Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Geographical background
South-western Nigeria is home to about 20-25 million Yoruba (figures projected from the Nigeria Fertility Survey 1984), who inhabit an area that stretches about 120 miles along the coast of the Gulf of Guinea, east from the Benin border, to about 200 miles inland into the savannah belt of West Africa. The Yoruba are the most urbanized and possibly the most industrialized ethnic group in sub-Saharan Africa. About 20 per cent of Nigerians and 10 per cent of West Africans are Yoruba. Yorubaland has at least nine cities with populations of more than 100,000, and has a 60-70 per cent rate of urbanization overall. Lagos, the home of our urban sample children, is the centre of a greater metropolitan area with a projected population in 1991 of 10 million (Federal Republic of Nigeria and UNICEF 1990). South-western Nigeria has the country's highest concentration of industries, with more than 50 per cent of the country's manufacturing output, predominantly in light industrial manufacturing products such as furniture, textiles, clothing, plastics, paper, leather goods, foodstuffs, confectionery, beverages, and tobacco products.
The Yoruba had a complex precolonial system of urban residence, economic production, and trade (Bascom 1969). Their precolonial town crafts, dating from the Middle Ages, were among the earliest developed in Africa. Yoruba towns are mentioned in written records of the sixteenth century (Gugler and Flanagan 1978). These towns were composed of enclosed compounds, with descent groups varying in size from 20 to 2,000 persons living together in each compound. The towns had divine kings who were selected from a royal lineage by governing bodies of chiefs and elders, who represented the different wards and constituencies of the town. Originally, most towns had broad, straight streets crossing at the centre, where a palace adjoined the most important market. The tendency of refugees from tribal wars and other inmigrants to locate and build within the towns later led to more compact, less systematically planned, residential neighbourhoods (Mabogunje 1968).
According to tradition, the Yoruba migrated from the north-east between the seventh and tenth centuries, establishing Ile-Ife as their city of origin and spiritual capital, from which the sons of their mythic founder, Oduduwa, were sent forth to found their own cities and kingdoms (Lloyd 1974). The most wide-reaching of these kingdoms, which evolved before the eighteenth century as the political capital of Yorubaland, was Old Oyo. A military coup, in which the commander of the Imperial Army called upon the assistance of powerful Fulani rulers to the north, brought down Old Oyo in a Fulani victory in about 1835, sending waves of refugees to swell the towns to the south and leading to the eventual establishment of Ibadan as the major Yoruba city. A period of internal wars followed, until the time of British colonial rule, which started at varying dates of conquest in the 1890s and lasted until 1960.
Today, about three-quarters of the adult men living in the smaller Yoruba towns are farmers who commute between their town compound and a hamlet at the farm perhaps five miles away (Lloyd 1974). Urban-rural distinctions are blurred by the fact that farm dwellers view the towns as their true homes, where they return for important ceremonies and for burial. Through most of Yorubaland, rural farm hamlets spring up as temporary shelters. Farmers working in hamlets near the town may spend five nights a week there and two in the town. As the radius of farmed land expands, the more distant hamlets become increasingly permanent but continue to be viewed as daughter settlements to the town. Many migrants to the large cities continue to view their towns of origin as the homes to which they will retire and locations in which they should invest in building family houses.
Class is not a particularly useful distinction in studying Yoruba families. Although Aronson (1980, 176-182) argues that the highly educated élite represent an emerging upper class, with lifestyles and attitudes similar to those of the middle class everywhere, these major shifts appear mainly among the small percentage who are university educated. Precolonial Yoruba society was not a class society, nor are class terms in widespread use today.
Differences in wealth and lifestyle between the educated élite and the uneducated exceed those between Western social classes. While the writers know members of the professional élite whose adolescent children have never visited villages near their homes, most of the élite retain an identification with their more traditional and poorer relatives. Despite status barriers, there are no class barriers, per se, to upward mobility. Disparity in status also was pronounced in precolonial days. At that time, status was based on prestige, power, and evidence of agricultural surplus (Tuden and Plotnicov 1970; Lloyd 1974), whereas wealth and consumer goods increasingly determine status today (Babatunde 1992). In precolonial Yoruba society, sex, age, descent group, and political role determined social rank.
