Nature, Kubo, Etolo, visible world, visible and invisible, dance costume, contrast, Papua New Guinea, primary forest, Topographic Survey, food production, activities, nature and culture, evolutionary ecologists, subsistence activities, evolutionary perspective, Contour intervals, Siane, spirits of the dead, spiritual beings, Siane language, invisible beings, lexical distinctions, Kubo The Kubo, forest spirits
Dwyer, P.D. "The invention of nature" (1996) in Redefining nature: ecology, culture and domestication, edited by R. Ellen & K. Fukui, pp. 157-186. Oxford: Berg. ISBN 1 85973 130 9
The Invention of Nature
Peter D. Dwyer
you people try and dig little bit more deep you bin digging only white soil try and find the black soil inside (Paddy Roe, an Australian') Modern thought treats nature as separate from culture and has assigned ontological priority to the former. This is analogous to the separation of environment and organism that informs much of biology. The analogy is unsurprising. In a materialist world digital codes offer powerful analytical tools [Wilden 19723. We are easily taught that nature is other than culture or that environment is other than organism; that cultures, like organisms, are emergent products. These understandings mesh comfortably with a tradition of thought that, for more than a century, has been underlain by an Evolutionary Perspective
. I wish to revise, and to some extent up-end, this tradition. If we must adhere to a logic that is digital, then, I shall argue, in the domain of human affairs culture should be taken as prior, nature as emergent. The sad truth may be that the idea of 'wilderness' - that supposed last refuge of nature - is no more than an attempt to represent an imaginary place as a concrete symbol. 'Nature' as Westerners know it is an invention, an artefact. These introductory words both contain and conceal a paradox. On the one hand 'nature' and 'culture' are opposed as discrete categories. On the other hand, they appear as moments on a continuum that may be either developmental or evolutionary. The paradox is well known: it concerns the tension between product and process, between entity and relationship [Wagner 1977: 3861.
Nature as a Cultural Concept
By and large, in anthropology, the paradox is ignored. The often elegant cross-culturalanalyses of those who study symbols, seeking rules of transformation or generative grammars, remain, at base, static. By contrast, the rich empiricism of ecological studies, which often focus on specific populations or circumscribed questions, is driven by an evolutionary imperative; the terms of reference of evolutionary ecologists allow no alternative and, for the rest, the awkward, ill-defined or undefinable concept of 'adaptation as state of being' is ever-present [cf.Dwyer 19881. Neither approach to the understanding of nature and culture is satisfactory. In each, ingrained categories of Western thought are assumed to underlie the mental processes, behaviour or environmental relations of other people. The assumption is not spelled out. In moving toward a different appreciation of nature and culture I will proceed as follows. First, I shall briefly examine conventional understandings. My starting point here will be an analysis of the dance costume of Kubo people of the interior lowlands of Papua New Guinea. I shall then turn to the substance of my paper. Using information from three Papua New Guinean societies I shall depict for each not one, but two human geographies and ask if and how these interconnect. The two geographies are those of the visible and invisible worlds experienced by the people concerned. I shall then generalize to an alternate perspective on nature and culture. And finally, though briefly, I will turn to global issues. The Papua New Guinean material raises challenging questions concerning the relevance of ecological science to the study of people and the impact of conservation,as strategy and philosophy, on their lives.
The Kubo Dancer Among Kubo, curing ceremonies, intercommunity prestations, initiations and the completion of new longhouses may be all marked by dancing. At these events one or two males, their bodies painted and elaborately costumed, dance to the slow, rhythmic beat of a drum that they themselves carry. They dance through the night until dawn and are accompanied by others, both male and female, who sing. The songs are nostalgic, of loss and longing, and the dancer, set apart from his audience, is
The Invention #Nature
withdrawn, his head hung, the tears painted beneath his eyes a concrete symbol of the collective mood of subdued sadness. Yet to those who watch, the dancer is beautiful, he is the Raggiana Bird of Paradise. The body painting, costume, drum beat and movement combine as a transcendent whole that reaches beyond the mundane world to communicatewith spirits. There are many possible interpretations of the dancer's costume (Figure 6.1). I shall refer to three.2
Figure 6.1. The Kubo Dancer
Nature as a Cultural Concept
From the palm skirt, trailing on the floor of the longhouse, to the hornbill feathers that bounce at the top of the head-dress there is an orderly shift from aquatic to aerial domains. The skirt is a waterfall with cleansing properties, the drum is crocodilian and, at the rear, the dancer wears a rattle of crayfish claws. The bark belt and cape, the arm ornaments of palm fronds and fibre, and the soot and red and yellow ochres used to paint the torso are all elements and colours of the earth. The shell ornaments around the neck and on the chin are acquired by trade. Their ultimate places of origin are not known to Kubo. They denote relations that connect all people who move upon the earth. The band of cuscus fur across the forehead is from a cryptic, treedwelling species. People say that the animal looks like leaves against a background of sky. Next are plumes of the Raggiana Bird of Paradise that displays in the tree tops. Above these - and, I note, above the dancer's body - are feathers from three species that are usually seen above the forest. The apical hornbill feathers are enigmatic; they may have come from the spirit of a dead person. This reading of the costume is simple. Both natural elements and natural order are transformed as culture. Another theme is evident, though, in these enlightened times, it is, perhaps, more risky. The dancer's skirt, front and rear, is made from the fronds of sago palms, palms that epitomize women's contribution to subsistence. The frontal skirt is that of a woman and the designs painted on the torso may represent female genitalia.3 But, moving vertically to the ornaments of shell, bone, fur and feathers we encounter objects from the domains of trade and hunting, domains that are largely the prerogative of men. Even the decorative clusters worn on the upper arms are made from the fronds of black palm, the source of men's bows, and not from those of sago palms. We seem to have a secondary transformation, paralleling the first, even reinforcing it, from 'natural' female to 'cultural' male. These interpretations are my own. By training I am a biologist and so it is not surprising that they are rooted in such basic concerns of biological science - sex, diversity of species and the proper ordering of niches. Nor is it surprising that I should perceive a flow from concrete source to abstract meaning, from literal to figurative.Anthropologists,however, are predisposed to grasp the immediacy of the figurative. Thus, Knauft [1985al offers a different reading of the similar costume of Gebusi people.
