Recent scholarship in archaeology, history and anthropology has begun to critique the deeply embedded link between the emergence of complex societies and evolutionary models of development. Rather than taking centralization as an inevitable outcome of complexity, these works have questioned the pervasive notion of ‘complexity as differentiation by political hierarchization’ and have focused instead on the various strategies African communities have employed in constructing arenas of collective action.Footnote 1 The narrow, linear lines of organic growth that once dominated the literature on state development and complexity are rapidly disappearing, and scholars are now grappling with the task of proposing alternative explanations. This growing body of literature points to the importance of diffuse, heterarchic and segmentary forms of power and to the significance of ritual, technological and other types of knowledge in achieving complexity.Footnote 2 These works emphasize the need to consider different sources of power and to look for voices that chiefs and rulers cannot express. In so doing, they acknowledge the ‘creativity of power’ by highlighting how efforts to establish control over material resources – an essential component of crafting polities – often relied on less visible resources such as the capacity to manage the ideas underlying human actions intended to affect the relationship between the natural, social and supernatural realms.Footnote 3
Perhaps not surprisingly, this skeptical thread of anti-social-evolutionist scholarship has been accompanied by a thorough critique of lineage theory. In a critical retrospect published nearly twenty-five years ago, Adam Kuper reflected upon what he considered the ‘long-evident bankruptcy’ of the lineage model, which he ultimately declared of ‘no value for anthropological analysis’. For Kuper, the model's bankruptcy resulted from the fact that it corresponded with neither how actors viewed their societies nor the actual organization of political and economic activities.Footnote 4 In a recent article published in this journal, Wyatt MacGaffey elaborated upon and refined Kuper's critique of lineage theory as part of an examination of what he described as the ‘historiographic burdens imposed by questionable anthropological models of kinship’.
According to MacGaffey, the significance accorded to lineality and corporate kin groups in discussions of African history stems more from late nineteenth- and early twentieth-century anthropologists' preoccupations with explaining social evolution and the maintenance of order in so-called primitive societies than from the centrality of kinship in African social organization.Footnote 5 Despite this critique of lineage theory, however, MacGaffey declared that Kuper was ‘almost certainly wrong’ in pointing to the lack of correspondence between actors' views of their social world and the lineage model. Rather than casting doubt on the existence or importance of descent groups, MacGaffey questioned the primacy accorded to these groups in prevailing understandings of African social organization, declaring that ‘there is not and never was any such thing in Africa as a matrilineal society’. Like races and tribes, matrilineal or patrilineal societies serve for MacGaffey as ‘imagined objects that made good myths for a while’ but ultimately require abandonment.Footnote 6
Jane Guyer and Samuel Belinga have offered a different yet related explanation for the primacy of kinship in the study of precolonial Africa. In their examination of community-building processes in equatorial Africa, Guyer and Belinga noted how the commonly encountered assumption that kinship served as the guiding principle of African social life may in large measure reflect the effects of conquest and the implementation of colonial rule. They pointed out how the disorder and demographic collapse that accompanied the establishment of colonial rule in many parts of Africa rendered inoperable previous processes of social mobilization based on the complementarity of different bodies of knowledge. Kinship, on the other hand, survived this period of despair and loss and therefore rose to the forefront of ethnographic endeavors undertaken during the early colonial period. Rather than representing the fundamental building block of African social life, then, kinship constituted ‘“what [was] left” in the organizational repertoire’ when ethnographers composed the earliest descriptions of African communities. When subsequent scholars turned their attention to examining the precolonial period, the models of community formation they developed preserved the primacy accorded to kinship in these early ethnographies. Whether based on the lineage mode of production prevalent in the 1960s or the concept of wealth-in-people launched in the 1970s, these models focused on kinship as a means of managing processes of control and accumulation, a theoretical insistence that preserved problematic assumptions from evolutionary theory.Footnote 7
Critiques of both social evolutionist scholarship and lineage theory correctly call into question the models scholars have most frequently employed to describe social organization and modes of order in Africa. In so doing, these critiques compel us to re-evaluate the meaning and significance of the various components that have constituted the building blocks for these models. Many of these components – such as the ubiquitous references to the existence of ‘clans’ in almost every region of the continent – have long occupied a substantial place both in the lives of Africans and in discussions surrounding political complexity and the expansion of scale in early African history. In this regard, we might take care to distinguish between the models we have employed and their constituent parts. For while matrilineal or patrilineal societies may represent ‘imagined objects’, matriclans and patriclans are certainly not a figment of professional historians' and ethnographers' imaginations – they are ‘out there’, in the words of Vansina.Footnote 8 This raises the question of how we should make sense of these phenomena in the absence of the models that once gave them meaning.
This article offers an example of how we might approach such a question. It examines the early history of clanship in Buganda by asking why our earliest glimpses into the distant Ganda past appear in the form of clan histories.Footnote 9 Over the course of an extended period of fieldwork in Buganda, I engaged in many informal conversations about the meaning of clanship as people walked me through their clan estates and explained the significance of the shrines and other sites located on these plots. These conversations, which revealed a connection between clanship and Ganda notions of individual as well as collective well-being, both suggested that we might best examine the clan concept from the perspective of public healing and pointed to a novel way of approaching the interpretation of more formal, structured clan histories. By bringing into focus the discursive, ritual and political connections between shrines located on discontiguous clan estates, the practice of walking through clan estates also brought into focus the significance of examining transformations in public healing for understanding shifts in the scope of political activities in the distant Ganda past.
