1. The rapidity of this change raises an awkward question for us, as well as for any anthropologist working in a similar rapidly changing setting: just what is it that we have studied and described? This is too subtle a question to deal with here. All we can say is that we have studied and described a set of social and economic relationships that existed on a particular small island at the time we were there—and even this is a simplification, for those relationships were changing as we watched. By observing this set of relationships we have learned something about the range of possible relationships, and we are able to use this to reflect on some issues that concern anthropologists. Whether we have learned about Ponam society, or indeed even what that phrase might mean beyond a convenient designation of a particular conjunction of people, time, and place, is an issue we leave to those whose familiarity with metaphysics is greater than our own.

1. This characterization, and the discussion that follows, is a simplification, and inevitably an exaggeration, of what we see as the development of an important stream in Melanesian ethnography. This discussion is not intended to be a comprehensive review of the literature, and at every stage instances can be found that contradict our assertions. Even so, we think we have identified an important aspect of anthropology in Papua New Guinea, an aspect whose importance is not dependent on the presence or absence of alternative and even competing intellectual developments.

2. Grossman intentionally rejects a structural approach to articulation (1983, 59-60), arguing that structures themselves cause nothing, that it is necessary to consider "the strategies and aggregate actions of individuals and groups" if one is to understand articulation. While such an approach can command support, Grossman ignores structure altogether and so does not tell us how a village society, as opposed to certain individuals, articulates with the larger economy.

3. Those familiar with Gregory's work may have noted that we omitted his point that gift exchanges involve inalienable objects, while commodity exchanges involve alienable objects. The reason for this omission is that we believe this characterization is incorrect. We find much more provocative the point made by a colleague, Marc Schiltz (1987), that the two systems are distinguished not by the presence or absence of alienation, but by the point at which this alienation takes place. In capitalist economies it takes place at the point of production, and hence it is relations of production that are most elaborate and receive the most cultural attention. Alternatively, in many Melanesian gift systems, it takes place at the point

of exchange, and hence it is relations of exchange that are most elaborate and receive the most cultural attention. This sort of alienation in exchange is described in Josephides (1985).

4. A different form of economic organization occurred among women in the Goroka area who were more familiar with Western social and economic ways. This was the Goroka Women's Investment Corporation, also described by Lorraine Sexton (1983). This was a fairly straightforward women's investment group that ran small businesses around Goroka. More Westernized women apparently did not feel the need of their less Westernized village counterparts to transform capitalist institutions to suit a precapitalist cultural idiom.

1. We became particularly aware of this protective function when it was informally passed on to us shortly after our arrival in late 1978, a time when we knew remarkably little about Ponam society or the wishes and opinions of islanders. This did not appear to disqualify us in islanders' eyes, and they showed a marked tendency to stay away from our house when we had such visitors and to be uninterested in the visitors themselves and what they had to say.

2. "Luluai" is the name given to village leaders appointed by the German and later the Australian colonial administration in New Guinea. Originally a vernacular word of the Gazelle Peninsula in what became East New Britain Province, it has passed into Pidgin and Papua New Guinean English.

3. This set of asi is structurally identical to what Mead (1934, 308) and Fortune (1935, 77-79 and passim) called the tandritanitani cult. They say tandritanitani was a curse that could be cast on ego by members of the father's matriline, which they see as evidence that ego maintained a submerged membership in that matriline. Ponams denied that ego's asi formed a descent group of the sort Mead and Fortune describe, a view that is supported by the extensive application of the term.

4. Ponam kamal appear to have been agnatic in a way that many Highlands agnatic clans were not. Ponams had no sense that a person's agnatic descendents could ever change kamal in the way that members of some Highlands societies saw that a man's agnatic descendents could change clans. This sense of unyielding agnation may have been possible simply because kamal were not the only important Ponam groups (a similar point is made in Harrison 1984).

1. Schwartz's classification of production specialization in Manus (1963, 77-79) is slightly more complex than the system we prefer to use.

2. Shell money was one of the two forms of money used in precolonial Manus, the other being dog's teeth. Both were used inside and outside the market to purchase anything—vegetable food, seafood, raw materials, pigs, turtles, or manufactures—as well as to make ransom, adoption, or

marriage payments, and the like (Mead 1930, 121-122). As Ponams explained it, these were to Manus people what gold is to Europeans: intrinsically valuable and inherently beautiful. Because of this they were suitable as money itself and for making special objects of ceremonial value.

3. Establishing plantations had less effect than these three main factors. According to the report to the League of Nations in 1921-1922 (Parliament of the Commonwealth of Australia 1922), 2.4 percent of the region's approximately 1000 square miles of land was alienated, much of that on the coastal strip that had been a no-man's land in many parts of the region before colonization, and by no means all the alienated land had been planted. As far as we can tell, Schwartz (1962, 224) is broadly correct in concluding that in being colonized "the people of the Admiralties suffered little physical deprivation. Most villages lost little or no land," though obviously some villages did suffer. However, plantations provided a demand for local foodstuffs to feed laborers prior to the switch to the standard diet of rice and tinned beef and provided a demand for pearl shell and trochus. Satisfying these demands diverted labor from old productive activities to new ones.