The Invention @Nature
From the outset each costume element is symbolic of the form in which a category of spirit people is manifestin the natural world. Knauft [ibid.: 257-601 writes: The items worn on the upper half of the body derive from the upper spirit world (tree tops and sky), while those worn on the lower half are from the lower spirit world (in the water and beneath the ground). . .. The dance costume is a collage of all the true spirit beings in their natural-world forms. The Gebusi dance is in fact an enactment of a dance these spirits are simultaneously attending in their own world. The transformation we see here is of elements retrieved from the natural world being employed as concrete symbols of spirit beings and unified 'into the living motion of spiritual harmony' [ibid.: 2611. These sorts of themes are a commonplace in anthropological writing. The domain of the familiar, here taken as 'nature', is the source of metaphors (or tropes) that inform understanding of the less familiar [Dwyer 19791. The human body is a source of countless 'natural symbols'. Through metaphor, the contrastive pairs 'self and other', 'culture and nature', are linked, and meaning arises. The transformation always flows from the perceived raw materials
of nature to the mysterious unity of culture. The process is one of cultural genesis. We are drawn to the generalizations of Mary Douglas , Claude Lkvi-Strauss [I9661and Victor Turner [19741.Yet I remain dissatisfied. Too often, I think, the artistic symmetries we perceive deflect us from important and challenging questions. We conceal the ambiguities and inconsistencies of our own interpretations. Thus, Knauft identified elements of the Gebusi dance costume as symbolic of the natural-world forms of spirits. But is it true that all spirit forms are represented? If they are not, then there is less unity than Knauft asserted. Or again, the representation of an underground spirit-being in the upper part of the costume disrupts the appealing logic of upper and lower worlds that Knauft proposed. My own interpretations fare no better. The transitions from aquatic to aerial, and from female to male, are less tidy than I suggested and the case for 'natural symbols' is seriously marred by the incorporation of trade items into the c ~ s t u m e . ~
Nature as a Cultural Concept
I am not comforted by the thought that symbols allow multiple meanings or, as Fredrik Barth [I9871 has asserted, that there is no striving for coherence. Such assertions may rationalize intractable data; they do not challenge underlying theoretical structures. The analyses are too simple. Nature, we argue, is transformed. But where precisely does the moment of transformation reside? Is there perhaps a zone of transformation, a 'culture-tone' (cf.ecotone)that is neither natural nor cultural? Or might we borrow from Ingold [1986,19881 and ask whether the act of retrieval from nature - the intentionality that underlies that appropriative act - serves to mark the transition? Ingold's concern was with the boundary between the social and the ecological. He proposed that the boundary marked 'the point . . . where purpose takes over from, and proceeds to direct, the mechanism of nature' [1988: 2851. I want to go further and dissolve the boundary. I have wondered whether, for Kubo, the component elements
of the dance costume may never have been natural. The transformation we, as outsiders, witness may be one in which a collage of what were always cultural eleinents assumes a new and different status. The transformation may be of meanings, a transformation wherein cultural elements are related analogically to make communication possible with the world of spirits, a transformation wherein elements of the visible world, properly assembled, mediate the separation between Kubo people and their perceived 'other'. A new approach is needed. Anthropological perceptions of nature have grown from a deep appreciation of and concern for the social. What might we learn of culture if our starting place is ecology or, as I prefer here, geography? It is this that I shall explore with reference to Papua New Guinean populations. But some definitions are needed. They must be sufficient but not inflexible. 'Commentaries' might be a better word.
Commentaries The focal referent of 'geography' is the landscape itself. My concern is with the ways people live within that landscape, with the impress of them upon it and of it upon them. This, of course, is part of the intellectual endeavour called ecology but it is not
The Invention $Nature
the only, nor even the primary, concern of modern ecological studies. Indeed, by using the word 'geography' I seek to avoid the methodological requirement of ecologists that 'organismenvironment' relations are uppermost. That dichotomy is not central to the analyses of this paper. Space and time will be integral to my geographic readings, but we should not forget the importance of human knowledge
that emerges from and feeds into each landscape. human geography
cannot be held apart from human cognition. I must distinguish also between the worlds that I call visible and invisible. In the first instance the visible world is material, the invisible world is immaterial. The former comprises plants, animals and people and the physical media within which or upon which these reside. The latter comprises mythological, spiritual and fabulous beings that are known to the people as, for example, creators, witches, ancestors, omnipotent forces or the disembodied essences of corporeal beings. In their 'purest' forms these beings and forces are usually unseen. Indeed it may be the case that to see them is to die; that the event may be inferred by the living but never directly reported. But spiritual beings may be often manifest in the material world. They mark their passing by signs and they assume the forms, or inhabit the bodies, of particular physical entities such as individual rocks, plants, animals or persons. These manifestationsmust be classed within the visible world. They provide one channel for communication between the two worlds. There are other channels as well. In sleep or when dreaming the spiritual essence of a person may cross to the other world, in trance the bodies of mediums may speak with the voices of the spirits of the dead, and through ritual all people may participate in acts that, at least temporarily, make connections between visible and invisible worlds.
