- 383

The description and analysis of symbolic systems are major concerns in anthropology, though not, of course, new ones. In 1930 Maurice Leenhardt described the symbolism of the traditional Melanesian habitat in New Caledonia and he subsequently integrated these ethnographic data into his more abstract works, notably Do kamo (1930:ch.1; 1947). 2 Leenhardt saw duality as the basis of Melanesian thought: “the couple, the pair, the duality, acts as the basic unit for all the... [Melanesian's] mental constructions....All their social life plays on the rhythm of exchanges. The language is studded with dual forms” (1979:101-2; 1930:59-60; 1935:87-8). In New Caledonian languages the dual form itself and the relationship between elements which comprised it were of prime importance. That is, the elements were incomplete alone, 3 and derived meaning from their relationship, which could take either of two forms: duality, implying an organic bond, as between mother and child, or parity, implying equivalence and reciprocity, as between maternal uncle and nephew (Leenhardt 1979: ch.7, ln.2, 97-102; 1930-60). 4

For Leenhardt, the paradigmatic duality in New Caledonian patterns of meaning and classification was the conjugal couple, embodying the relationship, complementary but not opposed, of male and female (1937:22; 1979:66-7). It was represented symbolically in the physical layout of the habitat, recurred systematically in cosmology and values and was reflected in the structure of society. New Caledonians did not engage in abstract speculation on the nature and meaning of their world and relationships within it, but neither was their thought purely concrete. 5 Rather, the world of the living was identified with and participated in the natural and spiritual realms and “lived myth” played the part of abstract concepts in enabling them to understand and explain the cosmos (Leenhardt 1979: esp. chs. 5 and 12). Though Leenhardt did not clearly distinguish indigenous models from his own analytic models, the latter were profoundly rooted in - 384 New Caledonian linguistic usage and patterns of thought and behaviour. Modern Melanesians, concerned to reaffirm the value of their culture and create a basis for nationalism, accept the accuracy of his data and the validity of his analysis (Melanesia 2000, Tjibaou and Missotte 1978).

Similarly, his data and analysis help to provide a historian of New Caledonia with a grasp, essential to understanding historical process, of Melanesian cultural structure and its symbolic and social referents, in the context of which past action occurred. In terms of the theoretical perspective used in this paper, partly derived from Sahlins, “process” involved “a permanent dialectic of structure and practice” (Sahlins 1981a:54, see also Barth 1966:14-21). According to Sahlins (1981a), a people's understanding and interpretation of experience are modelled by the received categories of their culture. In action, cultural categories and relationships between them may acquire “novel functional content” (1981a:50), thereby undergoing pragmatic revaluation and redefinition. Fed back into the cultural structure, revised categorical relationships provide altered bases for evaluation of subsequent action. Underlying such structural transformation, however, is the logic of the cultural system itself. Thus, in Hawaii, “the meanings of tabu violations follow from the system. In combination, then, with the perturbations introduced by practice, the tabu logic becomes the mechanism of revaluation of persons and objects it had otherwise originally defined” (Sahlins 1981a:52).

I begin this paper, using data drawn mainly from Leenhardt's works, by examining relationships between spatial symbolism, modes of thought, principles of classification and valuation, ideas of substance, totem and spirit, and social and political institutions. This structural analysis is related to ethnographic and historical evidence suggesting disjunctions between certain ideals and social and political practice. Examination of the dialectic between ideals and action, of the revaluation in practice of concepts and relationships between categories, and of the generation of new categories, provides a key to understanding some historical processes in New Caledonia, both before and after the onset of intensive contact with Europeans. In contrast, Sahlins analysed the transformation of Hawaiian culture only in the context of pragmatic innovations consequent upon European contact, though he noted explicitly that his theoretical approach was not limited to culture contact situations (1981a:vii, 68). Actual transactions, processes of negotiation by individuals and groups among perceived alternatives, and the specific means, including ritual, through which cultural transformation occurred, cannot be elaborated here. They remain opaque, precontact, because of limitations in the data available, and in relation to more recent contexts must await a study in greater depth. This paper attempts to reconstruct the underlying logic of Melanesian cultures in - 385 New Caledonia, to outline basic categorical relationships at very broadly delineated periods, and to examine aspects of their transformation, in action, over time.

Leenhardt spent 25 years in the field as the European pioneer of Protestant missionary activity in mainland New Caledonia. He later became a professional anthropologist and ultimately gained the chair in the History of Primitive Religions at the École Pratique des Hautes Études. His main ethnographic concern was culture, shared patterns of meaning, modes of thought and cognition, and cosmology. Though his general reconstruction of traditional Melanesian culture was distilled from profound observation of behaviour in particular societies and the resultant understanding he developed of some Melanesian worlds, he felt little need in his works to elaborate social context. “The meaning of his [the Melanesian's] behavior both old and recent is not fully explained by the play of the institutions to which his behaviour relates. Instead, it is revealed through mythic forms of life and the elusive myths which govern them” (1979:2).

As a result, his image of social structure was very idealised. Kinship, for instance, he analysed as a value system, as a blueprint for, rather than a model of, actual social relations. His books largely ignored the behaviour and relationships of particular individuals, categories and groups of people, the institutional bases of society emerged only indistinctly and violence and politics featured hardly at all.

There are several reasons for this emphasis. Leenhardt's theoretical interests were partly determined by his missionary vocation, by his conception of conversion as a dynamic process of mutual translation of culturally conceived meaning (1937:8; Clifford 1980). Moreover, he worked mainly in a region which in the immediate pre- and post-European contact eras had no large chiefdoms. As a result, most political action occurred at the interclan level, where the bonds of kinship were as much real as idiomatic and ties of blood and affection were likely to limit the scale and ramifications, if not the intensity, of conflict. Again, by the time he arrived in New Caledonia in 1902, Melanesians had become largely irrelevant to the wider colonial society: a half-century of French colonialism, increasingly oppressive after 1878, and some depopulation, especially among Christian converts, had reduced the scale and significance of Melanesian politics, and most mainland peoples had been compulsorily relocated in tiny reserves, under the supervision of local French administrators. Such overtly or potentially violent practices as war, cannibalism and sorcery had become less visible, though, as Leenhardt was well aware, they had by no means completely disappeared.

- 386

Leenhardt's interest in ethnolinguistics and culture is in the mainstream of several modern perspectives in anthropology. On the other hand, his prevailing image of traditional Melanesian culture and society as balanced, harmonious, stable, disciplined and tightly integrated might seem anachronistic, and compared with some of the historical record “his” Melanesians do appear too good to be true. For instance, the ferocious cannibals depicted by the Polynesian evangelist, Ta'unga, in the 1840s (Ta'unga 1968:86-95) bear no apparent resemblance to the gentle rustics of Leenhardt's works. Leenhardt himself explained this seeming contradiction in terms of the duality of male and female in New Caledonian society:

It was, moreover, on this “restricted domain,” which few earlier European commentators had perceived, let alone understood, that much of Leenhardt's work concentrated.

Compared with the many complex symbolic analyses in recent anthropological literature, Leenhardt's data on the traditional New Caledonian habitat seem sparse and his analysis perhaps simplistic. None the less, they are all we have and should not be dismissed because of the possibility that a more complex structure may once have been discernible, overlapping or integrated with the simple synchronic dualism he depicted. These Melanesians, he said,

Melanesians in traditional New Caledonian societies typically lived in - 387 small hamlets dispersed on crests, spurs or small hills among their garden land. Landscape, gardens, cultivated trees and buildings formed a harmonious unity, which blended into the surrounding bush, the spirit-place, and in which spatial, social and mythic (totemic and spiritual) dimensions merged:

Dominant spatially and symbolically, at the head of leafy avenues, was the great round house of the patriclan, which could extend to 9m. in diameter and rise up to 12m. high. In the Ajie language of Houailou the great house was called moaro, also the generic term for clan, the basic unit of social structure, and for the altar, ka moaro, “which is the moaro,” the ultimate “point of support for the social and spatial world” (Leenhardt 1979:1067).