David Schoenbrun's historical linguistic work on the history of spirit mediumship and banana cultivation in the Great Lakes region, along with Holly Hanson's examination of the transformations in productive, social and spiritual relationships that accompanied the transition to land-intensive banana farming in Buganda, offer a way to anchor the insights derived from Ganda clan histories and less formal conversations about clanship within a broader regional narrative. The combination of these various sources of evidence and scholarly insights suggests the following scenario: beginning around the fifteenth and sixteenth centuries, Ganda expanded upon earlier knowledge of banana cultivation to develop a land-intensive banana farming system. The opportunities for permanent settlement and the accumulation of wealth in the form of perennially fruiting banana trees rendered bananas a potentially lucrative long-term investment. The gradual transition to intensive banana farming, however, also generated social tensions surrounding property inheritance and succession that posed significant challenges to the communities living along the northwest shores of Lake Victoria. In the face of these challenges, Ganda developed dispersed therapeutic networks that reconfigured the boundaries of public healing. They did so by transforming previously territorial spirits into portable spirits capable of ensuring the health of disconnected groups of people. This process drew upon the skills of itinerant mediums and lay at the core of clan formation. The webs of shrines situated on discontiguous clan lands created therapeutic networks that drew together communities whose leaders possessed a variety of skills, thus forging a powerful connection between clanship, collective health and the composition of knowledge.
The early history of clanship in the area that would become Buganda reveals how the domains of public healing and politics were intimately entwined in practical ways. This observation provides new insights into the practice of politics in the distant Ganda past and also offers a novel perspective from which to view Buganda's subsequent development. In his important research on the role of public healers in precolonial East Africa, Steven Feierman has pointed to how public healing served as an alternate realm of authority that lay beyond the purview of chiefs or kings. Feierman has illustrated how public healers drew upon their standing in this alternative realm as a means of evaluating, critiquing and rebelling against political leaders.Footnote 10 In the periods preceding the emergence of an instrumentally powerful kingship in Buganda, however, political leaders acted as public healers who sought to ensure the well-being of increasingly complex and dispersed communities. This convergence of public healing and political authority in the distant Ganda past compels us to historicize the frequently observed divide between the authority of healers and the power of royals and their representatives. The gradual institutionalization of this divide occurred later as part of the violent and extractive character of Buganda's formation in the eighteenth century. Throughout these momentous developments, however, efforts to ensure collective prosperity and perpetuity – usually expressed in the idiom of health and healing and linked to shifting notions of clanship – continued to influence community-building processes.
‘THE KIGGWA BRINGS TOGETHER THE OWNERS OF SHRINES’: CLANSHIP AND THE FORMATION OF THERAPEUTIC NETWORKS
That the construction of clan histories depended upon the presence of clans perhaps seems like an obvious statement. Yet in simply scanning clan histories for clues about Buganda's past, most historians have failed to account for the fact that these histories exist in the first place. The question of what holds clans conceptually together – of what, essentially, enabled historical actors to compose clan histories – compels us to consider clans as something other than groups of people who share a common history. Treating clan histories as the outcome of historical processes, in other words, urges us to contemplate the circumstances that made possible the very construction of these histories. At the core of these matters lies the relationship between clanship and Ganda practices of public healing.
The convergence of clan name and totem, along with a distinctive series of subdivisions conceived through the idiom of kinship, together distinguish Ganda clans from similar structures found throughout the Great Lakes region.Footnote 11 These two features also provide the conceptual framework within which Ganda narrate clan histories. The presentation of these histories, which detail the movements and activities of clan founders and their descendants, suggests an enticing mode of analysis by offering the possibility of tracing the gradual dispersal of Ganda clans from their central estates (butaka bukulu) to their numerous scattered and discontiguous settlements. This reasonable prospect presents itself wherever people employ the idiom of kinship as a means of narrating clan histories and has indeed seduced scholars working in Buganda and its neighboring areas.Footnote 12 The problem with this approach, however, lies in its failure to distinguish between the language of clanship and the types of relationships embedded in this language, a failure that results in the privileging of a model of clan formation based on notions of kinship and descent to the exclusion of other, perhaps more revealing, formulations.Footnote 13
Ganda discourse surrounding clanship, as opposed to the narration of clan histories, suggests an alternate manner of conceiving the philosophical and historical underpinnings of clan formation. When Ganda speak about clanship, they offer a conceptual map that differs from the more familiar familial models. Whereas Ganda clan histories invoke the image of a gradually expanding family tree sprouting from the clan's founder, more informal and less formulaic discussions of clanship paint the picture of a ritual web consisting of dispersed yet connected therapeutic networks. In placing emphasis on the integrative capacity of the shrine centers located on scattered and discontiguous clan lands (butaka), Ganda squarely situate the concept of clanship within the realm of public healing. This discourse provides insights into the manner in which Ganda perceive clanship and opens the door for a novel approach to examining the clan concept.