Papua New Guinea I have lived with and worked among three Papua New Guinean populations (Figure 6.2). In 1972 I spent ten months at a village on the slopes of Mount Erimbari (2850 m) in Simbu Province. Leu village was at an altitude of 1975 m; the 100 people who lived
Nature as a Cultural Concept
Figure 6.2. Papua New Guinea, showing locations of the three communities discussed there were of Rofaifo clan and spoke the Siane language. Seven years later, in 1979-80, I lived for fifteen months with 109 Etolo speakers at the village of Bobole in the Southern Highlands Province. This village was at an altitude of 1100 m on the steep slopes of Mount Haliago (2689 m). More recently, in 1986-7, I spent fifteen months with twenty-five Kubo speakers at the small village of Gwaimasi, 80 m above sea level, in the interior of the Western Province. Kubo and Etolo are members of the same language family and cultural complex [Knauft 1985b; Shaw 19861. In moving west along latitude 6"S, and from higher to lower altitudes, my research emphases shifted from the biological to the anthropological.And progressively, despite the lapse of fourteen years, my encounters were with people whose lives had been less affected by contact with missionaries or colonial and national government
The invention @Nature
Subsistence modes in both traditional form and as observed by me were very different between these populations. The trend from east to west is of decreasing intensification: population densities decline, and reliance on domesticated resources is reduced as the importance of 'non-domesticated' resources increases. In parallel, the east-west axis or, better perhaps, the shift from higher to lower altitudes corresponds with reduced emphasis upon resources whose returns are delayed and with the expected concomitants of this: mobility increases, community composition is more fluid, disputes are more often resolved by departure than by negotiation, and both sharing and egalitarianism are more pronounced. These differences between the three populations are etched into the local landscapes. They may be depicted most simply, though I hasten to add superficially,by viewing topographical maps published during the 1970s. The figures that follow have been extracted from those maps. My focus will be with the actual places at which I lived but I will reverse my own historical trajectory by taking Kubo first, then Etolo and finally Siane. The shift will be from less intensive to more intensive systems of production.
Kubo The Kubo landscape is sparsely populated and particular assemblies of people often divide and relocate (Figure 6.3). At Gwaimasi, which did not exist when the topographical map was published, twenty-five people use 50 km2.But communities are from three to six hours apart and crude population density is much less than 0.5 people per km2.Only 1per cent of the land comprises current gardens or second growth forest and, in most directions, one would travel several days before this value increased. The tiny regrowth patches seen on the map are near primary waterways and it is the latter that orient people within the landscape. Altitude is a minor consideration. Aerial photography is not omniscient. When the land near Gwaimasi is viewed from the ground the apparent importance of gardens is diminished (Figure 6.4). An anastomosis of trails spreads across the land to connect village, gardens, sagoprocessing sites, hunting shelters and so forth. Other trails lead
Nature as a Cultural Concept
Figure 6.3. Map of Gwaimasi region (Kubo). Shading shows anthropogenic habitats recorded by PNG 1:100,000Topographic Survey, Sheet 7385 (edition 1)Series T601, Printed 1979. The 200 m contour is marked. to favoured fishing spots, tiny groves of fruit pandanus or to nut trees. Everywhere there are other signs of people: clumps of bamboo and okari and breadfruit trees indicate earlier gardens; the remains of bird hides, rotting logs or sago palms may be places where animals were captured; and some areas of inhibited regrowth show where canoes were made. With guidance other dimensions of understanding emerge. A fifteen-minute walk through the forest provides a list of a hundred species of plants and a companion list of their uses. At one place the collapsed frame of a shelter reminds people of a man's illicit elopement and, at another, a tuft of feathers prompts the story of a
The lnvention $Nature \
167 1 h.
0 village houses 0 sago lark sites -C t r a i l s Figure 6.4. Walking trails associated with Gwaimasi village: 1986-7 (Only the area bounded by the Strickland River and the Sigia and Dege streams was comprehensively surveyed) cassowary hunt. An ancient post hole recalls a longhouse where people once lived, the initiationsheld there or cannibal feasts,the sorcery that occurred there and the departures that followed. To Kubo the land is a web of past and present human action and interaction and it is of these events that men, women and children speak when they walk together in the forest. Village life is not the preferred style of these people. They range widely through the day and, on 25 per cent of nights, sleep at bush or garden houses away from the village. The common theme of their activities is food production but the pace is relaxed, the demands on time and effort seldom great. Different sorts of procurement activities are environment or
Nature as a Cultural Concept
season-~pecificG.~ardens are made on levee banks, sago processed in b e back-swamps and pigs preferentially hunted in the foothills. But these activities are widely dispersed in space and Cvenly distributed in time. The location of favoured fruit and nut trees shifts from year to year and the dispersion of productive gardens changes continually. Women range as far as men; they visit the same environmental zones though they cannot visit all the same places. There are no sustained gradients of either aggregate land-use or gender-based association with land. Because different domestic units select different locations for the production of food and are often engaged in different activities, potential gradients are diffused. If the depiction of land-use was founded in a classification of activity types, then the pattern would be a far-ranging and everchanging mosaic. The dispersal in space and time would, I think, greatly exceed that reported by Ohtsuka [I9831 for Oriomo Papuans. There is a single notable exception to this description. Contained within the landscape is a forbidden place, the toi sa. Except in special ritual circumstances, people do not venture there. Yet even this is a cultural artefact: it is the abode of the spirits of the dead. The toi sa serves to both introduce the invisible world and to characterize my description of it.7 The resting place of the spirits of the dead is contained within Kubo territory. Indeed, the customary land of each clan contains its own toi sa. The invisible world permeates the land. Fabulous beings are associated with specific environmental zones or even particular places; there is a giant hunter in the back-swamps, a huge eel in the Strickland River and a python without a tail that guards the toi sa. The spiritual essences of animals are unconfined by the habitats of their mundane forms and, in the form of animals, the spirits of the dead may be seen in both expected and anomalous places. These last, in spirit form, interact freely with the spiritual essences of people, sometimes assisting, sometimes harming, corporeal beings. Their journeys and encounters occur within the landscape of human action, for there is no other place. Only the threatening essences of some foreigners suggest a perception of geographic gradients. From the swamp lands to the south they come as spirit warriors, marking their presence and intentions with signs. From the mountains to the north they come on the wind, or follow large waterways, to infiltrate the air and
The Invention @Nature
insinuate their way through narrow gaps in the walls of houses. They too, however, travel wherever Kubo live. The geography of the Kubo landscape is not characterized by strong gradients of use value or spiritual association. The visible and invisible worlds are co-extensive. Jn each, the significance of particular places is pre-eminent but always transient. Through time, places of current significance drift across the land. The two worlds converge in a mutual dynamic that facilitates their intercommunication.There is no sense in which the invisible has been positioned toward the periphery of the visible, and yet this is what we shall find elsewhere in Papua New Guinea.
