- 167

A series of recent publications by Sahlins (1981a, 1981b, 1983) illuminates the relationship between history and structure in the Oceanic context by charting a dialectical course between two contrasting approaches. The first, popular with historians, holds that the organisational categories of a culture in place at any one moment are the cumulative outcome of the actions and interests of individuals; in this view, structure is statistical history. The second, more common among anthropologists, assumes that cultural categories are rigid, synchronic impositions upon individual experience and contextual action; thus, history is the inevitable expression of some immutable “culture-asconstituted”. The solution Sahlins proposes is an elegant one: by combining these two analytical concepts into a unified notion of “structural history”, he points to the diachronic interplay of cultural categories and particular events. As categories become functionally revalued in the process of social life, structure emerges precisely as the pattern of this transformation. So culture, in this way of thinking, is profoundly historical—constituted by real powers and their concrete interests—and also structural—revealed in the systematicity of diachronic processes.

This processual definition of structure as “a dynamic development of the cultural categories and their relationships, amounting to a world system of generation and regeneration” (Sahlins 1981b:111) sheds new light on two aspects of Belauan 1 culture, the linkage between which seems at first glance puzzling. The first aspect is a set of diagrams or - 168 models pervasively realised in traditional sociopolitical organisation, namely, linearity or ‘paths’, quadripartition or ‘cornerposts’, and dual opposition or ‘sides’. As examples of ‘paths’, chiefly houses in different villages are said to be associated by ‘paths’ or relationship dictating normative channels of social actions such as exchange, title inheritance, land ownership and co-operation; relations of mutual hostility between villages are labelled by the name of the forest or lagoon path connecting the two places; and individual claims to male or female titles are evaluated according to the strength of the matrilineal ‘path’ to the social unit holding the title. For ‘cornerposts’, the four highest ranking titles in a village council are called ‘four cornerposts’, referring to the corners of the meeting-house floor where holders of these titles take their assigned seats for political discussion; the four ranking houses of the village are also called ‘four cornerposts’, since patterns of social exchange, co-operation, and political alliance among affiliated houses revolve around these nodal units (Parmentier 1984). Finally, the model of two opposite ‘sides’ organises the structure of village men's clubs into competitive ‘sides of the mangrove channel’ for the purposes of recruitment, warfare, feasting, and labour; also, traditionally sanctioned divisions or moieties of high-ranking houses which express balanced competition over titles and ceremonial precedence are known as ‘one side leg’ and ‘the other side leg’.

The second aspect of the culture is a set of myths narrating a progressive sequence of cultural eras, each expressed as a distinctly shaped order of villages and chiefs. A single term, renged (found in the verb merrenged ‘bind together, tie up in a bundle’), combines the temporal and spatial meanings of the English glosses ‘era’ and ‘polity’. The political organisation of villages within a district is a renged ‘federation’, and the earliest period referred to in mythical narratives is the mechut el renged ‘ancient times’. These myths first describe the foundation of the Belauan cultural order in the ‘era of Chuab’ as a loosely knit string of eight villages located along the eastern side of the archipelago. Their political affiliation stems from the south to north path of the journey of the mythological figure Chuab, who institutes chiefly titles and councils at each village. After the destruction of this archaic order by a great flood ordered by the high god as punishment for lawless behaviour, a second cultural period called the ‘era of Milad’ emerges. The dominant villages of this second polity are aligned in a quadripartite pattern, which the myth explains as the result of the birth of four stone-children to the goddess Milad, the only survivor of the flood. When this order in turn - 169 becomes condensed in the historic period into two rival alliances focused on the militarily powerful villages of Oreor and Melekeok, political chants and ethnohistorical narratives speak of this arrangement as two opposed ‘side heavens’, referring roughly to the eastern and western sides of the archipelago.

Thus, this narrative sequence of cultural eras is coded in terms of the progression of spatial models, from ‘paths’, to ‘cornerposts’, and finally to ‘sides’, with each model originating and coming into prominence in a particular period. These three models are, then, both diagrammatic patterns (Parmentier 1985b) found in various sociopolitical institutions and also members of a transformational set which organises indigenous understanding of the islands' political development. 2 Now, in light of Sahlins' concept of structural history, the question which immediately arises is why should these three models be selected in mythical discourse, and why in this particular sequence? One could possibly view these models simply as reflections of independent social realities: that the first settled villages were, in fact, linked by genealogical and political paths tracing actual migration histories, and that the subsequent quadripartite and dualistic political orders developed as Belau grew into an increasingly hierarchical and stratified society. According to this naïve Durkheimian argument, Belauans were able to conceptualise things generally in quarters only after reflection on their quadripartite sociopolitical organisation. At the other extreme, one could claim that this progression of models is an entirely rhetorical, poetic or ideological device constructed after the fact and bearing little relationship to concrete factors such as interisland migrations, resettlement patterns, shifting political alliances, and contact with Western colonial forces.

The solution to be proposed here mediates these extreme responses by arguing that the narrative sequence of political orders demonstrates a coherent logic once the cultural significance of the underlying models is grasped, and that these meanings are in turn shaped by the contextual application of the three metaphors to particular historical circumstances. The first of these tasks is addressed in a companion paper which establishes specific meanings for ‘paths’, ‘sides’, and ‘cornerposts’ as organisation patterns for sociopolitical institutions (Parmentier 1985b); the present essay focuses on the second task of discovering the narrative motivation for the application of these models to particular kinds of mythical and historical contexts.

But before this project can begin, a brief introduction to Belau itself is necessary. Belau is an Austronesian culture occupying an archipelago of - 170 more than 200 islands in the extreme western corner of the Pacific Ocean (Figure 1). Although Belau is surrounded by the disparate Indo-Pacific cultures of the Philippines, northern Indonesia, western New Guinea and Micronesia, evidence from archaeology, linguistics and ethnography suggests that the culture developed in relative isolation for a millennium or perhaps two. 3 The original inhabitants were most likely part of the movement of Austronesian-speaking, seafaring horticulturalists which spread from mainland South-east Asia through Taiwan and the Philippines into Indonesia, western Micronesia and island Melanesia. 4 The islands themselves display a range of geological types, from high volcanic islands such as Babeldaob in the northern part of the group, high and platform limestone ‘rock islands’ in the central and southern sections, and small reef islands and atolls in the southern and northern extremities.

