Marriage patterns in Ankole, South-Western Uganda Marriage patterns in Ankole, South-Western Uganda

Abstract This paper studies marriage patterns and systems in Ankole. Discussed in the study are age at marriage, proportion married,marriage dissolution, remarriages, types of marriage and bridewealth. The paper finds that most of the marriage patterns are among the major causes of high fertility in the area. Abstract This paper studies marriage patterns and systems in Ankole, Discussed in the study are age at marriage, proportion married, marriage dissolution, remarriages, types of marriage and bridewealth, The paper finds that most of the marriage patterns are among the major causes of high fertility in the area,


INTRODUCTION
Recent studies have shown that East Africa is one of the sub-regions with the highest fertility levels in Africa (ECA, 1979). Of the many societies in East Africa with particularly high rates of fertility are th~Banyankole of south-western Uganda. The Banyankole have maintained the highest fertility rates in the country for the last three decades. Some research has been done to explain the high fertility situation in the area. Arya et al. (1973) studied the medical aspects of fertility in the area and found relatively. little incidence of sexually transmitted diseases which have caused secondary sterility in some of the neighboring societies.. For instance, they found that Teso in north-eastern Uganda had much higher prevalence, which was perhaps responsible for the low fertility in the area.
Using data from the same survey, Costello (1986) examined the nutritional factors in Anko1e and concluded that the status of nutrition was low, which could have been responsible for the high infant mortality levels, and indirectly (through shortened birth intervals) for high fertility in the area.
In another study Sheku (1987) recently investigated pathological sterility and indicated that it could be higher t~an~hat Arya et al. 2 (1973) had found and hence depress fertility.in the area.
But fertility still remains high, which renders Sheku's findings questionable.
Unfortunately, few of the cultural variables of fertility in Ankole have been studied. This paper uses recent data to study marriage patterns and systems in Ankole.
The study gives background characteristics of the area, briefly describes data sources and presents recent fertility experience. Various marriage patterns are discuss~d and in some parts of the paper related to fertility. We conclude by suggesting a research agenda to fill the present data gaps.

OCIO-CULTU~AL SETTING
Ankole is situated across the equator in south~western Uganda, about 100 kilometers west of Lake Victoria and immediately south of Lakes Edward and George. Its size is about 16,182 square kilometers. Administratively, the area was one of the four interlacustrine kingdoms of Uganda abolished in 1967.
In the 1970s Ankole was divided into two districts, namely Mbarara to the east and Bushenyi in the west (Map 1).
The people of Ankole are called "Banyankole" and speak "Runyankole", one of the major Bantu dialects in Uganda.
Economically, the Banyankole are divided into two main groups.
The "Bairu", who form the large majority, are cultivators of coffee and tea as cash crops and bananas, maize, millet, beans, sweet potatoes and groundnuts as cash and food crops. They are relatively permanently settled. The Banyankole are organized on a patrilineal clan structure and practice patrilocal residence. Although clans are exogamous, they tend to marry within the same ethnic group.   IN 1959, 1969, 1980, AND 1985 (United Nations, 1983). It is clear that tbe four data sources agree that fertility has remained bigh at all ages, giving total fertility rates of above 8 between 1959 and 1985. A comparison of 1959, 1969, and 1985 rates for the whole of Ankole shows that fertility had been constant or slowly increasing in most age groups.
Perhaps a more interesting analysis would be to study the age specific marital fertility rates of the area. They were computed by dividing the rates in Table 1 by the corresponding proportions ever married in Table 6. We then obtained the ratios of 1985 to 1969 rates. The results are given in Table 2 and exhibit even higher levels of fertility in all age groups, 9 leading to total marital fertility rates of 10.1 and 11.7 in 1969 and 1985, respectively. The ratios in the table represent a moderate increase in fertility over the period and an overall rise of between 10 and 20 percent. It is likely that the decreasing ratios with age reflect underestimation of the pr~portion of dissolved unions at old ages due to poor age reporting.
In a~d~~ion, Figure I displays age fertility patterns for 19~9, 1969, 1980 and 1985, which show a plateau shape between ages 20 and 35 for the last three surveys. The 1959 curve / displays~n early peak. As will be explained later, marriage " \' p~tterns are largely responsible for the shape of the curves.
Furthe~f it is remarkable that data from 1985 survey -'F "p ,ẽ x~Jb'~i't:e(J. in Table 3 I 1959.1969, 1980 and 1985. . Evidence of a gradual shift from an early peak in 1959 to a broad peak in 1969 and to a latebroad peak in 1985 fertility patterns shown in Table 1 and Figure   1 implies that this rise has taken place. The reasons for the increase are largely socio-economic.  It has been observed in some African surveys that childhood residence of women influences their age at first marriage (Republic of Kenya, 1980). Our findings indicate that the women who grew up in rural areas of Ankole married at a mean age of 16.8, which is lower than the 18 years reported by those who grew up in relatively urban centers. Those born in urban centers were less bound by traditional custom to make them marry early.
Besides, they were more likely to go to school.
The pattern of work was also related to age at first marriage. The 1984 data in Table 5 indicate that women who never worked outside their parents' homes married at lower ages than those who worked before marriage in all age groups. The former group had no employment to prevent them marrying at an average age of 16.5. Those who were employed delayed their marriage because they wanted to work for some time first and they were at school before their employment. On average, they first married at 19.4 years of age.
An interesting finding is that although the cultivators and pastoralists live completely different lives economically, the mean age reported by their women in 1984 is the same, 16.8 years.
This is perhaps because both groups are still highly traditional in attitudes towards marriage of their daughters. In addition, the roles of their women in marriage are the same, being housewives and producers of children.
Further, the modernization factors discussed above do not vary between the groups.

