In this volume, in addition to the paper on the Bali Aga, there are a number of other papers on societies in which some reflex of *banua continues to be an important organising category: three societies in eastern Indonesia—the Keo of Central Flores, the indigenous Masarete-speaking population of Buru, and the Alune of the island of Seram—and two societies in the Pacific—the Raga of North Pentecost and the people of the Kingdom of Tonga. In each instance, reflexes of this category (nua, fena, hena, vanua, fonua) are embedded with other spatial categories and take their meaning, in relational terms, in a matrix of senses. It is possible to trace various comparative pathways among these cases. Interestingly, these pathways, signposted by reflexes of *banua, intersect with other pathways, marked by reflexes of PMP *taneq (POc: *tanoq) which has similar meanings of ‘earth, land’.

Philipus Tule’s paper on the Keo of Flores offers an examination of the contextual significance of the *banua category. In the case of the Keo, it is of interest that the reflex nua occurs with another recognisable widespread Austronesian term for ‘land’—tana. The Keo identify themselves as ‘children’ of the land (tana) in both a spatial and a genealogical sense. The land is ‘feminised’ (’ine tana) as ‘mother’ and all Keo derive from this ‘mother’. In ritual language, the male complement to this ‘mother land’ is ‘father stone’ (’ame watu)—all that surmounts the land. Figures of custodial authority are identified with the land and are referred to by the dual name ‘mother land, father stone’ (’ine tana, ’ame watu). Within this context, the term nua refers to the autonomous settlements, each with its own ritual leadership (mosa daki—or, in ritual language, mosa nua, daki oda) whose rituals are conducted through the origin houses (sa’o pu’u) that make up a village.

Crucially, among the Keo, there are two constituent categories of people: the ’ana tana, an indigenous category of those whose origins are affirmed through their origin houses, and ata demba mai, incorporated ‘migrants’ of various sorts, whose origins are known to derive from outside the tana Keo. Ultimately, the constitution of Keo as a domain is based on an assertion of the knowledge of origins combined with the ritual acknowledgment and celebration of such origins. The nua serve as the named settlements in which the rituals of the origin of the land (tana) are grounded.

An examination of the use of reflexes of *banua among the Keo of Flores with either the indigenous population of Buru or the Alune of Seram, who also structure their social worlds by means of this category, is one pathway of comparison, but it is also instructive to compare the Keo with the Sikka and related population of Tana ’Ai who live further east from the Keo on Flores.

For each of these societies, tana functions as the ‘higher’ category. It is the tana that Lewis defines as ‘domain’—’the highest order of the classificatory categories of Tana ’Ai society’. A tana, Lewis writes, is ‘a loosely organised region defined by a centre and whose peripheries form no clear boundary. A tana’s centre is defined ritually (if not geographically) by its mahe, the domain’s central ceremonial site, and socially by the ceremonial office of the tana pu’an [its ‘source of origin’].’ The key to understanding a tana, in this context as in the case of the Keo, is its ritual definition and ceremonial continuity. In the case of the Keo, the head of a particular clan identified in relation to a specific nua is the local ritual celebrant; in Tana ’Ai, this celebrant is identified by clan affiliation alone.

In Central Sikka, where once there were more than 40 named tana, a political process of consolidation and redefinition has occurred. It is this historical process that Lewis sketches in his paper. At the heart of this process is an origin charter that transforms the basis of precedence allowing outsiders and other immigrants to establish their rule within an expanded domain. Similar myths by which an ‘outsider’ is installed ‘inside’—the idea of the ‘stranger king’—occur throughout eastern Indonesia and are, one could argue, a key epistemic component in the structuring of polities in the Austronesian-speaking world. ^[1]

In the comparison with Tana ’Ai, the rulers of Sikka have displaced the ceremonial figure of the tana puang as they established their own political centre at Sikka Natar. Thus in Sikka, the natar became a new ceremonial. In contrast with the Keo, where incorporated immigrants remain subordinate to the children of the land, in Sikka, the outsider has achieved political centrality.

Comparison with the cases from Maluku—Buru and Seram—reflects a different set of historical processes. Among the indigenous population of Buru, the reflex of *banua is fena. As Barbara Dix Grimes notes, the fena constitutes an inhabited, domesticated territory whose ritual custodians are represented by particular clans (noro) within specific locations (neten). These custodians are the geba neten duan. Under a regime of shifting cultivation, the ‘inside’ of this territory (fen-lale) is not a fixed location and may shift in relation to changing patterns among clans. Named territories thus consist of an intersection of relations between clans and specific areas.

Among the Alune of Seram, the equivalent category to that of the fena on Buru is the hena. Where the population of Buru uses the term noro for ‘clan’, the Alune use nuru. The fact that the languages of the two islands are closely related makes comparison between their social systems the more interesting.

Among the Maluku Islands, Seram is considered the ‘Mother Island’ (Nusa Ina) and the source of some of its most profound ritual traditions. Christine Boulan-Smit, in her paper, provides a glimpse of this complexity. Her focus is on one hena, Ma’saman Uwei, of the Alune, who inhabit a segment of the territory known as the ‘Three Large Rivers’ (’Wele Telu Batai) in the western region of the island. In comparison with Buru, the hena is a more ritually circumscribed component involved in a wider range of social and spatial relationships. The ritual ordering of each ‘domain’ is assigned to particular clans that hold specifically named ceremonial positions—seven such positions in Hena Ma’saman Uwei—whose responsibilities are directed within and beyond the domain. This ceremonial system also involves an allocation of land based on the shifting precedence among the constituent clans of the domain.

In the discussion of these various societies from the mountains of Bali to the islands of Maluku in eastern Indonesia, the focus is not simply on inhabited land but on ritually defined, ceremonially ordered social space. Equally important to an understanding of this ordering of space is a social dynamic that defines the categories of ‘inside’ and ‘outside’ and their relationship to one another.

Here, the case of Banda, ‘The Blessed Land’ (Tanah Berkat), presented by Phillip Winn, offers an instructive contrast. Banda is remarkable in eastern Indonesia for the total displacement of its original population. In 1621, when the Dutch East India Company’s (VOC) forces conquered the Banda Islands, they killed, enslaved, deported or drove into exile its entire population and then repopulated these islands as plantation with outside labour. As a consequence, Banda became a Malay-speaking social enclave witshin eastern Indonesia, a wholly immigrant society whose origin narratives link Banda to the foundations of Islam and as a ‘blessed’ land (tanah) directly to Mecca. In place of indigenous claims to origin, local sites imbued with ‘past’ significance are viewed as places (keramat) of ‘hidden’ meaning associated collectively with ancestral spirits (datu-datu). The discourse is that of the Malay world where religious sites, particularly tombs (makam), are the focus of veneration (Fox 1991, 2002). As a society of outsiders, the population of Banda treats the ‘inside’ foci of the land as spiritually powerful but vacant of any specific custodial claim. Yet, as Winn argues, through their ‘ritually based engagement’ with these sites, the population is able to ‘become demonstrably local’.

It is this same theme that is carried forward in the discussion of vanua in North Pentecost and the fonua in Tonga. John Taylor begins his analysis in distinguishing the category of vanua from that of tano, a discussion that has its counterpart in eastern Indonesia. On the one hand, he notes that vanua are relatively small—an acre or two at most. This situation contrasts with all of the Indonesian cases, where banua/nua/fena/hena can encompass a substantial territory. On the other hand, he goes on to note that vanua are identified with what are called bwatun vanua (literally, the ‘source’, ‘foundation’, ‘head’ or ‘roots’ of vanua): ‘Bwatun vanua represent the specific origin places of individual descent groups.’ These origin places are the sites of circular stone piles—grave sites associated with ancestors identified as ‘people of the place’ (atatun vanua). These ancestors are the invisible presence whom the living within each descent group must continue to propitiate.

Instead of holding a position of custodianship within a particular banua/nua/fena/hena, however, each descent group among the Raga ‘retains custodianship of many individual vanua that are scattered throughout the North Pentecost landscape’. Thus, while local discourse resembles that in eastern Indonesia, the scale and scatter of social groupings is strikingly different. Particularly remarkable is the use of the ‘land tree’ as an image of differentiation—an image that is also common throughout Indonesia. Thus, for example, the mythical banyan tree, Nunusaku, at the centre of the ‘Mother Island’ of Seram, provides a similar metaphoric image of differentiation and distribution among all of the peoples of Seram.

Mary Patterson’s paper on North Ambrym provides a counterpoint to John Taylor’s paper on North Pentecost. Where Taylor discusses vanua, Patterson discusses tan. As Patterson indicates, in Vanuatu, the use of vanua or tan in local discourse tends to be mutually exclusive—even to the point of political identification: the Vanua’aku Pati versus the Natui Tano and the Tan Union. Yet her description of local tan resonates with aspects of Taylor’s discussion of North Pentecost with its emphasis on origins and the ancestral spirits of the dead.

In the case of North Ambrym, tan refers to specific ancestral lands where custodial male kindred are described as the ‘doorways’ that provide access to local resources while women of the kindred are the ‘roads’ that ‘go forth and return’. In this metaphoric world, men—when they die—are ‘planted’ by means of sacrifice and marked by the transfer of tusked boars to members of the mother’s origin place.

Appropriately, Tonga offers a case study where the concept of fonua has undergone considerable social reconstruction but continues to figure prominently both nationally as well as locally. As Steve Francis points out, the fonua in contemporary Tonga embodies notions of nationhood. Whereas the fonua in pre-constitutional Tonga were territorial units under chiefly custodians ordered within a social and political hierarchy, the 1875 Constitution transformed this hierarchy and separated the fonua from ‘the geographical context of local territory, the economic context of agricultural production and the social context of ritual obligation’. Interestingly, the process that Steve Francis sketches bears resemblances to the processes described by E. Douglas Lewis in the transformation of central Sikka under the rulers of Sikka Natar.