Introduction^{Footnote 1}

This article presents the first attempt to understand the precolonial history of agro-pastoralism in the vast miombo or brachystegia woodland region of western Tanzania, particularly in Unyamwezi. Given the extensive literature on pastoralism and its bearing on politics and society in the Great Lakes region, it seems surprising that this phenomenon receives scant attention in an adjacent – albeit ecologically different – area. The article makes two additional interventions into the intertwined histories of East and Central Africa. First, it repositions aspects of the history of western Tanzania, specifically Unyamwezi, the extensive homeland of the Nyamwezi people, into a transregional historical context involving East and Central Africa including the Great Lakes zone. Within a Tanzanian context, the Nyamwezi are the second largest ethnic group (after the neighboring Sukuma) and were well known as inveterate travelers, traders, caravan operators, porters, elephant hunters, and migrant laborers during the long period of commercialization from the late eighteenth century into the early colonial period.^{Footnote 2} They were also migrants and colonists (in the neutral sense) par excellence, and influenced the history of large swathes of Central and East Africa from the Congo River to Zanzibar, and from Bunyoro in the north to Katanga (Shaba) in the south.^{Footnote 3} This wider perspective has not yet been utilized to examine their own domestic economy in Unyamwezi. Here influences from far afield – the great lakes kingdoms, Central Africa, East Africa, and the Indian Ocean – converged.

Second, scrutinizing agro-pastoralism in Unyamwezi, as we will see, not only extends the scope of research on this topic and region, but also involves reversals of the commonplace social, cultural and political readings of “Tutsi” pastoralism in the Great Lakes region and beyond. It is well known that in other societies in the wider region where Tutsi or Hima pastoralists lived in close proximity to more populous cultivating peoples (typically “Hutu” or “Iru”) it was the former who emerged to dominate society and kingship.^{Footnote 4} In Unyamwezi there was a marked difference: immigrant Tutsi cattle herders remained subordinate to the Nyamwezi and became clients of the Nyamwezi chiefly elite and “big men.”

Underlying everything is the theme of mobility, as people and livestock – particularly cattle – moved in and out of neighboring ecological zones, and from one state or chiefdom to another, continually constructing new networks, communities, and forms of economic, social and cultural interaction. Although the Nyamwezi and their forebears were familiar with cattle, it was only in the early nineteenth century in central and southern Unyamwezi that cattle became relatively common. At the same time the environment itself was affected as settlement patterns changed, populations grew (or diminished in some places), and larger extents of cultivation opened up more land for cattle keeping in districts made free of the trypanosome carrying tsetse fly. Late in the nineteenth century political instability in some areas led to conflict and declining populations as people moved to more secure concentrated settlements. In such localities the return of formerly cultivated areas to bushland led to tsetse infestation and consequently cattle herding became more difficult.

The overall result of these complex historical and ecological changes was the transformation of the productive economy of Unyamwezi from a concentration on cultivation and forest pursuits to agro-pastoralism. What can be termed a “great lakes” complex of specialist cattle keeping was added to the pre-existing economy based on extensive cultivation, forest pursuits, and regional and long-distance trade. Not only did people, livestock and trade goods move, but also ecological boundaries and economic activities rooted in particular environments. These changes were directly related to the larger context of commercialization and migration in nineteenth century East and Central Africa, processes in which the Nyamwezi people had central roles.

From the early nineteenth century into the colonial period and beyond the commercially successful Nyamwezi chiefdoms of western Tanzania hosted a remarkably diverse, even cosmopolitan immigrant population.^{Footnote 5} Traders, porters, settlers, and slaves came from all over Central and East Africa and the Indian Ocean region, traveling along the caravan routes of the ivory and textile trades that converged at Tabora, the main trading town in the East African interior. Their origins ranged from the eastern Congo to Zanzibar, from Buganda to modern Zambia, and from far off Oman, Kutch, and Gujarat. Among these migrants were cattle herders identified by the first European travelers in the region as “Watosi” or “Watusi,” predominantly from the neighboring states of Burundi and Buha but also from Rwanda and Karagwe.^{Footnote 6} Some were refugees from conflict related to violent state expansion, civil wars and drought, while others were encouraged to migrate by increasing population densities in their homelands or attracted by the opportunities of economic growth. The rising Nyamwezi and resident Muslim commercial elite invested a portion of their trading profits in cattle and so for the first time, especially in central and southern Unyamwezi, they found themselves owners of large herds. Thus began a new pattern of patron-client relationships, with Nyamwezi chiefs, big men, and Muslim traders adding Tutsi herders to their retinues in an increasingly complex division of labor. At the same time some ordinary Nyamwezi also accumulated cattle through income from their trading activities and caravan labor.