Concepts of descent group, lineage, family, and community
Lloyd (1974, 30) claims that descent groups are the antithesis of the nuclear family upon which industrial society is often based, but for the Yoruba the concept of family starts with the descent group or lineage, which stresses group loyalty rather than individual independence. The rights of the descent group are relatively permanent, so that an individual who returns home after many years in the city, for example, can reclaim entitlement to farm land and other privileges. The traditional descent group shares a compound, and typically has a patrilineal core group, a hereditary chief, and a set of selfregulating functions concerning internal disputes and inheritance, its own titulary deities and shrines, its own praise songs, and sometimes hereditary occupations (LeVine, Klein, and Owen 1967, 227). The descent group holds title to land and controls its distribution among members.
Among the Yoruba, descent is predominantly agnatic, through the father, although south-eastern groups also have varying degrees of cognatic descent residence patterns and inheritance through the mother's side. Hard and fast generalizations on this subject inaccurately categorize an intrinsically flexible (Eades 1980, 37-63) and increasingly fluid (Guyer 1990) situation. Lineages trace descent from a common member. New members are added by birth, while the dead are believed to retain their interests in, and influence over, the group, so that the group exists in perpetuity.
There may be a formal lineage or family corporation, composed of senior male members who hold monthly meetings to run family affairs (Gugler and Flanagan 1978, 131), such as holding and managing property, seeing to the economic welfare of members in need, caring for children of incapacitated parents, and arbitrating disputes among members. Lineages typically manage land allocation among their members, while competing with other lineages for tracts of land. According to Eades (1980, 60), it appears likely that the degree of corporate identity is correlated with the control of resources.
In addition to lineage membership, Yoruba town dwellers belong to many traditional community associations based on age, religion, and occupation. As the more traditional versions of these associations have fallen into decline, new associations take their place, based on such alliances as modern religious groupings, neighbourhood friendships, occupational associations, rotating credit associations, and town improvement unions (Aronson 1980; Eades 1980, 61).
Seniority system
Inherent in the concept of lineage structure is the system of seniority, which establishes a single hierarchy of reciprocal obligations in all situations (Aronson 1980, 93-94). Traditionally, any senior had a right to unquestioned service, deference, and submissiveness from any junior (Lloyd 1974, 35-36). Traditional rules assign age seniority according to order of entry into the lineage, either by birth or by marriage. Seniority also is derived from gender, hereditary titles, designated leadership roles, physical ability, and supernatural endowment (as in the case of the priesthood).
Seniority traditionally determined task allocation and resource distribution in the labour system of the household production unit. Distinctions defining seniority were, of necessity, elaborate and were expressed in the myriad terms by which individuals greeted and addressed each other (Fadipe 1970). Distinctions among these titles and greetings might claim in the old production system the same importance now attached in the modern sector to job grades and job descriptions.
By 1988, however, commercialization of the rural economy had led to major shifts in production systems such that seniority no longer entitled the senior party to significant productive labour, such as farm work, for example, from the junior party (Guyer 1990). According to Gugler and Flanagan (1978), the unrelated members of city housing and neighbourhoods re-create and make use of the seniority system in social discourse and in requests for minor errands or services. While Aronson (1980, 94) claims that the importance of seniority for purposes of social etiquette has not diminished in new urban areas, the modern employment structure introduces a time-bounded seniority system with new rules, sometimes creating uncomfortable conflicts in seniority, as, for instance, when a young, highly educated female supervises an older male driver or secretary. Overall, although seniority still is much honoured, the resources that flow through the system are more and more supplemental rather than crucial, except in providing access to such things as job opportunities that are not awarded on the basis of money or merit.
Living arrangements
The typical traditional Yoruba compound contains a large patrilineal and patrilocal extended family. The head of the family is usually the most senior male member, and the men are normally polygamous, with each wife having a separate room. Most houses are rectangular single-story buildings, with either a traditional central compound or a central corridor and a yard at the back (Eades 1980, 40). With the reduction of courtyard space, doors and windows have become larger. Before European contact, the rooms had no windows (Fadipe 1970, 98).
