FOREST KNOWLEDGE

FOREST KNOWLEDGE, FOREST TRANSFORMATION: POLITICAL CONTINGENCY, HISTORICAL ECOLOGY AND THE RENEGOTIATION OF NATURE IN CENTRAL SERAM
Roy Ellen
University of Kent at Canterbury

Since the mid nineteen-eighties Nuaulu swidden cultivators and sago extractors living on the edge of lowland rainforest1 in central Seram, Maluku, have become increasing active in countering threats to their traditional resource base. This latter has been dramatically eroded, mainly through government-sponsored settlement and logging. Nuaulu have successfully defended land claims in the courts, there have been violent incidents at a nearby transmigration area leading to their imprisonment, and in their representations to outsiders they have become articulate about the damage done to their environment. However, Nuaulu have a long history of contact with the outside world, of forest modification and participation in the market. They were politically engaged as early as the Dutch wars of the late seventeenth century and have been indirectly, and, more recently, directly, subject to the oscillations and economic fall-out of the spice trade ever since. The seventies and eighties of the present century have seen the expansion of cash-cropping, together with accelerated rates of land sale and forest extraction.
I wish to argue in this paper that as different material and social changes take place, so Nuaulu have renegotiated their conceptual relationship with forest in perceptibly different ways. In particular, I seek to ask why, given an apparent historic readiness to accept environmental change, they have now adopted a rhetoric which we would recognise as `environmentalist'. I claim that part of the explanation is that older, local, forms of knowledge which underpin subsistence strategies are qualitatively different from knowledge of macro-level processes -- `environmental consciousness' in the abstract -- which only comes with a widening of political and ecological horizons.

HOW THE NUAULU HAVE CHANGED THEIR ENVIRONMENT

Conventional Western conceptions of nature are usually of some unaltered other, of wilderness; and conventional views of traditional peoples living on forest margins or ecotypes, of tribes benignly extracting from an essentially pristine ecosystem. Such a view is, of course, now wholly unacceptable and there is mounting evidence of the ways in which humans dependent on forest actively change it. Much tropical lowland rainforest -- in Indonesia as elsewhere -- is the product of many generations of selective human interaction and modification (deliberate and inadvertent), optimising its usefulness and enhancing biodiversity. The outcome is a co-evolutionary process to which human populations are crucial. Indeed, particular patterns of forest extraction and modification are often seen as integral to its sustainable future. For some authorities, the evidence for intentional rather than serendipitous human influence is so compelling as to invite the description of `managed' forest [Clay 1988; Schmink, Redford and Padoch 1992: 7-8].
The empirical work supporting these claims comes mainly from the Amazon [e.g. Balée 1993, 1994; Posey 1988; Prance, Balée, Boom and Carneiro 1987]; but there is emerging evidence that it also applies to large parts of Malaysia and the Western Indonesian archipelago [Aumeeruddy and Bakels 1994, Dove 1983, Maloney 1993, Peluso and Padoch in press , Rambo 1979]. My own work, supported by recent botanical research, suggests that it is no less true for the forests of Seram, which have long been a focus of subsistence extraction, and where human agency has had decisive consequences for ecology. This has been largely through the long-term impact of small-scale forest-fallow swiddening and the extraction of palm sago over many hundreds of years [Ellen 1988a], but also through the introduction and hunting of deer, selective logging and collection for exchange in more recent centuries [Ellen 1985: 563]. Since sago is a frequent reason for venturing into forest beyond the limits of the most distant gardens, and since it illustrates so well the kind of co-evolutionary relationship I have just been discussing, it is helpful to say a bit more about it here.
Sago (Metroxylon sagu ) is extracted by Nuaulu both from extensive swamp forest reserves along major rivers and from planted groves much nearer to settlements. They manipulate vegetative reproduction by replanting and protecting suckers from recently cut palms, selecting suckers from some palms rather than others, and transferring root stocks to village groves. The result is an interchange of genetic material between cultivated and `non-cultivated' areas, even though there is no particular evidence of domestication through selective planting of seeds. Although most reproduction of sago palms in the lowland riverine forest areas of Seram occurs quite independently of human interference, in certain areas human involvement is highly significant, and the contemporary phenotypes of Southeast Asian sago palms are best seen as the outcome of a long-term process of human-plant interaction. Indeed, the historic spread of Metroxylon from its assumed centres of dispersal in New Guinea or Maluku suggest very strongly anthropogenic factors. Ecologically, the heavy reliance placed by Nuaulu and other indigenous peoples of Seram on sago has, over some hundreds of years, reduced the necessity to cut forest for swiddens. This has an important bearing on Nuaulu changing conceptions of their environment, as we shall see.
The distribution of many other useful trees throughout the lowland forests of Seram reflects patterns of human modification, and serve as convenient botanical indicators of settlement histories. Many are certain or probable domesticates and semi-domesticates. One of the most culturally salient of these is the kenari, Canarium indicum (=commune) . This is found so widely in lowland areas, and in particular configurations, that its distribution must almost certainly be explained as a consequence of human interference, both motivated and inadvertent [Ian Edwards, personal communication]. Kenari provides nuts rich in protein and essential oils, which are an important ingredient in local diet, but which for the Nuaulu also have a salient symbolic role, the precise character of which I shall return to later2.
Nuaulu practices of swidden cultivation and movement have, over several centuries, altered the character of forest vegetation in measurable ways: increasing the proportion of useful species, increasing the numbers of stands of particular useful species, decreasing the proportion of easily-extracted timber trees against those which are resistant to extraction, creating patches of culturally productive forest in more accessible areas, and creating dense groves of fruit trees in old village sites. Many of the trees nowadays found in areas otherwise not obviously modified by humans represent species introduced historically, and even prehistorically, for their useful timber, fruits, and other properties [Ellen 1985]. Indeed, approximately 78 percent of the 272 or more forest trees identified by the Nuaulu have particular human uses which make them potentially subject to manipulation through forms of protection and selective extraction. No wonder, then, that the distinctions between mature forest, different kinds and degrees of secondary regrowth and grove land are often difficult to establish. Although the contribution of non-agricultural activities, narrowly-defined, to overall Nuaulu energy expenditure and production is not to be under-estimated, and by comparison with other Indonesian swiddening peoples is rather high, my earlier contrast [Ellen 1975] between `domesticated' and `non-domesticated' resources was, in retrospect, drawn too starkly.