A great deal has been written about the significance of the category “Tutsi” in the Great Lakes region but we know relatively little about Tutsi in Unyamwezi. We will come back to the issue of the arrival of the Tutsi in Unyamwezi and make suggestions concerning the processes that brought them southeastwards from the Great Lakes region, although the question of how interlacustrine Tutsi relate to the Tutsi in Unyamwezi is beyond the scope of this article.^{Footnote 7} What we do know is that in Unyamwezi the cattle herders moved in from the northwest during historical – i.e. not mythical – times, particularly from the late eighteenth or early nineteenth century onwards, and that these Tutsi were always the junior partners and clearly subordinate to the local elite, unlike their Tutsi and Hima counterparts in Buha, Karagwe, Kagera, and Buhaya.^{Footnote 8} This was different from the situation in Buhaya, for example, where Peter Schmidt has described a process in which interactions between agriculturalists and herders led to “a transfer of rights over land in exchange for access to cattle and their manure.”^{Footnote 9} It also differed from the well-known case of Ankole and its smaller western Uganda neighbors, where for the Buhweju, for example, Justin Willis suggests “a common pattern whereby agriculturalist men worked for a while for wealthy pastoralists, non-ruling and ruling, in return for an eventual gift of cattle – often a bull, but sometimes a fertile cow.”^{Footnote 10} The history of Nyamwezi-Tutsi relations in Unyamwezi thus adds complexity to our understanding of pastoralist-agriculturalist interactions in the wider Great Lakes region. Further, it is an aspect of the emergence of new multi-ethnic communities in the East and Central African interior associated with commercialization, mobility, and the caravan system.

This article first sets out the Nyamwezi social, economic and ecological context, in which the position of livestock gradually evolved, resulting in an agro-pastoral complex by the mid-nineteenth century. In the second half of the article, I explain the arrival of Tutsi pastoralists and their integration in and contribution to this agro-pastoral complex both from a Great Lakes and from a Nyamwezi perspective. Overall I argue that the reconstruction of Tutsi mobility and agro-pastoralism in Unyamwezi not only demonstrates entanglement and transfers as such, but also draws our attention to the intertwinement with social, economic and ecological transformations, and to reversals in socio-cultural and socio-political patterns when moving from the Great Lakes to Unyamwezi.

Cattle in Nyamwezi Culture and Society

In eastern Africa the history of cattle keeping and milking is an ancient one.^{Footnote 11} Yet well after parts of East Central Africa had been settled by cattle herding southern Cushites and southern Nilotes, as well as pioneering Bantu speaking communities, most of western Tanzania likely remained the preserve of Khoisan speaking hunter-gatherers, a topic itself hardly researched. The gentle slopes and miombo savannah woodlands of this vast region remained inhospitable to livestock keepers due to widespread prevalence of the tsetse fly. Thus when the Takama forebears of the Nyamwezi moved into new areas well to the south of Lake Victoria sometime in the first few centuries A.D. they kept comparatively few cattle.^{Footnote 12} Even after Takama culture divided into its Sukuma, Wembere, and Nyamwezi groupings cattle remained fairly insignificant in the Nyamwezi economy before the second quarter of the nineteenth century, except in some northern localities such as around Nzega and Kahama, in Usumbwa, and closer to Sukuma territory.^{Footnote 13} Yet the Nyamwezi as a whole and their ancestral populations were certainly familiar with cattle, and it is not suggested here that the Tutsi first introduced cattle into Unyamwezi. The proto-Bantu root word for cattle has been retained in Kinyamwezi as moombe (from -gombe)^{Footnote 14} and for milk, mavele or maBeele, from the root for breast, *-bede, suggesting that the meaning of cow’s milk was a later change as knowledge of milking spread.^{Footnote 15}

In central Unyamwezi the chiefdom of Unyanyembe – which became the main centre of commercial activity in the region – was a tsetse infested wilderness before its founding by Nyangwila migrants from the south-east in the eighteenth century. But as more and more people settled in the area the bush was cleared for farming and the tsetse fly was pushed back to the margins. A large cultivated area was created and it thus became possible to keep domestic animals including cattle.^{Footnote 16} However, as in most parts of Unyamwezi, the social adaptations and skills necessary for agro-pastoralism were not present. Nyamwezi men typically had neither the opportunity nor the aptitude for cattle herding and in the same way dairying and trade in dairy products were marginal activities for Nyamwezi women.^{Footnote 17} Instead, from the late eighteenth century the Nyamwezi became the premier long-distance caravan operators, traders and porters of the East African interior. Expanding ivory markets in India, Europe, and North America and the demand for imported industrial goods in East and Central Africa caused massive change in Nyamwezi society and throughout much of the wider region. The impact of market forces, urbanization, the rise of a new commercial elite, new forms of politics, religious change, and cultural innovation were transformational.^{Footnote 18} The caravan trade, urbanization and Nyamwezi migration also led to a rising demand for foodstuffs, and changed the labor requirements of the local mixed farming economy. Nyamwezi women took on new responsibilities within their communities, and female expertise and labor became increasingly important for the management of agricultural production. Imported slaves were increasingly employed in agricultural and domestic labor in rural communities, larger villages, and in market centers like Tabora.^{Footnote 19} It was in this nineteenth-century context of commercial expansion, when for the first time wealth from the ivory trade and other sources was available for investment in cattle, that pastoralists from the north, locally identified as Tutsi, were attracted to Nyamwezi chiefdoms. Tutsi cattle herders with their families managed most livestock, extended the distribution of dairy products and helped develop the agro-pastoral economy of the region.^{Footnote 20} The arrangements many Nyamwezi cattle owners made with these Tutsi immigrants to care for their herds, enabled successful Nyamwezi elephant hunters, traders and porters to move further west with the expanding ivory frontier, operate caravans to the coast and elsewhere and, at the same time, invest in cattle. We will come back to the arrival of these Tutsi later, but I first present the rise and decline of cattle keeping in Unyamwezi as well as the socio-economic and ecological constraints within which an agro-pastoral complex emerged in the mid-nineteenth century.

