Differentiation and Integration in Western Kenya: A Study of Religious Innovation and Social Change among the Bukusu 9783110808070, 9789027976727

Citation preview

Differentiation and Integration in Western Kenya

CHANGE A N D CONTINUITY IN AFRICA

S

]

MONOGRAPHS UNDER THE AUSPICES OF THE AFRIKA-STUDIECENTRUM - LEIDEN

Editorial Board: J. F. Holleman, Leiden Ali A. Mazrui, Michigan I. Schapera, London

MOUTON • THE H A G U E • PARIS

JAN J. DE WOLF

Differentiation and Integration in Western Kenya A Study of Religious Innovation and Social Change among the Bukusu

M O U T O N • THE H A G U E • P A R I S

Publications in collaboration with the Afrika-Studiecentrum, Leiden:* Communications 1. M. L. Daneel: The God of the Matopo Hills. An Essay on the Mwari Cult in Rhodesia. 1970 2. M.L. Daneel: Zionism and Faith-Healing in Rhodesia. Aspects of African Independent Churches. 1970 3. P.M. van Hekken & H.U.E. Thoden van Velzen: Land Scarcity and Rural Inequality in Tanzania. Some Case Studies from Rungwe District. 1972 4. Robert Buijtenhuijs: Mau Mau: Twenty Years After. The Myth and the Survivors. 1975 5. Jan Hoorweg: Protein-Energy Malnutrition and Intellectual Abilities. 1976 Change and Continuity in Africa 1. Robert Buijtenhuijs: Le Mouvement 'Mau-Mau'. Une revoke paysanne et anticoloniale en Afrique noire. 1971 2. M.L. Daneel: Old and New in Southern Shona Independent Churches. Volume I: Background and Rise of the Major Movements. 1971 3. Network Analyses: Studies in Human Interaction. Edited by Jeremy Boissevain and J. Clyde Mitchell. 1973 4. M. L. Daneel: Old and New in Southern Shona Independent Churches. Volume II : Church Growth. Causative Factors and Recruitment Techniques. 1975 5. J.F. Holleman: Issues in African Law. 1974 6. H.L. van der Laan: The Lebanese Traders in Sierra Leone. 1973 7. B.E. Harrell-Bond: Modern Marriage in Sierra Leone. A study of the professional group.1975 8. Jan Kaayk: Education, Estrangement and Adjustment. A study among pupils and School Leavers in Bukumbi, a Rural Community in Tanzania. 1976 9. Law and the Family in Africa. Edited by Simon Roberts. 1977 10. Jan J. de Wolf: Differentiation and Integration in Western Kenya. A study of Religious Innovation and Social Change among the Bukusu. 1977 * The Afrika-Studiecentrum cannot in any way be held responsible for the views or opinions expressed in these books.

ISBN 90 279 7672 4 Cover design by Jurriaan Schrofer © 1977, Mouton & Co. B.V., The Hague, The Netherlands Printed in the Netherlands

To my parents

Acknowledgments

This study was conducted under auspices of the Afrika-Studiecentrum at Leyden. I am especially grateful to Professor Holleman, the director of research at that time, who assisted with the initial draft for the project. In addition I am indebted to several persons for contributing to the succes of my fieldwork. First of all I would like to mention the Medical Officer of Health in Bungoma in 1968, Mr. John Mahieu, whose professional care and friendly hospitality helped me recover from a motoraccident just before the start of my research. Mr. David Welime helped me with my initial contacts at Kimilili. Among the other people at this place were my host, Mr. Pascal Nabwana O.B.E., and my assistant, Mr. Vincent Wanjala. At Sirisia I was especially grateful to Senior Chief Jonathan Barasa, who helped me find an excellent assistant in Mr. Jackson Wepukhulu. Ex-chief Jeremiah Kukubo extended his hospitality to me in one of his compounds. There are many others whom I cannot single out individually. Hardly ever was I refused co-operation with my investigations, although at times my questions must have been rather tiresome. An earlier draft of this book was prepared as a Ph.D. thesis presented at the School of Oriental and African Studies, London University. Here I was especially helped by my academic supervisor, Professor Ph.H. Gulliver, and Dr. D. Parkin, both of whom continued to show an interest in my work and encouraged the preparation of the present book. During the final stages of the writing of my thesis I was in the employ of the Afrika-Studiecentrum. While in Africa, however, and during another year in Europe I was given a generous grant by Wotro, the Dutch Foundation for Tropical Research. Finally I wish to thank Mr. V. A. February of the Afrika-Studiecentrum who was responsible for correcting the imperfections in my use of the English language.

Table of contents

Acknowledgments

vn

List of tables

xi

List of maps

xm

List of abbreviations

xv

1. Introduction 1.1 The area 1.2 The approach to the problem

1 1 10

2. Families 2.1 Introduction 2.2 Control over resources and economic stratification 2.3 Family types and the influence of Christianity 2.4 Conclusion

16 16 18 39 54

3. Communities 3.1 Introduction 3.2 Communities and reputations 3.3 Churches at the grass roots 3.4 Self-help projects 3.5 Local politics 3.6 Conclusion

57 57 58 68 81 85 94

4. The wider society 4.1 Introduction

97 97

x

Table of contents 4.2 4.3 4.4 4.5

Government Politics Religious organizations Conclusion

99 105 111 117

5. Political role differentation 5.1 Introduction 5.2 The traditional political system 5.3 The establishment of an administration 5.4 Maintenance of law and order under Wanga rule 5.5 Tribal associations and economic development 5.6 Conclusion

122 122 127 133 138 147 157

6. Religious innovation and political action 6.1 Introduction 6.2 Introduction of Christianity through education 6.3 The Quaker faction in Malakisi politics 6.4 The Bukusu Union 6.5 Party politics and ethnicity

159 159 161 169 177 191

6.6 Conclusion

197

7. Conclusion

200

Bibliography

210

Index

212

List of tables

1. Distribution of size of hold ing and number of people supported in Elgon Nyanza 2. Reported size of farm of simple family units in North Kulisiru Sub-location 3. Land utilization in Elgon Nyanza 4. Acreage of temporary crops in Kabuchai Division 5. Average number of livestock per holding in Kabuchai Division 6. Average acreage of permanent crops in Elgon Nyanza 7. Average monthly rainfall in Bungoma District 8. School fees paid in Kulisiru in 1969 9. Teachers' salary-scales in Kenya in 1969 10. Interval between marrying first and second wives 11. Recorded marriages of men in Kulisiru and average age of men at first marriage by decades 12. Population by age and sex in Kulisiru and Bungoma District in 1969 13. Summary of family case studies 14. Mean number of years of education of sons of good Christians and sample in Kulisiru 15. Mean number of years of education of daughters of good Christians and sample in Kulisiru 16. Minimum difference between the mean number of years of education of children of good Christians and sample in Kulisiru at the 5% and 1% level 17. Status of fathers of children with skilled and professional jobs in Kulisiru 18. Decade in which married people in Kulisiru joined a denomination

24 24 24 25 25 26 27 31 36 41 42 43 48 52 52

53 54 73

xii List of tables 19. Age and religious status of men in Kulisiru 20. Formal affiliation at some time and full membership in 1969 by sex and denomination in Kulisiru 21. Age of bakasa in North and South Kulisiru Sub-locations 22. Age of land adjudication committee members in North and South Kulisiru Sub-locations 23. Age of KANU committee members in North and South Kulisiru Sub-locations 24. Age of people with multiple leadership positions in Kulisiru 25. Pattern variables and social contexts

73 74 87 92 94 95 206

List of maps

1. 2. 3. 4. 5. 6.

Western Province Bungoma District, Administrative Divisions Bungoma District, Communications Bungoma District, Altitude and Rainfall Kulisiru, Administrative Divisions Kulisiru, Spread of Christianity

2 3 4 5 19 20

List of abbreviations

ADC BU CEE CMS CPE CSC DC DYM EAISR EAS FAM HSC KADU KANU KAPE KAU KES KJSE KPU Legco LNC MHM MISR PAG PEFA SA

African District Council Bukusu Union Common Entrance Examination Church Missionary Society Certificate of Primary Education Cambridge School Certificate District Commissioner Dini ya Msambwa East African Institute of Social Research East African Standard Friends' African Mission Higher School Certificate Kenya African Democratic Union Kenya African National Union Kenya African Primary Examination Kenya African Union Kitosh Education Society Kenya Junior Secondary Examination Kenya People's Union Legislative Council Local Native Council Mill Hill Mission Makerere Institute of Social Research Pentecostal Assemblies of God Pentecostal Evangelical Fellowship of Africa Salvation Army

1

Introduction

1.1 The area This book is a study of religious innovation and social change among the Bukusu, whose home is the Bungoma District in the Western Province of Kenya (Map 1). Apart from the Bukusu there are other much smaller ethnic groups in the same district. The foothills of Mount Elgon are predominantly inhabited by Kalenjin, sometimes also referred to as Elgon Masai. In the north-east near Kimilili they have their own location where they are definitely in a majority, although there are also Bukusu in that area. In the north-west the Kalenjin do not have a location of their own where they constitute the dominant group. In this area relations between Bukusu and Kalenjin have been a constant source of trouble. On the other hand relations with the Tachoni who live in the south-east of the district, and also in the north of Kabras Location, have generally been amicable. The Tachoni speak a Luhya dialect which is not very different from the one used by the Bukusu. The other neighbours with whom the Bukusu had peaceful relations included the southern Gisu, or Masaba, the Kabras, and the Nyala. Their traditional enemies were the Wanga, Khayo, and Teso (Wagner 1949, p. 27). Administratively Bungoma District is divided into eight locations (Map 2). The number of inhabitants ranges from 21,000 to 58,000. These locations are further divided into 44 sub-locations. In 1963 the Naitiri and Ndalu Settlement Schemes were carved out from the White Highlands and joined the Bungoma District. In 1969 there were 40,000 people in these new territories, out of a total district population of 345,000 people (Kenya 1970). Although many people living here are Bukusu from Bungoma District, I only took those schemes into account in so far as they influenced life in the old African 'reserve'. In the rest of this introduction I shall limit myself to a description of this area.

