Marriage and Family in the Dedza District of Nyasaland

Author: Lucy P. Mair

Source: The Journal of the Royal Anthropological Institute of Great Britain and Ireland, Vol.81, No. 1/2 (1951), pp. 103-119
Published by: Royal Anthropological Institute of Great Britain and

INTRODUCTION

The Dedza District of Nyasaland has a population of 140,000. Within it seven Native Authorities are recognised, two Ngoni, three Cewa and two Yao. As will readily be imagined, members of the three tribes are not neatly sorted out into the appropriate Native Authority areas. The Ngoni are inextricably intermingled with the Cewa, with whom they intermarried from the time of their arrival as conquering invaders in the latter's territory; for practical purposes they are distinguishable from them today only by the fact that they follow the rule of patrilineal succession. In Dedza District a Ngoni village is one in which a considerable proportion of the men claim to be Ngoni; some can still make good the claim on the ground that they were born before the Ngoni left Domwe, in Portuguese East Africa, at the turn of the century. The Cewa were invaded also by the Yao, and the present boundaries of Yao and Cewa are those laid down when the territories of " Principal Headmen" were defined in 1924. There are groups of Yao villages unider Cewa chiefs and vice versa. Within a mile or two of the court-house of the Cewa N. A. Kaphuka there are a group of Yao and a group of Bisa villages, both established before the days of effective British occupation. Today there is some immigration into Dedza of Ngoni from the densely populated neighbouring District of Ncheu; some of these obtain land from Cewa chiefs. One might expect the result to be a bewildering variety of family structures and marriage laws. In fact, however, in this district of mixed population a more or less homogeneous custom appears to have evolved.


The fieldwork in the course of which this information was obtained was carried out at the headquarters of the Ngoni Native Authority Kachere and the Cewa Native Authority Kaphuka. I spent four weeks at the former centre and eleven weeks at the latter. Cinyanja is today the language of both Ngoni and Cewa. I was able to speak this, though not as well as I, could wish, and I had the assistance of a clerk, Mr. G. I. Ganizani, whom the District Commissioner, Dedza, very kindly placed at my disposal. Although he was lent to me as an interpreter he very rarely translated anything into or from English; he put my halting statements into better Cinyanja and repeated for me the remarks of indistinct speakers. Much of my material comes from cases brought before the Native Courts and could not have been obtained without his assistance, as the proceedings were at first unintelligible to me, and I never was able to follow a case completely.

When I refer in this article to "Dedza District," I mean to indicate the Ngoni and Cewa population of the District, but not to include the Yao.

SOCIAL GROUPINGS

Among both Ngoni and Cewa, every person bears, in addition to his personal name, another which in most parts of Bantu Africa would be called a clan name. Here, where the common possession of such a name does not now denote membership of an exogamous group nor of one observing ritual prohibitions in common, it would be contrary to accepted anthropological usage to speak of a clan. The Nyanja word is mfunda. Both patrilineal Ngoni and matrilineal Cewa take the mfunda of the father. With the Ngoni the name is normally used-in respectful address, and also playfully to small children. I have heard the Cewa Native Authority Kaphuka addressed by his mfunda name, but suspect that this is one of the ways in which some Cewa deliberately imitate the Ngoni. There is of course a presumption that persons with the same mfunda name are related; though the Cewa have so few that it is very common to find couples of one mfunda. I was told that a youth wishing to marry a girl of the same mfunda would find out from his parents whether they were too closely related. The same informant said that marriage would not be permissible between persons who could trace their descent to a pair of brothers or a pair of sisters.

It is considered important that a child should bear the mfunda name of its actual father, though there may well be cases in fact where the paternity of a child begotten in adultery is concealed. In the case of the Ngoni the mfunda names of a woman's children are a useful indication to the anthropologist of the number of fathers. When councillors in the Native Court are berating women for promiscuity they may refer to " having children of several mfunda."

I have never heard the word mfunda used except in relation to the name, and rarely except when I was myself asking for such names. Two words which denote kinship are pfuko and mtundu: mtundu in other contexts simply means " sort " or " kind." Neither describes a defined group, but Cewa informants, explaining- why they live where they do, may say "because the chief is of our pfuko," " because the mother's brother is of our own pfuko, our own mtundu." With the Ngoni the words would of course refer to paternal kin. One informant explained pfuko as a group with a common place of origin.

The two terms in common use to describe social groupings are banja and mbumba. In practice these are used more or less'indiscriminately, but in so far as they can be distinguished it would appear that banja refers primarily to a domestic unit and mbumba to a wider group based on kinship. Banja could in some contexts be translated " home." As generally used it denotes both the homestead, inhabited by a group of kindred and their spouses, and the single household within this. In the latter context it signifies something more than a building and its inmates; a new banja is not established until a wife begins to cook her husband's food separately instead of cooking with her mother as she does when first married. But banja can also be used to describe a man's children after they have grown up and left him.

I have not the genealogical data which would make it possible accurately to define the mbumba. It is not a descent group, for it is said that the head of a village should be the man with the largest mbumba, and this can hardly be taken to refer to his kindred without their spouses, while with the patrilineal Ngoni it would exclude his sons living elsewhere. At the same time the " head of the mbumba" is the man whom a group of kinsmen recognise as their senior, and in one case a group of Cewa brothers living in one home- stead stated that the head of their mbumba was an elder brother who lived in his wife's village. It is sometimes stated that the consent of the head of the mbumba is required for a marriage and that he may be appointed a marriage surety (nkhoswe) with the duties described below. But in these contexts people speak equally often of the head of the banja.

The banja in the sense of " homestead " consists of a group of huts on a cleared piece of land; they are not placed according to any recognised pattern. It is not always easy to see the line of demarcation between neighbouring homesteads, though at Kaphuka's each was separated from its neighbours by a space of cultivated or grass land. Sometimes the inhabitants of neighbouring homesteads were as closely related as those of a single homestead might be, but the interval between, however small, was enough to constitute them separate mabanja. The homestead is identified by the name of the male head of the kin group which forms its nucleus.

The Ngoni, while retaining the principle of patrilineal succession, have almost without exception adopted the marriage system of the matrilineal Cewa, namely, that the husband goes to the wife's home. As with the other peoples of this region, however, this rule is tempered by the right to take her away, with her parents' consent, when they are satisfied that he will treat her properly. He may then either build a house near by, to become in time the nucleus of a new homestead, or attach himself to that of a brother or maternal uncle, which may or may not be in the same village as his wife's home. It is rare In Dedza District today for a man to live with his wife's relatives until he is himself the father of married daughters, though my data include one case of this kind.

The husband's move to his wife's home and then back to rejoin his own kin would not be likely to involve a long journey except where, as sometimes happens in these days, a man-who has been employed a long way from his home village decides to settle permanently where he has been working. African ideas of what constitutes a long distance vary in these days with the experience of the speaker. In two cases at Kaphuka's people said " We never see " somebody living less than a quarter of a mile away. An informant' at Kachere's spoke of the village from which a son-in-law had come as " a long way away-two miles; " while at Kaphuka's a man who had spent several years in Southern Rhodesia, asked if he had " come from somewhere else to marry here" (the usual way of putting the question in Cinyanja), replied, " Not really somewhere else," and indicated a village about a mile off. The great majority of marriages are between people from homesteads within easy walking distance; but the idea that even a mile or two is a long way for a man to go from his parents is reflected in the preference for cross-cousin marriage, which is always commended by its advocates on the ground that the man need not leave home. 