Cattle and Environmental Change

In central Unyamwezi during the first decades of the nineteenth century the population density was low, there were few cattle, and cattle herding was relatively insignificant for overall production. In his discussion of the early economy of the region Alfred Unomah mentions as livestock “a few goats, sheep and poultry.” Only when migrants from Unyangwila and elsewhere settled the area and cut down much of the tsetse-infested bush was the human population able to increase. Success in agricultural production and commercial activity related to the ivory trade, combined with a reduction of tsetse infestation, then made investment in cattle possible, with the Nyamwezi elite and Arab traders running big herds. By the 1850s Chief Fundikira of Unyanyembe was said to have 2,000 cattle.^{Footnote 21} A similar situation almost certainly existed in the western districts of Unyamwezi inhabited by the Galaganza branch of the Nyamwezi. In 1858 explorer Richard Burton found that the westernmost Galaganza villages lacked cattle.^{Footnote 22} Twenty years later Swiss trader Phillipe Broyon wrote: “The Wagaraganza originally possessed no cattle; and this is shown by the repugnance which they exhibit to take charge of those they now possess, for they leave them, as a rule, to the care of the Watussi.”^{Footnote 23} Broyon believed that the Galaganza obtained their cattle by raiding the Sukuma. More likely they, or the local Tutsi, traded for them in Buha. During the early colonial period, at least, a cattle trade existed between the resident Tutsi in Unyamwezi and Buha and Ushirombo to the north.^{Footnote 24} To the south of Unyamwezi, in Ukimbu, there had also been very few cattle, although there may have been more by mid-century. They never became numerous except in one or two localities.^{Footnote 25}

Reports such as these and evidence collected by researchers suggest that cattle may have been more numerous during the middle decades of the nineteenth century in central and southern Unyamwezi than in earlier and later times. This argument contradicts that of Helge Kjekshus, who contends that cattle were always important in western Tanzania until the onset of colonization, an argument that flies in the face of the ecological realities. He provides numerous examples from the middle and late decades of the nineteenth century showing the presence of cattle in Unyanyembe, Msalala to the north, and western Unyamwezi, but none from earlier periods.^{Footnote 26}

My argument for a mid-century wealth in cattle is supported by further references to the period from 1830 to 1880. The Nyamwezi trader Lief bin Said reported “bullocks four for one dollar” in southern Unyamwezi in 1831. Arab travelers who left Bagamoyo in 1845 and reached Benguella in 1852 noticed much livestock in Ugalla, to the south west of Unyamwezi.^{Footnote 27} Burton commented that in 1858 the ruler of Rubuga in eastern Unyamwezi, “Sultan Maula or Maura (…) was celebrated for his wealth in cattle: the herds appeared peculiarly fat and well-grown.” The people of Tura, the first district in eastern Unyamwezi, possessed “numerous herds,” while the large village of Hanga, on the eastern border of Unyanyembe, was “rich in cattle.” Just north of Tabora John Hanning Speke noted the “small and short-horned” cows of Ulikumpura. More generally, he reported, the country between Tabora, at 5° south, and the equator “teems with cattle and goats.”^{Footnote 28} In 1878 the neighboring chiefdom of Ndala possessed “large herds of cattle.”^{Footnote 29} Cattle were also present in western Unyamwezi. I have referred to the Galaganza. In Usagozi Burton noted that cattle were plentiful enough that some were sold to travelers. In 1858, Ulyankulu, later one of the two chiefdoms together known as Urambo, the core of Mirambo’s empire, was in part “a highly cultivated flat, peculiarly rich in black cattle.”^{Footnote 30} Twenty years later Southon of the London Missionary Society (LMS) found two kinds of cattle in Urambo, one humped, along with sheep, goats, asses, geese, ducks, pigeons and poultry. Of the cattle he writes, “The oxen thrive fairly well in certain localities, but the presence of the Tsetse fly all over the country is not favorable to them. (…) King Mirambo has many large herds which in the vicinity of Urambo do remarkably well, but the cows never give large quantities of milk.” Southon’s colleague, Walter Hutley, also commented on the prevalence of tsetse in western Unyamwezi at this time.^{Footnote 31} The problem was a perpetual one. According to Burton, the Arabs had described the tsetse as “so fatal that cattle attacked by it are at once killed and eaten before they become carrion from the venomous effects.”^{Footnote 32} Yet its effects were not felt everywhere. Almost at the same time that Southon and Hutley reported widespread tsetse infestation in the west, W. Griffith recorded a “few tsetse on the road but not in the cultivated district.” In Usagari, close to Urambo, he found “numerous herds of cattle and milk even at midday – [a] well watered district.”^{Footnote 33} This was at the height of the dry season. We can see, then, that ecological boundaries were not as fixed or clear-cut as has often been assumed. Not only did people and cattle move, but also tsetse fly infestation as a result of changes in land use, political management, and vegetation patterns.

