Chapter II
The Ethnographic Setting

The People of the Above

The Foi number approximately 4500^[1] and inhabit the broad valley of the Mubi River and the area to the east of Lake Kutubu, near the border of the Southern Highlands and Gulf Provinces of Papua New Guinea. The Mubi originates in the high-altitude country south of Mt. Kerewa, approximately twenty-five kilometers north of Lake Kutubu, and flows from northwest to southeast into the Kikori River (which the Foi call the Giko). Near Harabuyu village, east of Orokana mission station, it merges with the Wage River which flows into it from the north. The Mubi is also fed by the Soro River which empties from the north end of Lake Kutubu and flows southeast past the lake, parallel to the Mubi and merging with it about nineteen kilometers south of Orokana.

The altitude of the Mubi Valley ranges between 750 and 800 meters in its upper portion and drops to less than 100 meters at the point where it flows into the Kikori. The Valley is abruptly separated from the Wage River valley to the north by a range of mountains between 1400 and 1600 meters high, the Tida and Masina Ranges, and is similarly separated by a parallel ridge of lower hills to the south—the Kube Kabe and Harutami Ranges—from the interior lowlands of the Kikori basin. The entire region between the Wage and Kikori Rivers represents a series of parallel synclinal valleys and anticlinal ridges of

karstified limestone (see Brown and Robinson 1977:4) which is covered by montane rainforest in the higher altitude and by mixed sago—pandanus—Campnosperma lowland rainforest in the Mubi and Kikori Valleys (see Paijmans 1976: part 2).

The Foi are separated from their closest neighbors on all sides by large tracts of uninhabited bush. All Foi villages are located on the Mubi River or around the shores of Lake Kutubu, and the area to the north of the Mubi is used only for hunting expeditions. To the east, the Mubi is separated from the westernmost reaches of the Erave River (a tributary of the Purari) by the Go'oma River valley. Up until approximately twenty years ago, this area was inhabited by several small longhouses of Kewa-speakers from the east who had fled their homes after defeat in warfare. They were given land and political asylum by the Foi men of Kaffa and eventually left the Go'oma River entirely and settled in the southern Foi area and later with the men of Harabuyu and Yomaisi villages (Patrol Report No. 2 1955-1956; Patrol Report No. 2 1961-1962).

The Foi view themselves as comprising three main subdivisions (see Map 1). Those inhabitants of the four Lake Kutubu villages are known as the Gurubumena, "Kutubu people," or Ibumena, "lake people." Those Foi living along the Upper Mubi River are called the Awamena, "above people" or "northern people." Those Foi of the southern Mubi are the true Foi people, the Foimena. All three populations refer to their language as Foime , "Foi speech," despite minor dialectical differences. Murray and Joan Rule, the linguists who initially described and analysed the language, used the term Foi to refer to all people speaking this language, and their designation has been adopted by the Foi themselves.

The Mubi Valley occupies an intermediate position between the Highlands and Lowlands in every sense of the term—ecologically, geographically, and culturally. Along the Upper Mubi, the Awamena maintain intimate trading relations with the Kewa- and Wola-speaking groups of the Nembi River region, the Sugu Valley, and as far as Kagua and Nipa stations. The Foi collectively designate all Highlanders by the term Weyamo or Fahai . The Southern or Lower Foi by contrast maintain trading links with those groups living south of the Kikori River, now known as Kasere or Some (see Franklin and Voorhoeve 1973) and who the Foi themselves call Kewa.

The pronounced ecological and ethnic differences between the Mubi Valley and the Highlands region north of the Tida and Masina Ranges

Map 1.
The Foi and Neighboring Peoples

are translated by the Foi and Wola-Kewa respectively as a set of cultural and environmental stereotypes each group has of the other. The Foi point out the Highlanders' hirsuteness, their distinctive woven pubic sporrans, grass-roofed houses built right on the ground, their countryside consisting of much secondary grassland and casuarina groves, their skills at hunting, and their aggressive, argumentative personality as the defining features of that cultural region. The Highlanders in turn peer down into the mist-shrouded Mubi Valley and see it as a place of unremitting sickness and frightening monsters, and the Foi themselves as uncanny sorcerers and cannibals. These stereotypes seem to typify interethnic relationships between Highlanders and Lowlanders all along the interface of the southern edge of the Highlands cordillera. The Hull of the Tari region fear the sorcery and cannibalism of the Etoro further south, with whom they trade (Kelly 1977:15-16). The Chimbu-speakers of central Simbu Province for precisely the same reason fear the Mikaruan-speaking Daribi of the low-lying Karimui Plateau to the south (R. Hide, personal communication).