A village is, of course, a territorial unit, but in the neighbourhood of a Native Authority headquarters villages may be -so close together that it is not easy to know when one crosses the boundary. What is known to the authorities as Kachere's village (and still referred to by the name Maonde, given it by his predecessor, which should actlually not be mentioned now that he is dead) consists, in the eyes of its inhabitants, of three separate units, Liwenga to the east of the chief's court and Matope and Tampangani to the west. The dividing line between Matope and Tampangani is imperceptible.


In addition there were 13 cases where the spouse's home was described as "N.A. Tambala's." As the boundary runs within a few hundred yards of Kaphuka's court-house, some of these may fall within the one-mile radius. Out of the total of 73, 22 were cross-cousin marriages.

In the majority of cases, therefore, the removal of the wife, though it means that the couple' are freed from close control by her parents, does not mean that she is taken out of reach of personal contact with them. The desire of parents to keep their sons as well as their daughters near them results, as- in other societies with matrilineal descent or matrilocal marriage, in a high proportion of cross-cousin marriages. In such marriages the couple often live with the husband's parents from the start, since, it is argued, his character is already known and there is no need to put him through a probationary period before entrusting his wife to him.

Whatever may have been the case in the past, or elsewhere today, in Kaphuka's village it is regarded as normal for children to be brought up by their own parents, and unnatural for them to go to the mother's brother while the parents' marriage is in being. The matrilineal principle is not carried to the point of denying the father all authority over his children. It is definitely he, and not the mother's brother, whose consent is required for their marriage, and for the removal of a daughter by her husband. It is clear that the modem tendency is for the father's authority and responsibility to increase relatively to that of the mother's brother. It is less clear whether in the past he was as completely without it as has been alleged among other peoples of Central Africa.

A man who has more than one wife and has not established his own homestead (or does not wish to have his wives live in the same place), visits them alternately for a week at a time. If they are inconveniently far apart, it is said that he should spend a month with each in turn. I do not know how far this rule is actually observed in the rare cases to which it applies; it would clearly be impossible for a man who looked to European employment as his normal source of cash income, and it is only such men who are likely to contract marriages with women living at great distances apart. Under this system of alternate residence the children are free, for part of the time, from their father's control: but today this is of slight importance in comparison with the long periods of temporary orphanhood which they experience while their fathers are away at work in the Rhodesias or South Africa.

Both among patrilineal Ngoni and matrilineal Cewa the constant element in the homestead is a mother with her daughters. But one would rarely find a homestead in which the nucleus of kin consisted entirely of women and all the male members were affines, since it is every man's aim to return to his own kindred.

To the Ngoni the kindred are the father's kin, and return to the kindred means return to the father's homestead. With the Cewa the situation is rather more complicated; a man belongs by descent to the group of kindred from which his mother has, in most cases, been removed by her husband, and when he in his turn is given permission to remove his wife from her parents, he takes her to the homestead of a brother of his mother. Later on the mother may rejoin her son, either on her husband's death or because she leaves him for some reason. In many of the cases in my records, however, adults living on the home- stead of a mother's brother had come there along with a mother who had subsequently died.

It has been pointed out before that a characteristic feature of peoples where marriage in its initial stages is matrilocal is the changing composition of the homestead. This appears most strikingly on the death of the head. Whereas among peoples with patrilineal descent and permanently patrilocal marriage the only person affected is the widow, and even she may not leave her husband's homestead, with the Ngoni and Cewa the whole group may break up, the widow returning with her daughters to the village where she was born and the sons either going with her or attaching themselves to other relatives as they choose. The case-history which best illustrates this was obtained at Kachere's, and is of interest because it appears to involve a scattering of the members among their paternal kin. The narrator, Bisenti, was the son of a village headman. On his father's death the children found that his successor " did not love us" (a situation regarded as typical with the Cewa where the successor is a sister's son). The widow returned to the village where she was born. Most of the children went to another village, Ncinji, where they had relatives, not specified. Bisenti however joined a sister of his father who was married at Liwenga. On the death of her husband she rejoined her nephews at Ncinji. Bisenti meanwhile had married a Liwenga woman, and when his aunt left the village he took over from her a large plantation of bananas.

Though there are cases among the Cewa where a couple live with the wife's maternal uncle although her mother is still alive, one finds more commonly that the married woman living in the homestead of a kinsman has either been born there or come there as a child with a mother who has since died. In other words, it is not now common for children to be sent to a mother's brother while their parents are living, though they may be left with maternal relatives if their mother is divorced and re-marries.

PREPARATION FOR MARRIAGE

With the Cewa, as with other Central bantu tribes, it was customary in the past for girls to be betrothed before puberty. From the time of betrothal limited sex relations were permitted between the couple, but the girl must not be deflowered until the initiation ceremony, which was held as soon as possible after her first menstruation. This ceremony was linked so closely with the marriage as to be almost a part of it; the husband provided part of the cost in the form of presents to the women instructors, and married life began immediately afterwards. It does not appear from the scanty records that domestic relations between the couple began from the time of betrothal as with the Bemba. A special feature of the Cewa ceremony was that the girl must be deflowered by a man other than her husband, known as the fisi (hyena).

Read describes the disgust with which the Ngoni view these practices and the strict watch formerly kept over Ngoni girls to ensure that they remained virgin until the relatively late age of marriage customary with them. This however is a thing of the past, and today the essential difference between Ngoni and Cewa is that the former have no initiation ceremonies. Ngoni chiefs and village headmen sometimes refuse permission for the dances, which are an essential feature of the Cewa ceremony, to be held.

Cewa custom has also been much modified. Mission teaching condemns the whole complex of early betrothal, sex play before marriage, and the initiation ceremony, which is perhaps too readily assumed to consist wholly in the stimulation of interest in sex. The few informants with whom I could discuss this were " school people," and they assured me that girls are not now bethrothed before puberty. Cases did come before the native courts, however, in which a man would support his claim to be a woman's rightful husband by the assertion that he had paid for her initiation, and one or two in which a woman, seeking divorce from an absent husband, said, " I am grown up now, I want a child." There was also a case in which a husband was awarded damages for adultery from the man who had performed the fisi custom; but it was alleged in this case that the latter was the lover of the girl's choice.

The same informants who condemn the dances and forbid their children to take part in them describe them as the feature par excellence that distinguishes the Cewa from other tribes in this region, where tribal intermixture seems to have led in many respects to a general assimilation of custom. That " We dance the Zinyau" is what marks the Cewa as Cewa. The Zinyau are dances performed by men at the girls' initiation ceremonies and at the conclusion of periods of mourning. The fiction is maintained that the dancers are wild animals, and the measures taken to preserve this give them some of the characteristics of a secret society. I have seen them at a Yao girls' initiation in a village not far from Kaphuka's, but I witnessed no Cewa ceremonies. Informants stated that they had been dropped at Kaphuka's itself because of the irafluence of the Dutch Reformed Church Mission, exercised through two resident teachers, because no new mistress of ceremonies had been appointed since the last one died, and because parents could not afford the expense involved. It appeared that an initiation had been held a year or two before.

The value of these ceremonies could be briefly indicated as follows. Through them girls received formal instruction on the main principles of conduct accepted in Cewa society, including the duties of a wife, and also, according to Marwick, information known only to women; such " revelations " whatever their actual contact, emphasise the entry into responsible adulthood. They fixed a minimum age for full sex intercourse, and forming, as they did, the immediate preliminary of the marriage ceremony, they bound together the beginning of adult sex life with the entry into legal wedlock. They reinforced the view that girls should be technically virgin at marriage and above all should not bear children before it. A magical belief which supported this view was that full intercourse with an uninitiated girl led to -sickness of a supernatural origin, and that for such a girl to bear a child was fraught with disastrous consequences. European influences have been opposed to the view that marriage should follow immediately upon puberty, but the age for initiation ceremonies has not been postponed along with that of marriage, since the aim which is regarded as of crucial importance is to ensure that uninitiated girls should not.conceive. Indeed the Yao have thought it necessary to advance the age of initiation further and further as the age of marriage is postponed.¹

At the weddings of adherents of the Dutch Reformed Church a substitute for these ceremonies is provided in that, when the couple enter their house, they are harangued by an elder of the church, and older relatives of both may come forward and accuse them of past misdemeanours (notably failing in respect to their elders).

