The Duna people with whom we work in the Lake Kopiago District of the Southern Highlands Province of Papua New Guinea have a valley and hillside environment that borders directly on the Strickland River (for further discussions of the Duna people, see Stewart and Strathern 2000a, 2000b, 2001a, 2002; Strathern and Stewart 1999a, 2000a; Marecek 1979; and Modjeska 1977, 1982, 1991, 1995). The Duna intermarry, visit, and trade with neighboring populations including the Huli people to the southeast (e.g., Glasse 1968; Frankel 1986), the Hewa people to the north (Steadman 1971), and the Oksapmin and Telefolmin to the west (Brutti 1997, 2000; Jorgensen 1981). The Bogaiya people (Sillitoe 1994) live southwest of the Duna and have become considerably integrated into the Duna area over the last decade through territorial shifts of population.
The Duna are swidden and fallow horticulturalists. Their population is approximately 20,000. They live in forested mountain areas that contain numerous lakes and streams. Limestone outcroppings and sinkholes are integral to the landscape. The Duna speak a language which is related to that of the Huli people as well as to the Bogaiya language (Wurm 1964; Foley 2000).
The Aluni Valley, the focus of our study, contains six distinct named territorial areas or parishes (rindi) that stretch westward to the Strickland River from the Aluni government aid post. The population of this valley, around 1,000, is scattered and sparse in density by comparison with the thickly settled groups of the central plain around Tari in the Huli area or with the wide fertile valleys of the Central Highlands that stretch between Mount Hagen and Chimbu (Brown 1978).
The territory of the Duna in the Aluni Valley backs onto large stretches of high forest or low-lying bush and high forest ridges, sloping in the south up to the Muller Range, that provide important supplementary sources of subsistence. Duna horticultural practices show a difference between gardens cut into thick forest or heavy secondary tree growth and those that are maintained in continuous long-term use in which women mound the soil and plant successive crops of sweet potatoes. (See Waddell 1972 for comparable practices among the Enga of Wapenamanda; Steensberg 1980 for discussions of Duna gardening in general; and Marecek 1979:42-50 and Modjeska 1977, 1982, and 1995 for more details on Duna production and women's labor therein.) Like the Huli, the Duna divide garden areas into those cultivated and harvested by women and those cultivated and harvested by men. Men harvest their own potatoes from their sections and cook them for themselves in communal men's houses or in their sections of divided houses in which women and men may have separate entry points and separate hearths.
Men are responsible for cutting down forest trees and for initial clearing of gardens, which is done with the assistance of female labor. Men also build the garden fences to keep out domesticated pigs, generally cared for and fed by women. New gardens cut into forest areas or into regrowth areas that do not lend themselves to mounding may simply be planted with flat expanses of crops. Once these areas have been harvested and the soil there reworked, women do most of the subsequent planting and harvesting at the site if it remains in use.
Duna practices with regard to rights over land are of some interest because of the intricate combination of principles of descent which they exemplify. Duna territorial parishes are inhabited by groups of people who have varying ties to the parish land through kinship and residence. Groups of agnatically related kin descended from the putative parish founders as authenticated in malu, or genealogical narratives (see Stewart and Strathern 2002), are considered to hold the primary rights to land, and are known as anoagaro (man-standing) members of the parish. The anoagaro line controls the authorized knowledge of the malu appropriate to the parish, which is supposed to be passed down in an unbroken line of agnatic precedence from father to son. Each parish also contains persons who have ties to the anoagaro line through female links. These are known as imagaro (woman-standing) members. The imagaro have a definite right to reside in the parish, to clear land for their subsistence, and to inherit such land. After some generations of residence in a parish, the descendants of an original imagaro line may become known as imagaro tseni (base/origin women-standing). That is, they are recognized as having ancestral claims. But in principle the precedence of the anoagaro line is maintained. Anoagaro men who claim such precedence declare that it is they who have the right to say who should cultivate what pieces of land within their domain in the parish; and we have observed that this claim corresponds quite well with what actually happens. However, in the area where we work there is no land shortage, so chronic disputes are not common. Individual disagreements, nonetheless, can lead to witchcraft accusations and expulsions of suspected witches from the parish (Strathern and Stewart 1999b, 2000a). In a case known to us, the garden area involved belonged to the semipermanent open area of sweet potato cultivation near the nucleus of houses in a part of the parish.
Furthermore, parishes also contain persons who have come to live there on various bases of kinship, friendship, and affinity. These might be described as guests or permissive residents, but they are not short-term visitors and on a daily basis act as parish members. The circumstances of the Aluni Valley Duna must be remembered here. Since their parishes cover wide areas of land and have relatively few people, they are short of people for their various social activities, and their anoagaro members welcome outsiders to come and live with them. At the same time this does not compromise the agnatic framework of precedence. Only if there are no remaining parish agnates do the imagaro tseni take over their status, in effect becoming anoagaro. Elsewhere in the Highlands, where land shortage is not an issue, unilineal descent as a way of constraining access to land tends to be stressed less than in places with heavy population density (Meggitt 1965). In the Duna case, however, the principle of agnatic succession is maintained, while otherwise the system is very flexible and runs on cognatic principles. The system among the Central Huli can be interpreted as operating in a similar way, although Glasse (1968) sees the system overall as cognatic, while Goldman (1983), discussing the Huli of the Koroba area, sees it as basically agnatic. We resolve such a potential discrepancy of viewpoints for the Duna by seeing how both agnatic and cognatic principles are at work.
This form of tenure therefore gives the system both stability and flexibility. For the Aluni Valley Duna, given the close proximity to their settlements of secondary and primary forest areas, there is scope for everyone to make the different kinds of gardens that make up the roster of garden types they recognize. These types correspond quite well with those that Sillitoe (this volume) outlines for the Wola people: shorter-term swiddens used for mixed vegetables; plots of wettish land planted with Colocasia taro; and longer-term semipermanent mounded areas used largely for sweet potato cultivation.
Discussions of the prehistory, settlement patterns, and social evolution of groups in the New Guinea Highlands have centered on a putative historical transition from reliance on taro (along with yams, bananas, sugar cane, and other cultigens considered to be indigenous to New Guinea) to reliance on the sweet potato (see, e.g., Golson 1982; Feil 1987). If such a transition took place, it did not mean the abandonment of these other crops, but only their displacement as the main staples. In all the Highlands gardening regimens of which we are aware, a distinction is made between gardens largely devoted to sweet potato and gardens where mixed vegetables, bananas, and sugar cane may be planted together. In the areas with higher population density, more pigs, and a greater stress on competitive exchanges, some see a historical trend toward the open-field system of continuous sweet potato cultivation, marked by intensive tilling either by mounding or ditching, such as is found in the Hagen and Enga areas (Strathern 1972; Meggitt 1965; Waddell 1972). Correspondingly, the mixed-vegetable gardens tend to decline as long-term fallow or forest areas with the kind of rich soil needed to support these crops decrease in availability.
These generalizations appear to have some validity in a broad comparative sense. However, the case studies and illustrations which we present here suggest some further complications in the picture. The two major complications which we adduce are: (1) even in areas where there is a great stress on exchange and population density is relatively high, as in Central Hagen, there is still an insistence on the importance of mixed- vegetable gardens wherever this is possible, both because such vegetables can be sold in markets nowadays and because luxury crops have historically been needed to entertain visitors; and (2) even in areas where there is not such a great emphasis on competitive exchange, the distinction is still maintained between intensively cultivated infields near houses for growing sweet potatoes, and mixed-vegetable or taro gardens in fallowed areas. We suggest that this distinction has been carried historically as a "package" from higher-density to lower-density areas and has therefore been culturally diffused from place to place in the same way as cult practices (Stewart and Strathern 2002). Gardening practices, that is, can be cultural markers.
Mount Hagen, as we have noted, illustrates the first point, while the Duna and Pangia (Wiru) areas illustrate the second. The Duna area fits into this set of ideas because, at least in the Aluni Valley where we work, a distinction between mounded infields used for sweet potato cultivation and other gardens used for mixed sets of crops is clearly maintained. At the same time, as we have seen, population in the Aluni Valley is quite sparse and there is little pressure on the land, although there can be intense use of specific gardening sites such as around the Aluni station. Since the Aluni Valley Duna represent the westernmost extension of the Duna population, we hypothesize that in their westward migration these people brought the practice of mounding as a cultural pattern, rather than as an adaptation to population pressure in situ.
Most of the work on sweet potato gardens is done by women. A woman also tends to make her own mounds, although sometimes she and her husband may hire a gardening team to do the work of reshaping mounds and planting new vines. An important feature of mounded areas is that they can be renewed piecemeal as the availability of planting materials and labor and the propitiousness of weather conditions dictate. The sweet potato vines can also he planted at any time, requiring only sufficient rain for them to take root in a new mound. Rainfall tends to come year-round. If it fails, the plants may wither, but replacements may be found in sheltered forest locations.
Since it is from the mounded areas that most of the sweet potato tubers are regularly harvested, and this harvesting is done by women, it is clear that it is women who have the major responsibility for feeding and raising the pig herds. They divide tubers into the better and larger ones, which they take to cook and feed their families, and smaller ones, which they give to the pigs. Although the tubers for the pigs might be seen as an ancillary part of the harvest, they must still be gathered every day and fed to the pigs, a practice which importantly helps to keep the animals attached to their home places and discourages them from wandering and getting lost or stolen. Looking after the pigs is therefore a daily chore for women. The pigs themselves are used in the life-cycle activities common in the Highlands: brideprice, funerary feasts, and compensation payments, for example. In the past, they were killed as sacrifices in periodic rituals to ensure the continuing fertility of the earth or to counteract disasters such as drought, floods, and epidemics of disease. In social terms, therefore, pigs were and are very significant to the Duna. Sweet potato provides the major part of the vegetable diet for pigs, which also forage in fallow and forest for grubs, insects, and worms. Because of the daily requirement to harvest the sweet potato, women like to have mounded plots near where they live. They must also maintain plots in varying stages of preparation, planting, weeding, and active harvesting. Tilled and mounded soil lends itself to this requirement and makes women independent of men's labor inputs. Correspondingly, men take it upon themselves to cut new gardens in bush areas.
Although the stone adze-axe was rapidly replaced by steel tools after they became available in the Highlands of Papua New Guinea in the 1930s, tools of the digging-stick type have remained in the tool kit. A woman uses her small, slightly curved, hardwood digging stick for harvesting sweet potatoes. It is much more efficient than a spade, because it does not tend to slice through the tubers as a spade would. Large digging sticks are also used by men to break up clods of soil and lift them from the surface of the ground. Steensberg (1980:72-80) describes these digging sticks (called kimbu) in some detail, with illustrations. The smaller one used by women is the tsenge (Steensberg 1980). Both digging sticks and either large or small spades may be used to shape and reshape mounds. Spades must be bought with cash or borrowed. Like steel bush knives, they become heavily worn down with use. Indeed, these more worn ones often become easier to use than the heavier ones that are new. Digging sticks remain the best tool for preparing the soil for mounds, whereas spades are best for leveling the surrounding areas of earth. Steensberg (1980:80) points out that mounding carries with it ecological advantages for high altitudes or cold conditions; these relate to drainage, soil depth, composting vines and leaves, weed reduction, increase in soil surface to encourage photosynthesis, and frost protection where relevant. Mounding is also a form adopted by cultural preference, and is used at both lower and higher altitudes in the Aluni Valley. Our interpretation here has stressed the convenience and flexibility of working with mounds near the house, matters which are important to Duna women. Sillitoe (1983:188-216) gives some detailed examples of sweet potato mounds in Wola gardens. He points out that the Wola women usually make mounds in neat rows. The same is true for the Duna. There is therefore an aesthetic aspect of gardening which is quite significant for the people, and gardens contribute importantly to the landscape. Sweet potato is by no means the only food that the Duna enjoy, although it is a major part of their diet and is of primary significance as the feed for domesticated pigs. Aluni Valley Duna have access to a wide range of forest foods and resources: marsupials, cassowaries, large grasshoppers, pythons, many varieties of edible tree leaves and ferns, earth and tree mushrooms, wildfowl eggs, and pungent fruits of the Pangium edule tree (in season). In their gardens they grow Colocasia and Xanthosoma taro, sugar cane, peanuts, maize, pumpkins, onions, ginger, yams, cooking bananas, sweet bananas, and varieties of cultivated greens. Both men and women grow these crops with the assistance of their children.
Women grow and harvest several types of green vegetables, maize, and red pandanus fruit, which men also plant and harvest. The individual who plants such a tree and tends it has the exclusive right to harvest its fruit or to give permission to another person to do so. In the high mountain crests men gather the nuts of wild nut-pandanus trees. They bring these to the settlement areas to he roasted and shared with others. These nuts have a sweet, meaty taste and can be smoked so that they keep longer.
The sweet potato is recognized as having been introduced into the region in the distant past. Hina na nayana (the time of those who did not eat sweet potato) is a way of referring to the original ancestors who lived before its introduction. Unlike the situation reported for the Huli (Glasse 1968), the Duna do not explicitly say that taro preceded the sweet potato as a more ancient crop, only that there was a time when the ancestors, who were seen as tama (spirit beings), did not eat sweet potato, a characteristic that might he taken simply to indicate their status as spirits.
Some taro is planted near water courses or marshy areas, hut not immediately next to houses. Types of wild Colocasia taro grow at water courses and are commonly found throughout the Highlands of New Guinea (see Matthews 1995). Colocasia cultivars are planted in sinkhole areas where water collects from the surrounding limestone, bringing soil and nutrients with it. Xanthosoma taro is grown near houses on small mounds of earth but not in large patches. It has a slower developmental cycle than other forms of taro. Both men and women are involved in the work of planting Colocasia and married couples share rights over its harvesting.
Xanthosoma is especially cultivated by men. Wherever taro is planted by the Duna it tends to he a monocrop, primarily because of the high demand for water in relation to other crops commonly grown by the people. The soil for planting is only lightly tilled and the planting tops of the taro are kept for no more than a few days before replanting takes place. Ash is sometimes added to the earth as fertilizer.
The leaves of the taro plant are not cooked or eaten by the Duna. Taro is cooked by boiling, roasting in ashes, or baking in an earth-oven. Xanthosoma taro and Colocasia are mostly baked or roasted in ashes. When taro is cooked in an earth-oven it is often cut into pieces to ensure complete cooking. Prior to baking, the skin is cut away and the corms are carefully washed to remove their potentially bitter taste. If Xanthosoma taro is roasted, the skin remains on until the corm is completely cooked and then the skin is easily scraped away.
Many different kinds of fruit pandanus are widely recognized, and people describe how these vary in terms of their length, girth, time of fruiting, consistency of the seeds, abundance of juice that can he squeezed from them after cooking, and the color, taste, and consistency of the juice. All this practical knowledge goes with the regular planting, harvesting, and preparation for consumption of the fruit. The tree grows best in the valley areas below 5,000 feet above sea level.
People may plant pandanus seedlings on any piece of land to which they have a claim. They may also receive permission to plant them on land belonging to a relative, in which case they retain rights to the tree but do not gain rights over the land around it. Usually, plantings are made on one's own garden land. The trees are long-lived and may even outlast their planter and so become subject to inheritance. Often they are planted in clusters which form distinct groves over time. Secondary fallow grasses and trees grow up around them, although immediately underfoot within a grove the grass tends to be sparse because of the lack of sunlight. The tree grows spiky leaves which are used for roofing thatch, making sleeping mats, and drinking spouts for water. The leaves wither and fall from the outside edges of the tree's foliage, making good tinder for fires. The tree is also characterized by its multiple aerial roots. Because of the tree's longevity, its groves help to demarcate ancestral claims to areas of land when people have shifted their houses away from them. One parish area near the Strickland River has been entirely depopulated since the 1960s, yet in the 1990s people still visited there seasonally to harvest their pandanus fruits, bringing them a day's journey home over a mountain ridge to cook and share them with their kinsfolk. (The parish is called Wakuni, and people of the Yangone group near to Aluni parish were those who exercised these claims.)
People check their trees in the course of other gardening work, or while foraging for forest fruits and vegetables, or hunting nearby in forest areas, or on special expeditions. They observe the emergence and development of the long, hard-shelled, reddish fruits within their tight sheaths, and see also if any have been stolen. When householders want to hold a cooking, they may send out a child or a young helper to harvest one or more of the fruits which they consider must now he ripe.
The fruits are almost always cooked as a part of a mumu, an earth-oven cooking of mixed vegetables and sometimes pork and/or marsupial meat. Both men and women know how to prepare the fruit for cooking. The mumu are minor public events, attended by a core of persons who live in the same small hamlet and close relatives or affines whom they invite. For these events, men usually are responsible for the work of pandanus preparation. The male householder or a helper splits the hard fruit open lengthwise with his axe and proceeds to scrape out the tough pith with a piece of sharp metal from a tin can (formerly this would be done with stone tools). The pith is set aside and cooked too, and can be dipped in the fruit juice and chewed, but it is not as highly regarded as the juice itself and is sometimes given to the dogs of the settlement area.
The mumu makers place the hollowed-out fruit cases on top of the mass of vegetable greens and tubers that form the bulk of the oven, interspersed with hot stones. Any meat that is being cooked is also placed over the rest of the vegetables. Both men and women help prepare the oven. After a variable period of hours, they decide to open the oven, clearing away the grass and earth they have packed over a covering of leaves which protects the food. The pandanus fruit shells are now soft and break easily in the hands. Younger men and youths take control now, squeezing and kneading the masses of fruits vigorously to separate the juice from the hard seeds. Water is mixed with the juice to give it a consistency similar to tomato catsup. Sometimes cooked pumpkin is added to make the mixture thicker. Strands of hairy pith from a banana tree stock may be dipped into the juice and squeezed to separate the juice entirely from the seeds, which are not consumed. The resulting sauce can he mixed further with green vegetables, or spooned onto taro or sweet potato pieces. Enamel and plastic bowls are used as receptacles; previously containers could he made from the spathes of black palm inflorescences, bound at the ends with vines.
People cut small spoons from pandanus leaves and use them as dippers into a shared bowl. Men and women eat separately, women sharing with their small children. Women sometimes cook and consume the fruit among themselves, in which case they carry out all the preparations. Men also may hold cookings for themselves and their sons; if so, they take the fruit to a communal or individual men's house to carry this out.
Our account here has been designed to give an idea of the rationales that underlie sweet potato production, arguing that mounding is both a cultural marker and a convenient modality of garden management for women. We have also shown that although the sweet potato is the staple, other crops such as taro and fruit pandanus are also highly significant.