African Studies Quarterly

Volume 11, Issue 1
Fall 2009

Social Organization and Social Status in Nineteenth and Twentieth Century Rukwa, Tanzania

Print/Download PDF Version

Abstract: Nineteenth century histories of Tanzania typically focus on "tribal" histories, customs, and military action. To a certain extent, this is expected. The story of how interior Tanzania came in contact with the Indian Ocean World is an exceedingly violent one. However, there are different ways of looking at interior history which highlight factors besides "tribal" histories. The story told here of Rukwa Region highlights alliances, status hierarchy, and fighting during the second half of the twentieth century. Such institutions emerged out of an "ecology of fear" which resulted in the re-organization of peoples, trade networks, and the emergence of a strong separation between common people and powerful rulers from different status groups even though they may have spoken the same language and had the same "tribal" affiliation. The fears generated by the clash of such institutions often shaped local responses to rapid social change.

This essay highlights what this re-organization meant for what is roughly Rukwa Region of western Tanzania in the late nineteenth and early twentieth century. Focus is on the peoples of the Fipa Plateau, and Rukwa Plains. Traditionally, these people are referred to as the Fipa, Pimbwe, Bende, Kimbu, and Konongo people. The Gongwe, a group previously not described in the anthropological or linguistic literature is also discussed.


Tanzanian history in part emerged from the accounts of nineteenth century explorers, and the needs of the colonial and post-colonial states.As such, history of Rukwa in western Tanzania, emerges from the needs of these literate intruders. Thus in the 1870s, explorers like Burton, Stanley, Livingstone, and Thomson told the story of conflict that hounded their movement across a social environment dominated by violent and charismatic figures.Each chief, they wrote, sought an advantage in the newly arrived world markets of ivory and slaves.German colonialists and missionaries arriving after the 1880s, told stories emphasizing the cruelty of chiefs to the local people as they sought to legitimize their own rule, and identify administrative units which could be adapted to colonial domination, taxable trade, and church building.The British arriving in Rukwa in 1920, took up where the Germans left off.Intent on imposing indirect rule in a fashion which would facilitate trade, they organized their east African peoples as "tribes," each having a language, and a chiefly lineage through whom the colonial power could rule.

Finally after 1961, the independent Tanganyika (later Tanzania) government looked to pre-colonial history to re-establish the legitimacy of African rule in the twentieth and twenty-first centuries.In this context, nineteenth century African leaders like Chief Mirambo of the Nyamwezi, are presented as figures resisting colonial intrusion.More problematic for post-colonial government and historiography has been the question of what to do about the "tribal" classification inherited from the British."Tribal" identities run counter to nationalist conceptions of a Tanzania.

This paper discusses Rukwa's history in the context of what I call "the ecology of fear" created by the reorganization of society in the context of the intrusion from the Indian Ocean world.This is done by evaluating the interests, capabilities, and reorganizations undertaken by the various "status groups" found in Rukwa during this time.

The New Nineteenth Century Trading System and the Ecology of Fear

Rukwa—or for that matter Tanzania in general—is not only understood as a retelling of tribal histories, but as a story of how subsistence societies organized and reorganized themselves relative to the outside world.As Iliffe notes, "Early nineteenth-century Tanganyika was not inhabited by discrete compact, and identifiable tribes, each with a distinct territory, language, culture, and political system," although the "need to describe makes the use of such collective names inescapable, even though they distort and oversimplify a vastly more complex reality."[1] This period is important, too, since as Illife notes further, it was during "the nineteenth century [that] Tanganyika's inland peoples made contact with the outside world through a long-distance trading system based on Zanzibar," and as a result, "Tanganyika experienced a transformation more intense than any other region of tropical Africa at that time."[2]

It is my thesis that reorganization in Rukwa happened in the context of an ecology of fear featuring frequent raiding, trade in new products like guns, slaves, and ivory, and the decimation of human populations by violence, disease, and famine.This "ecology of fear" is written into the modern landscape in the form of parks and conservation areas established by the modern Tanzanian government, in places first abandoned by humans fleeing violence in the nineteenth century.The ecology of fear typically pushed people living in the region to rely on apparently older forms of social organization, as they were confronted with new challenges created by intruders whether from southern Africa, or the Indian Ocean Coast.Notably, this happened not only in Rukwa, but in other regions of Tanzania as well.[3]

Social Status as an Analytical Category and the "Terrible Dilemma" of Peasants

Keeping in mind that, as Iliffe noted, collective tribal names are both imprecise, but also inescapable, I would like to experiment using Max Weber's term "status group" to describe social relations in nineteenth century Tanzania.[4] Status groups are normally communities rooted in shared honor, i.e. a shared style of life.Status groups can be dominant or subordinated in a wider social system. Weber developed this term in the context of his understandings of how a wide range of societies in Asia, North America, and Europe developed and changed in ancient and modern times.This essay will, I believe, make it clear that thinking about Africa in this fashion is also useful.

According to Weber, status groups (Staende) are rooted in an effective claim to esteem vis a vis others.Status groups share a style of life, formal education, putative hereditary lineage, and occupational status, among a range of other conditions, which in turn result in group-based action.[5] Sometimes claims result in a group of people who together assert rights to privileges such as trading monopolies, or even sovereignty over territory. Status groups seek to have privileges/honors become hereditary, and seek legal guarantees.In the context of nineteenth century Tanzania, salient positively privileged status groups includes chiefly lineages, inhabitants of a particular locality, clans, followers of a charismatic leader, rainmakers, blacksmiths, mystics and religious teachers, and shared national/tribal identity.Weber goes on to note, however, that such privileges imply status groups which are negatively privileged and in the African context include a variety of outsiders: forest people, hunter-gatherers, refugees, potters, and other outcasts.Status for negatively privileged is still expressed by pride in a particular style of life, putative hereditary lineages, occupational rights, etc., and a disdain for the trickery of the dominant group.The advantage of Weber's approach is that it reflects beliefs about stratification rooted in a range of unequal relationships, occupational monopolies, and so forth, not just tribal identity.[6]

As I think will become clear here, reframing pre-colonial society as one composed of status groups is particularly useful when evaluating social change in pre-colonial Tanzania.[7] Among the higher ranking status groups asserting monopolies over powerful symbols were the chiefly lineages who dominated through charismatic means; groups of traders who developed unique economic relations with local leaders; and European missionaries who inserted themselves into leadership.Beneath such high status groups were the sedentary farming, pastoral, warrior, and even slave status groups who while often victimized by the more powerful, still maneuvered in the same social environment.[8]

This paper re-evaluates what happened in what is (roughly) today's Rukwa Region.[9] What is striking from this analysis is that across "ethnic groups," "tribes," "occupational stratification," and status groups there were consistent caste-like relationships.The early Arab and European intruders arriving in the nineteenth century necessarily integrated themselves into this pre-existing status system, responding and reacting to a complex social world, even as their presence changed it. The stories Europeans wrote of nineteenth and twentieth century Rukwa thus are ones of violent chiefs, and violent intruders.But the vast majority of people were not the perpetrators of violence, but subsistence farmers growing the grains which not only sustained themselves, but also the chiefs and their retainers who ruled through their own armies and courts, making the accumulation and trade in ivory possible.[10] In this respect Rukwa was not unusual; it was the agricultural production of the peasants that underpinned the world of marauding chiefs, their armies, and their courts.This is because settled subsistence farming with its granaries and ripening fields of crops, by its very nature, makes farmers vulnerable to appropriation.Chirot calls this situation the "terrible dilemma" in which agricultural peoples are forced to make a choice between sacrificing individual freedom and control of the production of their labor on the one hand and ceding control to aristocratic castes for the security from raids that emerges from a powerful protector who also in return levies military drafts and appropriates foodstores on the other.[11] The implicit exchange is that the fruits of their agricultural labor are exchanged for the protection provided by armies in an otherwise anarchic world.This was very much the world of nineteenth century Rukwa.[12]


Rukwa Region is in western Tanzania and occupies four ecological zones.First, there is a coastal strip along Lake Tanganyika (elevation 768 meters).Secondly the Fipa Plateau at approximately 1500 meters is the center of the Region, and currently the most densely populated area.To the west and at the foot of an escarpment, is the Rukwa Plain at 900 meters.The Plain begins at the saline Lake Rukwa which drains an area to the north extending to what is now Katavi National Park.Lake Rukwa itself is a geographical anomaly; fluctuating water levels are recorded during historical times.[13] North of Katavi National Park is a fourth area, a vast miombo forest.

In the nineteenth century on the Fipa Plateau where there was higher rainfall, a cultural complex organized around horticulture and pastoralism developed.[14] In the Rukwa Plain, where rainfall is noticeably less, human ecology was focused by a mixture of grain cultivation (millet and maize), small animal pastoralism, hunting, and gathering.Today, there are large ungulates, including the world's larges herd of cape buffalo, in the area which is protected as Katavi National Park.Only in 1970 were cattle introduced by Sukuma-speaking pastoralists migrating from the north.[15] Beans, which require more rainfall, are not generally grown in the Rukwa Valley.The combination of horticulture, hunting, and gathering occured also further to the north in the miombo forests.

Disease and Population Ecology in the Nineteenth Century

The "diseases of civilization" like smallpox, measles and typhoid arrived in Rukwa during the nineteenth century, causing epidemics with high mortality well into the twentieth century.Rinderpest, a disease of cloven-hoofed animals also decimated herds of cattle on the Fipa Plateau in the 1890s, as well as wild buffalo and other cloven hoofed animals on the Rukwa Plain.The rinderpest epidemics contributed to human famine in the 1890s, particularly in areas dependent on pastoralism.

Population maps published in 1907 indicated that densities were highest along the Lake Tanganyika littoral, and in the Rukwa Plains.[16] Densities on the Fipa Plateau (which today are the most densely populated), and in the forested interior were low, perhaps the result of the rinderpest induced famine in the 1890s, and the violence and epidemics of previous decades. By the late twentieth century the density patterns reversed, with higher densities on the Fipa Plateau, while lowlands were given over to unpopulated game reserves.

<Traditional Tribal and Linguistic Divisions of Rukwa

Rukwa Region is at the intersection of three Bantu language groups: Nyamwezi-Sukuma from the northeast (including Nyamewezi, Konongo and Kimbu languages), Bende from across Lake Tanganyika (including Bende, Tongwe, and Holoholo), and Mwika/Fipa which has its origins in Zambia (including Fipa, Pimbwe, and Rungwa).[17] In addition, Gongwe is a language which has lexical similarities to Pimbwe, and many loan words from Bende.[18]

Today, patterned multi-lingualism between the sub-groups, including widespread use of Swahili, is common.For example, Pimbwe are also likely to speak Fipa or Konongo.[19] Status identity, today as in the past, is a function of where a person is living, identification with a particular leader, and at-home language use although an all-embracing Tanzanian identity is today very important. Frequent inter-marriage and multi-lingualism meant that particular individuals may assert membership in more than one ethnic group.Linguist Yuko Abe has undertaken preliminary lexical studies which reflect these relationships (see Table 1).

Bende speaking groups are found in the northern part of Rukwa, and include Bende, Tongwe, and Holoholo speaking people.[20] This is the most sparsely settled part of the Region, and the only place never organized into formal chiefdoms and incipient states.[21] Nor did the Bende build defensive pallisaded villages. As a defensive strategy they retreated into the miombo forest, abandoning planted fields to the raiders.[22]

A Gongwe-speaking population persists today in parts of Mpanda District, and maintains a royal graveyard which is at the site of the old royal village, in what is now Katavi National Park. The Gongwe are not mentioned in the language databases of the Summer Institute of Linguistics, nor in the ethnographic literature.However, they are mentioned in British colonial documents describing the dissolution of the Gongwe court in 1927-1928.[23] Joseph Thomson mentioned them in 1880, and they appear on German maps.[24] As with other groups living today in Rukwa, the Gongwe practice dryland maize cultivation, small animal pastoralism, hunting, and gathering.[25] They have a tradition of chieftainship.[26]

19th Century Status Groups: Rulers, Soldiers, Traders, and Missionaries

So far, I used the traditional tribal and linguistic divisions to describe Rukwa because they are the status labels used in the traditional ethnographic literature, especially British colonial records.Indeed, it was along such status distinctions that opportunities to rule (and be ruled) were distributed by the British between 1920 and 1961.[27] However, in Rukwa other status distinctions were important; and when they are used, a more nuanced description of nineteenth century Tanzanian social worlds emerges.