4. We are grateful to Paul Alexander for suggesting that we think of lapan this way. Such a view parallels, of course, the basic point that Ron Brunton (1975) has applied to inequality in the Trobriands, that those who are a crucial nexus in the flow of wealth are able to secure privilege thereby.

5. The same opening up of migration and remittance that undercut intergenerational inequality also laid the basis for inequality along a new dimension within generations: between employed migrants and unemployed villagers. By the time of fieldwork this had not taken on the fairly open and exploitative nature of the earlier relationship between financier- lapan and their dependents. For reasons that we mention later, there was no particular evidence that such exploitation was likely any time soon.

6. This connection between decreasing control over wealth by bigmen, and changes in the social hierarchy and the reproductive system, are not peculiar to Manus. Thus, for instance, Andrew Strathern (1982, 313-316) points out that as access to wealth—cash income in his example—broadened in Mt. Hagen, inequalities deteriorated, both among men (as the category of "workers," kintmant , disappears), and between the sexes (as women insist on greater control over the money they help to make and begin to participate directly in moka), though as he noted elsewhere (1979, 533-534), bigmen have tried to control and appropriate new sources of wealth, just as early Pere bigmen tried to appropriate the wealth brought back by returning migrant workers (Mead 1963 [1930], 234).

1. This productivity is roughly in line with the results of a study of angling productivity by the Manus Provincial Government (1980, Appendix 7): 0.87 kg/man-hour of actual fishing time. If we assume that each hour of angling requires one-third to one-half an hour spent preparing for fishing, cleaning up afterwards, and getting to and from the fishing grounds, the range of angling productivity is about 0.6 to 0.65 kg/man-hour, broadly comparable to Ponam figures.

2. One other marine activity of importance was turtle fishing, the techniques of which are described in J. Carrier (1982). It was of some economic importance prior to World War II and even more in the precolonial period. Ponams did not, apparently, trade very much in sea turtle in this early period. However, they did hire themselves out as turtle catchers, using a technique called haliki . Turtle had an important place in prestation and feasting in early Manus, and Ponams said that they were recognized as specialist turtle fishers. Someone needing turtle would request that a haliki be carried out, providing a quantity of valuables and foodstuffs as advance payment. Total payment was increased fairly often by the fact that the operation of the haliki was thought to be harmed if the person requesting it failed to maintain correct social relations with his own kin. If the haliki ran into trouble and if it were decided that the cause lay in the actions of the person requesting it, he was obliged to make a further large payment.

3. Around 1980 the leader of Kamal Kahu, a migrant normally living in Port Moresby, tried to reclaim the south and north Tonuf rights to this technique, claiming that the gift to south Ponam kamal was revokable. Members of these kamal successfully resisted this attempt, saying the rights were given in compensation and could be reacquired only if Kamal Kahu made a generous payment to the current right holders. This payment was conceived of as a new transaction, unrelated to the original compensation except by the Kamal Kahu leader's desire to regain old kamal property.

4. In fact, three generations ago there was no spear fishing, only what islanders defined as its antecedent. This is described in J. Carrier (1981, 213).

5. While the repetition of names across generations was a common feature of Ponam genealogies, it is exaggerated in figure 2 by our decision to use birth order names wherever appropriate. For first-born to eighth-born sons, these are: Tol, Ngih, Selef, Sepat, Soon, Kupe, Kuem, and Kalai. Ponams had birth order names for daughters, but as women figure so little in figure 2, we have not used them, except where they are the way a woman is commonly remembered: Aluf Kuem and Pindriniu Kalo.

6. Associate status occurred in the case of five resident men who had immigrated to Ponam and had no formal kamal membership.

7. Spear gun fishers typically provided their own small canoes for those techniques in which they participated, and it was considered very rude for a share of the catch to be offered to the owner of a borrowed canoe. Likewise, owners of reef and sea did not request compensation for allowing people to fish in their waters either normally or when permission was needed.

1. Manus coastal fish-for-starch trade resembled closely what Richard Salisbury (1970, 176-178) reconstructs as the precolonial market trade around Vunamami, near Rabaul. It is true that for much of Manus the fish-producing villages were on islands while around Vunamami they were on the coast, but the overall logic of the two systems is quite similar

with the important exception that Manus appears not to have had a second and third tier of markets inland of the initial fish-for-starch markets. This was probably because Manus was so much smaller, less fertile, and less populous than the Gazelle Peninsula.

2. These goods in particular illustrate the point, made by Gregory (1982) among others (e.g., A. Strathern 1984, 92-95), that a single item or class of items can be a commodity, purchased from a store, and an exchange good, given in formal exchange.

1. Here we restrict ourselves to males. Positions of authority on Ponam were a male preserve, with the exception of the leader of the Nai Club, the young women's club. We eliminate the Nai Club from our considerations here.