In our study population, fewer than 10 per cent lived in traditional compounds, whereas 74 per cent lived in the more compact style with the narrow central corridor, referred to as "face-to-face" housing; 71 per cent of families occupied one room. This distribution of living arrangements contrasts with data from 799 families of slightly higher socio-economic status in the Ebute-Metta section of Lagos studied by Olusanya (1981). In this sample, only 1.6 per cent lived in "face-to-face" flats and only 59 per cent in one room in a house. Of Olusanya's (1981) élite sample of 1,002 families in SuruLere, Lagos, 63 per cent had self-contained flats, and 27 per cent lived in single-family houses.
A major difference between low-income urban dwellings and traditional compounds is the predominance of unrelated families, often of different ethnic groups, living together within the urban buildings on a rental basis. In studying Ijebu immigrants to the city of Ibadan, Aronson (1980) found that very few lineage members lived in the same urban compounds or housing complexes, although they often accommodated each others' children in their urban residences and automatically would have lived in the same compound in their own home towns.
The average family size in Olusanya's lower-income sample in Ebute-Metta was 6.5 persons, consisting of 87 per cent nuclear fami lies (mother, father, and children); 6.6 per cent polygynous families (two wives under the same roof); and 6.5 per cent modified-extended families (including one or more grandparents). All of these types of households could include house help and fostered children. For the élite in Suru-Lere, these figures were 7.5 persons per family, with 86 per cent nuclear, 2 per cent polygynous, and 11.2 per cent modified extended (Olusanya 1981).
In our groups who were poorer than Ebute-Metta residents, average household size of the rural, semi-rural, and urban subsamples were 4.5, 5.6, and 5.6 persons, respectively. Although we did not ask whether cowives lived under the same roof, indirect measures suggest that about 6 per cent of families had two or more wives living together under the same roof. The small household size of our poor samples confirms Olusanya's assertion that extended family members, who live separately under conditions of urban poverty, tend to regroup somewhat as socio-economic level rises. Aronson (1980), however, made a marked distinction in Ibadan between the highly educated élite, who continue to maintain small monogamous nuclear families, and the more traditionally affluent urban dwellers, whose households expand with co-wives and extended family members.
Sleeping arrangements also have changed. In the past, the husband and each of his wives had his or her own separate room. Female children slept with their mothers until adulthood, as did male children during early childhood, after which time they moved onto the veranda or into a separate boys' room (Fadipe 1970, 98). Now, when finances permit, husband and wife share a room, and the children sleep in the parlour (Aronson 1980, 34). If only one room is affordable, husband and wife share the bed and the children sleep on mats on the floor. The opening of these single-room dwellings onto a narrow central hallway may accentuate crowding. Fadipe (1970, 101102) described the intense interaction of the kinsmen and co-wives of the traditional compound:
A large part of the day is spent in the open ... everyone eats and drinks and talks in the full view of everybody else; and as the rooms are hot in the daytime..., most of the life of the compound has to be passed on the open veranda [now, in the yard] ... quarrels and rebukes take place within the full hearing of neighbors ... each individual's weaknesses and vices are open to the observation of other[s] ... People outside the immediate family are interested in its members and their welfare ... This makes exclusive family life in the Western sense impossible. Only a limited amount of privacy is possible.
According to Aronson (1980, 52) and Gugler and Flanagan (1978), the same lack of privacy tends to prevail in urban households with unrelated occupants, although larger windows and amenities such as electric fans make it more comfortable to seek seclusion behind closed doors. Aronson (1980) asserts that a major change associated with urbanization is the possibility of privacy, even if it is rarely realized. Olusanya (1981, 23) studied both lowand upper-income neighbourhoods in Lagos. In quoting a lament from a column in the newspaper Lagos Weekend, he notes a trend towards privacy and seclusion when it is affordable:
Once in the city of Lagos, this virtue [of openness, and being one's brother's keeper] is no longer to be found. This is because there has been a cold craze ... branded "fenceophobia" ... and its symptoms include a tendency for seclusion. Many people are getting secluded in Lagos and one cannot even know his neighbors again. This is all part of being Westernized.