2

Introduction

A

M o u n t Elgon

Mbale GISU

t -

Uganda

POKOT

Kitale ^

/KALENJIN Bungoma \ , \ TESO J District Settlement! n a c V » / schemes s (UASIN GISHU MASAI) ./ f TACHONI r I

/

BUKUSU

^jS

r

J . Eldoret ( Busla ^ v. District \ ^ / KHAYO..._-rrS W e s t e r n -. ^ - ' K A B R A S ^ Riftvalley Province MARACHI •/ NYALA x P r o v i n c e NANDI -_( K a k a m e g a D i s t r i c t ) TSOTSO SAMIA ^\ MARAMA KISA LUO MARAGOLI' I Nyanza / Province - - r v Kisumu

\

\ ;

7 ^ )

m

area above 8,858 feet area above 6,890 feet Lake Victoria major town

Map 1. Western Province

TESO

Scale 1 : 1,000,000

international boundary internal boundary name of tribe

Introduction

Administrative Divisions and Population Density per square kilometre in 1969 N.K. = North Kulisiru Sub-location S.K. = South Kulisiru Sub-location

Map 2. Bungoma District

Scale 1 : 400,000

3

4

Introduction

Roads Railway

•

Trading Centre •

Market

f

Mission Station

Map 3. Bungoma District Scale 1 : 400,000 Communications and settlements

Introduction

Map 4. Bungoma District Scale 1 : 400,000 Altitude and rainfall

5

6

Introduction

Before 1963 the present Western Province and Nyanza Province formed one province. For the greater part of the colonial period until 1956 most of what is now called the Western Province formed one district with its headquarters at Kakamega. It was first known as North Kavirondo and after 1949 as North Nyanza. Between 1956 and 1963 Bungoma District and the northern locations of what is now Busia District were known as Elgon Nyanza District with its headquarters at Bungoma. Before 1959 there were only four locations in Bungoma District: Malakisi (North and South Malakisi), Kimilili (Kimilili, Bokoli, Ndivisi), Elgon, and South Bukusu (East and West Bukusu). Before 1952 South Bukusu was known as South Kitosh, and before 1927 Malakisi, Kimilili, and Elgon were jointly referred to as North Kitosh and were under one chief. Kitosh was the name which the Masai gave to the Bukusu. Both the main motorway and the railway between Nairobi and Kampala pass through the middle of Bungoma District (Map 3). Coming from Eldoret one passes through the Nandi Hills and then descends to cross the Nzoia River, which flows into Lake Victoria. Here one enters Bungoma District. There is another 20 miles before one reaches Bungoma, the district headquarters, from where one has to travel an equal distance before reaching the border with Uganda. Halfway between the border and Bungoma one leaves Bungoma District without noticing it and crosses into the Teso Locations of Busia District. The area where the main road crosses the district consists of hill ridges, often several miles wide, which run in a north-south direction. They have been formed by rivers flowing down as small streams from Mount Elgon, a huge extinct volcano of about 14,000 feet in height. The base of the side visible from Bungoma is about 25 miles long. Immediately after crossing Nzoia River it is possible to turn right and ascend a steep ridge, which is in fact a continuation of the Nandi Escarpment. Following this ridge in a northerly direction along the main Kakamega-Kitale road one reaches Misikhu after six miles. A left hand turn here brings one to Kimilili after another six miles. Kimilili lies in the middle of an important agricultural area and is a flourishing trading centre. Four miles or so to the north of Kimilili are the foothills of Mount Elgon which rise rather abruptly and which have been called cliff's by some scholars. Generally the forest reserve starts almost immediately after ascending these cliffs, but at Kapsakwony in Elgon Location the inhabited area is several miles wide. The main road runs parallel with the foot of the mountain from

Introduction

1

Kimilili to Chwele. The slopes of the ridges are much steeper here than those to the south. The distance from Kimilili to Chwele is about 10 miles. From Chwele one can go straight southwards and reach Bungoma again after a dozen miles, passing on either side of Kabuchai, a prominent rocky outcrop which is almost the geographical centre of the Bukusu country. Before Bungoma became the headquarters of the district, Kabuchai, conveniently situated on the road from Mumias to Mbale, was the seat of the Divisional Court. Mumias was the first administrative centre in western Kenya and was near a ford in Nzoia river, where the caravans from the coast to Uganda had to pass. Mumias is about 18 miles to the south of Bungoma. The Wanga Locations of Kakamega District start to the south of Kabula. One can also continue along the road from Chwele to the west and cross the watershed between the Nzoia and Luakhakha basins to reach Sirisia after eight miles. Mount Elgon turns rather sharply north now. If one travels from Sirisa in a north-westerly direction one reaches the border at Luakhakha after ten miles. Malakisi, an old trading centre with a cotton ginnery on the border of the Teso Locations, lies to the south-west of Sirisia. From here one road leads back to Bungoma, while another runs straight south to link up with the main Tororo-Eldoret road after only five miles. This road is only 15 years old and was not of tarmac until 1969. Before its construction the way to Jinja and Kampala went through Kakamega, Mumias, and Busia, or from Kitale, through Kmilili, Chwele, and Malakisi to Tororo. The main north-south links were from Kakamega to Kitale and from Mumias to Malakisi and from there onwards to Mbale in Uganda. When the railway was completed in the late twenties new trading centres were opened where these roads crossed the railway: at Bungoma and at Broderick Falls. At that time Kimilili also became a trading centre. Before this Malakisi was the only officially gazetted trading centre. Asian traders who dominated the commercial life in the district lived in these trading centres. But since they were few in numbers and far apart, Africans were enabled to set up neighbourhood shops and retail daily necessities, while the Asians acted as wholesalers. In the early fifties the African District Council brought the building of shops outside the trading centres under its control. Plots were acquired and for a fee of shs. 200 per annum people were allowed to erect shops constructed of stone or brick with corrugated iron roofs according to a standardized design, in permanent materials. Shops were often arranged around fenced market squares, where people could sell their farm produce and buy goods from

8

Introduction

pedlars at the weekly markets. The best developed markets of this kind, where the greater number of shops are also found, are in prosperous areas between trading centres. Chwele and Misikhu are excellent examples. The markets which sprang up adjacent to the trading centres have become something of an anomaly now that Africans can easily acquire shops there. It has even become official policy to limit the role of the Asians in trade. But they still offer the advantage of a flat rate which is much less than what one would have to pay in a proper trading centre. Trading centres and markets with shops are the most conspicuous concentrations of permanent buildings in the district (Map 3). People do not live in villages but each family lives on the plot which it cultivates. The most prosperous trading centres and the largest markets are to be found in the areas with the highest agricultural potential and the greatest population density (Map 2). These are the foothills of Mount Elgon and the area immediately to the south of them. The soil consists of dark red friable clay with a deep humic topsoil with a carbon content of 3-7%. This soil is derived from both volcanic and basement complex rocks. The same type of soil can be found in the area to the north-east of a line running from Chwele to Broderick Falls, with this difference that the deep humic topsoil is limited to the summits of the broad flat-topped ridges formed by the rivers which come down from the mountain. The carbon content is less where the land slopes i.e. 3-5%. In the rest of the district the soil consists of dark brown sandy loams with a mere 2% carbon content. It is derived from sediments and basement rocks, partially covered with more humic ash and pumic soils derived from recent unconsolidated volcanic ash. In many places there are swamps which are badly drained and which prevent the growth of trees and shrubs, the natural vegetation of the district. In the centre, where they are most extensive, they consist of dark brown clays with light textured topsoils (Gethin Jones and Scott, 1959). Agricultural potential, however, does not solely depend on the composition of the soil but also on rainfall and height (Map 4). The rainy season is from March until the end of October, with seasonal peaks during the period April - May and again in August. Rainfall in the south-west is higher than in the north-east, but generally it is over 50 inches per annum. This allows for the cultivation of maize throughout the district. However, this crop fares better on the more fertile soils of the north and north-east. Where the soil is relatively poor maize is not generally grown in commercial quantities. If people do so they use fertilizers. On the other hand the less fertile areas are also lower and hotter. This means that grain grows

Introduction

9

faster and can be harvested several weeks earlier in these areas. Traders take advantage of this situation by selling grain from the lower areas to people in the higher areas during this period. The other important cash crops are cotton and coffee. Cotton does not grow well above 4,500 feet while Arabica coffee is not grown below 5,000 feet. Coffee can be grown even in the less fertile areas provided that the site is carefully selected, for example on the site of an old walled village. In the north-west onions have recently become an inportant cash crop. In the north and north-east maize is the staple food, but in the rest of the area millet mixed with cassava is also customary. Milk, eggs, poultry and beef are consumed everywhere. The diet is varied with bananas, sweet potatoes, beans, groundnuts, and rice, which is grown in the lowest river valleys. Pineapple and sugarcane are favourite snacks. Cabbages have to a large extent replaced local vegetables grown in the kitchen gardens. European potatoes are grown in the foothills of the mountain and consumed locally. Maize became an important cash crop in the nineteenthirties after the completion of the railway which facilitated export. In the area adjoining Trans Nzoia people started to use ox-drawn ploughs with which they had become familiar while working for European employers. In the lower areas adjoining Uganda cotton was vigorously promoted in the nineteen twenties and at Malakisi a cotton ginnery was built which is still in operation. The present coffee plantations were started between 1954 and 1962. Onions were first grown on a commercial scale in 1964. They were grown especially in the north-west, probably because the Malakisi Ginnery provided a good marketing channel. Cattle husbandry is also very important in Bungoma District. In 1960 it was estimated that there was approximately one head of cattle per person (Kenya, 1962). The County Council has six major auction rings. The cattle trade can be very profitable. During the colonial period some Africans accrued large fortunes through it. Here they did not face any competition on the part of Asians. The population density in the most fertile areas in the north and northeast is more than 200 per square kilometre. These areas attracted most migrants before 1950 (Map 2). The only exception is Elgon Location. This is partly due to the fact that part of its area is covered with uninhabited forest, partly to thejfact that the area was almost uninhabited in the past. While other areas of Bungoma District increased their population between 1948 and 1962 by 68%, in Elgon Location it increased by 184% (Kenya, 1950b, 1964). Between 1962 and 1969 Elgon Location

10

Introduction

increased its population by 44%, while the rest of the district, with the exception of the settlement schemes, increased its population by only 33%. Locations which had a considerably smaller increase of population during the latter period were South Malakisi by 23%, Kimilili by 22% and Ndivisi by 20% (Kenya, 1964, 1970b). These differences can partly be explained by a continuing movement of the population into the foothills of Mount Elgon, partly by a younger population and a higher birth rate in that area, and also partly by the greater pull of the settlement schemes for people who live closeby. 1.2 The approach to the problem I think that on the whole social anthropologists have shown less interest in religious innovation than in other aspects of social change such as labour migration, urbanization and political adaptation under colonial rule. Only a few anthropologists have given us sufficient material to enable us to judge whether such neglect was unintentional or deliberate. Among those anthropologists who have given us detailed information about their field work, there are two who worked in East Africa of whom one should take note in this respect. Beattie saw it as his task to understand Nyoro society as it was during his period of field work and did not attempt to describe the traditional culture and society which was no longer in existence (1965, p. 8). Yet he seems to have found it difficult to appreciate that Christianity should be studied and explained equally to traditional practices and beliefs. Beattie's visit to the local mission church on Christmas Day is only mentioned because the sermon revealed that people were still practising the traditional spirit medium cult, something which had always been denied (1965, p.22). Similarly a group of women converts to the revivalist and fundamentalist branch of the Anglican mission seems only of interest because they were willing to give him details of cult membership and initiation (1965, p. 29). Middleton (1970, pp. 56, 57) also admits that he did not do as much work among adherents of various break-away movements with their origin in Christian mission activities, as he now wishes he had. The role of the leaders of these sects as instigators and agents of radical social change was not realized by him until after his final return from Uganda. These examples show that an anthropologist who would be prepared to focus his research on the influence of Christianity might be able to elucidate aspects of social change which had hitherto remained obscure.