Over the course of my stay in Buganda, I engaged in many discussions, both formal and informal, about the meaning of clanship. These discussions differed from those in which I encouraged people to narrate the histories of their clans or their particular branch within the clan. They were more personal, more about how people located themselves, in an existential sense, within the discursive and practical boundaries offered by the clan concept. Some of these conversations occurred in casual settings, almost accidentally or unintentionally at times, while others unfolded as people guided me through their clan estates. Many of the people I spoke with had received a Christian or Muslim upbringing, a factor which, along with a justifiable skepticism about both my interest in such matters and my ability to grasp their significance, initially hindered some aspects of these exchanges. Yet after much persistence and a few stumbling demonstrations indicating that I did, in fact, have some knowledge about the rudiments of Ganda clanship, many of my hosts elaborated upon the importance of the shrines and other sacred spaces located on clan estates. The head of a lineage (mutuba) in the Colobus Monkey clan, for instance, related the following personal narrative:
I was living a good life but had one problem – I did not have a child. I had wives and money but no children. My father was still living and I asked him why I had money and other things but could not have children. He took me to Bussi, where our clan's main shrine is located. The people at the shrine told me to construct a branch of the Bussi shrine at my house. They instructed me that after putting up the roof I should slaughter a cow, a goat, a sheep, and a chicken to distribute to people. I was told that I would not have children if I did not do these things. I want to assure you that I did not take any medicine, and now I have thirty children. I constructed a branch of the Bussi shrine here at my home in order to provide an opportunity to those who cannot cross the lake to reach Bussi. I then became a medium without succeeding or inheriting the position.Footnote 14
While these sorts of narratives did not provide any specific details regarding the distant histories of a particular clan or segment of a clan, they did highlight the intriguing connection between the ideology of clanship, the shrines located on clan estates and Ganda concepts of personal as well as collective well-being. Over the course of listening to several similarly autobiographical narratives, I gained glimpses into how people perceived and experienced clanship beyond the conventional and much documented (though far from insignificant) discourse surrounding succession and inheritance disputes in twentieth-century Buganda.
When the people I spoke with reflected more generally upon the manner in which clanship operated as a guiding principle in their lives, they frequently pointed to the significance of biggwa (kiggwa, sg.) – the sacred sites located on each clan's primary estate. Many people described biggwa as places where individuals sought diagnoses of personal problems such as illness, infertility, poverty or faltering relationships. But they also described how clan members gathered at these sacred sites, more so in the past than today, to remember their ancestors at their place of origin.Footnote 15 Prior to the dramatic changes in public healing practices that accompanied the imposition of colonial rule and the spread of Christianity, most biggwa included a large conical shrine that housed a clan's most prominent spirits during these public clan gatherings. These spirits usually resided in nearby features of the natural environment and regularly manifested themselves in animal form. But they could also be called upon by trained mediums with the assistance of the priests who lived on clan estates and looked after biggwa. Ganda in fact drew upon the imagery of spirit possession in naming the locales at which these public meetings occurred. They derived the word kiggwa from the verb okuggwa (to be exhausted or worn out), thus designating these sites of (re)collection as places where mediums, following a common practice found throughout the region, fell to the ground in exhaustion once a possessing spirit had exited their bodies.Footnote 16
Clan members converged around biggwa to feast, drink and recite historical narratives. Most importantly, however, they gathered to seek solutions to collective problems (ebizibu ebwawamu) such as famine, epidemics and warfare.Footnote 17 These conditions required that clan members collaborate in a concerted effort to ensure their collective well-being. They pursued this objective by engaging the area's dominant spiritual entities, which maintained a connection to the clan's founding ancestors whose graves were located at the nearby burial grounds (kijja). These two features – a sacred site and a burial ground – together constituted a clan's primary estate, which ‘belonged to the whole clan’ and served as a meeting point for ‘those who believed in their totems’ to ‘beat the drums and call the spirits which protected the lives of all descendants’.Footnote 18 The activities undertaken at these gatherings, which included large-scale practices of spirit possession, drew upon the idea that the public recognition of common spiritual entities generated the conditions for collective prosperity.
As the sites of these collective endeavors, biggwa functioned as ritual arenas where clan members assembled to seek life, peace (bweza) and collective good fortune.Footnote 19 The shrines located at biggwa, which were often situated on hilltops, dotted the areas surrounding the northwest shores of Lake Victoria and anchored people, in both a discursive and a visual sense, within dense healing networks. A series of dependent shrines (masabo), each connected to a branch of a particular clan at either the maximal lineage (ssiga), minor lineage (mutuba) or minimal lineage (lunyiriri) level, lay scattered on discontiguous secondary clan lands. These dispersed shrines constituted ritual networks that shaped the framework within which communities sought their collective health. They formed the ritual rungs of therapeutic ladders that stretched across the Ganda curative landscape, with each upward movement representing both an act of ritual escalation – the shifting of ritual activities to ever more important shrines – and a metaphorical step towards the spiritual powers housed at the clan's original home. Phrases like ‘the kiggwa unites all [clan members] from different patrilineages (masiga)’, ‘the kiggwa brings together the owners of shrines (masabo)’ and ‘the kiggwa brings together people from different areas’, which surfaced repeatedly during my conversations in Buganda, served as a testament to this connection between clanship and public healing.