BETROTHAL

Informants insist that today the choice of a spouse is absolutely free; it would be impossible, they assert, for parents to impose their will. They imply that this was done, in the past, only in connection with the marriage of cross-cousins.² But although it is true that today no young man or girl could be forced into an unwilling marriage, the choice of a partner is rarely based on close personal acquaintance, for the opportunities for unmarried people of opposite sexes to develop such acquaintance are almost as strictly limited as they ever were. Normally the men and women of a homestead gather in separate groups for conversation or other sedentary activities such as cobbing maize, and in any large concourse such as the audience on a native court day men and women form separate groups. Considerable embarrassment was caused on an occasion when I was photographed with some Africans by the suggestion that one man should stand with the women, on the other side of me from the rest of the men.

A youth marrying for the first time generally chooses a girl he has seen, but it is still considered necessary to make enquiries about her character, and unwise to decide without this. He may himself question the women of the homestead or leave it to his parents. Some still ask their parents to find them wives. The consent of the parents to the marriage is still indispensable, as without it the necessary procedure to make it legal cannot be carried out. The head of the mbumba could also refuse his consent, either because he disliked the character of the proposed bride or bridegroom, or because he did not wish to be allied with the family concerned. Marriage prohibitions appear to be confined to the children of siblings of the same sex.

The marriage procedure is now almost identical for Ngoni and Cewa in the area where I worked. The Ngoni have largely abandoned the cattle payment, though court cases arising out of it show that it is still sometimes paid, usually by chiefs and their close relatives.³ An elderly Ngoni informant argued against it on the ground that it enabled husbands to take their wives to their own homes and so " destroyed the village." - He also asserted that only one man in his village had paid it and that his wife had run away. The word for the transaction, kuloola, is clearly connected with the Zulu ukulobola. Cattle marriages will not be discussed in this paper.

The first approach is made direct by the young man to the girl If she is willing she obtains her parents' consent. Then each side must find two " marriage sponsors " (ankhoswe). The formal meeting of the ankhoswe, who have a status something like that of diplomatic representatives, is the essential act that makes the marriage legal. One of the two is usually a maternal uncle, but there are exceptions. In one of the two marriages Which took place during my stay, the bridegroom's ankhoswe were his own brothers and the bride's were both young men. I suggested to the heads of both households concerned that the ankhoswe chosen were very young. Both replied, " As long as I have my strength, I will see to anything that goes wrong." A woman can be a nkhoswe, but this is rare. Cross-cousins are said to be allowed to marry without ankhoswe, no doubt because the close relationship of their parents makes go-betweens unnecessary.

One informant insisted that the reason why ankhoswe are indispensable is that, if either party falls ill, there must be some formal channel for letting the relatives know. In the days when all illness was treated by divination and frequently ascribed to sorcery it was considered essential that a person's own relatives should be present at this process; it was probably for them to engage and pay the diviner, though I have no data on this. From Mr. Marwick's work it would appear that this is commoner today than my Christian informants would have liked me to suppose.⁴ If a person dies without his kinsmen having been told that he was ill, this is taken to suggest that his death was sought by his relatives-in-law, and someone who refuses to go to a sick relative is guilty on his part of a serious breach of obligation. If, as does some- times happen, a couple had set up -house together without the consent of the relatives on one side or the other, a messenger bringing the news of such an illness would be treated as nul et non avenu. The relatives, according to my informant, reply, " We do not know that our child is ill; " the practical con- sequence being that they do not go to see,him, and, if he dies, hold the spouse's relatives responsible. " Further," they say, "we do not know that our child is married. Did you see us at that marriage, to summon the ankhoswe (kugwirizira, associated with kugwira, to seize, but only used in this context)? If not, you know that we have not- agreed on it with you, you arranged it by yourselves. You should have told us in the first place, so that we could agree and nominate the ankhoswe. Now don't tell us about this sickness, it is your affair, see to it yourselves; the one who is ill is your child, not ours." In a court case where a couple had been living some years at the homestead of the woman's relatives without going through the marriage formalities, the latter, who, as he pointed out, had seen him every day, used the standard formula, " We do not know this man."

In practice the important duty of the ankhoswe is to deal with what are always called " difficulties "- in other words, matrimonial quarrels., If a wife living in her husband's village leaves him and goes home he must not follow her but must send the ankhoswe, who put to the woman's ankhoswe the standard complaint, " We have no one to cook for us." In one case in my records, the woman's ankhoswe made the first move, asking the man's why they did not come for the wife. The wife on her side should not go home without telling the man's ankhoswe, who would then- of course try to patch up the quarrel; Kaphuka's court required a woman to pay £5 damages to her husband for omitting to do this (no doubt, as so often, adultery was presumed-it may even have been admitted, though the record does not show this). Both ankhoswe need not act in every case; the presence of the senior is essential only if divorce is under discussion. Ritual offences, quarrels involving insulting words or the use of excessive violence, or a single act of disobedience by the wife, all of which can be settled by the payment of compensation, may be dealt with -by the junior, who should tell the senior what he has -done. In Kaphuka's court a man told how his wife turned him out of the house at night and threw his sleeping-mat after him; he went to her nkhoswe there and then. After a divorce the ankhoswe are responsible for getting back any property that the man may claim. When a matrimonial case is tried in the native court the ankhoswe must be present, and their names must be entered on the marriage certificates which the native courts have issued since 1944.⁶ On one occasion a woman  brought a case for divorce against an absent husband whose nkhoswe was dead. Kaphuka was reluctant to pronounce the divorce, but the elders persuaded him to let the Village Headman act ad hoc, so to speak. A nkhoswe has been known to " resign " an digust at the trouble a couple are causing, and an informant at Kaphuka's told me that he had ceased to consider himself nkhoswe for a woman who had gone off with her husband without asking anybody's permission. There are also cases where an heir will not accept the nkhoswe obligations of his predecessor.

On obtaining the girl's consent the suitor makes her a present of two cloths, valued, today, at £1. This is called cikole or cigwiriro; " like a ring," said an informant, who also said the latter word meant evidence of a transaction. This might be considered an expense not to be incurred lightly, but it may happen that he then goes no further with the affair. If the girl decides to give him up and take someone else, he may claim the price of the cloths from the new suitor. iBut he may not assert a claim to the woman against the latter; a man who took from her husband a woman for whom he had first paid the cikole was adjudged to be in the wrong. The relationship initiated by the gift is called bwenzi, "friendship," a word also used of extra-marital associations.

The best evidence of serious intentions is to begin at once to build a house. This must be, for the first years of the marriage, on the homestead of the girl's father, who indicates to his future son-in-law the place to build it. The marriage takes place as soon as it is ready. If the betrothal is concluded during the planting season, the young man -should also do his share of work with his future wife in her mother's garden. This may prove to be a test he cannot pass. One of Kaphuka's wives told me she was such a terror -for work that she frightened away her daughter's suitors; and showed me the patch of tobacco that one had planted the previous year. When his betrothed, with her mother's backing, had remonstrated, " I don't eat tobacco, I eat grain," he had lost interest in the match.