In this spirit, I highlight status groups found in western Tanzania which patterned strongly feelings of loyalty, and made the pursuit of political and military power coherent. Notably these status groups went beyond simple feelings of kinship based on land and tribe, and included aristocracies, language groups, and occupational groups. Criteria included groups with educational, professional, linguistic, ritual, and kin-based criteria for membership. Each status group transcended loyalties rooted solely in economic or kin-based interests.

In nineteenth century Rukwa, relevant statuses included child soldiers/pages known as ruga-ruga and who were typically kept close to the person of the chief, and in European terms might be thought of as being the equivalent of pages or retainers in an aristocratic court. [28] As for the leadership itself, chiefship was rooted in aristocratic ranks which spread across geographical boundaries of "tribes," and were often rooted in claims of foreign origin, which, as Tambila notes, were "a ploy to acquire a greater measure of legitimacy" in order to rule effectively.[29] In Rukwa, the putative origins of such groups were the chiefly clans of the Tutsi with assumed origins in the regions near Rwanda and Burundi, or Nyamwezi clans from near Tabora.[30] Swahili speaking "Arab" traders formed a status group.The Catholic order of the White Fathers, which arrived in the 1880s, also in effect brought a status category for themselves from Europe, which, as is described here, developed in a fashion similar to other local aristocratic ranks.


The ruga-ruga were boys who were separated from their families at a young age, and raised to be the personal retainers/soldiers for a chief.[31] They were raised in chiefly courts to be loyal to their commander.The status was acquired as a result of initiation rituals, not birth. [32] Shorter describes these boys as . wild young men, a heterogeneous collection of war captives, deserters from caravans, runaway slaves, and others.without roots or family ties, and they owed no allegiance other than to their chief or leader.

He also notes that:

Nearly every Nyamwezi chief had [ruga-ruga] during the last quarter of the nineteenth century.[they were] standing armies of young, unmarried professional soldiers, especially trained to fight and terrorize their enemies.they often wore mutilated parts of the bodies of their enemies as ornaments, and their name is said to derive from this fact.Many of them wore belts made of human entrails and necklaces of human teeth.The ruga-ruga were encouraged to smoke Indian hemp to make them fearless and excitable.Their weapons were bows and arrows, spears, and muzzle-loading guns.[33]

Nyamwezi and Tutsi Ruling Castes [34]

The term Nyamwezi applies to a variety of status groups which are found in Tanzania today, and in the past.Today, the term refers speakers of a language living in the Tabora area.In the nineteenth century, the term was also applied to low status occupational status group, including the porters who carried goods from the interior to Bagamoyo on the Indian Ocean coast on behalf of Arab traders.Also, nineteenth century explorer accounts of Rukwa describe these chiefs of the Pimbwe, Kimbu, and Konongo as being "Nyamwezi."Indeed, Bennett even notes that there was a Nyamwezi chief in what is now Katanga Region of the southern Congo.[35] Well-known chiefs in Rukwa identified as Nyamwezi included Mirambo of the Nyamwezi, Simba of the Konongo, Nyungu ya Mawe of the Kimbu, and Kasogera of the Pimbwe.[36]

The term Tutsi today refers to the traditional ruling castes of Burundi and Rwanda.[37] There was also a Tutsi ruling class among the Waha and Wahangaza of Tanzania.[38] And as Willis points out, the Tutsi chiefly caste in Fipa has an origin myth indicating that they migrated from the north from the area of what is now Rwanda/Tanzania/Uganda.[39] Each share cultural characteristics including a focus on cattle herding, chieftainship traditions, and royal regalia.As Willis notes, however, it is not clear whether this is the result of an actual migration from the north, or cultural diffusion. [40] For the purposes of this paper, though, this does not matter.The point is that there is a chiefly caste in Rukwa that has putative origins in migration and symbolically identifies with other leadership castes ranging from Uganda, to the Fipa Plateau.The Tuutsi status groups of Rukwa monopolized the symbols of leadership before the nineteenth century.[41]

The chiefly castes of Rukwa—both the Nyamwezi and Tuutsi—controlled routes passing through their territories by the 1860s.This became more lucrative as the Arab-generated trade in ivory, guns, and cloth increased.[42] Control meant that chiefs and their ruga-ruga monopolized trading rights and demanded payments from the caravans in the form of guns, goods, and other supplies with which they could strengthen their own military position.Implicit to this ideal was the sale of the supplies that the caravans needed to feed and care for porters, slaves, and others.In this context, chiefly succession in Rukwa was negotiated in the context of alliances (and enmities) which spread across western Tanzania.[43]

Arab Traders as a Status Group in Rukwa

Zanzibari Arab traders were resident in the chiefly courts of the Fipa Plateau, Gongwe, and Simba at the time Europeans arrived in the 1850s-1870s.[44] They were involved with the practical governance of the area via the ruling chiefs.In many respects they served a function similar to that of merchant minorities such as Chinese in Southeast Asia, Lebanese in West Africa, or Jews in parts of Europe.[45] In this context, they often became the financiers and advisors to royal courts.For this reason perhaps, European explorers accustomed to observing the phenomenon elsewhere in the world, described the Arab courtiers as being "ministers" or "prime ministers" at the fortress at Karema, Simba's, and the Fipa Plateau.Whether this formally corresponds with such a modern bureaucratic category is improbable.After all, too often Arabs had their own ruga-ruga to guard the caravans.What is clear is that Arab traders by the nineteenth century established themselves as a separate status group with legitimated monopolies over various aspects of trade and governance in the area of Rukwa.Tambila regards the success of Arab newcomers as . proof of the fact that expanded trade opened possibilities to more people than established chiefs alone.But this access was also a source of instability because the rising class of merchant/hunter warlords with their ruga-rugas and hangers-on were potential power sources.[46]

European White Fathers as a Status Group in Rukwa

The Catholic "Order of Africa" (known as the "White Fathers" due to the color of their cassocks) established a mission station at Karema on the shore of Lake Tanganyika in 1885 after purchasing it from King Leopold's International African Association (IAA).The immediate goal of the White Fathers, many of whom were French, and all of whom were ritually initiated into a religious order, was to establish a mission station to spread the Christian gospel.

In the context of a dangerous world though, the mission station at Karema was far more than a church: it also provided a home for escaped and redeemed slaves, proselytisation, and the maintenance of a secure place for visiting Europeans.Necessarily, the priests fortified the station against military attack.Guns inherited from the departed IAA proved particularly useful for the White Fathers' "small military" when seeking compliance from nearby Bende villages, whether it was for protecting territory, housing redeemed and escaped slaves, attacking uncooperative villagers, or mounting defenses against potential raiders.[47] The White Fathers found recruits for their "military" among the escaped slaves they sought to protect, refugees, and famine victims, as indeed the Nyamwezi chieftains in the region also did.In the process, the mission at Karema became a "state unto itself," and began to negotiate with neighbors in much the same fashion as the other chiefly figure in the region. [48] For example, when the White Fathers negotiated with Chief Msuulwa of Nkansi/Fipa, they were admitted to his court on "equal status with Arabs and other Muslims."Towards the end of the nineteenth century, the White Fathers turned the tables, though, and began demanding political homage from the leaders of the Fipa Plateau.[49]

In 1893, the German colonial military arrived, which enhanced the role of the White Fathers in Rukwa.Understaffed German authorities delegated many of the responsibilities for governance, most importantly that of judging crimes, to the French-speaking White Fathers.The White Fathers in turn used this authority to further their own goals for the elimination of witchcraft (especially trial by poison ordeal), proselytization, agricultural improvement, and the establishment of churches and schools.

As Smythe in particular points out, the White Fathers were the most consistent foreign presence in a remote region where the outside political powers changed frequently.[50] Between 1870 and 1970, political shifts included that of indigenous chiefs, to German military officers, British colonial officers, and finally the independent Tanzanian government.The presence of foreign White Fathers did not dissipate until the 1980s.While few European White Fathers remain today, the church they established remains strong and influential.

Pre-Colonial Rukwa: Ngoni Raiders, Arab Traders, and "Little Wars"

The first Ngoni warriors invaded southern Tanzania in the late 1830s, setting the stage for an extremely violent nineteenth century.The Ngoni are typically described as having an origin in southern Africa.Using infantry techniques developed by Zulu warriors, they made their way into what is now Tanzania in the 1830s from bases in Malawi.As Koponen notes, though, presumably as with other marauders in nineteenth century Tanzania, the Ngoni (also called Watuta) were warriors who had origins in many places.[51] Indeed, Koponen believes that only a few hundred of the 16,000-20,000 invading Ngoni in the early nineteenth century actually originated in South Africa.As with ruga-ruga, most Ngoni were refugees, former slaves, conquered peoples, adventurers and others who had joined for any number of reasons as the invaders pushed northward.[52]

Using superior infantry formations, especially stabbing swords, the Ngoni entered Rukwa in the early 1840s, deposed the chiefs of the Fipa Plateau, and took the available cattle back to "Ungoni" to the south, probably Malawi.[53] The Ngoni invasion into Tanzania probably went as far north as Biharamulo near the southern shore of Lake Victoria over the next decade, but then receded.The Ngoni returned to Rukwa a second time in the 1850s; the English explorer Richard Burton passed through Tabora and northern Rukwa in 1858 and described the devastation such attacks left: "The route before us lay through a howling wilderness, once populous and fertile, but now laid waste by the fierce Watuta [Ngoni]."[54] Shorter makes the argument that the Ngoni activity stimulated surviving Africans to organize more systematically a defensive and offensive military, perhaps giving birth to ruga-ruga traditions.[55] He further asserts that it was at this time that surviving chiefdoms began to re-organize themselves into pallisaded fortresses to protect themselves against further plunder.

The next intruders into the region were Zanzibari Arab traders who reached Unyanyembe/Tabora in Central Tanzania by the 1830s and Ujiji shortly thereafter in their search for slaves and ivory.The purpose and organization of the Arab traders created a fundamentally different status group; they were not interested in plunder, but in the protection, staffing, and provisioning of trading caravans which would walk from central Africa (today's Tanzania and Congo) to Zanzibar on the Indian Ocean Coast.Trade was focused by world demands for ivory and for slaves for plantations on the Indian Ocean littoral. Trade goods purchased included guns, beads (which Burton and Cameron described as being used as currency), and cloth.[56] Guns in particular re-oriented power relations.

