Family’ is one of the words most commonly used in anthropological writings and discussion, and yet its meaning is neither always clear nor a matter of consensus. This is partly because in everyday use in Euro-American culture, the word covers a multitude of senses of relatedness and connection. It may for instance refer to the domestic group or household, to close kin who are not co-resident, such as parents and adult offspring, or to a much wider network or deeper genealogy of kinship, as in ‘the entire family attended the funeral’, or ‘the house has been in the family for seven generations’. People know what they mean when they use the word family, and the meaning is usually made clear to others by the context in which it is used, but most would find it difficult to define precisely what sorts and range of relationships the word covers. The same complexity of meaning often exists within anthropological writing and, furthermore, the way in which the concept is used, and the sets of assumptions it embraces, have changed radically as new theories of kinship, gender, and social structure have been developed.
A related point is the fact that in Euro-American discourse the concept of the family is politically and ideologically ‘loaded’, or imbued with sets of politically and culturally contested ideas about the correct or moral ways in which people should conduct their lives, and the people with whom they should conduct them. In the postmodern intellectual climate of the 1980s and 1990s, the term has been increasingly subjected to re-analysis, deconstruction, and radical redefinition: as is the case with many other cultural and social categories, emphasis has shifted from one meaning to a plurality of meanings. ‘Families’ have increasingly replaced ‘the family’ as an analytic concept, and the family itself, singular or plural, has come to be seen less and less as a ‘natural’ form of human social organization, and more and more as a culturally and historically specific symbolic system, or ideology.
Types of family
In anthropological writings, different congregations of kin and affines have been labelled as specific types of family. The conjugal family refers to a heterosexual pair and their offspring, while the extended family refers to at least two related conjugal families, and for instance may consist of a woman and man, their children, and the spouse and children of at least one of these, or two or more siblings, their spouses and children. The stem family includes a couple, their unmarried children and one married child with spouse and offspring. Other labels have also been devised to refer to specific types of situation. In most of these definitions the family overlaps with the household or domestic group: that is to say, the family is identified as those kin and affines who live together in the same dwelling, share a common hearth, and jointly participate in production and consumption. As will become clear, this identification of the family with the domestic group has given rise to various analytic problems.
Theoretical origins
Late nineteenth- and early twentieth-century writers, influenced by ideas of social evolution, saw the conjugal family as growing out of more complex systems of kinship, cohabitation or marriage, and divisions of labour. Engels for example explicitly linked the rise of the monogamous nuclear family to sedentary agriculture, the development of private property, and the elaboration of exchange relations between men. In his evolutionary scheme of things, a gendered division of labour which associated women with the domestic sphere and childcare, and men with the outside domain, was both natural and egalitarian; initially neither the domestic nor the external domain was valued over the other. The transformation of this balanced division into a hierarchical order in which women’s labour was undervalued and women’s persons and sexuality controlled by men was effected through the rise of the patri-archical nuclear family and property relations between men.
The three ideas, of first a ‘natural’ gendered division of labour, second a linked association of women with the domestic and men with the jural sphere, and third a narrowing of wider kinship relations to those of the patriarchal nuclear family concurrent with increased economic specialization, were to remain essential premises in theories of the family for a long time to come. Goode (in Goody 1971) for instance drew on Marxist theory to argue that with the growth of urbanization and industrialization there was a world change in family patterns, as the state increasingly took over kinship functions such as socialization, health and welfare, and the organization of labour and the extended family came to be replaced by the nuclear family. Talcott Parsons, following a positivist, functionalist line of thought, maintained that the nuclear family provided the basis for socialization of children and for the personal development and stability of the adult couple, and identified the ‘isolated nuclear family’ as the ideal family form for the mobile workforce of industrial society.
The conjugal family as building block
Early social theorists defined the family in terms of its functions, and implicitly saw it as a natural form based on a heterosexual conjugal pair and their offspring. Anthropologists developing kinship theories based on cross-cultural analysis, however, were more interested in identifying universals than in developing theories of social evolution and change. For "G.P. Murdock, the nuclear family was a universal feature of society, through which the needs of consumption, socialization, sexuality and labour were met and regulated. Malinowski, writing in rather a different vein about what he termed the ‘principle of legitimacy’ argued that universally the father is indispensable for the full sociological status of the child as well as of its mother, that the group consisting of a woman and her offspring is sociologically incomplete and illegitimate … [and that] … the father, in other words, is necessary for the full legal status of the family.
