Ethnography in Papua New Guinea:

A       A

Some may find it interesting to know what academic ethnography is like and how it compares to market ethnography. So, I am including a brief account of my experience in Papua New Guinea from 1969 to 1971 and again in 1976 and 1977.

Mine was one of the last anthropological expeditions. Nineteenth century expeditions were more adventurous and more arduous: these ethnographers traveled by canoe, or by outrigger, or by horseback, or on foot, deep into deserts or jungles or to the” back country of various kinds. I traveled by plane - ever smaller planes and ever more rugged landing strips - until I was within ten mile walk of the place I studied.

This portal was an administrative district office containing a clinic, an agricultural field office, a small police force, a Christian mission, several small trade stores with short hours and limited stock, and a post office. It received several flights a week from the Missionary Air Fellowship flying out of Wewak(a coastal town of 10,000, 100 miles away). There were also daily flights by a small commercial air service flying out of Vanimo (a coastal town of 1,000, 50 miles away). The level of commercial service was not really justified by the size of the post, but was subsidized because it was a frontier post - very remote from the capital of Papua New Guinea, but near to the border with Indonesia and an area of international tension.

A       A       A       A

As background to my expedition, I consulted a variety of sources. Several missionary linguists had lived in the area, studying local languages as part of a bible translation project. Several anthropologists had made a brief survey of material culture on the southern border of the area. I spoke or corresponded with each of these. In addition, the area was a UN trust territory administered by Australia through a series of District Offices. District Officers patrolled throughout the country, visiting every village annually. They conducted censuses and wrote sketchy reports on a variety of conditions - health, agriculture, local social organization and traditions. I spent several weeks at administrative headquarters in Konedobu, reading through 20 years worth of typescript reports on the area and locale I was intending to visit. There were discrepancies - sometimes serious - between the accounts of different officers at different times. Some of these revealed changes, but - based on internal evidence - some accounts were obviously more reliable than others.

In addition to local accounts, several good regional surveys were available to provide context. Thanks to the missionary linguists, regional language surveys identified the languages spoken in the area, described their historic relations, and mapped them village by village. Similarly, the Commonwealth Scientific and Industrial Research Organization had conducted surveys to identify and map the resources and eco-zones that comprise Papua New Guinea. They had not engaged in local studies in my intended area, but their reports placed my area within the range of eco types that existed in Papua New Guinea and their detailed descriptions of similar areas gave me a sense of what to expect and what to look for.

Each of these sources gave me a little feel of the elephant, and in the end I could begin with some information and some expectations, and a strong awareness that all of my ideas at that point were provisional.

From the district office/landing strip, a 50 foot wide dirt track had been cleared through the forest for 5 miles to the south. In wet weather it turned slick and often could not even be negotiated with the old Ford tractor - the only vehicle on the post. 2 miles out was a swift river, which carried away every bridge laid across it. Fortunately, at normal water, it could be forded. The cleared track ended at a small village and gave way to a narrow dirt path which wound down a deep ravine and steeply up to the shoulder of a mountain, around it, and then down into a country of knife edge ridges and deep ravines falling off to the Southeast and eventually to the flood plain of the Sepik river. On my first trip out I was not very fit, and still wearing heavy cleated boots, which became heavier when caked with mud. I arrived in the village exhausted, with huge blisters, and unable to move for several days.


It is easy to exaggerate the physical difficulties and travails of ethnographic expeditions. They are real, but adjustment is surprisingly easy. Eventually I lost 30 pounds and was able to walk the track barefoot and go and come in half a day, picking up mail and supplies. The hard part was the mental work.

Language is obviously an issue - there were and are no tapes for many of these languages. Neo-Melanesian - a pidgin of English - was an “intermediate” language. The people of the village had learned it as a second language, and I had studied it from books. It was deceptive in two senses. First it seemed easier than it was. Many words were English cognates and seemed familiar - but many did not mean in Pidgin what they meant in English, and most English words were not part of Pidgin. Plus, many local people spoke Pidgin poorly - it wasn’t their language either. I studied their language assiduously, learned many words of vocabulary, and came to understand in principle how it worked. Eventually I could often tell what people were talking about, but I never managed to speak it with any fluency. This situation is common in ethnographic settings and fosters a valuable humility about communication - a heightened awareness that what is meant maybe different than what is heard, and a heightened alertness for additional information to confirm or disconfirm.