Although the exact chronology of settlement is still unclear, archaeological research and ethnohistorical traditions both indicate that the early inhabitants of the ecologically rich volcanic islands, creators of impressive brimmed terraces, vast irrigated taro swamps and anthropomorphic stone carvings, abandoned their villages around A.D. 1200 and occupied (or reoccupied) the ecologically marginal rock islands. By the time of the first substantial Western contact in the late 18th century, 5 however, these rock island sites had been abandoned, and the early English voyagers who visited the islands describe a rigidly ranked society of between 30,000 and 50,000 people living in village complexes located primarily on the larger islands of Babeldaob, Oreor, Beliliou and Ngeaur. 6 Political life of this period is characterised by the sharp cleavage between high-ranking and low-ranking houses, endemic inter-village warfare and headhunting, and the monopolisation of trade by titled chiefs who enforced their authority through the manipulation of stone and bead valuables, the consolidation of power through rank-endogamous marriages, and invocation of powerful religious sanctions.

This rapid sketch is intended only to be suggestive of several points. First, Belau's central position in the Austronesian world makes it an extremely important comparative example, especially in light of Geertz's (1980:239) tantalising hints about the mingling of South-east Asian and Pacific political forms in Bali. But since Belau was not subject to an overlaying of Indic or Islamic institutions and never developed many of the features Claessen (1984) identifies in “early states”, the clarity of its Austronesian polity makes it a crucial case for comparative ethnographic reconstruction. Second, prompted by its geographical isolation

- 171

FIGURE 1.

Illustration.

- 172

and millennia of uninterrupted occupation, Belau developed a sense of itself as the original, authentic, and sufficient exemplar of humankind and yet as the product of an internally continuous process of cultural development. Living amid ‘reminders of the past’ (ngesechel a cherechar) such as terraces and megaliths built by miraculously skilful beings, as well as abandoned beaches strewn with pottery fragments and shell deposits, Belauans in what we presumptuously call the “historic” period knew quite well that theirs was a history not devoid of significant cultural transformation but still not disconnected from the time of origin. 7

The first segment of the myth of the origin of Belau—and not surprisingly the very first story I taped in the field—describes the emergence of Latmikaik, a giant clam who gives birth to the first creatures of the sea. 8 At the origin of Belau, the voice of the high god Uchelianged rings out across the empty sea to command that a piece of land be brought out of the depths. The land which arises from beneath the sea near Ngeaur, the southernmost point of the group, pushes to the surface a giant clam named Latmikaik, which, when buffeted by waves and wind, gives birth to many fish, the ‘children of Latmikaik’. Uchelianged then commands these fish to pile up coral rubble from the sea floor to serve as a towering bridge to the heavens. But in mid-construction this stonework begins to tilt and falls over to the north, forming a bridge reaching all the way to the village of Oikull (‘Measured’) on Babeldaob island. These original fish creatures migrate along this series of stepping stones, settle the high northern islands, and give birth to the first generation of inhabitants of Belau.

In this myth, the vertical cosmological axis constituted by the previously unmediated differentiation between ‘beneath the sea’ and the ‘heavens’ rotates to a horizontal axis along the surface of the sea formed by a linked or bridged differentiation between southern ‘lower sea’ and northern ‘upper sea’. In terms of the movement of the children of Latmikaik, the aspiration to climb to the heavens becomes the quest for a land-based cultural existence as distinct from an amorphous life beneath the sea. But both these transformations are grounded in a more fundamental theme of this myth. The account of the origin of Belau (uchul a Belau), in both geographical and cultural senses, is phrased in terms of the construction of a path (rael) along which motion takes place from a - 173 beginning point (uchul) near Ngeaur to an ending point (rsel) at Oikull. The key word which appears in the phrases ‘origination of Belau’ and ‘beginning of a path’ is uchul, which means not only ‘origin’, ‘beginning point’ but also more generally ‘source’, ‘point of growth’, ‘basis’, ‘cause’, and ‘reason’. The ground for these extended meanings appears to be the meaning ‘tree trunk’, so uchul is simultaneously the physical support for upper limbs and the point at which growth originates. 9 (Since bridges are frequently made from tree trunks, the bottom of the trunk becomes the beginning point of the bridged path, while the top of the trunk becomes the ending point.) These intertwined meanings provide the semantic motivation for Uchelianged's action of creating the Belauan race by providing simultaneously a path leading from the starting point at Ngeaur to the measured terminus at Oikull (‘Measured’). In other words, the creation of cultural order is expressed as the possibility of directional movement along a recognised path.

Migration is, thus, the paradigmatic cultural act, since it is the simplest form of presupposition: a journey from beginning point to ending point establishes a spatio-temporal linkage that can then be taken as the template or semiotic type for future action. In everyday speech the term ‘path’ (rael) can also mean a recognised social relationship (e.g., ‘a kinship tie’ [rolel a klechad]), an accepted political strategy (e.g., ‘the way of the firebrand’ [rolel a ngau]) or a technical procedure (e.g., ‘the technique of house construction’ [rolel a omelasech]). Numerous stories set in the archaic world share this same feature of describing the origination of social institutions, groups, customs and names in southern Ngeaur and tracing their subsequent passage north to Babeldaob. The brothers of Kerengokl discover fire and then distribute this miraculous phenomenon to northern villages; the art of carpentry originates beneath the sea between Beliliou and Ngeaur; and several northern villages trace their origin to sites in the ‘lower sea’ with identical names.

The next segment of this origin myth describes the travels of two children of Latmikaik, Rak and Chuab, who are responsible for instituting respectively the spawning cycles of fish and the system of political councils—the law of the sea and the law of the land. 10 First at Rois village on Ngeaur and then in eight villages in a line arranged from south to north along the eastern coast of Babeldaob, Chuab ‘plants’ chiefly councils (klobak), bodies of titleholders which govern the internal and external affairs of villages (see Table 1). Chuab travels to Ngerechol where she instituted a council under the leadership of Uchelchol; then in sequence Chuab appoints chiefs such as Secharaimul

- 174

at Imul, Tucheremel at Ngerusar, Rechiungl at Ngeremid, Obakeramechuu at Ngersuul, Ngirnguloalech at Ngeruikl, Ngirairung at Ulimang and Bdelulabeluu at Mengellang. This last chief, holding a title meaning ‘head of the villages’, is given the task of ‘holding steady’ (mengellang) all the other villages founded by Chuab.