One informant described the process of formal acts by which the domestic relationship is established, the bride saying to her husband, " Give me a water-pot, a pestle, etc.," as she begins to perform the corresponding activity on his behalf. But my attempts to find out in an actual case just what equipment the bridegroom was expected to provide were rebuffed with the explanation that the Cewa don't furnish their houses like Europeans. The irreducible minimum appears to be a sleeping-mat and a nsengwa (small shallow basket from which people eat maize porridge). For a certain period of time after her marriage the bride has no separate field or grain store, but helps her mother to cook food provided from the latter's store; the husband may eat with his father-in-law or alone. In the former case he need not even have his own porringer. For all other purposes the bride can use her mother's implements as she did before she married, and one statement made was that the husband is " ashamed " to buy such things while he is a mlendo (stranger or guest).

THE MARRIAGE CEREMONY 

My data do not indicate what is regarded as the standard pattern for pagan weddings today. It is clear that in some cases the marriage ceremony is still 'the final rite of the girl's initiation, and this may be commoner than my Christian informants wished me to believe. It is interesting in this connection that a marriage of Christians is called cinamwali, the word used for the initiation, and the couple namwali. Two marriages of Christians took place during my stay. At such a marriage the couple go to the church with their ankhoswe but not their parents, and are welcomed home by the relatives on both sides with feasting and rejoicing. The bridegroom's father should provide the feast. In one of the two cases, however, he refused to do so because, he said, he had spent so much money on building a new house for himself that he could not afford to kill a cow. The girl's father was very .angry, and he was generally considered very mean. It was shameful to save money ,on such an occasion, was the comment made. He himself explained to me that he could not have expected his son to receive enough in presents to recoup the price of a cow; also that the girl's father had borrowed money from him to pay for a cow for the marriage of a son of his, and had never repaid it.

The other occasion was the marriage of a son of Kaphuka, who was a mission teacher, to the daughter of another teacher.- For the feast two cows were provided, one by Kaphuka, the other -by a sister's son of Kaphuka who had named the bridegroom at his birth, thereby taking on a special relationship to him; this may be done on the initiative either of the parents or of the relative himself. This person took an active part in directing the preparations; the reason that he gave to me was that he was the only one who knew how a Christian wedding should be celebrated. He was on the spot an hour before the arrival of Kaphuka with another sister's son, the one who is expected to succeed him.

Among some 30 women engaged in preparing the food were a co-wife of the bridegroom's mother, and her daughter. Though I could not identify most of those present, there were enough people connected with the bridegroom through his father taking a prominent part to demonstrate that this is one of the occasions when, at any rate to-day, the father's position relative to his son is formally recognised.

The impending arrival of the bridal pair was announced by a messenger on a bicycle, riding at top speed, blowing a whistle and with green branches tied to his head. The women who had been cooking seized their stirring-sticks, others picked up pestles or porringers containing flour or grain, and there was a cry for hoes. Brandishing their implements or green branches, they advanced to meet the couple. When the two parties met they halted, and hymns were sung. Then a mortar was brought and set down in the path, and two women pounded grain in it, calling out, " A Kaphuka samadya gaga " (The Kaphukas7 don't eat grain with the husks), and then threw the flour over one another. The rest of the crowd called out, " Ngolima " (a form of the verb lima, to cultivate, which I have not heard elsewhere), with the names of the various plants, and some scattered grain or flour from porringers. All this time the couple stood motionless with eyes downcast, the bride holding a handkerchief to her mouth.

As the couple advanced towards the village two women threw grains of maize at them, again in token, so an informant said, that the wife must provide this for her husband. At the door of their new house they found two more women pounding grain; these were relatives of the bride, demonstrating that her family knew how to prepare food. The couple did not immediately enter the house, but sat down on chairs behind a table a little way to one side. The bridegroom's father sat beside him on a chair, his mother and one of her co-wives near by on the ground. The crowd gathered round and hymns were sung. A way was cleared through the crowd to the table, and a man who I was told was a church elder, and a classificatory father of the bride, whose own father-a teacher-was dead, came forward to preside over a collection, beating up the contributions almost as though it were a public meeting for some charitable cause. People came forward,and put small gifts of money-one gave a 5s. note, others shillings, Kaphuka himself six pence, some only a penny-in two plates, one for the bridegroom, one for the bride. The contents of the two, I was told, would be counted separately and then pooled to buy requisites for the house.

When the collection was finished the couple, preceded by another son of Kaphuka (not the bride- groom's own brother) carrying the plates of money, accompanied by the crowd, moved to the house amid more singing of hymns. Several persons went in with them, and this, I was told, was the moment when they would be instructed in their duties, reproached for past misdeeds or, if none were held against them, encouraged to go on as they had begun. There would also be a meal of tea and scones provided by the bride's relatives. For a quarter of an hour the rest of the men crowded on the verandah singing hymns. Then dancing began again. The feast was served about half an hour later; it was cold because the cooking had been started too soon, and was not very popular.

The wife's place in the domestic routine of the homestead is little changed at first. She still usually works in her mother's field; sometimes a girl is allotted her own field before marriage, but the marriage itself is not an occasion for this. Her husband should work with her, under the general direction of the mother. She helps her mother to cook the latter's grain, and her husband's portion comes from this.

This initial period should theoretically last two or three years, and in the past was associated with the marriage of girls at puberty. The allocation of a separate field with its corollary, a separate grainstore from which a separate food supply is drawn, was made "when the girl was grown up." Today informants lay more emphasis on the supervision of the husband by the wife's parents. It is for the husband, through his wife, who approaches her mother, to ask for " his own field " (it will never be his except to cultivate), and if they are satisfied with him, they will agree. It was also asserted that if they considered him thoroughly lazy they would take the initiative so as to make him dependent for food on his own labour. Today husbands are said to insist on a separate field from the start, and some obtain per- mission to take the wife away to their own village after two or three years.

It would be misleading to continue the description of married life as though the Cewa were still a peasant drawing his whole subsistence from his field and his game traps. The fact that the standard of living which is today considered normal involves the purchase of trade goods and even the payment of school fees obliges him to earn money. For the great majority this is only obtainable by wage-labour at a distant industrial centre, and many prefer wage-labour to growing cash crops, or find it more lucrative. The husband's duty to buy clothes for his wife is one of the essentials, failure in which justifies her in asking for a divorce. It is taken for granted that he will be away at work for periods of a year or two at a time, and so long as he sends money home no blame attaches to him for staying away even longer. In these circumstances there is a good deal to be said for a system which does not separate the wife from her own relations. It has been argued, as an advantage of patrilocal marriage, that male relatives of the husband are at hand to take his place in the essential work of the field and homestead, and also to watch over his wife's fidelity. With the Cewa, however, wives whose husbands are absent can sometimes look for help to their own brothers, and do not find themselves in an alien group, their only link with which is missing; they clearly do find infidelity easier, as their relations often raise no objection to their taking up with another man and they may even contract a new marriage without formally divorcing the absent husband.

It is not considered right, however, for a young man to go away to work between his betrothal and his marriage or for a year or two after his marriage. During this time he should be in the village proving his quality under the eyes of his wife's parents. " One who wants a marriage does, finish to build a house and all duties of the family, both gardens as well."

It is distorting the facts to assert, as is sometimes done, that young men go to work abroad largely in order to avoid their family obligations; as I have indicated, they must do so in order to meet these obligations. But it is common ground that employment away from home is an easy refuge from the consequences of seducing an unmarried girl. It has not been so much emphasised that the obligations involved in matrilocal marriage of the Ngoni-Cewa type may conflict with the economic ambitions which we are so anxious to instil into the African.