As indicated in the column, the communal quality of life is a social value, extending to the traditional town quarter or ward, and to the low-income urban neighbourhood. Gugler and Flanagan (1978, 7678) quoted an elder in a Mushin, Lagos, neighbourhood:
The action of a man cannot be hidden. If you are living in the neighborhood and you are bad, before you speak your neighbors will know what kind of man or woman you are ... If you go to them for help, they will pity you and help you. They may help you settle a quarrel.
The status of women
Yoruba women are both autonomous and subordinate to men. Autonomy arises through a fairly rigid sexual division of labour, which excludes women from most agricultural work, and means that traditional women work independently of their husbands and not jointly or cooperatively with them (Lloyd 1974, 37-38). Although a woman traditionally expects her husband to provide her with capital to start trading or to establish her craft, she is her own supervisor. Income derived from her labour is her own - to spend on herself and her children, after fulfilling her obligations to share in purchasing food, clothing, and sundries.
An aspect of the division of labour that is often ignored but which puts women at some economic advantage in Yorubaland is that, except in the case of cash crops, Yoruba social expectation and conjugal etiquette forbids the farmer to carry his own farm products to the market to sell. Any man who does is regarded as a miser. Since it is the wife who sells these items, whatever she declares as a sales profit is what the man will accept. It is not unusual for women to use some of such funds to begin petty trading in other goods as well.
The Western concept of the full-time housewife, who earns no income, is alien to Yoruba tradition, although women in urban areas who cannot find employment do become housewives, and the concept itself has been introduced by Christianity and Islam. In our sample, 10 per cent of mothers of two-year-olds described themselves as "housewives," but only 1 per cent reported earning no income. In Olusanya's study (1981), 34 per cent of the low-income and 15 per cent of the élite group described themselves as fulltime housewives (after reassigning women who identified themselves as both housewives and income earners to a non-housewife category).
In our sample, the mean ratio of father's contribution to food expenditure to the mother's income was 1:8 for the urban, 1:3 for the semi-rural, and 1:17 for the small rural sample. The lower contribution of women among the semirural may reflect their separation from rural agricultural production and from the major urban markets. Husbands and wives traditionally do not pool their finances. On the death of spouses, their individual property does not pass to each other, unless they have been married in the church (Fadipe 1970, 140146). In everyday life, each spouse comes and goes by himself/ herself, without necessarily telling the other one where he or she is going, although wives are expected to announce their intentions more frequently than husbands (Aronson 1980,115).
Despite women's autonomy, however, many aspects of the social system give men greater seniority and control than women. Men are permitted several wives, but women may have only one husband. According to the terms of traditional agriculture, the man controlled the farm labour of his sons by all wives until the time of their marriage. Moreover, according to traditional marriage conventions, a new wife was junior not only to her husband but to all of his lineage members born before the date of her marriage (Fadipe 1970, 114). She was also a subordinate in the domestic domain, where much unpaid labour was expected of her from her husband and his extended family. Young urban wives now prefer living away from their husbands' families because they no longer are willing to take so subservient a part in family life - being subject to running errands at any time, and shopping, fetching water, and cooking for the older women. Urbanism, education, and adherence to Christianity - with monogamy as its marriage tenet - have given women some measure of freedom from the control of the extended family.
Women's economic enterprises typically have been smaller in scale and subordinate to those of men. Despite very high female participation in petty trade, the trading structure has always been stratified to the relative disadvantage of women, with men in charge of most major longdistance trading enterprises. In addition, women were traditionally excluded from most, but not all, traditional political offices (Afonja 1990).
According to Afonja (1990), the Yoruba ideology of kinship and marriage, which operates to the disadvantage of women and which is relatively impervious to change, has greatly influenced the effects of modernization. Women have been denied access to, or control of, the new means of earning income introduced through contact with the West, and their relative position has deteriorated progressively since this contact began. Proportionately fewer girls than boys have attended school or studied to the higher levels. In our sample, the mean years of schooling for the fathers was 6.7, with 25 per cent having completed secondary and 10 per cent not having attended school. Corresponding figures for the mothers were 4.3 years, and 11 per cent and 35 per cent. Beginning with the slave trade, and increasingly thereafter with the introduction of coffee and cocoa, gender inequality among the Yoruba was intensified through increased male control over critical resources. Ye there have been counter-currents to this trend during the giddy years of the petroleum "oil boom," in which women competed and sometimes succeeded in landing bigger oil contracts than did men (Babatunde 1992).