Now what is important to note about this second segment of the myth is that the eight villages in the polity of Chuab are linked together by the fact that each has a council and a chief appointed by Chuab. The linear path traced by Chuab is thus marked by these identical deposits, which then function as external signs (olangch) of the creative journey. Other stories set in this archaic era mention journeys which deposit, for example, identical trees, parallel names and titles, similarly shaped stones, and tokens of the same class of money.

Stories set in the ‘era of Chuab’ are characterised by migrations along paths and by sets of identical elements marking the shared identity of the linked points. This model of the ‘path’ does not simply reflect the fact that, during this prehistoric period, migrations within the archipelago were frequently the way people and villages became affiliated. Rather, I think the ‘path’ is a narrative device to express the twin notions of origination, conceived of as putting things in directional motion, and typification (that is, making general regularities on the basis of occurring instances), seen as the possibility of repeating and thereby presupposing some previous activity. Nothing could be a more appropriate theme for an origin myth than an account of origination in general, nothing more fitting to present the first political order than a story about the creation of typifying paths. This interpretation mitigates concern that archaeological evidence does not provide substantial corroboration that these - 175 particular eight villages were dominant in the early settlement period. In fact, ancient population centres seem to have been located on the other side of the island, especially surrounding Ngeremeduu Bay between Ngeremlengui and Ngetbang districts and on the coast of Ngerechelong peninsula. In other words, the cultural significance of ‘paths’ is more important for understanding these narratives set in the ‘ancient world of Chuab’ than the particular political configuration of villages which the stories detail.

Krämer (1917-29:IV, 114-6) reports one of the few traditions which depicts the villages of Chuab acting as a solidary unit. In this story, a man named Madraklai from Ngerdmau district accidentally kills a woman from Ngerebkei, and the elders from this small village call upon the leaders of the villages of Chuab to assist them in demanding reparations from Ngerdmau. But when the requested payment is received, the chiefs award it to chief Rechiungl, who refuses it on the grounds that his village of Ngeremid is too close to Oreor village. Similarly, chief Tucheremel fears the proximity of Irrai, and chief Secharimul fears the proximity of Imeliik. Finally, Obakeramechuu pockets the money, since his village of Ngersuul is not under pressure from a potentially oppressive neighbour. Now the reason this story is fascinating is that it represents, by means of an inverted rhetorical anachronism, the villages of Chuab as being systematically threatened by a second set of villages, precisely those villages which emerge as part of the polity of the subsequent ethnohistorical period, the ‘era of Milad’.

The polity of Chuab comes to a sudden close when, at a feast held in honour of the eight titleholders appointed by Chuab, the food-server to the chief of Mengellang village violates his charge by stealing a ceremonial food portion. 11 The titleholders disperse in anger and the co-operative affiliation among Chuab's villages is terminated. This act of disobedience, appropriately assigned to the food-server at Mengellang, the northernmost village of the set whose mandate was to ‘hold steady’ the polity, is symptomatic of the lawlessness that troubles the high god Uchelianged throughout this era. In order to recreate a race of people who would obey the law, Uchelianged decides to destroy all the villages of the archipelago in a flood. 12 Before this took place, however, several of Uchelianged's messenger gods are befriended by a woman named Dirrauchulabkau, who prepares their lunches with a special dish of raw - 176 fish concealed in cooked taro. So Uchelianged decides to give her advance warning of the imminent deluge. She builds a large bamboo raft which enables her to float to the top of Ngeroach mountain, a rocky peak in the Roismlengui range on the west coast of Babeldaob. When the flood waters recede Dirrauchulabkau is resuscitated by the breath of the gods and then takes a new name, Milad (‘Was Dead’). Descending to the base of the mountain, Milad sets up her residence in a cave, where she gives birth to four children in the form of stones. Imiungs, the first son, is kept at his mother's side near Ngeroach mountain in Ngeremlengui district; Melekeok, the second son, is a stubborn and proud child and so Milad sends him away to the other side of the island; Ngerekeai (or Imeliik, in some accounts), the third child and only daughter, is placed to the south; and Oreor, the youngest son, is so restless that Milad sends him far way to a separate island across the water from his protective sister Ngerekeai. The four stones become four dominant villages, and as the ‘four cornerposts’ of Belau their political relationship depends on their siblingship, and their permanent rank derives from their divine birth. 13

Western sources from the 19th century indicate that the story of Milad was among the most widely known and frequently told. Captain Barnard, whose whaler Mentor was wrecked at Belau in 1832, learned from the people of Ngerechelong that “like most of the human race they have a tradition of a universal deluge”.

Thirty years later the trader Andrew Cheyne was told the following fragment by chief Ngirturong of Imeiong village.

- 177

Kubary, the Polish naturalist and ethnographer, who heard the story less than 10 years later, grasped more fully its political implications.

(In another passage in this same work Kubary states that Imeliik is the third child and Oreor the youngest.)

While there is little evidence to corroborate the assumption of traditional narratives that the eight villages of Chuab formed an actual archaic political regime, in the case of the ‘era of Milad’ these same traditional and historical sources abound in references to Milad, the four sacred stones, and the co-ordinated political relations among the four corresponding principal villages of Imiungs (or Imeiong), Melekeok, Ngerekeai (or Imeliik) and Oreor (see Table 2). And whereas the model of the ‘path’ and the various external signs linking Chuab's villages are only implicit in the origin myth, the four-part model and lithic markers implicated in the story of Milad are entirely explicit, even foregrounded, both in the extant texts and in the historical realities of political relations. While the federation of Chuab was regarded in the colonial and modern periods as a vague, archaic background standing in opposition to the ‘new thing’ (beches el tekoi) or ‘new world’ (beches el beluulechad) of Milad, the quadripartite order of four villages labelled ‘children of Milad’ played a powerful role throughout these historical periods by coding political action, social rank, and historical processes in terms of

- 178

the co-ordinated yet differential interaction among the ‘four cornerposts’.

