extension of his agnatic line through the production of descendants.
This aim was related to his lineage position. His lineage brothers who live nearby have larger numbers of adult sons who are becoming short of land for planting coffee, and R. has extensive coffee gardens tended by wives and co-resident kin. He wishes to have sons and grandsons to take over these claims and maintain them for the future, lest they be absorbed by his lineage mates. As a sponsor of the 1983¡V4 Female Spirit cult, R. achieved prominence for his lineage inside its clan, and the cult site is near to his own settlement; but this form of prestige has been on the wane for some time with the ever increasing ascendancy of charismatic forms of Christianity in the area. R.¡¦s son H. is educated and in 1997 had a paid job in the capital city, Port Moresby. He was at this time, therefore, an asset to the lineage, but he is also in general ¡¥disembedded¡¦ from his clan by virtue of his schooling and urban residence. R. indicated to us that he wanted to claim both H. and the grandson back for his group. In particular, he wished to create a set of conditions under which his son, H., would be more likely to return home and take an interest in local affairs, including his own obligation to look after R. in his subsequent old age, a theme to which R. frequently refers nowadays. R. was thus attempting to overcome the spatial separation between himself and H. by re-localizing H.¡¦s interests, by trying to ensure that H. had a stake in local concerns, since otherwise he might simply stay in the town. On the other hand, he had no wish to stop H. from succeeding in urban employment, since R. saw that as a useful potential source of monetary support for himself in future. Rather he was aiming to re-locate H.¡¦s familial concerns.
The ¡¥traditional¡¦ Hagen model involved here is that of the ¡¥men¡¦s house¡¦ and the ¡¥women¡¦s house¡¦.Within a local community, according to this idea, prevalent in the 1960¡¦s, men should build men¡¦s houses that become a focus for male sociality, while their wives¡¦ houses would be scattered around in proximity to gardens and areas of pig-pasture. But in times of change this model of centrality shifts. Men can be seen as migrants, living elsewhere, perhaps again with other men, in employment on plantations, in factories or institutions. Their women¡¦s houses now become symbols of centrality and rootedness in the areas from which they came. The question ¡¥Where is your women¡¦s house?¡¦ (niminga manga amb-nga nil morom) comes to mean ¡¥Where is your real home, the one you will go back to when you leave here?¡¦ R. was attempting to set up a manga amb-nga for H., using the pieces of culture that were at hand. Moreover, H.¡¦s sister E., along with her mother, assisted in this process in an attempt to create a woman-centred pull on H.¡¦s allegiances. Female agency, as well as male, was therefore involved in the collaborative effort to re-attach H. to his lineage by means of affiliating H.¡¦s own son to the same group. The affiliation of both was at stake, in different ways. Setting up a manga amb-nga in this way is a contemporary approach to the problems of dislocation and alienation resulting from urban education and employment. The manga amb-nga becomes a ¡¥symbol at a distance¡¦, used to 6 ANDREW STRATHERN & PAMELA J. STEWART refer to attachments back at home when someone is away from home, ¡¥not there¡¦. By contrast, idioms of male substance, co-residence and co-operation come into play when men are at home and are ¡¥in their place¡¦, exhorted to contribute to discussions and material exchanges of wealth on the basis of ¡¥being there¡¦.
In mobilizing support for this aim, R. also needed to overcome the divisive tendencies inherent in a polygynous household. In this regard the family¡¦s ¡¥misfortune¡¦ of lacking sons has been fortunate in the sense that it has minimized competition over increasingly scarce land. A great deal of the help received for the affiliation feast came from married daughters by three of R.¡¦s wives, all of whom attended and brought food and money, sharing also in the cooking of beef and its consumption. E., the immediate sister of H., gave the highest single contribution of 40 Kina (apart from R.¡¦s own 200 Kina) towards a total of 510 Kina in all. The presence of an agnatic line around which the interests of R.¡¦s wives and their daughters can cluster is essential for their own future interests, hence their support. (Daughters are indeed an asset, because bridewealth is received for them and they also return to their natal group to help their mothers in various situations, as they did here.)
The Kawelka as a whole are an enclave group, occupying a territory they had vacated in the early years of the twentieth century after defeats in warfare and to which they returned after colonial pacification in the 1950s. Their land is fertile and valuable and close to the town, allowing extensive cash-cropping and easy access to services and consumer goods. Around them are groups of a different local Government Council and parliamentary electorate. (As mentioned, they put up a candidate for election in 1996 and had spent considerable sums in campaigning on his behalf, only to fail.) They are a somewhat hard-pressed group, in need of numbers, and to this end R. was able to declare that his feast would provide the group with another male member who would be entered as Kawelka in the government ¡¥contract book¡¦ (census) for voting (political elections are an immensely important part of life in the Highlands) and tax-paying. (They themselves have become internally more complex in the last twenty years, taking into themselves immigrants from Tambul who have brought their own speech and customs, as we have noted above.)