CLANSHIP, COLLECTIVE HEALTH AND THE COMPOSITION OF KNOWLEDGE
The personal narratives and more general ruminations about clanship that I listened to over the course of my stay in Buganda offer insight into why our earliest glimpses into the distant Ganda past appear in the form of clan histories. In the period preceding the emergence of the Ganda state apparatus in the eighteenth century, Ganda employed the ideology of clanship as a means of constructing healthy political and social bodies. As the discursive and ritual cement joining people living in discontiguous territories, the ideology of clanship guided the historical formation of publics over which spirits – and their earthly representatives – sought to extend their efficacy. In this manner, clanship played a critical role in shaping the publics within which early Ganda communities sought to ensure their collective well-being, and the remembered founders of some of the earliest communities represent increasingly portable spirits (or their human vehicles) that joined people living in dispersed therapeutic networks. The history of the Otter clan, which claims indigenous status in Buganda, provides a revealing illustration of this process.Footnote 20
Members of the Otter clan trace their ancestry to the clan's founder Kisolo Muwanga Ssebyoto, bearer of the spirit Muwanga, father of the hearths, who transformed wild bananas into domesticated banana trees.Footnote 21 The clan's history begins with Kisolo's journey from his home at Nseke in Mawokota to Bweza in Busujju, where he established a settlement close to the powerful spirit Kintu's abode at Magonga.Footnote 22 Kisolo's capacity to possess the renowned spirit Muwanga attracted the attention of Kintu, who had himself recently settled in the area.Footnote 23 Eager to court Kisolo's patronage, Kintu granted him a plot of land at Bbongole opposite his home at Magonga. In return, Kisolo offered Lutaya, the oldest of his three sons, to perform the ritual weaving of the roof for Kintu's house, an exchange that suggests an alliance between the territorial cult of Kintu at Magonga and that aligned with Kisolo and the spirit Muwanga.Footnote 24 Lutaya's successors preserved this position within their patrilineage by performing subsequent weavings during the repair or rebuilding of Kintu's shrine, thus ensuring the clan's continued linkages to the ritual complex surrounding Kintu's shrine at Magonga as well as their control over the nearby estate at Bbongole.Footnote 25
Following a violent confrontation with Kintu at Magonga, Kisolo fled to the home of his second son, Ssenkungu, at Kyanja in Mawokota. When Kisolo reached Kyanja, his son Ssenkungu, inventor of the grooved mallet used for beating out barkcloth, directed his father to the home of his eldest son (Kisolo's grandson), Katwere Muganga, at Ffunvu in Mawokota. A talented doctor, Katwere Muganga gathered some herbs and treated the leg wound inflicted upon his grandfather by Kintu.Footnote 26 After his wound had healed, Kisolo returned to his original home at Nseke in Mawokota, where he ‘disappeared and became a musambwa (territorial spirit) [misambwa, pl.]’.Footnote 27
Despite his disappearance, or perhaps precisely because of his disappearance and subsequent new standing as a spiritual entity, Kisolo continued to exert influence on his descendants. His departed spirit, known as Muwanga, spoke through the medium operating at the shrine built for Kisolo at Nseke. The spirit Muwanga ‘looked after the well-being of the [Otter] clan, multiplied their cattle, and made their women fruitful’, and the shrine at which it possessed the resident medium also served as a site where mothers who had recently given birth brought offerings of beer, cattle and firewood.Footnote 28 Ssemwanga, another one of Ssenkungu's sons and Kisolo's grandsons, served as priest and caretaker of this shrine, a position he acquired after his father had planted a fig tree for him at Nseke and which has remained in his lineage ever since.Footnote 29
The Otter clan founder Kisolo's third son, Kinkumu Kitumba, remained at Bweza in Busujju when his father fled to Mawokota following his confrontation with Kintu. Kinkumu Kitumba succeeded Kisolo after his disappearance at Nseke and retained control over the estate at Bweza, which became the clan's principal estate and burial ground for subsequent heads of clan. Bweza's proximity to Kintu's shrine at Magonga further reinforced the clan's connection to the ritual complex surrounding Kintu. The estate also served as the clan's principal gathering spot and contained the clan's main shrine, around which clan members converged to honor the spirit of their founder Kisolo Muwanga Ssebyoto. In this respect, the shrine at Bweza provided a discursive and ritual connection to Kisolo's shrine located on his grandson's estate at Nseke as well as those situated on the clan's other dispersed secondary estates.