Etolo Etolo live to the east of Kubo, in mountainous terrain (Figure6.5). For them 5 per cent of the land is overtly shaped by people, though, to the northeast and west-south-west, it is only a day's walk to landscapes that are more intensively managed. Etolo communities of thirty to a hundred people are one or two hours apart and always separated by forest. Crude population density is two people per km2 though effective density in areas of primary usage may be 7.5 people per km2. The locations of communities are constrained by altitude and peculiarities of drainage. Here, therefore, the orientation of people is toward both major streams - the torrents that fan out from Mount Haliago - and the precipitous ridges that gather at its summit. Etolo food production is more intensive than that of K u b ~ . ~ This is evidenced by the facts that Etolo cultivate one and a half times as much garden land per head as Kubo, emphasise tubers rather than bananas, manage orchards containing hundreds of fruit pandanus, and operate long lines of mammal traps in positions that are fixed for many years. Their gardens are usually fenced, and abandoned plots are often maintained as foraging enclosures for pigs. Again, in contrast with Kubo, most sago palms have been planted by people. Forest products, both plant and animal, are less important than for Kubo though this statement needs qualification.The nutritional importance of wild animals to Etolo cannot be underrated; it is the quantitative representation of non-domesticated foods and the variety of species appropriated that are low relative to Kubo. In sum the
Nature as a Cultural Concept
Figure 6.5. Map of Bobole region (Etolo).Shading shows anthropogenic habitats recorded by PNG 1:100,000 Topographic Survey, Sheet 7485 (Edition 1) Series T683, Printed 1974.Contour intervals are 400m apart. landscape occupied by Etolo is more overtly managed, and the managed domain more permanent, than is the case among Kubo. Etolo food-producing activities display strong gradients in both space and time (Figure 6.6). Hunting, trapping, gardening and processing of sago are the dominant activities. Each occurs within a restricted zone, with most hunting in primary forest above 1200 m altitude and most sago-processing below 900 m altitude. In addition, there are shifts through time in the sorts of activities that dominate the productive process. Hunting gives way to the repair of traps, new gardens are made and sago work follows. There is a second wave of gardening and the cycle then repeats. Elsewhere I have argued that this sequence is not pre-
The Invention @Nature
ordained by seasonal constraints but, instead, is an outcome of choices made by people. I wrote that 'key subsistence tasks were organised such that people did similar things at similar times' [Dwyer 1990: 1571:by and large, they did them in similar places. The same could never be said of Kubo. The timing, sequence and synchrony of Etolo subsistence activities, and the specificity of the zones where these take place, underline my depiction of food production along the axes of altitude and time. The significance of these axes is reinforced by
H v 700
APR M Y JUN JUL AUG SEP OCT I*IV O K JAN Pee MR APR
Figure 6.6. Scheduling of primary subsistence activities at Bobole, 1979-80 (Figure adapted from Dwyer 1990)
Nature as a Cultural Concept
quantitative differences in the times allotted to specific productive modes. Gardening requires the greatest input and, with trapping, can be accomplished by people who reside at the village. To process sago, however, people move to lower altitudes for periods of a week or more. On average, these departures amount to about thirty days per person in twelve months. Hunting in the high forest is even less demanding with an annual average of perhaps eight days per person, though in this case nearly all input is from males older than fifteen years. Clearly, there are major differences in the intensity of both aggregate use and gender-specific use across the altitudinal span of the landscape. The invisible world of Etolo is populated by several categories of spirit being [Dwyer 1990: 13-16; Kelly 19771. The spiritual essence of people wanders alone when the body sleeps. There are witches that invade the bodies of some individuals and prey upon others. There are the spirits of the dead; forest spirits whose society invisibly mirrors that of people; and spirit bachelors that behave as vicious witches. These beings, assuming the forms or occupying the bodies of plants and animals, can appear anywhere within the landscape. But the forest spirits, in invisible form, dwell only in the high forest; people turn to these to identify witches and ratify subsistence decisions. Inasmuch as those decisions are ordered temporally, so too the role of forest spirits changes with the seasons. When people die their spirits move into the bodies of various species of bird. And when the birds themselves die, perhaps by being killed, the spirits descend to rivers and occupy the bodies of large fish. The geography of the invisible, as experienced by Etolo, displays gradients of the sort depicted for the visible world. There is not, however, a neat one-to-one correspondence. It is the underlying theme, not the details, that is the same. The spirits of the dead transfer through time from an upper to a lower world, from the tree tops to the largest rivers, from where they may pass beyond the domain of people. The forest spirits occupy the place that people visit and use least often. Not all Etolo spirit beings behave in these ways but, in contrast to Kubo, the peripheries of the visible world assume importance as the abode of invisible beings.