A       A       A       A

A bigger issue than language is content: even if I had known how to speak the language, I didn’t know what to talk about. My ability to have conversations evolved, and the shape and pace of that evolution was driven by events - especially at the beginning. I got acquainted by talking about what was happening - you have to observe to participate. This gave my ethnographic experience concreteness and intimacy, but at the expense of control. The whole process becomes very open ended. This trade-off was not a choice: the core truth about participation observation is that it makes a virtue of necessity.

Religion and Marriage

A good example of this is the area of religious beliefs and ritual life. I was much interested in exploring these topics, but there is no way to say - in either Pidgin or Anggor - “ tell me about the gods you believe in”. It happened that on one of my first nights in the village I witnessed and participated in an event that was actually the central religious activity of the community. This was lucky in one sense, and inevitable in another. But it was ten years before I realized what I had seen.

Hunters had killed a wild pig; this was announced by excited calls from deep in the forest, and triggered a buzz of excitement in the village that I could apprehend without language. People easily explained what had happened, and what would happen - a wild pig was killed, it would be butchered by men and boys at a nearby stream, they would apportion the meat, carry it to the village, and further distribute it to the households where some would cook and eat it, some would smoke it for storage. And this all happened exactly as described.

A       A       A       A

Shortly after sunset, some men began blowing horns in the village center, and then young boys began beating hour-glass drums and dancing in a line through the plaza. Eventually they were joined by older men who linked arms and followed the boys in a rank, wheeling around the village and singing. This lasted maybe half an hour and was punctuated by a cacophony of blasts produced on two wooden horns.

No one had included this in their description of what would happen. The event seemed low key, impromptu - it did not “feel” religious. People said it was dancing and singing about the pig, and how much they like pig meat. But they also said it was about the Sanind. The songs were about the Sanind, the horns were to summon and dismiss the Sanind, and the pig was a gift from the Sanind. And people thought the Sanind liked the dancing and singing.

As I lived in the village over time, I learned more about the Sanind. They are guardian spirits, one of many spirit communities in the landscape but with an especially close and protective relationship to the people. Individual Sanind have names and live in invisible houses, scattered about the landscape. Each enjoys a special relationship with the human family on whose lands they dwelt. At times they possess these people and speak through them - for they see much that is hidden. They see sorcerers in the forest and expose or hinder them; they interfere with thieves who steal from “their” humans, and they see and control pigs, and “give” them to people. If the Sanind wish, people kill pigs; if they not, they don’t. A pig kill is a sign of grace.

The dancing I had seen was called tupur, the singing hwil. These terms also referred to the dancing and singing that went on at other events that seemed far more “ceremonial” - costumes, feasting, dancing all night. In the end, I participated in many of these. I never learned more by observing them than I learned my first night in the village, but the more I saw the easier it became to have meaningful and detailed conversations.

The larger ceremonies were usually people from one village dancing for another. In effect, the visitors made a salute to the Sanind of the host village, so they received food in return. The songs - the hwil - actually formed a progressive series that stretched over 4 or 5 years. They marked different stages of preparation, and culminated in a ceremony in which the Sanind visited the village wearing elaborate costumes that included plaques painted with totemic representations. I left the village before this ritual was performed. But observing the initial stages of preparation, and several very different ritual performances that contained common elements provide a platform for detail discussion of what I never actually saw.

Eventually, I also learned that the hunter who kills a pig never eats any of its meat, but distributes the pig in such a way that every household in the village receives a portion. This distribution is egalitarian in quantity - but each household gets a different part of the pig depending on their relationship to the shooter. In the end, because of the way relationships are understood, the shooter’s wife’s “side” receives the breast and back, and his own “side” receives the extremities. In effect, the distribution of pig divides the village into two complementary halves, depending on their relationship to the shooter - “breast” people and “bone people”. The bone people are his brothers, the breast people his brothers-in-law. For any shooter, half the people in the village are “brothers” and half are “brothers-in- law”. This is the public institution of marriage in Wamu - an endless payment of breast meat from men to their wives’ “family”.