Folk etymologies explaining the meaning of the names of the four villages of Milad in terms of human behavioural characteristics are fairly uniform across sources. The oldest child, Imiungs, remains close to his mother at Ngeroach mountain, and its haughty superiority is expressed in phrases such as Imiungs el mad and mad el Imiungs ‘the Imiungs face’ which describe the severe and condescending countenance (imings) said to characterise people from Imiungs village. A man from this village told me that the name Imiungs implies that ‘we are very slow to speak up, and in meetings with other people we do not laugh or joke around’. As the oldest child of Milad, Imiungs village is proud to the point of being contemptuous of lower-ranking villages, whose geographical removal from their ‘mother’ Milad symbolises their subordinate positions.

The name of the second child, Melekeok, is derived from the word tekeok, which means ‘openly boastful’, ‘stubborn’, or ‘self-congratulatory’. In contrast with Imiungs' high rank derived from the unchangeable fact of birth precedence, Melekeok's second position among Milad's children correlates with its aggressive self-praise. The name Ngerekeai is derived from the word keai, a thick betel-nut fibre used for rainhats, basket coverings and, more important in this context, swaddling for babies. Oreor, the youngest child, was so hyperactive that Milad placed him farthest away on a separate island. The name Oreor shows the stem -reor which is found in oureor ‘to work’, ureor ‘labour, work’, and sureor ‘energetic, active’. Contemporary informants point out that Oreor has, in fact, been energetic as the centre of Western commercial activity and the site of the national government and that it was recently rash in its selection of a youthful man to hold its chiefly title Ibedul.

The qualitative gradation evident among the children of Milad corresponds to the antithetical forces of celeritas ‘swiftness, rashness’ and gravitas ‘heaviness, seriousness’ which Sahlins (1981b:121), following Dumézil, correlates with the difference between the active, conquering, warlike aspect and the peaceful, ceremonial, priestly aspect of Fijian chieftainship. Similarly, the boyish rashness of Oreor, Milad's youngest son, contrasts sharply with the dignity and passivity of Imiungs, the oldest child. The two middle terms of the series, boastful, stubborn Melekeok and the fruitful, nurturant female Imeliik or Ngerekeai, are viewed in terms of the inherent instability of the two extreme terms of the set. Structurally, the presence of Melekeok gives the energetic Oreor an - 179 additional term against which to exercise its active power, which derives from military strength, externally acquired influence, and achieved political alliance. Such activity directed towards its older brother Imiungs would be futile or counterproductive, since the ‘sacredness’ of all the children of Milad is said to remain with Imiungs near the ‘mother’ of the entire polity. And the existence of the female Ngerekeai gives the system the dynamic potential to generate a fifth term, Ngebuked, the capital village of Ngerard district, which, as the offspring of Ngerekeai, stands in the privileged relation of ‘sister's son’ to the three brothers.

These folk etymologies are at the same time explanations for the names of sacred stones and other objects which stood in the named villages as external signs of the political order created by Milad. The spot in the forest at the foot of Ngeroach mountain where Milad gave birth is marked by a huge volcanic plug approximately 15 metres high, called by contemporary villagers the ‘house of Milad’ or the ‘cave of Milad’. At this site can be seen the stone table Milad used, and the black soot on the ceiling of the cave is pointed to as evidence that she cooked taro there—taro she cultivated at nearby Ngeruuchel swamp. (Her stone cooking pots were later removed from the forest to the central square of Imeiong.) But the most important material sign associated with Milad found in the general Ngeroach area is Imiungselbad (‘Imiungs Stone’), the representation of Imiungs as the oldest child of the goddess. This sign is actually two stones, a circular mortar-like stone with a hollowed-out centre and a smaller spherical companion stone called Imiungseldui (‘Imiungs Title’) resting on the rim. Informants note that the function of the smaller stone is to ‘respect’ the larger stone, implying that the sacredness of Imiungselbad, and thus the village of Imiungs as a whole, depends in part on there being something or someone to hold it as sacred. Although there is no direct native exegesis to support the claim, it is certainly possible to speculate in light of Austronesian parallels about the connection between this circular mortar stone and the “feminine” quality of Imiungs as the oldest child of Milad (see, e.g., Fox 1924:223; Nooy-Palm 1979:69; Riesenfeld 1950:246). To be sure, Imiungs is the oldest son, but the qualities of stability, centrality, passivity, fertility and sacredness which characterise this high-ranking village are also female symbolic qualities found additionally to belong to the first chief of a village, who can be, in certain contexts, the ‘mother of the village’.

Lithic embodiments of the other children of Milad stood in the remaining villages as well. Olekeokelbad, the stone representing Melekeok village, had the shape of an oblong face and stood in a grove of croton - 180 trees at the centre of the village. The stone representing Ngerekeai is said to have had the form of a pot—pottery being an activity primarily associated with women—although no specific eyewitness accounts are extant. And finally, Surreorelbad, the stone representing the village of Oreor, stood next to a long flat stone called the ‘protecting mat of Milad’ at Irachel house; as the youngest of Milad's children, Oreor is in need of protection from the sun's glare.

In turning our attention from these folk etymologies and sacred stones associated with the story of Milad to the political implications of the ‘new world’ which emerged after the great flood, the most important point to begin with is that the villages of Milad treated one another in certain contexts as siblings. That is, not only does the story give grounds for an ideology of overall precedence for these four capital villages as the ‘cornerposts’ of Belau, a metaphorical analogy to the four cornerposts holding up the roof of a meeting-house, but it also dictates a specific pattern of interaction modelled after the relations between brothers and sisters. That the villages of Milad are called the ‘cornerposts’ of Belau and also conceptualised as siblings is entirely consistent, since the ‘cornerpost’ model is itself based on the idea of differentiated yet co-ordinated support, the same norms which govern cross-sex siblingship.

As the sister, Ngerekeai is supposed to be devoted to the financial support of her brothers, most especially of her youngest brother Oreor. Should a large or famous valuable enter Ngerekeai through the customary channels of marriage payment or fines, the three brother villages could scheme to acquire it, either peacefully through the inherent right of the brother to take control of his sister's earnings or forcibly through threat of attack. Traditional and historical evidence suggests, however, that Ngerekeai was not consistently supportive of her three brothers, and that she even took advantage of her symbolic status as a woman to wage war without risk of counter-attack. After its own failed attack on Oreor before 1783 (the so-called “War of Ngirakederang”), Ngerekeai began to co-operate with Oreor's efforts against the southern island of Beliliou. This support was, it seems, both insufficient and unreliable. When the Englishman McCluer presented a few beads to a delegation of Ngerekeai leaders in 1791, Oreor's chief Ibedul complained that his “sister” village did not deserve such valuable presents, “for when we [the English] were not here they [people of Ngerekeai] never come near him [Ibedul] but assisted his enemies with men and canoes” (McCluer 1790-92: February 5, 1791).

