My ethnographic research has revealed a confluence of Austronesian and Indic influences in highland Balinese culture. Annual ceremonies at the regional temples, which form the ritual centres of banua, for example, contain some elements of classical Hindu religion alongside other important rites, such as buffalo sacrifices, which are more commonly associated with cultures of the Austronesian region of which Bali is a part. In short, the cultural history of the banua of the Balinese highlands is in many ways typical of societies in the Indian-influenced part of the Austronesian world.

Megalithic stone carvings and other archaeological evidence suggests that some major contemporary regional temples of highland Bali have been places of worship from prehistoric times (Bernet-Kempers 1991). The earliest Indic kingdom on Indonesian soil was established in Kutai (Borneo) in the fifth century AD and others soon after in Java and Sumatra, while Indian trade and cultural influence in the region, including Bali, probably dates back to the first century AD at least (Ardika and Bellwood 1991).

The earliest written sources of local cultural history are ninth-century royal edicts from the first Balinese Hindu kingdoms (negara), which might have been centred in the region of the highlands and northern coast. These edicts depict several important and ancient temples that are the ritual centres of contemporary banua already at the heart of regional ritual organisations (Reuter 2002a). The prehistoric significance of these same sacred sites makes it seem very likely that ritual domains preceded the formation of early Hindu kingdoms on Bali as well as providing an organisational platform for their establishment, not just in the highlands but on the island as a whole (for evidence of similar ritual networks with royal patronage in southern Bali, see MacRae, this volume). By the same token, there is no doubt that the evolution of highland domains took a new direction under the influence of local Hindu kingdoms and again after their collapse in the 14th century, when Javanese invaders established new Hindu dynasties centred in the south.

Bali Aga society provides an ideal contemporary vantage point from which to explore some of the Austronesian foundations of Balinese culture that have become less visible in the polities and communities of the southern lowlands (Reuter 2002a). Highland people actively resisted the influence of Javanese-descended royal courts of southern Bali and the orthodoxies of the Brahmana priesthood affiliated with these courts for more than half a millennium (Reuter 2003). Studies of Austronesian traditions in highland Bali and in Balinese culture generally are also exceptionally valuable to comparative Austronesian studies because they were and still are better protected under the umbrella of Hinduism, with its great tolerance of local diversity, than societies in other parts of Indonesia and beyond whose local religious practices and belief systems have been devalued and abandoned under the influence of proselytising, monotheistic world religions. ^[2]

Many societies in the Austronesian-speaking world show patterns of regional and local organisation wherein a category of ‘place’—the kingdom, domain, village or house—serves as a key idiom of social identity in its own right, though often also in combination with a notion of common ancestry or ‘name’. In the Balinese case, and probably in other societies too, there are recurrent metaphors, idioms and concepts in how the idea of ‘place’ is defined across different levels of social organisation. Indeed, the distinctions house-village, village-domain, and domain-kingdom are all more or less fluid (see Reuter 2002a). In part, this conceptual similarity and fluidity is a reflection of historical processes. A house with a cluster of branch houses could grow into a village, which in turn could acquire branch villages and grow into a domain, which could—alone or in conjunction with other, allied domains—become the ritual foundation for the formation of a Hindu polity. ^[3]

Of the four most important terms for territorial categories in Bali, ‘desa’, the term for village, and negara, the term for a kingdom or its capital town, are of Sanskrit origin while umah, ‘house’, and ‘banua’, the contemporary designation for a regional ritual domain, are Austronesian words. My main focus in this paper will be on the regional institution of the banua, given that regional forms of social organisation in the Austronesian world remain relatively under-explored. There is a large body of literature on the comparative ethnology of much smaller social forms, namely the ‘house’ (*rumaq, *balay, but occasionally also *banua) and there has also been much research and debate on very large institutions, such as the classical Indic polities of South-East Asia (kerajaan, negara). The comparative lack of studies on regional organisation is regrettable especially because it is rather difficult to imagine how states or large chiefdoms in the Austronesian world could have been built directly on a foundation of localised houses and in the complete absence of ritual alliances between houses within villages, domains, or other, similar social structures of intermediary size.

Any comparison of the social organisation of a number of Austronesian societies inevitably leads to the discovery that speakers of Austronesian languages use similar metaphors and related words to describe their ideas of and relationship to a particular stretch of land, even though the meaning and social implications of the reflexes may vary widely. In this case, reflexes of the Proto-Malayo-Polynesian reconstruction *banua in the contemporary languages of this language family all tend to convey the general idea of a ‘place’ of varying size, namely, the ancestral territory of a ‘longhouse’, ‘hamlet’, ‘village’ or ‘domain’. For example, in Old Javanese, the term wanua connotes a ‘village’; in Iban, menoa rumah is the ‘territorial domain of a longhouse’; in Ngada (Sara-Sedu), a nua is a ‘village’ or ‘ritual territory’; and in Lio, nua is a ‘ritual territory, domain or polity’. Similarly, in Goodenough, manua connotes a ‘village’ or ‘dwelling place’, while in Vanuatu (East Ambai) and Fiji vanua, is a ‘land’ or ‘territory’. More sporadically, reflexes of *banua are used to refer to (the territory of) a ‘house’, as in Toraja, Banggai, Wolio, Molima and Wusi-Mana. Two other sets of terms with a similar emphasis on ‘place’ in this group of languages include reflexes of the contructs *tanah (tana, tanaq, taneq, taneh, tano, ’ano), of which the Balinese reflex is tanah, and *daya (darat, dare, dae, rae, rai).

Similarities and differences in the cultural meaning and social implications of these terms need to be established through detailed ethnographic and ethno-historical research, followed by ethnological comparison. This volume is dedicated to the task of an ethnological comparison of social categories based on the idea of a place or territory. As a contribution, this paper will unpack the meaning of the Balinese reflex banua by describing in some detail what a banua is or can be in the highlands of Bali.