ÁñÁ«ÊÓƵ

In a series of lectures given at the College de France in 1978, Claude L¨¦vi-Strauss proposed a model of social structure for societies that lack durable kinship structures. The idea emerged from his re-analysis of one of anthropology's favourite subjects, the Kwakiutl (Kwakwaka'wakw) of coastal British Columbia. In their world, shared descent from a common ancestor is not the most important principle determining one's social identity. The social groupings created by descent are relegated to the background, in favour of one's affiliation with great houses containing many families. The Kwakwaka'wakw house is "a corporate body holding an estate made up of both material and immaterial wealth which perpetuates itself through the transmission of its name, its goods, and its titles down a real or imaginary line...", L¨¦vi-Strauss wrote. In this way, it resembles the noble houses of Europe, such as that of the Plantagenets.

Rather surprisingly, Levi-Strauss' notion of soci¨¦t¨¦s '¨¤ maison , or "house-based societies", has been taken up enthusiastically by anthropologists working on the other side of the world, in the tropics of Southeast Asia. Since the 1987 publication of De la Hutte au Palais: Soci¨¦t¨¦s '¨¤ maison' en Asie du Sud-Est Insulaire , three additional edited volumes have appeared, of which The House in Southeast Asia is the most recent.

There is a certain irony in the vogue for soci¨¦t¨¦s '¨¤ maison in this part of the world, for it was here that anthropologists first encountered the complex patterns of marital alliance and cosmological order that inspired L¨¦vi-Strauss' vision of a "structural anthropology". In 1935, a Dutch ethnographer, F. A. E. van Wouden, published his doctoral dissertation on Types of Social Structure in Eastern Indonesia . According to Van Wouden, in these societies prescriptive cross-cousin marriage creates durable alliances among patrilineages. Such alliances form the basis for all-encompassing classificatory schemes that unite the social and cosmological domains. It was precisely the absence of these typically "Indonesian" characteristics - clans, lineages and prescriptive rules of marriage - among the Kwakwaka'wakw that led L¨¦vi-Strauss to formulate his concept of soci¨¦t¨¦s ¨¤ maison .

So it is puzzling, as one editor of this volume notes, to discover societies that seem to fit both these models. The Lio, of the Indonesian island of Flores, could serve as a textbook example for L¨¦vi-Strauss' concept of a social system defined by structuralist principles. Organised into patrilineal clans, they practise prescriptive cousin marriage and celebrate the cosmological harmony thus achieved in art and ritual. But "at the same time, houses are of the utmost significance in their understanding of reality. The Lio house may be thought of as a moral person... it organises and expresses kinship and marriage; it is a microcosm of the cosmos, linking the past with the present, it is imbued with tradition, biography and sacredness".

For L¨¦vi-Strauss, soci¨¦t¨¦s ¨¤ maison are stratified societies without unilineal descent groups, in which multi-family houses form part of a corporate estate. They represent a transitional phase, in which emerging elites still need to "borrow the language of kinship" to achieve their hierarchical ambitions. Ironically, as the vogue for finding such societies in Southeast Asia grows, this view looks less tenable, for few of them possess all these diagnostic features. Indeed, the discovery of so many purported societes a maison undermines the theoretical justification for even formulating the idea in the first place.

But most of the contributors to this volume engage only tangentially with L¨¦vi-Strauss' original argument. What is retained is a free-floating concept of "house societies" that highlights either the metaphorical bestowal of personhood on houses or the identification of houses with social groups, often by celebrating ties to the ancestors. Thus Jacques Ivanoff suggests that the Moken sea nomads of Thailand may be considered a "house society" despite the absence of real houses because they return to their islands of origin during the monsoon to reside in the company of their ancestors. Similarly, Janet Carsten sees the Langkawi of Malaysia as creating a "shadow" image of their community as a house society in communal feasts.

Still, some examples suggest L¨¦vi-Strauss was on to something. In Tana Toraja, aristocratic houses generally have genealogies with a depth of between ten and 30 generations, and serve as master symbols for the status claims of the elite. More modestly, the Kelabit of Sarawak define hierarchical relationships in terms of precedence in longhouses; indeed "in some sense the entire Kelabit social universe is one big 'house'".

The concept of a house as an embodied "person" gives rise to a wealth of architectural imagery. Houses may have faces, breasts, arms and legs; they may be male or female; the connection to their inhabitants may be expressed by burying the placenta of newborns beneath it, or hanging it in the rafters, and the bodies of the dead are often placed within or beneath the house. Many of these customs are remarkably similar among the Austronesian-speaking societies of the region, suggesting a common historical origin. But not all of the purported "house societies" discussed in the book are of Austronesian origin; the Tai Yong and Isan inhabit northern Thailand. Evidently, as the editors conclude, "much remains to be explored".

Although this book is written for anthropologists, the sheer richness of the ethnographic materials described in it will reward some general readers. Anyone intrigued by the broader questions posed by these discoveries will also appreciate Janet Hoskins' Biographical Objects (1998).

J. Stephen Lansing is professor of anthropology, University of Arizona, US.