The invention $Nature
Siane Most of the land used by and known to Siane people of the highlands is anthropogenic (Figure 6.7). Less than 25 per cent of the area near Leu village can be classed as primary forest. Villages of 100 or more people are common, often only a kilometre or two apart. Population density is thirty to forty people per km2.When people move across the land they skirt well-tended gardens, traverse lmperata grasslands and stands of Casuarina trees that were once gardens or, less often, encounter thickets of tall cane grass within which forest regenerates. Only
Figure 6.7. Map of Leu region (Siane). Shading shows anthropogenic habitats recorded on PNG 1:100,000 Topographic Survey, Sheet 7985 (Edition 1) Series T601, Printed 1978. Contour intervals are 400 m; dashes show vehicular roads
Nature as a Cultural Concept
at the peripheries of clan territories, especially at higher altitudes, does the forest seem unmodified. It is this altitudinal gradient that dominates the land and conditions people's orienbtion to it. At Mount Erirnbari the basal rock is cavernous limestone and large streams are not a surface feature. The intensification of Siane food production is an order of magnitude greater than that of E t o l ~T. h~e people are gardeners and pig-husbanders whose impact on the land is long term and permanent. They are familiar with techniques of tillage, composting, drainage and irrigation, obtain ninety percent or more of their food from gardens, and direct surplus production to pigs. Different procurement systems are often closely associated in space, and the productive activities of males and females are distinct and tightly constrained. These characteristics centralize the zone of primary food production to the vicinity of living places and establish strong gradients across the land in both use-intensity and gender association. Altitude delineates these gradients. Different garden types may be vertically zoned and, at Mount Erimbari, the high forest is a place where men hunt and, in particular seasons or suitable years, where fungi and pandanus nuts are harvested. The distance is not great but even the most enthusiastic hunters are unlikely to spend twenty days in the forest each year and will seldom remain there for two consecutive days. Average visitation rates per year may be only a few days. In the high forests of Mount Erimbari lives a white cassowary that should not be seen and a society of benevolent humans. These latter spirit people are almost always invisible; when they are seen their form is human. Indeed, the visible and invisible worlds of Siane do not routinely intergrade. Animals may understand human speech or, at particular times and in particular places, represent messages sent by invisible beings. But few animals are likely to be either transformed spirits or occupied by spirits. An identification of flying foxes with ancestors is the notable exception, though, in addition, some capricious spirit beings appear as snakes and fireflies [Salisbury 19651. Salisbury [I9651 discusses Siane religion at some length. He reports a god who existed before people did, who rules~overa land of the dead and 'takes the form of a circle of white light' [ibid.:551. The spiritual essence of people, pigs and some other
The Invention @Nature
animals animates them in life, can leave the body in sleep and appear as a white mucous secretion. After death these spirits persist for a time as ghosts, inhabiting the places they formerly occupied and appearing as humans in different form: 'they are white and insubstantial, emerge only at night and like cold and quiet while disliking strong smells and sexuality' [ibid.: 561. Through this phase of their existence they may be vindictive. Eventually, however, they are 'incorporated into the undifferentiated body of ancestral spirits
' who are well disposed toward people [ibid.]. Finally, there are what I will call 'spirit birds' [Salisbury 19651. Their only manifestation in the visible world is as a variety of flute calls and associated songs that are performed by men. They may have no other form, material or immaterial. Their songs record that they live in uninhabited forest to the southeast or with their mother, who is linked to the sun. The invisible world of Siane is unlike that of either Kubo or Etolo. The beings that inhabit it often assume forms that have no parallel in the visible world. They may reside far beyond the accessible domain of people or even in undifferentiated space. In the senses of both form and location they are peripheral beings who communicate with the living through intermediaries, through animals as messengers or bamboo flutes made by men. Similar themes are widespread across highland Papua New Guinea in regions where population density is high and agricultural production
intensified [e.g.Lawrence and Meggitt 19651. There are creators that appeared from distant or unknown places, made the material world, and went elsewhere. And there are societies of sky people, isomorphic with those of the earth, that, though they were the colonizing source of terrestrial populations or the originators of wealth, cannot be contacted by earthly beings.
Towards Theory Comparison of the Kubo, Etolo and Siane permits several generalizations. Starting with the visible worlds experienced by these peoples, I will summarize primary themes and extrapolate from them towards theory.
Nature as a Cultural Concept
tolo and Siane represent different points on a of intensification. That continuum, however, should *&+misrepresented as a unilinear, temporal trajectory. After m;&h of the three constellations of intensification is the current end-point of an ever-changing clade. The increase in intensification is correlated with the appearance of strong gradients in both the use of land and gender-based associations with land. Few such gradients are seen among Kubo, and these are transient. By contrast, their expression among Siane is evidenced in nearly all aspects of daily life. An outcome of the appearance of gradients of the sort described is the centralization of the productive domain. It adheres to the dwelling sites of people. Thus, the peripheries of local territories may be either little used or fulfil special purposes. Human impact on the environment is less overt in these places. Often, the peripheral domain will be demarcated topographically.This might be expected, for example,where the relief of the land varies a great deal; it will be contingent upon the varying suitability of different altitudinal zones to particular systems of procurement. There is, however, no necessity here. Distance from a central place or use-intensity alone may be sufficient to demarcate the periphery. The peripheral domain may be contained within, or dispersed throughout, the central domain. Peripheries originate through differential use but, ultimately, they are cognitive constructs. Inasmuch as the periphery of local territories is demarcated by use value, so it is likely that it will be demarcated lexically and, potentially, contrasted with the centralized productive domain. The literature seems full of examples but a major difficultyarises. The frequent equation of terms with, for example, 'domestic' and 'non-domestic', 'productive' and 'unproductive', 'village' and 'forest' or 'culture' and 'nature' may often entail imposition of the ethnographer's classificatory categories. Without extended exegesis such cases must be suspect. Strathern [I9801 provides a careful examination of Hagen terms that she translates as 'domestic' and 'wild'. Agricultural production and pig husbandry among these people of the Western Highlands of Papua New Guinea are more intensive than among Siane [Feil 19871. But the 'domestic-wild' differentiation identified by Strathern 'is in terms of the essence or
The lnvention @Nature