The history of the Otter clan provides a valuable glimpse into the early dynamics of Ganda clanship, the first hint of which appears in the clan's totem. The otter's anomalous characteristics lend the animal to particular metaphorical associations, and the animal's symbolic significance in the Great Lakes region situates it within the realm of healing practices involving acts of spirit possession. In explaining the significance of the totems associated with Ganda clans, Ganda healers describe a world in which all non-human forms of life fall into one of five categories: animals (ensolo); creatures that fly (ebibuuka), including birds; insects (ebiwuka); creatures of the lake, including fish (ebyenyanja); and plants (kimera).Footnote 30 A mammal that spends a considerable amount of time in water, the otter eluded categorization according to the Ganda taxonomy of non-human beings, and throughout the region Great Lakes historical actors viewed the otter as possessing metaphysical capacities.Footnote 31
The name of the Ganda Otter clan's founder further suggests that the clan played a prominent role in the area's healing complexes in the distant past. The full name of the clan's founder – Kisolo Muwanga Ssebyoto – points towards the clan's origins as a musambwa-type territorial cult aligned with a node of ritual authority in Mawokota. The name Kisolo derives from the Luganda word nsolo (animal), a reference to the common occurrence throughout the region of territorial misambwa spirits manifesting themselves in animal form. The name Muwanga, however, provides perhaps a more telling clue about the type of authority that the Otter clan's early leaders exercised in Mawokota and extending westwards into Busujju. Muwanga occupies a prominent position in the pantheon of spirits associated with the Ssese islands, and Ganda regard the spirit as a lubaale – one of the kingdom's national spirits. The Otter clan, however, holds a particularly close affiliation with this highly revered spirit. As mentioned earlier, Muwanga's main shrine operates on the mainland at Nseke, Mawokota, under the direction of a priestly guardian from the Otter clan. The name of the clan's founder therefore hints at the clan's origins as part and parcel of the spirit Muwanga's expansion from the Ssese islands, where the Otter clan claims indigenous status, to the mainland in Mawokota and beyond.
Building on this suggestion, we might regard Kisolo Muwanga Ssebyoto as a disembodied spiritual entity rather than the embodied person who founded the Otter clan. Viewed from this perspective, Kisolo's journeys as recorded in Otter clan narratives appear as the gradual movement of the spirit Muwanga from his ‘home’ in the Ssese islands to the coastal province of Mawokota and then further inland into Busujju. Made possible through the work of itinerant mediums, the spirit's journey prompted the emergence of dispersed therapeutic networks whose members turned to the powers of the spirit Muwanga for their well-being and drew upon the familiar idiom of kinship to describe their relationship to one another. In this regard, rather than preserving a record of the migration and segmentation of kin groups, the genealogy embedded in the Otter clan's history represents more the outcome of historical alliances between dispersed communities cemented by the intimately connected practices of clanship, spirit possession and public healing.
The spirit Muwanga's journey, however, resulted in more than simply the establishment of a series of dispersed shrines devoted to his continued perpetuity. The immediate descendants of Kisolo Muwanga Ssebyoto, who might also be regarded as a medium for the spirit Muwanga, together constituted a network of knowledge that included a priest, a critical figure in the ritual complex surrounding Kintu at Magonga, a talented doctor, a craftsman who made mallets used in barkcloth production, a well-known barkcloth artisan and a famous diviner.Footnote 32 The accounts of their individual talents and undertakings preserved in Otter clan narratives, along with the ritual activities undertaken at the shrines located on clan estates, provide further insight into the intellectual principles informing the clan concept. These accounts suggest that the general notion of collective well-being situated at the core of clanship drew upon the connection between collective health and the effective composition of knowledge. The practices of public healing undergirding the clan concept served as a form of practical reason that linked a community's collective health to the capacity of its leaders to bring together various sorts of knowledge and skills. In this regard, the ideology of clanship offered both a language and set of practices for incorporating and preserving new sorts of skills and statuses into social collectivities. Healthy communities, in other words, resulted from knowledgeable communities, and the networks of knowledge achieved through this compositional process connected centers of production and exchange whose successful coordination contributed to a community's collective well-being. These therapeutic networks also formed the skeletons around which communities composed clan narratives, the recitation of which ultimately became part of the clan formation process. In this regard, the composition of clan narratives constituted part and parcel of community-building processes and strategies.
SPIRIT POSSESSION, PORTABILITY AND INTENSIVE BANANA CULTIVATION IN THE GREAT LAKES REGION
The history of the Otter clan points towards the considerable intellectual and social changes surrounding a series of innovations in the practice of spirit mediumship. These changes accompanied the transformation of previously territorial spirits into portable spiritual entities through the work of itinerant mediums. To better understand these historical developments in Ganda public healing, however, we must locate them within a broader historical framework. How, then, do we set about situating narratives such as that surrounding the early history of the Otter clan within the region's social and cultural history? Taken on its own, the clan's history does not provide any sort of chronological anchor: we simply cannot determine when the developments described in this narrative occurred based on a reading of the narrative alone. We must therefore search other sources of evidence for a historical hook with which to secure the narrative to an identifiable temporal framework. David Schoenbrun's historical linguistic analysis of Great Lakes Bantu languages, which furnishes an independently derived body of knowledge, provides just such a hook.
If the early history of the Otter clan contains the historical residue from a period of innovations in the institution of spirit mediumship, then the linguistic evidence relating to the history of this practice provides an appropriate starting point in our efforts to set this narrative into historical motion. As Schoenbrun illustrates, this evidence indicates how spirit possession served as one of the oldest techniques Great Lakes historical actors employed to meet a variety of intellectual and material needs.Footnote 33 The beginnings of the practice stretch deep into the distant past, as evidenced by the wide distribution of the root *-bándwa (to be consecrated to a spirit) in Great Lakes Bantu languages. Forms of this root appear in all five branches of these languages with meanings related to spirit possession, suggesting that the experience of being possessed by a spirit probably constituted a part of the ancient Great Lakes Bantu cultural world.