Southwest of Lake Kutubu in the area between the Lake and the upper portions of the Kikori River (called the Hegigio at this point) is the territory of the Namu Po people who are known as the Fasu. They are the most closely related people to the Foi culturally and linguistically. Across the Hegigio on the west bank are located the groups of the Mount Bosavi area with whom the Foi of Lake Kutubu have regular but infrequent contact and whom they collectively term the Kasua. Southeast of the Foi in the vicinity of Mount Murray live the Samberigi, speakers of the Sau language (see Franklin and Voorhoeve 1973). The Lower Foi traditionally had contact with these people and called the area Foraba. The Upper Mubi people consider the Foraba region as the source of their most important traditional healing cults, primarily Usi, as well as the direction from which the Campnosperma brevipetiolata (Foi: kara'o ) oil tree was originally introduced to the Foi by the two mythical ancestresses, Verome and Sosame.

Linguistically, the Foi language has been classified by Franklin and Voorhoeve (1973) as belonging to the East Kutubuan Family of the Kutubuan Language Stock (located within the Central and South New Guinea—Kutubuan Super-Stock of the Trans—New Guinea Phylum [Wurm 1978]). The West Kutubuan Family includes Fasu, Some (Kasere), and Namumi. Linguistically and culturally, the Fasu have as many affinities with the groups of the Mount Bosavi area as they do with the Foi, and they can be considered middlemen in the borrowing

of cultural traits from Mount Bosavi. The Foi of the Lake Kutubu villages, the Gurubumena, obtained from the Fasu the costume associated with the gisaro ceremony which Schieffelin (1976) describes as general to the Mount Bosavi area, and which the Foi and Fasu call kawari . They also performed the gisaro themselves and called the 'burning of the dancers' siri kebora ("resin burning; scar-making burning") after the tree-sap torches used. The siri kebora did not pass further east to the Awamena, although men of the Mubi villages occasionally participated in it as visitors to the Lake villagers' ceremonies. The Gurubumena also traditionally accepted Fasu immigrants into their villages and intermarried with Fasu. Two clans of Hegeso in fact trace their ultimate origin to Fasu territory, though their more immediate ancestors arrived at Hegeso from Yo'obo Village at Lake Kutubu where, in 1938, seven out of the eleven clans represented were of recent Fasu origin (Williams 1977:208).

Other cultural traits from the Mount Bosavi area were also adopted by the Gurubumena, including cannibalism and boys' initiation. The former was restricted to enemies slain in battle (cf. Kelly 1977:15). The practice of boys' initiation did not involve homosexual insemination as it did for the Mount Bosavi groups, although the Fasu traditionally practiced a modified form of it (in which the semen of older men was rubbed into the navels of immature initiates).

To the south, the Lower Foi maintain trade links with interior Gulf peoples such as the Kasere, from whom they obtain cowrie shell, and the Samberigi, from whom they traditionally obtained stone axes in addition to cowrie (Patrol Report No. 2 1939-1940). The most important trade links, however, are undoubtedly between the Foi and the much larger populations of the north. The Foi have always been producers of the oily sap of the kara'o tree, used by many Highlands peoples for decorative purposes. In return, the Foi receive shoats, pearl shells, and in pre-steel times stone axe blades and salt. As is the case with other groups with whom they have traditionally traded, the Foi have imported several cults and ceremonies from the Highlanders. Although it is no longer practiced, I suspect that the Ma'ame Gai^~ cult was in fact the Timp Stone cult^[2] of the Mendi and Nipa region (see Ryan 1961: chap. 9; Langlas, personal communication). In more recent times, the Foi have adopted the Mendi—Nipa Sa or Ya pork exchange, which they call Dawa . Unlike their own traditional Usane Habora pig-feast that was linked to the performance of certain healing rituals (see chapter 4), the Dawa is a purely "secular" ceremony for

the Foi. Its primary characteristic is the establishment of long-term rotating debts and credits in pork and shell valuables.

The residential unit of Foi society is the longhouse community (a hua , literally "house mother"). This consists of a central communal men's dwelling flanked by smaller individual women's houses on each side. The smallest such community was Kokiabo longhouse, inhabited by twelve adult men and their wives and children. The largest long-house was Damayu village with a total population of 306 in 1980. This seems to represent the upper limit of longhouse community population. Two communities—Wasemi and Barutage—divided in the recent past with one portion of each village building a new longhouse shortly after passing that limit. Hegeso had a population of 266 in 1980. At several times during my stay, men discussed the possibility of building a second longhouse that would have been used by men whose land was then at some distance from the existing longhouse.

The Hegeso men's house contains twenty-two fireplaces, eleven on each side of a central corridor, and each one nominally used by two men. The longhouse is 54.25 meters long and 7 meters wide. It is built 1.5 meters off the ground, and the peak of the roof is 4.5 meters from the floor. The women's houses (a kania ) are each approximately 8.5 meters long and 6.5 meters wide, and the height of the roof peak is a bit over 3 meters. Each longhouse has a cleared area in front and back called a wamo . The front of longhouse is that end which faces the Mubi River, conventionally designated the "upstream" (kore ) end. The back of the longhouse is "downstream" (ta'o ) and faces the bush (see figure 1).