The house was raised on a carefully constructed circular mound of stones, itself named and often used as a designation for clans which claimed it as their original ancestral site. The house, built originally on the occasion of an important clan ceremony, the labour of several years and result of much physical and ritual effort, was the pre-eminent male space: it represented the power of an agnatic group and its ancestors. The massive central column, rhea in Ajie, symbolised the clan chief, “the center pole — the clan's pillar” (Leenhardt 1979:110, 117-18), the putative senior of the active generation of patrikin, conceived in fraternal terms. When an important man died it was proclaimed, “The rhea has fallen” (Leenhardt 1930:5). The house was decorated with sculpted representations in bas-relief of paternal ancestors and was crowned with a huge sculpture in the round, stylised image of an ancestor-god. Above this sculpture towered a spire decorated with white conch shells, the topmost of which contained the totemic plant of the clan, representing the female principle. All these decorations were designed to ensure the permanent presence of ancestors and totem. The entire habitat was infused with mythic presence, but the most sacred places were the house mound itself and especially the main altar. Sacred stones in which the spirits ko in Ajie of ancestors resided were kept on the mound and there also were placed two poles ensuring ancestral guardianship of the gardens. Nearby, in a thicket of poles, araucaria pines, cordyline ‘palm-lilies’ and crotons, was the altar, ka moaro, where the most important rituals affecting the affairs of the clan generally were conducted. It contained two hearths: one for the - 388 yam, dry and therefore male plant par excellence; the other for taro and other vegetables or fruits symbolically associated with women because of their moist quality (Leenhardt 1937:59-74). While men of the group shared the great house, there held their councils and received guests, women lived individually in smaller, more modest round houses. Behind the great house was a favourite haunt of the totem, which resided permanently behind the women's houses. 9 The basic building blocks of a binary system of classification encapsulated in the fundamental categories of sex were implicit, though not elaborated, in the purely male domain of the great house.

The great house was physically the culminating point of the habitat. It was approached via one broad and two, sometimes more, narrower avenues, evenly grassed and bordered by graceful rows of trees, especially coconut palms and araucaria pines. The symbolic significance of the habitat in terms of the complementary duality of male and female elements was most explicit in the context of these avenues. The central avenue represented the social aspect of maleness, the patriclan. The smaller flanking avenues represented the social aspect of femaleness, the wives of clan members and maternal kin, and were reserved for the latter during ceremonies. In these peripheral avenues some of the most intimate and important rituals were conducted, including those in honour of the maternal uncles, guardians of life of clan members. Life sprang ultimately from maternal totems. By totemism Leenhardt meant a sense of intense identification felt by Melanesians as members of a descent group with “a living or moist object: animal, plant of the moist regions, water, rain-laden wind, or thunder (...implying storm)” (1930:206). 10 A nephew's totem resided in the nape of the neck, the back and the chest of his maternal uncle. Life was transmitted through blood from mother to child in an endless flux over generations, which gave maternal kin a central role as life-givers to patriclan members, past, present and future (Leenhardt 1930:ch.5, 209, 1979:94-7).

The trees decorating the avenues and grown about the house and the altar variously symbolised the qualities of maleness and femaleness. Coconut palms and araucaria pines, crotons and the cordyline called diro in Ajie, which was always planted close to the house, were dry plants, male; erythrina ‘coral tree,’ banana and other moist, fleshy plants were female and were to be found along the flanking avenues and often at the end of the central avenue furthest from the great house. The actual species of tree grown to symbolise the complementary relationship of male and female could change, but the symbolic referents remained constant. The coconut palm was a fairly recent immigrant, known universally by its Polynesian name. Its utility and graceful, decorative aspect made it rapidly ap- - 389 preciated and assimilated, but it had acquired no ritual significance, unlike the much more ancient araucaria pine and cordyline, long-lived, highly resistant species which symbolised the permanence and endurance of the clan (Leenhardt 1930:22, fn. 1; 1937:20-1, 39; 1979:92).

Leenhardt commented:

An equation was manifest in the spatial arrangement, trees and plants of the habitat, man:woman: :dry:moist; it

On the social and cosmological planes the paradigmatic duality, that organic, complementary union of man and woman in the conjugal couple, provided a model for the relationship between the elements in other ensembles, (Leenhardt 1937:146-8, 181-5; 1930:63-79; 1979: ch.7) 12 for example:

In the affective, non-reflective realm which characterised Melanesian thought, these ensembles were not clearly differentiated, so that, for example, the social role of the couple merged with the mythic role of gods/totem: each participated in a common mythic reality (Leenhardt 1979:102-7). From an analytic perspective, however, the grouping of elements in a particular column does not imply identification or sameness. Rather the schema indicates that the relationship between juxtaposed elements was like that between man and woman: “each [element] plays its own role, but each lacks distinct boundaries. Outlines remain blurred. A closer look shows us an identity in each pair which cancels out any opposition” (1979:74, 41-2, 163; 1930:98, fn.3; 1937:148). Thus, the success of fertilisation in both social life and cultivation depended on the unit of the couple (see below), while on a wider, cosmic plane reproduction of all life, - 390 human, social and vegetable, required both the life force of the totem (rhē) in Ajie and the reinvigorating power of gods, ancestors (bao), to the extent that Melanesians often conflated the two in a single name or seemed to Europeans to confuse their qualities and roles (Leenhardt 1930:chs.9-10, esp. 231-4; 1937:39; 1979:67,71). Again, the notions of spirit, ko ‘vital essence of a living or dead being’ and totem seem frequently to have overlapped or to have been used interchangeably, while in languages in the north of the main island the same term applied to both:

In passing, however, Melanesians told Leenhardt that “Ko belongs to the moaro [‘clan’], rhē belongs to the maternal line” (1930:205), and “Gods are to do with men and totems proceed from women” (1930:234; Clifford 1980:16). Leenhardt, moreover, maintained that they clearly differentiated gods and totem in essence: gods were the stuff of legend, former men to whom hero or divine status was attributed, who served as prototypes and had fulfilled an important function as innovators and sources of change (1930:ch.10; 1937:181-5; 1979:ch.4) 13 totems were embedded in the timeless realm of myth, which “sums up human experience, encloses it in a symbol, and thereby indicates to men what their attitude to life should be” (Leenhardt 1937:181). 14

It is important to note that the binary relationship did not correspond to constant positive or negative valuation. Rather, each set could be cross-cut by value loadings depending on context. 15 Thus, the attitude towards both gods and totem moved between fear/respect and affection/respect: they were sometimes unpredictable and could cause misfortune and disease, but both were vital to reproduction and well-being. Both, therefore, were propitiated and invoked and their permanent presence ensured through appropriate ritual and behaviour (Leenhardt 1930:chs.9 and 10; 1937:172-80; 1979:70-2).

The classification set out above is suggestive to a historian. There are indications in the sources of disjunctions between ideals and practice, especially in the sphere of politics, which in some cases may be viewed diachronically as a dialectical relationship between cultural structure and historical action. Thus, the schema has potential for analysis in terms of the enactment of categorical relationships, and of structural transformation as their content altered over time, or new categories were generated, rather than as a simple synchronic catalogue of binary sets (Sahlins 1981b:110- - 391 11, 121, 128-32; see also 1976:39-46).

The ensembles power/life and paternity/fraternity 16 encapsulate an ideal of complementary equilibrium between lineal and collateral ties, between clan and generation, between paternal and maternal kin, between father (clan chief) and maternal uncle, between political chief and master of the land. These relationships were normally complementary in practice, but instances of tension or imbalance were common. In practice, societies 17 were strongly agnatic: the main corporate groups were localised patriclans; inheritance and succession were patrilineal; land and resources were held collectively by the clan, though individuals could acquire and transmit to their male descendants usufructuary rights. Yet, according to Leenhardt, the idealised conception of social relations was in bilateral terms. Concern for genealogy was minimal, except in the case of clans holding paramount chieftainships (a political concept, to which I shall return), and even these had relatively shallow genealogies (Guiart 1963:643-45). Clan members were conceived generationally as successive collectivities of “brothers,” permanently paired through bilateral cross-cousin marriage alliance with a complementary group of matrilateral kin (also “brothers”), who provided the wives (and husbands) of clan members. They were said to have had no concept of biological paternity. Life was transmitted by women; a husband (ideally the “brother,” cross-cousin, whom a woman married and referred to as such [Leenhardt 1930:56-62]) prepared his wife to receive the mythic seeds of life as he prepared his gardens for her work in cultivation, and he contributed to the successful outcome of pregnancy by observing sexual abstinence as he would during the maturation of the crops. “The ancestral blood which circulates in the woman,” said Leenhardt, “is reinvigorated by the action of the man. The father is not a generator, he is a fortifier” (1937:144; 1979:65; Tjibaou and Missotte 1978.28). Infertility in a marriage was usually blamed on the infidelity of the husband, whose family would offer shell money to the family of the wife (remembering that ideally father-in-law was also maternal uncle, embodiment of the totem and guardian of totemic life) to compensate for the misconduct of their son and guarantee anew the flow of life (Leenhardt 1937:125, 1979:65).