Malinowski had done his own research among the "matrilineal Trobriand Islanders, and he had argued that in this cultural context motherhood was perceived as a biological relationship, while fatherhood was seen as a purely social one. Thus, his insistence that both parents are necessary for the social legitimacy of the child must be seen in the context of his other claims, quite revolutionary for their time, about the ways in which biological and social kinship may be differentially perceived. Throughout such works, the nuclear or conjugal family, so ideologically central to the Euro-American social order of the time, was taken to be the basic ‘building block’ of kinship organization which, even when elaborated upon in such residential patterns as the extended family or the stem family household, still provided the basis for the regulation of consumption, production, socialization and sexuality.
The conjugal family and the developmental cycle
Meyer Fortes’s work on the developmental cycle of the domestic group shifted the discussion of the family away from function and towards comparison of structures. For Fortes, the minimal unit of kinship was the mother—child dyad, which he saw as ‘demonstrable’ and hence natural. Fatherhood and other kinship relations, it followed, were socially rather than obviously biologically based, and could be organized in a variety of different ways in different cultures. Fortes’s great achievement, however, was to distinguish between the domestic group and the family, and following this to recognize that any static picture of the domestic group was misleading. Following L.H. Morgan and other earlier thinkers, Fortes made a distinction between the domestic and the juro-political domain. The domestic domain was associated with the mother—child dyad, or the ‘matricentral cell’, and the juropolitical domain with wider, more instrumental social and economic relations:
Where the conjugal relationship and patrifiliation are jurally or ritually effective in establishing a child’s jural status, the husband—father becomes a critical link between the matricentral cell and the domestic domain as a whole. In this case the nuclear family may be regarded as the … reproductive nucleus of the domestic domain. It consists of two, and only two, successive generations bound together by the primary dependence of the child on its parents for nurture and love and of the parents on the child as the link between them and their reproductive fulfilment.
The domestic group, on the other hand, is a ‘householding and housekeeping unit organized to provide the material and cultural resources needed to maintain and bring up its members’ Thus, the family consisted of a heterosexual conjugal pair and their offspring, which formed the core but did not limit the parameters of the domestic group. Further, the domestic group went through phases, which involved different residential patterns and different personnel, beginning with the marriage of a woman and man, changing with the birth of their first child, and changing again when that child entered adulthood and set up her or his own household. The juro-political domain was implicitly organized around male relations of exchange, power and authority; the family around the mother and child and affective relations of love and nurture; while the domestic group, a residential unit based on production, reproduction and consumption, mediated the two domains through the husband—father.
While Fortes’s picture added flexibility and structural variation to the notion of the universal family, it still rested on an assumption of a conjugal family based on a heterosexual conjugal unit and, further, it promoted a gendered dichotomy between the natural matricentral family at the core of the domestic domain and the social male domain of jural polity. Variation and difference were added to the model, but the variation was cyclical and to a great extent uniform, as the term ‘developmental cycle’ itself implies. Fortes, like Malinowski, had done path-breaking research in a matrilineal society; Ashanti children received their social identity from their mother and belonged to their mother’s lineage, but their spiritual core, their ntoro, came from their father. For Fortes, as for Malinowski, two parents were therefore necessary for complete social identity. Both Fortes and Malinowski extended the idea of a socially complete person to their definitions of the family: if a person could only be complete with two legitimate parents, then the family should consist of two parents and their children, regardless of the lineage affiliation. Ironically, Fortes’s own work on Ashanti family structure and the developmental cycle of the domestic group showed how transitory and fragile the conjugal family could be, particularly in relation to the more enduring matrilocal extended family. In his later writings, however, he stressed the universality of "bilateral affiliation, and by extension the conjugal family. As Raymond Smith commented, this insistence upon the importance of bilateral affiliation as a moral principle of universal significance serves to reestablish the nuclear family (or something very like it) as the universally necessary matrix for the reproduction of social beings. It also forges a strong link in the theoretical chain between domesticity and kinship, and comes dangerously close to reintroducing the confusion between biology and kinship (Smith 1973: 122).
Even some quite early ethnographic accounts, however, attested to the fact that the conjugal family was not as ubiquitous as the dominant theories maintained. Evans-Pritchard had recorded the existence of ‘woman to woman’ marriages among the patrilineal Nuer, in which a woman could pay bridewealth for another woman, and by so doing attain rights over that woman’s labour and the right to claim any children she bore for her own lineage, just as a man would have the right to claim for his lineage any children born to a woman for whom he had paid bridewealth. While it could be argued that in such a case the woman paying bridewealth was acting as a ‘sociological male’, or that the rights to which she was entitled were closer to those over a servant or slave than a marital partner, the existence of the institution cast doubt on the universal and exclusive nature of the heterosexual conjugal union as the basis of the family. Other examples were even more problematic to the universal model because they represented not a sociological anomaly but the norm. Kathleen Gough’s account of the Nayar, a South Indian matrilineal group in which women underwent a ritual like marriage at puberty and afterwards took a number of lovers who fathered their children, while continuing to live with their brothers, cast doubt on the universality of the ‘fit’ between conjugal family and domestic group. These examples, and others like them, suggested that there was no universal natural family, nor any automatic correlation between family form and domestic group which could be taken as universal. While the assertion that two social parents give the child a complete social identity is probably correct in many if not most cultures, an automatic link between this bilateral affiliation and the practical, lived-in form of the conjugal family is less clear.
A further problematic dimension to the family in anthropological studies was its connection with an evolutionary bias in theory, and an unstated but omnipresent connection with modernity. Thus, non-industrial societies were depicted as having kinship systems, in which, a la Fortes, the family played the part of the basic building block, while Europe and middle-class North America, and at a broader level industrialized societies generally, were seen to have families rather than complex kinship systems. This harks back to the social evolutionists’ idea that, with increasingly complex divisions of labour and the development of formal political institutions, the importance of wider kinship structures such as the lineage diminishes and that of the nuclear family increases. One result of this perspective was that analyses of non-industrial societies concentrated on kinship, while those of industrial societies tended almost to ignore it, and to emphasize instead the conjugal or nuclear family.
The family and the household
This model of the evolution from kinship to nuclear family has been criticized from two quite different perspectives. Those associated with the Cambridge Group for Population Studies, most notably Peter Laslett, questioned the widespread existence of the pre-industrial extended family, and used census materials and other historical data to show that even in ‘past times’ households based on the small conjugal family were often the norm (Laslett and Wall 1972). Mac-Farlane (1978) specifically linked the conjugal family household, strong notions of private rather than corporate property, and an ideology of individualism to pre-industrial England. Other studies, however, such as Anderson (1971) on nineteenth-century family structure in Lancashire, showed that the family structure and household membership were far more flexible than a static census-based picture would imply. At times of crisis or hardship, families pooled resources among households, and kin moved between households, extending the nuclear family to the more complex forms. This was neither a regional nor a nineteenth-century phenomenon; the work of Wilmott and Young (1962) in East London in the 1950s revealed very similar patterns of resource-sharing beyond the boundaries of the household, and of very strong, regular ties between extended families, particularly between mothers and married daughters, and sisters. Grieco’s (1987) research in Scotland and northern England in the 1980s also suggested that the extended family was as important as the conjugal family in patterns of residence, household composition and access to resources during periods of unemployment. This research specifically refuted the Parsonian theory of the isolated nuclear family as the ideal, moveable unit for employment in the modern, industrial state: Grieco’s data show not only that local management recruit through extended kin networks — a syndrome also discussed by Wilmott and Young as the ‘lads of dads’ employment pattern — but that ties of kinship are used to find employment over long distances, to recruit labour from other regions, and to provide housing and other support when labour migration takes place.
An important implication of these studies is that family and household are not the same thing. While they often overlap, it is also frequently the case that households consist of members who are not family, such as servants or lodgers, while family membership, in terms of shared consumption, production and ties of intimacy, often extends over several households. An interesting similarity between the findings of Wilmott and Young and of Grieco is the significance of links between female kin and of matrilateral links generally in the workings of the extended family. In this they resemble what has been labelled the matrifocal family.
The matrifocal family
The matrifocal family occupies a curious position in anthropological writings, sometimes seen as a definite family structure based on a cultural valuing and centrality of the mother, and sometimes as a temporary or ad hoc response to poverty and exclusion. Raymond Smith’s study of family and kinship in Guyana showed that a high proportion of villagers lived in female-headed households, and that often the core of the domestic unit consisted of a woman, her children and her daughters’ children. Arguing that the ‘family is not based on marriage or the nuclear family’ (1973: 137), Smith describes a developmental cycle in which households are established when a man and woman, who may or may not be married and may or may not already have children either with other partners or together, set up house together. While the children of the household are young the woman is most dependent on the man economically, ‘but while men contribute to the support of the household they do not participate very much in child-rearing or spend much time at home’ (p. 124). As the children get older, leave school and begin to earn money or to work on the farm, they contribute more economically to the household. The woman has always been the centre of affective ties in the family; at this point she also becomes ‘the centre of an economic and decisionmaking coalition with her children … whether the husband-father is present or not (p. 125, emphasis added). Sons and daughters begin to engage in sexual relations while still living at home; if these result in children, they are incorporated into the maternal grandmother’s household, resulting in the three-generation ‘matrifocal’ family.
Smith suggests that in this family form the mother—child bond forms the affective and economic core, and the conjugal relationship is neither central nor necessary either to childrearing or to the family itself. He stresses that the term ‘matrifocal’ applies both to households where the husband—father is present, and where he is absent, and links matrifocality to three factors: a separation between the domestic and the poli-tico-jural spheres, and an exclusion of men from domestic tasks and responsibilities; an emphasis on the mother—child and sibling relationship, and an expectation that the conjugal relationship will be ‘less solidary, and less affectively intense’ (p. 141); and stratification and economic factors, notably the absence of property.
Female-centred families attracted the attention not only of anthropologists but also of policy-makers and politicians. In the United States in the 1960s, as racial tension was growing and the Civil Rights Movement was increasingly drawing attention to the poverty in which many Black Americans lived, the ‘dysfunctional’ family was increasingly singled out as the cause of poverty. In the early 1960s Daniel Moyni-han’s now famous report blamed the ‘pathology of the Negro family’ on the absence of men in the family, the high numbers of single-mother families, and the failure of unemployed men to take responsibility for their families and to provide role models for their children. A study by Stack (1974) of kinship and survival strategies among Afro-American families on a housing estate in Detroit showed that rather than being ‘dysfunctional’, female-centred households and kin networks provided reasonable and rational responses to poverty and racial exclusion. Stack argued that the combination of high rates of male unemployment and a national welfare system which denied benefit to women who were cohabitating with male partners favoured the development of female-centred households. In The Flats, the housing estate studied by Stack,high emphasis was placed upon the sharing of both scarce resources and childcare and domestic responsibilities among female kin and friends. Long-term affective relationships between men and children were encouraged and maintained, but the female-based extended family formed the core of economic activity, resource-sharing and consumption. While this type of kin and household interdependence was in many ways contrary to the dominant American ideology of the conjugal nuclear family, Stack showed that rather than being ‘dysfunctional’ it was a highly effective, strategic response to a situation of poverty and exclusion.
Families and ideologies
In studies of kinship and the family, anthropologists are increasingly looking at discrepancies between ideology and lived practice, at differences based on class, region, or race and ethnicity, and at the relationship between the state and the family. Goody’s (1983) historical work on the family in Europe specifically linked changes in marriage rules and in inheritance patterns and dowry payments within families to developments within the church and the state, thereby emphasizing the changing nature of family and property relations in the context of external power structures. Feminist social scientists have examined the relationship between gender and family, arguing that the ideology which represents the conjugal family and the household as ‘natural’ units serves to mask inequalities arising from an association of women with reproduction and men with production, a division which renders a great deal of women’s labour invisible. Similarly, social historians have shown that this association of women with the domestic sphere and men with the politico-jural or productive domain was historically specific, arising in Britain in the nineteenth century with the growing influence of a middle-class morality which promoted government legislation removing children from the labour force, relegating their care to women, and representing women as the core of the household, the ‘haven in a heartless world’ to which men could return after their long day in the brutal world of commerce and waged labour. These theories stress the links between capitalism, state legislation and changes in family ideologies, and emphasize the different ideologies of gender and family which obtained at various specific periods among different classes.
Increasingly, as single-parent families, same-sex unions, cohabitation without marriage, serial monogamy, and households based on friendship rather than sexual partnerships are becoming both more numerous and more visible to the public eye in Euro-American societies, established ideas about the ‘natural’ family are being challenged. Simultaneously, cross-cultural data which stress different types of family and household structure place claims about the universality of the conjugal family in question. While the family will doubtlessly continue to be a major focus of analysis in social science in the foreseeable future, the parameters of the subject are currently being redefined, and the emphasis of enquiry is increasingly on plurality and difference rather than universality.
