Rwezaura et al.

In 1991 the United Nations proclaimed 1994 to be the International Year of the Family (IYF) and further called upon member states to discuss various strategies for implementing the IYF objectives. The most important objectives identified for special attention include the stimulation of local, national and international actions as part of a sustained long-term effort to increase awareness of family issues among governments and the private sector. This was to be done by highlighting the functions of families, the importance of these functions and the problems faced by families in the performance of their roles in the community. Although the UN had in the past made similar proclamations aimed at drawing the attention of the world community to the specific circumstances of certain groups of people, we [page 26] think that a special focus on the family as a whole was not only appropriate but also very timely.

There are many reasons why this focus on the family as an entity is significant. First of all, a number of UN conventions, which have been ratified by states, aim at improving the status of certain members of the family such as women and children. Although these instruments provide a commendable basis for the protection of certain family members, when they are examined together the potential for conflict between them is immediately apparent. Perhaps realising this danger, the UN adopted a motto during the IYF indicating that the world community should aim at "building the smallest democracy at the heart of society."

Another reason why it is essential to focus specifically on the family is the fact that, for a number of reasons elaborated below, the family is today faced with unprecedented challenges. Thus although it is still described in many cultures as the smallest unit of society and therefore meriting protection, the need to provide this protection and support to that unit seems to have been forgotten. The family, everywhere, has always played significant economic and social roles which no other entity is equally suited to perform. However, the 20th century economic and social transformations have not only altered family roles but have also threatened the integrity of the family, especially the African family in its 'traditional' format.

The third reason why we must focus on the family is that, although the term 'family' remains an accurate description of the smallest unit of most societies, it has changed both in size and in the manner of its formation. Hence, we have today a variety of family forms differing radically both from the Western conventional concept and from the African traditional concept, which form the basis for evaluating and regulating individual behaviour at both state and family level. It is therefore timely for us to examine emerging family forms in order to enhance our understanding of the diversity and variety of this unique human institution. This will also assist us to reconsider the ways in which the law and various administrative bodies should deal with the family and, hopefully, to increase international understanding and tolerance of the needs of the family in its many forms.

The aim of this paper is to provide an analysis of the nature of the family and its functions in contemporary Eastern and Southern Africa. In doing this the paper focuses on the relationships between family members and the ways in which these social relationships reflect the underlying social and economic circumstances of this African region. The paper also provides an analysis of the relationship between family, law and the state. This is essential not only because the activities of the state and those of the family are inter-linked but also because the state [page 27] exerts a powerful influence on the family in many important ways. We put special emphasis on the historical transformations which have greatly influenced social relationships in this part of Africa. The changes, in which European colonisation and capitalist penetration of Africa feature prominently, provide a framework for a better understanding of contemporary family relations not only between the family members but also between the state and the family.

The paper is divided into five sections, of which the first is this introduction. In the second part we provide a conceptual framework which aims at elaborating on and putting into context the concepts of family and law which are central to the paper. In the third part we discuss the ways in which the family has become radically transformed both in terms of its formation and structure and also in terms of its functions. In this part we also consider the hardships caused by economic and social changes in society, the insecurity of families which change has brought about, and how all these changes generate tensions and conflicts between family members. It is also argued here that, although certain family members, especially women and young people, have now become more conscious of their individual needs and aspirations, this fact has not been fully recognised.

In the fourth part we explore the ways that the law, defined broadly to include formal and informal law, regulates and defines the family. In the fifth part the economic precariousness of the family is discussed, along with the role of the state in such circumstances. In the final part of the paper, by way of discussion and conclusion, we look at the possible approaches through which the conflicting interests of individual family members can be reconciled with those of the family unit. We also consider how the specific interests of the family unit can be reconciled with those of the state. In performing this task, the paper relies on the concept of democracy, which we believe provides a useful framework for examining many of the issues raised, most of which continue to confront the family in contemporary Africa.

Although the term 'family' is frequently used in academic literature and in statutory provisions, it is in fact rarely defined (Henaghan and Papp 1992; Gwagwa 1991). One possible reason why a definition is avoided is that the concept of family varies from culture to culture and from one historical period to [page 28] another.^{^[3]} We are, therefore, left to reconsider what the term might mean in our own time and culture. Moreover, besides the tantalising variables of time and culture, in order to define the term 'family', and particularly to delineate family membership, one has also to contend with the particular purpose for which the term is being used and the context in which it is being used. The family might be defined in one way for the purpose of inheritance, but in another way for the purpose of day-to-day decisions. One person may seek to define himself as part of the family, while other members of the family do not consider him a part. The answer to the question, 'Is so-and-so a member of your family?', might very well be 'Who wants to know, and why?' Thus, this paper seeks to look at the family, family membership, and the definition of the family in terms of the functions that a family performed in the past and performs today, and particularly the changes in those functions.

Finally, how 'family' is defined depends on whether it is being externally or internally defined: that is, whether individuals are describing the nature and form of the relationships that they consider define the family and its membership, or whether the family is being defined by some external body or agency for its own purposes. Thus, this paper seeks to look at both internal and external definitions of the family, and to explore the significance of these, particularly where they diverge.

To illustrate, in most African cultures the term 'family' is said to refer to a blood relationship arising from common descent. For the purpose of succession, descent may be traced through the mother or the father. In matrilineal societies, where descent is traced through the mother, one's family for the purpose of succession consists of uterine brothers and sisters, and the issue of such sisters (Bentsi-Enchill 1964). In patrilineal societies, where descent is traced through males, it includes a father, brothers, sisters (whether full or half) and the issue of such brothers. The descent-based definition of 'family' is closely linked to the pre-capitalist period and to the subsistence agricultural economy.

This definition for the purpose of succession, however, excludes significant individuals who are also 'family' such as the father and his family in matrilineal societies, the mother and her family in patrilineal societies and one's spouse in both cases. It also excludes grounded changes and reality-based permutations in the conceptualisation of the family in modern times. The problem which arises is [page 29] that the state, especially through its superior courts, may apply only the former conceptualisation of the family, and ignore the latter. For instance, the state may provide that only members of the legally recognized descent group may inherit property, while in practice the beneficiaries of inheritance may be a variable family group, including spouses (WLSA Zimbabwe 1994). The state definition has ignored the meaning of 'family' as internally defined in this specific historical context.

In today's world it is necessary to consider not only kinship ties, but also 'household' or common residence. This allows us to conceive of the family in a way which does not exclude persons connected by blood to other units. Family membership, then, includes both kinship links and residence links, with the latter including members temporarily absent, for instance, for work. As will be discussed later, there is some evidence that residence links are strengthening while kinship links are weakening, at least for some family functions.

Thus the framework through which we will examine the family in this paper is that of family function, looking at the specific cultural context, the specific historical context, and both the internal and external definition of the family.

The term law, like family, is not easily defined. In legal theory the definition of law has been historically contentious, and perhaps for this reason it is rarely defined in works about law. Its meaning is left to be gained from the context in which it is used. This is largely because in order to answer the question 'what is law?' one must ask a further question aimed at identifying the specific purpose for which the term 'law' is to be defined. Where the term 'family' is to be considered within the context of the law, the term law has multiple meanings. For example, the law in its formal regulatory and definitive function 'sets the parameters to what is considered normal' (Smart and Brophy 1985). In other words, law defines the conventionally accepted family forms by giving them legal recognition and by imposing and/or enforcing family obligations. In this way law plays constitutive and regulatory roles by spelling out what social relationships are to be defined as constituting a family: who is to be recognised as a member of that family and what rights and obligations attach to family membership. This approach sees law in terms of its functions.

But the term law can be expressed in other senses. In the second sense law can be taken to mean the agencies which enforce the agreed or imposed obligations between members of a family. Hence, to use common parlance, 'the law says a man must not beat his wife; and if he does, the law will get him!' This sense of [page 30] defining law stresses the machinery of law enforcement which is visible to the community and which expresses the practical manifestation of the law.

A third facet of the term law is the nature of the law, which in the context of power relations within a society involves examining the identity of those who make and enforce the law as well as their economic, social, gender and generational interests. Such individuals influence not only the content of the law in its first sense but also the manner by which the law is interpreted and applied.

These ways of looking at law, however, leave out the informal means by which individual conduct is regulated. In order to identify the forces that regulate, within the family setting, the behaviour of members, and compel compliance with obligations regardless of what the formal law provides, then we must expand the meaning of 'law.' The role and significance of the internal processes within other social fields in re-defining the African family is not a mere polemic. The fact that customs and practices have the potential to be formally presented as alternatives to the pre-existing concepts of the family within the law cannot be overlooked when the term 'law' is considered in the context of this paper.

Thus the role and force of state law as we consider it in this paper is, within the context of the family, hotly contested. There may be constant tension between the perceptions and expectations of those who define themselves as family, consider they have the power to enforce their hierarchical roles and have, according to past notions, an interest in a matter, and of those who perceive themselves as governed by a new system and a new set of rules. The latter will appeal to the formal system for affirmation and protection of their actions. Therefore, to effectively explore this terrain and the ways it affects the family, the legal centralist mode of consideration of the family and its construction in the context of African legal systems needs to be reviewed.

Law must, in the African context, be viewed in its pluralities. It is a mixture of principles, rules and procedures emanating not only from the state but also from non-state institutions such as kinship groups or other semi-autonomous social fields who have regulatory powers in constituting and controlling family relations in the various communities (Moore 1978, 1986).

When the family is considered in the context of African societies as they function on a daily, practical basis, account has to be taken of the various forces that influence the obligations and rights, relationships and affiliations that family membership entails. In such a re-examination, it is essential to recognise not only the plurality of systems of law that function within the formal state courts, but also the variety of norm generating forces that act as regulatory bodies within the [page 31] semi autonomous social fields such as the family and administrative bodies (Moore 1978, 1986).

Thus within the Southern and Eastern African context, law takes on many potential meanings. It is not confined to the so-called state law, but must also, often by virtue of the provisions of state law, take account of the customs and practices of the indigenous population when determining family related issues (Stewart et al. 1990; WLSA Zimbabwe 1994). Thus some investigations, such as the collection of information about practices within 'living law' (Cotterell 1992), might in other legal systems be regarded as generating little more than legal polemics. These in the African context become part and parcel of the formal legal process, as the research results can be directly channelled back into the formal legal system and become part of its formal growth and re-interpretation process (WLSA Zimbabwe 1994).

In this paper the question of what is law is raised in the context of the forces that affect the functioning or disfunction of the family in Eastern and Southern Africa. We are concerned with the inadequacies of the current definitions and conceptualisation of the family in dealing with changing family forms. This concern needs also to be located in an historical context, since the family forms that underlie the application of the law, and the decisions of administrators and independent agencies, are themselves within the Southern and Eastern African context dubious in their origin.

Despite the fact that we cast serious doubt on the hegemony of state law in defining and regulating the family, it is clear that state law, in its various forms, has a profound effect on the way the family organises itself. State law defines the family and in so doing influences the way in which individuals perceive and pursue their relationships. Whether explicitly or implicitly the state law influences the way in which the family is defined. It may define it as marriage-based, with recognition, for some purposes, being given only to formally registered unions. It may define it by biological links, especially where issues of support and relief of the state's potential supportive obligations are at issue. Maintenance, for example, extends beyond these formally recognised links to those who have created relationships by reproduction (Armstrong 1992). However, for other purposes that same biological link will not be recognised. The state in its extension or contraction of the family for purposes of its recognition and the imposition of responsibilities reflects the indeterminacies that society itself has in dealing with the definition of the family.