Introduction

11

Nevertheless there are some studies of this nature, although in many cases they were not written by professionally trained anthropologists, but, not surprisingly, by theologians. Of outstanding interest are undoubtedly the books by Sundkler (1961) and by Taylor (1958). Sundkler concentrates on the independent sects among the Zulu and Taylor on the Anglican Church in Buganda. Both authors try to explain the internal development of these institutions by relating them to traditional rituals, beliefs and social relations, and also to political and economic changes which transformed the traditional society. The same method was followed by professional anthropologists: Pauw (1960), Wishlade (1965), and Murphree (1969). It is interesting to note that both Pauw and Murphree were missionaries before they became anthropologists. Sangree (1966) and Long (1968) on the other hand were not primarily interested in explaining the acceptance of religious innovation and the development of new religious institutions. They wanted to know the social consequences of the introduction of these institutions and how participation in these institutions influenced social actions. Sangree tried to prove that Protestant sects, both mission affiliated and independent, may fill a transitional political need. Such sects provide a ready-made framework within which a leader can fulfill some of the particularistic relationships traditionally expected from a man in authority, but also at the same time introduce more universalistic and functionally specific relationships with the aid of Christian ritual and symbols. Thus people get used to the demands of bureaucratic government. Unfortunately, the Tiriki age group organization was an intervening variable which made the testing of the original hypothesis difficult. Long tried to establish how the members of one particular sect, the Jehovah's Witnesses, differed from the rest of the population with regard to their choice of residential patterns and the utilization of new economic opportunities. He came to the conclusion that there is a close connection between being a Witness and being engaged in certain forms of social and economic activity. Membership of the sect encourages people to repudiate certain social relationships (often of a so-called 'customary' nature), and sanctions the utilization of ties of a different kind. Thus people can gain a high status in a new and nation-wide type of social stratification, for example through concepts such as 'townsman' and 'villager'. Long shows that the doctrinal and ethical teachings of the sect influence the choice between alternative courses of action open to the individual. Sangree is more concerned with institutional adaptation to a new political system

12

Introduction

which was imposed on and influenced the life of all members of the tribe he was studying. Each author is concerned with different aspects of social change. Like Long and Sangree I was more interested in the influence of religious innovations on processes of social change than in new religious institutions as ends in themselves. But I did not consider it profitable to limit myself to a study of either individual choices or of institutional changes. Perhaps over-ambitiously I wanted to study the interaction between the two. Again in contrast to Long and Sangree I was not just interested in one sect or group of sects. I felt that too much attention had already been given to independent sects. Somehow they must have seemed more African than orthodox missionary churches and therefore a more appropiate subject for anthropological research, or perhaps more exotic and therefore more in need of explanation. I wanted to study both types of new religious groupings simultaneously so as to gain a more balanced picture. I deliberately looked for an area where a fair number of different denominations would be active and where an independent sect or two would also be present. The Luhya tribes of western Kenya contained the desired religious variety. Since Laura Bohannan and Walter Sangree had worked among two southern tribes of this group in the nineteen-fifties I decided to concentrate on the Bukusu in the north. I started my field work in March 1958 in Kimilili Location. Mr. Pascal Nabwana kindly agreed to accommodate me. He was also in other respects very helpful, not in the least by writing some autobiographical notes, which were very useful for an understanding of his career, culminating in his chairmanship of the African District Council 1958-1963 and the award of an OBE. One of his nephews was my assistant during the seven months that I stayed with him. Much of my time with him was used to learn the basic principles of the Bukusu language. In August the nominations for the local government elections were submitted. All candidates of the opposition Kenya People's Union, including Mr. Nabwana, were disqualified and all candidates of the governing party KANU were re-elected unopposed. In October the Government proscribed the Dini ya Msambwa sect when the leading prophet was sent to jail for assault. He had taken the law in his own hands by punishing the wife of one of his followers. I had hoped to find out more about the adherents of this sect after I had learned more of the Bukusu language and culture. This had now become impossible. I then decided that a change of scene would be beneficial for my research

Introduction

13

and I moved to Sirisia in South Malakisi Location. Here I was fortunate to find a house in one of the compounds of ex-chief Jeremiah Kukubo. His excellent memory proved a great help during my study of past conditions in Malakisi Location. My cook was assistant to the sub-chief in my area and he consequently knew a lot about my neighbours and community affairs in general. My assistant here had been a newspaper correspondent in 1963, when the present generation of politicians came to power. His knowledge of these matters was of great advantage to me during the national elections at the end of 1969. The Survey of Kenya was just finishing the registration of land around Sirisia. With the help of some of their up-to-date records I investigated the relationship between clan membership, ownership of land and the present settlement pattern. Genealogies were also recorded and the relationship between clan membership and political leadership at the grass root level in the past was investigated. I also made a point of participating in church services as frequently as possible. After four months at Sirisia I had a break of three months. One was spent in the archives in Nairobi. I went through Annual Reports, Handing Over Reports, Chiefs' Characters and several files on the Dini ya Msambwa. The rest of the time I stayed at Makerere College as an Associate of the Makerere Institute of Social Research. Here I tried to write the first systematic account of what I had hitherto established. I also went through the micro-filmed Kaimosi papers of the Friends' African Mission. These threw valuable light on the activities of the Quaker missionaries among the Bukusu. The late Father Walstra had already allowed me to read the Diary of the Kibabii Mission as well as some miscellaneous correspondence. This provided me with a good picture of the activities of the Catholic Mill Hill Fathers whose influence among the Bukusu equalled that of the Quakers. Upon my return to Sirisia I studied the organization, membership and local history of the various denominations in the area more closely. I also did a small-scale survey to get census data and to find out whether religious affiliation had made any difference with regard to certain social characteristics, the most important variable being education. That summer Mboya was assassinated and rumours about oathtaking ceremonies among the Kikuyu gave rise to concern. Later the KPU opposition was banned and the parliamentary elections were brought forward. They took place early in December 1969 and the new Government was announced just before Christmas when I left the country.

14

Introduction

I felt it was necessary to neglect some of the conventional limits generally accepted by anthropologists when deciding the scope and relevancy of the analysis of their material. Traditionally social anthropologists worked within a restricted geographical area and with a population numbering only some hundreds, or at most several thousands. At the end of their study, the material collected was used to provide a unified picture of a single entity. If they were faced with the problem of studying units which were much larger, they would concentrate on an arbitrarily defined segment. The chosen area had to represent a unit in which the people were in close social relations with one another. The factors which existed within that set of relations would be dealt with as though one were dealing with an isolated society (Colson, 1967). Bailey (1964) has examined some of the problems connected with this approach, especially the question of when and where to cut oif investigations of social relationships which fall outside the scope of the social system under investigation. He found that many relationships which connect the people of Bisipara village with outsiders - administrators, traders and merchants, religious experts - can be analysed by other disciplines and can be treated as 'given'. Investigations come to a halt at the point where these relationships tend to become single-interest and specialized. There are other relationships with outsiders which cannot so easily be treated as 'given'. For example in Bisipara there are people who are classified as untouchable and who have forged links with a nation-wide Harijan movement in order to advance their own interests. These links with the Harijan movement form the beginning of a new group, a new corporate group with political interests, which is not recruited solely on the basis of political interests. The social anthropologist, as distinct from the political scientist or the economist, is particularly equipped to study the basis of this recruitment, because such a study involves a knowledge of village structure and Hindu values. The same would not apply to relationships of the villagers with merchants or with the Deputy Commissioner, who do not constitute a group. The analysis of the Harijan movement, at least in its more local manifestations, is therefore a feasible target for social anthropologists. It is also logically the next step in social anthropological research, because the more such supra-village-level relationships come into being, the more difficult it will be to isolate a significant system of social relationships located roughly in a village. Bailey concludes that if the explanatory value of analysing the social structure of the village becomes very low, and a large number of factors have to be taken as 'given', then the social

Introduction

15

anthropologist must be prepared to select those 'given' elements which seem susceptible to social analysis by his own technique, or by techniques and concepts which he may have to devise in order to analyse them. I believe that this is a fair description of my approach in this book with regard to non-traditional religious groups. But there is not only a problem of spatial dimensions but also of temporal dimensions suitable for social anthropological research. Bailey considers that events which belong to history can be taken as 'given' because a wrong diagnosis of anterior causes will not invalidate the analysis of the effects of these events in the social system of the villages. Yet, Asad (1972) has argued, and in my opinion conclusively demonstrated, that a historical perspective is necessary for an adequate analysis of structures of domination in terms of relations between classes. He rightly postulates that social anthropologists are concerned with revealing the mechanisms through which particular structures are maintained, although it is completely inappropriate to specify a mechanism in accounting for class structure as a historical reality. A class perspective reveals the political structure to be a complex, developing reality, and not a logically closed system which happens to be located at a point in linear time. There are different combinations of factors operating at different periods, which can only be revealed by historical research because the process is a historical one. All one can say concerning a class analysis is that such factors must ultimately be seen in relation to the means of production and to the differential consciousness of self-identifying groups in formation and disintegration. The analysis of Bukusu history which I have given in this book is therefore not just a prologue or epilogue to the description of the Bukusu social system, but essential for a proper understanding of the dynamics of its power structure.

2

Families

2.1 Introduction The distinguishing feature of societies which are normally studied by social anthropologists is that social relationships are multiplex. The people with whom one is related through ties of kinship and marriage are the same as those with whom one combines to defend a common territory and to maintain the peace within such an area, with whom one exchanges labour and food, and with whom one joins in worship and the celebration of rituals. Families are embedded in this structure of multiplex relationships. They are encapsulated in the local group and in the kin group which mediate between the family and the wider society. This encapsulation diminishes and the mediation loses its importance as societies become characterized by single-stranded relationships. Not all relations with people with whom one does not have multiplex relations are equally important for the creation of direct contacts between families and the wider society. The fact that people sell some food or other crops to a trader for cash and buy some things from him which they do not produce themselves, does not necessarily affect their division of labour, their ways of solving conflicts or their religious activities. Similarly payment of taxes and prohibition of violence in respect of outsiders can leave the multiplicity of the relations with neighbours and kinsmen unchanged, although it is likely that not all the roles will remain exactly the same. For example, men may grow cash crops instead of making war. People will also have new resources at their disposal, both economic and political, and some will have better access to these resources than others. Yet as long as the use of these resources is limited in scope and does not change the multiplex character of roles, we cannot very well say that the people who come into contact with institutions constituting the wider society become integrated into that wider society.