All of these elements were “present” in the pig feast that I observed my first night in the village. Everyone knew all these things, but no one explained them. They didn’t imagine that I needed to know, and they would not have known where to begin. Eventually, as I observed, discussed, and imagined, I came to be in synch with how things happened and what they valued, to be familiar with their familiarities. Eventually, I became able to participate in their conversations about Sanind, and pigs and marriage. I was never in control, but I was also never passive. I had to push the conversation and to respond to opportunities, continually reformulating my notions and my questions.

Gardens and Time

A second area that was an important part of my study plan was the gardening process whereby the Wamu people made their living, and this required a rather different kind of observation.

Although some of the patrol reports claimed the Wamu people were hunters and gatherers, I knew they were gardeners because I could see extensive gardens cut out of the forest as I walked along the paths between the villages. The villages themselves were located at ridge crests, and floated far above the surrounding forest, which had been cut back and replaced with cultivars--pineapple, coconut, papaya, betel palm, rattan, bamboo, and many others that were unfamiliar to me but obviously planted and tended. These were all on steep hillsides, not laid out in regular fields or crop rows, and this may have misled the early patrol officers. Or, the gardens may have been much smaller in the 1950s, when steel tools were still rare in the area and stone axes still widely in use.

I purchased a shot gun and became a bird hunter and gradually became familiar with most of the territory within an hour’s walk in any direction. The process was gradual because the forest was very complicated and very closed in - hundreds of small twisting watercourses separated by short steep sided ridges, spreading out like the veins of a leaf and offering few panoramas. But as I learned the lay of the land, I also learned that it was full of gardens - new gardens, recent gardens, old gardens, ancient gardens - hundreds of gardens in different stages, with different plants with different uses at each stage.

A       A

Several months after I arrived, the dry season set in and I saw new gardens being made. Initially the work was almost invisible. It was carried out by families or individuals, working alone or in small groups. The cleared underbrush from steep hillsides in small patches, deep within the forest, and set it in piles to dry. You could easily pass by on a path and not realize garden work was in progress. But as time passed the patches grew and coalesced into large thinned areas, and the men began to clear the smaller trees and expose the trunks of the larger ones. The largest trees were cut nearly, but not quite through. After a month, a larger area of hillside was nearly bare, except for a scattering of very large trees and the rain forest canopy they supported.

A       A

All of this activity came to a head in a series of three dramatic events at the height of the dry season. For each garden, the owners chose a day for finishing. They recruited several friends to help them bring down the remaining large trees at the top of the hill - already cut nearly through and barely standing. With each man to a different tree, they timed their work so that the last bit of support for all the trees was cut simultaneously. This huge weight of trees, tied together by the canopy, collapsed the rest of the half cut trees further down the hill like a row of dominos and created a roar that was heard and cheered for miles. It also created a huge pile of debris which was left to dry out and then torched, creating a second dramatic event that could be heard, seen and cheered for miles around. This inferno cleaned the plot, leaving behind charred trunks and logs, and a lot of smoking plant ash to act as fertilizer. After several days the garden has cooled and was “opened” by ritual planting of 5 plants representing five key species, in the heart of the plot. This was a ritual event that happened at sunset, involved magical substances and spells, and - like the felling and the fire - was witnessed and cheered by the entire village.

A       A

Then the planting process began. Seedlings of taro and bananas from old gardens were collected and transplanted, interspersed with yams, pit pit, tobacco, and abika greens. Like the clearing process, this was accomplished by individuals and small groups working sporadically and gradually. They worked for several hours a day, replacing the under brush they had removed, filling in the space, one plant at a time - no rows or sections, everything mixed together. After several months, the scar of the clearing was closed over by a green fabric of cultivated plants, waist to shoulder high. This process would continue for years as tulip and pandanus were transplanted among the aging bananas. Fruit trees were planted to mature even later, and a variety of plants to provided useful materials - black palm for flooring, for baskets and for bows, bamboo for arrow tips and smoking pipes, rattan for binding, etc. Finally, the sago palm - harvested and processed continually to provide the staple starch of the diet - was planted in the wet ground at the foot of the garden.

