The Hagen people living north of Mt. Hagen township in the Western Highlands Province of Papua New Guinea are speakers of the Melpa language. Melpa speakers number in excess of 70,000 persons (see Strathern and Stewart 2000b for a contemporary examination of the Hagen people, and Strathern and Stewart 2000c and 2000d for retrospective examinations of these people; also see Stewart and Strathern 2000c). To the south and southwest live speakers of Temboka, which is closely related to Melpa (Merlan and Rumsey 1990). The town has greatly influenced the surrounding rural areas since the 1960s. For many miles around it people grow crops primarily in order to bring them to the town market for sale to urban workers. Coffee trees, introduced in the 1950s or 1960s, are planted throughout the area, and everyone is concerned to have some cash income.
The area of the Melpa speakers is ecologically varied. The Wahgi Valley, at about 5,100 feet above sea level, is the most fertile area, with rich black soil deriving from drained swamplands, vulnerable at times to drought. The Central Ogelbeng Plain rises to the west of the Wahgi Valley, with volcanic soil covered by grasslands reaching to the foothills of Mount Hagen and the Sepik-Wahgi Divide. The Wahgi Valley and the Ogelbeng Plain are densely occupied. Feil (1987:44) gives a density of 150 persons per square kilometer at Kuk in the territory of the Kawelka people in the Wahgi. Further to the north, in the area known as Dei Council, population density is by no means so high, and the people live scattered on hillsides cut by rivers and streams and dotted thickly with casuarina groves. The Kawelka people, a political group of some 1,600 today, were driven northwards from their territory at Kuk early in the twentieth century and sealed around Mbukl and Nggolke, there befriending members of the Tipuka tribe, with whom they became linked in alliances for major warfare. Pacification in the 1940s and the growth of the town in the 1950s and 1960s, together with smallholder cash cropping and the development of a government agricultural station at Kuk, brought many of the Kawelka back into the Kuk area, which is now once more their main population center (Strathern and Stewart 1998). In the 1960s, they were censused as having approximately 860 members, mostly living around Mbukl, Nggolke, and Mope (Sttathern 1972:32, 64- 65). Population growth has been steady since the 1960s.
The staple crop throughout the area is the sweet potato, and methods of cultivation for it and other crops are relatively uniform throughout the area. For the sweet potato, the so-called gridiron trenching system is predominantly used. In this system, trenches are first cut in a garden space after it has been cleared of debris and undergrowth, placed so that water will run off. Then trenches are dug that bisect the first ones to produce roughly square or rectangular patches of raised ground heaped with earth, which are then further lightly tilled prior to being planted. The sweet potato is usually planted in these tilled grids. Women take runner vines from existing plots and push them into pockets of soil, from which they develop new shoots and begin spreading within days. Men make the trenches, as well as participating in the work of clearing the garden area.
As numerous writers have noted, several characteristics of the sweet potato make it a highly advantageous crop for the New Guinea Highlands. It grows easily; it is propagated easily; it can be planted continuously the year round as long as there is some rainfall; it can he planted at a higher altitude than Colocasia taro; and its tubers develop faster than taro corms do. It is, in short, a perfect crop for the colonization and effective use of new areas of land. In addition, it can be planted repeatedly in areas of reasonably fertile soil. Finally, pigs like it and will happily eat the small tubers that women separate from the larger ones used for human consumption. All these considerations explain why sweet potato tends to he the staple crop in almost all Highlands areas, and why it is generally thought that its advent triggered a stage of further intensification of gardening and population movements.
In Hagen such movements are reflected at two different levels. First, the origin narratives of all the major groups tell how ancestors of the group migrated into their area from elsewhere or moved along a particular direction to arrive at their current settlement area (see Stewart and Strathern 2001b). Second, at an individual level, people use ties of kinship and marriage to shift from place to place in response to conflicts and new opportunities (see Strathern and Stewart 2000e). The major groups themselves are tribes, with populations ranging from a few hundred to several thousand people. These tribes are extensively segmented into subgroups down to small co-residential sets known as manga rapa, men's house groups. The large tribe segments are exogamous clans, which act as political groups in exchanges of wealth known as moka. Membership is determined by filiation, predominantly patrifiliation, but ties through the mother can also lead to acceptance (see Strathern and Stewart 2000f). The groups as a whole maintain stories that tend to represent them as expanded patrilineal groups, with male ancestors and their sons propagating the subgroups. However, matrifiliation is highly significant in these stories. For example, an ancestor's different wives may be represented as points of segmentation in a genealogy or a whole segment may be said to be descended from a sister of the group (Strathern 1971, 1972). The point here is that these rules and frameworks both provide an overall political structure and allow for considerable practical flexibility in affiliation.
Rules of precedence, however, hold in garden allocation and are reflected in myth also. A first-comer or original gardener of an area bequeaths to his descendants the primary right to cultivate that area in future. This category of people is called the möi pukl wamb, the 'root people of the ground." As a result of cumulative patrifiliation over time, most of these persons may be in practice male agnates of the group. But many groups also possess large percentages of nonagnatic members, who also have secure rights, and if their predecessors have been there for two generations or more they are in practice treated as indistinguishable from agnates. These flexible processes do not alter the generally patrilineal cast of the “dogmas of descent” (Barnes 1962) that define the group structure as such.
The gridiron system of gardening provides a simple way of dividing gardening strips in a garden. It is the owner of the basic rights (pukl) who apportions the area in this way. In turn, this person must recognize the claims of his household and others who have helped make the garden or whose kin gardened it before. A male owner typically allocates areas to female kin, especially to his wife or wives, and to male kin of his own men's house group who have an interest in the land and have helped in the work of clearing and trenching. If the garden is being made in fallow land, red cordyline markers may indicate where divisions were previously made, and these must be adhered to if the same persons are involved as before. Typically, wives are expected to be jealous of each other and to watch carefully over the divisions, as well as over equity of access to fertile areas. On hillside gardens this can he an important consideration. A central, low-lying area is fertile, while its higher sides are not. This is especially important for the planting of quick-maturing crops such as greens and maize.
The gridiron system is found throughout the Melpa area, from swampy valley flats to relatively stony hillsides with little topsoil. In the flats, its major function is to effect drainage; on the hillsides, trenching helps to provide sufficient concentrations of topsoil for crops to grow. In both contexts trenching can therefore be seen as adaptive. However, as with mounding elsewhere, we argue that an element of cultural preference enters here, and that preferred forms of gardens are imposed on the land to produce an aesthetic effect of landscaping that is found pleasing. The importance of the idea of trenching and its clear divisions of a garden is shown by the use of the term for a trench to represent a segment, usually at the clan level, within the tribe. While a tribe is mbi ou, “the big name,” a clan may be called pana-ru, "the garden ditch."
Gridiron ditching of gardens is found in Hagen and eastward to the Chimbu area. In the Fore area of the Eastern Highlands, sweet potato gardens are made of small mounds planted neatly in rows, sometimes with drainage spaces marked between the rows (Sorenson 1976:44, 233). South of Mt. Hagen at a certain point, larger mounds take over from the gridiron pattern and are found all the way south to Ialibu and beyond into Pangia (as we have seen above), through several ecological zones in which gridironing would be equally adaptive. Since mounding in the high-altitude areas of Ialibu and Tambul (at above 7,000 feet) has a primary function of protecting the sweet potato against the effects of frost, and mulching inside the mound increases its internal temperature through vegetation decay, it is reasonable to suggest that mounding was first developed in these high-altitude places and was subsequently diffused in different directions with the movements of people. Some of the groups in the Nebilyer Valley north of Ialibu do trace their origins to Ialibu or Tambul, just as some of the Pangia groups south of Ialibu do. A detailed examination of origin stories and contemporary gardening practices would be needed to examine this ethnohistorical pattern more closely. For the present, we suggest that gridironing began for the purpose of drainage in the Wahgi, and mounding began because of the need for mulching in Ialibu and Tambul. Both practices thus began as adaptations and spread with the spread of people.
As we have also argued, in no area is the sweet potato garden the only type. It is likely, therefore, that the specific forms of mounding and gridironing have themselves been produced over time in order to provide an effective environment for the sweet potato. Alternatively, of course, they may be older forms later adapted for the new crop. In Hagen and elsewhere, the other main type of garden is the mixed-vegetable plot. The name of this in Hagen is in fact the generic term for garden, pana (Strathern 1974:22); whereas the sweet potato garden is oka-pana (sweet potato pana). Pana gardens are planted with many different kinds of vegetables. They may have Colocasia taro and bananas as their main crops. The crops develop at different times, greens and cucumbers first, then maize, New Guinea asparagus, and later taro, then finally bananas and sugar cane. A pana garden therefore lasts for quite a long time. It needs fertile soil, and often is cut into tree fallow for this reason. Ditches are made to carry excess rainfall away, but not in a grid pattern. The drains are aligned as necessary to carry the water away in the most effective way. Men and women combine their labor in making these gardens also and in planting different crops. Pana gardens are much valued for the variety and flavors of their products, and people enjoy spending time in them or at their edges, weeding and harvesting, or simply sitting in the cool. Many of the pana crops are also taken to market and sold, as are specialized crops such as wing beans, peanuts, and fruit pandanus. When coffee was introduced, its seedlings were often surrounded by vegetable crops of the pana type, and this process continues with new plantings today. After all the vegetables and bananas have been harvested, the coffee trees take over completely. In this way some pana areas have been lost. People look for new areas within their land claims that they can use for pana crops. In the past, men in particular would offer sugar cane and ripe bananas, if available, to their guests, particularly valued exchange partners in the moka cycle. Today, especially at Kuk, if there are no such indigenous crops at hand, a man will hasten to the nearest store and buy timed meat or fish and rice and perhaps cigarettes for such a guest. If the guest does not trust food cooked by the host's wife, in case she might be from a hostile group or have some reason for personal enmity, the two men will cook the timed food and rice by themselves and consume it. At funerals, similarly, bananas and sugar cane were in the past contributed by visiting kin to help feed guests. Nowadays, cartons of tinned fish and bags of rice may be given, along with sweet potatoes. The money gained from cash cropping is thus in part returned into the maintenance and renewal of social relations. Cash cropping does not mean the end of these relationships, although in other ways it can lead to their weakening. For example, if people decline to make contributions to ceremonial events such as brideprices and funerals, this is a matter of concern for senior clan members as they monitor the conduct of the junior generation.
In Hagen, then, we find the same basic categories of garden types as among the Duna and Wim, with gridiron tillage taking the place of mounding. Taro gardens (me pana) may be a long-established type initially adapted to wetlands. As with the other cases, we suggest that there has been a diffusion of technical practices out from a center and that these practices serve as aesthetic, historical, and cultural markers. Gardening practices, then, contain traces of ethnohistorical processes. They leave marks on the landscape which express forms of identity as well as being records of adaptive ingenuity and local knowledge.
The practice of sharing male and female labor to make a large garden area and then dividing it into agreed strips is one that can hold for any type of garden made. It has also persisted, since apart from the introduction of steel tools in the 1930s, the work of making gardens is largely unchanged and the labor inputs required are heavy. One of the clan areas surveyed in 1964 belonged to the Yelipi clan of the Minembi tribe. This group was said to have ancient origins in the Ialibu area far to the south of Hagen and to have "come inside" the Minembi and joined them. It had also more recently been driven out of its former territory among the Minembi and had perforce to settle on a steep hillside at the western end of the former territory of a Tipuka clan, the Kengeke, to which some of its members were related. In these hard circumstances the Yelipi men and women habitually worked together to carve out adequate gardens for themselves. One sweet potato garden examined on September 8, 1964, had been prepared by five men, all of the same manga rapa (lineage) within the Yelipi, and each of these men's wives had planted her own strip in the garden. No share was given to the mother or sister of any of the men. One man had two wives, each of whom had her share, and another did not have a wife, so his share was to have access to the strip planted by one of the two wives of his lineage brother. This points up the fact that it is women who generally plant sweet potato vines. A man without a wife can plant his own if he wishes, but if not he must be incorporated along with a kinsman's wife.
This same survey turned up several cases of relatively newly established plantings of coffee seedlings. One of the men whose areas were recorded walked with a bent leg, a result of an old injury, but he still managed to do garden work, and in fact had been an innovator in bringing coffee seedlings from a commercial plantation some five miles to the east of his clan area in the North Wahgi. When he first brought the seedlings and planted them, people had asked him what this strange food was. He had later taken the ripe berries to the plantation and offered them for sale, only to be refused. Later he learned to pulp the cherry and to dry the beans and so was able to sell them. When his lineage brothers saw this, they all got seedlings and planted them too. The narrative is a microcosm of how smallholder coffee gardens spread throughout the Highlands. On July 14, 1964, this man had in one area 23 seedlings of coffee and nine plants bearing berries under the shade of some giant beans. In another area he had 48 older and 60 younger plants. In this case, the coffee had been planted where sweet potato had been growing earlier, so coffee does not always supplant mixed vegetables.
The garden survey continued into 1965, and some inquiries were made on pana gardens that had been made by families associated with a particular place, Kamb Kone, occupied by the manga rapa group of an old leader, Kambila. An account of this survey follows:
1. Pana made by three lineage brothers. One of these had in fact made three pana in addition to this one: one was up in the forest; another was in a different clan territory; and a third was in a low-lying and fertile area some six miles away to the west, where the Yelipi were allowed by members of a fellow Minembi clan to make gardens because of the paucity of their claims in their own area. In these pana he had given divisions to his brothers, but also to his mother and to wives of other clansmen, as well as to a man of a different clan who was living uxorilocally.
2. Pana made by an unmarried man. He divided the area among the wives of his brothers, and one of these planted vegetables for him and would share them for consumption. He had obtained the banana stocks from his mother's brother for no payment, and the sugar cane cuttings were from his own house yard.
3. Pana made inside the same overall fenced area as 1, planted by one man and his wife as a small separate share (oi). The ground was not previously used by his father; it was called emapana, "empty land."
4. Pana made up the hill by a married man, "on his own ground" (elimnga möi).
5. Pana made by the old man Kambila and one younger man; the undergrowth had been burned in preparation, but the garden was not yet made or its divisions settled.
6. Pana made by a man at his wife's place, in another Minembi clan area (not that listed in 1). This man bad a conflict with a fellow clansman over pana land in Yelipi territory. It was said he could settle the conflict if he wished by giving a pig or a pearl shell for the land, which had been used first by the other man, who therefore had precedence.
7. Pana made by a man who shared sections of it to the two wives of the local government councillor of the group.
8. Pana made by three men, who gave a share to the wife of a fourth man who had begun the garden but was involved in other court disputes.
9. Pana made by the coffee-garden innovator described above. Not yet planted or even its brushwood fired.
10. Five pana, each made by a different man, some in other clan areas as above.
Only three men of this group were said not to have made a pana. Of one of these it was said summarily that "he doesn't work," and of the other that “he goes around idly” with a kinsman from elsewhere. It was taken for granted that gardening work is something people just do, and making a pana was seen as an important part of the subsistence cycle. Those too idle to make their own might receive food from their kin, but they were still criticized. Gardening abilities and the willingness to do garden work were seen as essential attributes of people. With the rise of urbanization close to Hagen town, it is this attribute which some senior people fear is being lost.
Gardens are also an essential and pleasing part of the settings in which people live. They provide an intimate cluster of useful and decorative plants around houses. Settlement drawings made in the 1960s show this pattern. The settlement among the Kawelka Kundmbo of the ritual expert who set up the cordyline archway in our Hagen illustration no. 36 showed a clustering of houses surrounded by variegated garden patches of different kinds (see Figure). Settlements are also typically surrounded by plantings of flowers, bushes, and cordylines with their mixtures of bright red, yellow, and green, enlivening the appearance of the settlement and also fixing the claims of its occupants for the future. We have found that this observation applied as well to the Kuk area in the 1990s as it did to Nggolke in 1964.