From the 1860s the number of cattle fell in some parts of Unyamwezi, particularly in the east, but perhaps also in Urambo. The main cause seems to be the combined effects of conflict that facilitated the spread of tsetse. The Nyanyembe civil wars of the 1860s and instability along sections of the routes to the east and north of Tabora played a major role. Some districts in central Unyamwezi were abandoned during the 1860s. Again, during the widespread wars of the 1870s associated with the rise of Mirambo and his conflict with Unyanyembe, people moved from their kaya (small settlements) to larger and more defendable villages, allowing the spread of woodland, wildlife, and tsetse flies where there had been cultivated land.^{Footnote 34} The White Fathers believed that raiding by “les rougas-rougas” (rugaruga) was more significant than tsetse infestation, which they believed was exaggerated.^{Footnote 35} However instability led to increased infections from sleeping sickness as the fly returned to abandoned farmlands. Thomson of the LMS was told by Mirambo that:

Then, in the 1880s and 1890s, came rinderpest which killed approximately 90% of all cattle in many parts of Tanzania.^{Footnote 37}

The Agro-Pastoral Complex in Mid-Nineteenth-Century Unyamwezi

The mixed success of cattle keeping must be interpreted in relation to environmental and social-economic constraints. In Unyamwezi, as in other parts of East Central Africa, control over labor and the management of the often harsh environment was essential for successful cultivation and livestock keeping. This interplay of labor division and specialization, environmental control and limitations, agricultural techniques, and livestock keeping gave rise to an agro-pastoral complex in mid-nineteenth-century Unyamwezi. Nyamwezi farmers made the most of their natural and human resources, and their success is frequently attested to in early sources. Many European, Arab, and Indian travelers, missionaries, and colonial officials praised the industriousness and productivity of Nyamwezi farmers and herders.

Perhaps the first non-African to visit Unyamwezi was the Khoja merchant Musa Mzuri (Musa “the beautiful,” due to his fine silk clothing) who in about 1825 began a long association with his Nyamwezi hosts. Unyamwezi was at that time “richly cultivated,” he told Burton in 1858, when the wealth of the country was still evident.^{Footnote 38} Burton, in a well-known passage, forms an image of a satisfied peasantry, well provided for from their fields and herds.

Burton provides glimpses of Nyamwezi agro-pastoralism in specific localities along the road. Passing through eastern Unyamwezi in November 1857 during his journey from the coast, he found that the chiefdom of Rubuga was “celebrated for its milk and meat, ghee and honey.” Usaguzi, in the far west of Unyamwezi, the recipient of higher rainfall than the eastern side was, he said, “laid out in alternate seams of grassy plains, dense jungle, and fertile field. The soil is a dark vegetable humus, which bears luxuriant crops of grain, vegetables, and tobacco; honey-logs hang upon every large tree, [and] cattle are sold to travelers.”^{Footnote 40} In nearby Msene, in southern Usumbwa, Burton, who was no doubt influenced by prevailing images in Europe of tropical bountifulness and fertility,^{Footnote 41} waxed lyrical on the agricultural productivity of the locality:

Northern Unyamwezi was just as productive. In July 1858, early in the dry season and soon after harvest time Burton’s companion, Speke, was in Unyambewa, north of Unyanyembe. Here there were numerous villages, “and,” he wrote, “the country, after passing through the forest, is highly cultivated.” In Ukamba, a little to the north, he found that “the country here (…) abounds in flesh, milk, eggs, and vegetables in every variety.”^{Footnote 43} All these descriptions from the late 1850s point both to an outstanding agricultural production and to the presence of cattle and the sale of milk, ghee and meat to travelers.

Later travelers provide similar accounts. A few kilometers beyond Tabora, at Itetemia, the capital of Unyanyembe, the missionaries of the first White Fathers’ caravan found it easy to purchase milk with red same same beads. They noted how every evening the numerous herds of cattle were returned to the safety of the inhabitants’ tembes.^{Footnote 44} The districts around Unyanyembe were just as wealthy. The chiefs were “well-to-do,” and provided milk and fruit to a party from the London Missionary Society, while there was plenty of food for their porters to buy. Of western Unyamwezi between Tabora and Iselamagazi, Chief Mirambo’s capital, missionary wife Mrs. Hore writes: “The whole distance of this march, as indeed were several in Unyamwezi, was over cultivated ground.”^{Footnote 45} In 1893 Urambo, the core region of Mirambo’s empire, was described as “thickly strewn with villages.”^{Footnote 46} Broyon wrote more generally that “[t]he Wanyamwesi, although they are travelers and traders, generally remain at home at seed-time and harvest; they are for the most part good agriculturists, and for this reason provisions are always abundant in their country.”^{Footnote 47}