b) Proportion married
In all societies of sub-Saharan Africa, the primary purpose of marriage is procreation. Most societies in Africa do not tolerate women who fail to produce in the first few years of their marriage. Such women would be returned to their parents and bridewealth paid back. In contrast, motherhood outside wedlock is discouraged in most societies. Consequently, marital fertility is higher than non-marital fertility in these societies.
Similarly, the main objective of Banyankole marriages is to produce as many children as possible (Ntozi et al., 1986).
Their proverb says it all, "Omukazi ekishumba akitekwamu omwana" (the essence for a woman to stay in marriage is to produce children).
If a bride spends about one year without being pregnant, traditional doctors are called in to assist. After a few years she is declared barren and the husband is free to replace her with a "productive" wife.
If the "barren" wife is well behaved, hard working and liked by the husband's family, she may continue to stay with the man while she continues to try for pregnancy. But keeping her is entirely the husband's option.
He is free to send her back to her parents and demand his bridewealth back (Ntozi et al., 1986).
On the other hand, motherhood outside wedlock is frowned upon and used to be considered a curse.
In the past, a pregnant single girl was killed by her clan before it was engulfed by the curse. Even today, premarital pregnancy is considered shameful and often punished by banishment of the girls from home to look for the men responsible for their pregnancies and marry them.
Often the mother of the girl is reprimanded for failing in her duty of looking after her daughter well.
In Ankole, it is the wish of parents that every adult girl is married off as soon as possible. This desire is frequently expressed in two popular proverbs: "Enyamwonyo kwekura eriibwa" (when a banana ripens it is consumed) or "Omwishiki kwakura ashwerwa" (an adult girl should be married off  woman is found to be a witch, abusing the parents-in-law or guilty of equivalent misbehavior, the husband's family will not tolerate her. The husband will divorce her ("okubinga omukazi").
Divorce becomes effective if it is followed by return of bridewealth ("okuzimuura enjugano"). Return of bridewealth severs all marital bonds and withdraws sexual rights of the man over the woman. It is equivalent to modern dissolution of marriage in court. The woman is therefore free to marry another man who would pay bridewealth. In cases where the parents of the woman are poor, bridewealth from the new suitor is used to pay the former husband back. The children of the dissolved marriage will belong to the first husband and the woman is not allowed to go with them to the second marriage.

26
Because a lot of effort is put into saving marriages, Ankole has a low rate of divorces and separation. This is displayed by the low percentages in Table 7, which are in single digits for 1980, 1984 and 1985. Similar patterns have been observed in many African areas with high fertility regimes. Although there is no pattern consistent with age of women, the 25-29 age group seems to display relatively higher divorce rates than other ages. This is perhaps the age when women start facing serious marital problems, which they do not know how to handle. As they get older, they mature and learn to minimize their marital difficulties.
In addition, responses from ever-married women in 1984 reveal that only 10.7 percent had divorced their first husbands. On the other hand, the Banyankole practice levirate. Widows are inherited by a brother-in-law in most cases. In a few instances, a step-son who is equal in age or older may take over the widow. Since this is done immediately after the funeral rites of the deceased husband, the widowhood period tends to be too short to disrupt reproductive processes of the woman.
However, due to the erosion of cultural values, the levirate system is losing ground.
A surprising result in Table 7 is that although this society has high male mortality (Uganda, 1976) the percentage widowed is shown to be low in 1980, 1984 and 1985. This could be an indication of a strong presence of levirate culture. Most widows remarried in this system might not have reported loss of former husbands. For instance, detailed interviews of ever-married women of reproductive ages in 1984 reported only 12.6 percent widowed from first marriages.

d) Remarriages
The impact of marriage dissolution on fertility is reduced by remarriages. In Anko1e the system of remarriages is practiced in several ways. First, young widows are supposed to be inherited by a brother-in-law or step-son. This is because once married, women belong to their husbands' family and can only return to their parents if the bridewealth is also returned to the husbands' family. Since bridewealth is usually taken from family property, all the brothers of the man have some sexual 29 rights over his wife (Elam, 1973;Ntozi et al., 1986). When the man dies it becomes easy for one of the brothers to take over his widow. The levirate culture was evolved to benefit and protect a Similarly, if the first wife does not produce some sons, a Munyankole man has a traditional right to get another wife to produce them. A man needs sons to continue his lineage into the future, in this patrilineal society.
Third, for economic, political and social reasons, a rich man is not easily satisfied with one wife. He has the property to pay bridewealth for additional wives, whom he needs to produce many children, especially sons to inherit, expand and protect his riches.
In contrast, poor men cannot marry several wives because they cannot afford the required bridewealth. While the family is culturally obliged to pay bridewealth for the first wife, it is not supposed to meet the cost of the additional wives. A man has to use his personal resources as proof that he is rich enough to deserve more than one wife. Hence, polygyny is a symbol of wealth.