All of these aims can be seen as ones of creating difference at increasing scales. R. wished to differentiate his agnatic line from others; to secure his landed interests; to differentiate his lineage from others in the clan; and to assist in differentiating the Kawelka from their neighbours, avoiding their absorption into the political arenas of others.
His task presented some difficulties. The affiliation of a grandson would ordinarily follow from: (1) the marriage of a son by payment of bridewealth to the mother and her kin; (2) the proximity of the bride¡¦s group, at least within the same Council or language area; (3) the residence of the wife on her husband¡¦s territory, and (4) the continuing good relations with wife¡¦s kin expressed in matrilateral payments for children. In this instance conditions 1¡V3 were absent and 4 was partially secured through this feast. H. was not married to the grandson¡¦s mother and he had not helped much to look after the baby because of his other involvements.When W., R.¡¦s lineage brother, went down ANDREW STRATHERN & PAMELA J. STEWART 7 to Moresby he had a hard time, as he narrated in Tok Pisin at the affiliation occasion, to get to see the boy¡¦s mother and to persuade her father to let her and the child travel up to Hagen to ¡¥see the faces¡¦ of H.¡¦s people and thus lay the foundations for ¡¥kinship¡¦ with them. The boy¡¦s mother, as we have seen, did not belong to Hagen, but was of Kerema (Gulf Province) descent by her father and related to the Chimbu people through her mother. Potentially, the grandson could be claimed through any of these connections; whichever set of ¡¥kin¡¦ invested most in his upbringing would be able to claim him. The mother had lived in a squatter settlement area on the Brown River road, Erema, in Moresby, and now would be taking the boy to her mother¡¦s place in Chimbu Province to the east of Hagen. The Chimbu female kin, who attended without any of their male kin and without the Kerema man, had done most of the practical work of caring for the boy and feeding him. He was already ¡¥differentiated¡¦ in their direction: the money offered them was intended partially to re-balance this situation. The food and money therefore correspond to factor 4 above, to make good relations but also to demarcate the ties involved. Lacking factors 1¡V3, R. had to rely on a new application of 4, transforming it from a single part in a flow of exchanges into a mechanism to stop the pulls of overall homogenization of ties and to instigate a contrary differentiating pull in his own direction.
The new multi-site,multi-ethnic situation in which families find themselves in places like Hagen is therefore altering the conditions for the creation of kinship as differentiated relationship rather than as homogenizing substance ties. It is transaction that works on substance, differentiating persons through wealth. This was so in the past, and then as now the opportunities for individual agency led to complexity and flexibility, but these opportunities are further compounded or confounded by the unravelling of packages of practices (such as 1¡V4 above) and the need for improvisation and re-use of parts of them. The boy-affiliation described here also represented a re-scripting of an ancient custom, symbolically linked with the origins of groups, known as kng maepokla, the ¡¥maepokla pig¡¦ gift.
Vicedom and Tischner (1943¡V8 vol. 2: 243) described this as a payment to the mother¡¦s kin of a pig and for a first-born child a shell valuable also. The father killed one of his pigs and distributed its meat to the mother¡¦s kin, whom he had invited for the occasion. Vicedom thought that the necessity to make such a payment was a relic of matriarchy.We can see it, however, simply as an expression of the aim of differentiating the child from the homogenizing influence of its substance, shared with both father¡¦s and mother¡¦s side. Vicedom went on, however, to point out also a deeper significance. The pig killed was a sacrifice, and was eaten commensally by both sides of the kinship network. The child was taken under the protection of the spirits and the living kin all expressed goodwill towards it, so that its growth would be enhanced. By implication the spirits involved are the spirits of the paternal group, a point which is made more explicit in a separate text on the meaning of sacrifice in Strauss¡¦s study of Hagen religion (Strauss & Tischner 1962: 60¡V1). Strauss writes that those who have shared a sacrificial meal are described as ¡¥fed by the originary force (mi)¡¦, or as ¡¥those who have eaten the maepogla¡¦. Maepogla refers to the act of a parent bird in giving food to its nestling and the image relates to the idea of an ancestral 8 ANDREW STRATHERN & PAMELA J. STEWART bird, typically a Harpyopsis eagle, giving meat to its young or to the first baby boy born from an egg who becomes the founder of a given tribe (see also Stewart & Strathern forthcoming). In other words, the sacrificial meal also establishes the commonality of substance of the child with the paternal kin through the mythological anaphora to the primal scene of the ancestor bird feeding its young.