A       A

Unlike pig distribution, the finishing, burning and opening of the garden were “low-leverage” events. In a sense they had no back story, and so observing them did not really provide a topic of conversation. Understanding the gardening process required a different kind of observation. It required an inventory - of plants, of garden plots, of personal relationships. Counting and mapping the garden resources of the less than 100 Wamu people was literally an impossible task. But attempting it through a process of census and sampling provided a platform for conversations that enabled me to understand both the nature of the system and the history of the village.

In Wamu, in 1970, 12 gardens were made - some by single families, some by groups of 2 or three families - of widely varying sizes. One of my research goals was to compare gardening in Wamu with gardening in other parts of Papua New Guinea. Although many existing ethnographies lacked detailed information on agricultural production, there were enough studies to provide a frame of reference. Consequently, one of the tasks I assigned myself was to measure the Wamu gardens, in order to measure agricultural intensity - annual production per capita. My gut feel was that it was low. This proved true, but the measure proved difficult.

As noted, the Wamu gardens are situated on steep hillsides - up to 45 degree grades. And they have highly irregular shapes. They tend to follow ridge lines, but also contain ridge-lines. They are a kind of collage of irregular vertical surfaces. They are also full of tree trunks and branches, to which are quickly added, transplanted crops like bananas and taro, weeds, centipedes and snakes. Measuring the perimeter is difficult enough, but calculating the area from perimeter measurements was beyond me. Plus I was sure that the corrugations within the gardens would cause a very large error factor.

In the end, I devised a process of dividing each garden into a set of component triangles and threading a 100 meter tape along the sides in order to measure the area. To measure productivity, I roped off a stratified random sample of a dozen meter square plots within the gardens, and inventoried all the plants they contained. I then calculated the area of the triangles and added them together, to get the total area under cultivation, and multiplied that by an average crop count per meter to get the total number of taro, bananas, etc. cultivated. Dividing this by the population gave me a per capita measure of acreage and plants that I could compare with the relatively few reports available from other parts of Papua New Guinea.

This was a costly metric in terms of man hours. That is one reason there are relatively few comparable measures available for other parts of new Guinea. Was it worthwhile?

From one point of view, the metric does provide an index of a real and important difference between agricultural regimes. In a sense, the patrol officers were half right when the said the Wamu people were not gardeners. Their gardens were meager compared to many other gardeners in Papua New Guinea. The shifting cultivation systems of the New Guinea Highlands, for example, are certainly more intensive than the Wamu system by an order of magnitude. Since this is so, it would probably be revealed by almost any metric you could capture: diet surveys, time-motion studies, population density measures, or even court record analysis. This fact is useful to know, and the measure we obtained though costly, probably offers good value.

But observations and conversations generated by the measuring process were actually of greater significance than the measurements obtained. The critical and distinctive fact about gardening in Wamu is not what happens in initial garden planting - the phenomenon I could measure - but what happens for years after. Each household in Wamu has dozens of old garden sites, many of them dozens of years old, some cultivated by parents or even grandparents; each household in Wamu has hundreds of sago trees, scattered across the landscape planted at the feet of the old gardens. Observations of daily activity and daily food intake show clearly that most food comes from old gardens, and most labor goes into maintaining them. The garden conversation is about the rich tapestry of resources that comprise the landscape, and the processes that created it. Cutting, burning and planting were points of entry for the conversation, and measurement of these activities provided useful background. But in the end the conversation was about the intersection of the flow of daily life with the entire past and future of the village.

This was summed up for me, late in my stay, listening to an elder give advice to a group of young boys. He spoke in Anggor, and I actually understood him: “When you are men, you must make gardens so that there will be sago.” I understood what he said because I had observed the complex, endless, and wide spread process that leads from gardens to sago groves.