In traditional times men built their longhouses on the tops of ridges or spurs for defensive purposes, preferably on a ridge at the river's edge so that only one end needed to be palisaded against attack. This remains the pattern today. The cleared area in which the houses are built is bordered by dense stands of multicolored cordyline shrubs and crotons (as are all dwellings) as well as pandanus, banana, hagenamo (Gnetum gnemon ), and other tree crops, and nowadays including orange, lemon, and tangerine. Each longhouse has its own canoe harbor (merabe ) along the Mubi or one of its tributaries.

Each men's house is associated politically and spatially with between two and four others which together form a distinct unit. F. E. Williams, the first anthropologist to work among Foi speakers in 1938-1939 called these units tribes (1977:171). Charles Langlas, who carried out anthropological fieldwork in Herebo Village in 1965 and

Figure 1.
Diagram of Longhouse and Women's House

1968 referred to them as regions. I have chosen to label them "extended communities." The Foi themselves lack a term to refer to these units generically but usually call each extended community by the name of one of its constituent longhouses. For example, Hegeso, Herebo, Barutage, and Baru longhouses are known as Herebo by outsiders. The four villages of Lake Kutubu are known simply as Gurubu by the eastern Foi. There are four other extended communities in the east and south.

Foi men say that these communities formed units in warfare in earlier times. While homicide and sorcery were not uncommon between coresidential men, formal warfare did not take place between longhouses of the same extended community. Likewise, these allied longhouses also comprise units in competitive feasting and exchange, and their constituent adult men consider themselves collectively responsible for amassing stocks of shell wealth and pigs for such ceremonial occasions (see also Weiner 1982b ). Foi men also recognize that each group of allied longhouses comprises an inmarrying community: Roughly half of all marriages take place between clan segments of the same longhouse, but less than 10 percent of all individuals, male and female, from an extended community marry outside of it.

History and Contact

The Australians first entered the Mubi Valley along its southernmost stretch. In 1910, M. Staniforth Smith led a patrol from Goaribari Island up the Kikori River.^[3] The patrol reached Mount Murray by foot and crossed into Samberigi territory. It then continued along the Samberigi Creek in a northwesterly direction over several more arduous limestone ridges of the Murray Range. On the twenty-fourth of December, after a descent of 1200 feet, they arrived at a large river that "ran in a fierce rapid through converging mountains, forming a gorge 1200 feet deep. . .. The only conclusion we could come to was that this was the Strickland River." (Annual Report for 1911 :166). Smith later realized that he must have been in error and finally concluded that it was the Upper Kikori.

Smith and the expedition disappeared somewhere in the interior Gulf District along the Kikori, but Wilfred Beaver, who attempted to retrace Smith's journey the following year in the hopes of finding him, noted that it was in fact the lower Mubi River to which Smith had descended. On the twentieth of March, 1911, Beaver's party reached

Smith's No. Thirty-six camp, the last one they encountered. Beaver concluded that Smith had attempted his descent of the Kikori near that point, and he himself decided to descend the Mubi River, thinking it would flow into the Turama (Annual Report for 1911 :184). After encountering the same fierce rapids that Smith described, the patrol was forced to return to their Mubi camp on April 2 and return the way they had come. Beaver later became convinced that Smith and his party perished in an attack by "natives of the Kiko and others of the up-river tribes [presumably including the Foi]" (1911:178) somewhere along the river. However, on April 12 Beaver and his party learned from an officer who had come to meet them with fresh supplies that Smith and his group had arrived down the Kikori at Beaver's base camp (1911:185). In describing the rugged limestone country of the Lower Mubi and Upper Kikori, Beaver passionately wrote, "I can safely say, after an extensive experience of the roughest country throughout the Territory, that the portion traversed is the worst" (1911:186).

In 1923, Woodward and Saunders reached the Mubi in the course of their patrol through the Samberigi Valley (see Hope 1979: chap. 5). However, the Foi were not contacted again by the Australian administration until October 1926. In that month, a Kasere man arrived at Kikori station and reported that some Foi men under the leadership of one Poi-i-Mabu had crossed the Kikori River and raided the village of Sosogo, killing all of its inhabitants save for one adolescent boy. A punitive expedition was led by Sydney Chance, the assistant resident magistrate. He was, incidentally, the first white man to bring back a description of the waterfall that lies several miles above the confluence of the Mubi and Kikori, apparently the largest in Papua New Guinea (nearly 400 feet). Chance named it Beaver Falls after its discoverer (Annual Report for 1926-1927 :8).

The patrol followed Poi-i-Mabu and his associates who fled by canoe. For the next eight days the explorers traveled along the river and footpaths, confiscating canoes and eventually taking five prisoners at a village called Udukarua before returning to Kikori station. Chance also reported that most of the Foi men were in possession of steel axes that they obtained in trade from the Ikobi and Dikima (that is, Kasere) peoples south of the Kikori River.

This was the extent of European contact with the Foi for the next ten years. In early 1936, Ivan Champion made four reconnaissance flights over the area between latitude 5 degrees 50 minutes and 6 de-

grees 60 minutes and longitude 142 degrees and 144 degrees, the area between Mount Bosavi and Mount Giluwe, in the middle of which he viewed Lake Kutubu and the surrounding Upper Mubi River valley. Later that year, he and C. T. J. Adamson ascended the Bamu River and walked from Mount Bosavi to Lake Kutubu. They visited all five of the villages that existed at Lake Kutubu at the time but did not reach any of the Mubi River villages (Champion 1940). After crossing the lake, the party proceeded northeast across the Mubi and were shown a track leading across the Augu River and into the territory of Wola or Augu speakers.