Men used precapitalist systems of control to monopolize the technology, the new knowledge, and the products of capitalistic production of export commodities (Afonja 1990). Women were easy targets for "peasantization" and "proletarianization." With cash cropping, some women have entered commercial farming, but more have been co-opted to work on their husbands' cash crops, to their husbands' financial benefit and to the disadvantage of their own incomegenerating activities. Children's school attendance increases the women's work and financial burden. In short, Afonja states that the rubric of joint financial support for the household remains superimposed on a structure that denies women control of the most critical resources in modern capitalist society.
According to Afonja (199O), the nature-culture mystique for explaining gender inequality is relevant to the Yoruba view of women. This model holds that women are closer to nature by virtue of biological reproduction and that men, as creators of culture, are inherently superior because culture is superior to nature. Babatunde (1992) does not hold this view that procreation becomes the justification for women's inferiority. He demonstrates, instead, that the relatively lower status of females in day-to-day life is reversed in the symbolism of the fertility-enhancing Gelede cult, which honours the powers of older women who are believed to hold control over fertility. These women, past the age of menopause, have the right to speak their minds freely (Fadipe 1970, 116) and potentially are to be feared if they misuse their powers over nature.
Virtue in the form of `'good character" also is idealized as female. In myth she was the Supreme Diety's granddaughter, given in marriage to the oracle divinity. When, after some time, he mistreated her, she returned to her father's house. When the oracle divinity traced her to her hiding place and begged her to return, she refused to return in physical form but promised to abide with him in the invisible form of a helper - i.e. "good character." Good character brings success in life to a man who treats his wives with kindness and affection, takes care of his children, and does what is good (Babatunde 1992, 214).
The Yoruba value of children
Yoruba culture places extremely high value on children. It is safe to say that children are the summum bonum - the highest good - of the Yoruba. An economic rationale underlies this value, as has been pointed out by Caldwell (1976) and other "price tag theorists" (Olusanya 1987), who imply that parents bear many children because they profit from them financially. According to Olusanya (1987), these writers miss the point of the Yoruba experience and are in error regarding the profits as well.
When taken literally, arguments that parents bear children because of crude economic incentives are demeaning and inaccurate. The value of children to their parents may better be expressed in spiritual terms. A spiritual view explains why women, whom Olusanya (1987) refers to as "fertility martyrs," continue to bear children when further child-bearing is not only unprofitable, but also places women's health and financial security at risk. According to Babatunde (1992) and Hallgren (1991, 120-122), the very nature of the immortality of the soul flows cyclically through the lineage through the birth of children, and not primarily through the type of afterlife pictured by Christianity or Islam.
In traditional Yoruba religion, the various component parts of the soul can continue the good life eternally in a cycle of three states the living, the ancestors, and the unborn awaiting reincarnation. Children reincarnate ancestors of their own lineage. Continuing participation in this cycle depends on bearing children, living a long, full life, and being venerated by one's descendants. Although the majority of Yoruba now belong to the major world religions, their feelings about the value of children engendered by the earlier belief system do not appear to be greatly altered by new beliefs.
The old religious beliefs supported the old economic order. As discussed by Zeitlin et al. (1982), the traditional flow of investment, not only in Africa but in earlier decades in industrialized societies, was from child to parent, not from parent to child. This direction of flow could still be observed in the United States as late as 1907, when many children lived at home and contributed their incomes to their parents until they were married, and when children's earnings were more important than a wife's earnings in supplementing family income, as documented by Hogan (1985, 107) in Chicago. Over the course of economic development, the direction of investment reversed.