character of certain resources' [1.980:1931;it is not geographically bound though it is wedded to a distinction between 'social' and 'non-social'. Strathern shows how lexical distinctions in one domain may connect with, or connote, distinctions in other domains. However, her assertion that there is no geographic boundary between wild and domestic areas may be too strong. The landscape experienced by most Hageners is one of scattered gardens, fallow, grasslands and woodlands. The two last categories are not regularly used but they are anthropogenic. They are dispersed across the landscape and forest spirits are located there. Further, at the extremities of the known land, 'certain sparsely populated areas. . .which yield a concentration of [wild] things, come near to being considered a wild domain' [ibid.]. Among Etolo and Kubo I elicited no terms that were similarly contrastive. The Etolo word sege, meaning 'place', might connote 'somewhere else, not here' [Pwyer 1990: 121 and conjure up 'a distinction between non-domesticated and domesticated realms' [$id.: 121. But at other times, as in ne sege 'my place', the word 'embodied the sentiment of belonging - to the village, the gardens, the pandanus orchards, the sago groves, the forest, the streams and the mountains' [ibid.: 12-13]. Sege was more than these things, for it was also the Etolo word for 'rain' and for 'time'. There was no contrastive term. The Kubo word do is translated by some people as 'bush' but it too lacks lexical contrast. Its primary referent seems to be those portions of the forest that, at a given moment, are not in use. If the unforeseen happens and people are obliged to camp overnight before reaching a house then the temporary shelter they make is a do muson or 'bush house'. They do no.t generalize this label to more permanent houses or to those associated, albeit briefly, with specific subsistence activities. Nor to my knowledge are o wai and o fia, which I translate as domestic and wild pig, implicated as metaphors in the social sphere. It is true that for Kubo 'wild pigs' wander in the 'bush' - they are sometimes called do o - but this does not connote peripheral places. Far from it: the connotation is of potentially abundant production because wild pigs provide Kubo with most of their protein. My point remains. The potential exists to contrast periphery and centre where these have distinctive use values. The Kubo and Etolo languages reveal no such contrasts; at the least, the
Nature as a Cultural Concept
h q t q e s of many Highlanders move in this direction. Finally, among Kubo, Etolo and Siane the emergence of gradients in the use of land and in gender-based associations with land is matched by differences in the form and distribution of invisible beings. Gradients of these sorts are few among Kubo. Their invisible beings permeate the landscape, routinely appear in the form of animals and, in invisible form, interact freely with the spiritual essences of people. The visible and invisible worlds of Kubo are co-extensive. By contrast, gradients in the visible world characterize the landscape of Siane and, in parallel, their invisible beings are often located at or beyond the peripheries of local territories, assume nebulous forms, and interact only indirectly with either the material or spiritual manifestations of people. They may be propitiated, or appealed to for favours or blessings, but they can rarely be controlled. The visible and invisible worlds of Siane are neither reciprocal as with Etolo nor co-extensiveas with Kubo. If the correlations summarized here are not fortuitous, and I do not think they are, then important theoretical implications follow. Intensification of production will simultaneously alter people's perceptions of both the visible and invisible world. At the outset a landscape in which use values are generalized, extensive and ungraded and in which invisible beings are allpervading must be understood in totality as a landscape of human action and interaction. Hence, it is 'imltural'; there is no 'nature' and no contrast. But a landscape in which use values are particularized, intensive and graded and in which invisible beings are of nebulous form and peripherally located provides different opportunities. Here, through a process of cultural accretion, is the potential to invent 'nature'. Intensification and its correlates combine to categorize the visible world. Culture is internalized; it implodes. The created periphery of the visible world, increasingly divorced from human contact and understanding, emerges as 'nature'. It is imaginary and, in a materialist world, may stand for the 'other'. Differences I have read into the landscapes of Kubo, Etolo and Siane inform my grasp of this theoretical position. None of these people, however, should be taken to exemplify either a beginning or an end on what is a divaricating continuum of possibilities. Each is unique. Each offers to us a different appreciation of the relation between people and the landscapes within which they live.
The invention @Nature
There are places where 'culture' is the sum of the visible and invisible worlds and where it is these latter that provide sources of metaphor as people seek to comprehend 'self' by opposing it to 'other'. There are other places where perceptions have altered, where the invisible world has moved beyond the periphery of the visible and has lost relevance to the daily affairs of women and men or has been put aside. In these places the visible world may be conceived as the sum of culture and nature and these latter may, through metaphor, inform understanding of 'self' and 'other'. Little wonder, therefore, that for many Westerners the domain of the sacred has been relocated from the invisible world to the imaginary place that we call 'nature' - to the 'wilderness'. An evolutionary rider may reinforce the argument. It may also, perhaps, help fill the gap between the ecological and cognitive fields with which I deal. The gap, of course, is social relations
. The origin of self-aware, purposive and intentional beings - that is, I take it, of humans who are necessarily social can only have been located in a landscape of familiar relation- ships. The heightened perceptual awareness that marked this transition can only have been of those relationships - in space and in time and with all classes of interactive beings. Thus, it can only have been that constellation of relationships that was represented by self-awareness and permanently patterned purpose and intention. There was nothing else to grasp. The genesis of 'self' was itself an inverted metaphor because, at the outset, relationship was construed as 'self'. That which we, as Western thinkers, regard as 'other' was, in the beginning, the essence of 'self'; there was no 'other' as we understand it because the self-aware, purposive and intentional being was necessarily part of those relations. The domain of 'nature', which we have abstracted from that ever-changingpool of relationships, did not exist.
Other Lands, Other Lives Western thought has mistaken the periphery for the primal. To a large extent the conservation movement
has compounded the error by sanctifying the perceived primal.I0 The consequence is alienation because, at base, it was the domain of action and interaction - the substantive centre (which was total), not the
Nature as a Cultural Concept
imagined periphery - that was both primal and sacred. The mistakes are ontological. They impact on cognition. In acting upon, or seeking to understand, their landscape, which is assumed to be global, Western thinkers are informed by their cognitive categories. That is unavoidable. And, were it not that other people occupy that land, it might be harmless or merely self-destructive. After all, there is no a priori basis for claiming any one intellectual tradition as 'correct'; each may be judged, only in context, as apt or inapt. All such traditions have legitimacy as means of comprehending self and the world. Again, inasmuch as the conservation movement acts out its nostalgia for the 'vanishing past' - that is, the periphery - there may be no cause for alarm. No matter what changes occur to the earth there always will be a periphery, a place of attachment and a focus of action, for nostalgic longing. The rare and endangered building may assume the same value and satisfy the same need as the rare and endangered plant. Even individual people might attain heritage status. 'Nostalgia' is not a threatened species
. Within the bounds of Western thought the voices of dissatisfaction and protest will need more powerful arguments as they seek to change the world. Otherwise they will merely change with it. Other people do occupy the land and, I have argued, their perceptions differ from those of Westerners and from one another. I refer particularly to third- and fourth-world peoples who, so often, are underprivileged or rapidly becoming so. Victims of global hegemony, obliged to participate in global economies, they are dispossessed and effectively disenfranchised. These people, of course, have been and remain the 'objects' of anthropologicalinquiry. Increasingly their lands have become 'objects' of interest to conservationists. How should we perceive them or act responsibly toward them? To date our failures have been profound. I am not encouraged to think that things are improving. The methodology of ecological science asserts a separation of environment and organism. This conditions the questions asked, the analyses performed and the answers proposed. But if the 'objects' of inquiry are people who know no such separation, then ecological science is unable to reveal those 'objects' either to us or to the people themselves. My geographical analyses attempted to avoid the 'environment-organism' dichotomy.