Trading caravans in nineteenth century Africa were difficult and expensive to organize, because few areas were secure from raids by locals.Travel was by caravan which carried not only trade goods, but the arms and provisions needed to feed the porters, protect ivory, and prevent slaves from escaping.[57] Explorers paid tributary passage fees to chiefs who controlling stretches of the road. Two safe spots for the caravans were the fortified cities at Tabora/Kazeh, and Ujiji, which soon became notorious as slave depots during the mid-nineteenth century.[58]

Oral and written accounts indicate that by the 1850s, the chiefs of southern Rukwa, occasionally purchased slaves, traded foodstuffs to caravans, and most importantly, traded ivory for guns.[59] Arab traders settled in the region semi-permanently as hunters and traders, typically while acknowledging the sovereignty of a local chief.The good news for Rukwa was that there was no indication that peoples were systematically victimized by slave raiders in the manner that the Congo was; Zanzibari Arab presence in the region was instead focused on the trade in ivory, a function of the large elephant herds in the Rukwa Valley.[60]

The "Little Wars" Shift Status Arrangements (1860-1898)

Locals refer to the period between about 1860 and 1898—i.e. between the time of the Ngoni invasions, and the arrival of the German colonial army—as that of the "little wars" between chiefs.Shetler writing about the Serengeti where similar events occurred calls it the "Time of Disasters" when violence and famine dominated relationships, and a wilderness was created by human depopulation.[61] The accounts of locals as well as explorer journals from Rukwa reflect this view well.Richard Burton's 1858 travel journals focused on the violence he found in Rukwa, most of it stemming from the invasion of the Ngoni.Stanley's descriptions of Rukwa in 1870 also focused on the violence he observed.He also saw plentiful signs of buffalo, elephant, and rhinoceros, reflecting perhaps an abundance of forest abandoned by human horticulturalists terrorized during the previous decade.[62]

In response, fortified towns were built which served both defensive and offensive military purposes.These fortresses as is the case of horticultural empires elsewhere, were towns within walls which protected the power of the chief, including wealth in the form of ivory and foodstuffs.Privileged residents typically included the chief sorcerer, administrative chief, keeper of the royal regalia, the chief's wives, and the dwellings of the ruga-ruga.[63] The queen mother typically also had an honored dwelling, reflecting a common division of authority in much of central Africa.[64] Typically such fortresses also had a dry moat, and a palisade of logs.At least one of the moats described by Shorter was apparently filled with sharpened poisoned stakes.[65]

Such fortresses were a place for loyal vassals to retreat in the event of attack.But they were also a place for powerful chiefs to consolidate power by providing a location to concentrate soldiers, protect foodstores, store ivory, and hold court.Isolation of the chief from the common people added to the mystery and majesty of the chiefly position, and were symbolic elements used to assert authority.[66]

During the "little wars" alliances emerged and enmities developed between local chiefs.[67] Venturing unprotected into the forest to hunt animals, tend crops, or establish homesteads put men at risk.The net result in Rukwa was the emergence of frontier areas in which human population was sparse due to the dangers from enemies.The most important of these areas was probably between the Pimbwe, Gongwe, and Konongo in what is now Katavi National Park.Each grouprecalls accounts of attacks by the other.Perhaps the most vivid was an account of how Chief Kasogera's mother was impaled on a post after being skinned alive by Gongwe hunters, presumably in the late nineteenth century.[68]

A type of fluid power developed in the context of chiefs, Arabs traders, European intruders, and little wars is illustrated well by a document created by the elephant hunter Swahili Matumula from the Indian Ocean coast, with a representative of King Leopold of the Belgians, at the inland compound owned by the Nyamwezi/Konongo chief Simba in 1877. The agreement is a transfer of sovereignty for Karema, a lakeport traditionally occupied by the Bende-speaking clans, but which became a slave depot for the Swahili from Kilwa, Matumula, a few years previously "by right of conquest."The agreement reveals a great deal about the status stratification between the locals of the region, who were only recently "conquered" by Matumula's ruga-ruga, the role of powerful Arab traders (most of whom are nameless in the explorers' writings), and the European representative of King Leopold of the Belgians.Most importantly it was negotiated in the presence of Simba, the patron of Matumula.[69] But also what is not written is significant.Mirambo, who also asserted sovereignty over the area if not by right of conquest, on the basis of his reputation for ferocity, is missing from the agreement:

I Matumula a native of Kilwa [on the Indian Ocean Coast], landowner by right of conquest for five years past of Karema Territory.I give to the Sultan of the Belgians and to his subjects, the part they will choose, on the said Territory belonging to me, where they may build and cultivate.

I deny myself and refuse to my successors the right to send them away in future or to molest them.If they are attacked I shall defend them by force of arms and if they give way we shall die together.

The Sultan of the Belgians [i.e. King Leopold] and his subject will have an absolute sovereignty over that part of the Territory given away, the boundary of which will extend one mile in radius around the spot on which they will build their first settlement.

This act has been drawn up in the presence of Matumula, Cambier, a subject of the sultan of Belgians, Alijmasi and Mournie, written at Simba's on Friday the Fifth day of Ramadhan in the year 1298 [1877 AD] of the Heijira.[70]

This agreement reflects a mix of European, Arab, and African concepts of sovereignty, as well as bureaucratic rule.The Belgian was Joseph Cambier who negotiated on behalf of King Leopold of the Belgians, who was establishing the private companies which would bring devastation to the Congo after the Berlin Treaty in 1884.[71] As for Matumula, he apparently considered himself a vassal of the Nyamwezi Simba, rather than the other Nyamwezi Chief Mirambo, who was an enemy of both.

The Fortresses of Late Nineteenth Century Rukwa

Pallisaded fortresses, or "royal villages," dominated social organization in Rukwa and nearby areas during the nineteenth century (see cartographic essay).Such fortresses had dry moats, and wooden pallisades requiring the mobilization of large numbers of people to construct and maintain.In the case of the fortresses at both Pimbwe (in the Rukwa Valley) and Kisuumba (on the Fipa Plateau) (See Figure 1), they were apparently about four to five hundred meters across.The chief Simba's fortress to the north was, according to Thomson, larger, and the most impressive he saw in Africa.During the violent period when the maintenance of such fortresses focused social organization, the farming population clustered ever nearer, abandoning more distant fields as indefensible.When the power of the marauding brigands, chiefs, and others dissipated in the early twentieth century, the farming population retreated back into the more remote areas.[72]

Waters Figure 1

The fortified villages themselves typically lasted only a few decades before either indigenous raiders, or campaigning outsiders breached the walls.Nevertheless the fortresses were particularly important in the nineteenth and early twentieth century as farming peoples dealt with the changes associated with the invasion of Ngoni from the south, and the arrival of the first Zanzibari-Arab trading caravans bringing ivory, slaves, cloth, and firearms after the 1830s and 1840s.

Cameron in 1873 described a fortress built by who he called "Kimbu refugees" as a strong stockade in which the gates closed and guns and spears protruding through the stockade by which it was surrounded.The huts were flat-roofed and built in the form of long parallelograms, the whole being surrounded by a heavy stockade with only two entrances.Over each of these was a sort of crow's nest, where the defenders of the gate took up their position and were furnished with a supply of large stones to be used on the attacking party.[73]

This demand for military-like organization and engineering required large work parties, which in turn required leadership rooted in a capacity to command obedience through dispatch of ruga-ruga, control of the rains, and spiritual life.

Rukwa's Relations with Mirambo and Beyond

As the diffusion of political traditions with roots in Rwanda, Burundi, and Tabora demonstrate, Rukwa was never isolated from outside influence.There are in fact cultural clines and political traditions stretching across central Africa.In the nineteenth century, the best known political force in the region was the chief Mirambo, scion of a Nyamwezi clan from west of Tabora.Mirambo commanded an army of ruga-ruga who between the 1860s and his death in 1884, operated in an area between Lakes Victoria and Tanganyika.[74] Mirambo did this by negotiating alliances against the Arab traders of Tabora, with the Europeans who arrived in the 1870s, and 1880s, as well as chiefs ranging from the Fipa Plateau to Mwanza.[75]

Mirambo of course was not the only chief seeking fortune in the context of ruga-ruga, aristocratic ruling castes, Arab traders, and European explorers, although for a short time he was perhaps the most successful at doing so.Others in Rukwa who maneuvered within this system included Nyamwezi personalities like Nyungu ya Mawe (in Kimbu) and Kasogera (in Pimbwe); various Tuutsi chiefs on the Fipa Plateau; the Arab merchant minority; European White Fathers; and King Leopold's IAA. Chiefs in Rukwa had at least indirect relations with groups in today's Congo, Zambia, Burundi, Central Tanzania, and ultimately the Indian Ocean World and Europe.Their feudalistic system—which also had elements of anarchy—was not eclipsed until the German colonialists asserted bureaucratic control in the early 1900s.In many ways, the system persisting today in Rukwa reflects this history.

European Traders Arrive: the International Africa Association, the Lakeport of Karema, and the Rigidification of New Status Relations

The first of the pallisaded forts Stanley encountered was the small "Konongo" fortress called Mrera which he wrote controlled three or four villages.There were nine bleached human skulls at the entrance to this encampment, the result of feuds Stanley wrote, between Konongo and Wazavira.[76] They were armed with muskets, presumably obtained from the coastal trade system.[77] The next day, though, Stanley remarks upon the desolation of the villages that they passed through. The chief "Simba" he wrote had desolated these villages, and the inhabitants fled, leaving gardens for Stanley's caravan to plunder.

Joseph Cambier of the IAA was the next journal-keeping European to arrive in Rukwa, in 1877.As with previous visitors, he observed more warfare, and pointed out that the Nyamwezi Chief Mirambo was dispatching ruga-ruga to demand villages of Rukwa pay him tribute.Cambier proceeded to purchase the fortress at Karema from Matumula as described above, displeasing Mirambo who claimed the right to tributary payments from the region, despite its distance from Tabora.Bende homesteads outside Karema were attacked by Mirambo's ruga-ruga, raising anxiety within the IAA fortress.

It was in this violent context that King Leopold's IAA sent four Indian elephants to Karema for use in the Congo territories.Only one lonely elephant actually made it to Karema, and as a result, two Englishmen in Leopold's employ, Carter and Cadenhead, returned to the Indian Ocean Coast.[78] To avoid fighting between Mirambo and the Arabs around Tabora, they took a more southerly route through Kasogera's fortress at Pimbwe.Mirambo, allied with Simba, had in the meantime raided onto the Fipa Plateau, and swung back to attack the fortress at Maji Moto as well.

Allied with Mfundo, a former chief of Pimbwe/Maji Moto, Mirambo organized an attack on Kasogera's chiefly compound.[79] Kasogera in turn forced Carter, Cadenhead, and their armed caravan into the compound to assist in the defense of the fortress.But when the combined forces of Simba and Mirambo attacked, Kasogera's forces retreated.Only Carter and Cadenhead remained, and were killed by Simba/Mirambo's soldiers.[80]

The Court at Fipa-Nkansi and the Arrival of Europeans

The Fipa Plateau also experienced a period of wars in the 1870s and 1880s which created new alignments.The "kingdom" itself was split in two, Lyangalile and Nkansi, under the command of chiefs Kapuufi and Kimalaunga respectively.Kimalaunga became especially well-known for his capacity to deliver violence as he and his ruga-ruga sought a part of the ivory trade which emerged in the Rukwa Valley.He also apparently raided Lyangalile for cattle on a regular basis.In doing this, he displaced the nomadic Nyika, a group long in conflict with horticulturalists and pastoralists on the Fipa Plateau, and established alliances with groups of Nyamwezi.[81] According to the Nkansi chronicle as cited by Willis, in the early 1870s:

Kimalaunga set himself up as ruler of the Lyangalile part of Rukwa and the people declared themselves for him.Kapuufi sent for help to Kiyungi, king of the Nyamwezi [in Tabora].The king [Kiyungi] sent many soldiers under a man called Mwaana Katwe.But this force was defeated by Kimalauunga and fled back home.Kapuufi then sent elephant tusks as a peace offering to Kimalauunga, but he responded by killing the bearers.Soon after this the Nyika rose against Kimalauunga and defeated him.The Nyika routed Kimalaunga again in great battle in which he brought Kwa, Kuulwe, Wanda, and Konongo to his aid, but without avail.[82]

As described above, the young English explorer Joseph Thomson arrived in Rukwa in 1880 during the maneuvering between the local chiefs including Simba (Konongo/Nyamwezi), Mirambo (Nyamwezi), Nyungu ya Mawe (Kimbu/Nyamwezi) Kasogera (Mpimbwe/Nyamwezi), Kapuufi (Fipa/Tuutsi), and Kimalauunga (Fipa/Independent).More so than earlier explorers, Thomson had an eye for ethnographic detail, leaving more complete descriptions of the fortresses as he maneuvered his expedition in the region.On the Fipa Plateau, he described the compound of chief Kapuufi.Thomson's approach to the chiefly compound was negotiated by Arab intermediaries from Kapuufi's court at Kisuumba/Nkansi, who only invited him into the royal village after the caravan spent a cold night on the plain outside the village:

Early in the morning, after breakfast a royal messenger arrived, with the intelligence that his master was ready to give me an audience.Proceeding to the town, with my usual guard for state occasions, we passed a fine herd of cattle.Entering the town by the strongly fortified gate, we found ourselves a perfect labyrinth of inner bomas, or pallisaded quarters.Crossing over a variety of dunghills and filthy cesspools, which indicated that the cattle passed the night within the royal precincts, we reached Kapufi's [sic] palace.It differed from the other bomas only in size.[83]

Thomson goes on to describe the women of Kapuufi's court who he reports were all "plump and fat," but he said showed no signs of idleness. [84] He observed the weaving of cloth (presumably from locally grown cotton), pounding of skins, and preparation of food.[85] Willis emphasizes that the people of the Fipa Plateau at that time were prosperous, and able to support large herds of cattle—which of course was what attracted Kimalauunga's raiders in the first place.