Families

17

Relationships with representatives of the wider society can be singlestranded without diminishing the importance of the existing multiplex ties. This only happens if people do not merely come into contact with formal organizations which constitute the wider society, but actually participate in the affairs of such associations as members and leaders. Such formal organizations exist to achieve specific goals and can bring together people who are not already united through ties of kinship and neighbourhood. Such organizations normally become so large that face-to-face relations with all members are impossible. Through participation in such organizations people become integrated into the wider society. As members of these organizations they are no longer encumbered by the other social roles they play because most people with whom they interact directly or indirectly in such associations do not interact with them in other contexts. Even if all people who are kinsmen and neighbours and who have multiplex relations with one another become members of such an organization, they have to neglect these ties in so far as they might make co-operation with people with whom one does not share these attributes less succesful and the achievement of the desired goals less likely. When only a section of those people who are linked through multiplex ties become interested in such an association, it becomes even less possible to maintain the multiplex character of role relationships, because participants isolate a field of activity from their fellows who do not want to participate. Among the Bukusu the churches did play this double role in the process of social change. They brought people together in a formal organization which transcended the traditional limits of effective social interaction and they isolated members from neighbours and kinsmen with whom they could no longer share many interests. Those people who had this common interest in a formal organization which transcended local boundaries and diminished the importance of multiplex ties were often better placed to see opportunities for advancement in the wider society which seemed irrelevant to other people. This tendency was increased as membership was incompatible with certain traditional ways of gaining prestige. As a result economic differentiation occurred which led to an even greater isolation of their families from the immediate social environment of multiplex ties. With a view to studying the influence of the churches in this respect I selected an area where it was possible to compare families which were integrated into the wider society through membership of churches of various denominations with other families, living under approximately similar circumstances. The area which I selected is shown on Maps 5 and 6.

18 Families In the east and west the area is bounded by the borders of Kulisiru Sublocation as it existed before 1958. Afterwards the area was enlarged in the north and east and in 1960 it was split into two new sub-locations, North Kulisiru and South Kulisiru (Map 2). In the north the boundary of the area which I selected is formed by the Malakisi and Ndakalu Rivers, in the south by a part of the old road which connected Chwele with Malakisi. I shall refer to this area as Kulisiru, after the big hill which dominates the area. When talking about the present-day political units I shall designate them as North Kulisiru Sub-location and South Kulisiru Sub-location, large parts of which form the area of Kulisiru as I have defined it here. People who live in the other locations of Bungoma District often refer to this area as Sirisia, a term which I reserve only for the administrative centre of South Malakisi Location. Sirisia has a market with a row of shops, a health centre, a cattle auction ring, a Magistrate's Court, and the office of the chief with barracks for a detachment of Administration Police. The nearest official trading centre is Malakisi, where there is a sub-post office, a station of the Kenya Police and a cotton ginnery. There are still some Asian traders here but the big wholesale business, which supplies most of the smaller shops at the various markets in North and South Malakisi, West Bukusu and the Teso Locations, is owned by a partnership dominated by Africans. In order to get from Sirisia to Bungoma, the district headquarters, by bus or car one has to pass through either Chwele or Malakisi. A new bridge to the south of Bukukholo is expected to shorten the distance once the road has been improved. At Bungoma there are various Government offices, a hospital, a post-office, commercial banks, headquarters of the Farmers' Co-operative Union, and large stores of the Maize and Produce Board near the railway station.

2.2 Control over resources and economic stratification It is part of the rural development policy of the Government to have landholdings registered, so that owners can use them as security for raising loans. In Bungoma District, the colonial Government had already initiated and stimulated the enclosure of individual plots. In those days the aim seems to have been rather to decrease the amount of legislation over land. Registration still has a similar effect. Another important aspect of registration is the provision which is made for the construction of roads and the

Families

J /

r

~

V.

/

'

1

!

/ / S

' ,''sirisia

NORTH-KULISIRU

f |

•

> ^

-> i

N

MusiebcK \

/ r-N

"

- -

''

\ /

y

Kichawa

'

_

^

\

*

, \

Chepkutumi \

i /

\

Tunyo —

\

/

\

\

'

i

'

— ~

\

• u•t a s o L

N

*

\ \

Nabuloli

\ 1i

*

/ '

/ v

^

;

> \

, \ '

J

/

locational b o u n d a r y sub-Iocational boundary ward boundary Wokope •Jf

^

\ Chenjenyi

—

'

A d m i n i s t r a t i v e Divisions —-

I

Namawanga

\

/

/

'/ Chongoyi

\

I

J

_/

i

1

*

/

'

A ' '

*

/

^ :

S O U T H - KULISIRU.. i_

/ ' './ Wokape

'.

>

•

^

/

^f..

v

/ i--'' Yabeko

\ "v——.. / / "—" / Bitonge

, , Lukala

v

I

CK Sibanga

Ndakalu

/

v 'Mahnda*.

n a m e of ward ward outside sample area 'Kulisiru'

Map 5. North Kulisiru Sub-location south of Ndakalu River and South Kulisiru Sub-location Scale 1 : 50,000

, /

/

20

Families

major road minor road river boundary of sample area

Map 6. Spread of Christianity

I I primary school •j" church or mission house O former 'lines', catechumenate or 'bush' school

in Kulisiru

Scale 1 : 50,000

Families 21 selection of sites for shops and schools. If farms can be reached by lorries during the harvest season, cash crops which are relatively bulky can easily be transported. As part of the registration, local land adjudication panels are set up, together with a system of arbitration boards to which one can appeal in order to establish once and for all who the legal owners are. The area which I selected for detailed investigation had just been registered and only a few appeals were still waiting for a decision from a court at a higher level. The local panels had virtually finished their work. The records were completely up-to-date. Before 1908 the Bukusu were living in walled villages, some remains of which are still visible in Kulisiru. The people who used to live in these forts have spread themselves in the surrounding area or migrated elsewhere. On the basis of genealogies recorded by me I would say that if everyone whose father or paternal grandfather had once lived in such a fort had remained in the area, the number of descendants in Kulisiru would now be more than double. Until 1955, it was still possible to find unoccupied land in Bungoma District, particularly in the foothills of Mount Elgon. It was possible for people to go there, clear the bush, and start farming. Land did not become a closed resource until the late fifties, if we consider the whole of Bungoma District. People also went to the White Highlands to become squatters, or to Uganda, notably to Bukwa in Sebei and Yembe in Bugisu, where there was plenty of empty space. After independence many people went to the new settlement schemes. In 1969 11% of the population of Bungoma District was living there. Sons can claim land that has been occupied by their father, even if they are not actually living there. Brothers should get approximately equal shares. Fathers sometimes wished to register the plots of their married sons who were living with them in their own name. Sons would normally not dispute such a decision. Again, brothers would normally have arrived at some kind of amicable settlement when the fields were enclosed in the late fifties. In some cases brothers who lived together wished to register the whole plot under communal ownership. When sons took their fathers to court, or when a brother disputed the claims of other brothers much bitterness and ill feeling ensued. Land adjudication committees would only consider disputes between relatives other than fathers and sons or brothers, if the person who had originally invited this relative to live on a portion of his land was still alive. In the past it was not at all unusual that people would prefer to live with their brother-in-law, or with their mother's brother rather than with their own brothers or father. Fathers who had only grown-

22 Families up daughters, but no grown-up sons, might actually invite their daughters' husbands to come and live with them. People might even be willing to accommodate others who were not related to them but with whom they just happened to be good friends. Only this person would be allowed to ask the land adjudication committee to evict his relative or friend. He would probably act on the instigation of his sons who did not want to suffer shortage of land because of their father's former generosity. If the relative who was threatened with eviction had nowhere else to go, the committee would offer him a share equal to that of the other claimants and allow him to live there. I took down the name of 518 plot owners who had an officially registered holding and made diagrams of the position of these plots on the ground. My ultimate aim was to compare all families with a Christian head of the household, with a comparable sample of the total population of Kulisiru. For this reason I decided that a 10% random sample would yield sufficient data. Of the 53 plots I investigated, eight were without occupants. Four of these plots had been bought by people who already had a place to live elsewhere in the same area. Owners simply use them as pasture or for growing crops, but they have no houses built on them. Two other plots belong to people who have temporarily migrated to Uganda. Of the remaining two plots one is near the shops and belongs to a man who bought it with the purpose of building a rest house here. The other plot belongs to a widow who owns part of the plot of her late husband, but who lives now elsewhere with another man.* Of the 45 inhabited plots, there were only five on which the registered owner himself was not living. One plot was inhabited by a middle-aged divorcee, who had been given the plot by her brother, the registered owner of the plot. The owner of another was absent because he was employed in town. However, he had left his wife and children behind. The other three plots belonged to polygamists, who had some of their wives living there, but who were not living there permanently themselves. On all the other 40 plots the owner was living with his wife or wives. There was only one exception, a plot registered in the name of a small boy whose father had died and whose mother had not re-married. On 32 plots the owner and his wife or * The agnates of the dead men could not inherit this plot because he had given it to a sister's son before he went away to Uganda, where he died. If they had attempted to get an eviction order against him, the land adjudication committee would have left him in undisturbed possession. But the committee was found willing to give the two widows equal shares with the sister's son of their dead husband.

Families 23 wives were the only married people, but on seven plots there were also married sons and daughters-in-law or daughters-in-law whose husbands were absent because of employment in town. Apart from children of the inhabitants whom I have mentioned so far, there were also unmarried relatives. Widowed mothers were living with their married children on five plots; on eight plots grandchildren were living with their grandparents, their own parents being elsewhere. I found other unmarried or divorced relatives on six plots, mainly nephews or nieces of the registered owner. In 1960-1961 a sample census of African farms was undertaken (Kenya 1962). Some of the aggregate figures presented in the report apply to Elgon Nyanza, which consists of the present Bungoma District and that part of Busia District which is to the north of Nzoia River. The other aggregate figures are broken down to divisional level. Our area is in the Central or Kabuchai (Kavujai) Division, consisting of East and West Bukusu and North and South Malakisi Locations. Within each division a random selection of three sub-locations was made. In the Kabuchai Division 56 farms were investigated, in the Eastern or Kimilili Division 67 farms were investigated and in the Western Division, now Busia District, 66 farms were investigated. Farmers were selected from the list of tax payers. The holding of a farmer was defined as all the land under his economic control. Holdings might consist of more than one parcel. As I did not personally carry out an agricultural sample survey, I decided to use the officialfigures,supplemented with data from the AID Survey of Elgon Nyanza which took place in August 1962 (Wilson, 1962). North Kulisiru was one of the sub-locations chosen for the collection of data from families. One hundred and one simple family units were subjected to a questionnaire which covered among other things types of crop which were grown and income derived from cash crops and other sources. Simple family units were defined as consisting of a man, his wives and children and other relatives living with them. These units can be compared with the families which I identified in my own sample of registered holdings. But some discrepancy should be expected because some registered holdings contain more than the one family as defined in the AID survey. Again, the holding of the sample census is not exactly comparable with the plot in the registration book compiled by the Survey of Kenya. The latter is a contiguous stretch of land which is registered separately. It may be divided into farms which are under control of several tax payers. Moreover, a holding can consist of more than one parcel, each of which would have been registered separately. In Elgon Nyanza the average size of

24

Families

a holding was 18.5 acres, the average number of parcels 1.02 and the average parcel size 18.2 acres. Nevertheless, all three sets of data refer to very much the same kind of units. The total number of persons living on 45 plots in my own survey was 423, which gives an average of 9.4 people per plot. The census reports an average number of people supported per holding of 9.14 in Elgon Nyanza and of 8.57 in the Central Division. The AID survey reports an average of 9.7 people per simple family unit. Table 1. Distribution of size of holding and number of people supported in Elgon Nyanza Size of holding under 2.50 acres 2.50- 4.99 acres 5.00- 7.49 acres 7.50- 9.99 acres 10.00-14.99 acres 15.00 acres and over

Percentage

Number of people

5.7% 9.5% 11.4% 13.4% 17.5% 42.5%

5.00 7.12 7.90 8.51 7.80 11.23

African Agricultural Sample Census 1960-61 Table 2. Reported size of farms of simple family units in North Kulisiru Sub-location Size of holding less than 5 acres 5-10 acres 10-15 acres 15-20 acres 20 acres and more AID Survey Elgon Nyanza

Percentage 31% 9% 17% 16% 27%

Table 3. Land utilization in Elgon Nyanza Utilization temporary crops permanent crops uncultivated pasture bush

Average acreage 5.6 acres 0.3 acres 12.2 acres 0.1 acres

African Agricultural Sample Census 1960-61

Families 25 I think that we may conclude from the figures in Tables 1 - 3 that land is not really a constraint in Elgon Nyanza. Although there are some farmers whose holding is less than the average acreage of arable land, fewer people must be supported from these holdings. It is also very likely that young married farmers with only two or three small children live with fathers who have not yet divided up the land which is used as pasture. In North Kulisiru Sub-location 21% of the simple family units cultivated land that was reported to be owned by their fathers. The other figures also indicate that there is no serious land shortage in North Kulisiru. The median acreage of land under cultivation was 30% of the total acreage available.