At some point between the fourteenth and sixteenth centuries, however, historical actors in the northern Great Lakes region instituted an intellectual innovation that marked a critical transformation in spirit possession practices. By joining spirits to their mediums rather than to a specific territory, these historical actors enabled the transformation of territorial spirits into portable spiritual entities and severed the limiting conceptual connection between spiritual and territorial authority. This process overlapped with another critical transformation in kubándwa early in the second millennium: the institutionalization of spirit possession practices. As Schoenbrun notes, linguistic evidence for the portability of spirits appears in the fact that the term referring to this new type of professional medium – mbándwa – named both a medium and a spirit while the term for territorial spirit – musambwa – named a spirit and its territorial location but not its medium. No longer anchored in a particular locality, these transformed territorial spirits could extend the range of their healing capacities through the substantial efforts of professional itinerant mediums, who carried these portable spirits to new settlements and drew upon their expanded capacities to meet the needs of larger communities.Footnote 34
The innovations in spirit possession practices gleaned from the historical linguistic record and detailed in the history of the Otter clan constituted part of a broader set of historical developments in the area surrounding the northwest shores of Lake Victoria. Historical linguistic evidence suggests that these transformations in public healing occurred during a period in which Ganda were conducting critical experiments with intensive banana cultivation. As Schoenbrun demonstrates, the distribution of terms for various banana cultivars as well as those related to the cultivation process indicates that, while the sporadic cultivation of bananas had occurred in the Great Lakes region since at least the second half of the first millennium CE, the period of intensive development emerged considerably later.Footnote 35 The large number of completely unique terms for varietals in modern Great Lakes Bantu languages attests that the land-intensive practices familiar today in Buganda, Buhaya and parts of Kivu developed relatively recently, beginning sometime after the fifteenth century. The greatest number of these terms appears in Luganda, whose sixty unique terms for banana varieties suggest that cultivation may have been most intensive in the areas along the northwest shores of Lake Victoria that would become Buganda. Since Luganda and its closest genetic neighbor Lusoga began to differentiate into distinct languages in the sixteenth century, we can reasonably trace the development of Buganda's land-intensive banana farming system to around this period.Footnote 36
In sum, linguistic evidence related to both banana farming and spirit possession practices indicates that the gradual development of intensive banana farming in the area that would become Buganda coincided with the transformation of previously territorial spirits into mobile spirits capable of ensuring the collective health of larger, increasingly dispersed communities. Public healing practices, in other words, became de-territorialized as banana cultivation became highly territorialized. This paradoxical development requires closer analysis. Linguistic evidence merely suggests that these two processes coincided and does not indicate any sort of relationship. Holly Hanson's examination of the reorganization of social relations and settlement patterns that accompanied the transition to land-intensive banana farming, however, offers some substantive historical premises for connecting these overlapping developments.
LAND TENURE AND THE BANANA EcONOMY ALONG THE NORTHWEST SHORES OF LAKE VICTORIA
Between the fourteenth and sixteenth centuries, the communities living along the western shores of Lake Victoria gradually incorporated a system of land-intensive banana cultivation into an already diverse farming base.Footnote 37 The fertile soils along the lakeshore were far more conducive to banana farming than the grasslands further west. As Hanson and others point out, however, certain regions were particularly well suited to embrace this agricultural innovation, and the crescent-shaped stretch of land surrounding the northwest shores of the lake provided an exceptionally favorable setting. The upper and middle slopes of this area's many hills were covered with particularly productive soils, capable of supporting heavy cultivation in moderate climates with adequate and evenly distributed rainfall, both of which characterized the area surrounding the northwest shores of Lake Victoria. The relatively even spread of rainfall in this area made possible a kind of permanent cultivation that relied on a perennial crop.Footnote 38
As Hanson observes, the gradual development of a banana-based economy in Buganda prompted momentous changes in settlement patterns. Whereas farmers who practiced banana farming prior to the transition to land-intensive banana cultivation ‘broke new fields every few years and moved their homes about every decade’, those who came to rely increasingly on the banana as a staple food established new types of settlements. The fact that fully grown banana trees continue to produce fruit for many years meant that people developed more permanent settlements as they came to rely on bananas for a larger portion of their food. Banana farming represented a good investment and fostered deeper attachments to productive land than the cultivation of either grains or tubers. Hanson observes that, unlike crop land devoted to grains or tubers, which would be fallowed after three or four seasons, the land dedicated to the maintenance of banana groves remained under continuous cultivation for decades and even generations.Footnote 39
The development of more permanent communities in turn fostered the emergence of larger and more densely populated settlements. As communities increasingly came to depend on the banana as a staple food, the previously dispersed societies living along the western shores of Lake Victoria began to concentrate themselves around the best banana lands. Hanson describes how the growing importance of banana plantations made these lands attractive places for settlement and most likely resulted in increased population densities. While we cannot know how many people lived in the communities surrounding the best banana-bearing lands, Hanson's observations about the higher food yields resulting from land devoted to bananas suggest that these communities were certainly more heavily populated and densely settled than those located on lands unable to raise and sustain banana gardens. These higher food yields may also have resulted in an overall increase in population levels and the establishment of new settlements on previously unoccupied ridges.Footnote 40
The stability of the settlements surrounding the lands best suited to intensive banana cultivation led to dramatically increased stakes involved in securing continued access to these lands. As Hanson points out, however, not all land in this region was equally conducive to banana farming. Variations in soil qualities and rainfall in the region that became Buganda determined the differing potential of land for banana cultivation, and the best plots were not often contiguous but were separated by swamps and stretches of rocky hilltops.Footnote 41 The establishment of permanent settlements around the best banana-bearing lands prompted competition for these resources and generated newfound concerns regarding the related matters of inheritance and succession. The increased economic – and therefore political – value of banana lands meant that these plots were guarded more closely than other areas. While community leaders used their control over productive banana lands to attract followers who sought access to them, they simultaneously developed ideologies and institutions designed to preserve these repositories of wealth within particular social groupings. These efforts to secure and maintain access to banana groves resulted in an increased emphasis on lineality and patriarchal forms of authority.Footnote 42 Banana gardens represented ‘invested labor and generations of surplus value stored in the form of perennially fruiting trees’, and prominent farmers drew upon recently developed inheritance ideologies as a means of transmitting these valuable lands to designated heirs.Footnote 43
PUBLIC HEALING, CLAN FORMATION AND LAND-INTENSIVE BANANA CULTIVATION
The earliest processes of clan formation in Buganda coincided with the reorganization of communities that accompanied new land tenure practices stemming from the shift to banana farming. Perhaps not surprisingly, the few scholars who have discussed the role of clans in early Ganda history have presented functionalist interpretations of clanship that revolve around competition for material property. The related matters of inheritance and succession have guided these discussions, which have focused on how clans and their constituent patrilineages served as the preeminent vehicles for both converting disconnected stretches of fertile land into banana gardens and directing land-use rights. Scholars have for the most part conceived of clanship as a tool for managing social relationships and have treated clans as functional bodies that preserved sources of wealth and political standing for some people and limited the ambitions of others. According to this model, male leaders sought to employ their authoritative positions within patrilineages to control access to valuable clan lands and to distribute strategic resources, an enduring form of political power that would come to characterize the Ganda state.Footnote 44
Discussions of Ganda clanship that focus on the simultaneous emergence of patriarchalism, lineality and a land-intensive farming system have provided critical insights into the historical circumstances that prompted communities to develop social boundaries distinguishing insiders from outsiders, and they have shed light on some of the benefits that insiders could expect to receive. Yet the sorts of connections most often drawn between clanship, land-use rights and inheritance would apply equally to lineages, clans or any sort of social collectivity that drew upon notions of unilineal descent. As discussed earlier, however, Ganda clans were not simply lineages writ large. The ideology and practices of clanship in Buganda built upon transformations in public healing practices that posited a connection between health, wealth and the effective composition of knowledge. Understanding the relationship between the process of clan formation and the transition to intensive banana cultivation therefore requires that we situate the discussion of banana farming in Buganda within our earlier examination of the shifting boundaries of public healing.
While banana gardens anchored the land-intensive farming system in the areas that would become Buganda, their significance stretched far beyond their serving as a source of food or beer. Hanson describes how ‘the settled, fixed quality of intensive banana cultivation’ provided people with a new means – and a new vocabulary – for establishing relationships of obligation, both with the ancestors and between leaders and followers. Like their Great Lakes neighbors, Ganda had long recognized the importance of maintaining healthy relationships with their ancestors. But the development of more permanent communities both transformed the character of these relationships and amplified their significance. As Hanson observes, ‘the presence of … graves in a place that people stayed for decades … added a dimension to the way [Ganda] remembered ancestors’. In these circumstances, maintaining control over valuable plots came to revolve around the capacity to establish widely recognized connections with the ancestors whose ghosts hovered around burial sites, ensured the land's continued fertility and invited a host of prestational actions. People chose to live in close proximity to the graves of important ancestors in the hopes of securing such linkages, and as communities stabilized around the best banana-bearing lands, ‘the power of the dead over land’ provided the conceptual framework within which contests surrounding the control of these fertile estates occurred. Those who emerged victorious in these contests were the people who managed successfully to secure connections with the ancestors buried in a particular settlement.Footnote 45
Hanson's work on the social and spiritual transformations that accompanied the development of land-intensive banana farming allows us to perceive better the connection between agricultural transformations, the process of clan formation and innovations in the practice of spirit possession in the areas surrounding the northwest shores of Lake Victoria. The establishment of dispersed therapeutic networks that reconfigured the boundaries of public healing – the core of the earliest processes of clan formation in Buganda – served as a means of contending with the challenges and opportunities surrounding the development of land-intensive banana farming. The opportunities for permanent settlement and the accumulation of wealth in the form of perennially fruiting banana trees rendered bananas a potentially lucrative long-term investment. As discussed above, however, the gradual transition to intensive banana farming also presented significant challenges to the communities living along the northwest shores of Lake Victoria. The increased value of land generated contests surrounding land-use rights and property inheritance. These contests unfolded within the realm of public healing and were mediated by the historical actors best situated both to shape and to manage the shifting boundaries between the material and spiritual realms.