However, the reach of this successful agro-pastoral complex was under constant pressure and continuously needed to be reclaimed against environmental constraints. In the longue durée the borders of ecological zones in western regions of modern Tanzania were subject to change as a result of human activity. As we have seen, what were formerly preserves of hunter-gatherers became settled by Bantu speaking farmers, offshoots of the Takama culture. Even the difficult tsetse infested miombo woodlands succumbed in part to the interests of the new cattle keeping big men – elephant hunters, traders, caravan operators, and farmers. The movements of both cattle and people broke down older ecological patterns and created connections across ecological zones as well as political units and cultures during the nineteenth century. Yet despite the growing importance of cattle, perhaps it was crop selection and agricultural techniques, both shaped by environmental conditions, that ultimately better define the interaction between people and nature in the region, and help distinguish the Nyamwezi agro-pastoral complex from systems of cultivation in the neighboring Great Lakes zone.

Unyamwezi is part of the central plateau of Tanzania, with most land lying at an altitude of a little above or below 1,300 meters.^{Footnote 48} The rolling country is in places broken by high granite outcrops, but more usually consists of a succession of low ranges, shallow valleys, and broad plains. Beyond densely populated central Unyamwezi, settled areas alternated with strips of woodland or forest from five to twenty or more kilometers wide. The most popular grain crops in nineteenth-century Unyamwezi were pennisetum (bulrush millet, bubelee: Kinyamwezi) and the numerous varieties of sorghum (“Guinea corn,” busiga: Kinyamwezi), both of African origin. Pennisetum has the enormous advantage of producing a good crop in soils too poor for any other grain and is drought resistant. Sorghum was preferred in areas of heavier soils, such as marshland (mbuga). Although not as resistant to pests and disease as pennisetum, it was more reliable than maize, which was nevertheless entering many parts of western Tanzania from the coast during the nineteenth century. Maize had the advantage of higher yields than the African grains, and was suitable as an early crop and for intercropping with pennisetum or sorghum. According to most sources it remained a minor crop in this period, although in 1880 Ebenezer Southon of the London Missionary Society gave it third ranking in importance after millet and sorghum.^{Footnote 49}

The Nyamwezi, along with the neighboring peoples of the “cultivation steppe,” traditionally used grain-growing techniques well suited to their environment. Among these were the practices of shifting cultivation with crop rotation, intercropping, ridging, and green manuring. Only the first will be discussed here given its special relevance to cattle herding.

Shifting cultivation or “slash-and-burn” agriculture was widespread in the miombo woodlands of western and southern Tanzania. Shifting cultivation has many variants, and in Tanzania it was not the single widespread system assumed by most early European travelers.^{Footnote 50} In Unyamwezi the preferred fallow was just two or three years because fallow land, often used for grazing stock, became eroded if left longer as the animals wore down tracks around the resurgent bush. A short fallow had special advantages in the more heavily populated areas of central Unyamwezi where more extensive types of shifting cultivation were not practical. The alternative was a very long fallow, protected from cattle, to allow the soil to recover. The limitation was the declining fertility of the soil, and crops were rotated according to the soil’s condition. “This method,” Rounce writes of shifting cultivation in the miombo woodlands, “is in fact, a type of rotation, as the reversion to weed, grass and bush can be compared with the grass fallow in the European system.”^{Footnote 51} A typical rotation might be cassava, sweet potatoes and legumes in the first year, the last mentioned important for their role in maintaining nitrogen in the soil, followed by pennisetum or sorghum in the second. When the soil was exhausted it was left fallow for grazing.^{Footnote 52}

According to some colonial authorities the most significant technical weakness of Nyamwezi agriculture was the failure to systematically use cattle manure to improve fertility, although ashes from bush clearings and stubble were applied.^{Footnote 53} This is perhaps evidence for the comparatively late introduction of cattle and the short period during which cattle had been important prior to the German conquest in the early 1890s. It could also point to a decline in the use of agro-pastoral techniques after the devastations of rinderpest, the disruptions of colonial conquest and later the requisition of large numbers of cattle during the East African campaign of the First World War. But in late precolonial times, when the above described agro-pastoral complex impressed contemporary travelers, it was usually Tutsi cattle herders and their households who cared for Nyamwezi cattle and distributed the dairy products of the herds.

Mobility and Settlement of the Tutsi in East and Central Africa

In the Nyamwezi context of caravan trade and urbanization, complex human-environmental interactions, agricultural developments, and growing cattle herds, the arrival of the Tutsi seems unsurprising. However, their structural usefulness is not a sufficient explanation for why they migrated to Unyamwezi, nor does it explain the position of the Tutsi in Nyamwezi society – two questions which will be addressed in the remainder of this article.