The historical resonances of this sacrifice are found here also in R.¡¦s ceremony. It was thought very important that the kin meet and see one another¡¦s faces and, by implication, that all show goodwill to the child. This was a difficult feat, for the child had to be extracted from a different social matrix and brought expensively by plane to Hagen. And a pig had to be offered to the mother¡¦s kin, even though these were Seventh Day Adventist Christians who do not eat pork (in theory). The offering of a pig to the mother¡¦s group, even though beef was provided to the rest of the feast participants, marks the importance of pork as the food that comes into being because of the hard work of women who feed, nurture, and care for pigs, much as they do for children. It also represents in an attenuated form the element of the maepokla sacrifice. In this case the pig was not shared on the spot, but its recipients promised to distribute it to their kin back in Chimbu. R. thus aggregated the boy¡¦s kin on both sides, as far as possible, and disaggregated them by the differentiating payment of money. An ancient sacrificial form was pressed into service for a new and explicit piece of social engineering, made necessary by the vicissitudes of ¡¥disembedding¡¦. R.¡¦s effort was therefore directed at re-embedding and re-localizing what had been disembedded and de-localized by circumstances of change. He was also creating an interplay between homogeneity and heterogeneity that was based on maepokla logic, stretched to accommodate the contemporary world. The complexities of this situation were further reflected in the linguistic usages noted earlier.To the Chimbu kin it was not politic to declare that the payment was made as an affiliation, partly because more could have been demanded. Hence this was not stated directly in the Tok Pisin speeches made to them. Internally, however, among the Kawelka, R. had to enunciate his affiliative aim in order to explain his requests for support. The complexity of the linguistic situation was used in order to pursue subtly different rhetorical strategies within the same arena: a logic of differentiation in Melpa, and of commonality of interest in Tok Pisin.
The case history presented here lends itself to a contemporary reconsideration of issues that preoccupied ethnographers of the Papua New Guinea Highlands in the 1960s and 1970s. The narrative indicates ¡¥the perils of patriliny¡¦ in circumstances where people are disembedded from local clan life and the ways in which older persons try to ¡¥re-embed¡¦ younger ones by means of providing them with ties and obligations at home. This process has been going on ever since labour migration began in the Highlands in the 1950s, and as with other places in New Guinea (e.g. Manus, Carrier & Carrier 1989) the chief aim involved is to secure access to a flow of wealth from the urban sector back into the rural area from which migrants come.
The problems are heavily compounded when the migrants begin to set up ties of their own in the towns and engage in inter-ethnic sexual liaisons.The potential for them to be both dis-located and re-located increases exponentially in the circumstances. In our case-history, R. was attempting to reduce this complexity of H.¡¦s affairs by relocalizing him through the affiliation of his son. He was trying to pull the situation back into a frame that was local, governed by transaction, and was in accordance with patrilineal ¡¥rules¡¦. Yet these ¡¥rules¡¦ could not possibly operate in any ordinary fashion here. Hence the creative reversal of temporality and the use of a sacrifice both to mediate tensions between the inter-ethnic categories of persons at the affiliation event and to establish a tie between the grandson and the Kawelka¡¦s own sacred origins. R. was drawing on both old traditions of thinking about substance, sacrifice and affiliation, and his own contemporary understanding of the dangers of absentee sons and polyethnic permutations of ties between people. R. was using the old ¡¥rules¡¦, but he realized that the ¡¥game¡¦ was in a sense new, so he creatively reshuffled the elements in order to achieve his ends.
It is here that practice theory comes into play, since it is only when we recognize R. as a self-conscious player, working with norms and trying to match them with situations, that we can recognize his actions for what they are. Further, he was not alone in this enterprise, but was supported by some of his lineage brothers and in particular by the collectivity of his wives and daughters, all of them intent as he was to secure his lineage¡¦s land claims for the future. In symbolic terms, R. wanted the continuity of his lineage; in practical, material terms he wanted to secure his land claims. Both aims come to the same thing, since land and identity are linked.
The flexibility and creativity shown here by the actors are not new. They are expressions of the same capacities and tendencies observed by ethnographers in the 1960s. But the situation of historical change today brings the element of improvisation clearly into focus. This in turn could lead us to rethink the context of colonial control in which the ethnography of the 1960s was done as one that similarly elicited earlier responses of cultural and social improvisation following pacification and the expansion of exchange networks. Most generally, the approach to social action that we have used here corresponds closely to the idea of human action propounded at a general theoretical level by Holy and Stuchlik in the 1970s. In their Introduction to The structure of folk models, they stressed two viewpoints: first, that we must take very seriously what they called ¡¥folk models¡¦, which include models of the person. This part of their proposal seems strongly to favour the emic dimension. In their formulation folk models appear largely in the guise of knowledge, including understanding and information about rules. The second viewpoint, however, is that ¡¥the social sciences should be based on the concept of man as an autonomous, intentional and skilled agent¡¦ (1981: 26).This allows for a generative study of the emergence of social forms, somewhat along the lines of Barth¡¦s (1966) work, and takes as problematic the relationship between notions and actions, seeing that relationship as dialectical.Turning specifically to the domain of kinship they invoke Scheffler on Choiseulese practices:
¡¥Scheffler asserts that norms as such have no meaning unless they enter into transactions between individuals and groups¡¦ (1981: 27). The task, therefore, is 10 ANDREW STRATHERN & PAMELA J. STEWART to relate folk models and the performance of activities to satisfy goals and interests. Holy and Stuchlik¡¦s announced interest in seeing things in this way is also to set up comparisons between systems that reveal their differences and the reasons for these differences, rather than their similarities; but we can add here that such difference itself implies similarity in certain domains. Persons therefore instantiate and perform notions or folk models, and cannot do otherwise, even if they are breaking norms; but they also act as intentional agents with goals of their own. Intentionality in turn can be quite complex. In this vein we have discussed, both from R.¡¦s accounts and our interpretation of his accounts, the various motivations we think he had, in correspondence with different arenas of sociality, for sponsoring the boyaffiliation ceremony. The aims involved may be clear and simple, or inchoate and complex.We have been tackling here a case-history made complex by the flux of historical change in the community. R.¡¦s anticipations of results from his ceremony may prove to have had only a fragile foundation, and this because neither his agency nor that of others is encompassed within a set of binding institutional norms. Moreover, the ¡¥disembedding¡¦ processes of modernity are likely to act against his plans (cf. Gewertz & Errington 1996). What we have seen is improvisation and the creative reshuffling and rescheduling of pieces of folk models in the pursuit of a set of interests or goals.
At this point we can also learn by looking back at the literature of the 1970s with another purpose in mind. The questions we have dealt with by referring them to Holy and Stuchlik¡¦s notion-action scenario were signaled also in the title of a 1976 book, Transaction and meaning: directions in the anthropology of exchange and symbolic behavior (Kapferer 1976). In this volume, R. F. Salisbury raised an important point for contexts of change that bring together categories of people who do not necessarily share folk models or do so only partially. He first posed a very general question: ¡¥Can one understand the behavior of two individuals better by looking at each one separately, or by considering them as two poles of a dyad?¡¦ (Salisbury 1976: 41). The answer to such a question might be simply that it depends on what one is seeking to learn by using one or the other perspective, and also on whether such a distinction is itself part of the knowledge (folk model) of the actors themselves. But Salisbury¡¦s specific purpose is to apply his question to contexts of change, in which it becomes especially crucial to consider the transactors as well as the transaction (where transactor equals conscious goal-seeking agent and transaction equals a model of relationality that enters into goal-seeking action). So he asks another question: ¡¥what are the constraints on bargaining that prevent exploitative behavior. . . . when transactors are not members of a single moral community?¡¦ (1976: 45).
Salisbury¡¦s question was posed here along the lines of the formalist economic anthropology of the day, with emphasis on bargaining, but it is still relevant to events such as R.¡¦s boy-affiliation ceremony. Here, we cannot speak of bargaining in a strict sense, but of a kind of pre-bargaining, a setting of frameworks within which the participants might see themselves as bargaining within the future. For example, there was no stipulated request for an amount of money to be given and there could consequently be no telling how successful the actual amount given would be in influencing future events.
But the frameworks R. attempted to set were indeed intended as constraints on future behaviour, and his effort was precisely constructed in terms of making a semblance of moral community and giving it verisimilitude by metaphorical and metonymical means.
