Variation and Fluidity in Household Composition in Phoenix, Durban

Abstract

INTRODUCTION The purpose of this paper is to analyze the structural characteristics of a group of Indian households which do not conform to the conventional notion of patrilineal joint family. It attempts to delineate the variations in lower class compositions in an urban Indian dominated residential area, and to contribute to the ethnographic landscape of South Africa's lower class domestic formations. I emphasize this particular racial category because South Africa's stratified social ladder, consisting of Whites, Coloureds, Indians and Africans' respectively, had given rise to differentiated access to political office and economic resources. The effects of white hegemony, consolidated through the ideology of apartheid, are reasonably well documented in statistical and macro based researches, especially as it has affected Africans (Greenberg, 1980; Manganyi & du Toit, 1990; Thompson and Butler, 1975; Wilson & Ramphele, 1989). There is therefore a double edged criticism against these approaches. Most commonly they have articulated the ills of apartheid in terms of national oppression of the African alone, rather than in terms of degrees of racial and class oppression. They further ignored the variations of the historical experiences and social manifestations of other racial categories such as the Coloureds and Indians. This approach, in favour of examining the oppression of Africans, gave rise to research in anthropology which illustrated the link between the discriminatory effects of apartheid and the erosion of the most basic unit of their social structure, i.e. the family or the lineage. Some of the more careful and critical historical materialist analyses by Murray (1981; 1987), Sharp and Spiegel (1985) and Spiegel (1986) have provided mechanisms which demystify the complexity of domestic relationships in South Africa's rural areas. Constrained by the applicability of family, genealogical co-resident dwelling groups and conventional kinship structures, they produced out of their data the alternative analytical tools of household (Sharp and Spiegel, 1985; Murray, 1981; 1987; Martin and Beittel, 1987) and site (Spiegel, 1986). Collectively these writers view the as a conglomeration of economic marginalised individuals who are both kin and non-kin and who pool their resources for their mutual benefit. The meagreness of their material conditions, has produced a situation of frequent turnover in residents, and often complete dissolution of it as a social unit. The temporal nature of the structure has led Spiegel (1986) to adopt the site on which households set up their homesteads as a more reliable unit of analysis. The is characterised by male absenteeism - brought about by the export of labour to South Africa's core industrial areas; the emergence of female responsibility and the reversal of gender roles, wide geographical dispersion of genealogical kin - giving rise to differential de facto and de jure populations2; and dependence either upon remittances, or upon subsistence farming through various contractual agreements for their survival. Taking these factors into account, Martin and Beittel (1987:218) stated that Constituted by a small group the is the unit that ensures the continued reproduction of labour through organising the consumption of a collective fund of material - a unit therefore different from the family, co-resident dwelling groups and kinship structures. The may encompass these units, or be structured along their lines, but it may not and so is not identified with them. It is in the latter sense that I use the concept to refer to my data from Phoenix. It will demonstrate that the in an Indian urban complex constitutes a range of variables which are different from the characteristics of the rural African household. Not hampered by male absenteeism, or reliant upon remittances, or farming, the in Phoenix is meant to refer to at least 3 things: 1 institutions of kinship responsibility which include principles of matriarchy and matrilineality; and 2. …