A case which occurred at Nkhaza illustrates this. E. D., a church elder (an Ngoni), had a daugher whom he had brought up very strictly. He gave his consent to her betrothal with one James K. Before any of the formalities had been concluded, the latter went off to work. The girl was then found to be pregnant. Her father asked K.'s relatives to nominate ankhoswe, and he wrote to the same effect. They refused to produce sureties for one of his character, and E. set his face against the marriage. K. was ordered, in his absence, to pay a fine of £3 and £5 compensation to E. Eventually he got a week's leave from the printers where he worked, and the matter was brought into Kaphuka's court. He himself was anxious to marry the girl; her father retorted that he would simply abandon her again. The court elders exerted all their influence in favour of the marriage, one urging the disgrace of having children of several mifunda. The girl's relatives were sent away to think it over, and when they came back E. said he would agree provided K. remained in the village and began at once to build a house. Next day it transpired that he had gone straight back to work, as he obviously had to do if he were not to lose a good job. E. was very angry, and prepared to call the marriage off if money for building the house was not sent very soon. A solution by the girl joining K. in Lusaka was considered unthinkable; she would not dream of disobeying her father.

THE SON-IN-LAW

The position of the husband among his wife's kinsmen could not be better illustrated than by the answer given me on my first visit to a homestead by the first man I spoke to-" This is not my home, I'm only a husband" (strictly mkamwini, son-in-law). In the past he was expected to treat his wife's parents, particularly her mother, with formal marks of respect. If he met his mother-in-law in the path he would step aside not to pass close to her, and say "Zikomo," the formal greeting, too often translated "Thank you." He must not go up to her in the homestead without being called, and must never touch her or sit on the same mat. These rules are neglected today, though some people, 'including Christians, say they still observe them. The avoidance can be formally terminated by the gift of a fowl; in this either side may take the initiative.

The husband naturally looks forward to the greater independence of life with his kinsmen, if not of founding his own homestead, and the final proof that he has earned the confidence of his wife's parents is given when they agree to let him take her away. They must be satisfied that he will care for her properly. According to Marwick, she must also have borne him a child; my informants, in lengthy statements of the rule, did not include this point..⁸

The cases which I recorded are too few to give much indication of the length of time spent by husbands in the wife's homestead. One man had brought his wife home after 17 years, but she had later gone back to her own village and only visited him occasionally; he had not taken another wife, and apparently did his own cooking as well as hoeing, though a young girl on the homestead pounded the maize for him. Two wives who had come to Nkhaza with their husbands said they could not remember how long they had been married. One who had arrived the previous year said, " Since I was little;" the other had been six years in the'village, but it does not follow that she was married even longer at her own home. Two men had come home after two years, another after three. I noted one youth who was still working with his father-in-law after two years of marriage, and another who had cleared his own field for the first time in the fifth year.

If the parents refuse their consent it appears that the marriage frequently comes to an end; the son-in- law, once he has decided to go, is not willing to await their pleasure, and open defiance of their view is not usually considered a permissible course. Sometimes the wife will follow him; if her parents protest he will deny responsibility, according to informants, and will not keep her with him if they insist on her return. Nevertheless, on one occasion Kachere ruled that a husband might take his wife in spite of her father's refusal.

Marriage of this type has sometimes been called "trial marriage." I doubt if this is an appropriate description. The husband is not on probation as a husband, but as a person to whose sole care the wife can safely be entrusted. The wife's parents have no right to turn him out, only to refuse to let him take his wife away; if he consents to stay, the marriage con-- tinues in being. For the description " trial marriage " to be justified it must surzly be generally accepted that- the marriage as such has only been conditionally agreed to. I saw no sign that the Cewa regard marriage as conditional any more than other peoples among whom divorce is easier than it is-with us.

The significant feature of the system is that the wife's loyalties to her own relatives remain as important after marriage as they were before. Hence there is scope for conflicts of a type which is denied expression under the system of patrilocal marriage. They can go to the length of breaking up a marriage against the will of the woman. This may or may nof be the position when the relatives refuse to let a husband take his wife- away. A more complicated case was that of a man who was accused of neglect because he did not build a grainstore for his wife. His retort was that he would do so if her family would let him; they prevented him because they had always disliked him, whereas she herself had liked him enough to go with him when he worked in the next District. (By the time the case was brought, however, she had gone to another man.) Or the wife may give filial duties precedence over those towards her husband, like the one who refused to go to his village because her mother was old and would have no one to look after her; in this case, again, there was no love lost between the couple, but the. woman had a socially approved argument on her side.

There is also room for a conflict between economic interest and marital duty. A woman living in her own yillage is allotted a garden by her parents; often it is a portion of her mother's garden, which she and her husband may extend by clearing adjacent bush. She must leave this behind when she goes away with her husband. But there is a strong feeling that land should not " pass out of the family," and if, when her mother dies, there is no daughter in the village to take over the garden, she may return to prevent its being "hoed by other people." The husband is then justified in complaining that he has nobody to grow food in his village. Kaphuka's court tried to effect a reconciliation in a case of this kind, on the basis of a payment of compensation by the wife, but she refused.

Here too the traditional rules do not fit the situation of the modern breadwinner. An example was the case of M., who married a man from near Zomba when he was stationed near her home in the K.A.R., and went with him when he was discharged to Zomba, where he worked as a carpenter. He refused to live with his wife at her home and proposed that she should live with his mother; she went to her own mother to bear a child, and her mother-in-law came with her, expecting to bring her back again. Her mother would have agreed, though unwillingly, to her return: " the Europeans are always telling us,"' she pointed out, " that you can't eat money," and she wanted a son-in-law who would work in the fields. But the child died, and the husband ignored a telegram sent him when M. was dangerously ill. This- was regarded as the last straw, and M. is now considered to have been deserted.

Another case was that of Nelson, a foreman on an estate some forty miles from his wife's home. He had fulfilled all his obligations to her, though he had only been able to live with her a small part of the time, and he complained that there was " nobody to cook for him " while he was at work. Her relatives refused to let him take her away; he brought the case to Kaphuka's court, which offered him a divorce. Although the same court had put pressure on E. -to agree to his daughter's marriage they did not try to induce Nelson's relatives-in-law to waive their rights. A comment made on the case was that the latter had behaved badly; they were contrasted with people who let their daughters' go all the way to Salisbury. In the five months' records which I went through at Kachere's there was one case in which the court ruled in favour of a husband whose father-in-law wished to prevent his faking his,wife away.

OBLIGATIONS OF MARRIAGE 

The obligations which marriage is held to involve become clear from the claims which spouses bring before the native courts. The husband must provide his wife with sleeping-mats and clothes, clear a field for her to plant, and keep her -house and grainstore in repair; if he is away at work, sending money is regarded as the equivalent of these duties. If he has more than one wife he must divid-e his time equally between them; when he returns from work he should visit the senior first. While he is living at his wife's home he should not go away to work without letting her -relatives know. In one case where this happened, the mother- in-law indignantly told me that the youth was a bad lot and his action had put an end to the marriage; the nkhoswe, however, said he had written begging pardon and promising to send money to his wife.

A wife must fetch firewood and water, keep the fire going, cook for her husband and care for him if he'is ill; one man, bringing a case against his wife for neglect, complained that he had to burn the house- posts to make a fire. She must also of course work in the fields, but failure in this respect is not made a ground of complaint in the courts. She should not leave home without his permission and without telling himyvhere she is going; the presumption, a fairly safe one, is that if she does she is with some other man. Failure in any of these obligations may, be made a ground for divorce, which may also be obtained if either party has venereal disease or leprosy. A woman may obtain a divorce from a man who cannot give her- a child. Divorce may possibly be obtainable underr native custom for breach of ritual obligations, but the Dedza courts have been instructed not to take such cases, and I have little information on them. In one case in Kachere's records a divorce was grarited because the man did not come to the funeral of his child.