Table 4. Acreage of temporary crops in Kabuchai Division Crop Maize Sorghum Wimbi Cassava Rice Sweet potatoes Sesame Cotton Pulses Millet Groundnuts Potatoes

Average acreage 2.51 1.78 0.99 0.99 0.24 0.15 0.14 0.11 0.10 0.09 0.04 0.02

acres acres acres acres acres acres acres acres acres acres acres acres

African Agricultural Sample Census 1960-61

Table 5. Average number of livestock per holding in Kabuchai Division Livestock Cattle Sheep Goats Poultry

Number 7.09 0.77 1.46 9.34

African Agricultural Sample Census 1960-61

26 Families Table. 6. Average acreage of permanent crops in Elgon Nyanza Crop Fruits Bananas Coffee

Average acreage 1.00 acres 0.89 acres 0.36 acres

African Agricultural Sample Census 1960-61 From my own observations and the data supplied by the AID survey I would say that the pattern of farm activities which is tabulated in Tables 4-6 is also characteristic of Kulisiru. However, there was much less sorghum planted than the average of 1.78 acres given here. There were only a few scattered fields. On the other hand there is more coffee than the 0.36 acres average given for Elgon Nyanza, if only because in the Western Division coffee was not cultivated. Onions is a crop which has become exceedingly popular since 1965. In 1966 1,017 acres were planted in Bungoma District, in 1968 4,358 acres (Kenya 1970). Most of the seed was supplied by the Malakisi Cotton Ginneries, which also bought up the bulk of the harvest. Onions and coffee are the most important cash crops in Kulisiru. Although cotton is planted, the greater part of the area is over 4,600 feet and therefore too high for succesful cultivation. These are three crops which depend on external markets. The picture is different for the main food crops. Maize can be sold to a Marketing Board. Some farmers in South Malakisi Location who grow large quantities of maize deliver it directly to the stores at Bungoma. This saves them the charges made by co-operative societies or private traders who normally buy from the farmers locally and than organize bulk transport. This is the normal practice in Kulisiru, although most farmers would prefer to sell it locally at higher prices than those paid by the Marketing Board. They, however, make an exception in the event of the failure of other crops and if they are in need of ready cash. Maize and wimbi are sold locally to women who make beer from it and sell that or to people from adjoining areas where the main harvests are later because of the difference in height. It is also sold to families which have not grown enough themselves, either for cash or in return for their labour. Succesful farm management depends on the most effective use of available labour to satisfy the need for cash and food. If insufficient food is planted people are forced to use cash which is needed for other purposes in order to buy food. In the event of a shortage of cash, they would rather

Families

27

sell their chickens, goats or cows instead of using the eggs, milk and meat for improving their diet. On the other hand if there is sufficient or even a surplus of food but no cash available from export crops, food crops may have to be sold at a time when prices are low, or livestock may have to be sold at a time when there is little demand for luxury food. Clearly, producing food crops for the local market is most profitable if one can afford to wait until other people have exhausted their own resources. Table 7. Average monthly rainfall in Bungoma District Place Height Period January February March April May June July August September October November December

Sirisia 5,300ft 4 years 38.2 54.7 133.1 304.5 255.7 99.3 137.5 136.0 156.2 155.5 126.4 116.2

mm mm mm mm mm mm mm mm mm mm mm mm

Kimilili 5,500ft 20 years 36.9 50.8 109.9 197.2 229.0 141.2 125.7 132.0 139.3 119.0 81.4 68.0

Sang'alo (Bungoma) 4,500ft 33 years

mm mm mm mm mm mm mm mm mm mm mm mm

52.2 61.1 119.6 224.9 236.5 146.6 123.2 133.6 141.5 126.5 108.3 91.2

mm mm mm mm mm mm mm mm mm mm mm mm

Summary of rainfall in Kenya in 1966, EACSO, Nairobi Agricultural work is closely tied to the pattern of rainfall which is shown in Table 7. The first crop to be sown is wimbi. It is sown in January and February and can be harvested after about six months in June and July. It tolerates dry spells in the early stages of its growth. As wimbi is broadcast it must be weeded and thinned at least once. Sorghum, which is also very drought-resistant, is sown and harvested at the same time as wimbi. But sorghum may be ratooned and can yield a second harvest in December or January. Maize should be sown at the beginning of the long rains for optimal yields. It is usually planted from the end of February until the middle of April. Ideally maize should be weeded three times at two weeks' intervals. Pulses and groundnuts are planted at the same time as maize, often in mixed stands. Pulses can be harvested after four months, ground-

28

Families

nuts after three months. Maize can be harvested after six months. Since the long rains usually entail a period of food shortage the importance of crops which become available before the harvesting of maize is obvious. Onions are first sown in seed beds. After about a month they can be transplanted and will then be ready for harvesting after another three months. During this period they are weeded three to four times. The first seed beds are prepared in February, the last in June. People seem to prefer an early date, because the prices at the beginning of the season tend to be better and because of a serious shortage of cash after the long rains. The recommended sowing-period for cotton is April-May. The crop can then benefit from rains in August and September. Harvesting can take place when the weather becomes progressively drier. Yet, because people find other crops more profitable or need them more for food, most of them plant cotton in June or July, or even as late as August. Simsim and sunflowers are typically short-rain crops. They are planted in July and harvested in December. A second planting of pulses and groundnuts can also occur during this period. Famine crops such as cassava and sweet potatoes are also planted at this time. Sweet potatoes take three to four months before they are ready and can be left in the soil for a period of a year or so. Although it is a perennial crop, it is treated as being annual. If the major grain crops fail, the population can still eat sweet potatoes during the long rains when the granaries become empty. Cassava takes half a year before it can be uprooted. It can be left for two or three years in the soil before the roots become too fibrous and therefore unsuitable for consumption. Cassava does well on poor soils and is often eaten mixed with millet flour in areas where maize does not grow very well. Bananas yield throughout the year and can be weeded during slack periods. Coffee yields from the beginning of the long rains until the start of the dry weather in October-November. It is often picked on Saturdays and Sundays when children are not at school. Coffee should be weeded regularly in order to get a good harvest. The period of the greatest food shortage is from April until June. Little or no beer is made or sold during this period. There is also little cash income from the sale of crops. Animals can be bought cheaply. In June the first wimbi is harvested and the first onions can be sold. Maize can be sold locally in August and September to traders who take it to higher altitudes where it is harvested later. From November until February cotton is sold. In January or February the first and largest payment is made by the co-operative society which has the monopoly to buy up all the coffee.

Families 29 However, there are many people who process coffee illegally at home and sell it to smugglers who take it to Uganda. The advantage is that one does not have to lose time by going to the coffee processing factory. Moreover, one receives cash on the spot. But large-scale growers find it more advantageous to use the factories. They do not have surplus labour for processing and at the same time the overhead costs on transport are proportionately much smaller. In December large quantities of beer are sold and the demand for beef and poultry is great. Yet, people are unwilling to sell their animals which leads to a steep rise in prices. In January people have to pay taxes and school fees. Often the first payment for coffee is delayed and it becomes necessary to sell animals. Prices drop dramatically during this period. Heifers which could be sold for shs. 200 in December now cost no more than shs. 100. People probably believe that cash available in December would be used for buying new clothes and shoes for wives and children. Such purchases are always made shortly before Christmas. Husbands and fathers may therefore prefer to sell cattle at low prices in January, when such demands are no longer made, or when they have at least a good alternative way of spending money. According to estimates of the Department of Agriculture family incomes in Bungoma District averaged shs. 460 in 1966, shs. 360 in 1967, and shs. 860 in 1968 (Kenya 1970c). But it is admitted that these estimates are extremely rough. In 1962 people in North Kulisiru estimated their own average annual income from crops sold for cash at shs. 270, and other income, including salaries, remittances and sale of animal products, at shs. 320 (Wilson 1962). From coffee production statistics we can see that during a five-year period ending in 1968 the average earnings per coffee grower were shs. 170 per annum (Kenya 1970c). An estimated sale of cattle of 10% per annum would probably give about the same amount annually to each farmer. As we have seen, this would very much depend on the time of selling. But even in January a milch-cow or an oxen would seldom be sold for less than shs. 150. Onions yielded on average shs. 840 per acre in 1967 and in 1968 shs. 560 (Kenya 1970c). I am, however, unfamiliar with the size of the onion fields. Moreover, prices varied a lot throughout the season. But it seems very likely that most farmers who planted the crop earned more than shs. 150 from it. Farmers are requested by the Government to plant at least a quarter of an acre with cotton. This would give an additional income of perhaps shs. 30 - shs. 50. Thus it seems very reasonable to accept that the average farmer in Kulisiru earned a cash-income of at least shs. 500 per

30

Families

annum in the late sixties. Approximately the same is earned by a labourer in permanent employment by another farmer or a coffee co-operative society. It is very important that labour is available so that crops can be planted early. Wimbi must be sown before maize requires the attention of the farmer. Late planting of maize results in yields which are much reduced. The same is true of cotton. The yield of onions seems to be less dependent on the time of planting, but prices tend to be much higher early in the season. This is also true of the earliest maize and wimbi which are sold locally. One way of making sure that operations take place at the right time is to have the fields ploughed early. Oxen are generally used for ploughing, but farmers can only do so efficiently if there has been at least some rain to moisten the soil after the dry season. Not everyone owns a plough and oxen. The ones who do have them have a distinct advantage over those people who are forced to borrow or hire them. The same is true of the few tractors which are in use in Kulisiru. They are owned by individuals, partnerships or companies. With the help of tractors the owners can earn enough money to pay for the hire-purchase instalments or repay the loans with which they were bought. People who can afford to rent a tractor prefer it to ploughing with oxen. Planting, weeding and harvesting have not been mechanized. These operations depend on labour rather than appropriate equipment. People who can afford to pay for hired labour have a distinct advantage over those who are unable to do so in that they can now overcome certain crucial labour bottlenecks. At the time of the year that such bottlenecks occur there are sufficient people around who need cash immediately and who cannot afford to wait until their own crops mature for sale or consumption. It is possible to employ such people on contract terms and they can be paid either in cash or in kind. Other farmers employ labourers permanently. Such labourers are perhaps short of land or prefer the security of regular monthly payment to the uncertainties facing farmers who try to earn the same amount from their own farms. People who employ them often seem to use such labour as a substitute for family labour, especially when they or their wives have alternative wage-earning employment. The average farmer probably sees as the aim of farm management the minimizing of risks, rather than the maximization of profits. He wants to have sufficient food and cash all year round. Failing this he is forced to use his savings in the form of livestock and probably also run the risk of having to offer his labour for hire rather than invest it in growing crops for the following year, whereby he could still cover his losses. If he has a