The next year, Champion and Adamson established a police camp at Lake Kutubu. Champion made a number of patrols over the next two years, primarily for the purposes of suppressing warfare. In July of 1939, the men of Ku^~ hu^~ , Era'ahu'u, Harabuyu, and Pu'uhu'u (that is, Tunuhu'u) villages raided Ifigi village, burning eight dwellings and destroying twenty-two canoes and twenty pigs (Patrol Report No. 2 1939-1940:1). A head-man of Ifigi named Baiga reported the attack and enlisted the aid of men from Hegeso, Barutage, Herebo, Harabuyu, Yomaisi, and Yomagi to retaliate against Ku^~ hu^~ . Fighting continued and twenty men from the Herebo extended community were wounded. Champion finally led an armed party to Harabuyu where there had been more killings the night before. The Harabuyu men were armed and their village barricaded. One of the constables fired his rifle and the Harabuyu men responded with arrows. Champion finally arrested the culprits and brought the prisoners back to Kutubu station.

This was the only major fight that occurred before 1940, when the outbreak of World War II forced the Australians to abandon the station. However, in the three years of their initial administration, the Australians introduced a large number of pearl shell crescents into the Foi-speaking area, using them as payment for goods and services. It was during this period that the government anthropologist, F. E. Williams, spent four months gathering ethnographic information from the Foi villages of Lake Kutubu and the Upper Mubi. Prior to the administration's introduction of large numbers of pearl shells, the Foi had apparently been trading for them with their neighbors to the north and south. Williams reported that the Foi:

[spoke] almost with awe of the fine specimens which belong to the Grasslanders [to the north of Lake Kutubu]. . .. It would appear that there is in the upland valleys a great reservoir of shell some of which is occasionally traded down towards the south. (1977:176)

After the war, a new station was established in 1949 at Dage on the northeast corner of the lake, and regular patrols of the Fasu and Foi area began. Soon after the station was functioning, the patrol of-ricer appointed a village constable and village councilor for each long-house community. These men acted as intermediaries between their constituencies and the administration, organizing the villagers to carry out tasks ordered by the patrol officer, such as the construction of latrines and rest houses, footpath maintenance, and so forth. They were also given authority to settle quarrels within the village, though unless either of the village officials were head-men before their appointment they would have had little success.^[4]

In 1951, the Unevangelized Field Mission (UFM) established missions at Lake Kutubu and at Orokana, on the Mubi River east of the present-day site of Pimaga station. By the time I arrived in Hegeso, the Lake Kutubu mission at Inu was run entirely by Foi, and the single remaining European missionary and his family at Orokana who were there when I arrived in 1979 departed permanently in 1986. The entire Foi population considers itself Christian now, and many traditional practices, of which the most important include the healing cults and mortuary exchanges, have been abandoned as a result of UFM influence. The UFM became the Asia Pacific Christian Mission (APCM) around 1970, and it is now a branch of the Evangelical Church of Papua. In the national elections of 1982, the Evangelical Alliance, an association of fundamentalist mission organizations, wielded considerable influence in the Southern Highlands Province on behalf of the candidates it supported (Ballard, personal communication).

In 1968, however, Tugiri village invited the Catholic mission to establish a station. The mission is presently staffed by a Foi man who is trained as a catechist, and the mission is visited at regular intervals by a priest from Nembi or Mendi. The influence of the Catholic mission extends to Yo'obo village and parts of Wasemi, both at Lake Kutubu, but has made no further inroads on the rest of the still solidly APCM Foi population.

Although the Foi have had the longest history of permanent contact with the colonial administration of any group in the Southern Highlands Province, they have experienced only modest economic development. In 1959, the colonial government organized work on a footpath linking all the Upper Mubi villages which was completed in 1963. During this time, many villages, including Hegeso and Barutage, rebuilt their longhouses on sites along the footpath. By the time I arrived in the Foi area, the path had been upgraded to vehicular standards.

There are at present three airstrips serving the Foi area. Two are owned by the APCM mission and are located at Moro, west of Lake Kutubu, and at Orokana. The government airstrip at Pimaga was completed in 1974. The mission also runs two primary schools at Inu (Lake Kutubu) and Orokana. In 1974, the government opened Tanuga primary school located near Pimaga station.

Government aid posts were established throughout the Foi and Fasu area in the late 1970s. A larger health center, staffed by a provincial health extension officer, was opened at Pimaga in 1977.

Australian patrol officers introduced new food crops to the Foi at the very beginning of their administration. The most important among them were the Xanthosoma or Singapore taro, the Cavendish banana, choko, pumpkin, maize, cassava, pawpaw, citrus fruits, pineapples, and peanuts. In addition, birdseye chilies and Bixa orellana were introduced for their commercial value.