Kisolo Muwanga Ssebyoto, the remembered founder of one of Buganda's earliest clans, represented just such a figure. As either increasingly portable spiritual entities or their human vehicles, figures such as Kisolo sought to manage the opportunities and challenges surrounding land-intensive banana farming by securing a more widespread and increasingly diverse community of followers. This process drew upon the skills of itinerant mediums and lay at the core of clan formation. The webs of shrines situated on clan lands drew together dispersed communities whose leaders possessed a variety of skills. The proper management of these skills – the effective arrangement of various forms of knowledge distributed over wide geographic areas – resulted in the formation of more complex and efficacious social and productive networks. The individuals and communities recruited into these expanding networks of knowledge brought links to new territories, facilitated the establishment of trade networks and provided access to new forms of agricultural production and spiritual practices. This gradual process resulted in the reconfiguration of the publics involved in public healing and offered a means of alleviating the pressures generated by the transition to banana cultivation. In this manner, the establishment of these shrine networks forged a powerful connection between clanship, collective well-being and the composition of knowledge.
Rather than reproducing the instrumental character of existing arguments linking clanship to resource control in Buganda, this understanding of the relationship between clan formation and the development of intensive banana farming points to how public healing served, in Feierman's words, as a ‘form of practical reason’ designed to ‘reorient [people's] situation in the world for practical ends'.Footnote 46 The communities that emerged along the northwest shores of Lake Victoria might have been born of their productive bases, but rich agricultural yields – and the opportunities for political advancement that these yields provided – were a sign of successful relationships with the appropriate spiritual entities as much as of effective farming practices. Agricultural abundance, in other words, was both a yield and a sign: it was something in and of itself and also indicated a condition of collective health. In this respect, the maintenance of collective health revolved around the successful arrangement of the skills required for a community's continued prosperity, and the composition of skilled individuals and their followers into networks of knowledge served as a marker of effective political leadership. At the heart of these endeavors stood the ideology and practices of clanship, which shaped the historical formation of publics within which Ganda pursued collective well-being. The social organization of knowledge, the spatial organization of clans and material prosperity were intimately entwined in sixteenth-century Buganda.
CONCLUSION
This essay has explored the connection between clanship, collective health and the effective composition of knowledge in sixteenth-century Buganda. Faced with the social tensions generated by the increased value of land that accompanied the development of land-intensive banana farming, Ganda built upon recent innovations in the realm of spirit mediumship to develop dispersed therapeutic networks that reconfigured the boundaries of public healing and informed the earliest processes of clan formation. This understanding of clanship requires that we rethink the problem of scale identified, in the case of Buganda, by Christopher Wrigley. According to Wrigley, the reliance on ritual forms of authority necessarily limited the size and reach of the communities that preceded the development of states. For Wrigley, these communities, which included the many small kinglets in the Great Lakes region, ‘could not work unless most of the people were present at the rituals that were the reason for [their] existence’.Footnote 47 The dispersed clan networks described in this essay, however, illustrate the capacity for ritual practices to forge relationships among communities whose members did not necessarily share face-to-face interactions. By extending the territorial scale over which a particular spirit might ensure collective health, itinerant mediums managed the composition of communities consisting of people living in discontiguous territories. In this manner, the discourse and practices of clanship furnished a critical vehicle for the development of larger, more mobile and increasingly complex communities in the area surrounding the northwest shores of Lake Victoria.
The convergence of public healing and political authority in the distant Ganda past also compels us to reconsider the better-studied period of state formation in Buganda. Recent studies have reformulated the triumphant narrative of centralization and modernization that dominated an earlier generation of scholarship on Buganda. In their emphasis on both the decentralized nature of the Ganda economy and the manner in which the incorporation of significant numbers of non-free people affected the practice of politics during Buganda's territorial expansion in the eighteenth century, these studies have added valuable new dimensions to our understanding of Buganda's emergence and subsequent development.Footnote 48 Ganda clan histories, however, open up an entirely new territory for discourses that lie outside of dynastic narratives because, in part, they lie outside the world of royal politics. This territory includes discourses about spirit mediums, public healers and other wielders of authority on the local level and provides a novel perspective from which to view Buganda's development.
While the earliest processes of clan formation preceded the development of the Ganda state, Ganda clans and the clan concept continued to develop and transform over the course of Buganda's territorial expansion. The connection between clanship and public healing first found expression in Buganda during the transition to intensive banana farming, but Ganda would later apply the practical reasoning at the heart of this process to meet new challenges. The gradual emergence of a bureaucratic state apparatus and the militaristic nature of Buganda's territorial expansion from the late seventeenth to the early nineteenth centuries transformed deeply embedded concepts of collective well-being. Rather than effacing previous conceptions and practices, however, the state-formation process directed efforts to promote collective well-being towards a more muscular political center. These undertakings prompted transformations in clan practices and involved the creation of royal medicines, the nationalization of clan-based spiritual entities and the emergence of distinct forms of organized violence designed to meet the needs of shifting realities. The success of royal ideology throughout this process rested in its capacity to persuade the communities who came to constitute Buganda's clans to conceive of their collective well-being as intimately connected to the health and prosperity of the kingdom itself, a process that lies beyond the scope of this essay.