Thomson then visited the Gongwe "village" in the Rukwa valley a few days later.At the time that he visited in May 1880, he reported that Gongwe was an independent village, and owed feudal tribute only to the Arab governor of Unyanyembe (Tabora) not the neighboring chiefs like Kasogera, Kapuufi, or for that matter any of the "Nyamwezi" chiefs in the area like Nyungu ya Mawe, Simba, or Mirambo.Thomson using a variety of status-based terms described the population as being a mixed one of Wangwana (free men), Wapimbwe, Wakhonongo, and Wanyamwezi, rather than as an independent chiefdom. [86]

But most impressive to Thomson was Simba's town where he arrived a few days later.He described it as covering an area that was three-quarters of a square mile, and had large squares created by the construction of houses which had their doors facing inwards.Inside and outside the squares were "ordinary native houses."[87]

Simba personally met Thomson. This is the only written description that is left of this important leader, although it is apparent that he did occasionally meet with other Europeans like Cambier.Thomson was particularly impressed with the low quality of Simba's dress, and the fact that he moved within his village with a certain level of anonymity.Simba told Thomson that he was a classificatory brother of the Nyamwezi chieftains Mnywa Sele (Nyamwezi Governor of Tabora), and the Kimbu chief Nyungu ya Mawe (Kimbu/Nyamwezi). [88]

Irrespective of Simba's power, by1880 Mirambo sought to re-establish influence in Rukwa; indeed, this was not to be the end of Mirambo's depredations.Using the Carter and Cadenhead affair as a pretense to turn the tables on his erstwhile ally Simba, Mirambo infiltrated soldiers posing as refugees into Simba's village.These soldiers in turn opened a breach in the wall, which was used to sack the compound.[89] And so the town Thomson only a few months earlier described as "the largest I have seen in Africa" became a forgotten ruin, known not even to the local historians I queried in 2004.

Simba disappears from European accounts after 1880, and Nyungu ya Mawe and Mirambo died in 1884.The point of this is to emphasize that Rukwa bled severely since at least the time the Ngoni first arrived in the 1840s until the 1880s.The violence was intensified by the fact that no one group, leader, clan, tribe, or status group dominated politically or militarily.Rather, every putative chief sought political advantage by commanding followers who risked annihilation in their violent world, be they African, Arab, or European.In this respect, the weak European mission and trading settlements of about 1890 were not that different than the traders like Matumula (the elephant hunter), or other chiefs like Mirambo, Simba, Kasogera, Nyungu ya Mawe, or Kapuufi, all of whom sought some level of hegemony using violence.There was no peace for decades, as each sought advantage from inside walled-villages.Nor was there an easy peace for the intruders; every mission, hunting camp, explorer station, or trading post was built with an eye to military advantage.This was a logical response in a chaotic world in which refugees, freed slaves, and ruga-ruga bands were all common.[90]

Pax Germanica, Pax Brittanica, and Pax Tanzanica in a Remote Corner of the World

The Imposition of German Rule

It was in the context of the ongoing "little wars" that the German Empire first arrived in Rukwa in 1893 as a military force.In remote Rukwa, colonial hegemony meant acknowledgement of German sovereignty by accepting flags, payment of in-kind taxes to the German military government, providing soldiers to the new power, and for the chiefs, wearing German uniforms.In short, from the perspective of the subsistence farmer paying tribute, relationships to the new German power was similar to what had happened in Rukwa under the various marauding chiefs.

Demonstrations of brutality were also part of the German strategy.In Rukwa, the Germans attacked the weaker points, especially the unfortified Bende in their scattered villages, who typically responded by abandoning fields, and retreating into the forest.[91] Sakalilo Village, the fortress of the renegade Chief Kimalaunga near the shores of Lake Rukwa, was destroyed by German troops armed with Maxim guns and cannon in 1893. [92] Exhausted by their own fighting, and in awe of German firepower, the more powerful chiefs grudgingly gave in to the imposition of German rule.Chiefs of Fipa Plateau, Pimbwe and elsewhere, offered fealty to the representatives of the German emperor, provided labor to build roads, sent soldiers, and submitted serious disputes to German arbitration. Those who did not do so, were subject to whipping, fines, imprisonment, and execution.[93] For example, Kimalauunga who had terrorized the Fipa Plateau for a decade or more during the "little wars," and was widely believed to have super-natural powers, was finally imprisoned by the Germans, and then shot dead while "trying to escape" in 1899. The Germans tried to use his death to symbolic advantage:

He was [then] removed from his chains and his head cut off.The next morning the Wafipa were called to bury the body but they refused, expecting to see the body at any moment turn into a lion and attack them.The other prisoners were then made, much against their will, to dig a grave and bury the corpse.[94]

A second chief on the Fipa Plateau, Yuulamaasi of Lyangalile was also arrested by the Germans for stealing cattle, but was simply fined and released.He too died shortly thereafter, however, and was succeeded by Kuundawanantu, who after a few years of cooperation with the Germans fled to the bush to resist and to re-establish the traditional rights of chiefs.The Germans responded by holding his family members hostage, and mounting an aggressive search.After fleeing northward, Kuundawanantu surrendered, two to three months later.The Germans, after ordering the attendance of all chiefs from the Fipa Plateau, publicly hanged him.Willis considers this to be the actual end of any independent states on the Fipa Plateau.[95] German power was harsh and brutal, but its ubiquity also freed the people of the incessant warfare between the chiefly castes.

For the historian, the arrival of the Germans introduced the advantage—or perhaps crutch—of written records.Nevertheless, examining the nature of the conflicts emerging between the traditional rulers, the occupying Germans, and after 1916 the British, provides insight into what happened and how African leaders responded.Shorter writes particularly effectively about how active the German and British used their power to pick chiefs of the Kimbu in neighboring Mbeya.[96] The same happened across Rukwa as lineage-focused British administrators sought rulers who were both legitimate and compliant, and who could rule their "tribes."[97]

In the slow assertion of control, the Bende clans provided the greatest challenge.They were victims (successively) of armies raised by Simba, Mirambo, Fipa, White Fathers, and Germans.The Bende clans never responded by creating their own fortresses, however, but rather by retreating further into the bush, and with non-cooperation.This confounded the British who, thirty years after the arrival of German colonial power, still found it difficult to collect taxes from the Bende clans.[98]

Perhaps illustrative of difficulties in political incorporation are the clashes three Pimbwe leaders, Kalulu, Ngomayarufu, and Nsokolo had with the European colonialists over boundaries and ritual rights between about 1900 and 1940.Each resisted and cooperated with the Germans or British in attempts to maintain their own chiefly authority.But, their story illustrates how the chiefs' political power dissipated between about 1900 and the 1940s.

The Decline of the Pimbwe Chiefs in the Early Twentieth Century

The pallisaded fortress of the Pimbwe was built around a hot spring known today by the Swahili name "Maji Moto."Pimbwe tradition traces the origin of the settlement using a king-list that has approximately 23 chiefs.[99] Kalulu of Pimbwe, a young boy, became chief at the cusp of the German intrusion, probably in the early 1890s.[100] Succession though was handled as a regional affair requiring the approval of the Tuutsi chief at Nkansi on the Fipa Plateau.The Tuutsi chief in turn referred the case to the newly arrived Germans at Bismarkburg.As the scion of an older branch (Mfundo) of the Pimbwe royal family that pre-dated the Nyamwezi Kasogera, Kalulu was awarded the formal chiefship over another claimant, Kasogera's grandson Ngomayarufu.So while Kalulu prevailed at the hearing, the Germans also ordered the Chief of Nkansi to divide the country between the two claimants.[101] Ngomayarufu objected to this and was sentenced to two months imprisonment, and his claim to the throne discarded, at least temporarily.

Kalulu quickly became well-known for his aggressiveness towards the colonial powers, and used his command of ruga-ruga and trade routes to assert authority.Nevertheless, his fortunes were reduced "some years later" (presumably in the late 1890s) when he refused a German requisition of soldiers for a punitive expedition against the Bende to the north.Ngomayarufu complied with the request, and was rewarded by the Germans with the assignment of ten villages slightly to the east of Kalulu, probably about 1900.[102]

As for Kalulu he was subsequently arrested after . he was told by a witchdoctor named Mwana Kalembe that he must never shave his head without rubbing it over first with a human heart.Kalulu sent out two of his ruga-ruga who killed Wamikamba (Jumbe of Mikamba) and then cut out his heart.Word of this reached Kasanga [Bismarkburg].Kalulu was arrested, sentenced to a long term of imprisonment and to fifty lashes in two installments of twenty-five.The ruga-ruga were beaten.After his release Kalulu returned to Pimbwe, and only left when the White Fathers established a mission there.[103]

Other records indicate that the White Fathers established their mission at Pimbwe/Maji Moto in 1907, which is the first datable event since the departure of Thomson in 1880.The White Fathers built their mission station inside the Pimbwe/Maji Moto village walls, but soon had a falling out with Kalulu after a sacred stone was destroyed, a grove of trees around the royal graves was cut down, and the villagers restricted from using the hot spring.As a result

Kalulu took objection to the presence of the mission and gave it out that he would cause the spring to dry up. (The Mission admits this and says Kalulu was ordered by the Germans to move to a village four hours away). Kalulu had certain rites carried out by the witch-doctor Mwana Kalembe.The following year (1910) there was an earthquake and the spring dried up.Kalulu said he would return when the mission had left, which they were soon forced to do.Two years later there was another earthquake, the spring was re-opened and Kalulu returned to [P]imbwe [where he reasserted control of the villages to the east of Ngomayarufu].[104]

The British occupied German Bismarkburg on the shore of Lake Tanganyika in 1916 during World War I, which included Pimbwe/Maji Moto in its jurisdiction.Ever the schemer, Ngomayarufu requested permission from the British to build a new village.Kalulu quickly reported the fact that the new village was in his territory, and the British promised to send "the bwana" to investigate the claim.[105] The District Officer did not arrive, and Kalulu responded by ordering the new village burned.In response, the British deposed Kalulu, fined him, deported him from the area, and installed a more compliant Ngomayarufu as chief of the entire country.Ngomayarufu subsequently died in 1923, precipitating yet another succession crisis by a young alcoholic "half-wit" Zunda, who was reportedly addicted to marijuana (bhangi).Zunda displeased the British, who received repeated complaints from the villagers about his drinking and other "immoral habits."The British were relieved then when he suddenly died in 1928; it was rumored that his death was by poisoning.[106]

As for the pallisaded fortresses, the last were abandoned in 1927 under British programs to re-settle villagers into sleeping sickness settlements.[107] By this time, the pretense to chiefly rule dissipated further as the British staffed their district offices with European colonial officers.While chiefs were still occasionally borne on palanquins by courtiers, more and more they became creatures of British rule, subservient to the District Officers sitting in distant District headquarters.A final Pimbwe revolt of sorts occurred in 1944 when Chief Nsokolo, his court officers, and ruga-ruga prepared a stew using a sacrificed boy mixed with sheep meat, which in turn was fed to the villagers at a feast.Human sacrifice had long been assumed to be a common means for the Pimbwe royal lineage to gain mystical power, and Nsokolo and his sorcerer felt that this was an appropriate way to regain power they felt was slipping away. Villagers weary of such behavior, reported the incident to the British District Commissioner, who arrested Nsokolo and his court. They were then transferred to the jail in Kasanga, and eventually sent to Tabora to be tried and hung.[108] At least in 2001 and 2004 when I conducted interviews, the villagers remembered the execution with a great deal of approval—the rule of ancient chiefs was viewed as part of a cruel past and was little lamented.[109]

The Twentieth Century Demise of the Chiefly Rule, and Chiefly Caste

Marking the demise of an old phenomenon like the construction of pallisaded fortresses, chiefly rule, and chiefly caste is inherently difficult.Nevertheless, the archival and oral records provide some indication of how this happened in Rukwa.