Until 1979, the possibilities available to the Foi for obtaining money were limited. The most common method until recently has been contract labor at coastal plantations. Twenty of the men of Hegeso had by 1979 experienced at least one two-year term of labor, mostly on the cocoa and copra plantations of New Britain and the north coast. The sums that these men are able to save, however, are low and are primarily dissipated in gifts to relatives.

Men and women earn small amounts of money by selling fruit and vegetables to government and mission employees at Inu, Orokana, and Pimaga. Both the government and mission employ a certain number of men as laborers. During the mid-1970s, the Southern Highlands Province allocated funds under the Rural Improvement Programme (RIP) for the maintenance of roads, and a number of young men obtained regular employment in this manner (see Patrol Report No. 3 1974-1975). When I returned to Hegeso in December 1982, however, funding for local road maintenance had been discontinued.

Most men have obtained some income through the selling of chilies to government buyers. However, due to the lack of agricultural extension by government officers, the availability until recently of RIP funds for local labor, and the decline in chili prices, the production of chilies diminished considerably by the time I arrived (see Patrol Report No. 1 1975-1976).

The APCM mission has had considerable success starting cattle projects in Foi villages. Because of the relatively high costs of purchasing calves and fencing materials, most cattle are purchased cooperatively by an entire village or clan segment. The Foi view cattle

in exactly the same terms as they do pigs. In fact, they call them nami kau , which can best be translated as "the type of pig whose name is cow," just as they call the white cockatoo, for example, the ya namuyu , "the namuyu bird." The owners of cattle slaughter them to be sold piece by piece for shell wealth (and nowadays cash), the traditional manner in which they dispose of their own pigs, as I describe in chapter 4.

The most significant and potentially successful attempt at development in the Foi area, however, began in 1979 when the Department of Commerce introduced silkworms. Both the climate and the abundance of riverine garden sites in the Mubi Valley are well suited to the growth of mulberry shrubs. At the end of that first year, Foi growers had produced 1374 kilograms of raw silk, and by the end of 1981 the amount was 4022 kilograms. During the initial three-year period of the silk project's existence, Hegeso village led the Foi area in production, and in 1983 it had nineteen individual projects. The price for raw silk varied between K2.00 and K2.50 per kilogram (PNG K1.00 = U.S. $1.47 in 1979), and it was thus comparable to coffee in its potential for profit.^[5]

The money that the Foi earn from these various means is primarily channeled into bridewealth and other traditional payments, although money still accounts for only a small proportion of their total value. The Foi also spend money at the trade stores: these purchases are largely confined to tinned fish and rice, which they give to guests on ceremonial occasions such as weddings. Although the introduction of cash has not as yet had a significant effect on the composition of traditional prestations, other factors to which I now turn have altered them considerably.

Bridewealth

In the ten years between 1950 and 1960, the Foi and Fasu experienced a great inflation in their level of ceremonial payments, primarily bridewealth. This was a direct result of the large numbers of pearl shells introduced to this area and to neighboring Highlands regions by the Australian administration. In 1957, the patrol officer was approached by several village constables and councilors who requested him to set an upper limit of twenty pearl shells for all bridewealth payments. Men complained of betrothal payments ranging up to thirty pearl shells and subsequent bridewealth payment of forty or more

pearl shells plus an additional commensurate amount of cowrie shells in each case. The patrol officer reported:

Previously the initial bride price has consisted of somewhere around 40 items, but over the years if the wife has words with her husband and goes home to father, it is expected of the husband that he brings more items as compensation for the father's hurt feelings. In other words, under the present system, payment goes on for years. (Patrol Report No. 2 1956-1957)

By 1959, the head-men and appointed officials had apparently discussed the problem. They agreed to an upper limit of twenty pearl shells for all first marriages and to make the betrothal payment deductible from this total (Patrol Report No. 1 1959-1960). But the attempt was mostly a failure as they had no means with which to enforce their will. The patrol officers of this period noticed a ``large number of unmarried males" in the Fasu villages due to their inability to raise the necessary amount of wealth. It was reported that in the Fasu area "there is no basic payment, and the rule seems to be to get as much as possible" (Patrol Report No. 1 1960-1961). The same year, the patrol officer wrote a disparaging report on the high incidence of "child marriage" among the Foi, complaining of the frequency of death in childbed of immature mothers. Apparently, this was also linked to inflating bridewealth levels, for it was primarily the wealthiest men who were marrying immature girls as second and third wives (Patrol Report No. 2 1961-1962), and the girls' fathers were agreeing to it rather against their will in order to raise the wealth for their own obligations. "The universal distribution of such . . . complaints lays the basis of the basic economy of these people," the patrol officer wrote. "From the youngest child to the oldest man or woman everyone is in debt to someone else . . ." (1961-1962).

