When the miles came : land and social order in Buganda, 1850-1928


Material Information

When the miles came : land and social order in Buganda, 1850-1928
Physical Description:
x, 300 leaves : ; 29 cm.
Hanson, Holly Elisabeth
Publication Date:


Subjects / Keywords:
History thesis, Ph.D   ( lcsh )
Dissertations, Academic -- History -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1997.
Includes bibliographical references (leaves 294-299).
Additional Physical Form:
Also available online.
General Note:
General Note:
Statement of Responsibility:
by Holly Elisabeth Hanson.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028628352
oclc - 38863550
System ID:

Full Text





Copyright 1997


Holly Elisabeth Hanson


First of all, I want to thank many Baganda, now in the realm of ancestors, who

cared passionately about events in their kingdom, and made the effort to write about their own lives, to record successions, to write down in detail the arguments in a case, and to remember events in songs or sayings. I have relied on people's clear perceptions of how their society was changing with the coming of British overrule, and their brilliantly incisive criticisms of those changes. The Ganda thinkers who left their ideas where they could be found a few generations later have given my story both its structure and its life.

My second debt is to the scholars who have shaped my development as a historian. I am immensely grateful to Steven Feierman, who in his own work and in his interaction with me about mine gave me the tools to consider social change in the past. I want to thank David Schoenbrun, not only for making concrete the early history of the Great Lakes region, but also for asking me hard questions in a way that inspired confidence that I might be able to answer them. In his own teaching and also in his untiring efforts to bring scholars of Africa to the University of Florida, R. Hunt Davis created the learning environment which I enjoyed so much. I also appreciate the contributions of the members of my committee who were not African historians--Murdo M. MacLeod, Sheryl Kroen, Abe Goldman, and, at earlier stages, Goran Hyden, Carol Lansing, and Jeffrey Needell.

I am also thoroughly indebted to historians of Buganda whom I have encountered primarily through their written work. M.S.M. Semakula Kiwanuka's works on Ganda history have been my reference companions. Anthony Low's edited volume on Buganda thought gave me the idea for my topic, and John Rowe's vivid and insightful view of nineteenth century Buganda taught me how to pursue it. The clarity and rich detail of Michael Twaddle's biography of Kakungulu and Michael Wright's history of the civil war gave me entry into this difficult period of Ganda history. Henry W. West's precise and thorough explanations of what happened with land in Buganda were the keys that unlocked sources that would otherwise have been impossible to understand.

I could not have produced this dissertation without the generosity of other scholars of Uganda who shared source materials that were no longer traceable in Uganda. Foremost among these is John Rowe, whose work collecting archives in the early 1960s is evident to anyone who uses the Africana Collection at Makerere University Library; Professor Rowe graciously allowed me to look at his copy of the records of the Lukiko and his translations of the Ekitabo kya Kika kya Nsenene, and gave me a copy of the rare, privately circulated pamphlet "The Bataka Land Holding Question." Richard Waller gave me a copy of his unpublished S.O.A.S. Master's paper, and shared research notes. Glenn McKnight shared information on the Butaka controversy from files which had been misplaced in the Entebbe Secretariat Archives before I arrived there; I have also benefitted from materials directed to me by Mikael Karlstrom and Michael Tuck.

I am grateful to the Uganda National Council for Science and Technology for

permission to do research in Uganda, and to the Makerere Institute of Social Resarch for iv

providing me with an academic home. Innumerable people in Uganda received me with kindness and facilitated my work: among them were Christine Deborah Sengendo, Nakanyike and Ssegane Musisi, Judy Butterman, Mark Markwardt, the staff of the Africana Collection at Makerere University Library, the staff of the Mailo Land office, and the staffs of various courts. Anke Alemayehu, Rod and Dawn Belcher, Elizabeth Kharono, and others welcomed me and my children in ways we will never forget.

I received research support from the Fulbright Fellowship program of the

International Institute of Education, and from the University of Florida I received a Center for African Studies pre-dissertation research grant, a College of Liberal Arts and Sciences Dissertation Fellowship, a Ruth McQuown Fellowship, and Department of History travel funding.

I am deeply grateful to Kathryn Burns, Jan Shetler, Peter Von Doepp, Kiran Asher, Marcia Good Maust, Rebecca Karl, Tracy Baton, and others for their steady encouragement, and for helping me connect intellectual endeavor with the rest of life. I want to thank my family for giving me models of engaged scholarship. The support of my religious community --including Margaret Mattinson, Joanne Schwandes, Harriet and Sam Stafford, Jeanne and Jose Diaz, and many others--allowed me to meet the needs of my children and the deadlines of graduate school at the same time. Finally, I want to acknowledge my children, Corin Olinga Vick and Rebecca Margaret Vick, who avoided stepping on all the papers on the floor for many years, and who moved to Uganda and back in the middle of secondary school. I thank them for their courage, fortitude, grace, and good humor.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulffllment of the Requirements for the Degree of Doctor of Philosophy WHEN THE MILES CAME: LAND AND SOCIAL ORDER IN BUGANDA, 1850-1928 By

Holly Elisabeth Hanson

December 1997

Chairman: Steven Feierman
Major Department: History

The dissertation examines social change in the East African kingdom of Buganda in the decades preceding and following British colonial intervention, using documentation generated by an abrupt transformation in the pattern of land ownership in 1900. In considering the period of early Ganda/British interaction primarily from the perspective of Ganda written records and Ganda institutions, it challenges commonly held perceptions regarding colonialism and economic transformation in Africa. Ganda leaders interacted with British officials with expectations of mutual respect, incorporated new social forms such as private land ownership into Ganda structures of authority, and resisted the commodification of social relationships even as they adopted wage labor and commercialization of trade.


The dissertation identifies the political and social relationships encoded in land control before the mid-nineteenth century, arguing that Buganda was not a centralized despotism, but rather that overlapping and diffuse forms of authority characterized the kingdom prior to that time. Long distance trade in ivory and slaves undermined Ganda forms of authority, leading to a prolonged period of civil war. The dissertation asserts that this conflict ended not through British intervention, but when Ganda chiefs re-ordered the kingdom by associating control over land with religious allegiance.

The dissertation argues that class distinctions in Buganda did not emerge as a result of the creation of private land ownership: the chiefs who became land owners initially attempted to maintain relationships of mutual social obligation with the followers who became their tenants. Instead, new kinds of social distinctions emerged as a result of the excessive labor demands of colonial authorities, which altered the relationship of chiefs to followers. The dissertation demonstrates that the passionate protest against Mak (privately owned) land in the 1920s known as the "Butaka Controversy" was a direct critique of the new colonial order. The complainants asked for the restoration of all the positions of authority that had characterized Buganda in the past, and described comniodified social relations as a form of enslavement.



ACKNOWLEDGEMENTS ............................................ iii

AB STRA CT .................................................... ... vi


1 INTRODUCTION ............................................ 1


Households and Banana Gardens .............................. 15
The Logic of Kusenga: Attaching to a Chief ..................... 17
Production as the Enactment of Meaning ........................ 24
Remembering the Lives of Ancestors ......................... 25
Remembering Constitutional Events .......................... 31
Following Tribute Up: Overlapping Forms of Power ............... 34
Power at the Center of the Kingdom ........................... 44
Ebitongole: Kabakas Control Innovation ........................ 51
C onclusion ............................................... 53

3 CHIEFSHIP, LAND, AND CIVIL ORDER ......................... 55

Buganda and the Trade in Ivory and Slaves ...................... 57
The Dissolution of Authority ................................. 63
Buganda's Civil War: Social Violence with Religious
C ategories .............................................. 72
Re-allocating Land to Make the War End ........................ 85
C onclusion ............................................... 99


Mailo Allocation and Authority in Buganda in 1900 .............. 104
Ganda Meanings for Land Applied to Mailo ..................... 108
Mailo, Ancestors' Bones, and the Translation
of C ulture .............................................. 113
Mailo Allocation and the Locations of Power in Buganda ........... 122
Challenges to the Social Logic of Kusenga ...................... 128
Claiming Authority ...................................... 129


Cash and the Calculus of Kusenga .......................... 134
Tax in Rupees and in Labor .............................. 141
Conclusion ............................................. 149

Too Much Work: New Labor, Old Tribute, and the
Possibilities of Cash for Cotton ............................. 155
Work Obligations in the New Buganda ...................... 157
Responding to Overwhelming Labor Demands ................ 169
Immigrants and Independent Ganda Women .................. 173
The Missed Meanings of Labor Exchange ................... 181
The Deterioration of Kuseng ............................... 182
The Decline of Lineage Networks and the Threat of
"B ad H eirs .................. I ........................... 194
Innovations to Meet the Responsibilities of Chiefs and Lineage
N etw orks .............................................. 204
Conclusion ............................................. 209

6 THE ORDER OF MILES ON TRIAL ........................... 211

The Complainants and the Logic of their Case ................... 213
Mailo and the Young Kabaka's Power ........................ 222
The Grammar of Omusango ............................... 227
Daudi Chwa's Attempt to Rule and its Aftermath ............... 230
Colonial Power on Trial on a Colonial Stage: The Multiple Meanings
O f B utaka .............................................. 232
The Tenuous Intersection of Discourses of Power ................. 240


Mailo Shattering the Foundations of the Kingdom ................ 245
Critiquing Colonial Rule .................................... 249
The Harmful Constriction of Forms of Authority ................ 250
The Inappropriate Exercise of Power ..................... 257
Arguing about the History of Power in Buganda ................ 260
Arguing about the Ganda Social Order ........................ 264
People Turning into Things: Private Land Ownership as
Enslavem ent .............................................. 269
The Possibility of New and Old Together .......................... 278
Conclusion: Beyond Bakungu and Bataka ......................... 285


BUGANDA ................................................ 290

GLOSSARY ...................................................... 293

WORKS CITED.................................................... 294

BIOGRAPICAL SKETCH .......................................... 300


Lush banana gardens support the dense population of the ridges on the northern edge of the great East African inland sea, where the Buganda kingdom emerged about five hundred years ago. People used this fertile land not only to produce the means of their subsistence, but also to define relationships between people in the kingdom. Ceremonies of "showing the land" cemented the connection of wives to husbands, of followers to chiefs, and of regional leaders to the kabaka, the king of Buganda. Neighbors and their children gathered to witness when a newcomer was "shown the land." A bark cloth tree was planted, intermediaries received gifts, and children planted bushes that indicated the borders of the granted land. In Buganda, the power to allocate land meant power to rule the people who cultivated the banana gardens on that land, and each person who had the power to grant land had a place in the complex web of authority. More important ceremonies, carried out by a designated messenger of the ruling kabaka, marked a kabaka's decision to give power--and land--to one subordinate and to take it away from someone else.

A unique and profoundly significant series of ceremonies "showing the land"

occurred in every part of Buganda in 1900 and immediately thereafter. In the aftermath of the protracted war that followed Buganda's involvement in long distance trade in ivory,


slaves, and guns, the victorious Christian chiefs and their British allies had agreed to reallocate the land of the kingdom. In Ganda terms, new controllers of land logically followed new alignments of power. This time, however, the victorious chiefs (not the kabaka, then an infant) had made the allocations, and they had agreed to give half the land to the British, after all "ruling chiefs and notables" of Buganda had received their share. The leading Christian chiefs gave themselves the largest amounts of land and people, but took care to give estates of land varying in size from one to twenty square miles to almost four thousand other chiefs and figures of authority in Buganda. Since the British wanted to secure their portion of the land as a potential means of attracting European settlers and making a profit for their fledgling colonial endeavor, they insisted that the new allocations be marked and measured following European practices of landed property, as wen as Ganda practices of "showing the land."' People called the land allocated in this way majLQ (estates of land measured in square miles). b4AUQ land had implications beyond demonstrating a shift in the hierarchy of power in Buganda. Although Baganda giving and receiving mak did not recognize it at the time, the new rituals for land allocation, including land surveys, the distribution of certificates, and writing names in the land registry, gave maL the characteristics of private property, and this permanently foreclosed the possibility that any future kabaka might manage shifting power alliances by making new assignments of land.

'After waiting fifteen years for their turn to choose, the British got swampy lowlands, rocky hilltops, and Buganda's least arable land.


In creating Mak, Ganda chiefs and British colonial officers supported each other because they agreed on the importance of asserting their control over the land. Neither the Ganda nor the British recognized, in 1900, the huge gap in their intentions for the use of that land. The ruling Ganda chiefs focused on inscribing the new order of power, with themselves at the top, in the familiar form of chiefly control over land. The British focused on the creation of protectable private property, and on obtaining a share themselves. The coalescence of these two fundamentally different concerns generated unusually rich sources for African social history, ones that provide insight into the structures of an east African society before the tumultuous changes of the late nineteenth century. The documentation of mailo land contains types of information that are difficult to recapture from oral historical narratives.

The chiefs' distribution of land articulated their perception of the structures of power in Buganda, because they attempted to allot a mailo estate to each person in authority. Their decisions were recorded with due solemnity on land certificates, surveyors'maps, and land registry fists. Those who felt their authority merited estates and who had not received them immediately attacked the Mak allocation with statements explaining their right to land and power based on generations of remembered clan or family history. The British protectorate carefully recorded these counterclaims. Twenty years later, Ganda social discontent crystallized in a tumultuous public protest about mak land. The cuhTiination of this protest was a public enquiry under British colonial auspices in which people explained what Ganda society had been like in the past in order to prove


that the allocation of land--and the organization of society--had been utterly corrupted by mak. Their thorough critique of the consequences of colonial involvement in Buganda received a full hearing under British colonial auspices because British authorities felt obliged to investigate claims that involved the violation of property rights.

This dissertation uses Ganda actions regarding land to perceive social change in the kingdom at a crucial period in its history. It traces the outlines of the relationships which people had encoded in control of land, observes the impact on those relationships of Buganda's nineteenth century crisis in authority, and records the transformation of those relationships under the impact of British colonial demands for labor, tax, and obedience. Since Ganda chiefs, followers, and clan leaders made the decisions about land, my strategy focuses attention on Ganda ideas and Ganda intellectual endeavor. The dissertation discusses Ganda notions of the proper way to organize society, the creativity of Ganda leaders who used land allocation to re-create order during and after the seemingly interminable war of the late nineteenth century, and Ganda ability to integrate new and old ways of thinking when people chose to involve themselves with the British. By paying attention to the meaning of land, it is possible to see the 1920s conflict over land as an articulate Ganda assessment of the failures of British colonial intervention. In a sense, this is an intellectual history of Ganda use of the idiom of land as power.

My conclusions about the Buganda polity are quite different from those of most published histories of Buganda. Those works, based on research carried out during the transition from colonial rule to independence, emphasized the consolidation of central


power in the hands of kabakas, and argued that the militarism of Buganda in the late nineteenth century had characterized the kingdom for several preceding generations. In Chapter Two, I use clan histories, Ganda epic tradition, claims regarding land in the 1880s and other late 19th and early 20th century sources to argue that power in Buganda was diffused through a layered network of multiple forms of authority. I argue that the coexistence of overlapping forms of authority in the 19th century indicates a tendency to avoid conflict and seek compromise as Buganda institutions developed, and that people who held office oriented themselves to each other as weH as to the king.

In Chapter Three, I describe the collapse of Ganda structures of authority in the nineteenth century. Building on recent scholarship which identifies the introduction of caravan trading in ivory, guns and slaves as a transforming crisis for East African polities, I interpret the actions of nineteenth century kabakas as attempts to regain authority they had lost because of the fundamental disruptions initiated by the caravan trade. Using testimony of participants and a close reading of the records of Captain Lugard (who has hitherto gotten credit), I argue that Ganda chiefs ended civil war and reimposed order in their country through incorporating new religions in the arrangement of control over Buganda's ten provinces.

Using land allocation to understand social relationships also generates a new way of looking at the causes of the drastic social changes that occurred in Buganda in the early 20th century. Mailo land owners have been blamed for the emergence of vast differences in social class that happened at that time. One version of the story is that scheming, selfish


chiefs sold the nation to the British in return for huge estates of mailo land, and the masses who were not granted mailo land were shut out from the means of personal advancement. British Protectorate officials themselves accused the mailo owners of laziness and greed because they lived off the rents of their tenants instead of turning their land into profitable plantations. These explanations fad to consider the logical motivations Ganda chiefs might have had for re-allocating land, and later for refusing to turn their followers into wage laborers, as the Protectorate officials wanted them to do. Also, the story of greedy landlords leaves out the consequences for Baganda of British colonial exactions of labor and tax.

In order to develop a more complete understanding of the consequences of mailo, it is necessary to keep in mind that all interactions involving productive resources have both econornic and social (moral) components. People in Buganda produced bananas and other foods and manufactured goods in order to survive, but Baganda did not emphasize the economic aspect of production in their explanations of their society. Until the early twentieth century, Ganda production almost invariably had meanings and utility beyond the provision of subsistence: the production of things created and demonstrated connections between groups of people. British colonial officers arrived with a different way of thinking about production. Colonial decision-makers in Buganda attended to the economic dimensions of production and ignored the social ones. For most of the colonial era, British employers insisted that work was a purely economic transaction even when doing so drove workers away from their plantations and factories by the thousands.


Baganda, including mjik owners, did not stop seeing social obligations inherent in control of land even when the land came with a certificate of ownership. I argue that in 1900 chiefs wanted the authority and prestige that came from having followers, and the allocation of mals land was a statement of a new order of power in the kingdom, not a pre-emptive grab for resources that would soon have economic value. The labor demands that made life unbearable came from colonial exactions of one to two months' labor for tax and one month's obligatory labor on top of the obligations of tribute and labor for chiefs and the kabaka. Chapter Four describes the decisions made by Ganda chiefs in allocating mailo, and their attempt to include important meanings, such as the continuing power of deceased Kabakas, in the form of private property. Chapter Five outlines the effects of excessive labor demands on the relationship of chiefs to followers, and suggests that class differences emerged in Buganda as people with school-taught skills obtained exemptions from obligatory labor.

The massive social protest in the 1920s, known as the Butaka controversy, has

been understood as a further working out of the centralizing tendency in Buganda political development. In this view, the protestors were disgruntled clan elders complaining about having lost land to the chiefs appointed by the kabaka. A careful examination of the records of the 1 920s dispute over mailo land suggests that much more was going on. The complainants were not only clan elders, but also all the other kinds of people who had had authority in pre-colonial Buganda, and whose authority had diminished under the regime of mW1k. They asked for the return of butaka (clan graveyards) and their other lands, but


they also asked for the restoration of all the positions of authority and the patterns of decision-making that had characterized Buganda in the past. Chapter Six documents the participants in the case against mailo, and explains how their incisive critique of colonial power was misunderstood as an argument about graveyards. Chapter Seven explores the arguments made by the Ganda leaders who brought the case against mailo. They wanted a return of the Ganda pattern of rule which had more positions of authority, more participants in decision making, and more compromise. They claimed that their children were "enslaved" by the commodification of social relationships, and stated that people in power ought to consider the well-being of the people they were ruling. They claimed that progress would be most effective if it incorporated "the good customs of Buganda," and if change happened slowly. The epilogue sketches why these aspirations could not be met, describes the Busulu and Nvujo law of 1928 which defined the Ganda obligations of chiefs to followers in cash terms, and shows how Baganda retained their expectations regarding land and social obligation as the cash economy developed.

The sources I have used to examine the intellectual creativity of Baganda who reshaped their social institutions to make them work in new circumstances could also answer other interesting historical questions. The specific controllers of land before the time of mailo are discernible in Ganda epic tradition and in statements by clan elders and others who disputed the mailo allocation. These might contribute to an elaboration of Ganda history from the beginning of the kingdom until 1800. Who received mailo, and who it was passed down to, was recorded in carefully filed provisional and final


certificates of ownership in the Land Registry Office, in the Ekitabo kya Obusika (Book of Succession) of the Buganda kingdom and in land survey maps made between 1908 and 1914. These documents could be used to answer fascinating questions about the origins of a land market: what land was sold, and what land was not sold, and why? The ways that people infused meaning into plots of land can be glimpsed in scattered explanations of events in the epic tradition, in the chronicles of 19th century events written by Miti, Kaggwa, and Nsimbi, in reports of 19th century travellers and in the ethnographic works of Roscoe and Mair. Records of land cases heard in district level courts between 19 10 and 1970 contain transcriptions of pointed arguments between litigants concerning their mutual obligations in relationships mediated by land. In fieldwork in Uganda in 1993 and 1995, 1 participated in occasions of "showing the land" and listened to people's stories of the history of their land. These sources could be used to extend a study of social transformation in Buganda from 1928 until the present.

It is important to point out that my work takes place in a context in which people ask different kinds of questions of the past than the first generation of historians of Buganda. In the 1950s the semi-autonomous Buganda kingdom ran smoothly, managed from the imposing Lukiko (parliament) building by clerks with typewriters, and the Lukiko itself deliberated in a room with Westminster-like benches. Writing as the colonial era drew to a close, historians saw in the Buganda past a progressive centralization and bureaucratization which had allowed the kingdom to advance. Whether they intended to or not, their story of Buganda offered the suggestion that African nations might become


more modem through the imposition of strong central control. In contrast, this history of Buganda is written against a backdrop of decades of civil unrest and anxiety about African governance. It may not be a coincidence (although it was not my intention) that my story of Buganda describes institutions that once created civil order but have been irretrievably lost.


Ganda banana gardens were shady and cool, they produced lots of food, and the spirits of ancestors hovered in the play of light and shadow among the trees. In the nineteenth century, and for several hundred years before that, dark green banana gardens covered the middle heights of the hills and ridges in the region north of the great East African inland sea, the Nyanza. In these gardens, married women grew the food that fed their families, and in household compounds their husbands made the beer and barkcloth the family owed to a chief for the use of the land. In particular banana gardens, people gathered at the graves of lineage ancestors for ceremonies marking birth, growth, death, and inheritance. The gardens also supported the people who moved along the wide, straight roads from compounds of chiefs to the center of the kingdom, offering tribute and labor to the king and to the other powerful figures whose authority contributed to the rule oftheland.

For more than a thousand years before the Buganda kingdom emerged, people

living around the Nyanza had sustained themselves through a mixed agricultural system of cultivating grains and yams, herding cattle, and fishing. Then, during the period from 900 and 1200 AD, environmental stresses caused people to experiment with alternatives to


mixed agriculture, and intensive banana cultivation was one result.' The Luganda language broke off from its parent language North Nyanza at some time between 1200 and 1500, which is about the same time period that Ganda epic traditions are assumed to originate. According to Ganda origin myths, Kintu the first man and Nambi the first woman arrived in the land that would become Buganda with the first shoot of a banana tree.

How the people who grew bananas so successfully around the northern rim of the Nyanza created the kingdom of Buganda is a contentious question. In order to discern the early history of the kingdom one must interpret Ganda epic tradition, clan histories, and the placement of clan butaka (burial grounds of important people) on a foundation of awareness of the significance of banana cultivation. Banana cultivation made land more valuable than it had ever been in the past, and the historical linguistic record shows that new social institutions developed as bananas became central to people's subsistence strategies. Historians have come to various conclusions using this evidence. Perhaps Kintu, an immigrant, was the first king. Perhaps there had been five kings before Kintu, all members of Ganda clans. Perhaps Kimera, third in the dynastic list derived from epic tradition, was a Nyoro prince who founded a sub-Bito dynasty; perhaps he was a leader bringing hsown followers from the western region.'

'David Lee Schoenbrun, A Green Place. A Good Place: A Social History of the Great Lakes Region. Earliest Times to the 15th Centur, Portsmouth, N.H.: Heinemann, 1997, manuscript 295,290, 320. Note that references refer to page numbers in the manuscript, not in the published book.
2MS.M Semakula Kiwanuka, A History of Buganda from the Foundation of the Kingdomtol9QQ New York: Africana, 1972, 32, 39-41; Benjamin Ray, Mvil. RiitaL and Kingship in Buganda, New York: Oxford University Press, 1991, 101; Christopher C. Wrigley, Kingship and state: The Buganda dynasty, Cambridge: University Press, 1996.


However the kingdom originated, over the centuries people in Buganda found

ways to organize their fives and relationships with each other that ensured prosperity and cohesion for the kingdom. In this chapter, I argue that two social forms shaped the developing institutions of the kingdom. These were kuseng the attachment of followers to chiefs who gave them land, and the organization of production for remembering important aspects of the past. Using Ganda epic tradition and information on relationships among various parts of the polity in the mid- to late nineteenth century, I suggest that Baganda used kuseng and the pattern of remembering to create many overlapping forms of authority. As a result, the Buganda kingdom was characterized by both central order and diffuse authority. After describing production in the Ganda household, the nature of kuseng and patterns of using productive relationships to remember important things, I show how the structures of the Ganda kingdom described in epic tradition combined these forms. In the eighteenth century kabakas began to appoint another kind of chief, the ekitongole; this development of Ganda social forms allowed kabakas to control and manage innovation.

In contrast to the predominant understanding among historians of Buganda as a

despotic, highly centralized kingdom, I see the polity as one that had many nodes of power and authority. My conclusions are different because I have chosen to make different interpretations of three important sources of information. First, I have not assumed that nineteenth century travellers'and missionaries' accounts describe long-standing characteristics of the kingdom; those early visitors observed Buganda at a time of violent crisis. It is possible to construct a sense of Ganda social institutions before these


upheavals, which were caused by caravan trading, through a careful use of other sources. Second, I have not accepted at face value one of the pivotal documents used by historians of Buganda in understanding the development of the kingdom.? In the huge controversy over the morality of inaki land in the 1920s, which is the subject of Chapters Six and Seven below, Apolo Kaggwa justified his action in taking vast amounts of clan land by arguing that Ganda kings had been taking land away from clans for generations, and therefore his land-grab had merely followed long-established traditions. The group who brought the case against mailo claimed that power relationships in the past had been more fluid and subject to negotiation, and I have attempted to balance Kaggwa's statements about the development of the kingdom with those of his opponents. The third reason that my interpretation of power in Buganda departs from the predominant one is that I have tried to follow the record of exchange--to understand social relationships in Buganda based on how tribute flowed from followers to their superiors. Since some of the recipients of tribute were people whom foreign visitors would probably not have seen as powerful, their role was underemphasized in the descriptions of Buganda provided by missionaries and early colonial officers.

While my information about kuseng tribute, chiefship, and other forms of

authority comes largely from nineteenth and twentieth century sources, some evidence from earlier times confirms my suggestions about Ganda society before the late nineteenth century upheaval. Stories narrated in the Ganda epic tradition impart information about

31 show in Chapter 7 that a memorandum prepared by Apolo Kaggwa for the Butaka Land Commission regarding the relationship of kabakas and clans has been has been adopted uncritically by successive generations of historians of Buganda.


patterns of social interaction in Buganda several hundred years ago. Detailed drawings of the placement of compounds in the capitals of Kabaka Suna (who reigned from 1824 to 1857) and Mutesa (who reigned from 1857 to 1884) provide social maps relevant for an earlier time. David Schoenbrun's study of interlacustrine society until 1500 demonstrates the antiquity and historical antecedents of social institutions and relationships I describe.' Households and Banana Gardens

We can see from more recent descriptions something of the historically-rooted style of marriage and household production in the region. Ganda society was built on households in which women produced food and children and men produced manufactured goods and maintained formal networks of social connections. Near the cool, shaded quiet of each banana garden was the home of the family it supported. A man received a kibagj~ (a plot of land) from a chief when he wanted to marry, and once the land had been cleared and the banana shoots established, cultivating the growing trees was the responsibility of his wife. In Luganda the verb still used when a man marries is okuwa&s which literally means, "to cause (someone) to peel bananas." The verb used for a woman's marriage is passive, okfmbra and it means "to become the cook (for someone)."' Women in Ganda households were responsible for growing bananas and other crops, and for cooking; finding a wife could also be called "finding a hoe.",6 A man brought gifts of

4Schoenbrun, GrenPlac.

5 Lucy Mair, Native Marriage in Bugand IAI Memorandum 19, 1940, 13.

6John Roscoe, The Baganda: An Account of their Native Customs and Beliefs, 2nd.
ed., New York: Barnes and Noble, 1966 (first ed. 1911), 92; Mair, An African People. 8788.


barkcloth, beer, and food to his new wife's family to compensate them for the value of the labor that he was taking away. If thereafter a woman objected to the treatment she received from her husband, she returned to the home of her brother. From his home, the possibility of improving the marriage would be negotiated; if that proved impossible, the woman's brother would return the bridewealth to her former husband, and she would stay with her brother. Women who had been given away to chiefs, or captured during raids on Buganda's neighbors, and for whom no bridewealth had been paid, were "wives of the tired hoe," who could not be divorced because there was no bridewealth to return.

Husbands expected wives to produce children as well as food. Women who had been unable to have children were sometimes sent away by their husbands, and a woman might live apart from her husband after her children grew to adulthood.7 Baganda approved of widows who "remembered" their deceased husbands by continuing to live in the same place, cultivating the same gardens.8

Baganda families lived in large circular homes made of neatly trimmed elephant grass, with the roof sloping to the ground. In the same courtyard, there were smaller structures for cooking, for young men to sleep in, and for adult men to gather and talk. Nineteenth century travellers were struck by the sturdy construction and meticulous neatness of Ganda homes, which, at that time, were built in one or two days by

7 Jan Bender Shetler observed households of independent barren women in Northern Tanzania, personal communication.

'Cf. Buganda Law Reports, Civil Case No 38/46, p.41-44, which excoriates families that disturb widows.


neighborhood work parties.9 Bark cloth partitions divided a home into separate rooms. The homes in a neighborhood, separated by each family's gardens, stretched along the fertile middle ground of the hills characteristic of Buganda--swampy land at the bottom of the hill was uncultivated. A family's compound was distinguished from that of its neighbors by mpiwy, a border-marking bush that was planted by children during the ceremony in which the family was "shown" the land it was receiving. The street connecting houses in a mita neighborhood/ridge led to the compound of the chief, the figure of authority who had granted land for cultivation to each of the families in his area, who listened to disputes, and who marshalled the people in his (or her) area for tribute or service to the king.

The LD&giDL Kusenga: Attaching to a Chief Kusng the act of attaching oneself to a chief, was one of the fundamental forms of social cohesion in Buganda. The particular meaning the word kusng has in Luganda-indicating an exchange of service, allegiance, and tribute in return for a kibaqi (plot of land) and protection--is older than Luganda itself. According to Schoenbrun, the word emerged between 800 and 1000 A.D. in West Nyanza, the ancestral speech community that eventually split into Rutara, (the parent of Lunyoro, Runyankore, Ekihaya, and others) and North Nyanza (the parent of Luganda, Lusoga, Lugwere, and Orusyan). As Schoenbrun explains, "To ask for land, in West Nyanza societies, at the end of the first millennium A.D., was also to enter into a net of social obligations."' The mutual

9Mair, 123.

'OSchoenbrun, MS, 328.


expectations inherent in the kusenga relationship structured the productive activities of Ganda men and women and their aspirations for well-being and prosperity.

Allocation of a kibagi was the critical first element in the kung relationship. An oral tradition of the great loyalty of Nkalubo to Kabaka Ndawula, whose reign began sometime around 1700, describes this process. Nkalubo and another person emigrated from Sesse to the mainland, and became the men of a chief named Nawandugu. Nkalubo then decided to leave that chief, and went to serve the kabaka. He was given one plot, but then, because of the respect he showed, he was given a better plot, closer to the king's palace." As a result of his actions on behalf of the kabaka, a chieftainship was created for him, and successors to that chiefship assumed his name as their title of office. The sign of chiefship was control over land that one could allocate to people who would be followers. In the pattern we know from the nineteenth century, a person might attain this status by succeeding to the leadership of a branch of a clan, by appointment to a senior chiefship by the kabaka, or by appointment to a sub-chiefship by a senior chief. Chiefs attracted followers using their ability to allocate land. The Ganda likened chiefship to the light of torches that burned at night. Baganda chiefs were said to initially dismiss paraffin lamps, saying :"what will become of our torches? How will a chief be able to hang onto a torch?

""There were two brothers who went to live under Nawandugu at Lubu, having
immigrated from Sese. Their names were Nkalubo and Miingo. Nkalubo decided to leave Nawandugu and to become the Kabaka's man being presented before the throne by Sewankambo, and receiving plot of land near the palace from the Kabaka then reigning-Ndawula. From day to day Nkalubo paid a visit of respect to the Kabaka and soon he gained favour, the outward sign of which was a new plot nearer to the palace where Sebugwawo now resides." Basekabaka, is. version at Makerere, p. 73.


Surely it is the lamp-torch which is adhering to the chief?"'2 Chiefs, like torches at night, collected people around them.

After attaining a chiefship, a person might be demoted, but would never be left

with no followers.3 Part of the tremendous emotion attached to the creation of mailo land in 1900 came from the fact that chiefs lost their land, and had to become followers instead if chiefs. One person testified, "I had been a mutaka (a chief by virtue of clan leadership) but I settled down and became his man and cut reeds for him."'4 Another clan elder described how in 1900 he had been "turned out" of the historically important clan land that he had controlled, and had gone to the Lukiko (the assembly of Ganda chiefs, like a parliament) three times for the return of his land but failed to get it. He explained, "so I packed up my belongings and went and settled on the kabaka's land in Kikulu and became a mere peasant."'5

12 Apolo Kaggwa, Ekitabo kya Kika kya Nsenene, Mengo: AK Press, n.d., Manuscript translation of Professor John Rowe, 17.
13In the early 20th century, Ganda chiefs tried to explain to British Protectorate
officials that retiring a chief to the status of ordinary citizen was really unthinkable. A long correspondence about whether the man who had been an important chief, the Pokino, from the 1890s until 1924 ought to be required to pay tax demonstrates how seriously the Baganda considered this problem, and the failure of the colonial administrators of the 1920s to fathom Ganda political structures. Jarvis, the Governor's secretary, refused to acknowledge that having to pay tax would humiliate the once-great man, "The ex Pokino should be well able to pay the small sum demanded as Poll Tax. Personally I should like to see all exemptions abolished. When a European official retires he is still called upon to pay all his taxes!." ESA, A46/1315, SMP No. 4345.

14Commission, 425.

"5Commission, 458, Pasikale Bambaga.


Although on the surface kuseng may appear similar to feudal relationships in medieval Europe, its workings were very different. The relative abundance of land in relation to people shaped the character of the relationship. The Luganda vocabulary regarding kusng indicates that people chose to form these relationships, and could also undo them. A follower could kuseng "join a new master, settle, immigrate," and he could also k nguk "leave a master/chief, move away from." A chief could kusenza, "receive (newcomers into an area)," or he could kungus, "cause people to move away." Since a chiefs standing was dependent on having lots of followers, the terms of kusng favored the followers. Ganda proverbs speak of followers as people who had choices. "Musenze alanda"--"The follower often changes his master" and "Busenze muguma: bwe bukonnontera n'osongola"--"Service is like the digging stick: when it has become blunt, you point it again."'6 In the late nineteenth century, people left their chief if another chief seemed to present better opportunities; even a page in the kabaka's court could report that he had left his position in the palace because "they ruled him badly."'7

Baganda described kueng as beneficial in explanations to early twentieth century ethnographers; clan histories and the recorded epic tradition suggest people perceived balance and mutual benefit in the relationship in earlier periods.'" The forms of exchange

16 Ferdinand Walser, Luganda Proverbs, (Kampala: Mill Hill Missionaries, 1984) proverbs numbered 2938, 1034.

17Lubwuma interview, 6/4/1995; Commission, 540, Danieri Serugabi; Kaggwa, Nsenene, 14.
18 The sense that the relationship between leaders and followers was mutually beneficial has great antiquity: cf. Schoenbrun, ms. 186.


marking kuseng expressed the reciprocity people expected to experience in the relationship.'9 A chiefs men built elaborate reed fences that encircled his compound as part of their service to him; a chief protected his men from other powerful people who might claim their labor or service. The chiefs representative planted a barkcloth tree as part of the ceremony of "showing the land" in which the follower received his plot of land; the follower then gave back to his chief graceful, dark red cloths made of bark from that tree pounded and stretched in several days' careful labor. Followers took their chief part of every brew of beer they made, and chiefs offered beer to people who came to their compound.20

People spoke and wrote about kusng as an on-going exchange of gifts. Chiefs needed loyal service, and followers who served loyally needed to be rewarded. It is interesting to note that the Ganda epic tradition recalls a chief of Kabaka Tebendeke (the eighteenth Kabaka, who probably ruled just before 1700) who lost his position as keeper of the royal tombs and was killed because he asked too often for gifts, instead of waiting to receive them.2 Chiefs showed their gratitude to people who served them well with gifts of barkcloth, women, and cattle.22 Apolo Kaggwa, who held the highest chiefly office,

'9Schoenbrun notes that kiba4j means both debt and banana plantation, thus revealing "with stark efficiency the elision of inequality with access to land." 312, ins.

20Mackay, 197; Roscoe, "Enquiry."

21 Apolo Kaggwa, Basekabaka be Buganda. Typescript of English translation by Simon Musoke. Africana Collection, Makerere University Library, 58.

22A.M. Mackay, Pioneer Missionary of the Church Missionary Society in Uganda, by his sister, London: Hodder and Stoughton, 1891, 208-9.


Katikiro (Prime Minister) for forty critical years at the turn of the century, defined the logic of kuseng in his description of the actions of Christian refugees during the wars of 1889. According to Kaggwa, "Kabaka Ntale (king of Ankole) liked us, he gave us many estates and five tusks and about thirty or forty head of cattle." To show their appreciation (and also to further their war aims), the refugees raided Ganda cattle for the Nkole king. They then built Ganda bridges for him over the Luzi river. As a result "we became more loved by Kabaka Ntale who gave us more cattle."23 In the early twentieth century, people told Lucy Mair that they expected to receive "meat, beer, and politeness" from their chief; and records from the first decades of the century describe a chief giving a favored follower bridewealth contributions and barkcloth on important occasions.24

The highest reward a follower could receive was appointment to a subordinate chiefship. Ganda epic tradition describes chiefships that were created as rewards to loyal followers, and the practice did not end with the attempt by British colonial officers to rationalize chiefship.25 In 1924, the Omuwanika Stanislaus Mugwanya, who had been one of the Regents who received a large amount of land in 1900, made a point of emphasizing his relationship with a person who was testifying against him in the case against mailo. Mugwanya identified the witness as one of his followers, and remarked "You have

23Apollo Kaggwa, Basekabaka bya Buganda, Ms. in Africana Collection, Makerere University Library, 110/149 (doubl pagination).

24Mair, 183; Africana Collection, Makerere University Library, Kaggwa Papers, AR KA 43/52, Sitefano Serwange to Apolo Kaggwa, August 1913.

25Kaggwa, Basekabaka ms. version, 74.


cultivated a very nice garden at Buganga for which I thanked you and made you a chief in consequence. 26

The logic of kuseng and the bonds it created between Baganda of different statuses, can be glimpsed in a story of the failure of foreigners to understand how they were to behave as part of this system of reciprocal exchange. Sometime in the 1890s, a nine-year-old Ganda boy (who is not named) joined the household of the Protestant missionary C.W. Hattersley. He stayed there working for Hattersley for nine years. When he wanted to marry, he asked the man he had served for assistance in acquiring a plot of land. Hattersley told the young man to ask his father for help, but the young man said (as Hattersley remembered the conversation), "When I came to join your establishment I gave myself entirely to you. Since that time you are my father; I have no other. Were I to apply to my father, he would only refer me to you." Hattersley, however, did not think of himself as obliged to the young man. He had employed him for nine years, and now he was employing other boys, "and with my short pocket I cannot be always helping boys who have left me." He also explained to the young man that if he helped him with the plot of land, the young man would also want help with the dowry, then with wedding clothes, then with the wedding feast. He told the young man "It is very difficult to understand where such requests are going to end." Trying to get Hattersley to recognize his role, the young man explained

Sir, you altogether fail to understand the customs of the Baganda. Do you not know that the more requests we make the more we show our love for

26Commission, 424, Stanislaus Mugwanya.


you. Were it not that I greatly love you, I would never ask you for a single
thing. We never ask anybody we dislike to give us a thing.

The missionary replied, "Perhaps in this particular case less love and fewer requests might suit my pocket better." The young man responded

Sir, it distresses me much to hear you talk thus. I came to you because you
are my father. You have been in Uganda many years, and I thought you
knew our customs thoroughly. I hope you will never make such a remark to those who know you less than I do. At present they look upon you as a
great friend.2

However disappointed the young man might have been in his patron's lack of understanding, he did not give up on the missionary. Several months later, when Hattersley said he hoped the young man would serve as a housemaster at the Mengo High School, his former servant replied, "I have already told you that I am yours; that you are my father and I belong to you. If you say I am to come back and be a master in the High School, it is for you to command and for me to obey. ,28 The missionary, along with many other foreigners in Buganda, failed to understand that receiving the service someone offered created a relationship that did not begin and end with the payment of wages.

Production as the Enactment of Meanin Attaching to a chief through kusenga gave people a material place--a banana plantation to farm, a source of support, and a channel for their ambitions. Another fundamental process connected people to essential metaphysical realities--the act of remembering. In the long history of Buganda, people invested their productive energies in

27C. W. Hattersley, The Baganda at Home, London: Frank Cass and Co, 1968 (first ed. 1908), p., 189-90.

28 Hattersley, 190.


ways that enabled them to maintain the memory of meaningful people and events. The growing of food, manufacture of goods, and raising of children took place in configurations dedicated to, and named by, things that were important to remember. These acts of remembering had to do with the present and the future: they organized and defined relationships between people, and brought into peoples' lives the protection and assistance of able spiritual resources. One kind of remembering was the connection people maintained with their immediate and distant ancestors in lineage networks and clans. Another kind involved the continuation through generations of exchanges that had once taken place; these reenacted exchanges shaped Buganda as a cohesive entity. Remembering the Lives of Ancestors

Paths in Buganda took families from their own homes to those of their neighbors, to the compounds of their chief, and also further, to the home of the mutak~ (head of a line of descent from a remembered ancestor). In the banana garden of the person who had succeeded to the position of mutak relatives gathered to observe ceremonies marking birth, growth, death, and succession. These ceremonies took place in the nineteenth century, and people explained to early twentieth century ethnographers that remembering ancestors was critical to the well-being of Ganda families. Schoenbrun explains that in the society that preceded the kingdom of Buganda and its neighbors in the region, the gift of life (mwy-Q) and the physical force of life (lbugal ) were "joined together in the living," and when the body died, "what had been the life force of the living body, mwoyo, became the life-force of the disembodied spirit, muzimu." A xnuzirnu was a real entity, but it "could only be present in 'this' world (the land of the living) through acts of memory by its


descendants."29 Baganda remembered their ancestors in the banana gardens that contained the graves of generations of forbearers and asked their ancestors for protection. Before the arrival of Western religions, bzimu (spirits of ancestors) intervened in the lives of their descendants to assist and guide them, as well as to punish them.30

The power of distinguished ancestors over the living can be seen in the actions the kabaka had to take to free himself of the influence of the dead in an area that he chose as his capital. Whenever a kabaka moved his capital from one location to another all of the graves in the location of the new capital had to be removed. (The habit of moving the capital appears in Ganda epic tradition in the time Kabaka Mutebi, who probably reigned in the mid 17th century, and the statement that bones were always removed from the site of a capital appears in the epic tradition in a story about Kabaka Suna, who ruled in the first half of the nineteenth century).3 Since kabakas had sovereignty over all the land, it seems likely that graves had to be removed in order to eliminate the power over the kabaka's actions that the people buried there would otherwise wield.

29Schoenbrun, ins., 358-9.

301n the 1920s Mair was told that people paid attention to the spirits of ancestors to avoid their potentially malicious interference, but Gorju, writing about Ganda religion as people remembered it from the 19th century, described the impact of ancestors' spirits as beneficial. Mair, 225; Julien Gorju, Entre le Victoria. L'Albert et LEdouard, (Rennes: Oberthur, 1920), 220, ff., Schoenbrun, ins, 356-7. Schoenbrun notes that according to Welbourn, in Luganda the word zimu implied a long departed ancestor, and misambwa had the meaning of recently departed relatives. Schoenbrun, 364, citing Welbourn, "Some Aspects of Kiganda Religion." Uganda Iun 26/2(1962): 171-82.

31Apollo Kaggwa, Basekabaka bya Buganda, ins. translation in Africana Collection, Makerere University Library, 88/5; Kiwanuka, King, 44, 117.


Baganda conceptualized descent from specific remembered ancestors as a tree

branching out into branches and limbs. The kika (clan) divided into secondary units called 5sg" (branches) and smaller ones called mituba (twigs). A clan consisted of large numbers of people who thought of themselves as children of the same original forefathers, never married each other, and identified with each other by sharing names and a totem.3 The clustering of clan burial grounds suggests that several hundred years ago, members of each clan lived primarily in one particular area of the country.3 As more people moved into Buganda and royal institutions developed, the connection between clan membership and access to land became less direct. In the nineteenth century, and probably in the eighteenth, people maintained clan and lineage connections with relatives who did not live near them, and a muak~ had followers on clan land who were not members of the clan.3

After clans had dispersed so that their members lived all over the country, people relied on hospitality from fellow clan members--immrediately identifiable by their clan names-wherever they travelled.3 People observed important occasions with members of aig or millib and leaders of these units controlled succession. The clan as a whole provided labor to maintain the shrines of Ganda royalty who came from their clan, and

32Schoenbrun suggests that the kika (clan) was probably the means through which people obtained access to the best banana growing lands as the practice of intensive banana cultivation emerged between 900 and 1100 AD., is., 305.

33 Kiwanuka, 94.

34Z. Kisingiri, "Enquiry into Native Land Tenure in the Uganda Protectorate," Uganda (Kingdom); Bodleian Library, Oxford; Shelfmark MS Africa s 17.

35 Ndawula interview.


performed particular tasks for the kabaka. In Ganda epic tradition whole clans were responsible for the transgressions of one of their members, and evidence from the late nineteenth century suggests that persons who incurred debts or fines could approach clan leaders for help.36

The most important location for remembering ancestors was the butaka, a banana garden that contained graves of important members of the clan or lineage network. One witness before the 1924 Commission defined butaka as "the place of birth of anyone where his ancestors and forefathers have lived and were buried. And every chief whether Mukwenda, Sekibobo, or even the Katikiro himself when he dies he is buried on his butaka land."37 Banana gardens that contained important graves were controlled by the people whose ancestors were buried there, and not by chiefs who had authority over contiguous land. Before the twentieth century, only distinguished members of a lineage were buried in a butaka; the graves of ordinary people were not visited or honored.38 The mutaka Zedi Zirimenya Buga explained, for example, that his clan had always had a butaka at Mangira: "We were on that land when Kabaka Kintu [the first kabaka] came, and he found us there." In all the generations "from time immemorial" until 1924, however, only sixteen graves had been made on the land. Other butaka had graves "that cannot be numbered," but even those must have been the graves of important people, not all the

36Kaggwa, Basekabaka 69; Kiwanuka, Ki=gg, 67; Lukiko Record, 13, 29/5/1905.

37Commission, Aligizanda Mude, 333.

38Mackay, 196; Lubwuma interview; Commission, 540, Danieri Serugabi.


members of the clan.39 People who had been powerful during their lives belonged in the butaka after their death, where they would be remembered and invoked to continue to assist their relatives.

New butaka were sometimes created by the kabaka to commemorate the lives of very important people. People who had been chiefs in the nineteenth century described the process in 1904. They said that people who had gathered around a distinguished leader in his lifetime continued to live in the vicinity of his grave, and more people might choose to come to live there and remember him after his death. After three generations of descendants were buried in the place of the grave of a "man of importance," the area containing the graves--and the surrounding gardens occupied by people engaged in remembering the buried ancestor--became butaka. This meant that the ability of chiefs to require labor or service was diminished, just as it was on the ancient butaka that had been incorporated in the kingdom as it developed.' Since people might choose to live near the grave of a particularly powerful leader who had been a member of their lineage, the

39Commission, 438, Zedi Zirimenya; 443, Semei Sebagala Kyadondo.

40 Testimony of Apollo Kaggwa and Ham Mukasa, "Enquiry into Native Land Tenure." It is important to note that this document has been widely misinterpreted as meaning that butaka was created by the burial of three generations of one lineage in the same place. Both Kaggwa and Mukasa stated that the process was unusual and only happened in the case of an important chief. Roscoe, perhaps in eagerness to see an equivalent of private property in pre-colonial Buganda, wrote about the process as a general one, Roscoe, 134. Morris Carter, attempting to define Ganda land tenure for the Uganda Protectorate High Court, made the same inference from the "Enquiry" testimony, "The Clan System, Land Tenure and Succession among the Baganda," Uganda Protectorate Law Reports, 1(190410):99-120. Roscoe and Carter's interpretation has been taken as authoritative by other scholars.


kabaka and others with authority over land were especially careful about where such people were buried. If the kabaka, or a clan, did not want to surrender control of a particular area of land where an important personage had been buried, the successor to the important person would not be allowed to be buried in the same location.

Exactly this kind of conflict over the potential creation of a butaka took place in the early nineteenth century at Senge. The details of this conflict, which emerged in disputes over the allocation of mailo land, provide an important insight into social relationships and the process of creating butaka before the transformations that took place at the turn of the century. Kidza had been the Kimbugwe, one of the most important chiefs of Kabaka Suna. When the Kimbugwe Kidza died ("before the arrival of Mr. Speke the first European,") he had been buried on a plot of land that had been given to him by the head of the Mbogo clan merely for growing food. Since he was an important chief, the Mbogo clan elders feared that his relatives would begin to gather to live around his grave, successive burials would turn it into a butaka, and the land would be lost to the Mbogo clan. They negotiated with the Kimbugwe's clan to ensure that the important man's grave would not remain on their land permanently. They "were very anxious to have the body of Kidza, a member of the Nsenene clan, removed from our butaka land, but the members of the Nsenene clan begged us to allow them to keep it there until it was quite dry when they would disinter it and take it to their butaka land.'41 Baganda made significant efforts to bury important members of their lineage and clan in the appropriate butaka, where their memory would be best preserved and their enduring influence experienced by the group.

4Commission, 371-2, Luisi Majwega.


This involved travelling to take the body of an important person to the appropriate location for burial if he (or she) had died in another place, and also exhumation of the bodies of significant members of the lineage who had not been buried in the butaka.42 Men who had been leaders in the late nineteenth century described this kind of disinterment as normal; witnesses before the Butaka Commission in 1922 described both reburials and the tragedy of important men who had been buried in inappropriate places."3 It seems reasonable to surmise that people had given the same careful attention to the burial of important people in the eighteenth century. Remembering Constitutional Events

The strategic assemblage in a butaka of powerful people's graves had significance beyond the prayers and hopes of their descendants. As the institution of kingship developed, people found ways to incorporate the pre-existing centers of power and authority, manifested in butaka, into the emerging entity of Buganda. The butaka themselves, and the pattern of finding purpose and order in remembering, were extended and adapted to create the structures of Buganda as a nation. One aspect of this integration was pinpointing, and holding in memory, moments when the clan had demonstrated its support for an early kabaka, or a kabaka from the distant past had visited the butaka. The people who lived in an area reenacted in successive generations the actions that linked them with the center of the kingdom.L Gorju observed, in 1920, that a detailed

4' Africana Collection, Makerere University, Apollo Kaggwa papers, AR KA 1, CA 22 "Mugwanya to Apolo and Kisingiri, Rubaga 24 Jan, 1906; Commission, 357-8, 448.

43 Commission, 357-8, 361, 425, 448.


reconstruction of the entire history of Buganda could be made from piecing together clan traditions of their own contributions to the kingdom.'

The Ganda monarchy evolved through the interaction of pre-existing and

immigrant clan structures.45 This can be glimpsed at in the testimony in 1924 concerning the loss of butaka by the Nvuma clan. Kyadondo was an ancient butaka of the Nvuma clan; as their mutaka explained, "Kabaka Kintu found us there." Before the kingship came to exist (probably in the 14th or 15th century), the Nvuma clan had had a corporate existence and identified Kyadondo as their center. Kyadondo was the name of the butaka, the name of the large area that became a sa (province) of the Ganda kingdom, and the name of the mutaka himself. Kyadondo was a kaQly (principal) butaka, which meant that it had the quality of a charter: it was the place where the clan began and which everyone in the clan, no matter where they were born, called their birthplace. As one witness explained, butaka were "the origin or beginning of the baganda...the place which is hereditary during the reigns of all the Basekabaka of Buganda, and it is the place where the ancestors or forefathers of each clan are buried."46 Together, the ka1y butaka of all the clans were a kind of unwritten constitution for the kingdom, they were its first source.

"Gorju, 85.
45This process is considered by Julien Gorju, Entre le Victoria. L'Albert et LEdouard. Rennes: Oberthur, 1920, 133-4; Lloyd Fallers, "Social Stratification in Traditional Buganda," in Lloyd Fallers, ed., TheKing'sMen, New York: Oxford University Press, 1964:64-113, 76-81; Kiwanuka, 115; D. Anthony Low, Buganda in Modern History, Los Angeles: University of California Press, 1971, 15; Wrigley, Kingship and state, 64-5; Ray, 94-6. Schoenbrun's historical linguistic examination of the question provides the greatest time-depth supported by evidence, ms. 347-350, 369, summarized on 396.

46Commission, 342, Malaki Musajakawa.


The people who lived in subsidiary Nvuma clan butaka in Kyadondo remembered not only their ancestors buried in the butaka but also stories that linked their clan with the order of the kingdom. The clan members continued to work for kabakas to commemorate the way one of their ancestors had begun to serve a kabaka in the past. The Siga butaka called Sekagya at Bumbu, for example, was remembered because Sekagya, who had been the katikiro to Kabaka Nakibinge (the 8th kabaka, who probably ruled around 1500), took care of Nakibinge's wife, Nanono, at that place. All the generations of successors to both Sekagya and Nanono were buried there.47 A clan elder explained that another butaka, called Buwambo "had been given to us by Kabaka Nakibinge, who planted a tree there for us to tie on his cow which we look after there and which is called Nakawombe." For the hundreds of years since that event, according to the Nvuma clan, its members had continued to look after the kabaka's cows in that place, and "the present Kabaka Daudi Chwa came to this place and saw this very tree and he also gave us his own cow to look after." One of the mituba butaka of the clan, Jita, "was very important for it was in this estate that the kabaka's beer was brewed, and where the kabaka's big calabash called Mendanvuma was kept." 4 Other clans framed their relationship with the kabaka in similar ways. The Ngabi clan, for example, had a butaka called Kipapi, where the king's buffalo were looked after.49

47Commission, 446-7, Semei Sebagala Kyadondo.

48They said, "this estate had never been cut off before." Ibid, 442.

49Commission, 475b, Danieri Sendikadiwa.


Baganda spoke about the tasks or the remembered actions of their clan with a

strong sense of identity and purposefulness. When a mutaka claimed in front of the Bataka Land Commission, "we have always been fishermen for the Namasole (Queen Mother) from time immemorial," he asserted the importance for the people of their work not as a means of livelihood, but as a way of defining who they were and how they fit into the kingdom.50 Remembered relationships between clan ancestors and ancient kabakas connected people in Buganda to the central authority of the kingdom in ways that were meaningful and effective. Whether the remembered events represent co-optation by increasingly powerful kings, or clans' strategies of integrating themselves into a useful rising power, or both, are questions for further historical enquiry.5' For whatever combination of motivations, over the long dure people in Buganda created a resilient polity using production dedicated to remembering important people and relationships.52

Following Tribute Up: Overlapping Forms of Power

People in Buganda used the logic of kuseng and the pattern of production oriented to remembering relationships to create the structures of their kingdom. By comparing the patterns of exchange visible in the late nineteenth century with the forms of authority named in Ganda epic tradition, it is possible to reconstruct the order of the kingdom. In the time of Kabaka Namugala, the 24th kabaka, who probably ruled in the mid-eighteenth century, a new type of chiefship--ekitongole--appears in the Ganda epic

5Commission, 428, Makobo Kalonde.
"I "Schoenbrun, ms. 333.

5Wrigley, Kingship and state, 228.


tradition. Before the beginning of ekitongole chiefship, Buganda was organized in a complex and effective system of chiefs serving the kabaka, chiefs serving the Namasole (Queen Mother), chiefs serving extremely powerful chiefs, and partially autonomous clan elders and religious leaders.

The role of tribute in expressing political relationships may have contributed to Buganda's stability over hundreds of years. Followers gave tribute to the particular leaders whom they served: Ganda structures of power were the connections between large groups of less powerful people who had obligations to particular powerful people in control of land. Since political relationships were expressed through the exchange of gifts, new forms of authority (and new obligations for tribute) could be introduced without displacing older ones. Had rulers been contesting authority over territo one would have won and the others would have lost; but because Ganda rulers were competing for followers and their tribute, the allegiance of a group of people could be divided among two or more rulers with nobody losing out entirely. 53 The densely complex and overlapping patterns of tribute obligations that existed in the late nineteenth century suggest that in the past Ganda rulers divided power--in the form of control over tributegivers--among leaders who otherwise might have come into conflict.

Attention to the flow of tribute from households through chiefs to various leaders of the kingdom suggests that power in Buganda was diffused, not centralized, and the structures of power were overlapping and complex, not hierarchical in a linear way. The

13 Another reason that Ganda chiefs did not fight each other over territory in the eighteenth and nineteenth centuries was that they obtained wealth by raiding.


resolution of conflict by dividing followers among contenders for power created intense competition among chiefs to attract followers. Interpretations of Ganda history derived from royalist sources and from observations made in the nineteenth century have depicted a centralized state in which kabakas gradually became more dominant over other internal forces. However, the actual relationships expressed in the exchange of tribute, allegiance, and protection reveal power diffused throughout the structures of the kingdom. "4

According to Ganda epic tradition, Buganda comprised ten divisions called saza each ruled by a chief with a specific title at the time Kabaka Namugala established the first ekitongole chiefship in about 1700. The kingdom had grown through the incorporation of clan leaders whose territories became sazas of the kingdom under the kabaka and through the appointment of chiefs to rule newly annexed territories. These gradual changes are recorded in Ganda epic tradition, which concludes the story of each kabaka by naming the important chiefs appointed during his reign. The titles for the chiefships of the oldest, most central sazas are identical to the names of the clan elders of the clans which had

54Original sources that reinforce this point of view are Kaggwa's Basekabaka, his
contribution to the "Enquiry, and the memorandum that became part of the Butaka Land Commission records which listed every victory of a Kabaka over a clan leader. The point of view that kabakas attempted to systematically take over the power of other controllers of territory can be found in Martin Southwold, Bureaucracy and Chiefship in Buganda, East African Studies No. 14, Kampala, 1961, and D. Anthony Low, Moder Hisory, 30; and Low, "The Northern Interior, 1840-1884," in History of East Africa, Vol. 1, Roland Oliver and Gervase Mathew, eds., Oxford: Clarendon Press, 1963, 334; Fallers, "Social Stratification," 97, and Wrigley, Kings 65. However, Ray cites an informant who remembered seeing Kaggwa refuse to record clan traditions which named Kabakas before Kintu, 101. Relying on oral histories recounted by clan elders, Michael Wright disputed the view that clans and kabakas had been in conflict or that Buganda had been despotic, Buganda in the Heroic Age, New York: Oxford University Press, 1971, 2-4, 206; More locations of power and other non-royalist perspectives were published in Ebifa and Munno by Gomotoka and others; these are explored by Kiwanuka, 99-100, Ray, 96-13.


significant butaka in that area. For example, the Mugema was the head of the Ngeye clan, and also the chief of Busirro; and the Kitunzi was the head of the Mpologoma clan, and also the chief of Gomba. The titles for the chiefships of some of the areas which Buganda had annexed from its neighbors were identical with the names of the newly annexed sazas: the Kasuju was the chief of Busuju; and the Katambala was the chief of Butambala. Following the Ganda pattern of marking political relationships with remembered histories, each saza chief had particular obligations of service to the kabaka, as is illustrated in Table

2. 1.55

"All of Buganda" attended gatherings in the courtyard of the kabaka, nineteenth

century visitors were told; the people present were chiefs of every saza, and followers who had come from each saza to work for the kabaka.56 The compound of every saza chief connected to the courtyard of the kabaka in the capital with a wide, straight road. This real and mental picture of Buganda as a collection of ten sazas ruled by saza chiefs was valid, but it was not complete. Each of the saza chiefs nominally allocated land and received tribute over the part of Buganda that was his saza, but other powerful figures had claims within their sazas, and several of the saza chiefs also had authority over land in

5" Ham Mukasa, Enquiry; Roscoe, 233.The names and services of some saza chiefs
suggests that these chiefships originated with something like the remembered relationship of an individual with the king called obwengeze land tenure by Morris Carter in 1909, and by A. B. Mukwaya, Land Tenure in Buganda: Present Day Tendencies, East Africn Studies no. 1, Kampala: East African Institute of Social Research, 1953, 12-3.

56Maps of the capitals of Kabakas Suna and Mutesa, who reigned through most of the nineteenth century, and descriptions of the gathering of "all of Buganda" by Mackay and other nineteenth century observers confirm the order delineated in the epic tradition.


Table 2.1
The Chiefs of Buganda's Ten Sazas and Their Special Functions Name Name of Saza Special Work for Kabaka

Kg Kyadondo Built houses for kabaka; cared for his twins

Mukwenda Singo

Sekibobo Kyagwe Supervised people who came to work for kabaka
Collected tribute from Busoga

Kangawo Bulemezi Built the house of the kabaka's most important
wife; took care of her

Mugema Busiro Head of Ngeye Clan; built and maintained
shrines of dead kabakas

PoEQkino Budu Collected tribute from Koki and Kiziba

Kitnz Gomba Head of Mpologoma clan; took care of shrines

Kaina Mawokota Provided guides when the kabaka went to war

Katamnbala Butambala Head of Ndiga clan; carried the kabaka's charms

Kasju Busuju Took care of the families of the princes and
princesses; an elder of Ngeye clan.


other chiefs' sazas. For example, the Queen Mother and her subordinates controlled land in most of the sazas." Saza chiefs also lost control of areas that became populated by the followers of a particular Lubaale (god), when the devotees established farms and remembered the Lubaale under the direction of a mandw (medium). According to Apolo Kaggwa's explanation in 1906, the people who had had authority over the land on which the mandwa and his followers settled could not object to their presence, because the medium would say "You'l die if you don't let them stay.""8 Parts of every saza were butaka lands, and these were under the control of clan elders. A small number of powerful chiefs in the center of the kingdom, including the Katikiro, the Kimbugwe, and the Sabaganzi (brother of the Namasole) controlled lands in every saza.59

The power to collect tribute, or the inability to do so, demonstrated the extent of a chiefs authority. A Saza chief mediated between the center of the kingdom and people in authority below him; all these subtle relationships were enacted in the collection of tribute.

"Gorju, 139-40; Lawrence D. Schiller, "The Royal Women of Buganda." Internationl Journal of African Historical Studies 23/3(1990):455-473, also Lukiko Record, 239, 29/1/1917 on a conflict over the appointment of a Nanasole's chiefship that was being abolished; Lukiko Record 169, 30/4/1915, on a Namasole of a deceased Kabaka complaining about losing control of her land; and Customary Law Reports 1941-1951, 115-118, on a dispute between descendants of the Mugema and a Princess over ownership of land,
Civil Case No. 262/50.

58Kaggwa, Enquiry. Schoenbrun suggests that kubandwa became associated with
particular places, and their mediums were able to command considerable labor, at some time before 1000 AD. ms, 207.

9 Gorju, 136-7.


For example, people remembered in 1904 that only some of the people giving tribute in the suza of Kyagwe in the nineteenth century gave it directly to the Sekibobo, the saza chief of Kyagwe. Others paid directly to his deputy, because they were the followers of the deputy, not followers of the saza chief. Other people residing in the saza gave tribute only to the Mutaka of a clan, and in order to obtain tribute destined for the king from those people, a chain of collectors, representing the king, the saza chief, and his deputy, would appeal to the Mutaka to collect tribute. Some tribute items, such as barkcloth, had to go to the kabaka, but the collectors could ask people to supply other goods in order to have something to keep themselves. When chains of collectors asked for tribute in a group, each authority figure represented kept a portion of what was collected. 6 These complex and overlapping chains of tribute suggest that new demands had been layered on top of older demands over a long period of time. Mackay observed in the 1 880s that some forms of tribute were collected by responsible people going from household to household whenever the king or queen mother chose to collect tax. Other forms of tribute, such as beer from every brew or the obligation to provide a bark cloth for the burial of a very important man, were routine and the tribute was brought without a specific request."' Receiving tribute through chains of authority figures entitled to take a portion of what was given was a well-established pattern in Buganda: early in the twentieth century, Catholics

60Kaggwa, Enquiry.

61Mackay, 196.


and Protestants built their cathedrals and schools by collecting from followers using this technique.62

Evidence from the nineteenth century indicates that saza chiefs did not exercise

authority over all the people who lived in their sazas; a person could live in the area of one chief, but be the follower of a different chief. A follower's allegiance to the chief who was nominally in control of a region depended on a consensus regarding that chiefs authority over the specific land the follower occupied. A person might be independent of the saza chiefs authority because he or she was a follower of the Kabaka directly, of a Lubale spirit, of the Queen Mother, or of a different chief who controlled some land inside that particular saza. For example, the blacksmith Erenesti Kakoza claimed that in the 1880s he had not been a follower of the Kimbugwe, even though he had been "under" (living in the saza and of lower status than) the Kimbugwe. He demonstrated his autonomy by claiming that no intermediaries came between the holders of his title and the kabaka: "All the kabaka's messengers sent to Kakoza used to come straight from the kabaka to myself, but did not come through the Kimbugwe; and when the Kabaka used to come to my workshop to [ask me] to do some blacksmith work he used to come straight to me not through the Kimbugwe; and he would not have paid such visits to a private tenant...63

More evidence of multiple forms of authority that co-existed and sometimes

overlapped comes from testimony about control of Bussi Island from about 1880 to 1900.

62Africana Collection, Makerere University, Kabali Papers, AR KA 2/2, Budo Board of Governors' File, 23/10/1924.

63Commission, 451, Erenesti Kakoza.


The island had been given to the Gabunga (the kabaka's admiral) by Kabaka Mwanga in 1884, but not all the land came under his authority.' A witness before the Butaka Land Commission explained, "When Gabunga was given power to rule the islands of Sesse he found some important bataka who had power over their own land and he did not take away that power from them. But [he] took possession of all the estates which belonged then to some less important bataka and converted them into his own private estates."65 A series of clan elders testified that even though the island had been given to Gabunga, they had never become his followers. They proved their independence from the saza chief by describing their actions. Zedi Zirimenya said, "When my father died we were about 150 men who attended the funeral, but we never went to Gabunga first to apply for permission to bury him."66 Malaki Musajakawa challenged the Katikiro (Prime Minister), "Let Chief Gabunga point to any one of us [bataka from Bussi] who was his private tenant and who worked for him."67 Before it was given to Gabunga, Bussi island had been territory controlled by Guggu, the priest who controlled the shrine of the god Mukasa. Witnesses discussing the issue of authority on Bussi pointed out that no nineteenth century kabaka

64As we shall see in Chapter Three, Kabaka Mwanga's own authority was called into
question (and, by some accounts, entirely rejected) during the years for which control over the island was disputed, and the kabaka's own dubious position might have contributed to Bussi leaders rejection of the overrule of the Gabunga. However, the logic of the Bussi witnesses, who proved their autonc -ay by referencing their actions in relation to the saza chief, suggests that autonomous authority within the territory of a saza chief was not unusual.

65Commission, Yosiya Sajabi Semugala, 385.

66Commission, 359, Zedi Zirimenya.

67Commission, 345, Malaki Musajakawa.


had ever tried to call up the canoes controlled by Guggu for service on the lake.68 Kabaka Mwanga gave Bussi to the chief Gabunga, but large groups of people who lived in his territory gave their allegiance, tribute, and labor to others and not to Gabunga.

These multiple, overlapping relationships of followers with figures of authority shaped the political character of Buganda. In carrying out their obligations as leaders, Ganda chiefs had to pay attention to their peers as well as their superiors. A chief had to be constantly alert to the possibility that his followers might desert him because they decided another chief in the same neighborhood treated his men better. Successful chiefs had to be able to attract and maintain followers in a context of competition from other chiefs seeking followers. Effective chiefs had to be skilled in resolving disputes in ways that seemed just to all parties; they had to be able to obtain and redistribute goods in ways that satisfied their superiors and their followers; they had to develop working relationships with other rulers so that their control of their people would not be threatened.69 By the time foreigners arrived in Buganda in the mid-nineteenth century, the coordinating, balancing characteristics of chiefship had been replaced by a more aggressive style required by participation in an escalating trade in ivory, slaves and guns; but the Ganda

68Commission, 386, Gabunga. The chief Gabunga claimed that withholding of Guggu's canoes from service proved that Guggu had not been a chief, because "they were considered to belong to the Gabunga personally as his private property," but it seems reasonable to assume that Guggu's canoes were not utilized because they were considered to belong to Mukasa.

69Roscoe, 241; Martin Southwold, "Leadership, Authority, and the Village
Community," TheKing'sM., Lloyd Fallers, ed., New York: Oxford University Press, 1964, 211-255, 214.


ethic that ruling implied courtesy and cooperation among peers was still intact (see Chapters 3 and 7).0

Power at the Center of the Kingdom

All the broad, well-maintained roads that crossed hills and bridged swamps in straight lines from each saza converged in the courtyard of the palace of the Kabaka. In Buganda, the kibujg~ (capital) was a physical representation of the kingdom as a whole. It is possible to perceive the nature of power relationships at the center of the kingdom in the customs observed in constructing a kibuga, and in maps drawn by Apolo Kaggwa of the layout of the capital in the time of Kabakas Suna (1825-1852) and Mutesa (1852-1879).1 One half of the huge circular kibujg~ was the palace of the kabaka, including large houses for his primary wives, each one built and maintained by a saza chief. Immediately in front of the palace was a large courtyard where "all of Buganda" gathered to greet the kabaka and listen to cases. Elaborate fences, built by people brought to work in turns by their chiefs, enclosed each layer of buildings in the palace. Facing the palace across from its courtyard were shrines to the gods Mukasa and Nende. The chief of each saza had a compound at the top of the road connecting the capital to his saza. The central chiefs of

70Richard Waller, "The Traditional Economy of Buganda," Master of Arts Essay,
University of London, School of Oriental and African Studies, 197 1. Wright observed that "low tension" characterized Ganda political interactions: differences were not pursued to the point at which conflict would become necessary, 5 1. This perception stands in contrast to that of Lloyd Fallers, articulated in "Despotism, Status Culture and Social Mobility in an African Kingdom," Comparative Studies in Society and History 2(1959):4-32, that the lack of clearly delineated functions of each of the multiple office holders would have increased the power of the kabaka, 20.

Peter C.W. Gutkind, The Royal Capital of Buganda, The Hague: Mouton, 1963, 918; Kiwanuka, Histo facing 160.


the kingdom, such as the Katikiro and Kimbugwe (guardian of the kabaka's metaphysical well-being), also had large compounds in the kibug .

Kabakas lived at the center of the kingdom, at the pinnacle of Buganda's pattern of exchange. In the 1880s, Baganda told the missionary Mackay that "the axis of the earth sticks visibly out through the roof of the conical hut of their king."72 In theory, everything was owed to the kabaka and he had everything to distribute: he "ate" the nation when he became kabaka. Replicating the kusenga relationship on a kingdom-wide level, the kabaka allocated land to his chiefs and expected obedience and tribute in return. In the nineteenth century, kabakas demonstrated the power of the monarchy through rituals devoted to deceased kings in which large numbers of people were killed.73 As Ganda epic tradition mentions few cases of large scale killings in earlier times, it seems probable that these events were part of the kabakas attempt to counteract the collapse of their power induced by the caravan trade.

Kabakas marked the center of the kingdom but they were not by any means the sole wielders of power within it. The shrines to Mukasa and Nende in the center of the kibug represented the independent voice of spirit mediums that rulers of Buganda were obliged to accommodate.74 A confrontation between Kabaka Suna. who ruled from 1824 to 1857, and a famous spirit medium, Kigemuzi, arose when Suna ordered people not to defecate in the capital on pain of death. It may be that Suna made this impossible demand

72Mackay, 214.

73Ray, 177-181.

74 Wrigley, Kingshi 182-7; Schoenbrun, 368, 371-2, 464.


as an attempt to reassert control at a time when trade with the coast had begun to erode the kabaka's power. According to tradition, Kigemuzi objected to the new law and sent a message to Suna through the tax collector, "Ask him, where does he defecate?" The horrified tax collectors took him to the palace, but Kigemuzi refused to be humble. When the kabaka's men stuck his lips with sticks to make him be quiet, he said "You also will be stuck"; when he was burnt with irons he said "You also will be burnt." According to the remembered tradition, only a few hours passed before Kabaka Suna was struck by lightening, and his capital burnt down. Kigemuzi, who had been held in stocks, was released and taken to the kabaka. He told Suna, "Punishing a child does not mean hatred, you will soon recover," after which the Namasole and Kabaka Suna made sure that Kigemuzi got everything he might want.75

Two leaders of Buganda wielded such a high degree of power that it was

impossible for them to reside in the same place as the kabaka: the Namasole (Queen Mother) and the Mugema (saza chief of Busiro). The court of the Namasole was on another hill, separated from the kibuga's hill by a stream of running water. The Namasole was served by a coterie of chiefs in all parts of the kingdom that mirrored the set of chiefs serving the king." Namasoles exercised a kind of superintending power over the actions of a reigning kabaka through their independent material base in land and people, and through their influence over their other sons, who were potential rivals for the kabakaship. If a

75Basekabaka, Makerere MS, 128.

76Schoenbrun observes that words for queen mothers and other royal women emerged in ancient east African speech communities at the same time as words for kings, some time between 900 and 1200 AD, Green Plac is., 345.


kabaka wanted to remain in power, he had to act in a way that pleased the Namasole.7 A Namasole never entered the kibug ; when she wished to communicate with a kabaka, she sent messengers.7 The Mugema--who served the deceased kings all of whose shrines were all in his saza Busiro (literally, the place of shrines)--also resided in a compound that was separated from the kabaka's by a stream of running water. As the "Katikiro" of the deceased kabakas, the Mugema could speak with a voice of authority that challenged the reigning kabaka. The Mugema was the only chief who stood instead of kneeling in the presence of the kabaka; he did not eat food prepared by the kabaka's cooks, and he was not obligated to provide people to maintain buildings inside the palace." Their locations as well as the ritual prohibitions regarding interaction between the kabaka and both the Namasole and Mugema expressed the Ganda awareness that these leaders wielded power that could seriously threaten a reigning kabaka.

Although royal ritual and proverbs emphasized the absolute power of the kabaka, the patterns of interaction in the kibuga suggest that the Namasole and Mugema were not the only figures of authority who had the capacity to challenge the kabaka. The Kimbugwe, the chief who was responsible for the king's "twin" (a powerful ritual object that represented the kabaka's metaphysical well-being), could suggest to a kabaka that

77For example, Kabaka Semakokiro's success in overthrowing his brother Kabaka Junju was engineered by the Namasole, angered by Junju's murder of a pregnant wife.Kiwanuka, Kings, 92: Kabaka Mutesa told Mackay he felt obliged to comply with the Namasole's wishes, Mackay, 162; Wright gives another example, 3.
78 Schiller, 461; Nakanyike B. Musisi, "Women, "Elite Polygyny,'and Buganda State Formation". Signs: Journal of Women in Culture and Society. 16/4(1991):757-86.

79Roscoe, 253; "Enquiry into Native Land Tenure."


specific actions were essential for his well-being, and the kabaka had to comply.8 The Sabaganzi (uncle of a kabaka) and Kasuju (saza chief in charge of princes) also had leverage over a kabaka because of their connection with the princes who were his potential rivals. As Rowe has pointed out, the Katikiro who assisted a prince to come to power on the death of his father often had considerable influence over a new king.8' The Kibari was a chiefship that required the holder to voice objections to unacceptable actions of the kabaka. The Kibari took the king's place when he was absent, and was the only person who could try the king. According to Zechariah Kisingiri, one of the most powerful chiefs of the early colonial period, the Kibari in the past "could find the king was in the wrong, but he had no authority to punish him.'82 Succession to this position followed a unique procedure: the Empeo clan selected fifteen candidates from the appropriate lineage, then the kabaka chose four of those, and the final decision of who would succeed to the position of Kibari was made by "all of Buganda"--the chiefs who gathered to hear cases in the kabaka's court.83

The gathering of "all of Buganda" had a larger role in the government of Buganda than has been recognized by studies that have taken royalist traditions and rituals at their face value. In these gatherings, and in the "endless amount of oiusang (cases) going on," various chiefs and coalitions of chiefs constantly worked out their relative positions of

8oRoscoe, 235.

81Rowe, xx.
82Kisingiri, Enquiry.



powerY. Decisions that went in favor of one chief or clan at one time were decided in favor of another party at a later date, when their relative strength--the love the kabaka felt for them--had shifted.85 As Waller observes, the role of the kabaka was to balance and coordinate the actions of chiefs: he could not rule without them. The chiefs who met in the gathering place in the center of the kingdom had the power to offer or withhold tribute and labor, and to choose the peers with whom they would align. A kabaka's capacity to secure the allegiance of chiefs depended on his ability to re-allocate chiefships or create new ones, but he had to constantly be aware of the power of groups of chiefs who might favor one of his brothers over him.6 When kabakas moved the kibiuga every few years, they were able to consolidate the allegiance of some chiefs and make others more remote and less powerful depending on where they placed the kibug in the kingdom, and how they re-arranged the order of the compounds of particular chiefs within it. The reign of each kabaka was remembered, in Ganda epic tradition, by enumerating all the important chiefs which that kabaka had appointed; the list of chiefs encapsulated how he had ruled.

8Many early observers of Buganda commented on procedures for trying cases, this
statement comes from Mackay, 187-8. Schoenbrun suggests that powerful men gathered at the court might have developed into the chiefships that were considered to be the specific servants of the king. ins, 338.

85This is one of the themes of the testimony of clan elders before the Butaka Land
Commission in 1924; see for example the testimony of Daudi Basude, Commission, 352.

86The story of the rise and fall of the servant Kiyanzi, which carried on for several generations, is an example of this. Kiwanuka, Ki=g 146-9.


Though ritual and proverbs celebrated the absolute authority of kabakas, in practice they were obliged to cultivate the cooperation of chiefs.87

Kabakas who violated the moral imperative of kusng faced rebellion. Kabaka Kagulu, the twentieth kabaka who probably ruled at the beginning of the eighteenth century, was remembered by people for making impossible demands. According to the epic tradition, he made people carry reeds for his fences with the points sticking up, and kneel down where he had planted needles in the ground. He also made them dig trees all the way out of the ground, including the roots, so that some people got buried alive in the process. The tradition explains, "When the chiefs and the rest of the people came to hate being pierced by needles or buried alive, they rebelled against Kagulu." People gathered on a hill adjacent to the kibuga and jeered the Kabaka, saying "Sir, we your men have come to pay you a visit, Busiro greets you." When Kabaka Kagulu saw that the people were refusing to come to him, he made a drum and ordered it to be beaten: "Buganda is at peace: Kagulu does not now kill people: come and visit him." But the people did not come, so Princess Ndege Nasolo called the princes, organized a battle, and eventually killed Kagulu herself. Kabaka Kagulu was not buried in Busiro, and after his death, people of Njovu clan were killed for having produced such an evil kabaka. 88 The epic tradition's

87 In his written version of Ganda epic tradition, Kaggwa explained that people
evaluated the success of a kabaka by observing how many beads had been left in his shrine by successors to the chiefs he had appointed. If there were many, people would say "He died a happy many because he had many chiefs." Basekabaka, ms in Makerere, 58.

88Kiwanuka, Ki=gs, 62-68.


story about Kabaka Kagulu demonstrates that power in Buganda did not simply flow from the top down.

Ebitongole: Kabakas Control Innovation

Ganda epic tradition records that in the eighteenth century, kabakas began to establish chiefships dedicated to specific purposes. In contrast to positions of authority that commemorated important people or significant interactions in the past, these chiefships were named for what they were supposed to accomplish. While saza chiefships had developed gradually by accretion over long periods of time, a newly appointed ekitongole chief displaced the previous authority on that land as soon as the chiefship was created. A chief explained how this had been done in the nineteenth century, "when the kabaka appointed you to a Kitongole you would choose an estate at which you would make your headquarters, and you would then distribute the rest of the estates among your Batongole."'89 Kabaka Namugala (who also made innovations in the ritual installing kabakas) and who probably ruled in the mid eighteenth century, established the first two ebitongole, Ekigalagala (for the purpose of spreading out) and Kitamanyang'amba (for the purpose of knowing what is said). A generation later, Kabaka Kamanya established the chiefship Ekikinakulya (for the purpose of things to eat).9 Consolidating military victories was the purpose of some of the original ebitongole chiefships, and for this reason, they have been erroneously considered a form of military chiefship. Their purposes were actually much broader.

"Commission, Paulo Bakunga, 536.

9"Commission, letter of Lukiko, 563.


Ebitongole allowed a kabaka to orient productive labor towards a task which he wished to have carried out. Settling an area to incorporate new territory into Buganda could be one purpose towards which the kabaka directed productive resources, but there were many others. An ekitongole on Buganga was called Ekibukula Mabira (for the purpose of opening up of the forests) because the clan elder had asked the kabaka for hunters to drive away elephants and buffaloes that were attacking people, and the ekitongole was the land supplied to meet the food needs of the hunters.91 Another ekitongole in Buganga was Ekirwanyamuli, which was a place where the people of Chief Omulwanyamuli could obtain bananas when they visited the lake to fish. The ekitongole had been allocated as uncultivated land, and the chief had been obliged to bring people to cultivate the land in order to use it. Other ebitongole in the same area were Kikwekwesi, which was for the head of all the kabaka's servants to obtain labor and supplies, and Kisomose, the place where the makers of drums and mweso boards for the kabaka lived, grew their food, and carried out their work.92 Another ekitongole was responsible for brewing the kabaka's beer.93

The creation of ebitongole allowed kabakas to control innovation by placing the production of new things under chiefs who were directly obligated to him. Semakokiro, in the generation following Namugala, developed the innovation of a new type of hunting

91Commission, Mikairi Kidza and Stanislaus Mugwanya, 401.


93Cornmission, 53 1b, Siriwani Mberenge.


net, which brought him many followers who helped him defeat his brother Kabaka Junju.9 Kabakas attempted to control the social consequences of new productive possibilities-new commodities or production for new purposes--by confining them to ebitongole. One ekitongole was named Kirima Ntungo (to cultivate sesame seeds).95 Kabaka Suna assigned an ekitongole to "the Banyoro potters".96 The sesame seeds and pots produced went directly to the kabaka, without any chiefs who might trade independently as intermediaries; and also the producers owed allegiance directly to the kabaka. As sources of potentially disruptive innovation increased in the early nineteenth century, the number of ebitongole chiefships also expanded.


Buganda has been called a highly centralized kingdom, but this is not quite accurate. Kuseng, the fundamental component of Ganda production and political association, linked people in a reciprocal relationship which was premised on the possibility that followers could leave their chiefs. A similar premise--that chiefs might withdraw their support--shaped the relationship of kabakas with their followers, the chiefs. Remembering ancestors motivated the activities of people who produced food and reproduced communities; a remembrance of significant ties infused meaning and purposefulness into the relationship of different units of the Ganda polity. In people's

94Kaggwa, Basekabaka, 93.

95Commission, 395, Stanislaus Mugwanya.

96Commission, 472, Yosiya Sensalire.


minds Buganda existed as a network of chiefships that offered tribute to the kabaka who "ate" the kingdom--the existence of a center was fundamental. However, power wasnt centralized; authority and the will that caused things to be accomplished did not flow down from the center, dominating every other participant. Chiefs acted in ways that indicated their sense of their own power relative to their peers: sometimes they withheld obedience and labor, they made alliances that pressured kabakas into compliance, and at times they rebelled. Compelling reasons to act or to refrain from action came from peoples' ways of remembering the past, and from the connections to that past which they re-enacted in but aka. A map of the roads in the kingdom looked something like a spider web, and power in Buganda had some of the same characteristics. All the lines led to the center, but each connection in the circle had strength and integrity of its own.

CHIEFS HIP, LAND, AND CIVIL ORDER Civil war and social disorder convulsed Buganda in the late nineteenth century. Four kabakas (kings) were installed in less than a decade; tens of thousands died from famine and disease; and the institutions of the polity appeared to fall apart. By the end of the war Baganda appeared to have temporarily ceded the kabaka's authority to allocate land--the ultimate demonstration of his authority--to British officers. The war has been perceived as a "religious revolution" in which modernizing Ganda Christians and Moslems toppled paganism and then fought each other to make Buganda Catholic, Protestant, or Moslem: contemporary Ganda chroniclers, missionary and colonial observers, and historians have interpreted the war in these sectarian terms.' New religious categories were only one dimension of the war in Buganda from 1888 to 1896, however. It was also, fundamentally, a Ganda expression of the collapse of social institutions that affected all of

'Kaggwa recounts the plundering undertaken by the kabaka and chiefs, and also the provocations of "unnatural vice," but identifies the onerous burden of digging the lake as the cause of the revolt against Mwanga, 96/141-100/143. The conflict is cast in religious categories by Wright, 34,40, 164-5; Kiwanuka, 192-3; John Milner Gray, "The Year of the Three Kings of Buganda," UJgand Jojjrn~ 14(1949): 15-52; Christopher C. Wrigley, "The Christian Revolution in Buganda," Comparative Studies in Society and History 2(1959:33-48; and D. Anthony Low, "Religion and Society in Buganda 1874-1900," E=I Africa tudie No. 8, Kampala, 1957. According to Twaddle, Kakunguliu, it was less a religious revolution than a palace coup, 35.



east Africa as a consequence of trade in ivory and slaves.2 In Buganda, the exchange of cloth and guns for people both undermined the legitimacy of the kabaka and transformed the autonomous power of chiefs. The faction leaders in the civil war were chiefs who had new religious convictions, and also new wealth from independent trading with Arabs, new power from followers who attached to them instead of to the king, and an expanded set of potential foreign allies.

The overthrow of Mwanga in 1888 initiated a period of self-destruction in which Ganda religious factions raided and slaved against each other inside their own country. The war caused unprecedented devastation because most Ganda mechanisms for ending conflict could not function: the four successive kabakas were ineffective; conversion by the chiefs had undermined the mediating role of Ganda spiritual leaders; and there was no model for cooperation among the new religions. In these difficult circumstances, Ganda chiefs used the language of land allocation to forge compromises among the warring groups. In 1889, 1892, and 1893, a re-arrangement of control of land by the various factions sealed attempts to end the war. The failure of the first of these efforts, along with the further collapse of the authority of the kabakaship under Mwanga after he was reinstalled, appears to have motivated the Ganda chiefs to give Captain Lugard the kabaka's

2 This analysis relies on Richard Wailer's unpublished Master of Arts essay from the School of Oriental and African Studies, "The Traditional Economy of Buganda," John Tosh's description of the Ganda contribution to pre-colonial trade in "The Northern Interlacustrine Region" in Richard Gray and David Birmingham, eds., Pe-Colonial African Trade: Essays on Trade in Central and Eastern Africa before 1900 London: Oxford University Press, 1970, 103-18; and Steven Feierman's analysis of the consequences of long distance trade for a kingdom with a tribute-based economy in TIhc Shambaa Kingdom: A History (Madison: University of Wisconsin Press, 1974).


role of making land allocations to resolve conflict. The assignment of Catholic, Protestant, and Moslem provinces by the Ganda chiefs with the support of various British substitutekabakas resolved the political and social turmoil of the preceding decade by effectively integrating the new and potentially dangerous religious categories into the structure of the Buganda kingdom.

The first three sections of this chapter consider the unfolding violence of the late

nineteenth century in East Africa: the destructive effects of the caravan trade on the power of the kabaka, the development of chiefs who wielded autonomous military power, and the integration of new forms, of spiritual power into the growing conflict. The fourth section demonstrates how Ganda chiefs used re-allocation of land to end the spiral of violence.

Buganda and the Trade in Ivory and Slaves

Trade goods which had come from the East African coast were first mentioned in Ganda epic tradition in the time of Kabaka Semakokiro in the late 1700s, and successive kabakas managed for half a century to incorporate these new things into the circulation of goods that expressed Ganda social hierarchies.' They did this by making specific chiefs responsible for trade in various markets on the edges of the kingdom, and by circumscribing new economic possibilities inside the structure of ebitongole chiefships.' At first, goods from the coast flowed exclusively to and from the Kabaka. In 1861, after

'Items traded from the coast have been found in archeological sites in Uganda dated at several centuries before Kabaka Semakookiro. David Schoenbrun, personal communication. It is possible that Ganda traditions associate trade with Kabaka Semakookiro because long-distance trading expeditions began to reach Buganda during his reign.

4 Kaggwa, Baeaa 103; Chapter 2.


Speke's visit, Kabaka Mutesa killed a chief--the Mutongole of Karema--for acquiring cloth in Karagwe which he had not turned over to the king.5 Foreign traders were met at the borders of the kingdom and escorted to the capital, and food was provided to them in order to prevent them from interacting with people or trading on their own.6 Through most of Kabaka Mutesa's reign, foreigners--both traders and missionaries--were the guests of the kabaka at the capital: they could only acquire food or labor when the kabaka supplied it, and were forced to offer their goods to the kabaka on his terms.' The absolute nature of the kabaka's control over foreign travel and trade was illustrated by Mwanga's killing of Bishop Hannington in 1885; he had aroused suspicion when he failed to enter from the correct direction and changed his travel route without informing the kabaka .8

Neither the authority of the kabaka nor Buganda's well-developed forms of hierarchical exchange could withstand the negative effects of the caravan trade which reached to Buganda in search of sources of ivory that had been depleted closer to the coast. Traders acquired slaves to carry ivory tusks to the coast, and the existence of a market for human beings transformed the practice of utilizing the labor of war captives into more aggressive forms of slave raiding. Ivory harvesting easily merged into slave raiding, as guns were the tools of both trades, and sources of ivory were quickly depleted.9

5Kaggwa, Basekabaka, 64/123.

6Mackay, 216-7; Waller, 22.

7Mackay by his sister, 216-7; Mackay journal quoted in Waller, 30.

gAshe, Chronicl, 72-73.

9Steven Feierman, "A Century of Ironies in East Africa (c. 1780-1890), in Curtin,
Philip, Steven Feierman, Leonard Thompson, Jan Vansina. African History: From Earliest


The argument that late 19th century enslavement was different in degree and character from earlier uses of war captives contradicts a tradition of scholarship on Buganda that views Ganda slavery as static or as a phenomenon that increased in scale without having serious social repercussions.'0 The view that accelerating enslavement is central to the Buganda civil war is based on recorded memories of the nature of enslavement, the documented increase in raiding and captives taken into Buganda, evidence of slave buyers' participation in Ganda war making, and the ways that the kabaka's loss of authority to his chiefs was connected to slave raiding.

Baganda remember the time when people began to be sold to the coast. Selling people for cloth was entirely different than other kinds of nonfreedom (such as pawning) that people experienced in their lives. According to Kiwanuka, Kabaka Mutesa was the first to allow the selling of people. The impact of an Arab selling cloth in 1868 were recorded by Apolo Kaggwa,

[Mutesa] found an Arab by name Wamisi had arrived at the capital of Nakawa, bringing with him a lot of cloths and many other things. The
Kabaka distributed cloths to princesses and ladies .... [ to pages and specific
chiefs] ... later on he gave cloths to all chiefs and ordered them to buy.
Many people, boys and girls were sold to the Arabs in exchange for
cloths. "

Times to Independence 2nd ed., New York: Longman, 1995:352-376, 354.

'OFallers, "Social Stratification," 111-2; Christopher C Wrigley, "The Changing Economic Structure of Buganda," in Lloyd Fallers, ed., Th.King'sM New York: Oxford University Press, 1964:16-63, 19, 21, 25; Rowe; Twaddle, 59. Kiwanuka argues that there was no slave trade until after 1860, 167.

"Kiwanuka, 167; Kaggwa, Basekabaka, 66/124.


In 1883 OFlaherty reported that he had had a conversation with Kabaka Mutesa about the effects of trading slaves to the Coast. Mutesa said that two years earlier he had been trading ivory, but "such a thirst for cloth has caught hold of [the Ganda] that they will sell men and women for guns, powder, and shot, cloth, soap, etc." According to OFlaherty, Mutesa regretted the trade, but felt he could not prevent it.12 A son of one of the first Christian chiefs described the late 19th century as a time when "a piece of soap could buy a man, and a measure of bafuta [cotton cloth] could buy many slaves."'3

One indication that raiding for slaves to sell to Arab traders was changing the nature of Buganda's wars is the intensification of conflict during the 19th century.'4 Kaggwa's history describes not only more conflict, but also more conflicts resulting in the death of the leaders of the expeditions. He names "a lot of slaves" as well as women and cattle, as booty from battles in this period.'5 The huge increase in the number of royal wives also suggests that the nature of warfare was changing.16 Mackay wrote in 1881 "One army has been sent east to murder and plunder. Not even the natives themselves can call it war, they all say it is for robbery and devastation."7 He wrote to the Times in

120Flaherty, CMS Archives, quoted in Waller, 31.

'3E.M.K. Mulira, Sir Apolo Kaggwa, CKCMG, MBE Kampala: Buganda Bookshop, 1949, seen courtesy John Rowe.

14Waller, 31; Kiwanuka states that the power of the monarchy increased, 108; Rowe, Twaddle says patterns of plundering were formalized, 13.

15Kaggwa, Basekabaka, 63/122.

16Musisi, "Elite Polygyny", Sgm.

17Mackay by his sister, 185 (check page).


January of 1889 that Buganda and Bunyoro "have generally large armies in the field, in one direction or another, devastating whole regions of their inhabitants." Kabaka Mutesa attacked estates which had been protected from raiding 'from time immemorial'; these included the estates of Lubale, and also estates of the Namasole. These violations of Ganda morality may have been Mutesa's test of the power of Lubale, as Rowe suggests, but it is also possible that the estates became vulnerable as social disorder and the need to supply traders increased."8

According to Mackay, Arabs supplied the guns and powder for the plundering expeditions, and then received "women, children, and ivory" procured in the raids as payment.'9 Traders sent agents into the field with the armies to select the slaves they wanted.20 Kiwanuka observes that Ganda military success declined after 1880; this is perhaps because Buganda's neighbors were also participating in the exchange of cloth and guns for ivory and slaves. Richard Waller outlines the increasing importance of guns in Buganda: in 1875 there had been approximately 500; in 1882, Felkin complained "Mutesa's cry is always guns and gunpowder"; he calculated that there were 2,000 guns in the country, and guns and powder had completely replaced all other trade items. Waller notes that in the 1880s Mackay described traders bringing nothing but guns, and OFlaherty reported the arrival of a trader with 600 rifles.2'

18 For example, his raid on Batombogwe hill; Kaggwa, 82/133, also 65/123; Rowe.

'9Mackay, 435.

20Waller, 32.

2Waller, 29.


Kabaka Mutesa brought the caravan trade closer to his court in order to maintain his supervision of the distribution of goods, but eventually was overcome by forces inherent in the trade that he could not control. Waller identifies three stages in Buganda's external trade: a first stage in which royal agents traded on behalf of the kabaka in Karagwe; a second stage after the death of Kabaka Suna when trade shifted to the Ukerewe Islands and Kabaka Mutesa controlled access to Buganda by controlling canoe transport of traded goods across the Victoria Nyanza; and a third stage, in the 1880s, when the focus of trade shifted to Mutesa's court.22 Waller argues that the kabaka had used the distribution of prestige goods, such as guns and slaves, to enhance his power and his followers' obligation to him: the huge increase in both trade goods and plunder unbalanced the system and the kabaka's place in the center of it.23 Since the mutual obligations and relationships in Buganda society were expressed in the exchange of tribute and gifts, it makes sense that social relationships were fundamentally disrupted by massive increases in the goods being exchanged. The collapse of authority that characterized late nineteenth century Buganda was not merely a result of an enlarged market: it was also a consequence of the nature of the trade. The possibility of gaining wealth and power by selling slaves introduced a new kind of violence into the relationship of the kabaka to his chiefs, and of chiefs to their people. This is evident in the increasing autonomy of the chiefs, and in Kabaka Mwanga's ultimately ineffective attempts to re-assert control over them.

22Waller, 28.

23Waller, 32.


The Dissolution of Authority

A new kind of authority figure emerged in East African societies with the expansion of trade in ivory and in slave-taking and slave-holding. "Rugarug were followers of a powerful big man who broke the rules of social interaction and exerted power over others through military force.24 The way people lived in their environment changed in response: they grouped themselves into large defensive settlements, behind walls of stone and spiny cactus, whose ruined outlines are still sometimes visible in the rural landscape. In the well-developed bureaucracy of Buganda, the destabilizing potential of a new kind of trade manifested itself in changes in the action and role of chiefs, and in the total deterioration of the authority of the king, which began under Kabaka Mutesa and reached its culmination with the overthrow of Kabaka Mwanga in 1888. Most histories of the period explain the collapse of the kabakaship in terms of Mwanga's personal qualities: his youth and insecurity; his excessive attachment to pages who were his lovers and his consequent inability to value the advice of senior chiefs; his pagan small-mindedness and fear of the followers of new religions.25 While it is true that Mwanga did not lead Buganda effectively when he assumed the kabakaship in 1884, it is possible that elements of irrationality that were inherent in circumstances in the 1880s have been attributed to

24Iliffe, A Modern History of Tanganyika, 75.

25 Gorju, 120; Gray, 15; Wright, 28; Kiwanuka, 194. I see the growing power of chiefs as a descent into chaos fuelled by slaving, and not as a potentially competent emerging bureaucracy, cf. Low, Buganda and British Overrule, 4; and "The Northern Interior," 334; Fallers, "Social Stratification," 111; Wrigley "Changing Economic Structure," 25-6, Twaddle
Kakungulu, 38, 59.


Mwanga's personality. This was a time when things were turned upside down, when young men obtained power they did not deserve, and strong government from the center of the kingdom became impossible. Ganda chiefs did not manifest the inversion of all socially appropriate behaviour that characterized the rugarug, but they amassed and deployed wealth and force in ways that were fundamentally destructive.

Chiefs of the border sazas who came into unsupervised contact with traders were the first ones able to trade on their own account: the Pokino and Kago irritated Kabaka Mutesa by selling ivory and obtaining cloth without his permission and this may have contributed to the redirection of trade first to Ukerewe and then to Rubaga.

The contribution of trade to the growing power of chiefs is most obvious in the

position of the Katikiro. The Katikirro's responsibility for overseeing receipt of tribute and distribution of the kabaka's wealth made it a chiefship that brought wealth to the holder.26 Describing how wealth had derived from control of land in an earlier time, Stanislaus Mugwanya explained in 1906 that "The Namasole or king's mother, got estates, and originally was a person of more consideration and honour than the Katikiro."27 Mackay's journal gives an indication of the control the Katikirro exerted over trade: for five months Katikirro Mukasa had blocked Mutesa's orders that Arab traders be supplied with canoes, and Mackay told Mutesa that Katikirro Mukasa "was practically causing rebellion in the

26Twaddle associates the katikiros' wealth with the political influence of 19th century katikiros in issues of succession, 34.

27Enquiry, Rhodes House Af S 17.


country".28 He took on two of the most important chiefly titles as well as being Katikiro-the Sekibobo of Kyaggwe and the Pokino of Buddu. Both of these provinces were critical to long distance trade.29 The power of this Prime Minister over Kabaka Mutesa was formidable: in 1881 the Kabaka made blood brotherhood with Mukasa, and directed that Mukasa's sons should be carried like princes.3o Possibly the Katikiro was taking advantage of Mutesa's vulnerability because of his incurable gonorrhea, as Rowe suggests, but the economic dimension of his growing power cannot be discounted, either. Ashe noted in 1888 that Mukasa had an important role in directing the ivory trade, and foreign visitors commented on the wealth and impressive character of the Katikiro." Kiwanuka records that Mukasa had the reputation of a man who sold his own relatives into slavery.32

The ekitongole type of chiefship was also transformed by long distance trade. As we have seen, Kabakas Suna and Mutesa increased ebitongole chieftainships in order to channel new economic activities and to contain their effects. (Chapter 2). When the possibilities for trading became more available, and as guns became more significant in raiding, an ekitongole chief who had guns because the purpose of his chiefship was hunting or defense could become wealthy independent of the kabaka by raiding and

21Mackay, 150.

29Kiwanuka, 208.

3Kaggwa, Basekabaka, 85/135.
31Ashe, Chronicles, 116.

32 M.S.M. Kiwanuka in Kaggwa, Kin 184.


disposing of slaves on his own. Kaggwa explained the remarkable wealth of batongole chiefs in terms of their success in war:

Their areas carried great honour and people used to flock to them and they
were therefore well cultivated...When the Kabaka was at war, people in
such areas (Bitongole) excelled in capturing the booty for they were always
young men. From what they had captured their chief (Mutongole) would choose the best and consequently became a rich man. Such a chief would also act as the Kabaka's messenger and thus again become rich for he was
given presents.33

The presents given by kabakas to ebitongole chiefs may suggest that Kabakas Mutesa and Mwanga recognized the possibility that these chiefs might act outside of their control and attempted to maintain their allegiance.

Michael Twaddle's richly detailed biography of Semei Kakungulu documents the potential independence of an ekitongole chief. Kakungulu obtained an ekitongole for elephant hunting from Kabaka Mutesa in 1884. Kakungulu had arrived in Buganda with experience of elephant hunting, and Mutesa gave him "guns, gun caps, and bullets", and land in Buddu. This land had been attached to a different chieftaincy, but was reallocated to the new ekitongole, which was called "Ekirumba njovu"--'for hunting elephants', and Kakungulu's title was "omulumba njovu"--' hunter of elephants'. 34 Although nominally under the control of Kabaka Mutesa and required to turn over all the ivory he acquired, Kakungulu's control of one hundred guns enabled him to build up an independent following through predation. According to Paulo Kibi, at this time Kakungulu had a drumbeat:

33Kaggwa, Basekabak, 340/277.

34 Twaddle, Kakungulu, 37.


I eat what I choose:
I eat what I find:
I eat whatever does not belong to me.35

Kakungulu and his men raided Nkore on their own, without instructions from Mutesa, and disposed of the cattle and slaves they obtained in Kiziba. Even when Kakungulu and his men participated in a raid against Bunyoro initiated by the Kabaka, Kakungulu's men got in trouble for looting inside Buganda. That his activities went beyond the pale of appropriate behaviour for subordinate chiefs is evident in the story that Katikiro Mukasa either placed Kakungulu in stocks or threatened him with death; he was only saved by the intervention of his blood brother the Pokino or, in another version of the story, by the Kabaka.36

The power that men like Kakungulu created for themselves in the tumultuous circumstances of late nineteenth century East Africa was mercurial. As the chief of an elephant hunting ekitongole, located on a route along which guns were being brought into Buganda, Kakungulu raided people and cattle without passing them on to the Kabaka, and collected followers of his own.37 Kakungulu was able to attract followers by offering to arm them, and also by trading ivory for enslaved people. He was not, however, able to maintain the following he created for himself. Kakungulu lost his chiefship, the Ekitongole of Ekirumba Njovu, when Mwanga was deposed, demonstrating that in Buganda, the arrangement of the kingdom allowed the center to exert a degree of control over the

Twaddle 21.

36Twaddle, 22-3.

17Twaddle, 37.


destabilizing force of men with guns. As soon as Kakungulu lost the chiefship, seventy of his followers deserted him for the new Katikiro, Honorat Nyonyintono. These men had "belonged" to Kakungulu, but they chose to align themselves with the strongest leader available.3" Kakungulu's experience suggests the similarity of late nineteenth century ekitongole in Buganda and nugauga south of the Nyanza: in both situations, big men controlled unfree people who had guns.

The Kabaka's control over raiding deteriorated markedly as the amount of military hardware in the nation increased in the 1880s. Ashe reported that the escort taking him to the capital made an "impromptu slave raid" during the journey.39 In 1862, members of the party escorting James Grant to Buganda had been punished for raiding without permission. Waller also points out that Pearson estimated that 75% of the slaves taken in a raid were not reported to the Kabaka, "the rest having been secretly disposed of by the chiefs." 4' By the 1880s the Kabaka received only ivory, and chiefs retained cattle, women, and slaves.42 This represents a diminution in the Kabaka's share, and may have been a recognition that chiefs would retain slaves and cattle on their own whether or not the Kabaka gave them permission.

3Twaddle, 28.

39Waller, 32.

4James Augustus Grant, A Walk Across Africa. or Domestic Sceneries from my Nile Jounm (Edinburgh: Blackwood and Sons), 191.
41Waller, 32.

42Kaggwa, Mpisa, 157-160, quoted in Twaddle, 14.


The increasing social instability was expressed in allocations of land: Mwanga took large areas from saza chiefs between 1886 and 1888 in order to create four new Bitongole, which he placed under the control of young men. These were the Ekitongole Ekiwuliriza, the chiefship of listening carefully; the Ekitongole Ekigwanika--chiefship of wealth; the Ekitongole Ekijasi--the chiefship of guns; and the Ekitongole Ekiyinda--the chiefship of menacing noise.43 Not only did Mwanga take land that had been under the control of saza chiefs to make the new ebitongole chiefships, he told the new batongole to establish their chiefships in every saza, presumably by force. According to Fallers, these bitongole represented Mwanga's attempt to remove power from the saza chiefs and give it to young chiefs he could control more easily."44 However,in the highly disordered condition of Buganda in the 1880s, it is difficult to assert that Mwanga was actually creating new chiefships in order to advance the structure of the state.45 A more accurate assessment might be that in creating huge new ebitongole, Mwanga was merely naming as chiefs new holders of power who had emerged from circumstances of the violent exchange of ivory and people for guns, and attempting to claim power over them.46 As Kiwanuka points out, Buganda was in such turmoil at the time that chiefly authority over land was not readily discernible. Kiwanuka claims that people deserted other chiefs to become the followers of

43Kiwanuka, History, 198-9; Kaggwa, Basekabaka, 99/143.

"Fallers, 64-116.

45See references for footnote 24.

46According to Twaddle, the purpose of these new ekitongole chieftaincies was to guard Mwanga, but the information available about them suggests they were acting on their own, in their own interest. Twaddle, 59.


the batongole, and "by 1888 the four new Bitongole had nearly 100,000 men, all young and arrogant."47

The new batongole proceeded to plunder all over the country. Apolo Kaggwa's account of this period describes the Kabaka's lack of control:

the morals of the country became deteriorated as we young men adapted a bad habit of robbing people of their cattle and goats at random; and people
found on the way were killed for no just cause. The Kabaka knew of this
and he did not care for the well-being of his country at all. He liked the
young men more than his chiefs.48

Kiwanuka states that the batongole and their followers "became the rulers of the country"; they raided and took captives without any inhibitions.49

Kabaka Mwanga himself took part in the process of raiding and enslaving

Baganda. In 1888 the Kabaka had made a tour of the country. A royal journey to "show the kabaka" and to receive tribute was not an unusual thing, but this tour proved "nearly as disastrous to his unhappy subjects as a foreign invasion, since he ruthlessly robbed and raided his own people."5 In Kyagwe, Singo, and Buddu, he raided hundreds of cattle, and seized "vast numbers" of women and children. On his return to the capital, he distributed these as gifts to his pages. The right of a kabaka to sacrifice the lives of people in a "kiwendo" for the spiritual well-being of the nation was accepted; Mwanga's use of

47Kiwanuka, 199.

48Kaggwa, sekabak 98/142.

49Kiwanuka, 199.

5OAshe, Chronicle, 90; also James. S. Miti, A History of Buganda, n.d.; a manuscript translation in Makerere University Library Africana Collection, (N.B. pagination of this document is unreliable), 252-297.


captured subjects to enrich the members of his palace household, who had not even been introduced in the palace in the appropriate manner by chiefs, was not acceptable.

Mwanga's final, unsuccessful attempt to consolidate his authority was his demand that his people dig a large artificial lake in the capital: requiring unnecessary work was a way of making people demonstrate their allegiance that had been deployed by other kabakas.5" Everyone, of every status, was required to participate in this public work or face heavy fines. Ashe reports that

The chiefs came with extreme reluctance, many of them smarting from the loss of their wives and other valuable property extorted from them during
the King's progress.52

A royal drum was beaten calling people to work on the lake before dawn. Kaggwa wrote that anyone who did not arrive early in the morning was fined one woman and one head of cattle; Ashe reported that insufficient service at the lake resulted in fines of large numbers of women, expensive cloths, and guns; Zimbe wrote that they were fined, "women, slaves, livestock, and loads of barkcloth"; and that the treasurer's house "became a huge prison camp overflowing with alleged defaulters".3 Baganda remember not only the unreasonable fines, but also the horrifying humiliation forced on important, old chiefs who were made to sit in the mud if they arrived late.4 Mwanga's bizarre behavior in the last months before

5'For example, Kabaka Mutesa had required the Kaima to build a hill inside his palace in 1871; Kaggwa, 78/139.
52Ashe, Chronicle, 90.

53Kaggwa Basekabaka 100/143; Ashe C nil, 93, Zimbe, Buganda Ne Kabaka, 133-4, quoted in Kiwanuka, 200.
54Kaggwa, Basekabaka, 100/143.


he was overthrown can be understood as an attempt to demonstrate authority over his subjects which he had already lost, and also, as a means to obtain slaves by fining his subjects in people and in creating situations in which they would avoid humiliation by offering bribes.

Kabaka Mwanga was deposed because his chiefs withdrew the will to be governed by him: the firsthand accounts of the events in 1888 are reminiscent of the history of Kabaka Kagulu, a century and a half earlier, whose reign ended when his chiefs retreated to a hill overlooking the palace and jeered.55 This moment came for Mwanga when the readers refused to embark in canoes for a journey on the lake they suspected would lead to their deaths: Mwanga's Katikiro told him "All Buganda refuses to take you to Sesse.56 Mwanga, like Kagulu, was overthrown when people became fed up with entirely unreasonable demands. In Mwanga's case, we can recognize that profound social changes contributed to the Kabaka's unreasonable actions and unworkable relationship with his chiefs.

Buganda's Civil War: Social Violence with Religious Categories

The fall of Mwanga was one moment in an unfolding crisis of authority in Buganda which was much larger than a palace coup. Fundamental terms--of how to be a chief, how to express authority, and how and why to be productive--had been altered by the possibilities and also the violence of long distance trade. This moment of political, social, and economic turmoil also contained a crisis in ideology, because Arab traders and

55Kaggwa, B, 65, also Chapter 2.

56Ashe, Chroaic, 102.


European explorers had introduced new ways of thinking about the world in the form of Islam and Christianity. The conversion of large numbers of Baganda to these faiths in the nineteenth century was so unique, so attractive to observers, and so clearly genuine that it has tended to overshadow other aspects of the processes of change underway at the time. Without denying the significance of conversion for individuals and for their community, it is important to keep in mind that the people who became Moslems and Christians did not stop being Ganda. The late nineteenth century was an encompassingly difficult time, intellectually as well as on every other level. People expected the kabaka to be powerful and to be in the center of things, but no kabaka filled that expectation from Mutesa's reign onwards. World religions became a principle for organizing relationships at a time when other means of organizing them were not functioning effectively.

The new religions gave young men powerful spiritual resources. Reading and prayer gave them access to spiritual power without the mediation of their elders. The world views of Christianity and Islam offered comfort and security lacking in a difficult time. In the new religions, positions of spiritual leadership were open to young converts, who had been pages at the lowest level of the Ganda chiefly hierarchy.

Ganda Moslems, Protestants, and Catholics used their new sets of ideas to create social institutions that did the kinds of things that organizations of people had always done in Buganda. That is, a religion was not only a spiritually effective practice and a form of identity, but also a way of organizing economic activity and an instrument for wielding political power. "English religion" (Protestantism), "French religion" (Catholicism), and

57 See footnote 1.


Islam functioned like clans or important chiefships: they brought people together under well-respected leaders for political and economic actions as well as spiritual ones. In the highly unstable context of the late nineteenth century, the new religious categories gave Ganda readers a means to re-group the authority that had been dissipated by the actions of the kabaka and by some chiefs. As Michael Wright points out, Ganda categories of clan and family continued to be salient, and throughout the war people defended and protected family and clan relatives of different faiths. 58

Religions became alternative categories in which people could continue to make the social arrangements they had always made.59 It has been said that the Baganda were fighting each other fQr religion, but this ignores the larger East African context and the instabilities that would have led to armed conflict whether or not the protagonists had adopted new religions. A more accurate perception might be that they were fighting each other with religion, since the logic of the organization of the groups, their sources of supply, and their maneuvers to gain political power were all connected to their sets of beliefs. Fighting with religion made ending the war particularly difficult, because the new religions were bereft of the conflict- resolving role that had been played by Ganda spirit mediums, and because both the ideological and economic dimensions of the new religious communities facilitated a prolongation of hostilities.

"8Wright, 114-5.

"9According to Bakale Mukasa "They did not fight for religion but for chieftainship", quoted in Twaddle, 40-4 1.


Baganda followed the new religions in the ways that they had followed spiritual leaders in the past. Allegiance to a leader was an element of allegiance to a religion; even the Christian missionaries accepted the role of nurturing "their" Christians. When the Katikirro Honarat Nyonyintono was killed in a battle, his followers found it impossible to continue fighting, "they were not cowardly but were distraught and did not see why they should fight just for Protestants. "60 The connection of personal and religious allegiance meant that it was logical for people to switch religions in order to gain chiefships: Simioni Segutta had been a Catholic in 1886, but became a Protestant when offered the position of Kiryagonja, when he did not get the chiefship he had wanted as a Catholic. Yosefu Sebowa was promised a chiefship if he converted to Catholicism, which he did, and became Kisalosalo.6'

The new religious communities became arenas for competition over status in the same way that Ganda chiefs had competed with each other over relative status in other circumstances. The individual chosen to lead Christian groups in any given engagement had authority over the division of spoils. Before one battle in 1890, a messenger had to be sent back to Mwanga to enquire whether it was acceptable for a Protestant, Kakungulu, to take over the leadership of a campaign whose original Catholic leader was indisposed.62 The group of Catholics and Protestants who retreated to Ankole clashed with the Christian group that had retreated to the lake over the issue of seniority and control of

6Wamala, quoted in Twaddle, 44.

61Wright, 116.

62Twaddle 51.


spoils: these groups were known as the "grain-eaters" and the "fish-eaters" .63 Entirely new forms of authority fostered further conflicts over relative status. For example, a conflict arose between the Pokino and the Katikirro because the Pokino was lower than the Katikirro in the chiefly hierarchy but higher than him in the church council hierarchy: he did not want to take orders from someone who had a lower position than his in the church council.'

The new religious factions controlled the organization of production and of trade. In the past, the followers of a Lubaale occupied land associated with that spirit; the Mandwa was given sufficient land for his or her followers, however large that group became (Chapter 2). Catholics held a monopoly on canoes at times during the war, so that lack of access to lake transport was a problem for Protestants, and one of the great weaknesses of Kabaka Kalema.65 The Catholic and Protestant coalition suffered because they needed food, and essential supplies of food were controlled by Ganda chiefs who were not readers in Kyagwe and Bulemezi.6 Each faction had sources of supplies from the Coast. Moslems got their guns through Arabs, and Christians got their guns through the former missionary Stokes. Co-religionists who were not Baganda participated actively in the war through their efforts to provide supplies. Miti states that Kipanda, an Arab trader

63Twaddle, 5 1.

64Ashe, Chrnicle, 141.

65Ashe, Chronicle, 41 ;Twaddle, 47.

66Hamu Mukasa, Simuda Nyuma, 383, quoted in Twaddle 47.


at Magu at the south end of the lake, sent a dhow of guns and ammunition which he paid for himself, and told his people to attack and sink Stokes' boat if they found it."7

Ganda Christians and Moslems created organizations that had religious, political, and economic dimensions at a time when raiding and plunder had become one of the main occupations of groups of people. The destruction of the civil war, and the terrible calamity of people slaving inside their own society, have been underemphasized by historians who have focussed on the religious identity of the combatants, and described the war as a conflict between new and old ideas .68 The accounts of the war written by participants, and also the statements of non-Ganda observers, describe an effort to overthrow an unsatisfactory king that spun out of control in the volatile conditions in which young men with guns had power.

In 1888, Moslem, Catholic, and Protestant leaders had made blood brotherhood with each other before beginning the battle which caused Mwanga to flee: they were making an effort to overcome the potential conflict inherent in their different religious identities.69 Once they had installed Kiwewa as Kabaka, they assigned chiefships in a way that attempted to divide positions of high status between Moslems and Christians. This arrangement quickly dissolved in conflicts over which religious group ought to hold which

67Miti, 349.
68 Part of the challenge of the war for Ganda Christians and Moslems was to find ways to integrate new and powerful ideas into their organization of Ganda society, but to frame the conflict in a tradition vs. modernity dichotomy ignores the reality of fundamental change in the region in the nineteenth century.

69Kiwanuka, 205.

chiefships. Not long after a fight which caused many Christians to withdraw to Ankole, Kiwewa lost control of Mengo in conflict with Muslim chiefs, and he was replaced by Kalema, a son of Mutesa who had been considered by everyone a better candidate for Kabaka than Kiwewa. During the brief reign of Kalema (1888- October 1889), conflict between the factions escalated from raids on Kyaggwe cattle by Ganda Christians in Ankole to violence and plundering that led to the depopulation of Buganda.7

The civil war protagonists were the same people who had been involved in raiding and plundering outside of Buganda and sometimes inside it: the neglected role of elephant hunters was identified by Hamuli Suku, who remembered that the Moslem defeat was a result of the joint action of "all of them, the pagans, the readers, and the hunters'"7' Kaggwa acknowledged that in an engagement he led in Mawokota "the Mohamedans were defeated and their wives plundered", but he states that the wives were later returned.72 The Protestant missionary Ashe, who returned to Buganda during the war, wrote that probably not all the women had been returned after that engagement, and that loot was the main objective of the combatants." One of the Christian combatants later explained that in re-taking Mengo, the Christian army failed to capture Kalema because people stopped to plunder. "What saved Kalema was our poverty. Just when our victory

70Kaggwa, B skabak 110/149.
71Hamuli Suku oral testimony 1969, English translation at Department of Religious Studies, Makerere, by Abdul Kasozi, quoted in Twaddle, 58.

72Kaggwa, B, 116/153.

73Ashe, Chronic, 137, 139.


was almost complete, everybody went to the place of the coastal traders in order to plunder the cloth.",71

James Miti, who followed his participation in the war with a distinguished career in the Uganda Protectorate government, stated clearly that enslavement was a goal of making war, to "plunder and carry off men and women from the vanquished side on every occasion was the order of the day at that time" and "it was each warrior's ambition to fight hard in order to be able to return home with plunder and captives".'5 He acknowledged that he himself had taken "not less than seventeen female captives and some six male prisoners of war" in the attack on the Buvuma Islands in which Major MacDonald had participated and forbidden any enslavement, and suggested that many hundreds of captives had been smuggled away into Buganda by other Ganda warriors.

Miti also described an incident in the war which indicates that people were not only being enslaved, but that many of the captured people were being sold to Arabs. Kabarega, the king of Bunyoro, had sent an army to assist Kalema in 1890. This army got confused in retreating from a battle, and accidentally went further into Buganda, where "they fell into a trap and many of them were captured and made prisoners or slaves." Miti describes how the captured people attempted to prove their value to the Ganda captors, pleading that they should remain with the person who captured them:

A Munyoro potter or blacksmith would plead his case by assuring his
Baganda captors that his knowledge of pottery or of the manufacture of

14 Paulo Kibi testimony, quoted in Twaddle, Kakuigiulu, 55.

75 Miti, 409.


spears, as the case might be, would be found very useful if he were only
kept under their service.6

After the Moslems had been driven from Mengo in 1889, according to Miti, they began plundering food and property all over Ssingo, and into Ggomba and Busujju.

Women, children and even old men fell victims to the Mohamedan's acts of
cruelty, some of them being killed on the spot, others being carried away
for sale to his Arab friend as slaves.77

When Christian chiefs organized themselves to stop the raiding, they engaged in a battle in Lumanyo, when they surprised the Moslems and routed them so that the retreating army dropped their plunder. Captured women and children were abandoned along the road. The Christian army returned them to their families, and the children whose families could not be found (because the children were too young to identify themselves) were adopted.78 A praise song devised for Kakungulu during the war "Kangabaana, eyawangula abensambya"

--"the scatterer of children, the one who conquered those of Nsambya" suggests the social consequences of the war.79 That children and their mothers were targets for enslavement led to a problem encountered in Buganda some years later, when Ganda who had been children in the war were unable to successfully contract marriages because they did not know their real clans.

76Miti, 359.

77Miti, 368.

78Miti, 368.

79Twaddle, 78.


Baganda remembered the war as a time of unimaginable destruction. The

population of Bunyoro is said to have increased because so many people fled from war and the danger of enslavement in Buganda.80 An image that recurs in descriptions of the war is of corpses rotting by the roadside because no one was available to bury them. People stopped cultivating out of fear of fighting, and in the ensuing famine people dug up the stumps of banana trees in order to eat the roots. Warriors with guns "used to assuage their hunger by force of arms, carrying guns with them wherever they went and threatening to shoot anyone who would not give them food."8' An outbreak of bubonic plague followed the famine. Estimates of the death toll range from 7,000 to 400,001.8 After October 1889, Moslem armies had moved into and then out of Kyaddondo and Busiro because they were empty of people and animals and there was nothing left to raid; they then proceeded to plunder all of Kyagwe. Carl Peters passed through Kyaggwe early in 1890 and found "a desolation of destroyed banana groves, with vultures gorging on unburied corpses and the wind raising flurries of ashes in the burnt villages."v83

The civil war fundamentally undermined the institution of the kabaka.

Success for any of the factions in the civil war depended on having a prince of the drum-one entitled to become king because he was a direct descendant of a kabaka. Among the

'01. Nyakature, Anatomy of an African Kingdom 144, quoted in Twaddle, 60.

8tMiti 361.

82 The lower estimate is from Ashe, Chronicles, 144; the higher estimate is from Kaggwa, 119/155.

13 Wright, 10 1.


first actions of the Christian group after it withdrew to Ankole in 1888 was to try to acquire a prince of the drum. Bawmweyana, one of the sons of Mutesa, bribed his guard to allow him to escape to join the Christians, but Kalema had sent people to watch for him after he escaped, and he was captured. Kalema then decided to kill all the princes of the drum, and also all the princesses, because if they had no potential kabaka in their camp, they would have no means of regaining power.84 Princesses were killed as well as princes because the British were ruled by a woman, and therefore it seemed possible that Christians might put a princess on the throne. One generation earlier, Mutesa's mother the Namasole Muganzirwaza had caused the deaths of eleven of Suna's sons through starvation: only Mbogo, Mainja, and Kabaka Mutesa had been left alive.5 These two successive wholesale executions of princes may be an indication of the accelerating instability of the nineteenth century, because no earlier kabaka had considered it necessary to kill all his brothers. Even the practice of imprisoning princes of the drum had started in the late 18th century under Kabaka Semakokiro (who had killed his brother to obtain the kabakaship),6 The killing of two generations of princes (and one of princesses) was a disaster for Buganda because it created a dearth of potential effective leadership. When Kalema died of smallpox after his retreat from Mengo, the only possible kabakas were the Moslem leader Mbogo, two young sons of Kalema and Kiwewa who were out of the country with Catholic missionaries, and Mwanga, who had already been deposed once.

"Miti, 337; Kaggwa, B.skabak 114.

85Kaggwa, Jebak 76/130.

86Kiwanuka, 129.


The Christians turned to Mwanga because they had no other means of maintaining a credible bid for control of the kingdom. Since Mwanga had demonstrated his ineptness as a ruler and the chiefs had demonstrated their lack of respect for him the first time they overthrew him, his return to power inevitably entailed a further diminution of the authority of the kabaka. Unfortunately, the elements of Ganda society that served to balance the power of the kabaka were also in decline. The Kimbugwe, the chief who was officially the keeper of the kabaka's "twin" (an elaborate charm which contained the kabaka's umbilical cord), had the right to speak against the kabaka in the Lukiko and to try the kabaka for improper actions; but the Kimbugwe chiefship was abolished in 1892 because "during the Christian reign, we could not honor the traditional twin-god."87 Spirit mediums, who had served to safely focus legitimate criticism of the kabaka also lost their influence with the spread of Christianity and Islam, and with the general disorder of the time. Kakungulu was said to have a new drumbeat when the Christians and their allies gathered on Bulingugwe island

I eat whatever I find:
I eat whatever belongs to emmandwa.88

Kabaka Mutesa had been obliged to entertain and submit to the actions of the priest of the shrine of Mukasa, the Lubaale of the Victoria Nyanza. The son of that priest had lost his followers and his land was taken by the Gabunga in the 1890s; the grandson claimed to be "an important mutaka in Sesse as well as in Buganda" but, when questioned, he admitted

87Kaggwa, Basekabak 104/146.

88Paulo Kibi testimony, quoted in Twaddle, 50.


that he received land as a tenant of the Gabunga, and he performed services as the Gabunga's man.89

The crisis of newly strong chiefs and a thoroughly weak kabaka led to "wars that did not let go."' In 1888, Mwanga was driven out, Christians withdrew, and Kalema replaced Kiwewa as Kabaka. In 1889,a coalition of the forces of Christian chiefs and the forces of avowedly pagan chiefs brought Mwanga back to Mengo by the end of the year, after battles that had left fields where "skulls are as numerous ... as mushrooms."9' Groups that considered themselves supporters of the Moslem Kabaka Kalema, and after his death the Moslem Kabaka Mbogo, fought against the re-establishment of Mwanga all over Buganda. In 1892, fighting broke out in Mengo between people who identified themselves as Protestants, and people who identified themselves as Catholic. In 1893, negotiations for territory (see below) ended conflict among the new religious factions, but in 1894 and 1895, Baganda participated in fighting against the rebelling Sudanese troops. From 1897 to 1899, Mwanga and a coalition of chiefs fought against Baganda who allied themselves with "the Kampala European'. The "religious revolution" framework explains these continuing conflicts in terms of what religious group held Mengo or wanted to control it: but the triggers for each outburst of hostilities were so trivial that conflict seems

8gEntebbe Archives, SMP 6902, Butaka Commission Report, Guggu, Yosiya Ajabi
Sumugala, and Gabunga, 384-387. Cited hereafter as Commission, with witnesses names and page numbers.

9Mulira, 2.

91Solomon Wamala, Obulamu, 51, quoted in Twaddle, 52.


to have been prolonged for its own sake. More war meant more opportunity to acquire the wealth that came from raiding. Reporting on the volatile situation in Mengo in 1890, Captain MacDonald noted that chiefs seemed to be acting in their own interest, and the Kabaka "had little control over powerful, intriguing chiefs, ripe for any contingency that promised a chance of plunder."92

Re-allocating Land to Make the War End

At each shift in the control of the kingdom, the victorious group re-allocated

important chiefships. Naming the people who would control the sazas and the important functions of the kingdom was a way of stating the order of the nation. In the oral traditions of Buganda recorded by Apolo Kaggwa, the stories of each kabaka concluded with a list of the important chiefs during that reign; the named chiefs, in their named chiefships, constituted Buganda as it had been in that reign. In the civil war period, naming the chiefs was also a way of identifying how those who had taken control of the capital intended to manage the complicated problem of competing claims for authority between factions of chiefs. The problem faced by the groups of chiefs who came to power was to find a way to map the increasingly salient new categories of allegiance to Islam, and to English' and French' Christianity onto the structures of saza and subsidiary chiefships. Over and over again, these efforts failed. The order of Buganda, defined in chiefs of territories and chiefs of important functions, could not hold together at a time when organizing raiding had become the dominant occupation of chiefs. Furthermore, attempting to insert new religious categories into the structure of chiefships in some

92Macdonald Report (1893) PRO Series Fos (African) F02/60, quoted in Waller, 32.


logical way created even more instability, because the new categories always provided reasons for rationalizing renewed conflict. The various factions of chiefs attempted successively more radical techniques for combining religion and chiefly control of the provinces as the war continued.

In 1888, the Moslems and their Catholic and Protestant blood brothers assigned the katikiro ship to Christians, gave more than half the saza chiefships to Moslems, and gave two heavily armed bitongole to both Christians and Moslems; but this division fell apart within six weeks because an assertive Christian chief agitated for the post of Kauta, a chiefship that included the function of cooking for the palace and also land in the central provinces. His claim upset the precarious balance of power that had been negotiated and led to an armed skirmish which caused Christians to leave for Ankole.93 Those remaining in Mengo then had to carry out the action of identifying the chiefs of the nation again, in order to replace the Christian chiefs who had left. But this exercise in naming authority exacerbated tensions between Kiwewa and the Moslem chiefs who had brought him to power, resulting in a violent episode that ended in the departure of Kiwewa, the installation of Kalema, and a further naming of chiefships, to replace the ones who had just been killed. Both the Christian faction and the Moslem faction appointed chiefs for all the significant chiefships, even though they did not always control the sazas to which they were naming chiefs. Wright points out that the Moslems twice reappointed a chief for Kyaggwe when the intended Moslem chief was killed, even though the Moslem faction never held that area after they abandoned Mengo.

93Twaddle, 38.


When the coalition of factions that identified themselves as Christian regained

control of part of the mainland near Mengo in October, 1889, they named chiefs for all the sazas of the country. Two aspects of the chiefs' action reveal the intensifying crisis of authority in Buganda: the chiefs created an order of the nation that attempted to thoroughly balance the power of Protestant and Catholic factions, and they made the allocations entirely without the participation of Kabaka Mwanga. According to an eye witness, Mwanga "had no power" in the allocation; the chiefs made their decisions, and informed him afterwards.'4 This is the clearest possible evidence that the central place of authority in Buganda was empty. The structure of exchange in the country which, as Walker argues, required everything to flow into and out of the center no longer existed, and the figure who held the place that was also supposed to be that center no longer had real power.

The group of chiefs who had beaten Kalema defined Buganda as a balance of English religion adherents and French religion adherents at every level. The elaborate system they devised of alternating Protestant and Catholic chiefs appears to be an attempt to use the structure of chiefly control over territory and over subordinate chiefs to diffuse potential conflict. The system is described in detail in the report of Captain Macdonald in 1892:

The Estates, chieftainships and posts of honour and importance were
divided equally between the two parties on a system which aimed at
absolute fairness and justice but which was so complicated as in itself to
contribute a great and ever present source of danger. The principle adopted seems simple enough. Every holder of a post was to be under a superior of

"Wright, 117, 95.


the opposite party. Thus the owner of a Catholic sbhnba (estate) was
under a Protestant sub-chief, who in turn was subordinated to a Catholic
chief and so on and vice versa. In addition to this Buganda was divided into
ten districts--amasaza--five of which were headed by Catholic and five by
Protestant chiefs. Below these the alternation perpetrated [sic] but in
districts headed by a particular religious chief the estates belonging to him
were regarded as belonging to his party i.e. religious sect.'5

The alternation of Protestant and Catholic permeated Ganda structures of authority: even the estates of the Namasole, the Lubuga, and those of the Katikiro and Kimbugwe in every province were supposed to have subchiefs of the other religious persuasion under chiefs who held the same faith as the controllers of the estates.%

The orderly and logical plan imposed by the chiefs could not function in the actual conditions in Buganda at that time. In the imagined Buganda of the named chiefships, networks of chiefs expressed their relationship to each other by passing tribute up the hierarchy and receiving gifts down it, but actually the sazas were devastated by raiding, and decimated by famine and disease. The authority of chiefs should have derived from their submission to the authority of the kabaka and the kabaka's recognition of their role, but actually, there was no authority, only intense competition among powerful, armed chiefs, which was to be kept in check by their wiful intention to share power.

The ordering of the nation under layers of Protestant and Catholic chiefs implied that all people in authority were to be Christian. Since not a the Baganda were Christian, this decision indicates the connection people made between political and spiritual authority

95j.R.L. Macdonald, "Report on Uganda Disturbances in Spring, 1892" Entebbe Archives, quoted in L.L. Kato, "Government Land Policy in Uganda: 1889 to 1900, U~gand~a Jorl l~, 35, 2 (197 1) pp. 153-160, 153.

96Macdonald Report, quoted in Kato, 153.


--people would have to be found, of the appropriate religion, to rule in each particular locality, and the people below that chief would follow his religious lead. It is possible that the coterie of chiefs who made this plan believed that commitment to Christian faith implied a capacity to live peacefully. It is also possible that they chose complete powersharing in every direction as a means to prevent any accusation of unfairness. Some members of the coalition that had defeated Kalema were denied a leading role with the rationalization "bhiang is not religion".'7 Since bhiang smoking was identified with elephant hunting, (and had been forbidden by kabakas in the past) the exclusion of bhiang-smokers may have represented an attempt to eliminate the instabilities associated with new wealth and new military power.

Whatever the intentions and aspirations of the group of chiefs that devised the

ordering of Buganda in alternating layers of Catholics and Protestant chiefs, the system did not work. The Baganda were used to changes in chiefship being ordinary, fluid, and easily accomplished: this system required the order of chiefship to stay exactly as it was at the moment the system had been initiated. Any change of chiefs, or any chief's change in religion, became a source of conflict between the English' and French' factions. When a chief changed his religious allegiance, he lost his control of that chiefship. This was logical since the chiefs had linked the political authority inherent in control over land to prescribed religious allegiances, but it was impossible to carry out in the context of the Ganda practice of constantly reordering chiefships. Irresolvable problems arose when the Lubuga, Mwanga's sister, changed from Catholic to Protestant. The chief Yoswa

97Wright, 95-96.


Wasekere changed from Protestant to Catholic, provoking another confrontation. Catholic missionaries asserted that the principle of religious freedom required that people be allowed to retain land, even if they changed religions. In the volatile atmosphere of 1890, disagreements over who should control land quickly escalated into armed confrontation between Catholics and Protestants. When a minor chief in Kyaggwe who held what was supposed to be a Protestant chiefship became a Catholic, Protestants tried to evict the chief and Catholics agitated for him to remain. Semei Kakungulu, who was at that time the chiefs superior, travelled to Kyaggwe to resolve the problem, thirteen people were killed on the disputed estate, and shots were fired in Mengo.9"

Mwanga was entirely incapable of asserting the Kabaka's authority over land that might have resolved the disputes. The chiefs considered Mwanga to be someone who could be "herded like an ox" and they manipulated him to get the decisions they wanted." In 1891, a dispute arose because Mwanga had secretly given a village on Bussi Island to a Catholic, although this area should have been under the control of the Gabunga, the Saza chief of the Sesse Islands, which had been designated a Protestant chiefship. Two Catholics had been killed when they went to take possession of the land, because the Gabunga's men had refused to give it up, saying it was impossible that the land could have been transferred if the Katikirro's representative was not there to "show the land" and make the transfer. When Mwanga attempted to decide against the Gabunga, a Protestant chief stood up in the Lukiko and shouted at the Kabaka "No, sir Kabaka! You are wrong!

98Macdonald report, cited in Twaddle, 77.

"Wright, 99.