Families

31

surplus he may use it for consumption or may save it in one form or another. Savings can be used for buying oxen and ploughs, or a bicycle, which is a time-saving and therefore productive investment. An important item of expenditure are the school fees. Poor people can refuse or postpone sending their children to school, thereby saving much money. In the four lower standards in the primary school the fees amount to shs. 42 - shs. 53 per annum, in the three higher standards they are shs. 83 per annum. The amounts charged for secondary schools vary considerably. Some are fully state-supported, others only partially or not at all. In fully supported schools fees amount to shs. 450 per annum, in the other schools shs. 700 - shs. 900 or more per annum. The Graduated Personal Tax and the Local Rate also form an important part of expenditure. The latter is shs. 17, the former is shs. 24 - shs. 600 and is supposed to represent 5% of one's income. Tax assessment is difficult, but in practice it appears that the chiefs and sub-chiefs divide the burden in such a way that the estimates, based on previous collections, can be reached without too much difficulty. People who are ostensibly wealthier than the rest are placed in a higher category. Some old people may be exempted. There is little or no attempt to define what a farmer's income is and even less efforts are made to assess his actual income and to induce him to pay 5%. On the basis of my own survey in Kulisiru it is clear that the variations are great. Out of 47 families only 25 were paying school fees for their children. There are many young families where children have not yet reached the normal schoolgoing age. There are also some old families who do not have children at school. The average cost of school fees per family is shs. 317, the average amount of tax paid is shs. 56. If we only consider the families who pay fees we find that the average fee amounts to shs. 595, the average tax to shs. 67. Significantly, all families paying more than shs. 1,000 in fees were assessed in the highest class of tax payers recorded by me, that is shs. 89 per annum. This figure excludes such taxes already paid directly by employers. Table 8 shows the differences. Table 8. School fees paid Kulisiru in 1969 Number of families

Amount of fees

14 3 4 4

shs. 249 and less shs. 250-499 shs. 500-999 shs. 1,000 and more

32 Families School fees are generally paid by the father, although full brothers may assume the responsibility for paying fees for their younger siblings if they themselves earn sufficient money. Occasionally they pay the fees of the child of another wife of their father, or for a cousin. This is most likely to happen in the case where the father is dead and his son acts as guardian over the younger children. It is my impression that this happened more frequently in the past than at present. The same applies to widow inheritance. Bukusu say that elder sons can take over the younger widows of their father. I recorded several instances of this happening in the past, but I do not know of any recent cases which occurred during the last 20 years. If wives enjoy a cash-income of their own, their husbands call upon them to help pay the fees of their own children. Women may cultivate their own annual cash crops or obtain an income from brewing beer or trading on the market. A profitable sideline is to produce sprouted wimbi, which is essential for making beer. This can be bought for shs. 8 per tin and after treatment sold for shs. 18 - shs. 24 per tin of four gallons. Some women work as teachers or nurses and receive monthly wages. I do not know the extent to which men control such incomes, nor do I know how much work husbands can demand from their wives on fields with cash crops which are completely controlled by men. Perennial cash crops seem to be mainly controlled by men, who will ask their wives to cultivate the coffee for them. However, with perennial food crops such as bananas, this appears to occur less frequently. In monogamous families problems of control over resources do not seem to be great. There are no competing units. Yet it is my impression that if there are several wives, the cash income of each matrifocal unit is kept separate. However, the role of the husband in such families varies considerably. Of course, Bukusu will tell you that a man must plough the fields for his wife and provide her with a house, with clothes and other necessities which must be purchased in shops. He must also clothe her children and pay their school fees. Wives on the other hand are responsible for cultivating enough food for themselves and their children and husband and must prepare it. Men should treat their wives equally well, if they want to avoid jealousy and quarrels among them. Yet Bukusu will also tell you that if a man has two wives, there is one whom he loves and another whom he hates. He will always help the wife whom he loves with the cultivation and buy her nice clothes, while the other wife will be neglected. These are stereotyped notions about what should happen and what actually does happen. However, such notions do not necessarily give a true picture of the

Families

33

actual situation as observed by me. But such a picture in the form of a typology or a model can only be constructed through generalizations based on detailed investigation of a sufficiently large number of cases. It would have been impossible for me to conduct such an investigation without giving up my other research priorities. Control of land and cattle is not nearly so complicated. These resources are in the hands of the husband and father, with the exception of some cows or oxen which women have bought with their own money. I have already mentioned the case of a widow who had become the registered owner of the plot which her husband could have claimed from his sister's son. But in all other cases men control the land. A father is obliged to give his son a piece of land if he remains on the same plot as his father after having established his own family. A son may buy additional land or migrate and occupy land elsewhere, but if he insists he will always receive a share of the land his father owned as part of the inheritance. Bridewealth for the first wife has to be paid by the father, provided he is still alive, or at least that part of the bridewealth which consists of livestock. Bridewealth negotiations are the responsibility of the fathers of the couple. They are assisted in such negotiations by their relatives and friends. If the father is dead, his brother or another agnatic relative may take such a responsibility upon him. First of all the number of cattle to be transferred is agreed upon. This invariably comes to 14 or 15 cows and three goats, except in the case of divorcees or girls who have borne children out of wedlock. In such cases the number is less. The agreement concerning the payment of these animals constitutes the legal bond of marriage, although very often the couple will have lived together for some time already. Parents usually wait with the arrangements of the bridewealth, if the woman is pregnant, until after the birth of the child. Not all animals must be transferred at once. A first payment may consist of the goats and four or five cows followed by the rest over a period of years. Only wealthy people who desire a wedding in church or in the DC's office will pay the entire bridewealth before the celebration of the wedding. The son-in-law is required to give a raincoat, a blanket and some cash to his father-in-law. Lately, fathers-in-law have tried to get as much as one year's earnings for daughters who enjoy a regular income as teacher or nurse. Negotiations over the amount start after the number of animals has been agreed upon. The negotiations seldom lead to a conclusion. The matter is then left to the father-in-law and his son-in-law, but in most cases no agreement is reached and the matter is left at that. The son-in-law

34 Families may pay several hundreds of shillings. However, I have not come across cases where the several thousands which were claimed have actually been paid. However, disagreement about cash paymens does not invalidate the marriage. A husband himself must pay for any additional wives he wishes to marry. A father is to some extent helped with the collection of the bridewealth by brothers and full paternal cousins and also by his wife's brothers and these obligations are reciprocal. However, the bulk of the bridewealth is his responsibility. Marrying an additional wife entails a considerable investment and it means that there are less resources available for other investments, for example providing for the education of one's children. In the past it was natural to marry many wives if one wanted to become important and influential. In those days when there were no schools, or only a few, and the economy was still very much in the pre-monetary stage, additional wives probably meant larger food surplusses. This meant that more visitors could be entertained with food and beer. It also meant that there were more children and therefore more bridewealth arrangements. A polygamist could create a larger kinship-network through the marriages of his children than a monogamist. But even more important in this respect were his own marriages, especially in view of the strong emphasis on friendly and co-operative relations between brothers-in-law. Probably sons were older upon marriage than daughters and this must have made it easier for a man with many children to create a cattle surplus which could be used for his own marriages. To some extent it is still true that additional wives mean more available labour. But on the other hand, if children receive more than primary education, additional children probably become a liability rather than an asset, at least until such time that children with secondary education can find jobs and can start contributing to their father's cash income, especially in order to help him pay the school fees. But in this respect it is important to remember that normally sons will only help with the education of their mother's children. In this way cattle which are used to acquire a second wife cannot be used to pay for the education of the children of the first wife, while the education of the children of the first wife will not make it possible to get more money for the education of the children of the second wife. It is generally recognized these days that a good education brings with it a higher status and a higher income. A good job provides money which can in turn be invested in large-scale farming and in commerce. Money can

Families 35 also be used to further a person's political ambitions. Such jobs require an education which was not generally available until the late nineteenfifties. Consequently, people in these jobs are young. Older people may be able to invest money in farming and trading, but they are seldom able to become politically important. Such men earned money in the early days of cash crop farming or through the cattle trade. Often people who were able to set themselves up successfully as farmers or traders in those days had accumulated capital for buying equipment or engaging in trade because they were employed as teachers or local Government officials within the colonial framework. In the pre-colonial days younger men could not attain positions of greater influence and importance than older men. On the contrary, it normally took a long time for a man to build a network of relations based on marriages made by himself and his children. In colonial times this changed because people who had some education were preferred as local Government officials and they were on the whole young men. But one still finds that such people found it necessary or at least advantageous to marry many wives themselves. This was probably to legitimize their status in the eyes of the ordinary man, who tended to think of political leaders as people with many wives and children, and also perhaps because they could only be effective rulers if they were in a position to rely on a network of relations based on kinship and marriage. They certainly did have the resources to marry many wives and in this way they were able to extend oldfashioned hospitality when entertaining guests and visitors. This trend must have been partly due to the fact that there were also still some old polygamists around who had been able to retain the influence they had gained in pre-colonial days. These people were appointed by the British, who were looking for Africans whom they could use to maintain law and order in their areas. They undoubtedly set an example for the younger and educated generation of leaders. Nowadays the perspective has changed radically. With the dawn of independence the centre of power for Africans shifted from the local or tribal level to the national level. Instead of trading in cattle or cultivating a hundred acres, Africans came to take over shops and wholesale businesses from Asians and to acquire large farms in the White Highlands with thousands of acres. Ambitious politicians no longer compete for the j o b of chief of a location, but for that of MP. Moreover, they are no longer interested in becoming chairman of the African District Council, but set their eyes on being a minister or an assistant-minister. Fathers feel that

36 Families if they themselves cannot participate in this new type of social situation as politicians or entrepeneurs, they should at least try to give their children the opportunity to attain a life style which is commensurate with the new cosmopolitan elite status. And if not at the national level, at least at the local level within the district or location. Status considerations are not the only factors which fathers take into account in emphasizing the education of their children. If their children are able to find employment they can buy a farm for themselves and their families which means that they no longer have to claim a piece of their father's land. The father in turn will have to share his land with fewer children and he can thus go on farming as he was wont to. Again, in old age parents with children in employment find that they can be helped more readily than when their children do not earn salaries. In many cases these are the only advantages they are likely to enjoy. Although the older generation expected to continue to exercise political influence through their sons, this has proved to be an illusion, because the level at which political competition is relevant has in fact shifted so much from the local and tribal to the national level. In 19691 made a survey of all people who were in paid employment and who lived or who were born and grew up in Kulisiru. These were classified according to formal education, training and income. I did not make explicit enquiries as to the amount of money earned by people, but I inferred their income from the salary-scale which applies to most people who were considered (Table 9). Table 9. Teachers' salary-scales in Kenya in 1966 Basic academic qualifications Primary, no KPE KPE KPE + 2-4 years School Certificate School Certificate Higher Sch. Cert. University Degree