Not only did Ngerekeai fail to “come near” Ibedul, but other chiefs - 181 tended to avoid coming too near Ngerekeai. In a gesture that perplexed his English allies, Ibedul refused to come ashore when his party landed there in 1783. An informant explicated this behaviour in terms of the prescribed physical avoidance between brother and sister. This avoidance also implies that Ngerekeai could function as a place of refuge and protection. A story about a man named Beludes concludes with his flight from pursuing warriors; stopping successively at Ngerard, Ngerdmau and Ngeremlengui, Beludes is refused protection by the chiefs of these villages. Finally at Ngerekeai, chief Rengulbai welcomes him with the words, “Imeliik is the deep channel of Belau (mechesengelel a Belau).” This expression means that, as a woman, Imeliik (that is, Ngerekeai) is like a deep sea channel that one enters in pursuit only at risk of great danger.

Not only can brothers demand money from sisters, but older brothers can take charge of the financial dealings of their younger brothers. Imeiong played this role of ‘older brother’ (obekul), for example, when Oreor and Melekeok became embroiled in political intrigue in the late-19th century. Oreor jealously guarded its favoured position with respect to Western commercial and military influence and resisted attempts by several merchants to set up trading operations in Melekeok and Ngerard for fear that these villages would receive arms in return for fish, trepang and agricultural produce. When Kubary finally managed to visit Melekeok in 1871 he was given a valuable by its chief Reklai. Learning Reklai's village had given such favoured treatment to its foreign visitor, the leaders of Imeiong demanded that the titleholders from Oreor purchase back Kubary's money and have it returned promptly to Melekeok. And if Oreor did not have the necessary money on hand, chief Ngirturong of Imeiong offered to contribute an chelebucheb-piece to aid his ‘younger brother’, as Kubary notes:

In explicating the interrelationship between the ‘cornerpost’ model and the ideology of siblingship of the villages of Milad, my informants said that Ngebuked, the present capital of Ngerard district in northern - 182 Babeldaob, was traditionally regarded as the child of Ngerekeai, Milad's only daughter. This fifth term of the Milad polity is, however, not equivalent to the many Indonesian and South-east Asian examples of a fifth, central term surrounded by four quarters (see Tambiah 1976:102-11). Rather, in the Belauan case, the position of Ngebuked is discussed in terms of the ‘offshoot of the turmeric plant’ (chebedel a kesol), a calculus which specifies the linkage between a man and his sister's children (Force and Force 1961). This metaphor derives from the observation that the turmeric plant develops projections or buds on its sides, some of which in turn produce additional offshoots. This system of multiple projections from the same ‘mother’ root is a metaphor for the way in which female children of a house generate offspring who are in a strong matrilineal line (ochell), while offspring of men become increasingly removed from this line. As the sister's son, Ngebuked is granted special protection and support from its mother's brothers, and is lovingly spoiled by its mother Ngerekeai. Historical accounts bear this out, for Ngerard, the district containing Ngebuked village, constantly played off Ngeremlengui, the district containing the village of Imeiong, against Melekeok and refused to join Oreor's fight against the Ngetelngal federation headed by Melekeok.

In 1783, for example, chief Ibedul tried to convince Madrangebuked, the chief of Ngebuked, to join with him alongside the miraculous foreign power of Captain Wilson's English troops (Keate 1788:174). But even the gift of an English scarlet coat was not enough to impress Madrangebuked, who was clearly attempting to coax Wilson to travel north to Ngebuked for his own purposes. This impasse continued until the 1860s when Oreor and Ngebuked finally came to blows. In the role of the senior brother, Ngirturong, the chief of Imeiong, tried to arbitrate the dispute, as Cheyne reports in his journal.

A decade later Ngebuked once again established, peaceful relations with Oreor, but continued to court the alliance with powerful Ngetelngal district on the east coast.

In addition to these fragmentary references documenting the strategic - 183 application of siblingship and quadripartition to intervillage relations, there is more substantial, though less direct, evidence that the story of Milad is an expression not just of differential co-ordination but also of completed political transformation; that is, an institution characterised by a four-part arrangement of elements or terms is presumed to signal thereby an achieved state of coherent unity or staged maturity. This evidence comes from an important set of ethnohistorical narratives which describe the nearly simultaneous overthrow of neighbouring villages by Milad's three sons. Imeiong's victory against Uluang, Melekeok's defeat of Oliuch, and Oreor's triumph over Ngerekebesang each involves a co-ordinated campaign by allied villages and each results in the consolidation of political leadership in the respective districts. For example, after generations of being oppressed by the insulting behaviour of the people of Uluang, the leaders of Imeiong finally decide to take action when the chief of Uluang haughtily violates the wife of Ngiraklang, then chief of Imeiong, as she passed by Omekesebech meeting-house on her way to work in a distant taro patch. Ngiraklang arranges to have expert military leadership by requesting the aid of the famous strategist Ngirairung, who in turn recruits warriors from allied villages such as Ngerechelong, Ngeremetengel and Ngellau villages. Their attack from both land and sea succeeds in burning Uluang to the ground. Having achieved supremacy in Ngeremlengui district, Imeiong repays its allies by awarding the Imiungselbad (‘Imiungs Stone’), the symbol of its own sacred rank, to Ngirairung, who in turn presents it to the people of Ngellau, where it stands today.

This puzzling combination of an achieved victory and the alienation of the symbol of sacred rank suggests that there are two conflicting mytho-historical interpretations operative here. On the one hand, the story of Milad stipulates that the precedence of the four ‘cornerpost’ villages derives automatically from their descent from the original stone-children of Milad. These villages do not develop as a continuous process from an earlier state; rather, they preside over a fresh political slate, wiped clean by the flood. On the other hand, the parallel battle stories describing the overthrow of Uluang, Oliuch and Ngerekebesang imply that the supremacy of Imeiong, Melekeok and Oreor is an historically contingent success, part of a long process of district consolidation and shifting political alliance. These two alternatives are, I think, two ways those currently enjoying pre-eminent positions can talk about political rank, namely, as the result of inherent supremacy or as the result of acquired power. The ‘era of Chuab’ differs from the ‘era of Milad’ in exactly this respect: while a logic of migratory paths provides the domi- - 184 nant metaphor for the archaic world of Chuab, the new world of Milad substitutes for the diachronic, progressive connotation of linearity the idea of structural completeness without processual depth. Analogously, the ideology of quadripartition expressed in the story of Milad's constitution of the political order works to mask the historical implications of the warfare narrative involving these same villages.