Surveys conducted by Caldwell and Caldwell (1977) indicated that Nigerian children began to contribute substantively to the family's subsistence by the age of five or six. The parents estimated that each child born would remit an amount of money to the parents equivalent to 10 per cent of household income, in addition to providing child care for younger children and old-age and disaster security. Children also benefited the family materially at the lineage level. The larger the population of the lineage, the greater its claims on land and other resources, and the greater its chances of surviving in perpetuity (Lloyd 1974)
According to Guyer (1990), however, even in rural areas the direction of intergenerational transfer has changed since the 1960s. At the time of Guyer's first field work in rural Oyo in 1968, sons already were ambivalent about working on their fathers' farms, as tradition demanded, because their fathers were not under obligation to share any given amount of the farm proceeds with them. But a young man had few personal needs in those days, except the bride-wealth for a first marriage, for which the cash component accounted for over half a year's income from an average adult man's farm. The father was obligated to pay this sum. When Guyer returned to the same area in 1988, she found that teenaged and young adult sons no longer worked on their fathers' farms: on the one hand, a higher level of commercialization allowed the sons to earn ready cash; on the other, the custom of paying bride-wealth was less mandatory. Guyer states that while lineal identity and loyalties continue, the only material transfers that parents claim they can now rely on from their children are burial rites, nursing through their final illness (by daughters), and some crisis assistance.
The fact that bride-wealth has been commuted to cash payments denies the philosophy behind bride-wealth, which had to do with testing the endurance and patience of men who take daughters into marriage and keeping husbands in close touch with the wife's family. Although not all parents count on their children's support, the obligation for adult children to provide it is still strong. In fact, two factors that have made corruption a problem are the need to satisfy acquired expensive tastes and the need to keep up with the neverending entitlements of the extended family.
To support her observations on the changing economic value of children and hence changing family priorities, Guyer quotes a statement from Berry (1985), in Fathers Work for their Sons, regarding productive assets: "Low and uncertain returns to most forms of productive and commercial activity do not slow things down; rather they reinforce the impetus to keep moving, if only to avoid falling farther behind in the economic and political lottery of accumulating good connections (Berry 1985, 192)." Guyer (1990, 5) proposes that children now are valuable for the lateral networks of kin relations which they establish between a mother and the father(s) of her children, in a shift "from longer term to shorter term logic, on the part of both men and women." The parents provide support and a network of kin connections and opportunities to each of their biological children and the child's other parent during the present generation. A new survival strategy for women, therefore, becomes one of bearing children by several fathers in serial relationships. This pattern is similar to that described by Gussler (1975) among low-income mothers on St. Kitts.
In support of the hypothesis that survival hinges on lateral networking, Guyer noted a marked increase in ceremonial celebrations for professional launchings, such as freedom from apprenticeship or book publication, that create opportunities for occupational, political, and other forms of entrepreneurial networking by attracting heterogeneous crowds. With respect to lifestyle rituals, she found marriage less formal, naming ceremonies for babies much more lavish, and apprenticeship "freedom" ceremonies second only to funerals in expense.
Marriage and the husband-wife relationship
Marriage typically is prohibited between partners who can trace a blood relationship. The normal age of marriage is between 25 and 30 for men and between 17 and 25 for women (mean age 20 in our sample). A man's father traditionally was, and often still is, responsible for arranging and financing his sons' marriages. Although today the majority of young people choose their own partners, most obtain their parents' consent. Lengthy, discreet inquiries and introductions may still be made through a female intermediary who belongs to the groom's side of the family by marriage (Eades 1980, 5659).
Modern forms of marriage vary from the English-style weddings, under the Marriage Ordinance, to marriage by Yoruba customary law, to simple parental consent and blessing, down to casual and temporary mutual consent. Of our sample mothers, 75 per cent claimed to be married by customary law; 6 per cent had been married under the Marriage Act; 13 per cent were cohabiting, 2 per cent each were single and divorced, and 1 per cent were widowed. One pattern is for men with monogamous homes to have "outside wives" (Aronson 1980, 113114). While the existence of these women may not be known to the "inside" wife, the outside wives consider the man their husband and consider their children entitled to share in his inheritance. Another common pattern is for urban polygynous men to divide their time between wives and children living at different addresses.
In the traditional division of labour, the husband provides capital with which his wives trade or engage in crafts. With their profits, the women cover many of the costs of food, clothing, and sundry needs for themselves and their children, and take turns feeding the husband (Aronson 1980, 132-135). The husband provides housing, staple foods, and some money for education and children's clothing; the wife provides her own clothes, the rest of the children's clothes, and other items of food (Eades 1980, 68). In rural areas, the wives have a share of farm products. A farmer and his wife have a commercial arrangement whereby she markets his crops. When game was available, the husband provided meat by trapping or hunting. Housekeeping chores were the responsibility of the wife, who could delegate them to children and others junior to her in rank. Husbands disciplined their wives without the intervention of neighbours, unless the neighbours judged that the punishment exceeded the crime (Aronson 1980, 52).