The husband's right to beat, his wife is accepted by all, including Christians, but it must be exercised within certain limits. One Christian informant told me that a man should practise restraint when his wife is pregnant, " and not kick her; ' another that one should strike a woman only with the open hand, not with the fist, and kick her only with the bare feet, and that beating with astick would be a serious offence. Men are sometimes fined by the native court for beating their wives: the reason may be considered to be insufficient or the injury inflicted excessive. Reasons recorded for such quarrels are that the woman's child ate an egg of one of the man's fowls; that she was late in preparing the evening meal; that " she swore at me when I swore at her " (this involves the use' of impermissible language, notably mentioning the names of dead persons); that she refused to have sexual intercourse in the daytime (in this case the husband maintained-that she went off to drink beer, and the court rebuked both sides and dismissed the case). In one case a divorce was demanded because the husband was constantly beating his wife, but the court refused it on the ground that the woman had a lover; this presumably justified the beating. Of course no amount of admonition from ankhoswe or court will keep in being a marriage which one party is determined to break. There is no effective sanction against cohabitation without marriage; the cases of " wife-stealing " that come before the courts end in divorce if the wife prefers her lover to her husband, as is frequently the case, and those of adultery sometimes also have this issue, though they are brought by husbands primarily with a view to obtaining damages. Women do not however seek divorce without alleging some wrong on the part of the husband; men do not seek it, but may obtain it by refusing to support a wife.

MATRIMONIAL CASES IN THE COURTS

 The courts deal with accusations of adultery and "wife-stealing " (in practice, usually, setting up house with a woman in the absence of her legal husband at work) and with demands for divorce. These last are often made by women whose husbands have been long absent and who wish to be free to make another legal marriage, but they may also arise from some irrecon- cilable difference between a couple, which is expressed in persistent failure to perform marital duties. In almost every case where a divorce is granted for any other reason than desertion, the woman is required to pay a fine. This, I was informed, is: because her reason for bringing the case is that she has a lover and the fine will in fact be paid by him; it therefore represents the fine payable for adultery. It may be that in some cases the judges know this to be the fact; but while I admit that I was never able to follow every detail of a court case, I did not have the impression that any great pains are taken to establish it. Two. remarks by court elders (as recapitulated shortly afterwards by Ganizani) illustrate my point. " This is a simple case," Kaphuka said once, " because I know the woman is no good." And the other: " We have known some time that you were deceiving your husband. We heard that at Christmas a man brought you home from a beer-drink on a bicycle, and no doubt he is your lover."'

As in many parts of Africa, the great majority of the cases handled by the native courts are matri- monial. Out of 330 cases tried by Kachere's court in the first five months of 1949, 73 were for statutory offences, 52 for divorce, 48 for adultery or attempted adultery, and 19 for wife-stealing. Of 183 at Kaphuka's in the first six months of the same year 26 were for divorce, 6 for adultery, 16 for neglect of a wife by her husband, one for neglect of a husband by his wife, and 14 for wife-stealing. Where a husband is accused of neglect, the wife and her relatives may hope to induce the court to direct him to carry out his obligations, but there are not many cases on record where this is done. Kachere's court was much more inclined to do so than Kaphuka's, and sometimes refused to accept a'husbanid's plea that his wife was idle or unfaithful. Frequently the husband says, " I am tired of mitala " (the duty of alternate residence), or, " Two wives are too much for me," and the court then pronounces a divorce. In one case at Kachere's, the husband divorced one wife in order to return to the Catholic communion. In these cases it is naturally apt to be the senior wife who is divorced.

The migratory labour system is responsible for a large number of the claims by men for " wife-stealing" and nearly all of those by women for divorce on account of desertion.

The great majority of men in Dedza spend a large part of their time working for wages at the big industrial centres of the Rhodesias and the Union. Those who are recruited on contract for the Rand mines must be sent home at the end of two years, but those who make their own way and find their own employment may stay- away very much longer. Legislation has recently been introduced requiring the repatriation ofLall migrant labourers at the end of two years, and also making compulsory- the remittance home at fixed intervals of a part of their. wages.

Up to now, however, it has been open to the labourer to send home as much or as little money as he pleases. Large sums of money are in fact sent home; but some men do not make remittances and some wives set up house with other men. Such an action would always be defended on the ground that the husband had deserted the wife. If this is a purely private arrangement it is called " stealing the wife." But it is fairly common for the wife's relatives to agree to a formal marriage with ankhoswe although the existing marriage has not been dissolved; the offence is then " divorcing him behind his back " (kusudzula m'tseri).⁹ The man's ankhosw'e should watch his interests and may write to let him know what his wife has done, whereupon he will sometimes come home to bring a case against his successor; he rarely gets his wife back, but is usually awarded compensation of the order of £3. It is more correct for the woman's ankhoswe to approach the man's and ask them to agree to a divorce; they will do so if they too have lost touch with him, but they can insist that no decision be taken till they have written to him, even if he has been away five or seven years. Two cases are on record at Kachere's where divorce was granted after an absence of two years. But the courts generally follow the principle in force elsewhere in southern Africa, that three years' desertion is necessary to justify divorce, but is in itself a sufficient ground.

Applications for divorce are rarely refused. In one or two cases in Kachere's record, a woman's demand was considered to have insufficient ground and she was required'to pay compensation (either for adultery,  admitted or presumed, or as the appropriate way of settling the quarrel which had led to the demand). Kaphuka refused a divorce for neglect to a woman whose husband was lame, and until he became so had supported her; and ordered her to pay £5. He also refused divorce on another occasion when a woman complained of neglect and the husband wanted to divorce her rather than support her.

While modern economic forces have created a situation in which occasions for divorce have greatly increased, another modern influence is exerted in the opposite direction, that of the Christian missions. Their policy is not uniform. The Catholic missions do not countenance divorce in any circumstances, and even, so informants told me, require proof of death before they will permit the re-marriage of a woman whose husband has been lost sight of " at-the mines." The Dutch Reformed Church Mission allows divorce for desertion and also for repeated adultery b$y the wife. When a case involving a church marriage is brought into the native court the court informs the mission concerned, who give their views on it; if they are opposed to the divorce the court should not grant it, but I was told that they would divorce a couple who were constantly quarrelling, even against the mission ban.

It would be impossible to estimate. how many Christians resist temptation to infringe the sexual code inculcated by the missions, particularly in those respects where it is stricter than that imposed by native custom. There are certainly many cases of church members who are suspended for such offences; but also some who later seek re-instatement-for example, by putting away a - second wife. An informant described how his wife ran away but later wanted to re-join the D.R.C.M. She tried to conceal that she was married, but this was known to some of the church elders, who required her to return to her husband, and also wrote to him bidding him return to his wife, which he did, leaving another woman whom he had married when the first left him.

In one case where a- woman repeatedly left her husband and demanded a divorce, and the ankhoswe on both sides were willing to agree, the husband refused because there was no ground that would satisfy the mission; for this he had to wait till an adulterer could be identified. This was one of the cases where a nkhoswe refused to take any more responsibility and a younger kinsman had to take his place.

In nearly every matrimonial case the party adjudged to be in the wrong is required to pay a fine to the court as well as compensation to the injured party. At Kachere's, adultery was punished by £1 fine and a cow to the husband, £2 and two cows or £3 and 3 cows. For wife-stealing the penalty was usually £3 and 3 cows to the husband, sometimes £2 or £3 15s. Od. and 2 cows. Women who asked for divorce were frequently ordered to pay £3 (for proved or presumed adultery).

At Kaphuka's, where the Court Clerk prided him- self on the fact that the court revenues were the highest in the District, the rates were even higher. A woman who went home without telling her husband's ankhoswe had to pay £5 compensation; there are cases on record of women being required to pay £4 on getting a divorce, £4 and £5 for bringing unjustified demands for divorce. Adultery is punished with a fine of £3 plus £5 compensation, and in one case in the record, in addition to this sentence on the man, the woman was required to pay 30s. to the court and £2 to her husband; for adultery with the wife of a village headman £7 compensation was ordered. A man who abandoned his wife and did not reply to her letters asking for divorce was fined £3 and £5 com- pensation " for leaving his wife without cause." About half of these fines are recorded as paid. Few women could raise such a sum except by frequent brewings of beer, but penalties imposed on women are expected to be met by the presumed lover.

This system of fines and compensation has been introduced by the native courts in the vain hope of reduciing the number of breaches of marriage obliga- tions. Some people seem to interpret it to mean that a woman can now be freed from her husband by a money payment; one woman who was brought before Kachere's court alleged that her second husband had " divorced her with a penny " and produced the penny as evidence. At Kapjhuka's another woman, rather more au fait with the state of the market, tried to " divorce her husband with £2," and when he did not accept this as-freeing her, showed her indignation by rooting up his tobacco crop and destroying a sewing machine. In another case a man who had been fined for " stealing a woman " proceeded to marry her though her first husband had not divorced her 'the case came into court because he subsequently left her and she wanted to be freed from him (the court dealt with this as an ordinary application for divorce). It appeared that he had regarded the compensation which he paid to the first husband as " a payment in lieu of divorce," and the wife was seeking to gain her freedom by paying the same amount to his ankhoswe..¹⁰

That some Cewa reject this attitude towards marriage is evident from the occasions when a decision involving a money payment is indignantly described as " like buying a person."

Some of the arguments used suggest either that there is now great uncertainty about what the law is or that people hope to get through the courts something other than their legal rights-an attitude which of course is consistent with the idea of the chief's court as a place where differences are reconciled on the basis of whatever compromise the parties will accept. A youth, for instance, brought a girl into court for disobedience when she was not even married to him; and there are cases where a woman's relatives side with an adulterer or " wife-stealer " against her legal husband because the former has more money.

When a marriage is dissolved young children go with their mother, but the father may recognise his responsibility by sending money or clothes for them and even paying for their schooling. Kachere's court when granting a divorce sometimes required the man to pay a stipulated sum " to clothe the children," and in one case he claimed that by doing so, he had established his right to have them live with him, a right which his wife's relatives contested in the court. If the mother's new husband takes her to his home, the children may be left with her relatives. When they are old enough to choose-according to Ganizani, at about ten years old-they may return to their father "if they think he cares for them better."

Everything that a husband has bought for his wife is regarded as his property, and in case of a divorce where she is at fault he claims it back. His right to all household goods is generally accepted, but to try to take her clothes from her is not well thought of; " he forgets that she has worked hard for him." If there are no children he is entitled to the food in field or grainstore; if there are, he should allot some to them, and should also leave some of the household objects for them (four out of thirteen teaspoons for two daughters is an example of an actual distribution). The division is made by the nkhoswe in the presence of an elder -of the native nourt. A surprise visit is considered good tactics, lest the wife should conceal the property. At Kaphuka's a divorced wife brought a case against her former husband because he had harvested and removed the whole maize crop; the court awarded half to the wife, but she considered this unjust, as she maintained that she had worked- the field alone.

DEATH OF A SPOUSE

Marriage in Dedza District is not immediately dissolved by death. If the wife dies while still at her parents' home the husband is not free to leave until her kinsmen have given their formal permission, which should be after about three months; this is called kusudzula, the word for " to divorce." 'He may be offered a sister.

The wife at her husband's home is in the same position. It is not clear how far it is held to be the duty of the dead man's heir to offer her his protection, but she is not automatically inherited. In Kaphuka's village there is living with her daughter a widow of the late Kaphuka who, I was told, refused to marry the present one, and I have heard one of the present Kaphuka's wives discuss in a bantering way with his presumptive successor whether she will marry him. Another man in the village told me he had taken on the widow of his elder brother because " I ought to look after her." In comment on court cases 1 learnt that if a widow is held in especial esteem she may be invited to choose a new husband among the dead man's relatives (kugana). If an outsider wants 'to marry her before they have released her he must seekl their permission, or she herself may do so. Otherwise she " just waits." It is again a sign of esteem that she should not be quickly dismissed, but if the release is delayed too long she will resent it and may appeal to the court. Such a case was brought before Kaphuka by a woman who had waited three years.

Pagan ritual, on which I have insufficient information, involved the sacrifice of a fowl by the new spouse " to drive out the spirit." Christians refuse to do this, to the indignation of their pagan relatives- in-law, who threaten to bring them into court and have them heavily fined. But their teachers' concern on their behalf is misplaced, for the native courts do not hear such cases, and no village headman can enforce the payment of a fine.

THE FATHER AND THE MATERNAL UNCLE

The question of succession and inheritance can be most profitably considered in the wider context of the claims of the maternal kin as opposed to those of the husband and father. Today it is only in the rules of succession and inheritance that the customs of Ngoni and Cewa in Dedza are clearly differentiated, and it is only with regard to succession, where no compromise between the patrilineal and matrilineal principles is practicable, that there has been no modification in Cewa custom to meet the increasing recognition of the father's position in the family.

With the Ngoni a chief or headman is succeeded by a son or a brother; it can happen that where a brother has succeeded his heir will be, not his own son, but the son of the elder brother whose heir he himself is. With the Cewa he is succeeded by a sister's son, and it is fairly common for the heir to be on such bad terms with his cousins that they decide to leave the village and join maternal kinsmen elsewhere.

The disposal of a dead man's property affects commoners as well- as chiefs and headmen. With. the Ngoni the principle is simple; it is divided -among his sons. With the Cewa it is distributed by -a brother, who may make such modifications as he thinks fit in the traditional rule that it should all go to his sisters' sons. There is today great dissatisfaction with this rule, and a man's sons sometimes bring into court a claim for a share of the property. In one such case a son was allotted two cows which he had herded. Feeling is strongest about the inheritance of relatively expensive trade goods, such as bicycles or tools. It is assumed, no doubt correctly, that a man will-teach his sons the use of these, but also, perhaps less correctly, that his sister's sons will not know how to use them and will " just sell them.'" A case frequently quoted was that of a carpenter in Dedza, Township whose equipment, when he died, was all taken by his sisters' sons, so that his own sons were left without the means to carry on his trade; this caused especial resentment because, as the man had been long settled in the township, the heirs from his wife's distant village were more or less strangers.

A further contrast is in the position of children when either parent dies. In any village a man is either a mwini-the word so often inaccurately translated "owner "-or mkamwini, a son-in-law. With the Ngoni the son of a mwini is mwini in his father's village. With the Cewa he is mwini among ,his mother's kinsmen, which is apt to mean that he leaves the village on his father's death. If the mother dies, immature children may be brought up by a co-wife, but it seems to be considered normal for them to go to a sister of the dead woman, which would involve separation from the father unless he married this woman. The position is thus different from that in case of divorce, when both parents are living. Also, actual cases do not always square with this theoretical formulation. A man claimed in Kachere's court that his dead wife's relatives had harvested the maize from her field though he was left with a child to support.

