In the past the house has often figured in ethnographies as an item of material culture, as an object replete in symbolic meanings, or as the locus of the domestic domain and systems of household production. Various studies have revealed ways in which aspects of architecture are used to symbolize social relations and categories, whilst others have shown how gender representations involve the house, through the division of domestic space and the role of women in the domestic economy, itself frequently modelled on the house. More recently however, an anthropology of the ‘house’ has emerged which aims to unite these diverse elements, combining an analysis of the house as built environment with the ‘house’ as a category and idea central to the conceptualization and practice of social relations.
Much of this work forms part of current attempts to rethink the anthropological study of kinship, adopting a cultural approach which focuses on native categories as a better basis for analysing indigenous understandings of relationships (Schneider 1984), and in the process criticizing various elements of established kinship theory. Studies from Africa, for example, have used the concept of the ‘house’ to question the usefulness of descent theory (Kuper 1993; cf. Saul 1991), whilst the concept has been fruitfully employed by others working in Southeast Asia and lowland South .America (Macdonald 1987; Carsten and Hugh-Jones 1995). Although many draw inspiration from the writings of Claude Levi-Strauss, some see a need to move beyond his model of ‘house societies’, which is arguably mired in the sort of problems more recent work has attempted to overcome.
Levi-Strauss and ‘house societies’
Whilst Levi-Strauss’s thinking on ‘house societies’ (societes a maison) represents his most recent engagement with the study of kinship, he has written no extended account of the subject in the manner of his earlier work on elementary structures, nor does he fully contextualize what he does say in terms of that earlier work. Despite this however, Levi-Strauss’s writings on the house can be seen as an attempt to deal with an institution which lies somewhere between elementary and complex systems, combining elements of descent and alliance, and found in widely separated societies.
The initial discussion of the ‘house’ comes in Levi-Strauss’s re-analysis of Kwakiutl social organization (Levi-Strauss 1982: 163—87), where he attempts to overcome some of the complex problems encountered by Boas in trying to describe the kinship system of a society that exhibited both "patrilineal and "matrilineal elements, and for which all available analytic terms proved misleading. Using both comparative ethnography and history, Levi-Strauss comes to the conclusion that the basic unit of Kwakiutl society was similar to that of the Yurok of California (studied by "Kroeber) who had similar institutions named ‘houses’, which in turn closely resembled the ‘houses’ of medieval Europe and feudal Japan. Here Levi-Strauss defines the ‘house’ as a ‘moral person’ (sometimes taken to mean corporate group) which perpetuates itself through time, this continuity expressed in the ‘language of kinship or affinity’, and quite often both (1982: 174). These ‘houses’ exhibit organizing principles that are usually seen as mutually exclusive in kinship analysis: patriliny and matriliny, filiation and residence, hyper-gamy and hypogamy, exogamy and endogamy. However, it is the peculiarity of the ‘house’ as an institution that it in fact transcends these theoretically irreconcilable traits, rather than being undone by them.
In a subsequent work, Levi-Strauss (1987: 149-94) extends the discussion of the ‘house’ to various societies of the Pacific, Melanesia, Indonesia and Africa, seeking ‘common social-structural characteristics that might explain [the] appearance of the same institutions among peoples far distant in both time and space’ (1987:152). Having previously stressed the continuity of the ‘house’ in a ‘real or imaginary line’ (1982: 174), Levi-Strauss here suggests that the ‘house’ acts to ‘solidify’ the ‘unstable relation of alliance’, a form of ‘objectification’ born of the oppositions characteristic of ‘house societies’, in particular the competing principles of descent and alliance (1987: 155-9). This instability coincides with a situation where political and economic interests have begun to invade the social field without entirely undermining the importance of kinship. Indeed, Levi-Strauss suggests that the ‘language of kinship’ is borrowed to talk of political and economic interests, there being none other available to ‘house societies’ (1982: 186-7). This tension between kinship and political and economic interests is referred to in terms of the ‘duality of ‘blood’ and ‘land’, and is seen as typical of ‘house societies’ (1987: 181-2).
Although Levi-Strauss’s writings on the ‘house’ have provided inspiration for some, his theory arguably ‘raises as many problems as it solves’ (Carsten and Hugh-Jones 1995: 19). Whilst some have tried to refine Levi-Strauss’s theory (see e.g. some of the contributors to Macdonald 1987), most have used it as the starting point for a more holistic analysis of social relations, criticizing Levi-Strauss’s view of ‘house societies’ as a strict category. There are problems for example with the applicability of the model, for whilst Levi-Strauss ranges with breathtaking ease across vast tracts of cultural and historical time and space in the defining of a social type, some question the usefulness of a model which encompasses such diversity. In addition, some criticize the rigidity of the definition of the ‘house’, finding it a useful model outside the contexts in which Levi-Strauss has used it. Furthermore, although Levi-Strauss talks of the house as the ‘objectification’ of the relation of alliance, there are cases where the ‘house’ is associated with the conjugal couple rather than alliance per se, whilst elsewhere the ‘house’ is associated with siblingship, or a form of generational continuity akin to the notion of descent.