In the 19th century, warfare intensified in a series of conflicts between villages representing two hostile political confederations; roughly speaking, villages on the east coast (desbedall) of Babeldaob were allied to Melekeok's chief Reklai and villages on the west coast (kiukl) of Babeldaob as well as villages in the southern islands of Ngeaur and Beliliou were allied to Oreor's chief Ibedul. 14 While it is undoubtedly true that the steady influx of imported firearms and the increased competition over access to foreign trade significantly raised the stakes in the intervillage struggles of this period, the conceptualisation of the lines of opposition drew on the traditional geographical division of Babeldaob into two ‘sides of heaven’. The ‘heaven’ referred to in this expression is not the celestial realm but rather Babeldaob island itself, which, as was noted above in the discussion of the story of Latmikaik, was the destination of the original creatures who migrated from beneath the sea near Ngeaur. The island is symbolically split into two ‘sides of heaven’ by a north-to-south line called Raelkedam (‘Path of the Kite’) following a mountainous ridge running from the northern shore of Ngerdmau through the centre of Irrai in the south-east. The name of this dividing line comes from the story of the founding of Ngerdmau in which people from Ngerekedam village in Irrai pursue their lost kite (kedam) along this central ridge until they arrive at Ngerdmau. Some informants mention a more ancient term, Raelbalech (‘Path of Shooting Arrows’), the route along which the mythological character Belebalech habitually walked while pigeon hunting.

This application of the model of ‘sides of heaven’ to village confederation parallels several other examples of dualism in sociopolitical organisation (see especially McKnight 1960). Two ‘sides of the mangrove channel’ divide a village's men's clubs into competitive halves; two ‘sides of the meeting-house’ split the 10 titles of a chiefly council into balanced sets of political allies; and two ‘sides of the leg’ segment the various lineages of people related to several high-ranking houses. In these usages each ‘side’ contains one set of elements (clubs, - 185 titles, descent lines) which is exactly matched by an opposite set; and the two sets are not distinguished by any positive value or graded differential, such as right and left, high and low, large and small, or male and female. Not only is each set the perfect mirror image of the other, but two ‘sides’ together represent the totality of possible elements, since one ‘side’ logically implies the existence of the other. So once Belauan polity is described as two ‘sides of heaven’, no village can exclude itself from the extension of this categorisation. Furthermore, elements divided into ‘sides’ are assumed to be in a state of perpetual conflict, competition or rivalry (see Vidich 1949:50).

Two ‘sides’ do not, however, constitute a perverted Hobbesean war of each against the other, for struggle is channelled or focalised by the highest ranking element on the two sides. Low-ranking titleholders, for example, voice political dissent only through the mediation of the first two titleholders; similarly, hostility between smaller villages allied to either Oreor or Melekeok is politically meaningful only when expressed in terms of the focal conflict between the two confederations. A final feature of ‘side’ oppositions is their resilience in the face of internal and external pressure. In contrast with the relative fragility of ‘cornerpost’ structures, where the metaphorical roof comes crashing down when one of the supports is withdrawn, ‘sides’ are able to perpetuate themselves historically, since any collection of entitles can be divided into paired segments, especially if no positive value is involved, and since the opposition itself can become recontextualised without altering the fundamental pattern. In modern politics, for example, the opposition of ‘sides of heaven’ has been replaced by factional splits between Progressive and Liberal parties, between elected members of the House of Delegates and traditionally sanctioned members of the House of Chiefs, and more recently between pro-Constitution and pro-Free Association groups. In summary, then, the model of ‘sides’ expresses notions of totality as the sum of mutually implied sets and conflict as the easily recontextualised struggle between focal points of articulation.

The rivalry between Melekeok's Ngetelngal confederation and Oreor's Ngerekldeu confederation was immediately apparent to Captain Henry Wilson in 1783, when chief Ibedul of Oreor demanded and obtained the support of English firepower and fireworks in several assaults against Melekeok. After one successful campaign Ibedul's forces carried off the council's backrest stone (btangch) from the centre of the village, an insult that signalled Oreor's intention to treat Melekeok as an ‘enslaved village’ (ker el beluu). But by 1791 this intention quickly resolved into a more permanent reciprocal hostility (kaucheraro) in which - 186 disputes are temporarily resolved by the exchange of valuables. 15 The norms of institutionalised enmity between these focal villages of the opposed confederations were not followed, however, in Oreor's relations with other villages, where Ibedul's propensity for aggressive violence manifested itself in the capture of slave women, the burning of houses, the destruction of gardens, and the exacting of heavy tribute. Traditions mention successful campaigns by Oreor against Ngerekebesang, Ngeredelolk in Beliliou, Ngertuloech in Ngetbang, Chelab in Ngerard, Oikull in Irrai and Urdmau in Ngerdmau. Western historical records corroborate attacks on Ngerechelong in 1832, Melekeok in 1850, Ngebuked in 1860, Ngcheangel in 1872, and Ngersuul in 1875. 16 Through these wars of subjugation Oreor assured the subordination of villages in the western and southern ‘side of heaven’ and further exacerbated its traditional hostility with Melekeok. 17 In response, Melekeok succeeded in solidifying its rule over villages on the eastern coast of Babeldaob, including Ngiual and Ngersuul, and mounted an unsuccessful effort to attract Western commercial operations (curing sea cucumber, in particular) to its member villages. (Melekeok's early acceptance of Catholic missionaries may have been motivated by this same rivalry.)