As part of the proliferation of new arrangements, urban husbands and wives, according to Aronson (1980, 53), may operate more as a team than they would in the rural areas, where gender roles are more rigidly defined. In the city, as mothers work at greater distances from the home and as the costs of children's schooling increase, the father may look after the children more directly, and the couple may share expenses more cooperatively. As noted by Olusanya (1981), higher-status urban husbands might prefer that their wives did not work, if her work were not needed to cover extended-family obligations from the wife's side.
Polygyny and the relationship between co- wives
While polygyny was necessitated by the traditional structure of the economy, not all men can afford more than one wife, and it is questionable whether the majority of husbands ever had more than one wife (Aronson 1980, 115). In our sample, 33 per cent of our urban, 47 per cent of the semi-rural, and 89 per cent of the small rural sample of families were polygynous. Olusanya's estimates of 7 per cent polygynous families in Ebute-Metta and 2 per cent in Suru-Lere (Olusanya 1981) apparently counted only those cases of polygyny where cowives lived together under the same urban roof.
In traditional rural conditions, acquisition of a co-wife adds to the prestige of the household and of other wives living in the same compound, who are entitled to assistance with chores from the newcomer according to their seniority. According to Fadipe (1970, 115ff.), the socially approved young wife is extremely deferential to senior compound members for at least a year after her marriage, or until the arrival of her first child, after which time she can pay more attention to her own immediate business. Among other things, the new wife is a useful standby for children in trouble with their parents, rushing forward to plead for a child she hears crying under punishment. Wives of the compound, individually or collectively, ideally should never lose their role as a peacemaking force in the compound.
In rural areas polygyny remains profitable. Polygynous men have been found to have larger farms. More wives enable a man to manage scattered landholdings, which, traditionally, their sons would farm for their father. By helping with child care and household duties, junior wives in rural areas free the senior wife to trade (Eades 1980, 69).
Polygyny has been linked to a number of drawbacks, particularly in urban areas. As Aronson (1980) notes, if any misfortune befalls a child, the jealousy of a co-wife may be suspected as the cause. In addition, polygynous marriages in Ibadan were found by Olusanya (1970) to be less stable than monogamous marriages. Co-wives tend to compete through bearing children. In our study, the greater her number of co-wives, the higher the mother's stated ideal number of children. Sembajwe (1981) provides evidence from several studies that Yoruba women in polygynous unions have the same high fertility rates as those in monogamous unions. This occurs in spite of the fact that reduced coital frequency and greater age differences between spouses tend to make polygynous women less fertile. De facto or temporary unions, however, were less fertile than formal unions.
Bledsoe (1990) claims, based on field work in Liberia and Sierra Leone and a literature review that includes Nigeria, that female education exacerbates inequities between polygynous women who previously would have lived together. A man can now at almost any point marry a new wife who is more educated and more socially presentable than his earlier one(s), at which point he can effectively cut off, or greatly reduce support to, children by the previous unions. The previous laws that would have given an older wife seniority, regardless of education, no longer apply. Bledsoe states that men now sustain the costs of polygyny and high fertility by marginalizing lowstatus women, usually those with the least education, as outside wives and their children as outside children.
In our study, children from monogamous unions were significantly more likely to be in our high developmental group, even after controlling for ruralurban location, presence of the father, and economic and educational variables. After controlling for monogamy, children whose fathers lived with their mothers all or most of the time, rather than less than half of the time or never, were also significantly better in growth and cognitive scores. Fathers and mothers who had never attended school, and hence probably followed the rules of traditional polygyny, had children with marginally higher growth and development indicators than did parents with some primary or secondary education. At the other extreme, the 28 children with parents who were married under the Marriage Ordinance or listed "professional" occupations, or with fathers with more than 12 years of schooling, had significantly better status than all others.