Dating the arrival of Tutsi cattle herders in Unyamwezi is difficult given that virtually no previous research has been undertaken on them in this part of Tanzania. However, the movements of migrants from Burundi, Buha, Rwanda, and Karagwe – whether Tutsi or not – can be situated within two wider historical processes connecting Central and East Africa, with the Great Lakes region as the pivot. Here I exclude processes of commercialization which are discussed elsewhere in this article. First, Tutsi migrations eastward must be placed firmly within the context of migration out of core regions of Burundi, Rwanda, and neighboring states through gradual settlement of new territories – what Chrétien refers to as the gradual “filling up” of various parts of the Great Lakes region – state expansion, or the results of internal conflict and crisis.^{Footnote 54} Thus, in the case of Rwanda, even if the evidence is largely circumstantial, movements into Buha and Unyamwezi can in part be linked in timing and causation to numerous episodes of migration and expansionism in all directions. There were Rwandan migrations to the west and north including to the Lake Kivu area as well as into neighboring parts of eastern Congo. In areas adjacent to Lake Kivu this was followed by state expansion.^{Footnote 55} Those to the east were connected to the war against Karagwe in 1843 and the subjugation of Gisaka from 1860 to 1867, as well as many smaller-scale movements.^{Footnote 56} In the case of Karagwe, the rise of king Rumanyika with the assistance of Muslim traders from the east coast must have stimulated movements of Tutsi pastoralists into western Tanzania.^{Footnote 57} In Burundi, expansion by both state and non-state actors during the long reigns of Ntare Rugamba and Mwezi Gisabo continued into frontier regions, including into northern Buha. Once again, there were multiple factors behind these movements including the effects of population growth and land clearing, the mobility of chiefs, and military campaigns.^{Footnote 58}

Second, and with regard to Burundi in particular, there is a clear connection between mobility and the long-distance trade in salt. Many Rundi frequently visited the Uvinza salt springs near the confluence of the Rutshigi and Malagarasi Rivers on the southern edge of Buha, as did the Nyamwezi. Uvinza was a major source of salt and the trade linked the whole Lake Tanganyika region, including what is today western Tanzania, Burundi, and the eastern Congo. Uvinza also lay along the main east-west caravan route linking Tabora (and the Indian Ocean coast) to Lake Tanganyika and the eastern Congo, and was very well known to the Nyamwezi. The salt springs were a meeting place for many different peoples and a node in networks of mobility and migration and at the same time a further link connecting East and Central Africa.^{Footnote 59} Burton, in 1858, noted that quality salt from Uvinza “finds its way throughout the heart of Africa, supplying the tribes of both the Tanganyika and the Nyanza Lakes.” This was confirmed by other travelers, who also noted that Vinza salt was available in Burundi and “over much of the eastern Congo.”^{Footnote 60} Nyamwezi traders took salt, most likely from Uvinza, as far as Karagwe in order to buy ivory from Bunyoro and Ankole. Caravan porters from chiefdoms in western Unyamwezi also used salt from Uvinza to buy hoes to equip themselves for the long journey to the coast.^{Footnote 61} Large numbers of Rundi and Ha, among other peoples, came Uvinza to boil brine from the springs, make clay pots, and pack the salt into 25 kilogram loads. Others brought trade goods including livestock (cattle, sheep, goats), butter, grain, honey, and iron goods to exchange for prepared salt.^{Footnote 62} The participation of traveling traders and hawkers from southern Burundi is supported by traditions collected by Jean-Pierre Chrétien and Michelle Wagner.^{Footnote 63} In this context of mobility and connectedness throughout the wider region, we can also situate Tutsi migration. East and Central Africans in the nineteenth century were often more mobile than standard histories suggest, movements were complex and multidirectional, and there could be a myriad of routes by which small groups of Tutsi entered Unyamwezi.

The history of neighboring Buha is relatively well understood thanks to the work of James L. Brain, G.C.K. Gwassa, J.F. Mbwiliza, Michelle Wagner, and others.^{Footnote 64} Given what we know about the arrival of Tutsi immigrants in Buha it is possible to extrapolate and make some tentative suggestions for Unyamwezi. From about the late seventeenth century land feuds in the well-watered Buha highlands became common. This was the situation when Tutsi immigrants from Rwanda and Burundi entered Buha and established themselves through their control over cattle, formerly scarce in this region. Over time the Tutsi, along with wealthier sections of the Ha agriculturalists, emerged as a superior social strata. Tutsi immigration increased during the eighteenth and early nineteenth centuries as a result of conflict and state expansion in Burundi, Rwanda, and Karagwe, land shortages, and periodic famine. At the same time the existing population of Buha grew, shifting cultivation became less viable, and disputes over land increased. Tutsi immigration exacerbated the problems of access to good highland land. A complex situation emerged with changes in the division of labor and social structure, leading ultimately to the emergence of what Gwassa and Mbwiliza term “feudal relations,” a concept that has been, it must be said, misapplied to the situation in the Great Lakes region and western Tanzania.^{Footnote 65} Nevertheless, this process can to some extent be imagined as the export of Great Lakes social and economic patterns into a new region in East Central Africa.^{Footnote 66} Gwassa and Mbwiliza stress, however, that the Tutsi came in small numbers and were accepted by the agriculturalists because of the productive possibilities that they brought with them through their cattle.^{Footnote 67} These included the manuring of peasant lands, as the agriculturalists needed to intensify their land use and hence its fertility. Mbwiliza describes the economic relationship between the pastoralists and the agriculturalists:

During the first decades of the nineteenth century there was internal conflict and the main kingdoms including Heru splintered. From this time the Tutsi looked for further support from the Ha agriculturalists, cattle were redistributed to non-Tutsi cultivators, and a greater degree of mixing between Ha and Tutsi occurred.^{Footnote 69}

The Arrival and Integration of the Tutsi in Unyamwezi

Given the above, we can imagine that a trickle of Tutsi migrants from Buha entered the western and northern chiefdoms of Unyamwezi during the late eighteenth and early nineteenth centuries, in some cases through relationships initiated at the salt springs in Uvinza, or through other trading partnerships or migrations. With the break-up of the Heru kingdom, instability in Buha, and a loosening of Ha dependence on the Tutsi through their trading activities and increased access to cattle, many Tutsi probably took advantage of better opportunities in the successful trading chiefdoms of Unyamwezi further to the southeast. Another possibility is that cattleless Tutsi or those owning just a few beasts offered their services to Nyamwezi cattle owners or were invited into Nyamwezi chiefdoms to care for cattle obtained from other sources, perhaps Usukuma.^{Footnote 70} It is known from later in the century that Sukuma caravans periodically brought cattle to Tabora for sale.^{Footnote 71} Ambitious but poor Tutsi would then have had the opportunity to build up herds of their own, as well as care for the cattle of Nyamwezi chiefs, traders and porters. There is a hint of such a situation in southern Unyamwezi in the early 1880s: “Humpbacked cattle are never kept by the Mnyamwezi (…) but only by the immigrant Watusi,” Reichard writes. “The cattle owning Mnyamwezi always hands over his animals to a Mtusi, who possesses only a few of his own cattle.”^{Footnote 72} As we have seen, this situation differed from that in Buha, where the social position of immigrant Tutsi was based on control over their own cattle.

An alternative history concerning the arrival of the Tutsi refers to them as “slaves” of the Nyamwezi. According to an elderly informant, Mzee Shabani Kapiga, from Uyui chiefdom, north of Unyanyembe,

No doubt a subordinate relationship of a patron-client type is being described, despite the references to the Manyema, becoming “the child of that house,” and the perhaps exaggerated size of the accumulated herd.^{Footnote 74} Interestingly, Mzee Shabani Kapiga suggests that later Tutsi arrivals became clients of those Tutsi who had already established themselves in Unyamwezi and built up large cattle herds. This may have been a further pull factor encouraging mobility.

Around 1880, and referring to the larger Nyamwezi cultural region, Southon wrote: “The Watusi are scattered all over Unyamwezi and though occasionally owning a few head of cattle, they are generally found in the capacity of herdsmen to chiefs who own cattle.”^{Footnote 75} The numbers were never great, however, even almost eighty years later. Figures from the 1957 census, although far removed in time, may have validity because the Tutsi did not often intermarry with the Nyamwezi. At that time 7 percent of the population of Tabora district declared themselves as Tutsi, 2 percent of Nzega district, and 2 percent of Kahama district. The lower percentages for Nzega and Kahama, in the north of the region, might be understood in the context of the deeper tradition of Nyamwezi cattle keeping in those districts. No doubt a proportion of these figures represents further immigration during German and British colonial rule.^{Footnote 76}

Pastoral activities were further stimulated by processes of urbanization around political capitals, the growth of new market towns, and the demand for food supplies due to increased caravan traffic. In the new urban centers of Unyamwezi, notably at Tabora, the resident population included Arab, Swahili, and Indian traders and, from the late 1870s, Christian missionaries. Missionaries and Nyamwezi and Muslim traders brought with them or established large households of employees and retainers, each numbering from about twenty to perhaps two or three hundred people, including askari, servants, employees, clients, concubines and slaves.^{Footnote 77} The populations of each town included local people drawn by ties of marriage, clientage, or economic advantage, as well as slaves and freed slaves of heterogeneous origins. In Tabora, Manyema from the Congo of both free and slave origins settled close to the market, the main center for the caravan business, while White Fathers at nearby Kipalapala collected a retinue of redeemed slave children.^{Footnote 78} As Coquery-Vidrovitch notes, the emergence of such slave and ex-slave communities at the heart of many of the new urban centers paradoxically helped create a degree of “openness” in which kin and other social relationships could be recreated in a heterogeneous and multiethnic environment.^{Footnote 79} Tutsi immigrants were a part of these new communities. They settled close to the capital villages of the commercially active Nyamwezi chiefs and also attached themselves to the households of the rising class of Nyamwezi and Muslim traders and estate holders, the vbandevba, who were backed by followers armed with muskets, in Urambo, Tabora, and other important chiefdoms and new towns.^{Footnote 80}