A discussion of the principles involved arose in a very interesting way when G., a member of one of the newly created "group councils " of village headmen, proposed that from the 1st January, 1950, all Cewa should adopt the Ngoni rules of patriliny. His motive, as commentators explained, was a purely personal one. He was hoping to establish himself as a possible successor to the position of Native Authority. He had also quarrelled with his wife's kinsmen. Accordingly he decided to move from his wife's village and set up his homestead near Nkhaza, hoping that his twelve children, some of them adult and married, would come with him. His wife refused to move, and so did the children, so that his prospects of setting up as the head of a substantial "mbumba were shattered. It never occurred to him that the proposed change would rule him out as a successor to the Native Authority.

The case was argued on the usual ground-namely, that Ngoni loola their wives and Cewa don't. That most Ngoni no longer do so is beside the point. The principle that the cattle payment establishes the affiliation of the children is as clearly understood by the Cewa as it is by the anthropologist. In contrast the Cewa husband is described as " borrowed" (ngongole) or as tonde obwereka, a stud animal (literally, borrowed male). "'He begets children and goes home." But the conclusion reached is always the same-" We must follow the customs of our ancestors "-even though in the view of one speaker this is " a sore from of old that does not heal." 

How far it is true in any matrilineal society that the husband's status is as insignificant as these comments suggest is a matter for examination by anthropologists. There seems always to have been a good deal of divergence between the matrilineal Central Bantu peoples in the extent to which they recognise father- hood as a social relationship, but modern influences are all on the side of greater recognition. This recog- nition involves both claims and responsibilities, and the reciprocal relationship between them. Claims to the services of children are readily pressed, whereas there is not the same competition to pay their school fees. One might say that the maternal uncle is not usually prepared to do more for his nephew than he did in the old days, but will, if he can, make the same claims on him as before. This cannot be simply dismissed as a manifestation of original sin. It is a logical development of the father's traditional role. In the old subsistence economy the mother's brother was not called upon to make any specific outlay on behalf of his nephew except when he incurred a fine for some misdemeanour. The husband was specifically responsible -for economic assistance to his wife and thus indirectly for the maintenance of his children. When that assistance began to be given in cash rather than in labour it naturally fell to him to meet the increased cash outlay involved in the clothing and schooling of children; though the extension of this principle to make him liable for the same provision even after divorce is hardly implicit in the traditional system.

The claim of the maternal uncle on his nephew is usually for no more than an occasional day's work on building. Even this, of course, may, take him away from school, and a. more serious clash arises if his father wants to send him to school and his uncle wants him for a whole-time occupation such as herding. In a case of this kind in Kachere's court records the uncle based his claim on the fact that no cattle had passed for his sister, which suggests that the parties were Ngoni, since with Cewa the question could not arise.

The statement of Rangeley, that a father may not chastise his child without asking the mother's brother, was not confirmed by my informants, though one said that if it was a question of severe punishmen the would tell his brother-in-law (the one who was his wife's nkhoswe). If a youth got into serious trouble, say, over an adultery case, the father and -mother's brother would consult together, both would come into court with him, and, if the father could not pay the fine, the mother's brother would help. (In fact, if a surety is required for the payment of a fine, the court still always asks " Is the mother's brother there ? "). To this man these rules were connected with the fact that the mother's brother would be responsible for the man's children in the event of his death. In a later conversation he said that whereas in the old days it was right for the mother's brother'to have -authority over her children, today the mwini mbumba ought to recognise the mwini wobereka (literally, the genitor), the reason being that he takes responsibility for clothing the childien.

The latter comment was made on a case which came  up at Kaphuka's; a man had found his wife's brother beating his son and complained that he should have been told first. The reply was, " I don't know you, 1 only know my sister." A general fight ensued, and the child's father took out a summons against, his brother-in-law for " brawling" (ndeu). The court sided with the father, asking the uncle " Did your sister bear her child alone? " and variants on the theme: " We know he is your nephew, but you had no right to beat him on your own (kutsera, cf. kusudzula m'tseri) without telling his father. How could you tell the father you didn't know him? Is the child yours ? Who begot him-? You should have gone to the father and told him what the boy had done, and he should have punished him in your presence."

The degree of feeling that a man can have for a nephew was dramatically illustrated when a village headman urged that if his young nephew were flogged for stealing he would himself be distressed to the point of suicide, and offered to buy the youth off, an offer which was accepted.

The tie between father and son is strengthened, particularly where the father has paid for his son's education to a stage beyond the vernacular village school. " All school people," it is frequently asserted, "follow -the father." This is not solely because of the mission's insistence on this tie. The general principle that benefits should be repaid leads fathers to expect that when their sons begin to earn wages they should help to maintain them, and sons do accept this obligation. Ganizani told me that he had provided all his father's clothes since he began to earn wages. Kaphuka's clerk received a gift of £26 from his son in South Africa as a contribution to the cost of the new house that he had built. These contributions in turn create an expectation of return in the form of inheritance of the father's possessions.

 

GENERAL CONCLUSIONS

 

The general trends observable in Dedza District are thus in line with developments that have been noted in many other parts of Africa. An economic system that necessitates the separation of most of the men from their wives during the greater part of their young married life results in frequent marital infidelities, and casual changes of partner as well as formal divorce. Sanctions are imposed to counteract this tendency, by the native courts in the form of heavy fines and by the missions in the form of suspension from church membership. Both are largely ineffectual. The attitude of the courts implies the expectation that a woman should remain faithful whatever the difficulties, but, unlike some of the missions, the courts are prepared to release her from the bond of marriage when she has been effectively abandonied. Where this situation is so common it seems hopeless to expect that the ideal of a marriage dissoluble only by death can ever find general acceptance.

 

It is admitted that the traditional matrilocal system had characteristics which tended to make marriage unstable. Divorces occasioned by the parents' refusal to let the husband take his wife away may increase the total number, but in the present situation their significance is slight in comparison with that of forces which must militate against marital stability whatever the traditional rules.

 

Although it does not seem that the mother's brother among the Cewa ever had the dominant position which has been described among some other tribes, the relative importance of the father is certainly increasing, particularly as a result of his new financial responsibilities for his children.

Footnote

1. Mr. J. C. Mitchell tells me that that is the reason why Yao girls are now initiated some years before first menstruation.
2. The arguments for and against cross-cousin marriage are interesting. In its favour it is asserted that it prevents the son from leaving his father, and that the character of both parties is known. But against it is the argument that there is " no respect" between families who know one another so well. 'This was the view of an old man and a quite unsophisticated woman.
3. Possibly the " legalizing payment " required by the Native Authorities in Northern Rhodesia is regarded by the courts as a monetary equivalent. It is there required of both Ngoni and Cewa: it does not correspond to anything in Cewa custom, and it does not have for the Ngoni the legal consequences of the cattle payment.
4. Marwick, M. G., unpublished MS
5. Cf cigwiriro, a token.
6. I know of no case in which such a certificate has been invoked in a native court as evidence of marriage, but they are used where men in employment are entitled to housing or free transport for their wives.
7. The plural may be the frequently used plural of respect. Even then, however," it is of interest to note that, in this matrilineal society, the group who represent the bridegroom's relatives call the name of his father. Moreover, the line of people who could literally be called " the Kaphukas " (those who have taken the name on succeeding to the position of chief), is a matrilineal one in which the bridegroom at this ceremony has no place. This is a striking illustration of the way in which the position of the father is asserted in contradiction to the logical consequences of the matrilineal principle.
8. Marwick, unpublished report.
9. Sometimes they are genuinely unable to get in touch with him, particularly in modem marriages between people whose homes are a long way apart.
10. It appears that with the Fort Jameson Ngoni a man can divorce his wife by giving her a sum of money; nevertheless, Kachere's councillors laughed at the idea.