For most who have attempted to use Levi-Strauss’s ideas there is agreement that some of the limiting aspects of his definition of ‘house societies’ must be abandoned, a move which particularly leads to a questioning of the notion of kinship underlying Levi-Strauss’s discussion of the ‘house’. Besides a latent functionalism in Levi-Strauss’s writings, such that the role of the ‘house’ is to overcome the problems of opposing principles, there is also a quasi-evolutionism, whereby ‘house societies’ exist somewhere between elementary and complex structures, as well as between preliterate and modern societies. Nowhere is this clearer than in his discussion of Africa, where Levi-Strauss links an account of the ‘house’ in the Niger delta region to the emergence of ‘kings’ who ‘based their power, not on ties of descent, but on military and economic success’ (1987: 186). When political and economic forces come into play, Levi-Strauss suggests that the institution of the ‘house’ overcomes the duality of ‘blood’ and the ‘genealogical chain’ on the one hand, and ‘land’ and ‘ancestral soil’ on the other (1987: 181-2). As with his earlier work on elementary structures, here Levi-Strauss assumes that kinship is principally about ‘blood’ and genealogy. But there is a missed opportunity in his discussion when he notes that the ‘representation of social relations in material form’ (1987: 160) may appear in shapes other than the ‘house’, a point he fails to develop by taking it to its logical conclusion: that whilst the ‘house’ is one form of the representation of social relations, so too is ‘blood’ or ‘land’, none logically prior to the other (cf. Carsten and Hugh-Jones 1995: 19). In his work on myth Levi-Strauss has brilliantly revealed the complexity and subtlety of the ways in which the human mind classifies and categorizes the world; however he fails to bring this understanding to the ‘house’, where instead of seeing in the ‘house’ the basis of a radical critique of the theory it fails to be classified by, he instead creates a model of an institution which, whilst transcending previous categories of kinship analysis, remains premised upon them. The fundamental problem with the model is thus that whereas Levi-Strauss claims that it is ‘house societies’ which employ the ‘language of kinship’ to talk about political and economic interests, it is in fact Levi-Strauss who employs the ‘language of kinship’ to talk about relations which a variety of societies phrase in the ‘language of the house’.
Although Levi-Strauss may have ‘reminded us of a sound anthropological principle: the priority of native categories’ (Carsten and Hugh-Jones 1995: 20), it is not a principle that Levi-Strauss himself fully adopts. However, those who have sought to go beyond Levi-Strauss’s theory have usually found his discussion useful in situations where the ‘house’ is a salient native category of social relations. But, as one frequently finds the word for ‘house’ employed as a term for a social grouping in a wide range of societies, the problem becomes ‘not one of discovering which are ‘house societies’ but of discovering which ones are not’ (Carsten and Hugh-Jones 1995: 18). Such classic examples in the field of kinship studies as the Lovedu, Nuer, Tallensi and Tiko-pia all use ‘house’ terms for groups related by descent and/or alliance. These and other examples therefore suggest that Levi-Strauss’s model of ‘house societies’ needs to be reformulated in the process of creating a ‘holistic anthropology of the house’ (Carsten and Hugh-Jones 1995: 4) which combines the analysis of native kinship categories with the study of architecture, symbolism, and indigenous models of political and economic organization.
Recent work in this vein aims to move beyond the outmoded categories and formalist typologies of kinship theory, taking a processual view of social relations. The study of the ‘house’ is thus not simply an attempt to explain the social organization of those societies which previously resisted typologies, for it can also be used to shed light on societies previously classified as descent-based or alliance-based, such as the Kachin. A brief look at Kachin ethnography reveals the house to be central to Kachin thinking about kinship, politics, economics and religion, and an account which focuses on Kachin categories and metaphors arguably gives us a better understanding of the Kachin sociological imagination than does an account framed in terms of descent and alliance.
‘Houses’ and the Kachin
Among other things, the Kachin (Leach 1954) are famous in anthropology as an example of a society that practises matrilateral cross-cousin marriage, an exchange relationship of alliance that takes place between two or more lineages related as wife-givers (mayu) and wife-takers (dama). Leach tells us that among the Kachin there exists a system of patrilineal clanship, with clans made up of a number of localized lineages (1954: 55), although he adds that within villages, people of one lineage are usually related to everyone else by affinal ties rather than common clanship, that is by alliance rather than descent (1954: 68).
When one looks at the Kachin terms for these groups and relations (see 1954: 108—36), some rather interesting things emerge. Those of one lineage are said be of ‘one household’ (htinggaw), and share a ‘household name’ (htinggaw amying) which they inherit from their father. Glossed by Leach as ‘lineage’, the term htinggaw literally means ‘people under one roof, and in addition Kachin refer to ‘lineage’ members as ‘brothers’ (kahpu-kanaii), those who are of the same ‘branch’ (lakung), or the same ‘hearth’ (dap), and those who are of the same ‘sort’ (amyu), a term which Leach glosses as ‘clan’. Although members of a htinggaw do not necessarily live together in a single dwelling, the social group is still referred to as ‘people under one roof, whilst Kachin refer to a ‘family’ as ‘people of the house’ (nta masha). Finally, the individual household is the basic unit of economic organization, working its own swiddens and gardens, a unit which comes into being through marriage, when a man is said to ‘extend the roof (htinggaw rawn), and the husband and wife become ‘owners’ or ‘lords’ (madu ni ) of their house.
The architecture of the Kachin house is also important, particularly as it relates to the differences between chiefs’ houses (htingnu) and those of commoners (nta). As well as being larger than other houses, their grandeur a marker of status and prestige, chiefs’ houses have oversized hardwood house posts (nhpu daw), a sign of chiefly rank, and some chiefs are reported to have surrounded their houses with elaborate stone walls. Within the house another feature differentiates chiefs’ houses from other dwellings, the possession of the Madai dap, a shrine to the chief sky spirit, Madai (dap, ‘hearth’). Members of a htinggaw are said to be ‘people who worship the same set of household spirits’, and all houses contain a shrine to the masha nat, spirits of the ancestors of the husband of the house. However, the right to sacrifice to the chief sky spirit is the prerequisite of chiefs, and it is only in their houses that one finds the Madai dap shrine.
As for the term for chief s house itself, htingnu literally means ‘house of the mother’ or ‘the mother house’ (nu, ‘mother’), for which Leach offers two interpretations: first, ‘that the chief s house is the parent house from which other houses of the village have segmented off; the other that the house of the chief is mayu [wife-giver] to the houses of his followers’ (1954: 112). The gendered nature of the term, however, appears confusing, particularly given that Madai ‘represents the male principle’ (1954: 112) and that the wife-giver/wife-taker relationship appears to be gendered male. It is therefore surely significant that although Madai is wife-giver to all the chiefly ‘lineages’, their common ancestor is Madai’s daughter, ‘mother’ to all those of chiefly ancestry.
A chief s house, Leach suggests, ‘is more than a dwelling-house, more than a palace, it is also a kind of temple to the Madai [spirit]‘ (1954: 113). But the Kachin ‘house’ is also more than a religious building and a place to perform rituals. It figures as the basic unit of economic organization, and is associated with the continuity of a group of ‘brothers’, men of the same ‘hearth’ who worship in their house their own ‘ancestor spirits’ and who pass on to their children their ‘household name’. The ‘house’ is similarly central to marriage, for besides the fact that alliance relations exist between groups known as ‘people under one roof, the conjugal union itself is said to ‘extend the roof when a man sets himself up with his wife as ‘owner’ of a house. Finally, the difference between chiefs and commoners is marked through differences in domestic architecture. Kachin categories therefore reveal a conceptualization of social relations premised on metaphors about the ‘house’, which rather than simply being an ‘objectification’ of the relation of alliance, emerges as an object significant in various domains. It is by trying to combine elements such as those used in the foregoing analysis that the anthropology of the ‘house’ has begun to rethink the study of kinship, focusing on connections between houses as items of vernacular architecture and the people who build and live in them.
Rethinking kinship
As well as attempting to bring together various domains of anthropological enquiry which often remain separate, recent work on the ‘house’ has joined a more widespread critique of anthropology’s own metaphors. In some of this work there has been a move away from seeing kinship in terms of a set of structured relationships lying at the heart of systems of descent and alliance, and an emphasis on the need to analyse native categories as a basis for understanding the practice and conceptualization of relationships as an ongoing process. In many cases the ‘house’ emerges as a category better suited to such analysis than our own metaphors of ‘lineage’ and ‘clan’. That the ‘house’ crops up so often as a metaphor of social relationships is perhaps not surprising, given that the house is the place where much that is usually taken to constitute kinship is practised. As Carsten and Hugh-Jones point out, kinship is about ‘sleeping together … living together, eating together and dying together, not just about bed but also about house, hearth and tomb’ (1995: 19). Nowhere are these relations more vividly realized than in some parts of the Austronesian world, where the house is a salient category and metaphor of social relations as well as being the dwelling place of the living built over the burial place of the dead.