The "little wars" which apparently began in the context of first Ngoni raids, and the penetration of Arab traders, probably ended in the 1890s in the context of declining ivory trade caused by the depletion of elephant, declining human populations as a result of famine, violence, and disease, and the establishment of the German military station at Bismarkburg.Most importantly, the alliance of German military and French White Fathers produced a level of political stability rooted in their capacity to assert what Max Weber calls the monopoly over the legitimated use of violence in a particular territory.Violent though this monopoly was, the rapid destruction of villages and execution of dissident chiefs, whether in Rukwa, or in more distant Iringa, during the Maji Maji rebellion (1904-1906), precipitated acquiescence of chiefs to German sovereignty.At least in the general picture, ruga-ruga were controlled.[110]

This "stability" of course eventually evolved into domination from British centers of power after they acquired sovereignty in Tanganyika in 1920.But, not until 1927 did the British have enough authority that they could order the concentration of the population of the Rukwa Valley into new village sites under policies emphasizing public health (especially sleeping sickness control), road construction, agricultural production, and wildlife conservation.Conservation became particularly important in Rukwa as an emerging system of parks and reserves in which wildlife and forests were established, and farming excluded.[111] As for the remaining Gongwe, whose traditional settlements were located in the heart of the new conservation area, they were dispossessed of their rights to conduct court trials by the British, and dispersed.The Pimbwe court was moved away from the fortress at Maji Moto, where it was dominated by the exiled Kalulu's followers, to Usevia where the successors of Ngomayarufu, including Nsokolo, taxed, collected, and judged on behalf of the British authorities.The remaining chiefly courts of the Fipa Plateau meanwhile followed the British to the administrative capital of Sumbawanga, abandoning the chief Kapuufi's fortress-village that Thomson described at Kisuumba-Nkansi.Pallisaded villages were no longer protected from bush fires, and the remnants of the walls and gates were destroyed by fire and the elements.When Aylward Shorter surveyed Kimbu for such sites in the 1960s, all that generally remained were the remnants of ditches.[112] Only in one place did he find a structure which, it was claimed, was the remnants of a gate.

The symbols of powerful independent chieftainships such as the ruga-ruga, right to conduct trials, installation ceremonies, etc., of course continued for some time after the disappearance of actual local rule.But the symbols were just that—symbolic in the context of a more powerful British colonial government.If there was any pretense of independent royal prerogative in the region, this ended with the hanging Chief Nsokolo and his courtiers in 1944.In turn, bureaucratic control was consolidated by the British, and passed to an independent Tanganyika in 1961.Chiefs at that time were given the option of cooperating with the new government, or confining themselves to ritual activity.While a few like Chief Nsalamba of the Konongo were elected to the first Parliament of independent Tanganyika, this in fact provided only a minor token of the local ritual authority their grandfathers previously had.

At the beginning of the twenty-first century, Rukwa is still remote from the bustling markets of the Indian Ocean world and the services offered by the central government.[113] Farmers are still by and large embedded in a subsistence lifestyle in which they grow what they eat, build their own houses, and have large numbers of children.Only minor vestiges of ruga-ruga and chiefly rule remain.So while lives focused by subsistence agriculture are similar, Rukwa is notably more peaceful then during the nineteenth century when every petty chief maintained ruga-ruga, epidemic disease stalked the land, and the human population declined as a result.Instead, since about 1950, the population has expanded rapidly, pushing farmers back into areas abandoned by fearful villagers in the nineteenth century.Just how peaceful the area has become is highlighted by Willis' 1989 article "The Peace Puzzle" about the twentieth century Fipa Plateau.[114] As he notes, the Fipa became a society with exemplary traditions of dispute resolution, despite the violence of the nineteenth century.In the context of a social order—externally imposed though it may be—norms for the peaceful resolution of disputes emerged among a people who only decades before were viciously attacked by neighbors and mounted raids of their own.

The enforcement of the peace by the state has created its own paradox, though.Peace means that human farming populations can expand back into the forest abandoned by their fearful ancestors.In the meantime, however, the forests took on a new value to the independent government in Dar Es Salaam—as game parks which have value not only locally, but in the global marketplace where conservation is a potential source of revenue for a cash-strapped government.[115] In this context, fear of arrest for poaching or farming restricts access to fertile lands, even in times of famine.[116] So in twenty-first century Rukwa, a population which could not occupy forest fringes due to the ecology of fear in the nineteenth and early twentieth centuries, now cannot do so because the powerful national government which guarantees the peace also conserves the forests.

Just how tenuous this arrangement is can be seen by the resurgence of vigilante movements in Tanzania during the late twentieth and early twenty-first centuries.Called sungu sungu, in Rukwa (and elsewhere), such self-help justice organizations reflect the weaknesses that government institutions have in such areas, particularly in terms of policing.Such village justice system, contain elements of "trials by ordeal" and emerge out of the desire for order in a society threatened by anarchy.[117] Sungu sungu share the same emphasis on internal solidarity, secret ritual, and righteousness that chiefs in nineteenth century Rukwa exploited.Fortunately, in modern Tanzania the state is not so weak that such freelance militia can terrorize as widely today as they did in nineteenth century Rukwa, or for that matter in twenty-first century eastern Congo.But sungu sungu still raise the ugly possibility of the anarchy of the nineteenth century.

A Contingent History of Twenty-First Century Western Tanzania

A point of this article is to expand the historical understandings of western Tanzania beyond traditional tribal, political, and military histories which have been the focus of a number of excellent tribe-specific ethnographies such as those of Shorter (1972), Willis (1981), Tambila (1981), and Smythe (2006).In doing this, the article has emphasized the status based relationships which structured the extraordinary violence, the ecological context, and the social psychology of fear in a rapidly changing society.It did so in the context of dominant leadership status groups which crossed tribal boundaries. Such a context made it possible for the Swahili Matumura's right to sell "the country" of Karema to the Belgians in 1877, and later the important role that White Fathers played in the political governance of late nineteenth century Rukwa.These roles emerged in the context of cross-cutting leadership traditions, not tribe-specific ideologies.

Nineteenth century social change in Rukwa occurred in the context of pre-existing traditions of leadership, military activity, and tributary relationships.What was observed was that in each of the more densely populated areas, traditions of chieftainship through the rule of putative outsiders emerged.This created a social space for population movements as new status groups became more salient, including those of a merchant minority, soldiers, missionaries, peasants, and even refugees. In the absence of any central power, outsiders, whether African, Arab, or European, asserted the trappings of the independent leadership caste by creating fortresses, controlling weapons, collecting tribute, protecting followers, attacking enemies, and controlling trade.From this position they entered into negotiations with established chiefs.None though imposed a tabula rasa on the societies that they eventually dominated.

Underlying this was an implied fear, whether of neighbors, Ngoni, European, or long distance raiders.This "ecology of fear" is of course common among farming societies such as those in nineteenth century Tanzania and elsewhere who, as LeBlancwrote, lived with a constant threat of attack from the morning raid, picking off a person who leaves the village, or inviting enemies to a feast where they are simply slaughtered.[118] In the historical record of Rukwa, such violence was common, and frequent.Stealth and trickery were routine, including Simba's mistake of admitting putative refugees to his compound in 1880, Mirambo's shifting alliances, White Father raids on the Bende, the impalement of Kasogera's mother, the German sacking of Kimalauunga's capital Sakalilo, and Nsokolo's bizarre cannibalistic feast of 1944.The evidence of routine violence is evident in what early explorers like Burton, Stanley, and Livingstone observed in abandoned villages, and especially skulls on pikes.As much as traditional geographical constraints like weather on settlement patterns, this fear provided the parameters in which nineteenth and early twentieth century history unfolded.

If there was an organizing tradition to political leadership in the pre-colonial Rukwa, it was found in caste-like leadership status categories, rather than traditional tribal units rigidified by the British. The leadership complex across the region was often composed of putative outsiders—be they Tuutsi, Nyamwezi, Arab, or European.Control of local population was also through a separate subordinated status groups of young people, be they called royal pages, ruga-ruga, Nyamwezi porters, slaves, converts, or freed slaves.Common peasants were at constant risk of attack and needed the promise of refuge in one of the well-defended fortresses; undoubtedly their "ethnic" or "tribal" loyalties shifted accordingly.For this reason it is more appropriate to think of their status as being a peasant, who because of their social status, risked becoming a refugee, slave, or Catholic convert, rather than a Fipa, Pimbwe, Gongwe, or Bende.The social system of Rukwa was ultimately rooted in a need to protect crops, land, and livestock.This meant peasants were faced with what Chirot described as "the terrible dilemma" between individual freedom and security.[119] In the case of Rukwa this "ecology of fear" patterned human interaction in a way that determined the limits of social activity, as surely as rainfall and crop selection do.[120]

Much can be surmised about what the introduction of nineteenth century trade meant for the ecology of the Rukwa region in the twenty-first century.The introduction of the ivory and gun trade made possible—necessary—the maintenance of pallisaded fortresses in the more heavily populated areas.The increasing supply of meat brought about by guns may well have led to a demographic expansion in the mid-nineteenth century.There was a nastier side, however, that of the emergence of the military rivalries in which predatory groups flourished.But this success in terms of population growth, even in the absence of slave raiding, was short-lived, probably only a matter of decades.What is more, the elimination of horticultural humans and probably elephants in the late nineteenth century led to the re-expansion of the forest fringe.It also may well have led to the expansion of the large herds of animals, particularly cape buffalo, now found in the Rukwa River Valley.An open question for further investigation is what key species were in the region in the past?Has Katavi National Park always been dominated by buffalo, or is this domination a by-product of the ecology of fear of the nineteenth century?

Cartographic Essay[121]

The five cartographic representations reflect political and social changes that occurred in Rukwa Region between about 1880 and 2000.

Map 1 is of German East Africa and reflects political boundaries, settlements, and the caravan routes which were of relevance to Rukwa in about 1912. The central railway from Dar Es Salaam to Kigoma was under construction by the German colonial power at that time.