2. As this is so different from the other tables in this discussion, it is worth spending some time explaining how it was produced. Using the life histories we gathered in our fieldwork, we constructed a migration chronology for each individual male (we did not gather these data for women, as they migrated to work less often and for fewer years than men). Our concern was for the decade they began to work and for the decade they quit work for good (that is, we ignored brief spells out of work). Then, decade by decade, starting with the earliest decade in which living Ponams entered work, we tabulated the number of Ponam men working at any time during that decade. The results of this are the various ns in table 18. For each of the men at work during that decade we computed their educational level. We used these to produce an average educational level for all workers, new workers, and retiring workers, for each decade, as shown in the table. Finally, we ranked each job on a scale of 1 to 10, 10 being the highest, and computed for each worker the mean rank of the jobs he held during the decade. From these we computed mean job ranks for all workers, new entries, and retiring workers, for each decade, again as shown in the table. Illustrative occupations and their rankings are: 1—plantation workers, cargo carrier; 2—trochus diver, factory hand; 3—ship's crew, fisherman, malaria control officer; 4—police and defence force private, domestic servant; 5—store clerk, all drivers, all self-employed; 6—primary school teacher, police and defence force NCO; 7—patrol officer, secondary school teacher; 8—hotel manager, provincial radio station manager; 9—middle-level public service; 10—senior-level public service, ordained priest.

3. Marilyn Strathern (1975, 44) argues that one should not see migration experience merely as a function of how long a set of people have been in contact with Europeans and the employment they offer—where they stand, in other words, in relation to the labor frontier. Accepting this point, it is nonetheless true that the migration experiences of groups with a similar degree of colonial experience, here prewar Ponams and Hageners around 1970, show many similarities.

4. Of course, sociological and ideological, as well as strictly agricultural, factors prevented cultivation. Even without using compost or fertilizers

available on the island, Ponams could have produced a reasonable supply of greens and a moderate number of sweet potatoes for personal consumption. Islanders tended to ignore this.

1. There was no simple reciprocal term for these asi and tama , collectively called tamatu . Formally, the reciprocal for tama was narok , and the reciprocal for asi was narohamerok , for males, and natuek , for females. Islanders commonly used the Pidgin ''pikinini'' (literally, child) as a collective reciprocal, which captures the relationship of dependence and nurturing well enough.

2. Migrants did have a range of celebrations of their own. These marked important events in the lives of their children—birthdays, first communions, graduations—and they celebrated church and state holidays. Clearly migrants were strengthening important social and symbolic elements in their lives. They did not, however, undertake the ceremonial prestations of the sort described in this chapter, prestations crucial for one's social identity as a Ponam, rather than as a parent, a citizen of Papua New Guinea, or a Catholic.

3. Certainly we would expect Ponam migrants to be more committed to the symbolic goods that home offered than were the Hagen migrants Marilyn Strathern described (1975). In contrast to Ponams, they were likely to have little or no education and were likely to migrate as an act of rejection of their social identity at home and of the system that gave that identity meaning. And Strathern notes that Hagen migrants, though obviously ambivalent, felt a burden of debt existed against them at home because of their absence, so that they did not want to go home without substantial money on their skins to discharge that debt and retrieve their social identities. This isolated them from home in a way that simply did not exist for Ponam migrants. The case of Kilenge migrants, though not as extreme, is closer to Hageners than to Ponams. These migrants were seen to contribute nothing of much consequence to Kilenge society, though there was always hope that a migrant would succeed in town. However, their migration was not an act of rejection of Kilenge society. Rather, some town polish was thought desirable among the Kilenge. Further, it appears that for many young village men, their migration and return were influenced by the wishes of their elders (Grant & Zelenietz 1980, 229) and hence are more likely to have been affirmations of their village identity than rejections of it.

1. We have other, more fundamental reservations about Meillassoux's and Fitzpatrick's work (not least the hyperbolic language in which it is couched). Fitzpatrick, for example, portrays Melanesia in terms of the two-dimensional stereotypes embodied in his chapter titles, "The Colonized," "The Peasantry," and so forth, lacking any sensitivity to the differences between the sorts of people, situations, and relationships that he condenses into his knockabout characters. This may, of course, be

inherent in any attempt to force recalcitrant real life into the neat, theoretically determined molds he uses, molds that may, in fact, be quite inappropriate (cf. Lloyd 1982). More fundamentally, Meillassoux generally ignores the complicating factor of productivity. When he does, fleetingly, consider it, he is led to admit that the greater productivity of labor in the capitalist sector may allow workers to be paid more than they would be worth in the domestic community, even while being underpaid in terms of the hypothetical cost of their reproduction in the capitalist sector (Meillassoux 1981, 128). This raises the possibility that, at least in terms of Meillassoux's economic calculus, circular migration may in fact benefit both sectors, a possibility apparent in the case of Ponam and one that would pose serious problems for the moral thrust of Meillassoux's theoretical view of the expanding and destructive nature of capitalism.