When I arrived in Hegeso, each village had appointed a special komiti to represent them in matters of bridewealth. On November 5, 1980, in a meeting of these komitis and head-men from the entire Mubi area, it was agreed that the bridewealth should be fixed at 37 pearl shells (of which no more than 11 would be of the largest size, worth K40.00 each), 111 ropes of cowrie shells, and 2 bridewealth pigs (no rules were set concerning the cooked forest-game portion of the bridewealth, the ka aso [see chapter 5]). It was also agreed that a separate payment of shells and/or cash to the amount of K100.00 would be made to the father and mother of the bride which would not be subject to redistribution to relatives of the bride's parents. In order to enforce this standardization, the head-men further established

that one of the bridewealth komitis would be present at each distribution to make sure that no "secret" payments above the set level were made.

I have gone to some lengths to describe the changes in Foi ceremonial-payment levels since the time of contact with the Australian administration because, as I discuss in chapter 5, Foi social categories are primarily embodied in the Foi's own rules concerning the distribution and collection of such payments. The oldest men of Hegeso and neighboring longhouses were able to remember quite clearly what ceremonial payments were like in their youth before the advent of large numbers of pearl shells. Some men claim to remember when pearl shells were so scarce that they were not a part of payments at all. Williams described a bridewealth transfer involving a man from Gesege and a woman from Fiwaga in which the total number of items given was "apparently seven pearl shells, seven strings [of cowrie], four shell frontlets, six steel axes, two trade knives and three pigs" (1977:222). Current bridewealth levels are not only higher in real value, they are also far more divisible—they consist of at least 148 separate items, and hence can be distributed amongst more people.

I thus wish to discuss rising bridewealth levels along with what was evidently a rapid increase in the Foi population over the last forty years. Unlike the groups of the Mount Bosavi area which experienced drastic reductions in population as a result of epidemics of measles and influenza (see Kelly 1977:28-31), the Foi population, although it experienced short-term declines in population levels due to introduced disease, rose steadily after contact. The census of the Upper Mubi villages (that is, excluding the Lake and Lower Foi villages) in 1950 accounted for 1638 people. The patrol officer reported that deaths totaled only 16, or less than I percent of the population, and he estimated that in some villages over 50 percent of all women were pregnant (Patrol Report No. 1 1950-1951). Another count carried out the next year accounted for 1119 individuals in the Upper Mubi villages and 852 at the Lake, showing that one set of figures was grossly inaccurate, probably the first one. In 1954, it was reported that between 70 and 80 people died of pneumonia and influenza epidemics in the previous two years and that 84 children between the ages of birth and four years old died in that same period. Altogether there were 208 births against 248 deaths in the two years prior to the 1954 patrol (Patrol Report No. 2 1953-1954). By the last census held in 1980, the population was 4030 for all Foi-speaking villages, an increase of approximately 100 percent over thirty years.

A senior head-man of the Orodobo clan once remarked that when he was a young man there were many fewer people, and that distantly related consanguines and agnates could be included in bridewealth and death distributions. Now, however, because there are more people, relatives related more distantly than cross-cousins or father's brother's children are usually not included.

In one sense, this is a misleading characterization. Bridewealth networks have obviously expanded rather than contracted since, with rising levels of bridewealth, men have been forced to seek contributions from more distantly related and unrelated men. What the head-man therefore meant (as I learned subsequently) was that distant relatives no longer receive a share of bridewealth automatically ; they must first have contributed to the donor at a previous time. This norm of reciprocity has, however, been extended to "close" relatives also: for the Foi, a relative is no longer considered a close one solely because of genealogical propinquity and the coresidence it usually implies. Such ties must be validated by material aid in raising ceremonial payments, primarily bridewealth. The exigencies of exchange and inflation have therefore seemingly served to attenuate the importance of consubstantial kinship among the Foi. In chapter 5, I describe the nature of such ties of reciprocity as define the Foi kinship universe.

Subsistence

The Foi subsistence economy, with the exception of sago production, is primarily based on seasonal variations that regulate the availability of certain animal and vegetable species. The major Foi productive activities in their order of importance are: sago making, permanent tree-crop cultivation (chiefly Marita pandanus and breadfruit), slash-and-bum gardening, foraging (mostly for wild vegetable species), fishing, and hunting.^[6]

The Foi of Hegeso recognize almost forty different varieties of sago palm,^[7] but they say that only one was native to their portion of the Mubi Valley (though there are myths accounting for the autochthonous origin of this and two other species). The other varieties are said to have been imported from areas further south. All sago is subject to management. There are no "wild" palms, and the clearest proof of this can be had by flying over the Mubi Valley and observing that not a single palm is in flower (the flowering of the sago palm corresponds to a drastic reduction in its starch content rendering it unusable).

Besides providing the staple starch flour, the Foi also use the leaves,

midrib, and bark of the sago palm in the construction of houses. The base and extreme top of the palm trunk are characteristically more fibrous than the midsection and hence are difficult to process for flour. For this reason, these portions are either left for pigs to forage or for the formation of sago grubs. Unlike the people of the Mount Bosavi area, the Foi do not reserve entire palms for sago-grub production. Although the Foi prize them highly, grubs do not constitute an item of important ceremonial exchange among the Foi as they do for the Onabasulu, Etoro, and Kaluli for example (see Ernst 1978).