The Invention @Nature The now blossoming field of evolutionary ecology, which seeks 'explanations for cross-cultural and historical variation in human social behaviour' [Smith 1988: 2221, fares no better. The rigour is impressive as is the disciplined approach to the formulation and testing of hypotheses. But the commitment to selection - 'natural' or otherwise - is the problem. It assumes a primal state and, thus, if we have misapprehended the primal then they are not revealed. Again, despite a controversial literature, I think that Marxist anthropologists speak, ultimately, with the same voice. The domain of the primal is coincident for these two schools of thought: it is 'nature' as we have perceived it and I have argued that some people - all, I suspect, in the category we like to call 'hunters and gatherers' - know no such condition. These different approaches may achieve great elegance, they may position other people within the categories of our thought and contribute to our imagined understanding. But they cannot elicit the essential and unique humanity of these people. The search for man - and woman! - in a state of 'Nature' is doomed to failure. The primal human, I have argued, was cultural with no experience of nature. The ecologist, perhaps, may intrude upon the lives of other people, seek answers to his or her questions, go away, and contribute to the excited dialogue of a select company of likeminded scholars. With luck he or she may be only an innocent participant in, and witness to, the hegemonic process. The conservationist cannot be let off so easily. The struggle for land is unequal. The categories of thought that inform the conservationist may have no analogue among those whose land is at stake. And, in these cases, to give precedence to our categories is, by fiat, to deny the land rights of others [Wright 19901. Again, where the strategies of conservation acknowledge the proprietorial rights of others, seeking only to 'manage' wildlife in joint arrangements with those others, then often, it seems, those strategies are founded in the perception of a 'proper' trajectory that passes from nature to culture. The people shall be.'tamedf, 'civilized', and made like us, for their own good and the good of the 'wilderness'. That is not good enough.
Nature as a Cultural Concept
c.1My intellectual debt to Roy Wagner is considerable. His book 7%&mention $Culture [I9811 influenced the title of this chapter. Blsewhere he has written: 'the dangers of working out our own prbblems on the soils and in the hearts and minds of other peoples should not be overlooked' [Wagner 1977: 4091. 1 have argued for more intimate connection
s between the geographies of the visible and invisible worlds than, I think, have been previously suggested. My excursions across the lands and minds of other people reveal much diversity and challenge central tenets
of our lives. It is our responsibility to those people to acknowledge and act on the challenges.
Notes I thank the University of Queensland
for granting periods of leave, the Papua New Guinea National Government for awarding research visas, and the University of Papua New Guinea and Papua New Guinea National Museum
for affiliation. I thank Rofaifo, Etolo and Kubo people who have been my friends, hosts and teachers and Monica Minnegal for her comments. 1. Benterrak, Muecke and Roe 1984. 2. See Beek 1987and Feld 1982 for other interpretations. 3. The interpretation of the painted design as female genitalia is from Beek [I9871 and was for Bedamuni people. 4. The sexual and gender imagery of the Kubo dance costume is multivocal. To Kubo observers the dancer is a fully plumaged Raggiana Bird of Paradise and, therefore, despite biological reality, is female. Females are attracted to the dancer and yet the sexual provocation seen in his movements is directed toward other males. Themes of transvestism, homosexuality and heterosexuality are merged in a costume that has higher purposes. The spirits of the dead, attracted by the beat of the drum and the singing, see the bird of paradise and are drawn to assist the living: to cure the ill, release game to hunters or assist novices in their transformation to men. Beek 1987 provides an extended discussion of the many layers of meaning that may attach to the similar Bedamuni dance costume.
The invention @Nature
5. The historical context
s within which I worked may, of course, influence my interpretations. I record them briefly here, referring to significant impacts, not to first contacts. By 1972, Rofaifo (Siane), of the Highlands, had experienced three decades of influential contact. They grew coffee as a cash crop, paid taxes, participated actively in national elections
and understood the requirements of modem law-enforcement agencies. Their familiarity with Christianity was of long standing; it was accommodated within, but did not dominate, people's everyday lives
. Indeed, there were signs of cynicism in response to unfulfilled promises and an expectation that desired change might be best facilitated through involvement in more secular arenas. By 1979, Etolo of the Southern Highlands Province had been hosts to Christian pastors for ten years and the impact of the church was pervasive. Government patrols, though intermittent, were accepted as sources of authority and as channels through which people might relay hopes and aspirations about future change. The commitment to change facilitated by missionization was strong and traditional ritual expression in abeyance. The experience of government help was one of frustration and abandonment. Through these ten years of contact Etolo communities had consolidated as larger villages. Though this had quantitative effects upon subsistence patterns there was little qualitative shift in the subsistence base. It was not until 1984 that people in the northern parts of Kubo territory (Western Province) enjoyed prolonged contact with the outside world. In that year a mission station (the Evangelical Church of Papua New Guinea) was established and this was followed, in 1985, by government-sponsored health facilities and, in 1988, by a primary school
. Earlier contacts with pastors from the Seventh Day Adventist Church had not been sustained and, at best, the influence of Government was indirect. Gwaimasi village, where I lived, was distant from these impacts. In 1986-87 people travelled for two days to seek medical attention and for three or more days to communicate with government officials
. Subsistence practices and ritual expression were little altered but people enthusiastically awaited profound, though barely understood, changes to their lives.