THE NUAULU IN THE WORLD SYSTEM

The patterns of ecological change indicated in the preceding section cannot be understood properly except in relation to the history of contact between the forest peoples of Seram and the outside world. The details of the early phase of the movement of biological species in and out of Seram [Ellen 1993b] is not relevant to the specific argument put forward in this paper, but that it happened is a part of the general background picture. Thus, the circulation of valuables, upon which the reproduction of Nuaulu social structure became effectively dependent over several hundreds of years [Ellen 1988a] was based on articles traded in from the Asian mainland; and what we know of the dynamics of the regional Moluccan system suggests contact which goes back much further than this, and which must have involved the export of forest products.
The most important single factor affecting Moluccan forests during the early period was the rise in the international demand for spices, which by the early sixteenth century had led to the spread of production from the northern to the central Moluccan islands. Expansion and fluctuation in growing clove in particular from this time onwards [Ellen 1985, 1987: 39-41 played a crucial role - both directly and indirectly - in the lives of inland and coastal peoples alike. The Nuaulu, for example, had an identifiable role in the relations of European contact as early as the Dutch wars of the late seventeenth century, as we know from the VOC archives and from the `Landbeschrijving' of Rumphius [Ellen 1988b: 118, 132n2]. We have a remarkably clear idea of the general location of their settlements in the mountains of central Seram from this time to the end of the nineteenth century, through oral histories, corroborated by surface archaeology, botanical evidence and eighteenth century maps [Ellen 1978, Ellen 1993a]. By the end of the nineteenth century, most Nuaulu clans had relocated around Sepa on the south coast [map 1], largely as a result of Dutch pressure, though they have continued an essentially highland, interior-oriented, way of life down to the present, relying on historic zones of extraction. In the eyes of the government, other coastal peoples, and in terms of their own self-definition, they have never ceased being uplanders and people of the forests.

[MAP 1 ABOUT HERE]

In the present century there has been renewed clearance, on Seram as a whole, for clove, nutmeg and other tree crops, such as coconut, cacao and coffee. The seventies and eighties have seen the expansion of market participation and cash-cropping (of clove, nutmeg and copra in particular), the planting of fast-growing pulp trees, together with accelerated rates of land sale and forest extraction. This has mainly taken place through logging and in-migration, first spontaneous and then official. Forest is being destroyed through unplanned slash and burn cultivation by non-indigenous pioneer settlers, and by the expansion of transmigration settlements into surrounding areas. There is no doubt that rapid forest clearance of this kind is damaging, and that long-standing sustainable practices are being eroded by technological innovation, population pressure and market forces. Local populations are encouraged by government to deliberately cut mature forest for cash crops, and commercial estate plantations are spreading widely. Logging is a particularly serious threat in the area where the Manusela National Park meets the Samal transmigration zone. Here and elsewhere so-called `selective' logging of Shorea selanica has led to water shortages, serious gully erosion and soil compaction; and has undermined existing forest ecology, resulting in more open canopy structures, Macaranga dominance, a greater proportion of dead wood, and herbaceous and Imperata invasions. In terms of fauna, there has been an obvious reduction in game animals. These effects have been systematically inventoried in the Wahai area by Ian Darwin Edwards [1993: 9, 11], but it is instructive to compare his description with that provided in the Nuaulu text discussed later, and which is appended to this paper. However, it has been transmigration and its various knock-on effects which -- more than anything else -- have been responsible for forest transformation

NUAULU RESPONSES TO INTRUSION SINCE 1970

The phasing and character of indigenous responses to the kinds of change I have highlighted depend very much on local perceptions of government policy and on the character and extent to which law and policy are interpreted by officials and translated into action. It is now widely acknowledged, for example, that the Basic Agrarian Law of 1960 and the Basic Forestry Law of 1967 are fundamentally contradictory and overlapping, and viewed differently by different government departments and in different situations. Sometimes they are used to defend the rights of indigenous peoples, but more often they override adat, legitimating the confiscation of land, and criminalising those local inhabitants who insist on asserting long-established rights of use [ Colchester 1993: 75, Hurst 1990, MacAndrews 1986, Moniaga 1991, SKEPHI 1992, SKEPHI and Kiddell-Monroe 1993, Zerner 1990]. Where there are doubts, national interest is invariably placed above local interests [Hardjono 1991: 9]. Up until recently, Nuaulu have been beneficiaries of an, on the whole, advantageous interpretation of the law, though as I go on to explain, this may now be changing.
During the period covered by my own fieldwork, the Nuaulu population has continued to grow dramatically: from 496 in 1971 to an estimated 1256 in 1990. This has led to greater pressure on existing land, intensified by competition along the south Seram littoral with people from traditional non-Nuaulu villages, and due to unplanned immigration, mainly of Butonese. Growth along the south coast has been facilitated by the extension of a metalled road during the early eighties. At about the same time the government began to establish transmigration settlements along the Ruatan valley [map 1].
The overtures by provincial government authorities to the Nuaulu, with respect to these developments, were, at least initially, benign and paternalistic. In part they have been guided by the special administrative status of the Nuaulu as `suku terasing' [Koentjaraningrat 1993: 9-16, Persoon 1994: 65-7]. Thus, the government recognised uncut forest in the vicinity of transmigration settlements as `belonging' to the Nuaulu, following the widely-held view of many non-Nuaulu inhabitants of south Seram. They then encouraged them to move into one of the new transmigration zone settlements along the Ruatan river, at Simalouw [map 1], an area which abutted sago swamps long claimed and utilised by Nuaulu. Although by 1990 only the villages of Watane and Aihisuru had moved permanently from their earlier locations on the south coast (about a quarter to one-third of all Nuaulu households), many Nuaulu established temporary dwellings, used the improved transport facilities to reach ancestral sago areas, and began to cut land for cash crop plantations. Moreover, two clans (Matoke-hanaie and Sonawe-ainakahata) moved even further inland and upland, out of the transmigration zone altogether to a place called Tahena Ukuna. Many Nuaulu saw these shifts as a return to traditional land, and for outsiders it confirmed Nuaulu status as upland forest peoples rather than lowland and coastal. Although Nuaulu had been located around the Muslim coastal domain of Sepa for the best part of one hundred years, and subject to the tutelage of its Raja, their self-image and the image of them held by non-Nuaulu, had never been otherwise. Moreover, implicit government recognition of Nuaulu preferential rights to over one-and-a-half thousand square kilometres enabled them to sell land in the Ruatan area to other incomers. This unusually positive approach was reflected in a successfully defended land claim in the courts at Masohi, the capital of Kabupaten Maluku Tengah.
The practical consequences of all this were alleviation of the growing pressure on Nuaulu land generally, and an opportunity to sell land along the more crowded south coast, most of which was sold to the inhabitants of Sepa itself and to incoming Butonese. This latter land, mainly old garden land and secondary forest, was a mixture of land gifted by the Raja of Sepa since the late nineteenth century, and land further inland which had always been regarded as Nuaulu. As I have argued elsewhere [Ellen 1993c], altogether, this created a rarely reported situation whereby an indigenous forest people appeared to be endorsing further forest destruction (both in the interior and along the south coast) by themselves and by others, for short-term gain.
Nuaulu cash incomes certainly increased through sale of land and trade with immigrants. Moreover, the practices which accompanied this were not dramatically contrary to any locally-asserted principles of indigenous ecological wisdom. However, there has recently been increased conflict with other autochthonous villages over rights to land, disenchantment with the effects of logging, and, since 1990, serious conflict with settlers resulting in convictions for murder of two Saparuan migrants being brought against three residents of Rohua. This incident was widely reported in the local press, who made much of the manner of death (decapitation), and of the removal of heads back to the village and their burial near a `rumah adat'. The episode has understandably been viewed by some government officials and other observers as a reversion to head hunting, or confirmation that it had never ceased, though the protagonists themselves strenuously deny such interpretations. Whatever the case, this narrative amply highlights the fundamental ambiguity in the concept `suku terasing', seemingly indicating both the vulnerability of a people so labelled, their need of special protection and advancement by the state, as well as their primitive threatening character, which the state must subject and change. Either way, Nuaulu are frequently viewed as prime candidates for `pembangunan' (development) in its moral and ideological sense [Grzimek 1991: 263-83]. Moreover, recent events reinforce a particularly pejorative local Ambonese stereotype of interior peoples as `Alifuru', and have made it easier for the government to explicitly expropriate territory when the occasion arises .

RENEGOTIATING NATURE

I argue here that as different material and social changes have occurred - changes which have accelerated over the last 20 years - so Nuaulu have renegotiated their relationship with forest, and nature more generally, in perceptibly different ways. How people conceptualise nature depends on how they use it, how they transform it, and how, in so doing, they invest knowledge in different parts of it. I have argued in another paper that concepts of nature have underlying pan-human cognitive roots, all people appearing to derive them from imperatives to identify `things' in their field of perception, situate these in terms of a calculus of self and other, and identify in discrete bits and aggregations essential inner properties [Ellen in press ]. However, identifying these commonalities is not to deny that such concepts are everywhere ambiguous, intrinsically moral in character and a condition of knowledge [Strathern 1992: 194]. Nature is not a basic category in the sense specified by Pascal Boyer [1993], and means different - often contradictory - things in different contexts. It is constantly being reworked as people respond to new social and environmental situations [Croll and Parkin: 1992: 16], and provides in the guise of something all-encompassing what I have elsewhere [Ellen 1986: 24] called a 'theory of selective representations'. Ambiguity itself, as Bloch [1974] has pointed out, can be socially useful. In the Nuaulu case there is an evident underlying tension between an oppositional calculus of forest and `village' (nature versus culture) and a non-oppositional calculus which draws much more on the lived experience of particular strategies of subsistence which unite what we loosely call nature and culture. Such an ambivalent conception of nature is wholly consistent with the difficulties faced in classifying the Nuaulu mode of subsistence according to conventional anthropological criteria [Ellen 1988a].
Before examining how these different concepts and their relative balance might be the outcome of a particular sequence of past events, and before highlighting contemporary patterns of change, it is necessary to sketch out in general terms the substance of the two apparently competing models or orientations. I do so on the basis of ethnographic data acquired by me at various times between 1970 and 1990. Since it is so obviously central, I start with the Nuaulu category of forest.
The Nuaulu use the term wesie to refer to forest of most kinds, but the term belies a complex categorical construction. Nuaulu relate to different parts of the forest - indeed to different species - in different ways. This mode of interaction is inimical to a concept of forest as some kind of void or homogeneous entity, and certain parts require different responses and evince different conceptualisations. Some bits of forest are protected, others destroyed without thought. Forest is never experienced as homogeneous, but is much more of a combination (rather than a mixture ) of different biotopes and patches. As such it well reflects the complex historical ecology which I referred to at the beginning of this paper. With its emphasis on human acculturation, it fits comfortably into a non-oppositional model of the kind we more usually associate with hunting and gathering peoples [Ingold in press ].
On the other hand, the generic term wesie exists, and is linked into general symbolic schemes such that it stands for some kind of conceptual exterior, a natural other. In some significant respects it is rather like the received twentieth century English concept of nature. Although subject to degrees of effective control through practical and supernatural mastery, wesie is associated with essential qualities of danger and otherness, and opposed to an unmarked category of `culture', most palpably evident in the category numa, `house'. As such it is intricately linked with gender imagery [c.f. Valeri 1990]. This forest : house : : nature : culture logic is evident in a whole raft of rituals, and in the symbolic organisation of space. In some ways it is not what we might expect given Nuaulu lived subsistence, with its heavy reliance on extracting forest resources, where gardening is traditionally rudimentary, swiddening practised on a forest-fallow basis, where regenerated growth supplies many `forest' resources over the longer term, and where - consequently - there is a definite blurring of anthropogenic and other forest.
The two somewhat contradictory models we find with respect to forest are repeated at the level of interactions with specific parts of nature. Thus, Nuaulu are primarily vegetative rather than seed propagators, and most of their starchy garden crops are tubers (taro, manioc, yam, Xanthosoma ). Such agricultural regimes are widely associated in the ethnographic literature with notions of continuity between nature and culture, in contrast to seed propagators who tend to emphasise a sudden transition between nature and culture [Coursey 1978]. In particular, Nuaulu are sago palm starch extractors, and as we have seen, this species is ambiguously wild and domesticated. Such a view is reinforced by the highly reliable character of palm starch as a staple, with a stable output little subject to fluctuation, lack of economically significant pests [Flach 1976] and considerable potential as a food reserve. In these ways, not only does sago contrast with grain domesticates, but is superior to tubers such as yams and taro, and is, therefore, an even better symbol of the continuity between nature and culture.
Given that many `forest' trees show evidence of human manipulation, occur simultaneously in cultivated and uncultivated areas, and provide long-term supplies of particular resources without continuous human attention and susceptibility to hazard, they too reinforce the applicability of the non-oppositional model. However, `trees' are only homogeneous as a category if we ruthlessly simplify it to some common cognitive morphotype (woody, foliaceous, rigid) and different modes of extraction, use and characteristics involve different relationships with people, different social profiles and potential symbolic values. This often leads to classificatory patterns which appear to cut across conventional logics, and which are almost provocatively ambiguous. I have already indicated that two extremely important sources of food - the sago palm and the kenari tree - are ambivalent in terms of the nature : culture :: forest : village logic. Both show evidence of proto-domestication, incipient cultivation, and their distribution is heavily affected by human use, despite the fact that they are for the most part culturally 'of the forest' and reproduce without much human interference. The problem is accentuated by the symbolic complementarity of the two: sago is the everyday starch staple and the product of - almost always - male labour, while kenari is collected for special festive occasions (such as female initiation ceremonies), when it is combined with sago to make maia [Ellen and Goward 1984: 32]. Thus, in certain contexts sago and kenari are linked together in opposition to products of the garden; in others they are contrasted in terms of an implicit gender distinction. Similarly, in the sphere of interaction with forest animals, I have [Ellen in press ] been able to demonstrate how a single ritual associated with killing (asumate) can simultaneously reflect a perspective which stresses the unity of all living things, and one which stresses human opposition through killing [c.f. Wazir-Jahan Karim 1981: 188]. Nature, I repeat, is not a basic category in the sense that it has a rooted perceptual salience, but though it may be symbolically deployed in radically different ways, it is still able to convey notions of logical primacy.
In developing a model which will help us understand how social and ecological changes have influenced Nuaulu conceptions and representations of forest and nature, we also need to recognise that in almost every instance this will have been motivated by an alteration in the character and intensity of relationships with the outside world, and how the Nuaulu deal with this socially. As I have indicated, ecological change has almost always been a consequence of exogenous factors: whether this involves the introduction of new species, outside appropriation of endemic resources or clearance of forest for extraction or agriculture. But whenever there is an environmental interface of this kind, there is also a cultural and social one. Transfer of new cultigens is not just about the movement of genetic material, but of cultural knowledge as well, knowledge which always carries a social burden. Contact with the outside world, in particular, seldom involves actors operating on equal terms, and the relationship is always mediated by considerations of power and control. For their part, the Nuaulu, inevitably, represent changes of all kinds in terms of the interplay of principles of opposition and continuity, complementarity and hierarchy3, symbolic schemes as opposed to practical experiences, outside influence versus persisting tradition. To show how this might work, we can, I think, provisionally identify three historical periods which are likely to have been associated with somewhat different conceptualisations of the natural world: pre-European contact, the VOC and early colonial period until about 1880, 1880 to 1980, and 1980 to the present.
From what we can reconstruct of pre-European Nuaulu social organisation, clans appear to have occupied separate dispersed settlements and had considerable autonomy, entering into loose alliances only for the purpose of intermittent political negotiation and to manage hostilities with outsiders. Thus, that subsistence placed less stress on gardening than became the case later was wholly in keeping with what we know of political arrangements. We might, therefore, expect here a concept of nature which focusses much more on the symbolic logic of vegetative propagation and the systematic harvesting of forest trees, and which involves a less oppositional conception of wesie. Moving around in forest is not conducive, after all, to developing an enduring opposition with it. Historically, we know gardening on Seram to be very underdeveloped, and even at the present time gardens are relatively unimportant in many areas, while in describing Nuaulu subsistence the distinction between `gathering' and `cultivation' is very fuzzy [Ellen 1988a: 117, 119, 123, 126-7]. There is no new evidence, as yet, ethnobotanical or archaeological [Stark and Latinus 1992], to suggest that horticulture amongst the native peoples of Seram was once more important than it is now [c.f. Balée 1992], except the general ethnological observation that pioneer migrant Austronesian speakers, their linguistic if not directly genetic precursors, depended on domesticates, including - in all probability - seed cultigens [Blust 1976, Bellwood 1978: 141].
The new embeddedness in the world system which developed from the sixteenth century onwards opened-up new pan-Pacific links, cut-out intermediary connections, and intensified exchange with Oriental, Asiatic and European centres. It also had immediate economic consequences in terms of spice production, and longer term implications for subsistence ecology. With the introduction of maize, manioc, Xanthosoma and Ipomea , reliable garden yields increased making these cultigens competitive with sago in their reliability and superior in the effort required to harvest them. This appears to have led to a greater dependence on gardens [Ellen 1988a: 123]. Almost all the new garden crops were vegetatively reproducing tubers, therefore sustaining a pre-existing relationship with nature; but they also began a longer term process of decentering sago from peoples conceptions of nature. Although sago is still culturally salient for the Nuaulu, amongst many present-day peoples of the central Moluccas sago (an indigenous crop) is nutritionally crucial but widely seen as inferior to (imported) rice. The same crops, because they decreased dependency on sago and other forest resources, encouraged greater emphasis on the symbolic opposition between gardens and forest. Increasing attention to cash-cropping, which both required high yield cultigens to offset the reduction of time and land available for subsistence extraction, and which provided opportunities to purchase - for example - rice, further accentuated this division.
The next major change came when the peripheral areas of Seram were formally drawn into the administrative system of the Dutch East Indies in the eighteen-eighties. From this time onwards environmental and social distinctions which had hitherto been implicit became underscored by administrative fiat. We have seen that from at least the late seventeenth century, the Nuaulu have had a distinct political identity. They have identifiable leaders, and were in various alliances, always including Sepa. Indeed, this long history of contact has made Nuaulu ultra-sensitive to questions of identity vis-a-vis other cultural groups, even though that identity has not always been reflected in any degree of permanent political centralisation. Formal incorporation into the Dutch administrative system, however, required that this identity and arrangement of traditional alliances of mutual advantage be regularised [Ellen 1988b: 118-9], both for administrative convenience and to provide the Nuaulu themselves with an effective channel of political communication. It is not, therefore, surprising, that at this time, when the Nuaulu clans were relocating around Sepa, when Sepa was - in Dutch eyes - becoming administratively responsible for Nuaulu `rust, orde en belasting', that there emerges a line of Nuaulu `rajas'. This, in turn, changes the terms of the oppositional relationship between Nuaulu and Sepa into a more hierarchical one. Clans begin to lose some of their autonomy, even though the line of rajas effectively terminated after only a few generations. And ever since, the question of a Nuaulu raja and his possible reinstatement has been an issue which has periodically become the subject of heated debate, most recently at the time of the establishment of the Nuaulu presence at Simalouw. The same necessity for formal mechanisms to communicate with the outside world is reflected in Nuaulu involvement in rituals of the Indonesian state.
Nuaulu movement to the coast meant a shift from a pattern of dispersed clan-hamlets and swiddens to concentrated multi-clan villages with large connected areas of garden land. This, in turn, led to a reconceptualization of the forest : farm boundary (wasi : wesie (juridical), nisi : wesie (technical), and therefore a move towards a more contrastive forest : village scheme. I have discussed on several previous occasions the changes in Nuaulu social relations of land use which followed from this [Ellen 1977, Ellen 1993c]: land sale, cash-cropping, individualisation, permanent occupancy - all of which still further emphasised a concept of nature in which contrastive properties predominated.
So, it is at least plausible that the apparent contradiction between oppositional and non-oppositional models of nature, the one more concordant with external relations of exchange, the other with internal subsistence experience, is a dialectical function of a particular transitional history. It might also be connected with the historic emphasis on exchange and the influence and internalisation of Austronesian symbolic schemes otherwise more amenable to seed-cultivation. Whatever the case, the balance is tipping in favour of an emergent, more oppositional, reified, concept of forest/nature. Amongst the coastal peoples of Seram (such as the inhabitants of Sepa) the enduring perception of the Nuaulu has been of a forest people - the opposite of themselves. Forest is a much stronger exteriority for coastal Muslims than it has traditionally been for animist Nuaulu, but it is towards this view that the Nuaulu are now progressing. Similarly, the Dutch colonial government, and thereafter the Indonesian government, created forest as a strong official category, establishing bureaucracies to manage it, a component in a wider state administrative division of labour, which encouraged implicit linkages between the geographical designation of forest and the social category `suku terasing'.
Moreover, as forest has been reduced in extent, so its representation as some kind of ether in which humans are suspended has been transformed into a much more restricted environmental category, just one ever-diminishing part of a wider non-afforested dwelling space. Not only does the small size of Moluccan islands make the forest more vulnerable physically, but also, as forest disappears, so it is reconceived as a fundamentally limited, rather than limit-less, good. Thus, both material experience of environmental change and the necessity to participate in a state level of discourse are reifying Nuaulu concepts of forest, just as environmental degradation and the ecological movement have done in the West. In order to protect their own lives, Nuaulu find themselves adopting the discourse of officialdom and national politics, responding to agendas dictated by the state. From a history of commitment to environmental change, they have now adopted a rhetoric which we would recognise as broadly `environmentalist'.

NEW RHETORICS AND RAPID SOCIAL CHANGE

What I have in mind by this new conception of nature and its relation to a more reflexive, globally-situated understanding of Nuaulu identity is well exemplified by two empirical cases: the first is a video-recording (cassette 90-2, 8-3-90) which I was asked to make by the people of Rohua in 1990 and which was prompted by Nuaulu concerns of state non-recognition of their religion; the second is a text recorded and transcribed in 1994 by Rosemary Bolton addressed as a personnel appeal to me.
The first - the video recording - consists of three parts, all of which refer to performances which occurred on 8 March. The first is a formal address given by Komisi Soumori (the kepala kampung and most senior secular clan head). It is an impassioned assertion of the legitimacy of Nuaulu core beliefs, showing how Nuaulu identity is rooted in land, forest and sago. This is unashamedly broadcast to an outside, unseen, audience. What is significant about this is its presentation: it is given in Nuaulu, because to speak of such things in any other language is to deny Nuauluness, but also because Komisi is most comfortable in Nuaulu. But the oratorical style and the physical props - rostrum etc - indicate the acceptance that discourse is at a state level [figure 1]. The second part is a short dramatic performance by adolescents about discrimination of Nuaulu customs and religion at school and in the labour market. This is conducted entirely in terms of a performance rhetoric which is alien to Nuaulu, and which is, appropriately, spoken in Indonesian - the language of the state. Paradoxically, such conventions (and the education through which they are acquired) inevitably result in the further attrition of Nuaulu identity, and perhaps its eventual disappearance. The third part is a speech by the ia onate (kepala) pemuda, Sonohue Soumori, again in Indonesian, which pulls the various themes together. Such reflections can also be cast in a more traditional idiom - such as the kepata ararinae and kepata Sepa - though on this occasion they were not.

[FIGURE 1 ABOUT HERE]

The transcribed text, the English language version of which is provided here as appendix 1, is a rather different kind of document. It was dictated by a long-standing acquaintance to Rosemary Bolton, and is separated in time from the 1990 performance by the harrowing events of 1993 in the Ruatan transmigration area, which I have already referred to. These events are structurally significant in Nuaulu representations of themselves, because an attempt to defend legitimate interests resulted in defeat. The rugged independence and assertiveness so typical of the seventies and eighties, and so well exemplified in the 1990 videotaped events, has - it would seem - been replaced by a new quiescence and passivity: `we are quiet and obeying them' (section 5). From a position in which Nuaulu saw themselves negotiating with the Indonesian state, they are now simply citizens of that same state. There is an acceptance that events are no longer under their own control, that they can no longer take them or leave them. As it happens, Nuaulu have a history of accommodating certain kinds of pragmatic change. This may explain their cultural survival, when most other groups of tribal animists on Seram have all but disappeared. But Nuaulu now claim not to want anything to do with the outside agents of change: government or logging companies. There is a realisation that the government does not keep its promises (7).
We can also see from this text how it is that the rapidity of environmental change has forced the Nuaulu to redefine nature, to see connections between microclimatic change, deforestation and erosion, and game depletion; between land clearance, river flow, impacting caused by logging vehicles, and fish depletion. We can see in it how Nuaulu now identify their forest as a whole as a commodity, something which has exchange value, when previously it was inalienable. We find an equation between big trees and profit (5-6), and governmental prohibition on sale. To begin with, Nuaulu accepted the advantages brought by the lumber companies: vehicles used the tracks and kept them clear, the tracks and trucks facilitated hunting (1,2). We also find recognition that replacement of large stands is in a time scale that is beyond the use of Nuaulu, that sustainable use has been superseded by something which Nuaulu would never seek to sustain (6), that old secondary forest, based on the cutting of patches and individual stands [Ellen 1985] has been replaced by wholesale clearance, which results in quite different patterns of regeneration, including more noxious vegetation (e.g. thorns). And the blame for these changes is placed quite squarely at the feet of logging companies and the state
So, recent Nuaulu reworking of their conceptions and responses to nature show that state patronage and state categories are no less central to understanding what is going on at the forest frontier than they are for lowland agrarian processes [c.f. Hart 1989: 31]. `Bringing the state into the analysis....entails understanding how power struggles at different levels of society are connected with one another and related to access to and control over resources' [ op cit p. 48]. As the forest frontier reproduces the inequalities of the wider state and its economically dominant groups, and as short-term production for use arises and is sustained by production for exchange [Gudeman 1988: 216], as Nuaulu move from being semi-independent `tribesmen' to dependent peasants, so their conception of nature reflects this. There is, in an important sense, an ecological, economic and conceptual continuity between forest modification and farming, and redefining forest extraction as a kind of farming may help us appreciate its similarities with the agrarian process.
In the Nuaulu case, intensification of subsistence agriculture, cash-cropping, forest extraction, commercial logging and transmigration combine to threaten an existing relationship with the forest. But Nuaulu attitudes have always been tactical, depending on their perceived material interests, and their conceptualisation of nature reflects this. The initial response to forest destruction and consequent land settlement was seen to have advantages in terms of a traditional model of forest interaction, based on implicit notions of sustainability of reproductive cycles of tree growth and animal populations. When this logic failed, complacency was replaced by uncertainty and bewilderment, eventually translating into hostility and decisive actions to defend their subsistence interests. Punitive actions taken by the state in response to this, have engendered further uncertainty and bewilderment.

CONCLUSION

What I have tried to demonstrate in this paper is that there is a connection between shifting Nuaulu conceptions of nature (particularly of environmental change), their social identity and the way they interact with the outside world.
There is nothing intrinsically problematic about environmental change for the Nuaulu. As we have seen, their cultural history is full of it. There is no overarching `ecocosmology' which rules it culturally illegitimate. Indeed, during the early phase of transmigration and logging in the eighties it was regarded wholly positively. What we need to recognise, however, is that there are different kinds of environmental change. The crucial distinctions here are between change which you can control, and change which is outside your control (and more specifically, is controlled by outsiders), and between change which is readily recognised as bearing unacceptable detrimental risks and that which is not so recognised. In terms of both distinctions it is the scale of change which provokes direct or delayed political responses and conceptual rejigging. The older, local, forms of knowledge which underpin Nuaulu subsistence strategies are qualitatively different from knowledge of macro-level processes, `environmental consciousness' in the abstract, which only comes with a widening of political and ecological horizons to a global level.
This global consciousness is no better symbolised than by the arrival of electronic means of communication in Nuaulu villages, first radio and then television. Television has not only enabled Nuaulu to keep in touch with the world by watching English league soccer matches and Thomas Cup badminton, but - and this is the reflexive twist - to watch David Attenborough eulogise tropical rainforest in its death throws. Despite a long history of contact with the outside world, changing patterns of environmental modification, patterns of subsistence and the conceptual modulation of these things, it is major changes associated with cultural globalisation which have forced a really radical response from them. It could be said that the aggressive individualism of the eighties, the selling of land and market engagement represented both the end of an old small-scale conception of nature in which resources and forest are infinite, and the beginning of a new conception of participation in an open global ecology of limited goods. The changes, therefore, are a response to a different problematic, to a different social and political agenda, rather than a rejection of environmental change itself or an a priori endorsement of ecological holism. Nuaulu constructions of environment are changing to accommodate a new level of discourse, and it is no coincidence that those who currently complain that their schooled children are unable to obtain appropriate employment in the Indonesian state because they are told that the doctrine of Pancasila is an impediment, also -- though paradoxically -- adopt an environmentalist rhetoric which seeks to keep the state from from their land.

ACKNOWLEDGEMENTS

The writing of this paper was supported by ESRC grants R000 23 3028 and R000 23 6082, and EC contract B7-5041-94.08-VIII, Avenir des peuples des Forêts Tropicales. Administrative permissions and financial backing for fieldwork on Seram between 1970 and 1990 are fully acknowledged at Ellen 1993a: xv-xvi. I am indebted to Rosemary Bolton of the Summer Institute of Linguistics in Ambon, and to its anonymous Nuaulu author, for permission to reproduce the text presented here as Appendix 1.

ENDNOTES

1. Lowland is used here to refer to a forest type generally dominated by the dipterocarp Shorea selanica , in contrast to the montane vegetation of higher altitudes. In fact, the lowland forest of Seram covers, on the whole, hilly country and may extend to an altitude of some 1000 meters.

2. Another striking case of human management of forest trees (though not one which I have observed in the Nuaulu area) is reported by Soedjito et al [1986] for higher altitude forests in west Seram. Here, seedlings of the resin-producing Agathis dammara , important as a source of cash, are systematically planted to replace older, less productive, trees.

3. Nuaulu symbolically represent their relations with the outside world, dialectically, in two ways: in terms of relations of complementarity, and in terms of hierarchy. The first is exemplified in the relationship between most local clans, in pela partnerships (that is between individuals linked through historical blood siblinghoods between villages) and through common membership of the patalima grouping [Valeri 1986]. The second is reflected in their relations with Sepa and the Indonesian state. Here they manage to assert, simultaneously, a mythic superiority (usually expressed in the conventional adik-kakak metaphor) and a pragmatic political submissiveness. The articulation of the two principles, however, is on their terms. They insist that they are prepared to accept the benefits of a good raja, but equally prepared to withdraw into their own autonomy when it suits them.

CAPTIONS

Map 1. The eastern part of the Amahai sub-district, Seram, showing historical, recent and present Nuaulu settlements, and other places mentioned in the text. The numbered locations are as follows (Nuaulu settlements in italics ): 1. Wai Ruatan transmigration zone (Kilo 5 - Kilo 11), 2. Simalouw (Kilo 9), 3. Tahena Ukuna (Kilo 12), 4. Makariki, 5. Masohi, 6. Amahai, 7, Bunara, 8. Watane, 9. Aihisuru, 10. Sepa, 11. Hahuwalan, 12. Rohua, 13. Tamilouw.

Figure 1. Three frames from a videotape recorded on 8 March 1990 and discussed in this paper. Komisi Soumori (ia onate Soumori and kepala kampung) addresses the world. His props include a young clove seedling and sago palm. Note the makeshift rostrum behind which he stands.

LIST OF REFERENCES

Aumeeruddy, Y. and J. Bakels 1994 Management of a sacred forest in the Kerinci valley, central Sumatra: an example of conservation of biological diversity and its cultural basis. Journal d'Agriculture Tropicale et de Botanique Appliquée 36(2), 39-65

Balée, W. 1992: People of the fallow: a historical ecology of foraging in lowland south America. In Conservation of neotropical forests: working from traditional resource use. 35-57, (eds.) Redford, K. H. and C. Padoch. New York: Columbia University Press.

Bellwood, P. 1978: Man's conquest of the Pacific: the prehistory of southeast Asia and Oceania. London: Collins.

Bloch, M. 1974: Symbols, song, dance and features of articulation: is religion an extreme form of traditional authority?. European Journal of Sociology 15, (1), 55-81 .

Blust, R. 1976: Austronesian culture history: some linguistic inferences and their relations to the archaeological record. World Archaeology 8, 1 19-43.

Boyer, P. 1993: Cognitive aspects of religious symbolism. Cambridge: Cambridge University Press.

Clay, J. W. 1988: Indigenous peoples and tropical forests. New York: Cultural Survival Inc.

Croll, E. and D. Parkin 1992: Cultural understandings of the environment. In Bush base:forest farm : culture, environment and development, pp. 11-36. (eds.) Croll, E. and D. Parkin. London: Routledge.

Dove, M. R. 1983: Theories of swidden agriculture and the political economy of ignorance. Agroforestry Systems 1, (3), 85-99.

Edwards, I. D. 1993: Introduction. In Natural history of Seram, Maluku, Indonesia, pp. 1-12. (eds.) Edwards, I. D., A. A. Macdonald and J. Proctor. Andover: Intercept.

Ellen, R. F. 1975: Non-domesticated resources in Nuaulu ecological relations. Social Science Information 14, (5), 51-61.

Ellen, R. F. 1977: Resource and commodity: problems in the analysis of the social relations of Nuaulu land use. Journal of Anthropological Research 33, 50-72.

Ellen, R. F. 1985: Patterns of indigenous timber extraction from Moluccan rain forest fringes. Journal of Biogeography 12, 559-587.

Ellen, R. F. 1986: Microcosm, macrocosm and the Nuaulu house: concerning the reductionist fallacy as applied to metaphorical levels. Bijdragen tot de Taal-, Land- en Volkenkunde 142 (1), 1-30.

Ellen, R. F. 1988a: Foraging, starch extraction and the sedentary lifestyle in the lowland rainforest of central Seram. In History, evolution and social change in hunting and gathering societies, pp. 117-34. (eds.) Woodburn, J., T. Ingold and D. Riches. London: Berg.

Ellen, R. F. 1988b: Ritual, identity, and the management of interethnic relations on Seram, pp. 117-35. In Time past, time present, time future: essays in honour of P. E. de Josselin de Jong (Verhandelingen van het Koninklijk Instituut voor Taal-, Land- en Volkenkunde 131). (eds.) Moyer, D. S. and H. J. M. Claessen. Dordrecht-Holland, Providence-U.S.A.: Foris.

Ellen, R. 1993a: Nuaulu ethnozoology: a systematic inventory of categories. (CSAC Monogr. 6) Centre for Social Anthropology and Computing and Centre for Southeast Asian Studies: University of Kent at Canterbury.

Ellen, R. F. 1993b: Human impact on the environment of Seram, pp. 191-205. In Natural history of Seram, Maluku, Indonesia. (eds.) Edwards, I. D., A. A. Macdonald and J. Proctor. Andover: Intercept.

Ellen, R. F. 1993c: Rhetoric, practice and incentive in the face of the changing times: a case study of Nuaulu attitudes to conservation and deforestation, pp. 126-43. In Environmentalism: the view from anthropology. (ed.) Milton, K.. London: Routledge.

Ellen, R. [in press]: The cognitive geometry of nature: a contextual approach. In Nature and society: anthropological perspectives. (eds.) Palsson, G. and P. Descola. London: Routledge.

Ellen, R. F. and N. J. Goward 1984: Papeda dingin, papeda dingin...Notes on the culinary uses of palm sago in the central Moluccas. Petits Propos Culinaires 16, 28-34.

Flach, M. 1976: Yield potential of the sagopalm and its realisation. In Sago-76: First international sago symposium: the equatorial swamp as a natural resource. 157-77, (ed.) Tan, K.. Kuala Lumpur: Kemajuan Kanji.

Goss, J. D. 1992: Transmigration in Maluku: notes on present condition and future prospects. Cakalele, Maluku Research Journal 3, 87-98.

Grzimek, Benno R. 1991 Social change on Seram: a study of ideologies of development in eastern Indonesia. Thesis submitted for the Degree of Doctor of Philosophy: London School of Economics and Political Science, University of London.

Hardjono, J. 1991: The dimensions of Indonesia's environmental problems, pp. 1-16. In Indonesia: resources, ecology and environment. (ed.) Hardjono, J.. Oxford: Oxford University Press.

Hart, G. 1989: Agrarian change in the context of state patronage, pp. 31-49. In Agrarian transformations: local processes and the state in Southeast Asia. (eds.) Hart, G., A. Turton and B. White. Berkeley: University of California Press.

Hurst, P. 1990: Rainforest politics: ecological destruction in southeast Asia. London: Zed Books.

Ingold, T. [in press]: Hunting and gathering as ways of perceiving the environment. In Redefining nature: ecology, culture and domestication. (eds.) Ellen, R. and K. Fukui. Oxford, New York: Berg.

Karim, W. A. 1981: Mah Betisek concepts of humans, plants and animals. Bijdragen tot de Taal-, Land- en Volkenkunde 137, 1 35-60.

Koentjaraningrat 1993: Pendahuluan, pp. 1-18. In Masyarakat terasing di Indonesia (Seri Etnographi Indonesia 4). (ed.) Koentjaraningrat. Jakarta: Gramedia Pustaka Utama.

MacAndrews, C. 1986: Land policy in modern Indonesia. Boston: Lincoln Institute of Land Policy.

Maloney, B. K. 1993: Climate, man, and thirty thousand years of vegetation change in north Sumatra. Indonesian Environmental History Newsletter 2, 3-4.

Moniaga, S. 1991 Toward community-based forestry and recognition of adat property rights in the outer islands of Indonesia, pp. 113-33. In Voices from the field: Fourth Annual Social Forestry Writing Workshop, J. Fox, O. Lynch, M. Zimsky and E. Moore. Honolulu: East-West Center.

Moran, E. F. 1988: Social reproduction in agricultural frontiers, pp. 199-212. In Production and autonomy: anthropological studies and critiques of development (Monographs in Economic Anthropology 5). (eds.) Bennett, J. W. and J. R. Bowen. Lanham: University Press of America.

Persoon, G. A. 1994: Vluchten of Veranderen: processen van verandering en ontwikkeling bij tribale groepen in Indonesie. Leiden: Riksuniversiteit te Leiden, Faculteit der Sociale Wetenschappen.

Padoch, C. and N. Peluso [in press] Changing resource rights in managed forests of West Kalimantan. In Borneo in transition: people, forests, conservation and development, N. Peluso and C. Padoch (eds.)

Rambo, A. Terry 1979 Primitive man's impact on genetic resources of the Malaysian tropical rainforest. Malaysian Applied Biology 8(1), 59-65.

Schmink, M., K. H. Redford and C. Padoch 1992: Traditional peoples and the biosphere: framing the issues and defining the terms, pp. 3-13. In Conservation of neotropical forests: working from traditional resource use, (eds.) Redford, K. H. and C. Padoch. New York: Columbia University Press.

SKEPHI and R. Kiddell-Monroe 1993: Indonesia: land rights and development, pp. 228-63. In The struggle for land and the fate of the forests. (eds.) Colchester, M. and L. Lohmann. Penang, Malaysia: World Rainforest Movement.

SKEPHI 1992: Logging and the sinking island. Inside Indonesia 33, 23-5.

Soedjito, H., A. Suyanto and E. Sulaeman 1986 Sumber daya alam di pulau Seram Barat, Propinsi Maluku. Jakarta: Lembaga Biologi Nasional.

Stark, Ken and Kyle Latinus 1992 Research report: the archaeology of sago economies in central Maluku. Cakalele: Maluku Research Journal 3, 69-86.

Valeri, V. 1989: Reciprocal centers: the Siwa-Lima system in the central Moluccas, pp. 117-41. In The attraction of opposites. (eds.) Maybury-Lewis, D. and U. Almagor. Ann Arbor: University of Michigan Press.

Valeri, V. 1990: Both nature and culture: reflections on menstrual and parturitional taboos in Huaulu (Seram), pp. 235-72. In Power and difference: gender in island Southeast Asia. (eds.) Atkinson, J. M. and S. Errington. Stanford, California: Stanford University Press.

Zerner, C. 1990 Community rights, customary law, and the law of timber concessions in Indonesia's forests: legal options and alternatives in designing the commons. Jakarta: FAO Forestry Studies TF/INS/065.