Looking at the event as a social performance or drama and the acts within it as intended performative acts can therefore help us to conceptualize the event itself in theoretical terms. Some of the performative elements were overt, as when the gift of money was said to set up good relations between the parties, enabling them to ¡¥see one another¡¦s faces¡¦, to produce later performances along similar lines. Other elements were more covert, but none the less cogent, in terms of the maepokla-symbolism of the pork gift. But with respect to these less obvious aspects of the communication involved, the problem of the lack of shared folk models arises. Unless the Chimbu recipients shared the same implicit logic as the Hageners in this regard, the communication would fail. Herein lies both the genius and the risk of what we have called improvisation, which occurs in contexts when there is no script and the actors are creatively cobbling together a constructed set of actions to make the script as they go along. This cannot work unless there is both the basis for achieving a structuring of elements and sufficient open-endedness for people to test their rhetorical strategies.The trope of ¡¥improvisation¡¦ therefore applies very well here, in an arena where there is less emphasis on a set of spectators who make an evaluation and more on the next anticipated performance by the actors themselves.
The affiliation drama here, which was executed in the performance modalities of paying, giving and sharing, can be further illuminated by reference to debates about gifts versus commodities in Melanesia generally. J. Carrier (1992: 185) has pointed out that in some instances ¡¥the very designation of a transaction as one of gifts or of commodities fails to capture important points of ambiguity or dispute¡¦ between the parties to the exchange themselves.As we have seen, R.¡¦s transactional acts were presented alternatively as ¡¥buying¡¦ the boy¡¦s affiliation or as simply creating a basis for shared good will between the boy¡¦s Hagen and Chimbu kin. It is worthwhile to note here also that transactions aiming at producing an affiliation can be seen both from a ¡¥commodity¡¦ and from a ¡¥gift¡¦ viewpoint, particularly in circumstances of commoditization of social relations, such as are in train in Hagen today, but also in the precapitalist contexts themselves: indeed, Gell (1992: 143) has argued that commodity exchange was an important feature of traditional Melanesian society, and along with ¡¥sharing¡¦ was distinct from gift exchange. Finally, it is clear that individuals do attempt, even if obliquely, to ¡¥bargain¡¦ over the results of transactions, both by negotiating the frameworks involved and by manipulating the amounts of goods given. R., for example, wished to make a relatively small payment to secure an affiliationclaim on the boy while avoiding or postponing the question of a larger brideprice to be paid to the boy¡¦s mother. Although the precise context and the methods of bargaining employed reflected the contemporary circumstances of inter-ethnic relations between Hagen and Chimbu, the element of ¡¥bargaining¡¦ itself is not new, since it was present at least in the manipulation of temporal sequences of transactions in the past also, so that the timing of a particular event was crucial.
Folk models in these contexts become tools whereby people attempt to define the stage for future performances. Notable here is the use of ideas about food and its distribution. Food and its giving through nurturing is an important component in Hagen folk models of kinship and identity, as it is also in many parts of the world (A. Strathern 1973). It can represent both similarity, because of the cross-cultural prevalence of ideas linking food, the body, identity and nurture, and difference, because the precise ways in which nurture is said to operate vary between cultural contexts. Pace Holy (1996: 10), it can therefore serve as a useful dimension for comparisons. The point here is that R. was attempting to use a food gift in order to promote a particular value and a claim expressed through that value. In other words, he was asserting similarity through it, performing an act that contained within itself a principle of efficacy and thus an assertion of community. But the act may have been received in quite a different spirit by the recipients, depending on how they categorized it. Douglas (1996: 128) has recently stressed that ¡¥similarity is relative, variable, and culture dependent¡¦, and if the Chimbu recipients of pork classified the transaction differently from R., his implicit intentions would neither be communicated nor would they be effective. It is likely, however, that there was sufficient similarity for a modicum of communication to have taken place. Had the act been within an already bounded moral community sharing a symbolic universe, the level of communication and effectiveness would of course have been higher. Future transactions will also depend on this same issue, as well as others, and a shared folk model will not guarantee agreement, but it will help to define the parameters within which any agreement could be negotiated. R. was also using a money gift as a bridging mechanism, but the same point applies as for pork. While both money and pork are clearly important for both Hageners and Chimbu people, what matters is how they, as transactors, saw the specifics of their transactions, and this in turn depends on both the details of their respective folk models and their intentions as these were embodied in their actions. Overall, therefore, a final theoretical conclusion is apparent: that we cannot separate transaction and meaning. ¡¥Interpretivist¡¦ analyses of exchange are sometimes counterbalanced against ¡¥transactionalist¡¦ ones, as though these were alternative frames of analysis. Our discussion here negates this putative antithesis, insisting on a synthesis that is encapuslated in a view of action which recognizes both strategizing and the creation of meanings: strategic meaning-making and meaningful strategies.
We wish to thank both the University of Pittsburgh and the James Cook University of North Queensland for financial help with fieldwork in 1997 and 1998.We thank also the anonymous reviewers of this paper for their numerous helpful comments.