Nature as a Cultural Concept
6. For details of Kubo subsistence, see Dwyer and Minnegal [1991, 19921. 7. There are many similaritiesbetween Kubo and Gebusi in the beings that inhabit their invisible worlds; for details of the latter, see Knauft 1985a. See also Shaw 1990 on the related Samo. 8. For details of Etolo subsistence, see Dwyer 1990 and Kelly 1988. 9. My anthropological study among Rofaifo was limited to a minor investigation of hunting and more extended work in ethnoclassification [Dwyer 1974, 19761. I have drawn on Salisbury 119621, who lived with Siane speakers a few kilometres south of Leu village, on an extensive literature concerning Highland agricultural system
s, and particularly on Hide's [I9811 major work among Sinasina, about 25 km WNW from Leu. My own information on the invisible world of Siane is meagre and I have relied on Salisbury 1965. 10. The biblical 'wilderness' was peripheral and essentially unused. It presented challenges from, or permitted communion with, the invisible world. The rise of Western materialism, entrenched in the values of the JudaeoChristian tradition, has progressively shifted this perception. Through several centuries peripheral domains were tamed, conquered and colonized and their diverse physical forms collected, displayed and subjugated. In these ways the geographic extent of the perceived periphery was diminished and fear of the 'wilderness' abated. This paved the way i for the current shift to attitudes of nostalgic longing and / sanctification that have adhered to the supposed 'wilder- \ ness'. See Wright 1990.
Bibliography Barth, F., Cosmologies in the making: a generative approach to cultural variation in inner New Guinea, Cambridge: Cambridge University Press, 1987 Beek, A. van, 'The way of all flesh: hunting and ideology of the Bedamuni of the Great Papuan Plateau (Papua New Guinea)', Unpublished Ph.D. thesis, University of Leiden, 1987 Benterrak, K., S. Muecke and P. Roe, Reading the country,
The Invention $Nature
Fremantle: Fremantle Arts Centre Press, 1984 Douglas, M., Natural symbols, Middlesex: Penguin Books
, 1970 Dwyer, P.D., 'The price of protein: five hundred hours of hunting in the New Guinea highlands', Oceania 44,1974, pp.278-93 --, 'An analysis of Rofaifo mammal taxonomy', American Ethnologist 3, 1976, pp.425-45 -, 'Animal metaphors: an evolutionary model', Mankind 12, 1979, pp.13-27 --, 'The perception and reconstruction of the apt-organism', Journal of Social and Biological Structures 11,1988, pp.221-33 -, The pigs that ate the garden: a human ecologyfiom Papua New Guinea, University of Michigan Press: Ann Arbor, 1990 Dwyer, I? D. and M. Minnegal, 'Hunting in lowland tropical rain- forest: towards a model of non-agricultural subsistence', Human Ecology 19,1991, pp.187-212 ---, 'Ecology and community dynamics of Kubo people in the tropical lowlands of Papua New Guinea', Human Ecology 20, 1992, pp. 21-55 Feil, D., The evolution $ highland Papua New Guinea societies, Cambridge: Cambridge University Press, 1987 Feld, S., Sound and sentiment: birds, weeping, poetics and song in Kaluli expression, Philadelphia, University of Pennsylvania
Press, 1982 Hide, R.L., 'Aspects of Pig production
and use in colonial Sinasina, Papua New Guinea', Unpublished Ph.D. thesis, Columbia University
, 1981 Ingold, T., The appropriation of nature: essays on Human Ecology and Social Relations
, Manchester: Manchester University Press, 1986 -, 'Notes on the foraging mode of production', in T.Ingold, D. Riches and J. Woodbum (eds),Hunters and gatherers 1: history, evolution and social change
, Oxford: Berg, 1988 Kelly, R.C., Etoro social structure
: a study in structural contradiction, Ann Arbor: University of Michigan Press, 1977 --, 'Etoro suidology: a reassessment of the pig's role in the prehistory and comparative ethnology of New Guinea', in J. Weiner (ed.), Mountain Papuans: historical and comparative perspectives from New Guinea fringe highlands societies, Ann Arbor: University of Michigan Press, 1988 Knauft, B.M., Good company and violence: sorcery and social action in a lowland New Guinea society, Berkeley: University of California Press, 1985a
Nature as a Cultural Concept
-, 'Ritual form and permutation in New Guinea: implications of symbolic process for socio-political evolution', American Ethnologist 12,1985b, pp.321-40 Lawrence, P. and M. Meggitt (eds), Gods, ghosts and men in Melanesia: some religions of Australian New Guinea and the New Hebrides, London: Oxford University Press, 1965 L6vi-Strauss, C., The savage mind, London, Weidenfeld and Nicholson, 1966 Ohtsuka, R., Oriomo Papuans: ecology $ sago eaters in lowland Papua, Tokyo: University of Tokyo
Press, 1983 Salisbury, R., From stone to steel: economic consequences $ a technological change in New Guinea, Melbourne: Melbourne University Press, 1962 --,'The Siane of the Eastern Highlands', in P. Lawrence and M.J. Meggitt (eds), Gods, ghosts and men in Melanesia: some religions $ Australian New Guinea and the New Hebrides, London: Oxford University Press, 1965 Shaw, R.D., 'The Bosavi language family', Pac$c Linguistics (Series A), 70,1986, pp.45-76 -, Kandila: Samo ceremonialism and interpersonal relationships
, AM Arbor: University of Michigan Press, 1990 Smith, E.A., 'Risk and uncertainty in the "original affluent society": evolutionary ecology of resource-sharing and land tenure', in T. Ingold, D. Riches and J. Woodburn (eds),Hunters and gatherers 1: history, evolution and social change, Oxford: Berg, 1988 Strathern, M., 'No nature, no culture: the Hagen case', in C. MacCormack and M. Strathern (eds), Nature, culture and gender, Cambridge: Cambridge University Press, 1980 Turner, V.W., The ritual process, Harmondsworth, Middlesex: Penguin Books, 1974 Wagner, R., 'Scientific and indigenous Papuan conceptualizations of the innate: a semiotic critique of the ecological perspective', in T. Bayliss-Smith and R. Feachem (eds), Subsistence and survival: rural ecology in the Pac$ic, London: Academic Press, 1977 -, The invention $culture, Chicago: The University of Chicago
Press, 1981 Wilden, A., System and structure: essays in communication and exchange, London: Tavistock Press, 1972 Wright, J., 'Wilderness and wasteland', Island 42,1990, pp.3-7