The House: Domestic and Public

As briefly noted in chapter 3, I treat inequality narrowly as ideological inequality. That is, I focus on how the Lese pattern social and cultural differences among human beings so that they are unequal to one another and are vertically arranged. I follow J. Rousseau's definition of ideological inequality as "conceptual sets which establish the superiority of some categories over others" (1990:163). Forms of political and economic inequality (for example, control over decision-making processes and the means of production) often accompany ideological inequality, but these more objective bases are to be found in detailed studies of politics and the relations of production. I am concerned primarily with the images and ideals of difference as they appear in a particular Lese discourse about the Efe.

It is in light of inequality that the house holds promise for a critical anthropology of Africa, a field that has relied so heavily on analyses of clans, descent groups, and other units that are, sometimes by definition, egalitarian social organizations, and often have little to do with gender or ethnic relations. The tendency to explore clan relations, for example, as if they could explain a social totality has undoubtedly affected the data we choose to gather and the interpretations and conclusions that ensue from them. Neither Winter nor Turnbull, for example, ever approached the signs and practices of inequality because they did not explore the contexts in which they exist. Clans tell us something very important about the ideologies framed by men about men, but they often tell us mainly about sameness and equality, egalitarianism and

solidarity. It is in the context of egalitarianism, of course, that the clan, descent, and lineage become so important to the Lese. The organization of descent lines into clans is especially important in times of conflict in which the clan ideology of organization is invoked. Clans unite in opposition to other clans and can mobilize for collective action in the case of dispute or illness. This is the principle behind segmentary opposition: there are groupings, some of which may be short-term, which must be solidary and egalitarian for the purposes at hand. But while the clan is essential to Lese social life, it is simply one level, one location, at which to analyze social organization. One location is not more important than the other, but an accurate anthropological picture depicts society as the product of the relationship between these two coexisting models.

The salience of the house also tells us something more general about egalitarianism and the so-called "agnatic principle" in patrilineal societies (Evans-Pritchard 1940, 1951). Evans-Pritchard explained Nuer social organization by positing a structural principle that achieved a degree of social equilibrium and integrated all persons into agnatic social units: "tribes," "clans," "lineages," and "lineage segments." I noted, however, that Evans-Pritchard translates the Nuer terms for these social units as "house," "hearth," or "entrance to the hut." Kathleen Gough has located other inconsistencies with the agnatic principles. In her reanalysis of Evans-Pritchard's data, she found that just as the whole of Nuer society could not accurately be termed egalitarian, so, too, not all Nuer were equally "agnatic." Aristocrats conformed fairly consistently to the agnatic pattern, but commoners varied considerably in how they reckoned descent and formed marriages and households. The variation was due, in part, to the fact that the Nuer had been expanding at least since the mid-nineteenth century and had absorbed numerous captured or immigrant Dinka into their households (Kelly 1985); by establishing a number of customary legal fictions, many Nuer could interpret the new social situations to fit the patrilineal principles, but "the less successful layers had to abandon strict patrilineal principles and adhere to the former [more successful layers of the population] through cognatic and affinal ties in order to reintegrate themselves into society" (Gough 1971:117). Despite his repeated descriptions of the integration of Dinka into Nuer households in his Kinship and Marriage among the Nuer , Evans-Pritchard, Gough argues, found an unstable and internally differentiated society in transition but sought to explain it with a timeless, homogeneous, and egalitarian structural principle

that obscured the unequal distribution of status and property, and the social composition of households (see also Hutchinson 1992). Gough's reanalysis suggests to me that a focus on the house might have revealed the social and historical complexities masked by Evans-Pritchard's specific focus on patrilineages.

The relationship between the house and clan brings us back to the conventional analytical distinction between domestic and public domains, a distinction that led to a methodological choice by most anthropologists to study one domain or the other, or to describe the domestic domain somewhat superficially and to employ more sophisticated analyses for the public domain. Those with an interest in politics studied bureaucracies, descent, and corporate groups, while those with an interest in domestic issues (a small minority in social anthropology) studied the household and relations between men and women. Indeed, the numerous studies of households and their economic functions in sub-Saharan Africa (see Guyer 1981) tend to isolate the household from "exogenous" social, political, and economic processes (Heald 1991:131). The infrequency with which the two domains have been systematically linked in anthropological analysis is in large part due to the tendency to bring such an analytic dichotomty with us to the field—a dichotomy that has been a part of Western social thought at least since the late nineteenth century (Coward 1983)—but it may also be due to the tendency for anthropologists to accept informants' renderings of their world at face value. The separation between the domestic and the public sometimes reflects local models of society as represented by our informants quite as much as it represents our own concerns or ideologies. But we must compare those models with ethnographic observation and experience, for ethnographic analysis cannot simply mirror local models, and the empirical and the ideological rarely agree.

Fortes himself made numerous caveats about this analytical distinction, remarking that the domestic and public domains are usually linked empirically. "The actualities of kinship relations and kinship behaviors are compounded of elements derived from both domains. . . . Every person is an agent, actual or potential, in all domains" (1969:251). Yet, as Yanagisako says of the Fortesian dichotomy, "There is, in fact, a tendency for the terms to be employed to refer to whole social relations (rather than to their contexts and implications) and to entire social institutions (rather than to facets of social institutions)" (1979:187). In addition, where the split is expressed by informants, it has to be contextualized as an indigenous or local model. In some societies, as in South

Asia, the domestic/public split will refer to the sexual purity and impurity of women, while in some, as in the United States, the same ideological split will refer to the opposition of love and money (Rapp 1979:510). In still others, such as among the Lese, where the private is to the public as wives and the Efe are to Lese men, "domestic" relies upon an ideology of ethnicity as well as gender.