Waters Map 1

Map 2 reflects the approximate boundaries of the different chiefdoms/linguistic groups present about 1880. As described in the text, Ujijji and Tabora were Arab trade depots. The station at Gongwe paid tribute directly to Tabora. Karema was a station of the Belgian International African Association, Simba was a Nyamwezi chief with hegemony over what is now considered the Konongo area, and there were two Fipa chiefdoms (Lyangalile and Nkansi). The Nyamwezi area to the north was dominated by Chief Mirambo, while the Kimbu area was dominated by the Nyamwezi Chief Nyungu ya Mawe. Sakalilo was the fortification of the Fipa Chief Kimalaunga. The Bende area was occupied by dispersed clans who spoke Bende dialects.

Waters Map 2

Map 3 reflects activity in the region in 1904-1905, and is in part based on the first systematic cartography done in Rukwa.[122] The German established the region as the military district of Bismarkburg, with headquarters at Kasanga/Bismarkburg on the shore of Lake Tanganyika. Lake Rukwa waas inexplicably beginning to dry up, and the human population was probably declining rapidly as a result of disease, famine, and violence. The Germans and the Fipa court established the town of Sumbawanga which in turn was to become the population center for Rukwa Region. Uruwira and Karema were two of several Catholic Mission stations maintained by the White Fathers.

Waters Map 3

Map 4 reflects Rukwa Region in 1930-1948 when the British reorganized their new colonial possession. Rukwa was part of the Western Region of Tanganyika Territory, with headquarters in Tabora. Sumbawanga became the population center, and a network of gravel roads was established. Pimbwe/Maji Moto was abandoned in 1927, and the Pimbwe court moved to Usevia. Gongwe was dispersed at the same time. In the late 1940s, a railway spur to the new town of Mpanda was established in order to take advantage of the gold mine established there.Lake Rukwa's levels had risen since 1905.

Waters Map 4

Map 5 is a representation of Rukwa Region in approximately 2000. Three areas were established as either a Game Reserve (Ugalla) or National Parks (Katavi and Mahale Mountains). In these areas, settlement was prohibited by the central government. The population centers were in Sumbawanga, Namanyere, and Mpanda. Maji Moto village was re-established on the old site of Pimbwe. Two new major population concentrations established in Rukwa were the Burundian refugee settlements established at Katumba and Mishamo in 1973 and 1978 repectively. The road networks (not shown) reflected this organization.

Waters Map 5


[1] Illife 1979:8.

[2] Illife 1979:40.

[3] For example, Shettler (2007) describes similar nineteenth century violence and abandonment in what is now Serengeti National Park.

[4] Weber 1946:305-306.

[5] Ibid.

[6] In Tanzania, distinctions were in turn maintained through rituals symbolized, as Iliffe (1979:38) describes, by symbols of authority such as horns, drums, chiefly regalia, royal retinues, and in some places umbrellas (see also Shorter1972:101 for a description of the royal regalia of the Kimbu.

[7] Such an approach could also perhaps be used with Uganda.Reid (2002) uses the term "class" to describe pre-colonial status distinctions. However, as Weber (1946:146-147) points out, the term "class" presupposes the presence of a market economy, while "status group" does not.

[8] Weber explicitly refers to the dispossessed as a negatively privileged status group, albeit one marked by a lack of privileges.See Weber 1946:190-191

[9] Rukwa Region as a political sub-division of modern Tanzania is a late twentieth century appellation.Earlier in the twentieth century the region was made up of the political sub-divisions of what was earlier Bismarkburg and Ujiji (German times until 1916), Western/Tabora Region of British Tanganyika, Mbeya, Kigoma, and Tabora Regions (British Tanganyika), etc. Rukwa Region as an administrative region of independent Tanzania was not gazetted until the early 1970s with a Regional headquarters at Sumbawanga.

[10] Writing this paper has forced me to choose terms for royalty.In the literature, the word "king" is often applied to larger polities, and "chief" to smaller ones.However, usage is in fact inconsistent and in many respects creates artificial distinctions between polities which are more similar than different.In this paper, I have opted to use the word "chief" as a rough translation of the Swahili words "mtemi" (which is related to the verb "to cut") and "mwene," a term specific to interior chiefs in a number of polities of the Great Lakes Region which may or may not use Swahili.All of the polities discussed here interacted directly or indirectly, and the fortunes of particular chiefs and polities fluctuated quickly—a polity which was small one decade could become large the next, and vice versa.

[11]Chirot 1994:128.

[12] See Schoennbrunn 2006 for a discussion of violence and vulnerability in East Africa before 1800 CE.

[13] In 1939, lake levels returned to historical levels (see Kjekshus 1977:77 and Dean 1967:45).

[14] Willis 1979:30-31.

[15] Paciotti and Borgerhoff Mulder 2004.

[16] see Koponen 1994:586-587.

[17] The place name Pimbwe in Tanzanian common usage is known as Mpimbwe, and the people of the area are the Wapimbwe.In this paper, I have chosen to use the term "Pimbwe" throughout. For Rungwa, see Willis 1981, and Walsh and Swilla 2001:279.

[18] According to Pimbwe sources, the chief of the Pimbwe married a Ukonongo woman from a specific clan. The throne passed to a son of this woman. (Interviews with Daniel Kasike and Chief Nsalamba, July 2004).

[19] Sources: Kibaoni Interviews 2001, Abe personal communication, and interviews with Mtemi Beda July 2001, and Gongwe elders July 2001.

[20] Bende, Tongwe, and Holoholo are all considered variations of Bende languages. Tongwe is still used for villages near Kigoma on the shore of Lake Tanganyika. Holoholo is considered to be an archaic term, although it is used in British colonial records to describe particular villages in the northern part of Mpanda District, and southern part of Kigoma.A Bende Dictionary was recently published (Abe 2006).

[21] Tambila 1981.

[22] Interviews with Chief Beda, July 2001.

[23] See Lamb 1929.

[24] See Moisel 1905.

[25] Interviews with Gongwe elders in Sitalike Village, July 21, 2001, and trip to Gongwe village site, July 24, 2001.

[26] The chief list for the Gongwe as described to me in July 2001 is as follows: Before the arrival of the Arabs (about 1850): Kongwe, Miombe, Vibe, Sishambuka. After the Arab times: Shambwe, Tende I, Sunga I, Kakamba, Sigulu (about 1900), Lukandamila, Tende II, Sunga II, and Kakamba II (Tende Sigulu).Sunga II died about 1999, and the Gongwe were given permission to bury him in the traditional royal graveyard, which is now inside Katavi National Park. Kakamba II (Tende Sigulu) in 2001 lived in Mpanda where he was active in local town politics. Source: Interview with Kakamba II (Tende Sigulu) in July 2001.

[27] However, as Benedict Anderson (1991) pointed out, they are also the imaginings of a European belief system that equated territory, mother tongue, and political leaders as being congruous and redundant distinctions. This may have been a convenient short-cut for Europeans dividing up the Ottoman, Russian, Austrian-Hungarian, and German Empires after World War I. But the assumption that territory, mother tongue, and political identity are inherently related is not necessarily good social science when assessing the rule of organizing society in Europe or Africa.

[28] Smythe (2006:15, 152) describes "ruga-ruga" as being a word that Europeans used it to describe the "King's" soldiers. What is clear, is that the capacity to draft and command ruga-ruga was key to the power of anyone who wanted to participate in nineteenth century trade.Those who did not command such a military force were unable to protect trade stores. In contemporary twenty-first century terms, the ruga-ruga were child soldiers.Meaning they were boys taken at a young age from their parents, and raised to have primary loyalty to their ruga-ruga brothers, and their commander.Such young men, raised by their brother ruga-ruga, were lethal tools in the hand of a charismatic leader.

[29] Tambila 1981:59.

[30] Willis (1981:45-48) also discusses older aristocratic statuses, such as the Twa who ruled Ufipa in the sixteenth to eighteenth centuries.

[31] See Shorter 1972:276-279 and 1969:11.

[32] A Catholic missionary establishing a station in the Fipa highlands in 1911 claimed that villagers were attracted to the station not out of commitment to the Catholic faith, but because close association with the missionaries made it much less likely that the king's soldiers would take their goats (see Smythe 2006:15).

[33] Shorter 1972:59.

[34] The term "caste" is used specifically here to describe groups whose ideals assumed a superior raking, and excluded marriage with groups regarded as inferior (see Weber 1946).The term is used here even thought the discussion is not of India.Nevertheless, as Weber notes, while the phenomenon is strongest in India, the term can be used to describe similar social phenomena elsewhere.

[35] Bennett 1971:28; kin charts typically identify each of these chiefs as having their origins in one clan of Nyamwezi (Nyayembe) who had their traditional home in Unyayembe (Shorter 1969:9).

[36] See e.g. Shorter 1969:7-11, and Bennett 1971:33-36.

[37] The term "caste" of course comes from India, and has long been used in sociology and anthropology to describe endogamous status groups which have occupational monopolies within a system of social stratification.Weber (1946) explicitly used this word to describe this system in both India and Europe. Despite the dated nature of the word, I think that it is still the best word to describe the relationships between leadership groups in nineteenth century Rukwa, and the masses of horticultural peasants.

[38] Waters 1995.

[39] Willis 1981:29-32; the Nyamwezi of Tabora also had relations with the Tutsi who lived on the Malagarasi Plains herding cattle (Bennett 1971:5).

[40] Willis 1981:34.

[41] It was probably in this context that the Tutsi King Kapere of Ufipa married the Tutsi Queen Theresa of the Waha (Kigoma Region) in the 1950s. For a more complete discussion of the term Tutsi, see Waters 1996. Bennett writes that the Nyamwezi also had relationships with Tutsi living in the Malagarasi Plains as cattle herders in the nineteenth century.As with the term "Nyamwezi", the term various forms of the word "Tutsi" have many meanings across central Africa.Lemarchand (1993) elaborates this in his description of Burundian caste relations. Waters (1995 and 2003) discusses how the term "Tutsi" was used in a variety of circumstances, including in western Tanzania, in Burundi, Tanzanian refugee camps for Rwandans in the 1980s, and urban Burundian refugees in Dar es Salaam in the 1990s.

[42] Tambila 1981:82.

[43] Besides control of trade routes, chieftainship in Rukwa included a capacity to try miscreants accused of witchcraft and other crimes by poison ordeal which involved taking a poison—if the miscreant vomited, he was assumed to be innocent, and if he died, he was assumed to be guilty.An alternative to this was to place the fruit of the umwaafitree in a pot of boiling water.If the accused was able to retrieve the fruit without burning his hand, he was innocent.If the hand was burned, he was guilty (Willis 1966:21).

[44] Iliffe (1979:41) indicates that the Arab traders may have penetrated as far as Ukimbu by the 1825.They were present as an independent community under the protection of the Buganda King (Uganda) in the 1870s (Reid 2002:27-28).

[45] See Hamilton and Waters 1996, Weber 1946, Schuetz 1944.

[46] Tambila 1981:87.

[47] Smythe 2006:5, Tambila 1981:96-98, 100-102.

[48] Smythe 2006:6.

[49] See Smythe 2006:5-7.

[50] Smythe 2006:1.

[51] Koponen 1988:77.

[52] See also Burton 1858:75-77

[53] Willis 1981:78-81.

[54] Burton 1861:34.

[55] Willis 1981:80, Shorter 1972:276.

[56] A number of the blue beads described by Burton (1858:398) were recovered from an archaeological excavation in Kibaoni Village at the northern end of the Rukwa Valley in 2004.The excavation was done inside a small fortress which had been occupied in the nineteenth century by officers of the Pimbwe chiefdom (see O'Brien, Waters, and Mapunda 2004).

[57] See Waters 2007a:163-164.

[58] The explorer David Livingstone was "found" by Henry Stanley in Ujiji in 1869. Livingstone, who was always steadfast in his opposition to the slave trade, was staying in Ujiji where ironically he was hosted by the Omani Arab and Swahili slave traders who were receiving ivory and slave shipments from the Congo across Lake Tanganyika. See also Livingstone (1874).