Thus, despite the existence of other sources of law which provide alternative ways of forming and ordering family relationships, state law constitutes an important, and enduring source of law. State law also provides a wider framework for the [page 32] processing of disputes within the family. Despite the fact that the state legal system has to compete for business with other informal systems at various levels, its role continues to be important for all members of the family. These functions of the formal legal system provide an important opportunity for examining how state policy translates into law and also how the existence of multiple systems of law enables certain individual family members to manipulate these various systems during intra-familial disputes.

In the preceding section we argued that there is a close relationship between the functions performed by the family and the way or ways in which family membership is defined. Although most anthropological literature defines family membership in terms of blood relationship or marriage, studies suggest that in practice there are a number of different factors, corresponding to various functions performed by the family, which influence whether individuals recognise family obligations and family membership. These factors include economic production, reproduction, co-residence, decision-making, and rituals. The changing character of these traditional family functions show the influence of the great changes in the lives of women and men in Africa beginning with colonialism and lasting to the present day, and paint a picture of the family in transition. The changing functions of the family, and the relationship of these functions to the definition and role of the family, will be discussed in this section.

In pre-colonial times the family's economic functions consisted of the production of goods and services which were essential for the survival of family members. Families produced their own food, clothing, utensils, shelter, and other daily needs. Some aspects of production, such as agriculture and small animal husbandry, were performed largely according to family units consisting of a wife and her children, while others, such as house-building, cattle-keeping, hunting and fishing, were usually organised in terms of a larger family unit consisting of a number of related husbands and their wives (in patrilineal systems) or related wives and their husbands (in matrilineal systems).

Gluckman (1969, 1950) argues that the ideology of kinship was a means of translating property relationships into family relationships. Thus, a wife's agricultural work was institutionalised into a wife's service to her husband and his family, or into a mother's support for her own children. In the same way, the claims of males on the services of wives, whether sexual or otherwise, were [page 33] translated into kinship obligations befitting the woman's role as wife and mother. These kinship obligations might be towards the husband and children, or towards the larger family unit, as in ethnic groups where the wife was required to work for her mother-in-law for the first few years of her marriage (Holleman 1952). Women were socialised from infancy to accept these positions and these roles, and benefitted from them as they got older and became senior wives, mothers-in-law and grandmothers with claims on the services of younger females in the family (Collier 1994; Lamphere 1994).

Similarly, a man's agricultural or other productive work was institutionalised into the labour obligations of kinship. The family made claims on the labour time of an individual in a variety of ways in different ethnic groups. For instance, in matrilineal societies, according to tradition, a man had to work for his future wife's parents for several years. In patrilineal societies, a son might be required to provide labour to build new huts for relatives, to clear new ground, and in some societies, to work on the fields of the chief, headman or King (Marwick 1940). The reciprocal exchange of labour and property between family members not only fulfilled their material needs but also performed the important social function of strengthening and reaffirming their kinship ties and mutual obligations.

The role of the child towards its parents and other relatives was also institutionalised through kinship ideology. Children performed productive labour from an early age and were taught to accept and fully embrace their obligations towards their parents, elders and each other. They were taught respect for their elders and hard work. A responsible son was one who respected his parents and worked hard to support them. Similarly, a good daughter was one who obeyed and respected her husband and her in-laws. Thus the training of children aimed to socialise them to accept their position in society and to play a pre-ordained role in the family. They were also taught that the ancestors would reward respectful children and severely punish those who were disrespectful (Armstrong 1995).

The vital economic roles played by women in this setting have been the subject of many accounts of farming systems in Africa. Boserup (1970) and others (Huntington 1975; Davison 1988) show that women in traditional societies are greatly valued for both their productive and their reproductive roles. Women were needed to reproduce and care for children, to care for the elderly, and to train children for the tasks they would perform as adults, as well as to undertake essential productive activities in agriculture, fishing, etc.

Women and children were considered to be resources which men wanted to amass, as illustrated by the fact that in many local languages the term used to describe a man's wealth did not draw any distinction between people and material possessions. For example, the Haya of Northeastern Tanzania use the word [page 34] o'mutungi to describe a rich person and his wealth is called o'muntungo which includes his family and material wealth. Thus studies by some social anthropologists use the concept of 'wealth in people' to describe claims which individuals are permitted to make on other people's time and resources (see Bledsoe 1980: 46; Schneider 1968, 1970).

Land played a pivotal role in the control of kinship obligations. In patrilineal societies land was transmitted through the male line to the men of the lineage. In the case of matrilineal societies, although land was transmitted through females, control was in the hands of men. The means to survival, land, was controlled by either the family of birth or the family of marriage. Thus in the case of a woman, in order to get access to land she had to associate with a man, either as husband (in patrilineal societies) or some male relative (in matrilineal societies).

Aside from control of land, two important traditional institutions served to enforce kinship obligations: bridewealth and spiritual beliefs. Bridewealth, in patrilineal societies, was provided for a young bridegroom by family members, who controlled his selection of a wife and acquired claims on his labour time. Since he needed the support of his family to pay bridewealth for a wife as well as to perform the necessary ceremonies, he remained under their control. Similarly, the bridewealth given for a young bride was acquired by her family of birth, and used to provide wives for the males in the family, giving the family an incentive both to find a husband for the woman and to ensure that the marriage survived. Bridewealth has thus been described as a means of transferring the productive and reproductive powers of a woman to another family and facilitating her replacement in her natal family (Hay and Stichter).

Spiritual beliefs were another way of controlling family members. For example, since belief systems inculcate the conviction that ancestral spirits control a person's fortune and health, widows and divorcees in patrilineal societies have a powerful incentive to maintain good relations with the family of the father of their children, lest the ancestors punish their children with sickness or bad luck (WLSA Zimbabwe 1994; Armstrong 1994). Research has shown that many women will trade off their property rights and other entitlements to avoid confronting their in-laws, to protect their children from, often unspoken, threats of witchcraft (Cheater 1991; WLSA Swaziland 1992).

The most important change in relation to the economic functions of the family has been the incorporation of subsistence economies into cash economies and the world economic system. The incorporation of a family-based system of production into the wider national and international market put new demands on the labour time of family members, introduced new needs for the individual and the family, [page 35] and provided means of economic production which were not controlled by the family.

New demands on the labour time of families originated with the imposition of taxes under colonial policies to acquire labour in European farms or mining establishments (Chanock 1985; Rwezaura 1985). In some countries each family was charged a 'hut tax' payable by cash, which required the family to send some of its members into wage employment in order to pay the tax. Soon, though, new needs for industrial goods and services which could not be produced within the family provided an even stronger pull into wage employment. It became necessary to send younger male members of the family into wage labour in order to purchase agricultural implements, clothes, foreign medicine and building materials, to pay school fees, and to meet the myriad other expenses that began to emerge. Thus, through the coercive measures of the colonial state as well as through market forces, most families in Eastern and Southern Africa became steadily drawn into wider economic relationships.

This pattern had and is having a massive impact on the functions and organisation of the family. First, the movement of people from villages into urban, commercial farming and mining centres caused kinship ties to weaken as they were no longer the only framework for production and distribution of essential goods and services. Individuals had opportunities to acquire necessities through their own individual labour, providing the chance to escape the control of the family. Many of these individuals, realising that they could reduce their total dependence on the family without risking their survival, began to aspire to some degree of individual autonomy and fulfilment of individual as opposed to family needs.

The individualization of needs and aspirations in turn transformed many of the customary practices which underpinned kinship ideology. An important example is marriage and bridewealth. As young males began to earn money outside the family, many began to feel free to make their own decisions about marriage independently of the control of the family, and others established informal sexual liaisons with women in the urban centres without the approval of the family, while still maintaining customary wives in their home villages (Mafeje 1991). The possibility of moving into urban centres also provided opportunities for young men and women to elope without having to undergo traditional ceremonies or to pay bridewealth. The influence of the church provided the possibility of non-customary marriages and an ideology of individual rather than family consent to marriage. Marriage, according to the church, was defined as a union between two individuals and the nuclear family was projected as the ideal family type (Weinrich 1982; Morris and Read 1972). The payment of bridewealth was condemned by missionaries and a few colonial judges as a form of wife-purchase. Other customary practices which reflected the family-based, rather than [page 36] individualised customary marriage, such as widow 'inheritance', and sororate and levirate marriage, were also discouraged by Christian missionaries (Morris and Read 1972; Kirwen 1979; Cheater 1986). Similarly new 'European' laws allowed state-recognised, as opposed to family-recognised, marriage. All of these influences had the effect of narrowing the involvement of the family and the wider community in the process of marriage, and consequently reduced the community's mutual support for married couples. In some cases as a result of these changes in perceptions of the family, the family's power, in so far as it exists, is based on the spiritual elements of life and the possibility that the couple may need support in the future.

The possibility of earning money outside the family also introduced the ability to acquire bridewealth without the involvement of the family, and gave bridewealth a new economic and social character. Today cash has become a major component of most bridewealth transactions. In the past marriage payments in pastoral and semi-pastoral communities comprised mainly livestock. In other communities it consisted of agricultural subsistence goods and services, such as iron hoes and cowrie shells. Today bridewealth in most communities consists of money, industrial consumer goods, or a combination of these and livestock (Cheater 1986: 145; Rwezaura 1985: 70-94; Cutshall 1991: 57-74). As no rules existed about how to dispose of the cash received, the recipients did not distribute it to other relatives, as they would have distributed livestock. Bridewealth become an increasingly individualised transaction.

Second, the entry of family members into wage employment and the commercialisation of agriculture caused the gradual translation of reciprocal labour obligations into money obligations, particularly for people involved in wage labour. Thus in matrilineal societies where, according to tradition, a man had had to work for his future wife's parents for several years, it now became possible to commute the bride-service into cash which was payable either immediately or over a relatively short time (Chanock 1985), and could be earned either through wage labour or by selling agricultural produce. One outcome of this practice was that the wife's family was denied the opportunity to cultivate closer ties with their son-in-law which the institution of bride-service promoted, thus making future understanding between the husband and the wife's family more difficult and again weakening family ties (Rwezaura 1994: 94).

Another outcome was that demands were made on the wage earner which the new economic system could not accommodate. The family of a wage earner, because they felt entitled under custom to his labour time, began to make claims on his wages. But the wages of unskilled workers were set, at best, at levels sufficient barely to support a nuclear family. In most colonial societies they were deliberately kept low in order to discourage workers from living with their wives [page 37] and children in the farms, mines or cities. In some cases workers were specifically prohibited from bringing their wives and children with them. Today, the problem may not be specific prohibitions but the absence of suitable accommodation and the higher costs of urban living. Thus, even though a worker's family members may consider themselves to have a claim on his wages, a conflict emerges because these obligations cannot possibly be met. The wage earner may therefore begin to perceive his primary obligations to lie towards his wife and his children, even though his parents, grandparents, uncles, aunts, brothers and sisters may consider themselves entitled to a share of his wages based on their labour entitlement within the traditional economy. While his labour obligations could be seen as owed to the family unit as a whole, his financial obligations are individualised towards members of his nuclear family. The conflict becomes even more pronounced when he acquires a second, non-customary, wife in town, in addition to his family-approved wife in the rural area. Even greater conflicts may be presaged when he marries the rural wife according to custom and the urban wife according to the general/received law.

Third, the entry of family members into wage employment and the commercialisation of agriculture also affected the land-based power of the family elders, who traditionally distributed land to family members. Under tradition family members were dependent on family elders for land, and thus remained under their control. When land ceased to be the only source of survival, individual family members could afford to break away from the family as long as they had another source of income. Land became something capable of being bought, rather than being accessible only through a family elder. An individual family member could buy his own farm, and produce crops for sale. At the same time, the appropriation of land by colonialists, the mass removal of people from their traditional lands in some countries, and the transformation of land into a commodity, reduced the amount of land owned under customary land tenure and the control of the family over it. Many families either became landless city dwellers (Moore 1991), or found themselves on rented land. With no land to control, elders had lost one of the powers which kept family members dependent on them. In countries where land ownership was individualised, irresponsible family heads could sell land and disappear with the cash, or mortgage the land and see it sold to satisfy the debt. An irresponsible heir appointed to take over land according to the general law provisions might just strip a farm of its assets or sell the property literally under the feet of the dependant family. (On this see the Zimbabwean case Seva v Dudza SC-131-92, WLSA Zimbabwe 1994.) Thus not only were the chances of economic survival for family members reduced, but the territorial cohesion of the family was destroyed as members were forced to move into other areas in search of land or wage employment.

Fourth, the disproportionate entry of young males into wage employment affected the position of women, children, and elders in the family. When kinship ties provided the framework for production and distribution of essential goods, the economic contributions of women and children were valued, and the elderly controlled resources. As money became valued over subsistence agriculture, the production by men in the wage employment or cash crop sectors became valued more highly than the contributions of women and children on the homestead, and the control of the elders was lessened, as discussed above. As the need for money inputs grew with the introduction of better seed varieties, fertilisers, tractors and other agricultural equipment, clothing and school provision for children, the position of the money earner was enhanced and the position of those who did not earn money was lowered. Women became dependent on men, particularly their husbands, increasing the individual power of the husband over the wife in the family (Armstrong 1993a). In place of a system in which the young had been dependent on the old for land, the old now became dependent on the young for cash. A system in which children were once viewed as a resource to be used as labour in the home and fields was transformed into one in which children required cash outlays in order to be fed, clothed and educated.

Research by the Women and the Law in Southern Africa Research Project (WLSA) shows that today there is a struggle over scarce 'family' resources, and that 'family' is sometimes defined to coincide with a person's self-interest in this struggle. The wage-earner may define 'family' as the nuclear family, and perceive his or her primary obligations as being towards a spouse and children, while the unemployed cousin of this same wage-earner may wish to define 'family' to include his uncle's children and thus bring the wage-earner into the circle of family obligations (Armstrong 1992). A mother may recognise a woman to be her son's wife during the son's life-time, but refuse to recognise her as a widow when the time comes to divide up the deceased son's property. (WLSA Botswana 1994)

If one takes as one's point of departure the economic functions and reciprocal obligations of the family in customary law, there is evidence that a working definition of 'family' may have to incorporate the concept of reciprocity. There are indications that the criteria for incurring obligations towards an individual today may be more likely to be related to affection and past assistance than to blood alone. For example, a person may feel more obliged to support the cousin he grew up with and called 'brother' than the true blood brother whom he never knew. The WLSA Lesotho research revealed the concept of lefa lea sekeltsoa which literally translated means 'inheritance is worked for'. This was reflected in a trend to appoint as heir the person who had contributed to the family or parents' welfare and not to follow the notion imbedded in the state law that the eldest son would be the heir.

WLSA research, among others, has found a great variety of working definitions of the family and family obligations (Armstrong 1992). Research on maintenance showed that the 'nuclear family', even when it is the operational unit in terms of residence, is not a financial unit. Women and men, in general, keep their finances separate and divide financial responsibility for the household in gendered ways, with women responsible for food and children and men for larger investments such as the house. Research also found a variety of financial units other than the 'nuclear family', including two sisters living together and pooling resources so that one could care for the children and the other work for wages, and the daughter working in town and supporting her mother who looked after the children.

Residence or organisation into 'households' has often been used as a factor to help in defining the family by sociologists in Africa. Yet there are many people who consider themselves members of the same family who do not reside together. There also those who reside together but would not consider themselves to be members of one family.

In pre-colonial times extended families comprising brothers and their wives (patrilineal) or sisters and their husbands (matrilineal), plus their parents and children, resided together on land passed down from their common ancestors. However, along with wage labour came the phenomenon of migrant labour, urban migration, and the splitting up of the traditional extended family. Molokomme notes, for example, that

as early as 1943...almost half of the young Batswana men between the ages of 15 and 44 years were employed away from their homes, and 28% of all adult Batswana males were absent working in South Africa. (Molokomme 1991)

This pattern continues. A 1982 study in Zimbabwe shows that 82% of the people living in communal lands in Zimbabwe were women and children (Government of Zimbabwe 1982).

Today there is a vast variety of residence patterns. 'Extended families' still live together on land held in common and farm that land. However, these extended families are seldom complete. There are almost always members living apart, doing or looking for paid work. There are men living alone in town while their wives and children live in the rural areas on traditional land. There are women living alone in town with or without their children. There are sisters living [page 40] together, and nieces living with aunts. And, of course, there are also 'nuclear' families.

In addition to the great variety of residence patterns and the (mostly) male migration for labour, the modern African family appears to be characterised by fluctuating residence. Several studies, for example, have examined the fluctuating custody of children. Russell (1989) examines the fluctuating residence of Swazi children, who move from relative to relative to secure economic advantage. Armstrong (1994) identifies a similar phenomenon for children in Zimbabwe, where children change residence for economic reasons, for practical advantages (such as to be near a school), or simply to reinforce family ties. Family membership for these children does not change with the household in which they are living for the time being. Such residence of children is normally linked to the extended family and to family obligations and fluctuates almost exclusively between blood relatives.

Because of the fluctuating living arrangements of many modern Africans, the term 'residence' itself may be misleading. Although it has something to do with family definition and family obligations, it needs re-examination in the context of separations endured by family members in Southern Africa. Armstrong (1995), in support of her argument for re-examination of what is meant by 'living together', cites her interview with a man who worked six days a week in Harare, Zimbabwe, but told her that his daughter "lives with me and my wife...in the Nyanga rural areas" 200 kilometres away. Residence therefore is sometimes defined to mean not where one sleeps and works, but where one's family, or one's ancestors reside.

Disruption of co-residence of family members has not obliterated family obligations between family members who no longer live together. Remittances by wage earners are often a major source of income for the families they leave behind in the rural areas (Murray 1981), although some studies suggest that when men are not physically with their families, they are less likely to contribute to the needs of the family (Standing 1987).

This points to the fact that residence alone is not sufficient to define family membership, and that laws and policies which assume that family members reside together, or that family residence is stable, are misguided. The concepts of the 'female-headed household' is thus called into question, since it assumes that when there is no man in the 'household' it is headed by the resident woman. This ignores the role of non-resident family members such as the husband, father-in-law, brother, or mother-in-law. Whether it is appropriate to seek a single family head in order to define power and categorise units is also another issue that needs to be reconsidered. Similarly, custody laws which favour the western-style nuclear [page 41] family in which a husband, wife and child live together under one roof fail to take account of the reality of residential patterns in Southern and Eastern Africa (Armstrong 1994).

Another function of the family is reproduction of itself: this concept includes both physical reproduction and social reproduction.

The pre-capitalist economy in Africa was characterised by a low-level technology which was sustained by almost total reliance on human labour for the production of essential goods and services. This reliance on human labour in turn underlies Africa's strongly held pro-natalist values, which survive to the present (Nhlapo 1991: 135; Hellum 1993: 243). Labour needs are changing as the economic system changes. Labour-intensive subsistence agriculture is in some areas being replaced by industrialisation, with its lower labour needs. Modern health care has meant that child survival rates have improved. The reproductive function of the family, therefore, is changing.

Population studies throughout Southern and Eastern Africa indicate that physical reproduction is no longer (if it ever was) exclusively within what individuals themselves define as the 'family' (e.g. Letsetedi et al. 1989).

First, they indicate that births to single women appear to have increased. Although Roberts and Comaroff (1977)'s informants told them that in earlier times extra-marital pregnancy in what is now Botswana was rare and considered a disgrace, and babies which resulted were sometimes put to death, in modern times some studies find as many as 50% of women in Botswana are single mothers (Brown 1984).

Second, adulterine births also appear to be on the increase. Because migrant labour causes husbands and wives to be separated for long periods of time, they are tempted to form extra-marital relationships, from which children may be born (Brown 1984).

Some of the statistical surveys presenting this data are suspect due to the definition of 'marriage' which is used. That issue is taken up in the discussion later in this paper of the conflict between customarily defined marriage and marriage as defined by the State. Nevertheless it is clear that there is something new happening here. While reproduction was once controlled by families through the institutions of marriage and bridewealth, today families have much less control. Men and women can decide to have children without the involvement of the family, by having children outside relationships sanctioned by the family, and can also make independent, individual decisions to use contraceptives and limit their families. Some studies also indicate that some women are choosing to have [page 42] children without forming a family with a man in order to increase their independence and retain their legal rights to the child (WLSA Botswana 1992; Hellum 1993).

Similarly some aspects of social reproduction, such as the education of children, have largely been taken out of the hands of the family and are now controlled by the State. In pre-colonial times children learned the skills they needed to survive from their parents and other relatives in their family. Today, those parents and other relatives such as brothers and sisters contribute to the education of children by paying school fees.

Decision-making is another family function which is in flux. The simple fact of being physically separated means that decisions which may once have been made by the family group, now must be made either by a reduced group, or by an individual (WLSA Lesotho 1992).

The inter-dependence of family members was a part of traditional life. The family members lived together, tilling the same fields or adjacent fields, participating in rituals together, assisting each other in childbirth, sharing food and labour. In patrilineal societies, these were usually relatives joined by blood descent from a common male ancestor, and in matrilineal societies from a common female ancestor.

This inter-dependence made it necessary for family members to make decisions as a group and to reach agreement on issues in ways which preserved the peace. Negotiation, therefore, was made easier because all parties needed a solution which allowed them to continue to live with and be dependent on each other. Consensus of values and principles was easier as people shared a common way of life and common perspectives on life. The family had a variety of "legal" functions as a semi-autonomous social field making and enforcing its own laws and settling disputes within the family and representing the family's interests in the community.

As discussed above, in most of Africa the family members who once lived together and tilled the land together no longer do so on a day-to-day basis. These far-flung family members are not dependent on each other in the same way as formerly. This has weakened the power of the family to make joint decisions and to produce negotiated settlements of disputes within the family, and to enforce these settlements. A family member who does not want to conform can merely leave. It also destroys the consensus of values upon which the traditional law was built. Family members who travel to far-off places encounter different ideas and values and make decisions at family level even more difficult.

Decision-making by the family was based on hierarchies of gender, age and position. In most decisions, men took the forefront. In some societies, such as the Tswana, women were not supposed to attend community meetings (kgotla). A woman needed a man to acquire access to resources, although this practice may have been, and clearly is today manipulated by women (Rose 1988). Older men and women had authority over younger men and women, based in part on the power of the older generation to distribute resources, including land.

Access to wage employment has meant that younger family members, and women as well as men (although usually to a lesser extent) have the power that comes with money. With regard to access to resources, this has reversed dependencies, with the old often now dependent on the young for resources rather than the other way around, weakening age hierarchies. It has also meant that women may have access to resources independently of men, weakening gender hierarchies. Again, this means that the decision-making function of the family has changed. The old have trouble enforcing their decisions on the young, upon whom they may depend. Men have trouble enforcing their decisions on women who have independent access to economic resources. Family members not living together may have less influence on each other's lives than family members who live together. This may mean, for instance, that a woman has more influence in her 'nuclear' family, if that is how she is living, than in her family of birth, if they are living scattered throughout the country.

However, while age, gender and position hierarchies are weakening, they are reinforced by ideology and by spiritual beliefs. Respect for elders is still one of the key values in southern Africa, and is seen particularly in the way that men and women take seriously their obligation to support their parents financially. Spiritual belief systems also point to respect for elders, both living and dead, as the key to happiness, success and good luck. Similarly, male dominance is still a key part of customary ideology in southern Africa. The conflict between ideology and practice produces tensions, particularly between the old and young.

The power of the family in the traditional setting was also dependent on the recognition of its power by the community. When communities were organised under chiefs, the family's power to arbitrate most disputes and to speak for its members was delegated by the chief. Today, the 'community' also includes the State. With modern transportation and mobility, a dispute which once had to be resolved within a confined geographic and legal sphere, now may move to other geographic and legal spheres. Most importantly, a legal issue may be moved out of the semi-autonomous sphere of the family, and into the court, the administrative machinery of the state or simply out of the reach of the power of the family and the community.

For instance, while in the past one might have needed the consent of the family to get married, today a couple may contract a civil marriage without regard to the family. Even if the marriage is not recognised by the couple's family, it is recognised by the larger community, the State. In the past, if a couple had a problem, they were supposed to discuss it within the family. Today, they may merely approach a court for a divorce. In the past, questions of custody of children were decided by the family alone. Today, a parent can apply to a state court to determine the issue of custody. The pluralism of the systems of law under which people operate today has the effect of weakening the power of the family to make and enforce law, as its members have other options.

Although the decision-making function of the family is changing, it has not disappeared altogether. WLSA research shows that families still control some decisions, particularly with regard to issues such as inheritance, and that the family may interact with other institutions such as the traditional authorities (chiefs, headmen, traditional courts) or the State legal system. (WLSA Swaziland 1995)

The formality of family meetings and collective decisions seems to be lessening, and individually made decisions appear commoner (WLSA Lesotho 1992). This is probably necessitated by modern conditions under which it is difficult to bring together family members who may have relocated all over the country and even in neighbouring countries for work. It is a consequence also of the reduced authority of traditional family leaders, who may now be dependent for support on the very family members who were once dependent on them. The informality of decisions may allow more opportunities for individual family members to manipulate situations to their own advantage without the mediating role of others to control them. Thus a handful of greedy family members in Zambia may purport to act for 'the family' by taking a widow's property for their own use, in spite of the fact that other family members condemn their actions (WLSA Zambia 1994).

Finally, another function of the family is to perform rituals and other customary ceremonies. Again, this family function is in flux. The introduction of Christianity and modern beliefs has meant that, for some people, many of these rituals have been abandoned or transformed. People now marry in church, are buried with Christian services, go to hospitals and clinics. However, traditional ceremonies are often performed alongside the modern ones, and for these traditional ceremonies the family is still in control. Many people marry both at church and according to traditional rites (Nhlapo 1992). They go to both clinics and traditional healers (for whom many rituals must be performed by the family). They have life insurance and pension policies. They may make wills favouring spouses, but also ensure that some items of property are distributed in accordance with customary practices when someone dies (WLSA Zimbabwe 1994).

In conclusion, an examination of the functions of the family in modern Africa shows why it is so hard to 'define' the family. Functions which were in the past performed exclusively by the family, and which served as a way of defining family membership and mutual obligations, may be today performed in different ways (as in the case of economic production, and residence), or outside the family (as reproduction), or alongside other, non-family alternatives (as rituals, customs). The changing modes of production, belief systems, residence patterns and technology now circulating in Africa have meant that the family itself is transforming, re-fitting itself to deal with changes in the society around it.

It is to be expected from our foregoing discussion that social relations were undergoing qualitative transformation not only at the family level but also more widely. What we have chosen to characterise as the new family law is a 'distilled' product based on an analysis of the interaction between, on the one hand, the colonial state legal system, which through the courts intervened in disputes between family members, and, on the other hand, the customary law and the received laws which were in various degrees applicable to these disputes. Mediating between this interaction is the new money economy, state administration and the missionaries which, as we have noted, provided the spark for some of the disputes and social transformations.

It is not possible to deal with all the major causes which led to the emergence of a new family law in Eastern and Southern Africa. Only two will be considered here. The first is the creation of a colonial legal system with courts and a local administration which had power to hear disputes between family members. These new structures provided an additional forum for family members to bargain over various entitlements. The second is the introduction of new ways of establishing family relationships through church and civil ceremonies. These not only expanded people's choices but also introduced them to new values such as monogamy and new rules about family property and the parent and child relationship. These two aspects seem to have played an important part in the emergence and growth of a new family order. Even for these we can only provide an outline in the hope that specific case studies might be made of other forces and to encourage in-depth studies of some of the issues covered in this paper. We shall seek to identify common themes which seem to run through the colonial history and beyond.

The first is the alteration of the notion of marriage as a union of families into a relationship between two parties. This was the premise of monogamy promoted by the Christian churches, and to a lesser extent by the colonial state through the [page 46] introduction of the civil form of marriage. Christian and civil forms of marriage also implied other related principles including the notion of spousal consent to marriage, the husband's obligation to maintain his wife and children, and the right of the wife to divorce her husband if she was not happy with the marriage relationship.

In most countries in the region the colonial state introduced its own marriage and divorce laws. As noted above, these new forms were at first intended to cater for Europeans living in these countries but were later extended to Africans, and in particular those who converted to Christianity. These new marriage laws, however, did not abolish or seek to replace customary forms of marriage. On the contrary, the two systems operated side by side within the same legal system. This is what has come to be known as the formal 'plurality' of laws.^{^[4]} As we noted earlier, the plural or dual system of laws came into being because western law was received into these states while at the same time limited recognition was granted by the colonial authority to existing indigenous systems of law and religious laws (Woodman 1989). This process resulted in something of a legal cocktail subsumed under a single legal system. Judges who were empowered to administer all these systems had to perform the complex task of determining the proper law applicable to a given transaction or dispute, particularly in the area of private law. (Allott 1970: 112, aptly describes the field as 'internal conflicts of laws', partly to distinguish it from, but also to compare it with the subject of private international law.)

As would be expected, the majority of people in relation to whom the indigenous laws were applicable were Africans while Europeans and other 'non-natives' were subject to the received law. Some Africans who converted to the Christian or Islamic faiths also acquired an additional system of law, besides African customary law, which could be applied to them in matters relating to their personal law.^{^[5]} Thus because of social and economic transformation, including the [page 47] influence of Christianity as well as the growth of urban centres, a number of Africans became partially incorporated into the 'western' system. Hence, although they remained largely traditional in outlook, they occasionally engaged in transactions which were governed by the western law. These include cases where an individual would contract a Christian monogamous marriage or would opt to transfer his property on death by testamentary disposition under the received law (Poulter 1981; Nhlapo 1987).

The colonial state was interested in the creation, preservation or fusion of a system or systems that minimised state responsibility for day to day welfare of the individual, but facilitated the imposition of standards and values that regulated the behaviour of the individual to the maximum benefit of the state. This technique of indirect control ostensibly left the indigenous population to regulate their own lives. It meant that the colonial state seemed to be supportive of the institutions already in existence. Yet it was based on the desire to impose the main tenets of its own value systems. Capitalism and Christian values increasingly put in place alternative mechanisms that had the effect, with time, of infiltrating and at times breaking down the structures of pre-colonial society.

The fascinating aspect of this historical development, which forms the centrepiece of this section, is the fact that many people did not confine their actions within one system of law. They tended to draw from all the various systems whatever suited them best. Perhaps no one should have expected otherwise. A man, for example, might contract a Christian monogamous marriage and yet several years later contract an additional customary marriage without first seeking a legally recognised divorce. The result was that, contrary to the couple's expectations, the second marriage would be void under the general law and the resulting children would be illegitimate.

As much of the WLSA Inheritance research has revealed (WLSA 1995), problems created by the juxtaposition of conflicting marital forms arise at points of crisis in the family, particularly at a man's death. At this point the various forms of law, state law, customs and practices, as well as the normative orders within the semi-autonomous social fields, become as it were competitors. The state courts would, in fact could not recognise that the latter marriage was valid. But this would not prevent the family of the deceased male strategically planning, if they considered it to be in their best interests, to give recognition to the latter wife. This would be particularly pronounced if this kind of tactic might be used to gain them a financial advantage. The widow from the first marriage would often respond by [page 48] using state law to protect her interests and those of her children. Formal determinations might be made in her favour, but whether all their provisions were fully and faithfully implemented, and all the property of the deceased made available to her was another matter (WLSA Zimbabwe 1994; WLSA Zambia 1994).

What might be somewhat surprising to outsiders, but quite normal in our region, is the fact that a man would go through two ceremonies, one under custom and a later one in church, with the same woman. However, as Nhlapo observes:

A marriage cannot be both monogamous and polygamous at the same time. It can include or exclude the marital power of the husband over the wife, but it cannot do both. (Nhlapo 1992)

In some countries, should the man die without leaving a will, the courts had to determine whether his way of life was such that he considered himself a member of his customary community. If so, customary law would apply; if not, the general law would apply.^{^[6]} This is indeed what happened in the famous Kenya case of S.M. Otieno, in which his widow and Otieno's clan fought bitterly over the right to bury his dead body (Virginia Wambui Otieno v Joash Ochieng Ougo & Anor., Court of Appeal, Civ. App. No. 31 of 1987; Cotran 1987: 331-345.). On the other hand certain crafty individuals took advantage of this plural system of laws to dispute their obligations to their customary law wives by claiming that they were not legally married to them. Very often the more experienced players in [page 49] this choice of law game victimized the less experienced, especially the women and their children (Rwezaura 1994-95).

Another remarkable feature of this development is that these various systems of law provided different remedies in respect of comparable relationships. Courts needed to establish which form of marriage the parties had contracted in order to determine, first, whether there was jurisdiction to hear a dispute and, second, the appropriate legal remedies. For example, a woman married under customary law or Islamic law could be divorced by the husband extra-judicially, a procedure which would block the state courts from determining matters such as the reallocation of family property or the custody of children. It was rightly thought, therefore, that the pluralistic legal system was discriminatory because it tolerated the existence of double standards in the administration of family law. It is from this angle that most of the more recent family law reform efforts have been undertaken (Rwezaura and Wanitzek 1988; Rwezaura 1990, Ncube 1991, 1990, 1987).^{^[7]}

But this criticism is sustainable only if we believe that one system of law provided better remedies than the other, or indeed that state courts were better forums than those available under traditional arrangements. From the perspective of most people, however, the plurality of laws, like the plurality within the economy and the plurality in social relations, was part of a new era in which individuals had a wider choice and could sometimes play one system against the other. What cannot be denied, and this will be stressed later, is the fact that these choices, whether in the economy, or in the forms of marriage, were not equally available to women and junior men within the family. Research has consistently shown that, more often than not, family heads had an advantage over the other family members. This advantage has not been substantially removed. We shall propose that the concept of democracy be applied to suggest a possible solution to this problem.

What all this plurality of laws and remedies seems to have produced is what we call the new family law. But this new family law did not come without some initiative from family members. As we have noted, much of this initiative was provided by the changing economy, the changing social relationships, and the disputes which these changes generated. Some of these disputes found their way into the state courts and were recorded. We cannot provide any useful account of the substance of these disputes. Here we can only describe the disputes as having been engendered by change and by a desire for accommodation of the realities of change. They occurred under circumstances where the elders and the junior men and women bargained for various entitlements. This bargaining, which was conducted under the shadow of the colonial state, did not terminate with the end of the colonial era but continued into the post colonial period.

We can now sum up the foregoing pattern of inter-familial bargaining in this way: Women and junior men tried to maximise the opportunities presented by change to gain greater individual autonomy while at the same time seeking to minimise the constraints and economic hardships brought about by change. In the case of family heads, their main effort went into trying to enhance their power using tradition and kinship obligations which were owed to them by the junior men and women. An interesting feature of this process however, is that neither the elders nor the juniors were opposed to the opportunities which tradition or change made available to them. Depending on the circumstances, each side would deploy an argument either from within custom or from some other source such as the general law to press a case.

Hence, when we look closely at these developments we see emerging a pattern which explains the rather contradictory positions sometimes taken by these actors. First, it is clear that because the elders had a greater political stake in the traditional system and a better understanding of it, they tended to support traditional claims more than their juniors. Elders also engaged in the manipulation of fact situations to bring the matter within the traditional ambit. And whenever it suited them, elders used tradition to gain certain advantages opened up by change (Rwezaura 1985). For example, as we have noted, elders had no difficulty in translating the labour obligations of their migrant sons or prospective sons-in-law into cash (Murray 1976, 1977). Equally elders were the first to step forward as the sole owners of family land when land title registration laws enabled them to register individual title (Okoth Ogendo 1982, Davison 1988).

The use of ritual power and its manipulation by the elders also has become a potent bargaining chip in this region. For example, a divorced couple can be confronted with a conflict of norms regarding custody of their children after divorce. On the one hand the formal law calls for application of the principle of the 'best interests of the child'; while the customary law is based on the principle [page 51] that a child belongs to his father's lineage in patrilineal systems (Armstrong 1994). Although, at the formal level, a divorced woman is more likely than the father to receive custody, the belief system supports the opposite principle. Thus, a woman is either told, or believes without being told, that if she insists on having custody the rest of her ex-husband's family would refuse to perform traditional ceremonies necessary if the child falls ill, or for a marriage, or for burial.

Women who have economic power, or support from other family members, may find ways to retain their bonding with their husband's families with new formulas to replace the old. For instance, in Zimbabwe, widows would choose a sister of their deceased husband to be their new 'husband' during kurova guva or umbuyiso ceremonies. These take place approximately one year after the death of the deceased and are the time when according to local custom the estate is distributed and the widow may remarry into her husband's family. Modern approaches favour looser and more pragmatic relationships that serve as bonds. Some widows choose their son as a 'husband', indicating so by presenting him with the symbols of his father's authority.

Of course there are some widows who, for strategic reasons, decide to remarry in the more formal sense into their husband's families and select one of their husband's brothers as a new husband. Other widows and widowers, when faced with harassment, may abandon clear legal claims in favour of the restoration of some kind of harmony between themselves and their former in-laws. This is often based on the belief that harmonious relations are in the best interests of the children. (WLSA Zimbabwe 1994) What individuals are able to do in such situations is to exploit the intersections between the customs and practices and modern economic realities and personal empowerment opportunities.

The strategies of women and younger men may have a slightly different focus than those of other men within the family. It can be argued that, since women and younger men perceived themselves as having more obligations within the traditional system, they were more attracted to the freedom which change could offer. Moreover, they were also young, strong and adventurous and more suited to meet the challenges of the external world. Once they had earned their cash, most young males were not willing to give it all over to their parents. They perceived it as their personal incomes. Whenever possible, these young men were able to purchase land or select wives and pay for their marriages. But having done so, these new husbands would activate tradition to require their wives to perform wifely obligations and perhaps even attempt to control their incomes from the sale of fruits and vegetables or the other products of the woman's labours.^{^[8]}

Yet despite these efforts towards individual autonomy, migrant men also needed to maintain links with the lineage in order to retain their inheritance rights. This aspect made them amenable to traditional demands which, grudgingly or otherwise, they acceded to, in the hope of greater economic benefits and autonomy. For example, a woman would divorce her husband in order to gain freedom but she would subsequently use tradition to assert her kinship claims for economic support from her sons and daughters-in-law (Collier 1974; Bledsoe 1980).

We should conclude this section by noting two points. The first is that although recent accounts of colonial legal history point to the emergence of a new customary law which was more favourable to men than to junior men and women, we should never overlook the fact that both men and women used the colonial legal system in the process of intra-familial bargaining and this resulted in some benefits on both sides (Mbilinyi 1988; Rwezaura 1985). This is, of course, not to deny the fact that male elders had an edge over women and male juniors, and therefore had more bargaining 'chips'. It is equally important to see women as actors, striving for economic and political advantages. These are actors who occasionally use both the traditional and the modern system to gain favourable results, i.e. to increase their opportunities and widen their choices, but also, where possible, to minimise or redefine their obligations (WLSA Zimbabwe 1994). Women are, therefore, not to be seen as passive victims of a male dominated political system but perhaps as disadvantaged participants who sometimes use other women to attain their political goals (Rose 1988; Comaroff and Roberts 1981; Bledsoe 1980: 179 Rwezaura 1985: 179). What seems significant, however, to the practice of inter-familial bargaining in the Eastern and Southern Africa region is that these inter-familial negotiations and struggles are best understood if located in the economic reality of the region and it is to this that we now turn.

The notion of subordinate family members (in which also women are included), bargaining with male elders (the family heads) over a range of family entitlements and claims may not have sufficiently stressed the economic and social precariousness of families in contemporary Africa. We feel that this reality must be stressed in order to remove any impression that disputes between family members might be motivated by greed, presumably under circumstances where there is plenty for everyone if only a fair system of sharing existed. The more realistic picture however, is different. It is one of scarcity of essential resources for the majority of families in this region. In such families shortage of resources can reach desperate levels where claims for family/kinship support may be the only alternative available to avert a crisis.

Much has been written about the declining role of the extended family in providing economic security for its individual members in modern Africa. This literature cannot be considered here. However, common themes can be drawn from research done in this region. For example, Rwezaura's study examines a link between the increasing economic insecurity of women (both rural and urban) and their inclination to find alternative security through both traditional and non-traditional avenues (Rwezaura 1985, 1988; see also Bryceson and Mbilinyi 1978). Today many families, especially in the rural areas, cannot fulfill their essential obligations without cash remittances from relatives employed in the urban centres. There are known cases where a child may not get money for school fees unless a capable relative volunteers to meet these costs, which, in some cases, might involve taking over the child as a ward (Armstrong 1993b). Comparable situations exist whereby a daughter from a poor family might be withdrawn from school and given away as a bride so that her bridewealth can be applied to obtain a wife for her 'overgrown' wifeless brother or to discharge an outstanding family debt. Thus the notion of an economically precarious family provides, in our view, the more accurate view of the context in which family members bargain over scarce resources and the new family law is reproduced.

The more recent context of economic and political crises that characterise life in Africa are too well known and indeed are still with us. We shall therefore not attempt to discuss them here (but see Isaacman 1990). Although some families have been hit much more than others, there are common features from which can be drawn helpful analyses and conclusions. The first is that the individualisation of support obligations among family members has created serious economic risk for most families. And within these families, vulnerable members such as children and the elderly seem to have suffered the most. Studies have shown, for example, that one of the reasons why reforms in family law face strong resistance from elders and their supporters is the fact that such laws have an adverse economic [page 54] consequence on the elders (Rwezaura 1988, Ncube 1987, Himonga 1987, Griffiths 1987).^{^[9]} This is because these laws are seen as empowering the young to escape from traditional obligations of support to the elders and thus as tending to undermine the economic security of the older generations.

Besides noting the economic risks faced by elders it is important also to think of cases where the economic welfare of the family is in the hands of a single person such as a husband. In this situation not only is the husband too powerful in the home but also there are additional risks for the whole family of economic deprivation in the event of premature death or debilitating sickness of that individual. Cases of 'property grabbing', where the dead man's estate is snatched by his rural relatives leaving the widow and her children economically destitute, must be viewed in that context (Armstrong 1992, Himonga 1987).

Often it will be argued that the solution may lie in every husband or perceived head of family ensuring that he leaves a valid will. But there is often great reluctance on the part of men to write wills because they fear that preparing a will might hasten one's death (Rwezaura 1994-95). Nor is there any guarantee that the will would favour his nuclear family (WLSA Zimbabwe 1994). If a will is eventually written its validity may be legally or socially contested by relatives who feel entitled under custom to an inheritance denied. Perhaps the solution does not lie in wills and adjusting the laws of inheritance, but in widening the ownership base to accommodate the real contributions of the other family members, which in many cases go unrecognised both in law and in fact.

The concentration of a family's economic welfare in one person has other negative consequences. There are cases where the family head, being the sole breadwinner, loses his job, or where because of high inflation his income ceases to meet the basic needs of the family. The contemporary illustration of this problem is to be found in the negative economic consequences caused by the structural adjustment programmes unleashed on Africa by the World Bank and the IMF. The cut-backs on public expenditure and state subsidies have led to soaring prices of basic consumer goods and essential services. The cost of food, medicine, education and transport have risen beyond the means of most families. As would be expected, [page 55] the low-income families have been hit harder by the lack of state subsidies. Also massive retrenchment of public servants has taken its toll on wage-earners.

The absence of state welfare benefits for disability, sickness or unemployment in these countries means that the family has to go without support. Thus studies made in urban centres have shown that a increasing numbers of women have entered the informal sector in order to supplement to the family income. In doing this, many mothers have sought and obtained the help of their children. In her research in Tanzania, Aili Trip found that the economic role of women had changed drastically over the last two decades. Whereas surveys conducted in Dar es Salaam in the 1970s showed that a large percentage of women (i.e. 66%) had no source of income, surveys conducted in the same city ten years later showed a remarkable change. In a Dar es Salaam survey conducted in 1987-1988 Trip found that many of the surveyed women (i.e. 66%) had income-generating projects in which they sold food-stuffs, charcoal, secondhand clothes and local beverages:

[A]s many as 61% of self-employed women said that they were helped by their children and according to teachers and principals, small family enterprises had become a single most important reason for poor school attendance and unsatisfactory progress in the classroom. (Trip 1992: 7)

The engagement of children in the economic undertakings of the family, however detrimental to their school work, is often viewed as an extension of the traditional practice whereby children were active participants in the subsistence economy. Yet unlike the traditional era where future career and training were synchronised, the extracurricular activities of the pupils clearly conflict with the objectives of formal education. And yet, on the other hand, these activities are in many cases vital for the economic survival of the family. But although the contribution of women and children to the household budget is vital for family survival, these young producers, like their foremothers, are often barred by reference to custom from asserting their right to make vital decisions, which are regarded as being reserved to the family head.^{^[10]}

Another problem facing families in this African region has resulted from civil wars, some of which have been instigated externally. The Mozambican and Angolan wars are the best examples. There are also countries still suffering from the effects of wars, such as Uganda and to some extent Tanzania. And more recently, the Rwandan civil war has created an unprecedented refugee crisis, quite [page 56] apart from the total devastation at home, which countries in the region will have to cope with for a long time to come. There are also reliable indications that the number of dead bodies dumped into Lake Victoria has triggered a serious environmental problem for the neighbouring states. The extent of water pollution will affect the supply of low-cost fish which are an essential component of family diet.^{^[11]} Thus the effect of the war on families includes not only the loss of breadwinners but also the disintegration of the family and additional strains on even seemingly unaffected families' budgets.

Other natural disasters such as drought have had a similar effect on the family in Africa. More generally, there has been increased environmental degradation, caused by poor farming methods, deforestation, desertification, and the concentration of population in particular areas of the country, accompanied by the depletion of livestock. All these factors have negatively affected the economic base of rural families. At the national level states have had to import food instead of purchasing essential inputs for development.

It is clear from this discussion that the task of supporting family members has become increasingly difficult within the context of a cash economy in which wages are just enough to support only one or two members of the family. Despite this, the state continues to use the ideal of the 'traditional, extended' family, which supports its members economically, physically and emotionally, to justify its neglect of those functions. Thus the African family is still expected to care for its elderly and this in turn justifies the failure to provide social services for the sick and elderly. The family is expected to care for its children and this argument is used to justify the lack of provision for child care facilities, or orphanages. Similarly, the family is expected to open its arms to accept and care for retrenched workers from the urban sector and those who are retired, and hence little or no severance pay is given to them. Wages, as we have noted, can be kept low because each African family is assumed to have land on which to grow food for the family.

Within this context we would contend that the question, whether the state has the resources to provide for all vulnerable members of the family, is at this stage not material. The important question of principle and political responsibility, is that [page 57] the state does not at the moment see itself as being under a moral or legal obligation to provide these services. There is ample justification for insisting that the state and all those who benefit from the cheap labour of industrial workers and peasants must make adequate provision for them, and cannot instead hide under the obsolete system of family reciprocal support which existed during the pre-capitalist era, since it is now utterly transformed and transcended. Indeed with the AIDS pandemic ravaging its way through the region like a bush fire, it is difficult to see how families will be able to cope with all these disasters. The effect of this on the family is that the competition for scarce resources enters naked (i.e. without shame) into the family itself. What may once have been a unit that supported and sustained its members, has now become a unit in which members compete for survival.

Perhaps to demonstrate that scarcity of resources should not be taken as a reason to exempt family heads from their legal obligation towards the family, in 1990 WLSA began a research project on maintenance laws and practices. The objective was to study those existing laws in the region which impose an obligation on spouses to maintain one another, as well as their children, to find out how these laws are applied in practice and to suggest appropriate changes. Two important points were noted by WLSA. The first was that "since it is usually men who have most access to resources ... maintenance becomes a struggle within families over cash". The second point was that it was necessary to seek strategies with a dual focus: both to increase family income levels and to distribute scarce resources equitably within families (Armstrong 1992).

We continue to support the view that, although scarcity of resources is a serious problem in our region, much can be done to establish and promote equitable property relations within the family. We believe that this will lead not only to better and more efficient distribution of that which is available but also will form a concrete foundation for promoting improved family income and greater security for every family member. In the next section we consider how a beginning can be made by applying the concept of democracy.

We stated in the introduction to this paper that one of the objects of the IYF was to promote strategies for building the smallest democracy at the heart of society. We suggested that the notion of democracy provided a useful conceptual framework for analysing the conflicting interests not only of individuals within the family, but also between the state and the family. We also think that some attempt can be made to take the analysis a step higher by looking at how the same concept can be applied to examine inter-state relations. Also in this section we shall [page 58] consider the role of state law in facilitating the process of building democracy at the family and national levels. We shall argue that a balance has to be struck between two competing political values; namely, the autonomy of the family on the one hand, and the need, on the other hand, for state intervention to promote the concept of democracy within the family. These issues are considered in this final section of the paper.

By democracy we do not refer to the system by which citizens of a given polity determine by whom and how they are to be governed. We are using the concept of 'democracy' to refer to conditions existing at three levels, i.e. the family, the state and the inter-state levels. These conditions either promote or frustrate the following three elements, namely, participation in decision-making, a fair sharing of responsibility and resources, and a recognition of individual autonomy. These three elements are central to the recommendation of the United Nations that during the IYF and beyond, member states should be committed to "building the smallest democracy at the heart of society"(UN 1991).

There is a sense in which the concept of democracy as defined above can be used to analyze and evaluate past, present and future relationships, on the one hand between individuals within the African family unit, and on the other hand between the family unit and the state. At another level, this concept of democracy can be used also to evaluate the relationship between states as international actors. Inter-state relationships fall into the political, economic or cultural fields and these relationships are crucial in providing us with a complete and realistic picture of the major players in the field and the factors which either determine or greatly influence the outcome of the game.

To consider the term 'democracy' adequately, it is necessary to look beyond the form in which decisions are made. We must look to the underlying power relationships among those who make decisions regarding the allocation of resources and responsibilities, and regarding the regulation of different forms of individual autonomy at these levels. It is at these three levels also that ideology, in its various forms including traditional values, religious beliefs, international aid and free trade, plays an important role in generating the required legitimacy and rationalisation of resource allocation and distribution of political power.

This framework enables us to challenge systematically power hierarchies, and statutory provisions and administrative or judicial decisions which either conflict or fail to pass the criteria of democracy. The framework also operates as a powerful political tool for demanding the elimination of certain economic and [page 59] social structures or arrangements which are exploitative or inequitable and which work against the three essential elements of this notion of democracy.

We have noted that the concept of democracy can be envisaged to include three elements: participation in decision-making, sharing of responsibility and resources, and recognition of individual autonomy. These three elements are applicable at each of the three different levels. The concept of democracy therefore implies the difficult dimension of autonomy at all the three levels. In the same way that the rule of the majority at the state level should be limited by certain parameters within which decisions may be made, often by a Bill of Rights, so there are limits to democracy at the family level.

The family should not be free, for instance, to engage in acts which injure its weaker members. We have seen, for example, how school children are involved in the economic activities of the family at the expense of their formal education goals. We have also discussed the removal of a girl child from school to facilitate the payment of a family debt or even the allocation of family resources and labour time favouring men and male children. The custom of homicide brides in Zimbabwe also falls in this category. By this custom a young woman is given away by her own family to compensate for the death of a member of another family killed by someone in her family. Thus, it might be argued that the state must intervene in the family to protect a child who is being abused, exploited or pledged. The more difficult question, however, is the extent to which the State may intervene in the family to create and police the concept of democracy as we have just described it. If democracy requires that hierarchies of age and gender be broken down, is the State justified in entering into the 'smallest democracy' to help break them down?

The same questions apply at the level of the State itself. It is necessary to balance the objective of democracy at the State level with the objective of democracy at the family level. While on the one hand political democracy may mean that the will of the majority rules, on the other hand the family must be given autonomy to make its own decisions without state interference. The line at which state democracy ends and family democracy begins is very difficult to draw. Finally, the same dynamic applies at the international level. While democracy requires that the community of nations have norms to which individual states must comply, it also demands that states maintain a degree of autonomy to decide on its own norms. Thus many international conventions are enforceable on the principle of pacta sunt servanda and, to confirm their voluntary character, many provide for reservations to be made by a state signatory.

The most important principle, however, is that neither the state in the international arena nor the family in the national arena must be allowed to hide behind the concept of autonomy or non-interference, to reproduce power hierarchies and unequal entitlements and responsibilities. There is a role for the international community to ensure democracy within the nation-state, and there is a role for the state to ensure democracy within the family. Some aspects of what that role should be are discussed next.

One of the recurrent themes of this paper has been the consequences of change on families. We have shown in different ways how change altered relations between family members. Change also provided opportunities for younger family members and women to challenge and ultimately to undermine the authority of male elders. One seemingly unifying thread which runs through the historical processes of the colonial and post colonial periods, and to which reference has been made earlier, is the fact that change also increased opportunities for individual choice. By opening up the traditional economy, it was possible for individuals to assert a degree of autonomy from the wider family. The kinship ideology or glue which, among other things, had served to mask property relations and to discourage tendencies towards individualism among family members, was also greatly weakened. All these changes, as we have noted, gave rise to and came to be expressed in the new family law.

But change also resulted in serious insecurity for vulnerable groups in the community such as women, children, and the elderly. We have also shown that the economic and social vulnerability of these groups has nothing to do with the fact that they contribute less to the family welfare. On the contrary, we argued that women and children contribute greatly to the welfare of the family. Indeed, we have even declared that it is their vital contributions to family survival which have been the cause of their bondage (Nhlapo 1991).

In this context, the concept of democracy can be used to mitigate these effects of change as well as to reduce, if not eliminate, the problems arising from unequal bargaining positions between family members. It can be used to begin a process of constructing a 'level field' on which family relations can be negotiated from a position of relative equality. However, the question remains as to what role the law can play in effecting the democratization of the family, while at the same time delineating the role and defining the functions and obligation of the state to the family.

We have argued that there is a need for greater family autonomy. However, we have conversely made a demand that the African state should accept greater responsibility for family members who are also its citizens. How can the law be harnessed to harmonise these seemingly conflicting aspirations? An even more fundamental issue is the very nature of law itself within the post colonial state. Despite some efforts towards gender neutrality this law remains undeniably patriarchal in character (Hellum 1993: 243; Paliwala 1993). The law in its functioning is often characterised as being primarily directive and determinative. But the law also reinforces values and preexisting norms and hence, in its reform measures, proceeds from the perspective of the existing power base (WLSA Zimbabwe 1994).

Thus the question is whether an inherently patriarchal institution, such as the family now is, can be used to transform patriarchy. A problem that needs to be addressed is that the law is commonly treated as if it is some remote abstraction that is soulless and gender neutral. The reality, however, is that the law is created and administered by human beings with their human idiosyncrasies, who are, besides their law-related roles, members of groups and political factions within society, and that these multiple roles and interests affect their orientation. In other words, if the judiciary and the legislature are predominantly male with socially imbued patriarchal values, will the law be responsive to the changes that are proposed? It might be argued that just as the pre-colonial power blocks manipulated and tried to control the transformation of the traditional society in the colonial state, so it can be anticipated that the present incumbents of power will act to preserve their own power base against the intrusive concept of democracy.

It is not enough, therefore, to change the law. We have also to target the institutions that control and administer the law in order to create ideal conditions for effective reforms. Even without legislative reform, one might argue that to cultivate a more flexible and a more socially and economically sensitive approach to the application of the law might ameliorate some of the present problems. The WLSA research has shown that the law is an influencer of attitudes which affect the options that individuals believe, rightly or wrongly, are available to them. This might be described as the power of rumour in the semi-autonomous social fields that surround and interact with the law (Moore 1978).

There is already some evidence of positive change at the level of state law and state institutions. In some countries there have been legislative changes which appear to improve the position of women in the family and in society. However, as discussed above, the power of legislation is limited in societies where the non-state 'law' is so strong. Legislative reform may seem to be addressing evident social needs, but in fact it is only a relatively limited sector of the society whose needs in the matter are actually articulated and addressed (Hellum 1993: Paliwala [page 62] 1993: Stewart 1993).^{^[12]} It is clearly not enough to target only state institutions. We must find a way of targeting non-state institutions such as the family as well. We have discussed how the family itself is transforming. This transformation has both positive and negative aspects in so far as democracy is concerned.

On the positive side, there is evidence that change has taken place in transforming and undermining, albeit slightly, the power and gender relations in the family and the state. As discussed above, some of the economic changes taking place have upset traditional power heirarchies of gender and age. Some women and young people are able to influence family decisions and make their own individual decisions in ways that appear to be coming closer to 'democracy.'

On the negative side, there is also evidence that the power of the individual to make independent decisions and take independent actions has reduced 'democracy' in the family. Male family 'heads' may now have opportunities to abuse their position in ways that were not available in the past. Husbands may have more individual power over their wives and children, without the mitigating influence of the extended family.

How can we direct change towards, rather than away from, democracy within the family?

In this paper we have discussed the ways in which the family exercises its hegemony over its members, constraining their actions and delimiting their choices. Unless the family is democratised, its operations opened up to scrutiny, the state can rail against the inequities but the shutters of the family will be firmly closed against its intervention. Equally, if the disadvantaged are to be helped, the state and the family must work together to build this new social order which starts with the smallest democracy. However, our efforts will have been little better than mere rhetoric or an exercise in populism if the only conclusion that can be derived from our work is that 'something needs to be done about the family'. Equally suggestions for reform of the law must be achievable, effective and above all practical. Thus we do not suggest wholesale legislative reform along idealist [page 63] models, nor do we demand that the state takes over the family and its roles. We advocate a system by which the state can provide mechanisms for the support and review of the family where and when necessary, respecting the autonomy of the family, yet having the capacity to intervene as and when necessary.

The law-maker's task of attempting to provide a practical framework for transforming family forms is akin to housework, when you have seemingly caught up with it, the mess is piling up again behind you! An approach which supports the autonomy of the family has to adopt or be open to the adaptation of the existing systems to a reflexive model of the handling of family issues. Family determinations would be given judicial approval not on the basis of their accordance with predetermined prescriptive rules but whether or not the proper processes and procedures have been adhered to and the parties' legitimate interests and expectations have been addressed.^{^[13]} Such an approach requires a transformation of the adjudication structures as far as the family is concerned into structures for mediation and conciliation. Such a system must also have an in-built flexibility to permit, in the event of deadlock, the intervention of some neutral arbiter or adjudicator not necessarily being the formal state court. Such intervention might be for the purposes, either of imposing a solution or by reference to principles of natural justice after which the matter would be res judicata. Rwezaura (1986) discusses some of the problems inherent in such an approach by reference to mediation in family matters in Tanzania.

We argue that this is not a radical proposition and that the mechanisms may already be in place to carry out such procedures. However, the procedures are ignored or have been buried under years of judicial and administrative practice. There are obvious reforms that would assist in the democratization of the family both in relation to gender equity and the fair distribution of assets. Recognition of the contribution of members to the family resources and, where there is dissolution of the family, consequent distribution of family assets among the members would be a programme to pursue. However, such distribution must be predicated on a major dislocation of the family, not merely a response to ad hoc or temporary crises.

The patriarchal order, which preserves vested economic and political interests, is inimical not only to women but to many men, who are excluded from this power base and are limited in their capacity to control their destinies and that of their [page 64] families. Our main challenge is to transform the patriarchal character of the economic and social relations within the family and society and the patriarchal character of the law, which reinforces these pre-existing relations. In the Eastern and Southern African context, where the new economic order with its structural adjustment policies, is leading our countries to a debt crisis and institutionalised poverty, a process which reinforces patterns of inequality between and within nations, social classes, individuals, families and households must be rejected. Changing the situation may depend on the resource base and historical conditions in each country, and at the same time on the different survival strategies that individuals, families and households may lead.

If we claim that state strategies can create new collective opportunities for organising and negotiating for change, we do not forget the capacity which political regimes have to secure their own hegemony, based on patriarchal and undemocratic relations. At the same time the different survival strategies can also lead to more egalitarian relations not only between women and men but also for individuals, mainly those that have been excluded from access and control over resources. How can we influence power structures and the choices of people if choices and decisions are often made through interaction with a range of social and economic institutions and individuals that represent competing, conflicting, as well as shared interests? The changing of this situation is linked, on the one hand with laws and conventions at the international level and on the other hand with the capacity of African states to cope with this challenge.

The major issue as we see it is to identify the role which law can play in building democracy within the family, in a changing world. We must take into account that state intervention may not work, that the family may continue to operate and evolve in its own ways despite state intervention. We must also take into account the fact that state intervention in the family in Africa has sometimes produced more harm than good. Acknowledging the importance of unofficial law presents problems for policy makers. Is it best to limit state intervention, in order to allow the family to act in a democratic manner? Or to refine and improve state intervention? Or to seek other kinds of interventions, perhaps using the powerful unofficial law which is evolving along its own lines in Africa, sometimes in conjunction with and sometimes in opposition to, the state law. How can the state contribute to the task of creating democracy within the family where non-state law is often more powerful than state law?

It is premature to suggest solutions at this point. We are instead building a framework for concrete investigation of the family, its changing functions, its relationship to the law, and its role in building true democracy. It is necessary to look at how people themselves define family, family function, family membership, family forms. We suggest more study, using the framework [page 65] presented above, of what is happening within the family, particularly the extent to which our concept of democracy is reflected there, and, within that notion of democracy, the extent to which women have the power and possibility to express, advance and protect their interests. We suggest a multi-tiered approach, which combines the creation of space to manoeuvre under present conditions while at the same time developing strategies and ways forward. And we suggest building upon the unofficial law as the cornerstone of an understanding of the ways that the family is responding to changes in Southern and Eastern Africa.

1992 Struggling Over Scarce Resources: Women and Maintenance in Southern Africa. Harare: Women and Law in Southern Africa Trust, University of Zimbabwe Publications.

1993a Different Women, Different Laws unpublished Ph.D thesis, University of Copenhagen.

1993b "School and sadza: custody and the best interests of the child in Zimbabwe." International Journal of Law and the Family 8: 151-190.

1995 A Child Belongs to Everyone. Florence: UNICEF, Innocenti Centre (forthcoming).

1993 "Uncovering reality: excavating women's rights in African family law." International Journal of Law and Family 7: 314-369.

1983 "State Power and the Regulation of Marriage in Colonial Zambia." Theory Culture and Society 1: 181-210.

1985 Application of Customary Law in Southern Africa: The Conflict of Personal Laws. Cape Town: Juta.

1980 Women and Marriage in Kpelle Society. Stanford, California: Stanford University Press.

1984 "Report on the child maintenance laws." Unpublished paper presented to Ministry of Home Affairs of Botswana.

1978 "Neo-traditionalism and customary law in Malawi." African Studies 16: 80-91.

1982 "Making customary law: men, women, and courts in colonial Northern Rhodesia." Pp. 53-67 in Margaret Hay and Margaret Wright (eds.), African Women and Law: Historical Perspectives. Boston University Papers on Africa VII.

1985 Law, Custom and Social Order: The Colonial Experience in Malawi and Zambia. Cambridge: Cambridge University Press.

1989 "Neither customary nor legal: African customary law in an era of family law reform." International Journal of Law and the Family 3: 72-88.

1991 "Investigating women's legal rights and entitlements." In Perspectives on Research Methodology. Harare: WLSA.

1974 "Women in politics." In Michelle Z. Rosaldo and Louise Lamphere (eds.), Woman, Culture and Society. Stanford: Stanford University Press.

1981 Rules and Processes: The Cultural Logic of Dispute in an African Context. Chicago: University of Chicago Press.

1987 Casebook on Kenya Customary Law. Abingdon, Britain: Professional Books and Nairobi University Press.

1991 Justice for the People: Community Courts and Legal Transformation in Zimbabwe. Harare: University of Zimbabwe Publications.

1988 "Who owns what? Land registration and tensions in gender relations of production in Kenya." Pp. 157-176 in J. Davison (ed.), Agriculture, Women and Land: The African Experience. Boulder: Westview Press.

1988 "Death African style: the case of S M Otieno." American Journal of Comparative Law 36: 329-350.

1990 "Measuring equality: a comparative perspective on women's legal capacity and constitutional rights in five Commonwealth countries." Berkeley Women's Law Journal 5: 110-138.

1950 "Kinship and marriage among the Lozi of Northern Rhodesia and the Zulu of Natal." Pp. 166-206 in Radcliffe-Brown and Forde.

1969 "Property rights and status in African traditional law." Pp. 252-265 in Max Gluckman (ed.), Ideas and Procedures in African Customary Law. London: International African Institute and Oxford University Press.

1983 "Legal duality: conflict or concord in Botswana?" Journal of African Law 27: 150-161.

1984 "Support for women with dependent children under the customary system of the Bakwena and the Roman-Dutch common and statutory law of Botswana." Journal of Legal Pluralism 22: 1-15.

1977 "The emergence of the 'stranger-permit marriage' and other forms of conjugal union in rural Sierra Leone." In Simon A. Roberts (ed.), Law and the Family in Africa. The Hague: Mouton.

1993 "Gender and legal change in Zimbabwe: childless women and divorce from a socio-cultural and historical perspective." Pp. 243-269 in Sammy Adelman and Abdul Paliwala (eds.). Law and Crisis in the Third World. London: Hans Zell for Centre of Modern African Studies, University of Warwick.

1987 "Property disputes in law and practice: dissolution of marriage in Zambia." Pp. 56-84 in Alice Armstrong and Welshman Ncube (eds.), Women and Law in Southern Africa. Harare: Zimbabwe Publishing House.

1991 "Family law reform and the integration of the laws of succession in Zambia." Verfassung und Recht in Übersee 24: 340-361.

1992 Family Law and Succession Law Reform and Socio-Economic Developments in Zambia, Research Report. Bayreuth, Germany: University of Bayreuth.

1975 "Issues in woman's role in economic development: critique and alternatives." Journal of Marriage and the Family 37: 1001-12.

1986 "The Maintenance of Law and Order in British Colonial Africa." African Affairs 85: 411-37.

1979 African Widows: an empirical study of the problems of adapting western christian teachings on marriage to the leviratic custom for the case of widows in four rural African societies. New York: Orbis Press.

1970 The Barbaig: East African Cattle Herders. New York: Holt, Rinehart and Winston.

1994 "Love and Hate Begin at Home." In M. Rosaldo and L. Lamphere (eds.), Woman, Culture and Society. Stanford: Stanford University Press.

1987 "The love of a mother: problems of custody in Zimbabwe." Pp. 137-163 in Alice Armstrong and Welshman Ncube (eds.), Women and Law in Southern Africa. Harare: Zimbabwe Publishing House.

1991 African Households and Prospects for Agricultural Revival in Sub-Saharan Africa, Working Paper No.2/91. Dakar, Senegal: Codesria.

1991 Family Law Through Cases in Lesotho. Roma: National University of Lesotho.

1988 "Runaway wives in colonial Tanganyika: forced labour and forced marriage in Rungwe District 1919-1961." International Journal of the Sociology of Law 16: 1-29.

1990/91 "Disseminating family law reforms: some lessons from Botswana." Journal of Legal Pluralism 30 & 31: 303-329.

1991 'Children of the Fence': The Maintenance of Extra-marital Children under Law and Practice in Botswana Leiden: Center for African Studies.

1978 "Law and social change: the semi-autonomous social field as an appropriate subject of study." Pp. 54-81 in Sally F. Moore, Law as Process: An Anthropological Approach. London: Routledge & Kegan Paul.

1986 Social Facts and Fabrications: 'Customary Law' on Kilimanjaro, 1880-1980. Cambridge: Cambridge University Press.

1991 "From giving and lending to selling: property transactions reflecting historical changes on Kilimanjaro." In Mann and Roberts.

1972 Indirect Rule and the Search for Justice: Essays in East African Legal History. Oxford: Clarendon Press.

1976 "Marital strategy in Lesotho: the redistribution of migrant earnings." African Studies 35.

1977 "High bridewealth, migrant labour and the position of women in Lesotho." Journal of African Law 21: 79-95.

1987 "Released from legal minority: the Legal Age of Majority Act in Zimbabwe." Pp. 193-209 in Alice Armstrong and Welshman Ncube (eds.), Women and Law in Southern Africa. Harare: Zimbabwe Publishing House.

1990 "Re-allocation of matrimonial property at the dissolution of marriage in Zimbabwe." Journal of African Law 34: 1-8.

1991 "Dealing with inequalities in customary law: action, reaction and social change in Zimbabwe." International Journal of Law and the Family 5: 58-79.

1989 "International protection of human rights and the family: African variations on a common theme." International Journal of Law and the Family 3: 1-20.

1990 "The legal situation of women in Swaziland and some thoughts on research." In: Stewart and Armstrong.

1991 "The African family and women's rights: friends or foes?" Acta Juridica 135-146.

1993 "Family transformation and family law: some African developments in financial support on relationship breakdown." Pp. 270-300 in Sammy Adelman and Abdul Paliwala (eds.). Law and Crisis in the Third World. London: Hans Zell for Centre of Modern African Studies, University of Warwick.

1981 Legal Dualism in Lesotho: A study of the Choice of Law Question in Family law Matters. Morija Lesotho: Morija Sesuto Book Depot.

1950 African Systems of Kinship and Marriage. London: International African Institute and Oxford University Press.

1982 "Tradition and travesty: chiefs and the administration in Makoni District, Zimbabwe 1960-1980." Africa 52: 20-41.

1977 "Marriage and extra-marital sexuality: the dialectics of legal changes among the Kgatla." Journal of African Law 21: 97-123.

1988 "A woman is like a field: women's strategies for land access in Swaziland." In J. Davison (ed.), Agriculture, Women and Land: The African Experience. Boulder: Westview Press.

1989 "Kinship, Homestead and the Custody of Swazi Children," unpublished report to UNICEF.

1983 Traditionalism and Law Reform in Africa. Saarbrucken: Europa Institut der Universität der Saarland.

1985 Traditional Family Law and Change in Tanzania: a study of the Kuria social system. Baden Baden: Nomos Verlagsgesellschaft.

1987 "State law and customary law; reflections on their relationship in contemporary Tanzania." Vorträge, Reden und Berichte aus dem Europa-Institut, Saarbrucken 83: 1-21.

1988 "The changing role of the extended family in providing economic support for an individual in Africa." Bayreuth African Studies Series (BASS): 57-89.

1989 "The changing community obligations to the elderly in contemporary Africa." Journal of Social Development in Africa 4: 5-24.

1990 "The integration of marriage laws in Africa with special reference to Tanzania." Pp. 139-161 in Abun-Nasr et al. (eds.) Law, Society, and National Identity. Hamburg: Helmut Buske Verlag.

1990-91 "Tanzania: family law and the new Bill of Rights." Journal of Family Law 29: 453-461.

1994-95 "Tanzania: building a new family law out of a plural legal system." Journal of Family Law 33: 523-540.

1988 "Family law reform in Tanzania: a social-legal report." International Journal of Law and the Family 2: 1-26.

1990 "Conflict between common law and customary law in Kenya: regarding burial rights." in Abun-Nasr et al. (eds.) Law, Society, and National Identity. Hamburg: Helmut Buske Verlag.

1950 "Kinship and marriage among the Tswana." Pp. 140-165 in Radcliffe-Brown and Forde.

1990 "The legal situation of women in Lesotho." Pp. 47-73 in Stewart and Armstrong.

1978 "Legal innovation and social change in a peasant community: a Senegalese village police." Africa 48: 231.

1981a Capitalism and Legal Change : An African Transformation. New York: Academic Press.

1981b "Colonialism and legal form: the creation of 'customary law' in Senegal." Pp. 90-121 in Colin SUMNER (ed.), Crime, Justice and Underdevelopment. London: Heinemann.

1987 "Gender relations and social transformation in Swaziland." In Michael Neocosmos (ed.), Social Relations in Rural Swaziland: Critical Analyses. Kwaluseni: University of Swaziland.

1991 "Burying Otieno: the politics of gender and ethnicity in Kenya." SIGNS: Journal of Women in Culture and Society 16: 808-845.

1989-90 "Who gets the money? Some aspects of testate and intestate succession in Zimbabwe." Zimbabwe Law Review 7-8: 85-103.

1990 The Legal Situation of Women in Southern Africa. Harare: University of Zimbabwe Publications.

1993 "The dilemmas of law in women's development." Pp. 219-242 in Sammy Adelman and Abdul Paliwala (eds.). Law and Crisis in the Third World. London: Hans Zell for Centre of Modern African Studies, University of Warwick.

1989 "Women and the changing urban household economy in Tanzania." Journal of Modern African Studies 27: 601-623.

1990 "Responses of urban women to economic reforms and crisis in urban Tanzania: new role of women's organisation." Paper Presented at African Studies Association Conference, November 1-4.

1990-91 "Legally unrepresented women petitioners in the lower courts of Tanzania: a case of justice denied?" Journal of Legal Pluralism 30-31: 255-271.

1982 African Marriage in Zimbabwe and the Impact of Christianity Gweru: Mambo Press, and Edinburgh: Holmes McDougall.

1987 "The bride price, revolution, and the liberation of women." International Journal of the Sociology of Law 15: 369-392.

1981 "'I'm hungry Mum': the politics of domestic budgeting." In Kate Young et al. (eds.), Of Marriage and the Market: Women's Subordination in International Perspective. London: CSE Books.

1989 "The peculiar policy of recognition of indigenous laws in British colonial Africa: a preliminary discussion." Verfassung und Recht in Übersee 22: 273- 284.

[1] This article is a product of a week-long seminar convened in February 1994 by the Women and Law in Southern Africa Research Project (WLSA) to discuss the state of the family, its role in modern Africa and the extent to which the concerns of the International Year of the Family could be localised and made relevant to families in the Eastern and Southern Africa region. It is based upon WLSA Working Paper No. 11 (1995).

[2] Bart Rwezaura deserves special thanks for the additional research and revision that he did. The editorial team of Rwezaura, Ncube, Stewart and Armstrong made substantial changes to the first draft of the paper which was written at the above-mentioned seminar.

[3] H.F. Moore has noted that it is futile to try to analyse the family without some discussion of the existing variability in family forms and responsibilities, and the manner in which different social systems and ideologies of family life encode particular definitions of the rights (Moore 1972).

[4] Sebastian Poulter (1981) has correctly noted that this pattern of legal pluralism is a widespread phenomenon around the world and is extremely noticeable in former British dependencies.

[5] See the Tanzanian Administration (Small Estates) Ordinance (Cap.30), s. 19 (1) (a), which states:

The estate of a member of a native tribe shall be administered according to the law of that tribe unless the deceased at any time professed the Mohamedan religion and the court exercising jurisdiction over his estate is satisfied from the written or oral declarations of the deceased or his acts or manner of life that he [page 47] intended his estate to be administered according to customary law.

[6] Perhaps to provide a smooth transition from traditional to modern life, it was provided in the Tanzanian Judicature and Application of Law Ordinance (Cap.453), s. 9(1):

Customary law shall be applicable to, and courts shall exercise jurisdiction in accordance therewith, in matters of a civil nature - (b) relating to any matter of status, or succession to, a person who is or was a member of a community in which rules of customary law relevant to the matter are established and accepted... except in case where it is apparent, from the nature of any relevant act or transaction, manner of life or business, that there is or was to be regulated otherwise than by customary law....

See also In Re Innocent Mbilinyi, Deceased (1969) HCD 283 (Tanzania) where the 'manner of life' test was applied to exclude the application of customary law.

[7] The existence of western family law remedies had certain interactive effects on the traditional law. One of these was that western family law remedies tended to seep into the decisions of local court judges or administrative officers when hearing family law related cases under customary law. For example, in cases involving the guardianship of minors, courts were able to apply the concept of the welfare of the child which was foreign to most customary laws. Also courts were able to inject western ideas into customary law by nullifying marriages where the woman had not consented or where the widow had refused to be inherited. Other examples include cases of divorce which were granted by colonial officers to women petitioners without going through the traditional system (Chanock 1985).

[8] Whereas in the traditional family a wife might be subject to decisions [page 52] made by her father-in-law and the other members of the family, and within this context she had an opportunity to negotiate, this is not the case any more. Under modern conditions, where a wife resides with her husband and children in an urbancentre, her bargaining power may be greatly reduced because she has to deal with her husband who has assumed the status of the new family head. But equally she may have more freedom to plan her family budget without too much control from her in-laws. Her contribution is also likely to be more easily identified and claimable at the time of divorce than was the case before.

[9] Most of the opposition to the Zimbabwe Age of Majority Act emanated from the realisation by elders that the new law permitted young people (especially

daughters) to be married without seeking consent from their parents. It is also reported that the delay in reforming the law of succession in Zambia has comparable sources of resistance (Himonga 1990: 163). A study of the Tanzania National Assembly Debates on the Law of Marriage Bill also reveals such opposition.

[10] The material relating to the participation of women and children in the informal sector has been derived from (Rwezaura 1994-95).

[11] According to recent media reports, fishermen in the lake Victoria have stopped fishing because, in a few incidents some of the larger species they have caught, especially the Nile perch, are reported to have had human remains in their bellies. Whether or not this is true, the fact remains that people are not willing to buy fish. Consequently, even as far as Dar es Salaam, the price of beef has gone sky high.

[12] This failure may not be a deliberate political exclusion but perhaps an oversight arising from the present inability of subordinate groups to articulate their political goals and interests more effectively and even aggressively. Sometimes the oversight may be a result of adopting legislation from another jurisdiction without fully considering the local conditions (Rwezaura 1994-95).

[13] In a recent paper Rwezaura has shown how progressive judges in Tanzania have creatively expanded the provisions of the Law of Marriage Act 1971, to provide certain remedies which might not have been anticipated by the Act itself (Rwezaura 1994).