Divorce and other forms of family stress
Traditionally, a Yoruba woman had only one marriage ceremony, without rituals to mark remarriage after being widowed or divorced (Eades 1980, 5859). Her husband's death did not mark the end of her marriage, which would continue according to the levirate system with a junior member of his descent group. Now, the ease of divorce varies with the legal status of the marriage. While English-style ordinance marriages can be dissolved only in the High Court, marriages contracted under customary law, which permits polygyny, easily can be dissolved in the local courts. Yoruba men rarely sue for divorce, and only on grounds of adultery. More commonly, wives leave husbands who have stopped supporting them, move in with a lover or with their parents, and start divorce proceedings from there. The main issue in these proceedings is repayment to the husband of marriage presentations and trading capital. Fathers traditionally have the right to keep the children, but do not usually do so.
It is easy for a woman to remarry, in part because it costs less to marry a divorcee than a first-time bride. In a study by the Okedijis (Okediji and Okediji 1966) in Ibadan, cited by Eades (1980, 58), the most common reasons given by women for divorce were non-support by the husband (71 per cent), trouble with co-wives (32 per cent), trouble with in-laws (20 per cent), and lack of children (20 per cent).
In the old days, when marriages dissolved, the parent generation apportioned blame and dictated conditions of settlement according to custom. Their role now is more and more limited to giving advice. In court cases, custody of children is increasingly awarded to the mother, and instead of receiving refunded bride-wealth, the husband may find that urban courts require him to provide child support (Gugler and Flanagan 1978). According to Guyer (1990), casual, easily broken alliances are increasing, although almost one-half of the men in her 1968 sample still had the same wives 20 years later. She states that women increasingly are able to keep the children after divorce and to maintain their claims on the father.
The changing pattern, according to which it is common for men to require girls to prove their fertility by becoming pregnant before they marry them, starts a chain reaction of marital dysfunction. The girl who proves herself in this manner enters into a depreciated role as wife and, hence, continues a demanding rather than a respectful relationship with her husband. She judges him on his ability to prove himself to her by providing support. Lacking emotional satisfaction, he then courts other women. His wife accepts his infidelity as the price she pays for sexual abstinence during breast-feeding. Her main concern is that he does not give the new woman too much money. Aronson (1980, 45), however, reports that urban women may also prefer to be pregnant before marriage to avoid the many problems that a childless married woman encounters, and therefore the woman also may initiate this process.
Both husbands and wives suffer from the fact that the urban environment tends to reduce women's earning capacity in comparison with the costs of living. Both expect, as under rural conditions, that she will be economically independent and able to contribute substantially to the family income. Yet she may not be able to find a paid job, sufficient capital, or sufficient child-care assistance to trade profitably in the urban setting. The husband is then forced into the unfamiliar role of breadwinner for his wife and children, and is alone held responsible for their financial support. He, meanwhile, may not regard his wife and children as having first claim on his wage or salary, often feeling that his mother, siblings, or cousins have prior claims.
Babatunde (1992) notes that difficult economic circumstances and the new value placed on consumer goods tend to make women value men for their money, rather than for traditional virtues. The breaking and the forming of marital ties is less regulated than before. Husbands may simply turn away a barren wife who earlier would have been kept on, although, according to Guyer (1990), the wife might leave first to attempt to bear a child by another man. Wives also may leave a husband who cannot pay the rent. A man may bring a new wife into the home without prior warning, and wife-beating may ensue from these stresses. This sequence of observations highlights the preoccupation with lateral transfers of support noted by Guyer (1990).
Modes of conflict resolution
According to Fadipe (1970, 307-308), anger is given very little overt expression or is expressed diplomatically in noncommittal sarcastic words whose literal meaning is the opposite of the true meaning of the speaker. Personal problems of all kinds commonly are believed to be caused by the jealousy of enemies within the family, whose identity may be secretly revealed through divination, and against whom religious protection is needed.
When conflict is overt, according to Aronson (1980, 115-116), Yorubas externalize it by direct and indirect insult and resolve the issue by involving outsiders. He claims that the first of three phases is an outburst of hostility, usually in the form of insults, ridicule, or, most seriously, imprecations or curses. The second phase brings an audience of friends and supporters for both sides of the issue. Resolution, the third phase, results from mediation by neutral parties, during which responsibility for the conflict is fairly apportioned to both sides. The above process is very public.