[59] Slavery was legal in mainland Tanzania until the British abolished the institution after World War I.Villagers in Kibaoni VIllage reported that the last redeemed slave in their village had died in about 1997.

[60] Today, of course, there are substantial elephant herds in what is now Katavi National Park.It is not known what the status of such herds were in the mid-nineteenth century when the area which is now the park was occupied by humans who cultivated grains, and hunted. See Situt et al., (2003) for a description of human/elephant ecology.

[61] See Shettler 2007:135.

[62] Stanley: [1913(1969)]:357-58.

[63] Interviews with Daniel Kasike, Msago Omari, and others about the compound of Maji Moto, 2001 and 2004.

[64] See Willis 1981:70.

[65] Shorter 1972:118; I heard a similar story about sharpened stakes being placed at the bottom of the dry moat in Pimbwe in 2001. Shorter's account refers to "snakes" instead of "stakes." I assume (along with one of the reviewers of Shorter's book) that this is a typo.

[66] Tambila 1981:54; oral descriptions of the compound at Pimbwe/Maji Moto indicate that there was a similar design.

[67] See LeBlanc 2003 for a general discussion of this phenomenon.Also Shetler 2007:135-166.

[68] Interview Daniel Kasike, 2004.

[69] Simba a nom de guerre of an important Konongo/Nyamwezi chief, meaning "lion" in Swahili.The same figure, who guaranteed the sale of Karema in 1878 to the International African Association of King Leopold, is known as being a scion of the Nyamwezi royal family who left the Tabora area, and organized the Konongo into a substantial walled town, probably in the 1860s (see Stanley 1913 (1969), Thomson 1889 (1968), Shorter, 1972:245 and 1969:8-9; Bennett 1972, Tambila 1981).The Scottish explorer David Livingstone (1874:233-237) passed along the lakeshore in 1872 and noted not the depredations of the Ngoni, but the villages burned by a chief "Simba" who lived inland from Lake Tanganyika.Konongo and Pimbwe informants when asked about Simba in 2004 did not know of him. Instead they attributed the attack on Pimbwe to Mirambo.

[70] Quoted in Tambila 1981:85.

[71] See Hochschild 1999.

[72] Koponen 1994:588-589, and Shetler 2007:144-146.

[73] Cameron 1877, 1, 128, 129-130; Stanley's (1872:257) description of such fortresses was similar, though focused more on military utility: "Their bomas [walls] are so well made that one would require cannon to effect an entrance, if the villages were at all defended.They are skillful also in constructing traps for elephants and buffaloes.A stray lion or leopard is sometimes caught by them."

[74] Bennett 1971.

[75] Mirambo's European acquaintances were impressed with his capacity to command—Stanley compared Mirambo to Napoleon—although it is unclear that the hegemony he exercised was greater or lesser than his contemporaries like Nyungu ya Mawe, Simba, or others.However, Mirambo's approaches to political dominance are better known to Europeans, and for this reason, he makes a stronger impression on the written record.As a source, he provides an important context for understanding the more remote and lesser known chiefs such as those in Rukwa.

[76] Stanley 1913[1968]:358.

[77] Stanley refers repeatedly to the threat posed by the Wazavira. There are no other references by other writers to this group so far as I can see.Presumably they were eliminated as a threat between the time that Stanley passed in 1871, and the arrival of other explorers some two to ten years later.It is not clear whether the term applies to a clan, secret society, tribe, or brigand group.However, they did seemingly inspire terror in the anarchic societies Stanley found.

[78] The final lonely elephant lasted only a few months in Karema before it too died.

[79] Kasogera was described by my informants as Pimbwe (Interviews of Daniel Kasike, and Zakaria S. Kalulu July 2001). Quoting Carter's diary, Bennett describes him as Nyamwezi, as does Shorter 1972, p. 270 n. 11.

Lamb (1929) describing the accession of Kasogera indicates that he was probably a commoner, and/or caravan leader. All sources agree that he was outside the normal line of Pimbwe succession. His reign was particularly violent.After defeating the Konongo, probably in the 1860s, Kasogera was deposed by Mfundo.Kasogera in turn went to live in Sakalilo with the Fipa Chief Kimalaunga. Mfundo reigned cruelly for seven or eight years before in turn being deposed by Kasogera, who then returned to Pimbwe.Mfundo responded by seeking refuge with Chief Mirambo in Unyanyembe, and he accompanied Mirambo in 1880 when he invaded Fipa, Pimbwe, and Simba's polity. Despite Mirambo's success at Pimbwe, Mfundo refused to stay, and Kasogera retained the chiefship and immediately purged potential challengers, putting a number to death.

[80] See Bennett 1972:111-129, and Lamb 1929.

[81] The Nyika (or Nyiha) lived on the Fipa plateau at the time that the Fipa first arrived before the eighteenth century.The chronicles of the Fipa collected by Willis (1981:49-52 and 62-64) describe them as nomadic hunters and gatherers who fought with the horticultural and pastoral Fipa until the nineteenth century. A population of Nyika lives today in Mbozi District of Mbeya Region.

[82] Milansi Chronicle quoted in Willis 1981:92.

[83] Thomson 1880[1968]:218-219.

[84] Fatness in women was considered a sign of beauty in several other places. Thomson (1880[1968]:221) mentions Karagwe.More recently, Mushikwabo and Kramer (2006) describe the phenomenon in the pre-colonial Rwandan court.

[85] Thomson 1880[1968]:221.

[86] Thomson (1880[1968]):231; villagers born in Gongwe who were from Sitalike showed us the remains of the ditch and wall of a Gongwe village in 2001 and 2004.

[87] Thomson 1880[1968]:233.

[88] Willis (1981:93), writing from a Fipa perspsective, describes the alliances that Simba created as being "a predatory, slave-raiding state in Ukonongo," although it is not clear that his capacity for slave trading was particularly notable in the context of greater depredations by Zanzibari Arabs in the Congo.

[89] Bennett 1972:134-135, and Shorter 1971:117-120

[90] Shorter (1972) also describes the slow demise of the fortresses in Kimbu during the late nineteenth and twentieth centuries. The stockades of the Pimbwe, according to one report, also had by 1897 changed character, and were of the type to repel animals, rather than protect from human enemies (Kjekshus 1996:77).

[91] Tambila 1981 and Kjekshus 1977[1996].

[92] In the Rukwa Valley there was also an ecological change as Lake Rukwa inexplicably dried up (as did other lakes of East Africa at the time, see Calvin 2002:157-158).As a result, between 1905 and 1929 in particular, vast herds of animals moved into the Rukwa Valley which became a grassland. See Kjekshus 1996:76-77. For the German assault, see Willis 1981:206, and Shorter 1972:267-268. The German commander apparently attacked on the basis of a tip-off from an Arab trader. This tip-off was related to Willis decades later with some glee by the Fipa story-teller who resented Kimalaunga's maurauding more then the duplicity of the Arab, or the brutality of the Germans (Willis 1981:206).

[93] Interviews with Daniel Kasike.

[94] Quoted in Willis 1981:211.

[95] Willis 1981:211-214, see also Shorter 1972:267-269.

[96] Shorter 1972:321-363.

[97] British archival records on microfilm and the Tanzania National Archives are replete with kinship charts which were used to assess decisions they made about inheritance of "chiefdoms," and other offices they needed to administer the colony.British ideology about the legitimacy of kinship in determining chiefs reflected a need to promote compliant tax collectors to serve the colonial state.

[98] See Kigoma District Annual Report 1927, quoted in Waters 2007a:173.

[99] Willis (1966:54) lists twenty three chiefs in his list, and apparently used Lamb (1929) as a source.Other lists provided by oral informants in 2001 were shorter reflecting both memory compression, and variations in accounts from different lineages. Presumably, the ancestors of the Pimbwe lived at Maji Moto at least since the eighteenth century, if not earlier.Pottery shards apparent in ditch walls up to 100 cm. below the current surface would indicate horticulturalists were there at least several hundred years.The remains of the dry moat were measured at 315 m. on the north-south axis, and 488 meters on the east-west axis in 2004.

[100] Kalulu was probably born in the early 1880s. The death of Kasogera led to a succession crisis in the late 1880s due to the fact that potential claimants had been purged following the sacking of Pimbwe by Mirambo/Simba and their ally Mfundo in 1880. As described in the text, Kalulu was from the house of the deposed Chief Mfundo.Kalulu was chief of the Pimbwe until the 1920s when he was arrested and exiled by the British for failing to show them deference in a legal dispute.He pressed his claim to the chiefship for at least ten years after this. According to his grandson Isaac Lyoba Kalulu, Chief Kalulu died in 1968. Other sources include Daniel Kasike and Chief Nslamba. All commented on the role of using human hearts as a way to preserve power.

[101] The Pimbwe at this point recognized the authority of the Fipa at Nkansi over them. On the other hand, there was also a tradition that the Pimbwe chief must always marry a Konongo bride. Sons of this marriage were eligible for the chieftainship upon the death of the chief (Interviews with Daniel Kasike, 2001).

[102] See Moisel 1905.

[103] Lamb 1929.

[104] Ibid.

[105] The southern part of Bismarkburg Region came under British occupation beginning in 1916.The northern part of Bismarkburg was under Belgian occupation from Kigoma between 1916 and 1920.In 1920, the Belgians retreated from western Tanganyika and handed sovereignty over to the British who were to administer the area as part of its League of Nations Protectorate. Belgium received what is now Burundi and Rwanda as a Protectorate. Sources: Various oral interviews in 2001.

[106] Lamb 1929, and interviews with Daniel Kasike 2001.

[107] Multiple oral sources, 2001 and 2004. Especially Daniel Kasike, Zakaria Lyoba Kalulu, and Msago Omari.

[108] Interview with Daniel Kasike and Msago Omari 2001, see also Willis 1966:54n, Lamb 1929.

[109] Likewise though, there is little nostalgia for British rule; interviewees without exception were Tanzanian nationalists.

[110] As Shorter (1972:337) heard his interviewees in the 1960s remark, "after the white men stopped the Kimbu-Nyamwezi wars, the only people to make war again in Ukimbu were the white men themselves."

[111] Interviews with Chief Nsalamba, Mtemi Beda, Daniel Kasike, and others 2001, and 2004.

[112] Shorter 1972:117-120.

[113] Waters 2007a:193-198.

[114] Willis 1989:30.

[115] Neumann (1998), and Brockington (2002) have written critical evaluations of how park policies in Tanzania have expropriated traditional land rights from indigenous peoples.Shetler (2007) has written about the interaction between human ecology, and the development of parks in northern Tanzania.

[116] Waters 2007a:197.

[117] See Paciotti and Hadley's (2005) description oftrials conducted by Sukuma sungu sungu in the villages of Mpimbwe sub-division in the early part of the twenty-first century. Abrahams (1989) also discusses the sungu sungu in the Nyamwezi areas around Tabora.

[118] LeBlanc 2003:147.

[119] Chariot 1994:128.

[120] The effect of human violence on ecology is apparent when the presence of the massive herds of Katavi National Park are considered.Katavi today is what in the nineteenth century was a frontier area in which few of the Konongo, Pimbwe, Gongwe, or Bende could establish themselves and instead abandoning the land.This abandonment created the ecological conditions necessary for the large herbivores to flourish, particularly as human populations declined in the context of repeated onslaughts of disease and violence.

[121] Sources include : Moisel 1904-1905, Willis 1981:57 163, 166, Koponen 1988:82, 198, Kjekshus (1977)1996:11, 123. Cartography by Janna Waligorski.

[122] Moisel 1905.

[123] See O'Brien, Waters, and Mapunda 2004.