Each generation of “brothers,” then, was conceived as offspring of the wives of the previous generation. The brothers of these women were the maternal uncles, who blew life into the ear of the newborn or ailing child, were honoured in the great life-crisis festivals of the clan, were rewarded or recompensed with shell money at the birth and death of their nephews and merged with them in dual terms of reference and address which symbolised the parity (reciprocity) of their relationship (Leenhardt 1930: 49-50, 54, 79-83, 213; 1937:122-5, 138-42, 174; 1979:95, 98-100; also see below).

- 392

The widespread preference for bilateral cross-cousin marriage between members of permanently allied clans illustrates the value these people placed on equivalence and stability, in this case between reciprocal wife-givers and wife-takers in on-going relationship. 18 In practice, the preference seems to have been largely unrealised, especially in the southern half of the island. In the eastern central region, clans were typically divided between exogamous moieties, but even here marriage was more dispersed in practice than the ideal of long-term alliance between paired clans (rather than moieties) would suggest (Guiart 1957:28-39). On the ground, the theoretical symmetry which divided each generation of “brothers” equally between intermarrying sections was not borne out. For a male or unmarried female, the agnatic group was the focus of daily interaction — admittedly constrained by fairly rigid rules of deference, respect and avoidance (Leenhardt 1930:63-9). In practice, matrilateral “brothers,” dispersed geographically among often quite distant clans, did not usually form a coherent group and were encountered only on ceremonial occasions, as Leenhardt acknowledged (1930:66-7). Ties of co-residence and co-operation, then, reinforced the patrilateral bias. Furthermore, political developments of some antiquity, but for which clear evidence exists only in relation to the periods immediately before and after the onset of intensive European contact, tended to turn the axis of the society through 90°, at least in the case of chiefly clans, so that an increasingly patri lineal emphasis appeared.

Despite the high cultural value attributed to women, in most public areas of social action they were, at least to European eyes, marginal or irrelevant. Moreover, despite, or more likely because of, their crucial roles in the totemic domain of life, in procreation and cultivation, women were peripheral to the day-by-day society of men, and vice versa. Women avoided the great house, while men avoided the women's houses, through respect for the totem. Women in general played secondary or onlookers' roles in ceremonies. Yams, male food, were valued (by men at least) above all others. Most major and certainly the most visible social, religious and political roles were male or were those generally filled by men — chief, maternal uncle, sculptor, warrior, most priests. 19 Wealth objects and valuables such as bark cloth, shell money, greenstone axes and necklaces and other regional specialties were mostly produced and owned by men and exchanged between them, though pottery-making was largely a female preserve and both sexes participated in the regular exchange of foodstuffs which occurred between coastal and inland dwellers (Leenhardt 1937:51-8, 79-82, 92-6). Residential patterns (generally patrivirilocal) and patterns of land holding (by males) and inheritance (usually patrilineal) reinforced this asymmetry on the ground by ensuring that married women, however much - 393 respected, were normally outsiders and performed their key roles in cultivation on land which they did not control. The situation of wives may have been less secure in the north, where clans were divided between two mutually and permanently hostile networks, labelled “phratries” by Guiart. A preference for network exogamy, especially for chiefs, meant that wives frequently belonged to enemy clans (Guiart 1957:21-7; 1963:647).

Maternal kin, especially maternal uncles, were owed respect amounting to reverence, to the extent that overt conflict with them or disrespect to maternal ancestors was strongly disapproved and feared. Maternal kin might also provide refuge and more practical aid for individuals and groups in conflict with patrikin. None the less, potential for tension between affinally related clans, greatest perhaps in the north, existed everywhere. On an individual level, young chiefs, firmly committed to the element of power, often found irksome the restraint demanded by their older maternal uncles. An ambitious clan could find itself in ethical conflict with matrikin, whose concern for the discipline and ritual observances necessary to ensure continuity of life might be at odds with the pursuit of power, prestige and riches (Leenhardt 1937:184-5; 1979:71-2). The potential for tension reached its peak in male funerary rites, the customary culmination of which was the ceremonial despoliation of the deceased's gardens, houses and property by the bereaved element in the dual form uncle/nephew. All maternal uncles and nephews had the right and duty thus to mark death's interruption to the flow of life, and to ensure that the dead man's spirit was accompanied by the trappings of the living in its transition to the state of bao ‘ancestor,’ for the items destroyed were believed to follow the dead into the invisible realm. Nephews, being younger, less assured and less affected by rupture to the flow of life, were generally more restrained in their actions than uncles (Leenhardt 1930:85; 1937:35-7). In the north, especially, war could occur after the final mortuary rites for a high chief, should the behaviour of his maternal kin be seen by his own clan and his allies as excessive ([Gagnière] 1905:28-33; Lambert 1900:245-49; Rochas 1862:268-69).

In cases such as the above, political considerations, in which the element of power predominated, would seem to have affected both the actions of the marauding matrikin and the responses of the host clan and its allies. Power and life were here in asymmetrical relationship, the norm of complementary equilibrium was ignored and schismogenic tensions were apparent. In passing, Leenhardt referred to developments in Fiji, as a result of which “their society gave primacy to the element of power.” 20 He - 394 illustrated this by reference to the particular stress in certain dual forms in the Bauan language, contrasted with equivalent forms in New Caledonian languages in which the element stressed was “not governed by a priority of power but by that of the life potential and the reality grasped through totemic symbolism, represented by the women, children and nephews.” Anachronistically, in terms of most modern scholarship, he associated the growing stress on power in Fiji with “Polynesian culture spreading out over the old Melanesian base” (1979:98-9). Leenhardt's usual attribution of significant social and cultural variation to the absorption of separate cultural strata resulting from external migration (e.g., 1979:4-5, 10, 43-59, 101, 118-119, 124) is ultimately untestable either way, but it is logically unnecessary if one accepts a dynamic notion of culture stressing inherent potential for change. Be that as it may, I suggest that there is evidence of the emergence of larger, political chiefdoms, sometimes following military conquest, in parts of New Caledonia before significant European contacts; as a result, the male element of power expanded in import in relation to the female element of life.

Change in the scale of political action seems, therefore, to have corresponded to structural transformation: to altered relationship between fundamental cultural categories (power and life). The widening of political scale seems also to have paralleled the generation of a major category, patriliny: succession to chiefly office in large chiefdoms was ideally and often actually by primogeniture. This contrasted with the more general ideal at clan level, where the incumbent of the sacred office of chief was conceived as the elder of the active generation of patrilateral brothers and was not necessarily the biological son of his predecessor (Leenhardt 1937:145, 149; 1979:111). The Ajie term used by the elders to refer to the chief was a plural form, paxani, meaning literally ‘the rising generation,’ but as a result of concentration on the first-born it had by Leenhardt's time become a singular (1935:222; 1979:108). Leenhardt, hinting at historical process, said: “The social succession of the elder brothers called [another dual term] father and son, preceded the organic succession called from father to son” (1937:145). An ideology of seniority of descent and the practice of primogenitural succession, suggested Meillassoux, were conducive to the development of more complex hierarchies and the emergence of classes in a kinship-based society, because they increasingly differentiated senior lineages from junior, and could lead to monopoly of political, economic and religious authority by groups or classes claiming higher genealogical rank and status, rather than the division of authority between separate office-holders within a kin group on the basis of age (Meillassoux 1978:146-50). 21

The longest chiefly genealogy recorded by Jean Guiart in the 1950s was - 395 of 11 generations and he collected at least a dozen of more than six generations (1963:643-45). Though genealogies are a notoriously unreliable method of dating, their existence and moderate depth in New Caledonia suggest that concern for patriliny in chiefly clans, probably associated with an expansion of the scale of political action, was established well before the advent of regular contact with the outside world after 1840. Documentary evidence confirms this presumption. In 1840 relatively large, powerful chiefdoms existed in several areas, especially in the Loyalty Islands, on the Isle of Pines and on parts of the east coast, while others were in process of expanding their influence, such as in the south of the main island. 22 There is no reason to suppose that these were the first of their kind, since political expansion was not historically a one-way process and political alignments were by no means permanent or immutable. In 1840 there was, moreover, marked regional variation in the scale, style and potential of political action. The Houaïlou valley, for example, was fragmented into numerous small, quasi-independent chiefdoms (Guiart 1963:ch. 1). Elsewhere the spheres of influence of large chiefdoms tended to be separated by semi-autonomous clans or small chiefdoms, which acted as buffers and intermediaries, and whose alignments at particular times usually reflected the contemporary political balance between their powerful neighbours. 23

This uneven pattern might be interpreted as reflecting the impact of several parallel processes of attempted “state formation” by ambitious chiefdoms, which sought, with varied success, to gain political sway over entire geographical or linguistic regions. 24 A number of structural variants can be discerned. A unitary chiefdom dominated the Isle of Pines and exercised considerable influence in the southern third of the main island (Goujon MS. 1848-54; Guiart 1963:ch. 5; Lambert 1900:256-66; Ta'unga 1968:43-54). Several more or less complex hierarchies existed on each of the Loyalty Islands (Guiart 1963:chs. 7-9; Leenhardt 1979:118-23). In the centre and north, a characteristic political and ritual dualism operating on a regional basis limited hierarchical development, but the opposed networks included influential chiefdoms (Douglas MS. 1972:197-202; 1979:35-42; Guiart 1966), several of which were to engage in spectacular political expansion during the early years of the colonial era. In the Canala-Kouaoua region, complex hierarchies, increasingly dominated by the Bwaghéa chiefdom, had developed in a context of ritual dualism (Doumenge 1974:48-68; Guiart 1963:ch. 4). Here, a possible concomitant of political consolidation appeared in a tendency to more concentrated village settlement. As political authority became more centralised, polynuclear, multiclan villages grew around the great houses of the four paramount chiefly clans which contended for regional supremacy (Doumenge 1974:67-8). 25

- 396

Though processes of political expansion were undoubtedly established well before European contact, the advent of traders and missionaries after 1840 gave further stimulus to several aggressive and ambitious chiefdoms. Europeans provided manufactured goods, especially steel axes and sometimes firearms, and the promise of access to new and powerful avenues of ritual and knowledge. Both could enhance military potential and prestige. The Isle of Pines, Canala and parts of the Loyalty Islands experienced early and fairly continuous contact with Europeans and paramount chiefly clans took advantage of this to further their political ambitions. 26 On the other hand, the smaller Kamba chiefdom of Paita, which in the 1840s was actually in process of expanding (or possibly re-expanding) its influence in the sparsely populated and impoverished south, had more fleeting and often violent relations with traders after 1847, and later collapsed during the first military campaign of the new French colonial régime in the 1850s. 27

In none of these cases can a simple cause and effect relationship between European contact and indigenous political expansion be invoked. For instance, Andrew Cheyne, a member of the first sandalwood expedition to visit the Isle of Pines in 1841, reported that the chief was “sole Monarch of the Isle of Pines he likewise claims Sovereignty over the S.E. end of New Caledonia, which he gained by conquest — being a great warrior and a very brave man” (1971:38). The reciprocal factor which helped attract Europeans to powerful chiefdoms where well-disposed chiefs were likely to provide orderly conditions and reasonable security, was of considerable importance. Further north, it was no accident that Hienghène and Pouébo, locations of two of the most powerful rival chiefdoms in the region, became centres, respectively, of trader and missionary activity, even though Hienghène was not particularly well endowed with sandalwood or bêche-de-mer, the main commodities sought by traders (Douglas MS. 1972:197-222; 1979:41-6). An indication that European presence alone did not ensure political success is the case of the Pouma chiefdom of Balade. This was the site for much of the period 1843-1858 of a mission station or a military post. The presence at Balade of potentially powerful Europeans seems to have helped discourage the Pouma's neighbouring enemies from overt aggression, but the chiefdom, always fragmented and numerically insignificant, grew steadily weaker and politically more ineffectual, until by 1858 it had virtually ceased to exist (Douglas MS. 1972:ch. 2, 202).

The nexus between power and politics, largely ignored by Leenhardt, and the expanding scale of the political arena in many areas of New - 397 Caledonia involved a different form of conceptual dualism from that represented paradigmatically by the conjugal couple. The latter, an ensemble of organically related, complementary elements, was itself a primary conceptual category, in which “identity...[cancelled] out any opposition” between the constituent elements, and “there...[were] no contradictions..., only contrasts” (Leenhardt 1979:74,41). With a widening of political scale, however, another mode of thought, more akin to the “apparently unlimited series of dualistic oppositions in Western thinking” (Willis 1975:125; Leenhardt 1979:41-2), became apparent. It involved complementary opposition of discrete, usually antithetical categories, which could no longer be subsumed within dual forms, though their meaning remained relative within a binary relationship. This mode was exemplified in processes which Leenhardt saw as rooted deep in the legendary past, but which accelerated during the colonial era: there were indications of an increasing stress in ritual on ancestors and gods; to Leenhardt this represented the beginnings of individuation and differentiation of a more sharply defined concept of self. The meaning and referents of some kinship terms changed and key concepts were extended to cover new dimensions of experience and political activity; concepts and relationships between categories were transformed as novel contexts of action emerged.

The degree of definition and ritual significance of gods varied from north to south. Melanesians told Leenhardt that ancestor worship was a relatively recent development which overlaid a more ancient mode. Once, they said, the dead had merged with the living landscape, the total space of great house, dwellings, avenues, trees, burial grounds, natural features, land, gardens, resources and the living (Leenhardt 1930:ch. 10; 1937:176-83; 1979:ch. 4), encapsulated in the beautiful Ajie term maciri, ‘the peaceful abode’ (Leenhardt 1930:24-5; 1935:173-4; 1937:43-4). Leenhardt explained that

The people of Houaïlou, who may have had a more ambiguous notion of gods than prevailed in societies further north, none the less claimed that “Eurigourou [founding ancestor of clan of that name] was the first to stop praying to the mountains in order to address the ancestors” (Leenhardt 1937:180; 1979:53).

In Leenhardt's view, the differentiation of gods from a unified sociomythic realm, where man, ancestors, totem and nature merged, was part of an intellectual process by which the concept of a person as a single entity - 398 became more clearly defined (1979:chs. 11 and 12). 28 Politically this was perhaps exemplified in the career in the 1850s and 1860s of the chief Gondou. His expansionist activities in the centre-north were on an unusually wide scale, to the extent that his attempted conquest of an entire region drove several defeated or apprehensive groups into the camp of a rival imperialism, that of France. He was killed in 1869 by local enemies, after his betrayal to the French, probably by a uterine “kinsman” (Guiart 1957:34). It is tempting to explain the latter's action in terms of an attempt by an agent of life to control the excesses of an agent of power by recourse to an outsider. Gondou's role and prominence may have been exaggerated by the French, in whose eyes he was an uncompromising opponent of civilisation and France, while locally he was regarded as an aberration, more dangerous to the previous status quo than the French, whose presence in that part of the colony before the 1880s was sporadic and uneven. He was certainly the most outstanding of several notable Melanesian personalities of the early colonial era, but was by no means an isolated phenomenon (Douglas 1980:28,40-9).

Exploration of ways in which concepts were elaborated to encompass new experiences and ideas or in response to changing circumstances provided a basic theoretical underpinning to Leenhardt's ethnographic method and to his notion of conversion, though he gave little attention in his published works to actual context. For a historian, evidence of changing concepts and relationships between categories is doubly significant. It helps explain local perceptions of changes in social and political institutions and values and it provides an index of structural transformations occurring in new action contexts. Reference has already been made to a relationship between alteration in political scale and in the kind of dualism which underlay Melanesian thought. In Leenhardt's view, processes of individuation in both the precontact and colonial eras resulted in “a new view of the world,” related to changes in linguistic usage, notably a tendency towards singularisation of dual and plural kinship terms:

In Ajie the term kanya, ‘maternal uncle,’ tended to replace rha, a reciprocal term indicating the parity of maternal uncle as father-in-law and nephew as son-in-law. According to Leenhardt, the use of kanya reflected a growing asymmetry between uncle and nephew, with the authority of the former supplanting the reciprocity, intimacy and familial ritual implied by - 399 rha (1930:60-1). Similarly,

Not only did nemui lose its dual usage, but also it came to be a singular, though literally a collective term. In the singular, however, it retained collective implications: “[These] ancient plurals passed into the singular state...recall groups from which a representative member has been detached, carrying with him all the qualities of the group” (Leenhardt 1930:62); 1935:201-2, 219, 221). 29 Transformation of concepts in novel action contexts thus occurred in terms of, rather than in opposition to or negation of, enduring cultural values.

Even more interesting to a historian, because of its political implications, is the extension of the term maciri ‘the peaceful abode’ to incorporate the lands, resources and people within a chiefdom. Later still, after annexation and Christianisation, it came to mean ‘government,’ ‘empire,’ ‘kingdom of god’ (Doumenge 1974:48-9; Leenhardt 1930:24-5; 1935:173-4; 1937:43-4). 30 In these contexts, however, it retained for users some of the affective connotations of its primary meaning, and its use is suggestive not only of the expansion of the category power, in conjunction with the expansion of politics, but also of very real constraints on such expansion. Despite evidence of asymmetry in practice in the power/life relationship and the emergence of a level of political action somewhat apart from the affective domain of kinship, none the less the concept of complementary opposition remains most useful, explicating the profoundly entrenched dualism of Melanesian ideological constructs and social and political structures.

In New Caledonia by the mid-19th century the development of a degree of political hierarchy had not been accompanied, as in Tahiti before 1767, by general transition to a class-based society, in which hierarchy had replaced dualism as the main social structural mode. Meillassoux's model (above), in other words, is only partly applicable in New Caledonia. Similarly, processes of state formation remained at a fairly rudimentary level, though the case of Gondou suggests that this might not always have been so had the development of indigenous polities not eventually been short-circuited by superimposition of an entirely new, exclusively alien level of political integration, with the gradual consolidation of the colonial régime after 1853.

- 400

Even in the Loyalties and the Isle of Pines, where hierarchy became the characteristic political mode, its further elaboration into a class system was limited by several factors, both cultural and social, which also applied more generally. The ideological basis of chieftainship remained the same, from the level of the clan, essentially a descent group, to the territorial political level of the chiefdom. Whatever the actual mode of succession, ultimately the status of every chief came from his putative position of seniority, an extension of the norm that authority within a group of siblings rested with the first-born male, though the notion of first-born was not always explicit in the terms used for chief. Chiefs were thus “elder” in relation to their dependants, “junior kinsmen,” over whom they could exercise considerable authority and from whom they received profound respect, deference and certain services: 31

The same set of values, however, placed real limitations upon the arbitrary or autocratic exercise of chiefly power. Chiefs always risked the ultimate sanction of replacement or desertion should they infringe their “kinsmen's” rights or be consistently unsuccessful, ipso facto evidence of ancestral displeasure (Guiart 1956:24; 1963:641; Lambert 1900:8-9; Leenhardt 1937:152; Rochas 1862:246-47). Politically, the key concept of relative seniority was not strictly genealogical. Over time it came to be a reflection of actual patterns of authority, respect and service as much as a determinant:

On the other hand, senior and junior in any group, however inclusive, were complementary, since some of the latter controlled the land or were responsible for ritual communication with the ancestors, who could affect the outcome of every activity (Doumenge 1974:48-68; Leenhardt 1935:140-41; 1979:121). Throughout the archipelago, chiefs were commonly said to have been recent arrivals, incorporated at the apex of the political structure but often enjoying few personal or clan rights to land. 32 Related to this, almost everywhere a clear distinction existed between - 401 political chiefs and masters of the land, the chiefs of clans acknowledged as first occupants (Doumenge 1974:49,56); Guiart 1963:35, 41, 645-46; Leenhardt 1935: 140-41; 1937: 151). This was one manifestation of a characteristic “complete dissociation of material power from affective prestige” (Leenhardt 1979:122; Guiart 1966:27) 33 inherent in the structure of Melanesian dualism in New Caledonia. Leenhardt explained the complementary dualism of chiefs and masters in relation to the Loyalties, but the basic principle applied generally:

In complementary opposition to the power of chiefs, then, masters of the land represented the female principle of life, and could, like maternal uncles, restrain the unbridled exercise of power. A conqueror, for example, would normally leave the masters in possession of their lands in order not to offend the mythic forces with which they were in communion (Leenhardt 1930:45; 1935:141; Rochas 1862:242-43). 34

On the mainland of New Caledonia, especially in the centre and north, hierarchical development was restricted by the operation of characteristic institutions of ritual and political dualism. This led on a regional basis to identification of local groups with one or other of two parallel networks, for example in the north where clans were linked in patchwork fashion to either of two mutually hostile, intermarrying networks known as Hoot and Waap (Douglas 1980:29-30). The lateral importance of alliance and the relative balance regionally of the power of opposed networks meant that political hegemony was difficult to achieve and tended to be short-lived (Guiart 1963:630-33, 639-40, 646). In the north, indeed, conquest of territory seems traditionally not to have been a feature of network opposition (Guiart 1957:23-7; 1963:647), which, though often manifested in war, implied mainly symbolic and ceremonial prestige. The most striking historical attempts to extend political control over a wide area seem either to have transcended network identification (the expansion of Paci speakers in the centre-north, culminating in the activities of Gondou [Douglas - 402 1980:47-9]), or to have been affected by factors related to the presence of missionaries or colonisers (e.g., the successive expansion of the already powerful Mwelebeng chiefdom of Pouébo and Bouarate chiefdom of Hienghène during the colonial era [Douglas 1979:49-56; 1980:41-7]).

Related to all these factors in constraining hierarchical development and emergence of a class-based society was the high incidence of internal migrations, usually on a very small scale, and originating in part from the norm that the weaker party in a conflict — usually a “junior sibling” — might withdraw and take his allegiance elsewhere (Guiart 1953:93-97; 1963: passim, esp. 640-41; 1966). While this may have created opportunities for ambitious chiefs to recruit a following, it also provided a sanction against chiefly autocracy, the absence of which in turn meant that the various marks of deference shown to a chief, in gesture, posture and modes of address and reference, remained, in Leenhardt's terms, a “lived” ritual of respect (1979:111-12), with no suggestion of sterile etiquette enforced by fear. A fairly high level of mobility also contributed to the development and maintenance of supra-local networks, seen most clearly in the institutional dualism of the centre and north, and necessitated elaboration of mechanisms for the absorption without prejudice of newcomers, whether strangers, distant kin, refugees or conquered enemies. Stereotypically, this absorption was often said to have occurred at the highest political levels, and resulted in the complementary opposition of chiefs and masters of the land.

An overriding impression to emerge from reconstruction of some precontact New Caledonian cultural, social and political processes, using data drawn from Leenhardt and the range of historical, ethnographic and traditional sources available, is of equilibrium, persistently re-emerging through changes in political context and cultural structure. I do not mean this in a functionalist sense of homeostasis or absence of conflict, but see it rather as a corollary of dynamic opposition of complementary concepts, values and institutions, of processes which contained or channelled, effectively for the most part, potential excess or extremism. 35 The balance between the changing and conflicting demands of power and life was to be severely tested during the colonial era, but it remains a characteristic motif in Melanesian culture, social relations and political style.

The remainder of this paper briefly sketches some implications for Melanesian culture and society of aspects of European contact and French rule, examined in the light of the theoretical perspective outlined, with particular reference to chiefly authority and relationship between the categories, power and life. The historical processes in question were of a - 403 different order from those already discussed, which had occurred in entirely Melanesian contexts. From the 1840s, the gradual spread of European technology and settlers created new action contexts for Melanesians, while the uneven but inexorable extension of colonial rule led to an entirely new level of political and economic integration, in which Melanesians scarcely participated, but which affected their lives in numerous ways. To a greater or lesser extent, new contexts involved the reworking of old and the formulation of new categorical relationships, in respect to other Melanesians as well as to Europeans. However, Melanesian understanding of and responses to European intentions rarely, if ever, matched the latter's expectations, and structural change was predicated upon the internal logic and style of a stable, yet dynamic culture.

Reference to notable individual Melanesians was common in trader, missionary and colonial records, most often, predictably, in relation to larger chiefdoms. This reflected a pre-existing trend to sharper definition of gods, ancestors and their living manifestations, chiefs, related to pre-contact processes of expanding political scale. It also reflected Western expectations about the ways in which primitives were governed and the European search for an individual locus of authority through which they could operate in dealings with local peoples. It was probably inevitable that, finding in chieftainship an institution which looked like the apex of a feudal structure, many Europeans should ignore or discount complementary institutions and dignitaries — councils of elders, masters of the land, maternal uncles — which functioned in part to offset and control the exercise of chiefly authority, and to subject potential excesses of power to the restraining requirements of life. Of particular significance, especially in the first three decades of colonial rule, was the extent to which some chiefs took advantage of European presence and misconceptions to gain greater freedom of action, thereby extending an already apparent asymmetry in practice between the categories, power and life.

The factor most stressed by Leenhardt as contributing to the articulation of a more sharply defined concept of self was the impact of Christian ideology, exclusively Catholic, except in the Loyalty Islands, until the end of the 19th century. Certainly, all Christian missionaries did seek to impart a notion of the individual spirit in personal communion with God. However, Leenhardt's most respected informant, the pastor and former sculptor Boesoou, provided a Melanesian interpretation of this process in an oft-quoted epigram. In response to the missionary's question, “We introduced the notion of spirit to your way of thinking?”, Boesoou retorted: “Spirit? Bah! You didn't bring us the spirit. We already knew the spirit existed. We have always acted in accord with the spirit. What you've brought us is the body” (Leenhardt 1979:164, chs. 11 and 12; R. - 404 Leenhardt 1976). Furthermore, as Leenhardt came to realise, and as modern Melanesians confirm, a profoundly collective and mythic mentality remains, whereby a person retains a sense of intense participation and identification with the groups of which he or she is a member, and with nature and the extrahuman world (Leenhardt 1979:70-3, ch. 12; Clifford 1980; Tjibaou and Missotte 1978).

Though individualistic in spiritual terms, however, most missionaries acknowledged pragmatically that their cause would only prosper with the support of a strong, well-disposed temporal authority. 36 In the early years of missionary work they enjoyed most success in areas where they were able to enter into a mutually beneficial relationship with politically ambitious chiefs, who, for their part, may have hoped not only to gain access to another source of extrahuman power, but to obtain new ritual experts who would allow them greater autonomy than the old. The presence, esoteric knowledge and ideas of missionaries could help strengthen the position of such chiefs in relation to complementary dignitaries within their own chiefdoms, as well as provide tactical advantages against rival clans and chiefdoms in a wider political sphere. In both the Mwelebeng chiefdom of Pouébo and the Belema chiefdom of Belep, for example, coalitions of leading chiefs and missionaries achieved greater concentration of political authority, in opposition to priests, sculptors and masters of the land, who restrained chiefly power and adhered persistently to older values and ideas (C[olomb] 1890:9; Douglas MS. 1972:204-28; Guiart 1966:24-30; Lambert 1900:10-12). Again, one of the most successful programmes of military expansion in the north before 1862 was pursued by the Mwelebeng, whose political chief, Bonou, had been an early convert to Christianity. Bonou and the missionaries both benefited from the campaign: within his chiefdom, Bonou's authority was confirmed; he extended political control or influence over neighbouring clans and chiefdoms; most remaining Mwelebeng pagans and many adjacent peoples professed Christianity following the demonstration that missionaries could provide useful support in local wars, which missionaries phrased as a Christian crusade against pagans, but which were generally seen locally as traditional conflicts (Douglas MS. 1972:216-22, 342, 387-8; 1979:49-50; 1980:42,45). Aggressive political expansion was not confined to Christian chiefs, as the examples of Gondou, and Bouarate of Hienghène indicate. The latter, however, seem to have concentrated on extending their political authority over other groups, in a manner already noted as widespread before European contact. In alliance with missionaries, some Christian chiefs pursued courses of action which redefined the relationship between power and life within their chiefdoms. At Pouébo, certainly, and probably also at Belep and Wagap, the pragmatic coincidence of chiefly and missionary intentions - 405 enabled internal government to become more centralised and autocratic.

Redefinition of categorical relationships, correspondence between chiefly aims and missionary prejudices, and centralisation of authority within wider political units are all implied in an account written in 1861 by an administrator visiting Pouébo. At missionary suggestion and under Bonou's orders, the people were about to begin production of coconut oil for sale to a European trader:

Missionaries applied a treasury mentality to such customs as the pooling of resources within families and clan, and the offering of firstfruits to chiefs, in a gesture of respect by those dependent on or beholden to them. The subsequent ceremonial redistribution of such offerings at festivals symbolised to missionaries all that was most excessive and wasteful in Melanesian culture, its so-called “primitive communism.” 37 In consequence, they encouraged pooling of saleable resources under chiefly control, on a village-wide or “tribal” basis, for approved projects such as church-building, and valorisation of individual labour to inculcate a work ethic and enable people to buy small items of European manufacture. 38

From the onset of trade with Europeans in the 1840s, some chiefs had sought to augment their power by monopolising commerce and controlling their “juniors'” access to European goods. In larger Christian chiefdoms, church-building projects demonstrated publicly the extent of a chief's authority and capacity to co-ordinate production of key items on a far wider scale than would formerly have been possible or necessary. Traditionally, the great house, symbol of the power of a clan and its ancestors, had been constructed by clan members, aided by allies and kinsmen, under the leadership and direction of priests and sculptors. It was inaugurated by a major ceremony, at which great quantities of yams and other foodstuffs were distributed to guests by the chief on behalf of the clan. The church, symbol of the power of an alien god and his chiefly and - 406 missionary advocates, was built by foreigners with the fruits of the labour of members of numerous clans, under the leadership and direction of chiefs, and with missionary advice. It served a wider congregation, to most of whom it offered reflected prestige rather than direct participation. All these procedures altered traditional exchange relationships, in which even pool-sharing and redistribution were phrased in terms of reciprocity, because, to the chiefs and missionaries concerned, social and affective aspects were secondary to political and economic considerations.

The best that can be said of both the policy and practice towards Melanesians of successive colonial administrators before the Second World War is that they were profoundly expedient (Douglas MS. 1972: passim, esp. chs. 4, 6, 7; Saussol 1979). In the first three decades of the colony, in particular, when substantial areas remained outside effective control, official policies and practice tapped a congruent, though by no means identical, strain of expediency in some chiefs (e.g., Douglas 1979:52-6; 1980:45-7, 50-1). There was an inherent contradiction in officials' attitudes to chiefly authority, especially in relation to large chiefdoms. On the one hand, entirely ignorant of the key role of the category, life, and the importance of the affective domain, dismissed as “superstition” and “fetishism” by most Europeans, they deplored the traditional subservience of “the people,” seen as slaves to the whims of ruthless and arbitrary despots; on the other hand, they promulgated policies which could be used by chiefs to enhance their authority within “tribes” (the basic Melanesian administrative units in the colony, bearing increasingly little relationship to traditional clans and chiefdoms) and to diminish the role of consultation and consensus. 39 For example, a ban on Melanesians leaving their reserves without official authorisation, progressively imposed as tribes were formally constituted and reserves delimited, 40 circumscribed one of the main customary sanctions against chiefly abuses, the threat of desertion.

Again, the colonial administration denied the existence traditionally of personal usufructuary rights to land and legislated communal ownership of reserve land, under the control of tribal chiefs. In each tribal reserve a separate terrain was legally set aside solely for the use of the chief, who was thereby freed from reliance on the masters of the land and the offerings of his dependants. 41 The administration appointed or ratified chiefs, whom it regarded and treated as intermediaries responsible to the governor of the colony and his representatives. This created a context in which the translation into French of the term kamosari ‘juniors,’ as “subjects” (see above, fn. 30) could be imbued with some practical content by chiefs so disposed. Especially before 1880, there was enhanced potential for arbitrary and autocratic behaviour by chiefs, notably within the new administrative unit of the tribe, though that potential seems to have been unevenly - 407 realised. Reports of veritable satrapy in some tribes are balanced by others which suggest that the nominee of the administration was a figurehead, pushed forward to deflect the attention of the Government, while customary leaders continued to exercise authority in the background. 42

By the end of the 19th century, most mainland clans had been resettled on limited reserved land to which they had no traditional entitlement (Saussol 1979:281-304, 356). The enforced disjunction of human groups from the land with which they identified and in which their ancestors resided, from their maciri, brought great dislocation to people to whom land was not “a region — a term which implies territorial continuity — but a heterogeneous ensemble of places animated by human groups joined together by the bond of altar and clan” (Leenhardt 1979:106; Anova-Ataba 1969:207; Tjibaou 1976:284-85). Legally all Melanesians were under the authority of the tribal chiefs endorsed by the French, and had communal rights to land in the reserves to which they were assigned. In fact, newcomers to reserves were usually dependent for use rights to land on the goodwill and generosity of those clan chiefs acknowledged locally as masters of the land (Bélouma 1958-9:21-2; Doumenge 1974:203-4). Thus, masters, largely unknown to the administration, often retained considerable authority within reserves, in continued, or more probably renewed, complementary opposition to political chiefs. Since the Second World War, councils of elders have reappeared, but on a tribal or village, rather than a clan, basis.

Modern Melanesian habitats have greatly changed. Traces of the old order remain, including a fondness for araucaria pines and coconut palms, but settlements are larger and more concentrated than the precontact norm, and contain several clans. A modern village is commonly aligned between the two poles of the church and the house, rectangular in shape and usually constructed of cement brick, of the tribal chief, thus separating the religious and political functions which, at clan level, had traditionally been combined in the great house (Avias 1953:141-50; Doumenge 1974:106-11; Guiart 1956:40; Leenhardt 1979:93). More recently, as representative government has come to the local context, the mairie ‘town-hall’ has in places begun to rival the tribal chief's house as a focal point, psychologically if not physically. Perhaps a new version of complementary opposition, balancing the customary authority of chiefs (now chosen by the appropriate tribal or village council rather than nominated by the administration) against the wider political potential represented by elected mayors and local members of the territorial assembly, is now in process of formation. This needs to be explored empirically, however, and such relationships would clearly undergo redefinition should Indépendance Kanak be achieved.

- 408

It seems appropriate to conclude that mission and official policies and practice coincided in some cases during the 19th century with the aggressive extension of chiefly authority by ambitious political chiefs, and corresponding weakening of customary constraints against chiefly abuse. During the 20th century a relative balance between power and life seems generally to have been restored in Melanesian communities, or has emerged in new guises, though serious tensions exist, especially over land, economic development and, latterly, independence. The coming of Europeans and more particularly the development of colonial rule introduced to dynamic but closely integrated cultural and social contexts a new technology, new ideas, new social relationships and new frameworks for action. These were evaluated, interpreted and imbued with meaning by Melanesians in terms of their received concepts and categories. Aspects of the new were incorporated and enacted differentially according to the (culturally motivated) interests and relations which individuals, categories and groups of people brought to particular sets of circumstances. Differential action in new contexts could lead to altered relationships between Melanesians, as between political chiefs and complementary dignitaries. As a result, relationships between basic cultural categories, such as power and life, were transformed. Thus, cultural structure changed, but it did so in terms of its own logic, not as an automatic reflex of encounters with an “objective” external reality. On another level of practice, colonial rule led ultimately to the development of archipelago-wide social, political and economic hierarchies. These, however, owed nothing to Melanesian initiatives or control and Melanesians remain in the lower strata of colonial society. Beyond the tribal level, colonial rule progressively contracted and ultimately removed the freedom of action (and movement until the late 1940s) of all Melanesians, including chiefs. Melanesians did not “lose” their culture, though its structure changed. Especially on the mainland, however, they lost control of most of their lands and much of their destinies. Many today are asserting the primacy of their culture, and are trying to regain control of their islands and their lives.

- 409 Page of endnotes

- 410 Page of endnotes

- 411

• Annales de la propagation de la foi, 1858, 30.
• ANOM = Archives nationales, section Outre-mer, série Nouvelle-Calédonie, Paris.
• ANOVA-ATABA, Apollinaire, 1969. “Deux exemples de réflexions Mélanésiennes: 1. L'insurrection des Néo-Calédoniens en 1878 et la personnalité du grand chef Ataï; 2. Pour une économie humaine.” Journal de la Société des Océanistes, 25:201-37.
• APM, ONC = Archivio di Padri Maristi, section Oceania, Nova Caledonia, Rome.
• APO = Archives of the Province of Oceania, formerly Sydney, now incorporated in APM.
• AVIAS, Jacques, 1953. “L'évolution de l'habitat indigène en Nouvelle-Calédonie de 1843 à nos jours.” Journal de la Société des Océanistes, 9:129-50.
• BARTH, Fredrik, 1966. Models of social organization. Royal Anthropological Institute Occasional Paper No. 23. London, Royal Anthropological Institute.
• BATESON, Gregory, 1958. Naven. Stanford University Press.
• BELOUMA, G., 1958-9. “Problèmes économiques et sociaux autochtones à Gomen.” Études Mélanésiennes, nouvelle série, 10-11 (12-13):20-31.

- 412

• [BOUTIN], Félix, 1951. “Essai d'établissement des missionnaires à Yaté (Oct.1849-Janv.1850).” Études Mélanésiennes, nouvelle série, 3(5):5-39.
• Bulletin officiel de la Nouvelle-Calédonie. Noumea.
• CHEYNE, Andrew, 1971. The trading voyages of Andrew Cheyne, 1841-1844 (ed. Dorothy Shineberg). Pacific History Series No. 3. Canberra, Australian National University Press.
• CLIFFORD, James, 1980. “The translation of cultures: Maurice Leenhardt's evangelism, New Caledonia 1902-1926.” Journal of Pacific History, 15:2-20.
• ——, 1982. Person and myth: Maurice Leenhardt in the Melanesian world. Berkeley, University of California Press.
• C[OLOMB], Antoine, 1890. La tribu de Wagap, ses moeurs et sa langue, d'après les notes d'un missionnaire mariste coordonnées par le P.A.C. Paris, Publications de l'oeuvre de Saint-Jérôme.
• CUNNINGHAM, Clark E., 1972. “Order in the Atoni house,” in William A. Lessa and Evon Z. Vogt (eds.), Reader in comparative religion: an anthropological approach. New York, Harper and Row.
• DOUGLAS, Bronwen, 1970. “A contact history of the Balad people of New Caledonia 1774-1845.” Journal of the Polynesian Society, 79:180-200.
• ——, MS. 1972. A history of culture contact in north-eastern New Caledonia, 1774-1870. Unpublished Ph.D. thesis, Australian National University, Canberra.
• ——, 1979. “Bouarate of Hienghène: great chief in New Caledonia,” in Deryck Scarr (ed.), More Pacific islands portraits. Canberra, Australian National University Press.
• ——, 1980. “Conflict and alliance in a colonial context: case studies in New Caledonia, 1853-1870.” Journal of Pacific History, 15:21-51.
• DOUMENGE, Jean-Pierre, 1974. Paysans Mélanésiens en pays Canala-Nlle Calédonie. Travaux et documents de géographie tropicale No. 17. Talence (France), Centre d'Études de Géographie tropicale.
• ERSKINE, John Elphinstone, 1853. Journal of a cruise among the islands of the western Pacific...in Her Majesty's Ship Havannah. London, John Murray.
• FEIL, D. K., 1980. “Symmetry and complementarity: patterns of competition and exchange in the Enga tee.” Oceania, 51:20-39.
• [GAGNIERE, Matthieu], 1905. Étude ethnologique sur la religion des Néo-Calédoniens. Saint-Louis, Imprimerie catholique.
• GOUJON, Prosper, M.S. 1848-54. Journal de l'île des Pins. TS copy. Archives du Vicariat apostolique de la Nouvelle-Calédonie, Noumea. MF, Mitchell Library, Sydney.
• GUIART, Jean, 1953. “Nouvelle-Calédonie et îles Loyalty. Carte du dynamisme de la société indigène à l'arrivée des Européens.” Journal de la Société des Océanistes, 9:93-7.
• ——, 1956. L'organisation sociale et coutumière de la population autochtone. Published with Jacques Barrau, L'agriculture vivrière autochtone de la Nouvelle-Calédonie. Noumea, Commission du Pacifique Sud.
• ——, 1957. “Les modalités de l'organisation dualiste et le système matrimonial en Nouvelle-Calédonie.” Cahiers Internationaux de Sociologie, nouvelle série, 22:21-39.

- 413

• ——, 1963. Structure de la chefferie en Mélanésie du sud. Travaux et Mémoires de l'Institut d'Ethnologie 66. Paris, Institut d'Ethnologie.
• ——, 1966. Mythologie du masque en Nouvelle-Calédonie. Publications de la Société des Océanistes No. 18. Paris, Musée de l'Homme.
• HALLPIKE, C. R., 1976. “Is there a primitive mentality?” Man, 11:253-70.
• HORTON, Robin, 1962. “The Kalabari world-view: an outline and interpretation.” Africa, 32:197-219.
• ——, 1967. “African traditional thought and Western science.” Africa, 37:50-71, 155-87.
• ——, 1976. “Stateless societies in the history of West Africa,” in J. F. A. Ajayi and Michael Crowder (eds.), History of West Africa, Vol. I. London, Longman.
• HOWE, Kerry, 1979. “The fortunes of the Naisilines: portrait of a chieftainship,” in Deryck Scarr (ed.), More Pacific islands portraits. Canberra, Australian National University Press.
• LA HAUTIERE, Ulysse de, 1869. Souvenirs de la Nouvelle-Calédonie. Paris, Challamel.
• LAMBERT, Pierre, 1900. Moeurs et superstitions des Néo-Calédoniens. Noumea, Nouvelle Imprimerie Nouméenne.
• LEENHARDT, Maurice, 1930. Notes d'ethnologie Néo-Calédonienne. Travaux et Mémoires de l'Institut d'Ethnologie 8. Paris, Institut d'Ethnologie.
• ——, 1935. Vocabulaire et grammaire de la langue Houaïlou. Travaux et Mémoires de l'Institut d'Ethnologie 10. Paris, Institut d'Ethnologie.
• ——, 1937. Gens de la Grande Terre. Paris, Gallimard.
• ——, 1946. Langues et dialectes de l'Austro-Mélanésie. Travaux et Mémoires de l'Institut d'Ethnologie 46. Paris, Institut d'Ethnologie.
• ——, 1947. Do kamo. La personne et le mythe dans le monde Mélanésien. Paris, Gallimard.
• ——, 1979. Do kamo: person and myth in the Melanesian world (tr. Basia Miller Gulati). Chicago and London, University of Chicago Press. (English translation of Leenhardt 1947.)
• LEENHARDT, Raymond-H., 1976. “Un sociologue canaque: le pasteur Boesoou Erijisi, 1866-1947.” Cahiers d'histoire du Pacifique, 4:19-53.
• Le Moniteur de la Nouvelle-Calédonie. Journal officiel de la colonie. Noumea.
• LENORMAND, Maurice, 1953. “L'évolution politique des autochtones de la Nouvelle-Calédonie.” Journal de la Société des Océanistes, 9:245-99.
• LÉVI-STRAUSS, Claude, 1966. The savage mind. London, University of Chicago Press.
• ——, 1972. “The bear and the barber,” in William A. Lessa and Evon. Z. Vogt (eds.), Reader in comparative religion: an anthropological approach. New York, Harper and Row.
• LÉVY-BRUHL, Lucien, 1923. Primitive mentality (tr. Lilian A. Clare). London, George Allen and Unwin.
• MEILLASSOUX, Claude, 1978. “‘The economy’ in agricultural self-sustaining societies: a preliminary analysis,” in David Seddon (ed.), Relations of production. Marxist approaches to economic anthropology. London, Frank Cass.
• Melanesia 2000. Festival des arts Mélanésiens, Nouvelle-Calédonie. Nouméa du 3 au 7 Septembre 1975. Noumea.

- 414

• METGE, Joan, 1976. The Maoris of New Zealand: Rautahi. London, Routledge and Kegan Paul.
• MONCELON, Léon, 1885. Rapport...sur les mesures prises et à prendre à la Nouvelle-Calédonie dans l'éventualité d'une nouvelle insurrection des indigènes. Mayenne, Imprimerie A. Derenne.
• MONTROUZIER, Xavier, 1860. Notice historique, ethnographique et physique sur la Nouvelle-Calédonie. Paris, Lahure.
• OLIVER, Douglas L., 1974. Ancient Tahitian society, Vol. II. Canberra, Australian National University Press.
• PATOUILLET, Jules, 1872. Voyage autour du monde. Trois ans en Nouvelle-Calédonie. Paris, E. Dentu.
• ROCHAS, Victor de, 1862. La Nouvelle-Calédonie et ses habitants. Productions, moeurs, cannibalisme. Paris, F. Sartorius.
• SAHLINS, Marshall, 1976. Culture and practical reason. Chicago, University of Chicago Press.
• ——, 1981a. Historical metaphors and mythical realities. Structure in the early history of the Sandwich Islands kingdom. Association for Social Anthropology in Oceania Special Publications No. 1. Ann Arbor, University of Michigan Press.
• ——, 1981b. “The stranger-king: or Dumézil among the Fijians.” Journal of Pacific History, 16:107-32.
• SARASIN, Fritz, [1917]. La Nouvelle-Calédonie et les Îles Loyalty. Souvenirs de voyage d'un naturaliste (tr. Jean Roux). Bâle, Georg.
• SAUSSOL, Alain, 1979. L'héritage. Essai sur le problèe foncier Mélanésien en Nouvelle-Calédonie. Publications de la Société des Océanistes No. 40. Paris, Musée de l'Homme.
• SMITH, Jean, 1974. Tapu removal in Maori religion. Polynesian Society Memoir No. 40. Wellington, Polynesian Society.
• SHINEBERG, Dorothy, 1967. They came for sandalwood. A study of the sandalwood trade in the south-west Pacific, 1830-1865. Melbourne University Press.
• STEPHEN, Michele, 1979. “An honourable man: Mekeo views of the village constable.” Journal of Pacific History, 14:85-99.
• TA'UNGA, 1968. The works of Ta'unga. Records of a Polynesian traveller in the south seas, 1833-1896 (ed. R. G. and Marjorie Crocombe). Pacific History Series No. 2. Canberra, Australian National University Press.
• THIERCELIN, L., 1866. Journal d'un baleinier. Voyages en Océanie, Vol. I. Paris, Hachette.
• TJIBAOU, Jean-Marie, 1976. “Recherche d'identité Mélanésienne et société traditionnelle.” Journal de la Société des Océanistes, 32:281-92.
• —— and Philippe MISSOTTE, 1978. Kanaké: the Melanesian way. Papeete, Les Editions du Pacifique.
• TURNER, Victor, 1974. The ritual process. Structure and anti-structure. Harmondsworth (Middlesex), Penguin Books.
• VINCENT, J.-B. Maurice, 1895. Les Canaques de la Nouvelle-Calédonie. Esquisse ethnographique. Paris, Challamel.
• WILLIS, Roy, 1972. “The head and the loins: Lévi-Strauss and beyond,” in William A. Lessa and Evon Z. Vogt (eds.), Reader in comparative religion: an

- 415

• anthropological approach. New York, Harper and Row.
• ——, 1975. Man and beast. London, Paladin.
• WOODIN, Edward, 1850. “Sandalwood voyage.” Nautical Magazine and Naval Chronicle, 19:298-303.

- 416 Page is blank