Especially in the late 19th century, realities of military and economic power made the application of the ‘cornerpost’ model to Belauan polity the distinctive, if self-consciously archaic, claim of Imeiong. Although generally supportive of Oreor's confederation, Imeiong continued to express its autonomy by engaging in reciprocal headhunting raids on Melekeok, thereby bypassing Ibedul's own strategies. But by the time Western interests forced a peace treaty between Melekeok and Oreor in 1883 and put an end to intervillage raiding by the end of the century, the ‘sides of heaven’ model had achieved ideological dominance. During the successive Spanish, German, Japanese and American colonial administrations, Ibedul and Reklai became recognised as twin “paramount chiefs” of the islands. At the intervillage level, pooled labour, sporting competition and fishing co-operatives all organised activities around the dualistic model. In 1974 this political model was codified in a Belauan document published by the Palau Community Action Agency, which depicts the seating arrangement for the Rubekul a Belau, that is, the leaders of all the districts in the two ‘sides of heaven’ (see Figure 2). According to this scheme, on the ‘side of heaven’ under Reklai of Melekeok the second place of honour is given to Madrangebuked of Ngebuked; and on the ‘side of heaven’ under Ibedul of Oreor the second place of honour goes to Ngirturong, the chief of Imeiong. Rengulbai of Imeliik,

- 187

FIGURE 2.

Seating Pattern of Rubekul a Belau (1974).

- 188

representing what in the era of Milad had been one of the four ‘corner-posts’ of Belau, is relegated to a minor position next to Ngirturong, the final ‘cornerpost’ position going to Madrangebuked of Ngebuked.

Does this account of the progressive unfolding of models of political development constitute no more than comparative support for Sahlins' thesis? In fact, I think that the Belauan case shows clearly that “structural history” must consider not only the revaluation of cultural categories but also the recategorisation of cultural values. Focusing on the former task, Sahlins' work on the conjunction of myth and history in Hawaiian, Fijian and Maori cultures notes, for example, the realignment of binary oppositions (e.g., male/female, internal/external) and privative oppositions (that is, marked/unmarked categories) and the recontextualisation of various paradigmatic symbolic structures (e.g., type-token and universal-particular relations). The analysis offered here, pursuing the latter task, concentrates primarily on the development of different kinds of organisational categories, ‘paths’, ‘cornerposts’, and ‘sides’, each with distinctive cultural implications apart from the particular values or terms being organised. This development should not, however, be viewed as a simple matter of concrete or rhetorical replacement, with ‘paths’ giving way to ‘cornerposts’, which in turn disappear with the advent of ‘sides’. Rather, as abundant ethnographic data reveal, each of these models is simultaneously present in a range of overlapping sociopolitical institutions. The point, then, is that ethnohistorical appropriation of the models in sequences is based on the projection of the corresponding cultural meanings in reflection on the past.

And finally, and here I enter the realm of informed speculation, this appropriation does not exhaust the significance of the mythological and historical narratives discussed above. On the surface it would seem that, in the use of the ‘cornerpost’ logic to tell the story of Milad's resurrection and repopulation of the islands, quadripartition is a metaphor for an historical process. But, in a sense at least, the narrative structures of these stories can also be seen as metaphorical means for talking about more fundamental matters. Stories about migration, for instance, are ways of discussing the notion of typification, since movement along a path is the paradigmatic act which separates cultural from precultural existence. The story of the flood and the birth of four stone-children is a highly condensed way to express basic ideas about internally generated structural maturity and presupposed rank. And the labelling of 19th cen- - 189 tury political confederations as ‘sides of heaven’ suggests that this holistic or totalising form of opposition encompassing the entire archipelago is one response to the impact of externally derived political regimes whose massive presence rendered impossible fragmented or localised political action.

Field work in Ngeremlengui district of Belau was carried out from 1978 to 1980 with the aid of a research grant from the Center for Psychosocial Studies in Chicago. Data analysis and writing were facilitated by the William Rainey Harper Memorial Fellowship at the University of Chicago. The Committee on Aid to Faculty Scholarship at Smith College provided funds for the acquisition of research materials and for technical support. I thank with pleasure the people of Ngeremlengui for allowing me to live and study with them, and I dedicate this essay to Ngiraklang Malsol, who supervised my investigations of Belauan folklore and history. Earlier versions of this paper were presented at the University of Massachusetts (March 11, 1985) and at Harvard University (March 18); I thank the departments of anthropology at these two institutions for giving me the opportunity to discuss my paper with colleagues. The present version has benefited from substantive and editorial comments of Nina Kammerer.

- 190

• Aoyagi, Michiko, 1979. Bitang Ma Bitang (Two Halves), Eual Saus (Four Corners), and Mechanical Confusion in Palauan Socio-Political Organization, in H. Kusakabe (ed.), Cultural Anthropological Research on the Folk Culture in the Western Caroline Islands of Micronesia. Tokyo, Tokyo University of Foreign Studies.

• Bamler, A., 1900. Vokabular der Tamisprache. Zeitschrift für ozeanische und ostasiatische Sprachen, 5:217-53.
• Barnard, Edward C., 1980. “Naked and a Prisoner”: Captain Edward C. Barnard's Narrative of Shipwreck in Palau, 1832-1833. Sharon, Massachusetts, The Kendall Whaling Museum.
• Barnett, Homer, 1949. Palauan Society: A Study of Contemporary Native Life in the Palau Islands. Eugene, University of Oregon Press.
• Bellwood, Peter, 1979. Man's Conquest of the Pacific: The Prehistory of South-east Asia and Oceania. New York, Oxford University Press.
• — 1980. Plants, Climate and People: The Early Horticultural Prehistory of Austronesia, in J. J. Fox (ed.), Indonesia: The Making of a Culture. Canberra, Research School of Pacific Studies, Australian National University.
• Bender, Bryon W., 1971. Micronesian Languages, in T. Sebeok (ed.), Current Trends in Linguistics, Vol. 8. The Hague, Mouton.
• Blust, Robert, 1979. Proto-Western Malayo-Polynesian Vocatives. Bijdragen tot de Taal-, Land- en Volkenkunde, 135:205-47.

- 191

• Capell, Arthur, 1941. A New Fijian Dictionary. Sydney, Australasian Medical Publishing Company.
• Cheyne, Andrew, 1863-1866. Journal Aboard the Brigantine “Acis”, November 1863-February 1866. Photocopy of the original manuscript in the files of Palau Community Action Agency, Koror, Palau.
• Claessen, Henry J. M., 1984. The Internal Dynamics of the Early States. Current Anthropology, 25:365-70.
• Craib, John L., 1983. Micronesian Prehistory: An Archaeological Overview. Science 219:922-7.