31
A socio-political reason is power and prestige. Since only rich men can afford to marry several wives, it is considered prestigious to be a polygynist. It is partly for this reason that past kings and chiefs of the area had more than one wife, to display their prestige, power and riches to their "subjects".
Another social justification for polygyny is the practice of levirate.
As discussed above, men are culturally bound to inherit wives of their deceased brothers. This is part of the mutual obligations between brothers in an extended family. In true Kinyankole marriage the wife was shared by her husband's brothers in sexual relationship though privately.
The husband did not stop his brothers going to his wife privately and he did not say anything about it, though he knew it (Bitashoborokire, 1974).
The custom is partly justified by bridewealth, which is paid out of the family central pool.
The wife paid for in this way is supposed to belong to all brothers. Even the sisters of the man are respected by the wife who frequently refers to them as har "husbands".
This practice is part of the extended family system which is strong in the area. Brothers strongly believe they are of the same blood and share all things, such as clothes and pipes and hunt together. The spirit of brotherhood is so strong that if an animal attacked or killed one of them while hunting, the rest would fight it until they killed it or it finished them all.
Likewise it is strongly believed that wives and children belong to the extended family, just as their ancestors are common.
It is a way of keeping the family together. This tends to raise the age at marriage of those poor men who must work for a long time before finally getting enough property to pay. But there are very few cases of this.
Banyankole is modest and manageable by most families. The Etesots of north-eastern Uganda pay between 20 and 30 cattle. In contrast. until recently two cows and three goats used to be the Recently, bridewealth has become commercia~ized by some families.
Due to the influence of the neighboring societies of Bakiga and Bahororo, wh~often value their girls' education from primary school to university and pass on the expenses to the suitor, some Banyankole families have hiked bridewealth beyond many men's financial ability. Fortunately, the young men have responded to this new development by teaming together and raising funds from relatives and~riends in order to marry when ready, rather than waiting and' losing their fiancees. Given the spirit of communalism which already exists in Ankolean culture, it was easy for the society to embrace the idea of raising bridewealth and wedding expenses this way. One can therefore not claim that the increase in bridewealth has delayed women's marriage.
The traditional response of a young man to unfairly high bridewealth is to elope with his suitor. There are several types of elopement in the society which this paper does not intend to discuss.
Suffice it to mention that elopement is common in Ankole. Its major advantage to young men is that it allows them time to enjoy some of the benefits of marriage while looking for bridewealth. However, elopement seldom persuades the girl's parents to redu~e bridewealth. In most cases, it is costly because a fine has to be paid in add~tion to the bridewealth.

CONCLUSION
In this paper; marriage patterns and systems in Ankole have been discussed. They include mean age at first marriage which has increased from 18 to 22 between 1969 and 1985, mainly due to dramatic increases in the proportion educated females and a wider spread of Christianity. Evidence was presented to show th~t variations in levels of female education, religion, childhood residence and pattern of women's work were largely 'responsible for the differentials in the mean age at marriage.
Further, it has been found that a high percentage of women is in unions because of strong pro-marriage traditions in the area. On the other hand, divorce and separation occur despite efforts by the society to discourage them. Fortunately, there is a high rate of remarriage. In addition, although there is a high adult male mortality rate, the level of widowbood is reduced by levirate, which exists in the area.
Also, the study has identified and explained three types of marriages in th~area. Most women are married to monogamous men, however polygny is still strong, especially among the cultivators.
In addition, a large proportion of the pastoral women claimed to have had sexual relations with their brothersin-law, in accordance with a custom that gives the latter sexual rights over the former.
Last~y, the paper discusses bridewealth. In general, the level of bridewealth is low. Hence, the cost is not 'prohibitive enough to delay marriages of young men and women. \ I 38 It seems ap~arent from the study that most of the patterns of marriage discussed above are largely responsible for sustaining high fertility in Ankole. However, the,impact of the different types of marriage on fertility in the area is not clear.
Our foregoing discussion has raised some questions which require more information to answer them. It 1s necessary to collect more data on pathological sterility of women, especially its differentials between economic groups in the area.
In addition, there is need for studies that trace a woman from the time she is married through dissolution and remarriage and the number of children she gets from each marriage. Also, the type of union and its timing in a woman's marriage life and childbearing career will help to determine the influence of polygamy on fertility. Lastly, research into how marriage factors of fertility are changing should be conducted. . .