Length of training 2 years 2 years 2 years 2 years 3 years 1 year 1 year

Kwenda (K.) 1967, p. 275 * One £ equals 20 shillings

Trained salary £*

Untrained salary £

P4 120- 180 P3 162- 264 P2 240- 456 PI 348- 726 SI 582-1110 SI 582-1110 Dip. Ed. 847-1710

84 96 108 240 240 300- 350 804-1710

Grade

Families

37

There are three people with university education or similar professional training: an advocate, an industrial development officer at the Ministry of Commerce and Industry and a Divisional Engineer with the Post and Telecommunications Services. The group below this top of the elite was defined by the educational achievement of the Higher School Certificate or some such similar training. They are four PI teachers, a district liverstock officer, the chairman of the Bungoma District Farmers' Co-operative Union, a clerk with the railways, a senior salesman of a drug firm, and a personnel officer of a medium-sized firm in Nairobi. The next category used by me were those employed people who had attained a Cambridge School Certificate or similar training. Their total number is 20. There are 7 P2 teachers in this group, 9 clerks, 2 medical assistants, an inspector of police and a district markets inspector. According to Lloyd (1966, p. 1) the term elite can appropriately be used to denote those who are western-educated and wealthy, suggesting an annual income of £250 as the lower limit. This definition would exclude those people with KPE and additional training, most important among whom are P3 teachers. But I think that we should also take into account whether the person concerned is living at home in a rural area with sufficient opportunity for subsistence agriculture and cheap housing, or in town, where the cost of living is much higher. Although no recent research on family budgets is available, it is my impression that an income of shs. 200 per month in town equals an income of shs. 50 per month in a rural area like Kulisiru. In other words a P3 teacher in a rural area can expect to save as much money as a P2 teacher in town, at least at the bottom of the salary-scale. The total number of people in this category originating from Kulisiru is 74. They include 19 P3 teachers, 16 clerks, 11 policemen, 5 artisans 4 nurses/midwives, 4 drivers, 3 salesmen, 3 prison-wardens, 2 veterinary assistants, 2 subchiefs, 1 bus-conductor, 1 Soil Conservation Service foreman, 1 National Youth Service instructor, 1 railway-engine driver, 1 locational marketing master. Of this number 23 live in South Malakisi Location, another 11 elsewhere in Bungoma District on their own farms, or at least with the advantage of the low cost of living in rural areas. These people are in a much better position to make investments in farming than others who have no regular employment. They enjoy a regular income and can save, and they are also much more likely to be given credit for the development of their farms. Significantly, six out of the ten partners in Kulisiru, who raised money for buying a tractor,

38

Families

belonged to this category. Because they have a regular income they can take advantage of the seasonal shortages of cash with which the other farmers are forced to cope. They can also buy additional land and employ labourers to cultivate it for them. They are also better placed to pay fees, even for expensive private schools, than most of the other farmers amongst whom they live. If we consider the elite in rural areas as a group which differs from the rest of the population in respect of the weight given to them by criteria which measure prestige and the way they conceptualize these differences, my objective categorization appears to be very relevant (cf. Long, 1968, p. 199). Although I did less systematic research into these aspects than I probably could have done, while observing and participating in everyday life, I became aware that there were marked differences in the way people dress, the consumer durables which they have at home, the type of house in which they live and perhaps most important the way in which they spend their leisure time. For example, drinking beer takes place in two different settings, at least around Sirisia where there is as yet no canteen for native beer. First of all there are old women who prepare beer every fortnight or so, sufficient to entertain a large number of people. People from the immediate neighbourhood sit around large pots sipping beer through long tubes. They do not normally include women younger than 35 years, who are not expected to attend beer-drinking sessions at all, nor young men under the age of 30 years. The latter drink their beer mostly from small tins and sit together in a separate hut or room. A popular day for beer brewing is market day when people on their way home drop in at the parties held. People at these parties speak Bukusu and may be entertained by musicians who play a fiddle or harp and who sing traditional or topical and satirical songs. In the larger shopping centres there are bars which sell bottled beer. The staff consists mainly of young women and entertainment is often provided by people who play guitars and at times there is a record player. People who frequent these bars are normally much better off than the ordinary peasants, being generally teachers, policemen, chiefs and sub-chiefs and medical assistants and nurses who work in local health-centres. The language of communication on such occasions is preferably English. Especially on teachers' pay days there are large gatherings at such bars. The use of English is not confined to bars, however. In fact, it is used frequently by secondary school students and clerks and teachers from the lowest grades upwards. This is one way of demonstrating that they con-

Families

39

sider themselves better than the ordinary uneducated Bukusu. Many of them are said to be proud people who do not show the ordinary civility which one could expect from one's fellows. It is also usual for female teachers and nurses to marry other teachers, clerks, health assistants and so on. Economically, such families are important because they enjoy a double income. Women are not expected to give up paid employment when they get married. People who marry women who are much less educated than themselves may even feel ashamed. I know a case of a young PI teacher who obtained a transfer from his home area to the farthest corner of the settlement scheme, at his own request, where nobody knew him. He had first intended to marry a P3 teacher, but her brother objected to it. He then married one of his pupils instead, who was in Standard V or VI. Apparently he felt that he might be exposed to ridicule if he remained in his home area.

2.3 Family types and the influence of Christianity Continual social change during the last two generations has had its effects on the development cycle of domestic groups. It is impossible to think in terms of a recurrent process in Kulisiru. The ideal of a large family with many wives and children was abandoned by at least some of the converts to Christianity. They were among the first who invested resources in education for their children rather than in additional wives. Others could afford to do both. They were often also Christian converts but reverted to polygamy once they had become local government officials. Yet their income made it possible for them to give their children also a good education. Although this second educated generation, which now constitutes the cosmopolitan and most of the rural elite, is still rather young, it seems likely that they will not find it necessary to set themselves up as polygamists with many wives in the future. Instead of gaining prestige in the eyes of the ordinary Bukusu they are likely to lose it among their fellow elite members. They may still do so, if they think that a second highly educated wife may be an advantage, but those who have already married a well-educated wife are unlikely to do so. These women are even more against sharing their husband with an additional wife than ordinary Bukusu women. They can also stand on their own two feet, because a job will give them economic independence. They would, therefore, rather risk a divorce. Elite members too would certainly suffer a great loss of prestige

40 Families if they were to neglect the education of their own children. Polygamy on a large scale would soon make the burden of school fees too heavy. Yet, there are still many polygamists in Kulisiru. In my own sample of 10% of the registered plots I recorded 48 married men and 87 married women: 19 men, 40%, had more than one wife. There were 13 men with two wives, 2 with three wives, 1 with four wives, 2 with five wives and 1 with eleven wives. Some of the men had wives who did not live on the plots selected for the sample. They were included in the analysis of polygamy. In the same area I selected two wards for intensive investigation of kinship links. In Namawanga there were 38 married men and 55 married women: 17 men had more than one wife, 45%. In Namawanga there were 11 men with two wives and 2 with four wives, including those not living in the area. In Ndakalu there were 54 married men and 91 married women: 16 men had more than one wife, 30%. There were 7 men with two wives, 4 with three wives, 4 with four wives and 1 with eleven wives, including those not living in the area. In the AID survey of North Kulisiru there were 97 married men and 168 married women: 42 men had more than one wife, 43%. There were 28 men with two wives, 5 with three wives, 3 with four wives, 3 with five wives and 3 with more than five wives. Women are almost invariably against their husband taking a second wife. They put up a resistance and may even decide to seek a divorce. But normally after a stormy period things simmer down and first wives acquiesce in the situation. Marrying other additional wives never causes such an upheaval. If a man has more than one wife it is customary for women who are no longer child-bearing and who have married sons to retire from active married life and to go and live with one of their sons. Sometimes a first wife may even tell her husband to look for another wife if she prefers such an arrangement. I have recorded quite a few cases of men who had spent the greater part of their life faithfully obeying Christian rules concerning monogamy but who took a second wife when their first had become too weak and old to cook their food, draw water, wash their clothes and help them with cultivating. In Kulisiru the average number of years which elapsed between the marriage of the first and second wives was 17 years, the median was 10 years. These figures refer to extant marriages only, not to those which ended in divorce or death.

Families 41 Table 10. Interval between marrying first and second wives Number of men

Duration of interval

3 5 5 2 3 1

0 - 4 years 5- 9 years 10-14 years 15-19 years 20-24 years 25-29 years

If it is true that women generally speaking prefer to be the only wife of a man, why are there still so many women willing to become the partner in a polygamous union? This is because divorcees or women who have given birth at their parents' home, i.e., without being married or at least living with the man who feels responsible for her pregnancy and is willing to marry her properly, find it very difficult to find a bachelor who would be willing to marry them. The few cases of which I know concern men who are quite old and who seem to have failed to convince other girls to marry them. Their only opportunity seems to be to marry a woman whom other men do not want. But most of these women have no other choice than marry a man who already has a wife. Other girls seem to be of the opinion that to become the second wife of a wealthy and important man is preferable to getting married to a young and perhaps poor bachelor. Divorce is not unusual. In Kulisiru I recorded the particulars of 109 unions. Of those 87 were still extant, 22 had ended in divorce and 10 with the death of one of the spouses. This gives the following ratios: A (22/109) 20.2%, B (22/32) 68.7%, C (22/99) 22.2%. I have no reliable data on premarital births, but again they are not at all unusual. While premarital intercourse is perhaps not entirely accepted, it is certainly expected.Once a girl gets pregnant, her parents will try to find out who is responsible and the girl is likely to be sent to the home of the suspected father. If he refuses to marry the girl her parents may acquiesce and be satisfied with receiving one cow, which establishes a man's legal right to the child. Or they may simply do nothing and arrange a marriage later on with a man who already has another wife. Others go to court and claim damages of up to a thousand shillings. Unless explicitly excluded bridewealth is also paid for establishing legal rights over any children born previously to the woman, if they are not claimed by another man. Quite obviously this would happen when bridewealth is arranged after a child

42

Families

has already been born to the couple that wants to be legally married. While concubinage occurs I again do not have reliable figures. However, I believe that it happens infrequently. Such cases as do occur are mainly due to the fact that parents have not yet settled the bridewealth for a young couple. In the case of divorcees the difficulty may be the arrangements concerning the return of the bridewealth, whereby the previous marriage would legally be dissolved. In such cases women will live with a man and be treated as if they were one of his own wives. Some women may prefer to live on their own and have more or less permanent relations with men without getting formally married to them. Children born out of such unions belong to the clan of the mother. The examples I know concern women who had jobs of their own and who were economically independent. One of the women had become separated from her husband by whom she had had two children, another had been deserted by the man who had promised to marry her and by whom she had become pregnant. They preferred such an arrangement rather than becoming an additional wife in a polygamous household. If we look at the age at which people get married for the first time we find that the figures obtained from the Kulisiru sample are consistent with the prevalence of polygamy. Men are on average 22 years (median 23), women on average 17 years (median 17) old, when they get married or start living together with the intention of getting married. Moreover, there are more women than men in Kulisiru. There are 96 men of 15 years and older, and 61 men of 23 years and older. There are 120 women of 15 years and older, and 102 women of 18 years and older. Table 11. Recorded marriages of men and average age of men at first marriage in Kulisiru by decades Decade 1910-19 1920-29 1930-39 1940-49 1950-59 1960-69

First marriage 1 10 8 11 7 11

Subsequent 1 0 12 9 7 15

marriage

Average age 16 23 21 24 20 26

The correlation between the average age at which men marry for the first time and the incidence of polygamous unions as shown in Table 11 is

Families

43

suggestive at first sight, but the numbers are too small to draw any definite conclusions. For example, if I would leave out one man with eleven wives as a statistical rarity the number of subsequent marriages made between 1930-39 would drop to six, and for the period 1940-49 to five. Nevertheless, the increase of polygamy in the nineteen-sixties is probably due to the fact that many young men are now forced to spend a much longer time at school than was previously the case. Students generally do not get married and girls who become pregnant by them are now more likely to be left in the lurch. Table 12. Population by age and sex in Kulisiru and Bungoma Kulisiru Age group 0- 9 10-14 15-19 20-24 25-29 30-39 40-49 50 and over

male% 37.0 15.0 11.5 10.5 4.0 6.5 3.5 12.0

Bungoma District

female %

male %

female %

36.5 11.3 11.3 7.0 6.5 6.1 9.1 12.2

39.2 14.6 11.1 8.2 5.3 7.6 5.3 8.8

37.7 14.3 10.4 8.0 5.7 9.7 5.7 8.6

If we compare the population of Kulisiru and Bungoma District (Kenya 1970b), divided by sex and age, we see that the proportion of women to men in Kulisiru is larger than for the whole of Bungoma District and the discrepancy is particularly large in the age group 40-49. This probably reflects the inadequacy of the sample which dealt with only 200 men and 230 women. Nevertheless, in Bungoma there are also considerably more adult women than men. As I have said before in a situation such as obtained in Kulisiru we cannot make very definite conclusions about the typical development cycle of domestic groups. Their composition cannot be solely explained in terms of recurring cycles. They are no longer recurrent because of the diversity of opinion about the desirability of polygamy, the differentiation in wealth and the changing influence of these factors during the last two generations. Even in the precolonial past there must have been a difference between those influential and wealthy notables who married 10 to 15 wives, and the majority of the population among whom polygamy was perhaps even less

44

Families

widespread than it is nowadays. I believe that it is more appropriate in such a situation to give a few cases which show the range of the possible variations. Case 1. A. was born in 1902. He was employed as a teacher by the FAM 1923-1928, was chief's clerk in Malakisi Location 1928-1936 and chief of Malakisi Location 1936-1944. A. married his first wife in 1924 and he paid the bridewealth for her himself because his father was deceased. Between 1931 and 1937 he married six wives and between 1941 and 1946 another four wives. He was and still is a wealthy cattle owner. He has his family in three different compounds, one to two miles from each other. A. has 50 children. Seven small children are still at home, sixteen are at primary school, eight children are at secondary school, one of whom is in Form V. Eight daughters are or were married. Two of them are midwives, the others have no job. One nubile girl is just at home. Seven of his sons had found jobs, but two of them are at present abroad for further training. One has been in East Germany since 1958 for medical studies, the other studies accountancy in Glasgow. Of the others one is a graduate of the Kenya School of Law and now works for a firm of advocates, another is clerk with the Kenya Police in Nairobi, two other sons are clerks with commercial firms in Kitale and Kampala respectively, while a third is a sign-painter employed by the Kisumu Municipal Council. One son was trained as a mason and is now self-employed, another never finished his primary school education and is now working for a teacher as an agricultural labourer. He is not yet married and lives at home with his mother. In 1969 A. paid shs. 89 tax and approximately shs. 3,600 school fees. Some of his children were paid for by others. The son who works in Kampala paid for his younger brother's stay there and also for his secondary education. A brother of A. who works with the County Council paid the secondary school fees of another son, while the son of this brother, who is a teacher, pays the primary school fees of a daughter. She also lives in his home. Another daughter stays with her sister who has married a medical assistant who pays her primary school fees. Yet another daughter stays with her mother's brother in Uganda. He also pays fees for his sister's daughter. Case 2. B. was born in 1900. He worked as a farm labourer in the White Highlands 1923-26. He married his first wife in 1923 and had to pay the bridewealth himself because his father was deceased. Between 1930 and

Families 45 1939 he married two other wives and again another two during the next decade. He has 24 children. Four small children are still at home and six children are at primary school. Het has seven married daughters, none of whom received any education. Four of his sons are married; one is a P3 teacher, the others are only farming. One son is trying to find employment in Kitale and one daughter is expected to marry soon. She has just finished her primary school education. In 1969 B. was assessed to pay shs. 41 tax, but he hoped to be exempted because of old age. He paid shs. 262 school fees. Three of his children have their fees paid for by their elder full brothers or the half-brother who is a P3 teacher. Case 3. C. was born in 1897. He worked for periods up to six months as a farm labourer in the White Highlands before he got married. Later he was shop assistant with an Asian trader in Kimilili from 1938 to 1943. He is a faithful Quaker who remained monogamous. He married in 1926 and had to pay the bridewealth himself because his father had died. His youngest son is still at primary school. One daughter is married and his other three surviving sons are also all married and have jobs. One is a medical assistant, another a clerk at a post office and the third is a P3 teacher. In 1969 C. paid shs. 65 tax and shs. 83 school fees. Case 4. D. was born in 1907. He was a farm labourer in the White Highlands from 1930 to 1947. He lived there with his family. He married in 1926. He had to pay the bridewealth himself because his father had died. He has three married daughters. His two married sons work in Kampala, but they are not well educated. They got no more than four years of primary education. One son left four small children with his parents, the other son has his three small children with him in Kampala. In 1969 D. was exempted from paying tax because of his old age. Case 5. E. was born in 1914. In 1950 he found work with the Department of Agriculture and since 1954 he has been working as a road-maintenance worker with the Locational Council. This gives him a regular income of shs. 100 per month. He was able to pay the fees of a private secondary school for his eldest son, who is now 29 years. He has been unable to find employment and is as yet unmarried. Another daughter is married, four children are still at home and four children are at primary school. In 1969 E. paid shs. 101 tax and shs. 291 school fees. Case 6. F. was born in 1919. He married in 1943. He and his wife are still

46 Families practising Catholics. He has eight children. Four small children are still at home and two sons are at primary school. A girl of 16 had to stay at home in 1969 because there was no money for fees. The eldest son, who is 22, was sent to a private school but had to give up for the same reason. He now lives with relatives elsewhere. In 1969 F. paid shs. 41 tax and shs. 125 school fees. Case 7. G. was born in 1918. He worked as a voluntary teacher in an ungraded FAM-school 1937-1945. In 1950 he attended a one year smallholders' course at Bukura. He was assistant to the sub-chief 1941-1960. In 1965 he became manager of a tractor acquired by a partnership of ten people. He married his first wife in 1935, his second in 1945 and his third in 1963. He has 17 children. Two small children are at home, ten children are at primary school, two sons are at a private secondary school, one son completed secondary school and is now working with the East African Airways in Nairobi. He is single but two other daughters are married. In In 1969 he paid shs. 89 tax and shs. 1,892 school fees. Case 8. H. was born in 1921. He married his first wife in 1941, a second in 1956 who left him in 1958, and a third in 1969. For the latter two no bridewealth arrangements have been made. He has nine children. Four small children are at home, two sons are at primary school and three daughters are married. In 1969 H. paid shs. 41 tax and shs. 125 school fees. Case 9.1, was born in 1921. He worked as a mason for the African District Council 1950-1960. He is a good Anglican and married his first and only wife in 1946. Six small children are at home, three children are at primary school, one son is at a government secondary school. One daughter is married. In 19691, paid shs. 65 tax and shs. 858 school fees. Case 10. J. was born in 1926. He worked as a farm labourer in the White Highlands 1947-1955, and 1962-1965. In 1965 he became guard of the dam near his house, which brings in shs. 35 per month. He married his first wife in 1945. She died in 1965 and he re-married in 1966. Two small children are at home, five children are at primary school and one daughter is married. In 1969 J. paid shs. 41 tax and shs. 209 school fees. Case 11. K. was born in 1930. He worked in Kampala as a tractordriver for a road building firm 1954-1968. Here he became a Muslim. He married in 1966 and had to pay all the bridewealth himself because his father had died. He has one small child. In 1969 K. paid shs. 41 tax.

Families

47

Case 12. L. was born in 1936. He had eight years of education but failed his KAPE in 1955. He worked with an electrician in Nairobi 1955-1960. He is a good Salvation Army member and married his first and only wife in 1962. He has four small children who are still at home. In 1969 L. paid shs. 65 tax. Case 13. M. was born in 1936. Before he got married he worked for two years as a farm labourer in the White Highlands and also during a year for an Asian trader in Uganda, just across the border. He married in 1955. Two children are at home, two are at primary school and two died young. In 1969 M. paid shs. 41 tax and shs. 95 school fees. Case 14. N. was born in 1937. He worked in the White Highlands 1959-1962. He married his first wife in 1959, his second in 1964 but separated in 1965 and a third in 1969. No bridewealth has been returned for the second wife and no bridewealth has been negotiated for the third wife. He is his father's only son and he paid bridewealth in both cases. He has five small children who are not yet at school. In 1969 N. paid shs. 41 tax. Case 15. O. was born in 1942. He is a P3 teacher. He says that he is a good Anglican and that he intends to solemnize his marriage in church. In 1967 he married a fellow student, who is also a P3 teacher. He himself paid half of the bridewealth so far, his brother contributed a little less than half of it. His father is no longer alive. He has one small child. In 1969 O. paid shs. 203 tax and his wife shs. 125. Both of them are working. In the case studies I indicated that some people were 'faithful Quakers', or 'practising Catholics'. This is because I found it necessary to make a distinction between those people who formally joined a Christian grouping at one time or another, and those who are still 'good Christians'. This is the way Bukusu themselves talk about it. You ask if someone else is a good Quaker and get as an answer: 'No, he married more than one wife'. I mean by a good Christian someone who keeps the rules which distinguishes his denomination from other denominations and from nonChristians, who pays church fees and who goes to church regularly on Sundays, or in the case of Catholics, who attends Mass whenever the priest comes to the location to celebrate it. I collected data on all the good Christians in Kulisiru. Three-quarters of the population in this area which

48 Families

>

N N m

I

¡S m „ ^ ^ rn

|

I I

OS « -H

i

rn

I

00

I

N N rt VI as

I I

" a

I i

I

i i

i

i -< — i

I

I

•n N —< Os o cs

I

I

fl ID v-l vo in00

I

I

N ^

-- VI "•t