A single palm produces between ten and twenty suckers during its lifetime and matures in about fourteen to twenty years. Men transplant suckers to prevent palms from becoming too crowded in any one grove and in order to utilize all the appropriate swamp areas they own or acquire access to. In this manner, an adult man usually owns between five and fifteen groves in different areas in the vicinity of the longhouse and one or two at Ayamo, the hunting area to the north.

Seasonal variation in the Mubi River area is dominated by the onset of the southeasterly trade winds that begin in May and last until October (Brookfield and Hart 1966:9). In the eight-year period between 1956 and 1963, Lake Kutubu station received its highest mean rainfall between the months of August and October (1966: table 1). The driest months, by contrast, are November to January. The Foi themselves recognize five seasons based on the appearance of certain seasonal plant species. Between mid-December and mid-January, the Marita pandanus (abate ) ripens and it becomes the single most important food for this time. This is the abate base or "pandanus season." It is followed by the ripening of the Saccharum edule pitpit (gabia ) between February and mid-March, and the consumption of this vegetable eclipses all others, so that this season is called the gabia base . Following this, between mid-March and mid-April, the fruit of the waria tree begins to ripen. The Foi use this fruit in the construction of deadfall traps and the onset of waria hase coincides with an isolated period of increased rainfall in April. The hunting season however does not begin in earnest until the ba ^~ ngo fruit ripens in late April, commencing ba^~ ngo hase. This fruit is also eaten by forest animals that begin coming down from the higher-altitude montane forest in search of this and other tree fruit. The end of ba^~ ngo base in July ushers in the main rainy season lasting until late November or early December and called the me hase , "bush season," or the kagi hua hase , "mother of rain season." Traditionally during this season the Foi would abandon the

longhouse village, and each man along with his wives and children would occupy individual nuclear-family bush houses at Ayamo, the hunting preserve located in the valley of the Yo'oro River and the foothills of the Tida and Masina Ranges directly north of it (see Map 2). At this time, men set deadfall traps and snares and collected bush-fowl eggs. While at Ayamo women processed sago planted there for the purposes of consumption.^[8]

Nowadays, with the responsibilities the Foi have to send their children to local primary school, attend church, engage in Local Government Council work projects, and tend cash crops, the amount of time they can spend at Ayamo has been drastically curtailed. Younger unmarried men tend to spend the most time hunting these days, but all Foi agree that most families in traditional times spent between three and six months of every year at Ayamo.

With the cessation of the southeasterlies in December and the ripening of the Marita pandanus, the Foi used to return to the Mubi Valley to subsist on pandanus sauce for the next two months and to prepare new gardens during this relatively dry period. Again, the current gardening system differs profoundly from that which characterized pre-steel times. Older men told me that in the days when steel implements were rare, all gardens were made communally. A man who possessed effective gardening magic would assume the leadership of the other men and plant the first sweet potato (Williams 1977:198; Patrol Report No. 3 1953-1954). This man was known as the kusaga , literally the "spell-base" (ga : base; root; cause; origin). By 1955, however, Langlas reported that the men of Herebo village were making individual rather than communal gardens, due to the relative efficiency of steel axes (1974:27). The magic spells associated with garden construction have also become confined to private use as a result of mission influence. In chapter 3 I describe the differential roles of men and women in gardening and sago making.

Foi garden produce is supplemented by a wide variety of uncultivated fruits and vegetables of which the most important are several species of ferns and leafy greens. Unlike the groups of the Mount Bosavi area who cultivate sweet potato as a staple along with sago, the only Foi staple is sago and it is made all the year round.

Besides Marita pandanus, the Foi depend heavily on the seasonal production of other permanent tree crops: breadfruit; ko^~ ya (a Bambusa species that bears an edible inflorescence similar to Setaria palmifolia ); and the new leaves of hagenamo , the tree from which the

Map 2.
Hegeso Village and Surrounding Area

Foi also obtain the underbark used in making string for net bags. These foods all ripen during the relatively dry months of the year between November and April. Each one assumes temporary importance as it appears, becoming the single most commonly eaten vegetable along with sago for a one- or two-month period.

The remaining important Foi subsistence activity is fishing, which is also seasonal to some extent. While the major rivers such as the Mubi, Baru, and Yo'oro are unowned, the innumerable small creeks that flow into it are. Men construct dams at the mouths of their individually owned creeks and obtain similar though temporary rights in other creeks that are corporately owned by individual clan segments. These dams are repaired annually in anticipation of the seasonal flooding of the Mubi River system at several points in the year, primarily between February and April, and September and October.^[9] When the water level rises, men place their weirs in gaps made in the dams for this purpose. With the lowering of the water level, the fish that have been carried into the smaller tributaries are trapped as they attempt to reach the Mubi again. Smaller creeks may be dammed completely and Derris fish poison prepared. In addition, the Foi now use hook and line at all times of the year.

As with the Eskimo who Marcel Mauss eloquently described (1979) and the Nuer who E. E. Evans-Pritchard spoke of in similar terms (1940), the Foi provide an example of what Mauss first defined as the seasonal variation of social morphology. The rainy season for the Foi is one of dispersed nuclear-family settlements and minimal group sociality, while the drier half of the year involves the Foi in their collective longhouse community. It is not just economic activities alone that are regulated by this dual scheme but ceremonial activities as well, for the Foi mythologically view the time of their Usane pig-feasts as coincidental with the germination of garden food (regardless of when they are actually held). It also follows that such events as marriage, healing-cult initiation, and mortuary ceremonies could not (in traditional times) have been practically held during the hunting season if, as my informants said, most families resided permanently away from the village at this time. Neither could the collective tasks of, for example, housebuilding and cooperative fishing be reasonably planned during this time.

The Foi's methods of animal husbandry are also consonant with this seasonal residential alternation. Although all pigs are owned by men, women take care of them on a daily basis. Mature pigs are al-

lowed to roam the bush, and women give them just enough food scraps occasionally to keep them coming back to the bush houses regularly. In the bush, domestic female pigs mate with feral boars and bear their litters, after which they usually return to the houses. Shoats are fed solicitously and kept in houses with the women until they can forage for themselves. Since domestic male pigs are castrated to prevent them from going wild, it is clear that the feral boar population is necessary to the maintenance of the domestic pig reserves.

The feral boar population has, however, declined substantially as a result of the introduction of numbers of shotguns into the Foi area during the 1960s as well as from the contraction of disease by Foi pigs, which is still endemic. It is of course difficult to estimate the effects of this decline on the impregnation rate of domestic animals, especially as the Foi have always imported shoats from their trading partners to the north to supplement their own domestic population (so that varying rates of shoat importation would be of little value in estimating the size of domestic reserves). The haphazard methods of pig tending are also little affected by prolonged absences of women in the hunting lodges (though smaller pigs might be taken along at this time).

The Foi themselves make a strict division between their alternating residential and seasonal arrangements. They say that while in the Mubi Valley they live in their sabu a , "bush house," and that these houses are where a man looks after his pigs, makes sago, and plants gardens. A man's sabu a is never more than forty-five minutes away from the longhouse, and most of the Hegeso men built theirs along the banks of the Mubi itself, upstream from the longhouse. There the houses are easily accessible by canoes and adjacent to the major sago swamps owned by Hegeso men. A man rebuilds his bush house every three to seven years, relocating it occasionally in response to the varying maturation times of the different sago groves he owns. In contrast to the sabu a , the houses that men build at Ayamo are used only for hunting and are called aya a , "hunting houses." Men usually plant small patches of certain cooking leaves used in preparing the earth oven in which the Foi steam-cook most meat, and also occasionally plant other greens habitually eaten with meat, but the Foi do not call these gardens.

It is clear that the social-morphological variation I have described for the Foi is, as Mauss himself originally intended the concept, both ecological and conceptual. The progression of the seasons for the Foi

simultaneously embodies an ideational and material aspect, and it would be more appropriate to see both aspects of this variation as merely different refractions of a single cultural conceptualization of space and time. The major difference between the Foi on the one hand and the Eskimo and the Nuer on the other is that residential alternation is more a matter of individual choice for the Foi. No one is obliged to leave the village during the rainy season, and men of advanced age may elect to stay behind. In similar fashion, young unmarried men today often abandon their responsibilities in the village to go on hunting trips at any time of the year. It is the manner in which the Foi themselves interpret the variation in seasonal rhythms that is important. They know, for example, that bush fowls lay their eggs only during the kagi hua hase , and since bush-fowl eggs constitute one of the major categories of forest game obtained at this time, it would be pointless to seek them at any other time. The Foi also "know" with just as much certainty that marsupials and cassowaries descend from their homes in the sky when the cloud cover forms during the rainy season, and it is perhaps just as fruitless to go hunting during the time of the year when they have returned to the sky.

Furthermore, I have already noted that the Foi economic system has undergone great change in the last twenty years, and the benefits that the Foi have enjoyed as a result of economic development have largely come at the expense of their hunting activity. To a very small degree, the nutritional gap left in their diet has been filled by store-bought food, purchased with money obtained from cash-cropping. Even in 1938, however, Williams wryly reported the following:^[10]

Meat diet, here as so often elsewhere, is extremely restricted, amounting to a rarely enjoyed luxury. . .. As in the case of fishing, interest in the chase seems out of all proportion to the results obtained. . .. A great deal of magic goes towards this difficult and not very productive phase of the food quest. (1977:199-200)

But it is precisely Mauss' point that economic rhythms are intelligible in terms of the social and moral oscillations of a particular society:

We have seen examples of this rhythm of dispersion and concentration, of individual and collective life. Instead of being the necessary and determining cause of an entire system, truly seasonal factors may merely mark the most opportune occasions in the year for these two phases to occur. (1979:79)

In the next chapter, I demonstrate that seasonal alternation—and the

subsistence system that such alternation represents for the Foi—is part of a single symbolically constituted conceptual domain that attributes a sexual and social dichotomy to the Foi universe.