Tutsi herders were certainly present in Unyanyembe and Msene in 1858. In the latter place, once an important caravan and market center but now long forgotten, there was “a large floating population of (…) Watosi, and fugitives from Uhha.” In both places they lived, according to Burton, by selling cattle, milk, and butter.^{Footnote 81} But their relationship with local Nyamwezi and Sumbwa was more complex than this and clearly it was of an entirely different nature than the situation in Buha. Tutsi diviners played a significant role in Nyamwezi life, and had a reputation for skill in settling disputes. Such roles may have been useful to the Nyamwezi elite. In this regard, it is very likely that Tutsi herders had brought the bacwezi possession cult with them from the north-west – northern Buha, Usumbwa, Buzinza, and Buhaya – considerably earlier.^{Footnote 82}

However, it is clear that the Tutsi in Unyamwezi never formed a significant element of the ruling class, let alone a dominant one, despite some claims to the contrary.^{Footnote 83} The kinds of relationships based on tribute and cattle redistribution usual in Buha could not develop in Unyamwezi, especially in Unyanyembe, where the vbandevba was established. There, any chance of Tutsi power was snuffed out when chief Mnywa Sele had some killed and others driven out of the chiefdom in about 1859.^{Footnote 84}

Despite Mnywa Sele’s ruthless actions the Tutsi remained in Unyanyembe. In early 1861 Speke and his fellow explorer James Augustus Grant were guests of Musa Mzuri in Tabora and Grant recorded various details about Musa’s Tutsi clients in his journal. These were cattle herders from Karagwe who kept half the milk produced by his herd with Musa retaining the other half. The “Karagwa [sic] people,” Grant noted, “live on nothing else but milk from morning till night: [they] are robust, muscular ‘fit for war’ men – he [Musa] has them here for his cows. They live in (…) circular huts.”^{Footnote 85} Later Grant noted that Musa had:

He also described how Musa’s Tutsi did the milking twice a day at 8 am and 7 pm:

Livingstone also noted the presence of Tutsi in 1872, commenting: “The Batusi are the cattle-herds of all this Unyanyembe region.” Cameron made a similar observation.^{Footnote 88} Their habitations were ubiquitous around Tabora and Kwihara where the milk, butter and meat production of the area was largely in their hands. In 1890 Tutsi still cared for the cattle in Unyanyembe.^{Footnote 89} Elsewhere, in eastern Unyamwezi in 1891 there was a Tutsi settlement near a chiefdom’s capital village.^{Footnote 90} In northern Unyamwezi numerous Tutsi were in Ukune in 1861, and there was a Tutsi village in Uyui in 1881.^{Footnote 91} The Tutsi presence was significant in western Unyamwezi, as already noted, an area strongly involved in the caravan trade. In Uyowa Hutley noticed Tutsi bomas close to the cattle enclosures and separate from Nyamwezi villages. The large herds of cattle at Mirambo’s ikulu (capital), the spoils of war, were in the care of Tutsi herdsmen under their leader Mkalya.^{Footnote 92} Ernest Cambier of the Belgian Assocation Internationale Africaine reported that these Tutsi had followed Mirambo’s rugaruga to Urambo when the latter had captured livestock under their care. They were, in his words, “voluntary slaves.”^{Footnote 93} During the German period there were still “a lot” of Tutsi, originally from Buha, in Urambo and especially nearby Usoke. They were described by missionary Löbner as “cattle breeders” and did not mix much with the Nyamwezi population, maintaining their own language, customs, and distinct culture.^{Footnote 94} These social and cultural distinctions further highlight the contrasting position of the Tutsi in Unyamwezi compared with the Great Lakes region, where there was no difference of language or culture.

“A Great Social Failure”?

Oral sources as well as evidence from the early British period are also suggestive about the relationship between the Tutsi and the Nyamwezi, although there are dangers reading back into the late precolonial period accounts derived from oral testimony, vulnerable as it is to the deceptions of human memory, or derived from the colonial archive, tainted as it was by “tribal” ideology and the Hamitic thesis. However, the detailed account by Bibi Kalunde Kongogo, herself a former chief, points to the characteristic patron-client relationship with the Nyamwezi elite retaining power:

The situation described by Bibi Kalunde was not so different from the kinds of patron-client relationships established several decades earlier.

In this colonial description from the 1920s of the Nyanyembe Tutsi, British obsession with rank, class and “tribe” is clear. There are also interesting comments on the lingering slave trade in the wider region:

“Nyarigwa” was no doubt the same “Nyaligwa” described by Bibi Kalunde very much in the terms of a supplicant, no doubt before his rise to leadership of the Tabora Tutsi.

The District Book also contains information of a less ideological type. Counts of two cattle herds, probably from the late 1920s (no date is given), suggest similar ownership arrangements to those of precolonial times but indicate that it was more complicated than merely a division between a Nyamwezi patron and Tutsi client. In one herd of forty six including calves thirteen head belonged to Nyamwezi, two to “Swahili,” three to “other natives,” and twenty eight to the Tutsi herdsman. A second herd of twenty six was divided into ten belonging to Wanyamwezi, six belonging to a “Tabora town native” and ten to the Tutsi herder. Thus Tutsi herders were entering into relationships with a variety of cattle owners, perhaps by doing so reducing their dependence on a single patron.^{Footnote 97} This suggests, along with the evidence for at least some mixed marriages, an increasing tendency for inclusion and connectedness in wider society in the region.