• Delano, A. 1817. Narrative of Voyages and Travels . . . in the Pacific Ocean. Boston, E. G. House.
• Dempwolff, Otto, 1834-38. Vergleichende Lautlehre des austronesischen Wortschatzes. Vol.3. Berlin, D. Reimer.

• Force, Roland, 1960. Leadership and Cultural Change in Palau. Fieldiana Anthropology 50.
• — and Maryanne Force, 1961. Keys to Cultural Understanding. Science 133:1202-6.
• Fox, C. E., 1924. The Threshold of the Pacific. London, Routledge.
• Fox, James J., 1971. Sister's Child as Plant: Metaphors in an Idiom of Consanguinity, in R. Needham (ed.), Rethinking Kinship and Marriage. London, Tavistock Publications.

• Geertz, Clifford, 1980. Negara: The Theatre State in Nineteenth-Century Bali. Princeton, Princeton University Press.

• Hidikata, Hisakatso, 1973. Stone Images of Palau. Micronesian Area Research Center Publication 1. Agana, Guam.
• Hockin, John P., 1803. A Supplement to the Account of the Pelew Islands. London, G. and W. Nichol.
• Hogbin, Ian, 1963. Kinship and Marriage in a New Guinea Village. London, Athlone Press.

• Josephs, Lewis, 1975. Palauan Reference Grammar. Honolulu, University Press of Hawaii.

• Keate, George, 1788. An Account of the Pelew Islands . . . from Journals of Capt. Henry Wilson, 1783. London, G. Nichol.
• Krämer, Augustin, 1917-29. Palau, in G. Thilenius (ed.), Ergebnisse der Südsee-Expedition, 1880-1910. Hamburg, Friederichsen.
• Kubary, J. S., 1873. Die Palau-Inseln in der Südsee. Journal des Museum Godeffroy, 1:177-238 (English translation in HRAF).
• — 1885. Die Sozialen Einrichtungen der Pelauer, in Ethnographische Beiträge zur Kenntnis der Karolinischen Inselgruppe und Nachbarschaft. Berlin, Asher. (English translation in HRAF).

• Lessa, William A., 1975. Drake's Island of Thieves: Ethnographical Sleuthing. Honolulu, University Press of Hawaii.

• McCluer, John, 1790-92. Voyage to the Pelew Islands in the H. C. Snow Panther, 1790-92. Typewritten copy of original British Museum manuscript in files of Palau Museum, Koror, Palau.

- 192

• McKnight, Robert K., 1960. Competition in Palau. Ph.D. dissertation, Ohio State University.
• Masse, W. Bruce, David Snyder, and George J. Gumerman, 1984. Prehistoric and Historical Settlement in the Palau Islands, Micronesia. New Zealand Journal of Archaeology, 6:107-127.
• Meares, John, 1790. Voyages Made in the Years 1788 and 1789 from China to the North-West Coast of America. London, Logographic Press.
• Murdock, George P., 1968. Genetic Classification of the Austronesian Languages: A Key to Oceanic Culture History, in A. P. Vayda (ed.), Peoples and Cultures of the Pacific. Garden City, N.Y., Natural History Press.

• Nooy-Palm, Netty, 1979. The Sa'dan-Toraja: A Study of Their Social Life and Religion. The Hague, Martinus Nijhoff.

• Osborne, Douglas, 1958. The Palau Islands: Stepping Stones into the Pacific. Archaeology, 2:162-71.
• — 1966. The Archaeology of the Palau Islands: An Intensive Survey. Bernice P. Bishop Museum Bulletin 230.
• — 1979. Archaeological Test Excavations, Palau Islands, 1968-1969. Supplement No.1 to Micronesica.

• Palau Community Action Agency, 1974. Dui Ma Klobak er Belau. Koror, Palau.
• Parmentier, Richard J., 1981. The Sacred Remains: An Historical Ethnography of Ngeremlengui, Palau. Ph.D. dissertation, University of Chicago.
• — 1984. House Affiliation Systems in Belau. American Ethnologist, 11:656-76.
• — 1985a. Times of the Signs: Modalities of History and Levels of Social Structure in Belau, in E. Mertz and R. Parmentier (eds), Semiotic Mediation: Sociocultural and Psychological Perspectives. New York, Academic Press.
• — 1985b. Diagrammatic Icons and Historical Processes in Belau. American Anthropologist, 87:840-52.
• Pätzold, Klaus, 1968. Die Palau-sprache und ihre Stellung zu anderen indonesischen Sprachen. Berlin, D. Reimer.

• Riesenfeld, Alfonse, 1950. The Megalithic Culture of Melanesia. Leiden, Brill.

• Sahlins, Marshall, 1981a. Historical Metaphors and Mythical Realities: Structure in the Early History of the Sandwich Island Kingdom. Ann Arbor, University of Michigan Press.
• — 1981b. The Stranger-King or Dumézil among the Fijians. Journal of Pacific History, 16:107-32.
• — 1983. Other Times, Other Customs: The Anthropology of History. American Anthropologist, 85:517-44.
• Shutler, R., and J. C. Marck, 1975. On the Dispersal of the Austronesian Horticulturalists. Archaeology and Physical Anthropology of Oceania, 10:81-113.

• Takayama, J., 1979. Archaeological Investigations of PAAT-2 in the Palaus, in H. Kusakabe (ed.), Cultural Anthropological Research on Folk Culture in the Western Caroline Islands of Micronesia. Tokyo, Tokyo University of Foreign Studies.

- 193

• Tambiah, S. J., 1976. World Conquerer and World Renouncer: A Study of Buddhism and Polity in Thailand against a Historical Background. Cambridge, Cambridge University Press.

• Useem, John, 1949. Report on Palau. Coordinated Investigations in Micronesian Anthropology, Report 21. Washington, D.C., Pacific Science Board.

• Vernant, J.-P., 1983. Myth and Society in Ancient Greece. London: Routledge, Kegan Paul.
• Vidich, Arthur, 1949. Political Factionalism in Palau. Coordinated Investigations in Micronesian Anthropology, Report 23. Washington, D.C., Pacific Science Board.

• Wilson, James, 1799. A Missionary Voyage to the Southern Pacific Ocean . . . in the Years 1796, 1797, 1798. London, T. Chapman.

- 194 Page is blank