Notes on Sources: Interviews in Mpanda District, 2001 and 2004

Mpanda District was visited in June-July 2001, and July 2004 for research purposes under the auspices of COSTECH. A short week long visit was made to Kibaoni (Rukwa) in February 2004 for logistic purposes. The 2001 trip was focused by oral history. The 2004 trip was focused by archaeology, the results of which are reported elsewhere.

Formal research interviews for which notes were taken are below. Research assistants in Kibaoni were Mr. Michael Sungula (2001 and 2004), and Mr. Renatus Kaanzyemu (2001). Mr. Omari Msago played a very important informal role in the development of research protocols, and facilitation of interviews during both trips.

In Mpanda town, Mr. Gadiel Sindamenya was a collaborator, and we wrote a Swahili papers "Historia na Kabila ya Wagongwe", and "A History of the Bende" together on the basis of our interviews. These were distributed locally as photocopies. The assistance and collaboration of these four men are gratefully noted.

Data and impressions were also gathered in informal settings involving many participants on trips to the sites of former fortresses at Maji Moto, and Gongwe. The dates of the trips are listed below.

Formal Interviews

Chief Nsalamba, June 15, 2001, June 19, 2001, June 25, 1991, July 25, 1991, July 6, 2004 (all in Mpanda)

Mtemi Beda, July 4 and July 9, 2001 (Mpanda)

Batromeo Chundu, June 30, 2001 (Kibaoni)

Beda Shauritanga, June 29, 2001 (Kibaoni)

Brazio Kasumbi, June 29, 2001 (Kibaoni)

Chief Nsalamba and Phillip Mbogo July 26, 2001 (Mpanda)

Clement Mkalala, June 26, 2001 (Kibaoni)

Daniel Kasike, June 20, 2001, June 29, 2001, July 7, 2004, July 14-15 (Kibaoni and Maji Moto)

Adolph Kikwala and Emily Kapama, July 11, 2001 (Maji Moto)

Malko Katala, June 20, 2001 (Kibaoni)

Moris Mapelani, July 2001 (Kibaoni)

Mzee Maruko Katala June 20, 2001 (Kibaoni)

Mzee Zakaria S. Kalulu, July 26, 2001 (Mpanda)

Mzee Zakariah-Founder of the AICT church. June 18, 2001 (Mpanda)

Victory Kalelembe June 22, 2001 (Kibaoni)

Pius Magazi June 21, 2001 (Kibaoni)

Trip to Maji Moto, July 11, 2001 (Kibaoni)

July 14 and 15, 2004: Tour of house of Tadeo Ngomayarufu (Usevya) and Maji Moto (Daniel Kasike)

Emily Kapana and Mzee Adolph Pigachai Kikwala

Petro Kanyegere (Kibaoni), June 30, 2001

Mzee Isaac Lyoba Kalulu in Mpanda, June 18, 2001

About July 18, 2001, Chief Kakamba II (Tende Sigulu) of the Gongwe People (Mpanda Town).

July 20, 2001-Sitalike Village Office

Lazaro Katabi-Igongwe

Luka Milunga-Igongwe

Paulo Mbulu-Sibwesa

Abel Tende-Igongwe

Bazilio Eduard-Igongwe

Ramadhani Mohamed-Igongwe

July 24, 2001-Visit to Sigulu's Ngoma, Katavi National Park with Gongwe Villagers from Sitalike.


Abe, Yuko.  A Bende Vocabulary.  Bantu Vocabulary Series No. 13.  Research Institute for Languages and Cultures of Asia and Africa (ILCAA).  Tokyo: Tokyo University of Foreign Studies, 2006.

Abrahams, Roy. "Law and Order and the Ste in the Nyamwezi and Sukuma Area of Tanzania.  Africa. 59, No. 3 (1989):356-370.

Anderson, Benedict. Imagined Communities. New York: Verso Books, 1991.

Bennett, Norman Robert.  Mirambo of Tanzania ca. 1840-1884.  New York: Oxford University Press, 1971.

Burton , Richard (1861) The Lake Regions of Central Africa

Calvin, William. A Brain for All Seasons.  Chicago: University of Chicago Press, 2002.

Cameron, Verney Lovett.  Across Africa.  London: Daldy, Isbister & Co., 1877; reprinted New York: Negro Universities Press, 1969.

Chirot, Daniel. How Societies Change.  Boston: Pine Forge Press, 1994.

Dean, G. J. W.  "Grasslands of the Rukwa Valley."  Journal of Applied Ecology.  4, No. 1 (1967):45-57

Hamilton, Gary and Tony Waters. "Ethnicity and Capitalist Development: the Changing Role of the Chinese in Thailand." In Essential Outsiders: Chinese and Jews in the Modern Transformation of Southeast Asian and Eastern Europe, ed. by Daniel Chirot and Anthony Reid (Seattle: University of Washington Press, 1997), pp. 258-84.

Hochschild, Adam.  King Leopold's Ghost: A Story of Greed, Terror, and Heroism in Colonial Africa. Boston: Mariner Books , 1999.

Iliffe, John. A Modern History of Tanganyika.  Cambridge: Cambridge University Press, 1979.

_____.  Tanganyika under German Rule, 1905-1912.  Cambridge: Cambridge University Press, 1969.

Kjekshus, Helge.  Ecology, Control and Economic Development in East African History.  2 ed.; London: James Currey, 1996

Koponen, Juhani.  Development for Exploitation: German Colonial Policies in Mainland Tanzania, 1884-1914. Helsinki: Finnish Historical Society, 1994.

_____.  People and Production in Late Precolonial History.  Helsinki: Finnish Society for Development Studies, 1988.

Lamb, J. E. S.  The Wapimbwe.  Document in Tanzania National Archive, Mpanda Book, 1929.

LeBlanc, Steven A..  Constant Battles: The Myth of the Peaceful, Noble Savage.  Boston: Harvard University Press, 2003.

Lemarchand, Rene. Burundi: Ethnic Conflict and Genocide.  Cambridge: Cambridge University Press, 1996.

Livingstone, David.  The Last Journals of D avid Livingstone, in Central Africa, from 1865 to his Death.  London: J. Murray, 1874; reprinted Westport: Greenwood Press, 1970.

Maddox, Gregory, James L. Giblin, and Isaria Kimambo.  Custodians of the Land: Ecology and Culture in the History of Tanzania.  London: James Currey, 1996.

Moisel, Max Karte von Deutsch Ost Africa. Karema D2 and Bismarkburg D3. (Scale 1:300,000).  Berlin, 1905.

Mushikiwabo, Louise, and Jack Kramer.  Rwanda Means the Universe.  New York: St. Martin's Books, 2006.

O'Brien, Christopher, Tony Waters, and Bertram Mapunda. "Preliminary Archaeological Investigations at the Kibaoni Mpimbwe Site, Rukwa Region, Tanzania."  Paper presented at the Africa Studies Association Meetings, Washington, DC, 2004.

Oliver, Roland, and Brian Fagan. Africa in the Iron Age, c. 500 BC to AD 1400.  Cambridge: Cambridge University Press, 1975.

Paciotti, Brian, Craig Hadley, Christopher Holmes, and Monique Borgerhoff Mulder.  "Grass Roots Justice in Tanzania."  American Scientist, 93 (2005):58-65.

Paciotti, Brian and Craig Hadley. "Large Scale Cooperation among Sungusungu "Vigilantes" of Tanzania: Conceptualizing Micro-Economic and Institutional Approaches.  Research in Economic Anthropology. 23 (2005): 117-145.

_____ and Monique Borgerhoff Mulder.  "Sungusungu: The role of preexisting and evolving social institutions among Tanzanian vigilante organizations."  Human Organization. 63 (2004): 112-124.

Reid, Richard.  Political Power in Pre-Colonial Buganda: Economy, Society, and Warfare in the Nineteenth Century.  Oxford: James Currey, 2002.

Schoennbrun, David.  "Violence and Vlunerability in East Africa before 1800 CE: An Agenda for Research."  History Compass. 4/5 (2006): 741-760.

Schuetz, Alfred.  "The Stranger, an Essay in Social Psychology." American Journal of Sociology. 49, No. 6 (1944): 499-507.

Shetler, Jan Bender.  Imagining Serengeti.  Athens, Ohio: Ohio University Press, 2007.

Shorter, Aylward.  Chiefship in Western Tanzania.  Oxford: Oxford University Press, 1972.

_____.  Nyungu-ya-Mawe: Leadership in Nineteenth Century Tanzania.  Nairobi: East African Publishing House, 1969.

Situt, N. W., M. J. Walpole, J. Smight and N. Leader Williams.  "Predicting Spatial Aspects of Human-Elephant Conflict."  Journal of Applied Ecology.  40 (2003): 667-677.

Smythe, Elizabeth.  Fipa Families.  Athens, Ohio: Ohio University Press, 2006.

Tambila, Anselm.  "A History of the Rukwa Region (Tanzania) ca. 1870-1940: Aspects of Economic and Social Change from Pre-Colonial to Colonial Times."  Ph.D. diss., University of Hamburg, 1981. 

Thomson, Joseph. To the Central African Lakes and Back Vol. 2.  Boston: Houghton, Mifflin, 1881; reprinted, London: Frank Cass and Company, 1968.

Vansina, Jan.  Paths in the Rainforests.  Madison: University of Wisconsin Press, 1990.

Walsh, Martin T. and Imani N. Swilla..  "Linguistics in the Corridor: a review of research on the Bantu languages of south-west Tanzania, north-east Zambia, and north Malawi."  Journal of Asian and African Studies.  61 (2001): 275-302.

Waters, Tony. "Assessing the Impact of the Rwandan Refugee Crisis on Development Planning in Rural Tanzania, 1994-6."  Human Organization. 58, No. 2 (1999): 142-152.

_____.  "Ethnicity and Burundi's Refugees."  African Studies Quarterly. 7, No. 1 (2003):68-75.

_____ (2007a).  The Persistence of Subsistence Agriculture: Life Beneath the Level of the Marketplace.  Lanham: Lexington Books, 2007.

_____.  The Social Construction of Tutsi in Modern East Africa.  Journal of Modern African Studies. 33, No. 2 (1995): 243-48.

_____ (2007b).  When Killing is a Crime.  Boulder: Lynne Rienner Books, 2007.

Weber, Max (1946). "Class, Status, Party." In From Max Weber: Essays in Sociology, ed. by H.H. Gerth and C. Wright Mills (New York: Oxford University Press, 1958), pp. 180-195.

Willis, Roy. The Fipa and Related Peoples of South-West Tanzania and North-East Zambia.  London: International African Institute, 1966.

_____. "The Peace Puzzle in Ufipa." In  Societies at Peace: Anthropological Perspectives ed. by   Signe Howell and Roy Willis (London: Routledge, 1989), pp. 133-145.

_____.   A State in the Making: Myth, History, and Social Transformation in Pre-Colonial Ufipa. Bloomington: Indiana University Press, 1981.

Tony Waters is a professor of sociology at California State University, Chico. In 2003-2004 he was a Fulbright Scholar at The University of Dar es Salaam, Tanzania. He has written extensively about Tanzania, refugees, economic development, and public education. Among his books are The Persistence of Subsistence Agriculture: Life Beneath the Level of the Marketplace (Lexington Books 2007), When Killing is a Crime (Lynne Rienner 2007), Bureaucratizing the Good Samaritan (Westview 2001), and Crime and Immigrant Youth (Sage 1999). He lived and worked in rural Tanzania between 1984-1987, and 1994-1996.

Reference Style : The following is the suggested format for referencing this article: Tony Waters, "Social Organization and Social Status in Nineteenth and Twentieth Century Rukwa, Tanzania," African Studies Quarterly 11, issue 1: (Fall 2009) [online] URL: