Introduction^{Footnote 1}

A highly diverse group of actors originating from a vast area in eastern Congo began to move in the nineteenth century from the western to the eastern side of Lake Tanganyika for varying reasons even before the lake became a state border between the Congo Free State and German East Africa. These migrants, who came to be widely known as “Manyema” after they left their place of origin, are now deeply rooted in Tanzania. The mainly urban-based Manyema network stretches from the Kigoma-Ujiji area along the old central route of caravan trade from the interior of the continent to the Indian Ocean coast, most prominently including the towns of Tabora, Dodoma, Bagamoyo, Dar es Salaam, and Zanzibar.^{Footnote 2} Who is included under the umbrella term “Manyema,” is an on-going debate among Tanzanians with Congolese origins and academics alike.^{Footnote 3} For the Manyema Mosque elders in Dar es Salaam, the Manyema are not a “tribe” (kabila), but rather a “nation” (taifa) or just a mix (mchanganyiko), and every Manyema has his own kabila.^{Footnote 4} At the core of this fluid identification are ethnic groups like the Bangubangu or Kusu from Umanyema, a vast forest region around Kasongo and Nyangwe on the Lualaba River in eastern Congo. Ethnic groups from the western shores of Lake Tanganyika are depending on place and time eventually included under this label in Tanzania while it was never used as an ethnic category in colonial or postcolonial eastern Congo.

In his comparative historical study of Western African identities in the borderlands of Senegal and Gambia, respectively Ghana and Togo, Paul Nugent argues that “the boundaries of self-identification have constantly mutated in time with the shifting configurations of space and power.”^{Footnote 5} His appeal to bring the “history back into ethnicity”^{Footnote 6} lends itself perfectly to the Manyema case with its multiple changes in ethnic identification over space and time. At first, African-Arab and European visitors to precolonial eastern Congo started from the 1870s to refer to people from Umanyema as “the Manyema.” They used this designation especially for those young men who had left their home villages in order to join caravans or to work as hunters, soldiers, or agents for African-Arab traders.^{Footnote 7} African-Arab raids in the wider region and the war between African-Arab traders and the Congo Free State from 1892 to 1894 caused massive population movements, and people from Manyema who moved to the northern and eastern shores of Lake Tanganyika adopted the designation “Manyema” as one possible self-identification.^{Footnote 8} The ethnonym Manyema and the stereotypes associated with it were – as Paul Nugent describes for the Jola and Mandinka in precolonial West Africa – already in use before it became a category in colonial official use.^{Footnote 9} “Manyema” was the most common ethnic category for Africans with Congolese origins in German East Africa and Tanganyika Territory, while it never became an ethnic category in Belgian Congo. In reality, however, many of the people moved back and forth, taking their Manyema identity with them. Regional historiographies extensively cover the long-distance caravan trade between the Great Lakes and the Indian Ocean coast, the African-Arab domination around Lake Tanganyika, and the colonial conquest in the region. However, there exists not a single ethnographic or historical study exclusively dedicated to the Manyema, who came into being as a group within this regional as well as historical context. From an East African perspective, the term denotes the idea of a “foreign” origin from eastern Congo. In studies on eastern African history, there are a few references to the Manyema, as in John Iliffe’s comprehensive Modern History of Tanganyika.^{Footnote 10} Central African research traditions, in contrast, only mention Maniema as a region, and not as a people.^{Footnote 11} Thus, in separated epistemologies, the Manyema occupy a doubly marginal place in historiographical traditions.

In this contribution, I will explore the mobility and identity of this group of people that is home to both regions and plays an important role in connecting East and Central Africa as one single region and history. In this sense, the Manyema are an intriguing case to study how people at the same time thwart the distinction between an assumed “East” and “Central” Africa and are identified as a group based on this regional distinction. If one epistemologically overcomes the East/Central African divide, Manyema are by no means marginal but stand out as the link of the wider region.

Manyema migration between what later became the Democratic Republic of Congo (DRC) and Tanzania as well as the ongoing mobility of many of the historical actors challenges the researcher, even if multi-sited field research has now become rather the rule than the exception in migration studies. Tracing the life histories of Manyema people required access to various field sites in the two African countries and the availability of Manyema interview partners, who due to their mobile lifestyle had sometimes just left for another town or were new arrivals not yet conversant with local history. While all Manyema interview partners spoke Swahili as their mother tongue, the written sources were in Swahili, English, French, and German.^{Footnote 12} Thus, the long-term approach of the study with its multiple historical caesura combined with the objective to include perspectives from both sides of the “Manyema world” stretching from the forest region in eastern Congo up to the Indian Ocean coast, depended on multi-sited archival research in the two African countries as well as in the countries of the three former colonial powers.

The Manyema in Tanzania today do not use the label “diaspora” or a Swahili equivalent to describe their community.^{Footnote 13} However, every Manyema in Tanzania is descendant of a migrant, who, voluntarily or not, had left his home village in eastern Congo and crossed Lake Tanganyika by boat or walked around its northern end to settle to the east of the lake. Reasons for migration in precolonial times included the following of fish, trading, joining caravans, being taken as a slave, or fleeing wars and revolts. These individual accounts on the migration of ancestors to what is now Tanzania are supplemented by myths of ethnic origin, which are remarkably similar on the Congolese and the Tanzanian shores of Lake Tanganyika.^{Footnote 14} According to oral and archival evidence, Manyema mobility clearly goes beyond one-time unidirectional migration from Central to East Africa or even a mere mythic history of origin. I am aware of how difficult it is to ascribe mobility as a general feature to a highly diverse group such as the Manyema. Some interviewees, for example, stated they would never go back to their families’ areas of origin in Congo because of their hali ya kuja, the often painful circumstances of their coming to Tanzania, or simply because of the on-going violence in eastern Congo as compared to peaceful Tanzania.^{Footnote 15} Nevertheless, most interviewees mention frequent travels (be it to the DRC or within East Africa), mobile lifestyles and multiple homes as important part of their life as this enthusiastic quote by a retired Manyema in Tabora illustrates: “I visited a lot of towns in Tanzania. There are only five towns in Tanzania left I did not get to yet. I love to travel, I really love it and if I get a lot of money, I will travel. I love traveling a lot; it’s in our blood maybe.”^{Footnote 16} Interviewees spoke about male and female trading activities, work migration, changing government positions including military service, (Islamic) teaching and education, marriage, visits of relatives, and pure curiosity.^{Footnote 17} According to oral information, therefore, mobility – together with living in urban contexts and being Muslim – constitutes in addition to eastern Congolese origins an important marker of Manyema group identification. German colonial files mention the Manyema most often in connection with the spread of Islam.^{Footnote 18} British officials were more concerned with migrant labor issues and frequently reported on Manyema migrant workers in railway or road projects or on sisal plantations, and Belgian officials paid even closer attention to the movement of “natives.”^{Footnote 19}

Manyema moved within the wider Lake Tanganyika region in precolonial times and later connected the two adjacent colonies and the concomitant “Central” and “East” African spheres in terms of language (Swahili as well as local languages), religion (Islam), kinship, and through commercial integration. The Manyema network spreading from eastern Congo to the Indian Ocean coast was upheld by shared memories of migration and the evolution of a Manyema group consciousness, but also through continuing individual mobility between its various, mainly urban nodes. In order to show to what extent Manyema mobility continued to serve as a bridge between eastern Congo and Tanzania also after initial migration, I will highlight different phases of Manyema mobility corresponding to different colonial constellations and indeed differing views and regulations of mobility. These concrete histories of cross-border mobility in the first part of the paper follows a discussion at the more abstract level of colonial perspectives on “alien natives” between “East” and “Central” Africa and its implication for the Manyema in Tanganyika and Zanzibar. There were demarcations among the Manyema between people from the forest region and the lakeshores, for example, but also between early Manyema migrants oriented to the Sultanate of Zanzibar and later migrants oriented towards the Belgian Congo. Individual Manyema or whole groups situationally crossed these identity boundaries, which partly had concrete implications for colonial jurisdiction. Thus, also far away from the border region and long after initial migration, the Manyema continued to live with and across boundaries through shifting identifications with one or the other side of the border.

Crossing Borders in East Central Africa

The histories of Manyema mobility include 1) precolonial or early colonial long-term migration, which led to the constitution of a Manyema migrant community between Lake Tanganyika and the East African Coast, 2) border crossing as routine practice around Lake Tanganyika in colonial and postcolonial times, and 3) circular migration between and within the countries that originated during colonial rule. However, there were not only various forms of and reasons for physical mobility across (colonial or state) borders; seen from a historical long-term perspective, Manyema crossed multiple divides.

The bridging (i.e. both separating and connecting) role of Lake Tanganyika, which is at the heart of this section, changed its political nature various times. In the pre-colonial period, there was no clear-cut political border along the lake, but different shifting spheres of influence around its shores. The lake was, in fact, the center of an intense regional trading network. Early Swahili and Arab traders described the lake as bahari (ocean or sea) and its shores can thus, according to Achim von Oppen, be seen as “Tanzania’s other coast” in analogy to the Indian Ocean and therefore as space of connection rather than of separation.^{Footnote 20} On their way towards the Indian Ocean coast, people from eastern Congo crossed various geographic divides in the East Central African region. Umanyema, the homeland of the core group of Manyema, is a vast forest region, still some 600 km away from Lake Tanganyika.^{Footnote 21} It is in ecological terms quite different from the lake area, from the woodlands in central Tanzania, and from the Indian Ocean coast and islands. Bridging these different physical divides within East Central Africa required different modes of navigation, logistics, and provisions. Moreover, on their way towards the east, Manyema crossed boundaries of chiefdoms or political spheres of influence, which meant coping with changing safety issues as well as languages and religious and cultural specifics.

After the establishment of colonial rule in the region from the mid–1880s onwards, Lake Tanganyika was first envisaged and then gradually turned into a border with all its implications and institutions. Political changes altered the exact significance of the territorial border. From 1890 onwards, the lake was internationally recognized as the border between German East Africa (including what are now Ruanda and Burundi) and the Congo Free State, and from 1908 onwards between German East Africa and the Belgian Congo. Between 1916 and 1921, the political significance of the lake border diminished because of the Belgian occupation of western parts of the German colony. After most of the occupied land was handed over to the British in 1921, the lake boundary divided the Belgian Congo from the mandated Territories Ruanda-Urundi and Tanganyika, and since 1960–1962 the lake marks the state boundaries between Congo/Zaire, Tanganyika/Tanzania, and Burundi, with Zambia (formerly Northern Rhodesia) touching the lake at its southern tip.

People who came to be known as the Manyema crossed the same physical border at different times in history out of varying motives and reasons. Regarding the changing political nature of the border, they had to cope with different requirements and regulations, from certificates and passes over currency, language, and transport modes to ideological and labor policy inspired considerations as to who was allowed to cross or not. People from Maniema migrated not only eastwards, but also to the North/Northeast into Kivu region and along the river Lualaba to the West/Northwest. Furthermore, eastward Manyema migration was not confined to the Tanzania mainland: for example, the Sultanate and from 1891 on British Protectorate Zanzibar was an important destination as well.^{Footnote 22}

Transborder Mobility around the Onset of Colonial Rule

In the early 1890s, Lake Tanganyika did not yet constitute an effective barrier for mobility despite European attempts to interrupt the on-going “Arab” caravan and slave trade between the Great Lakes and the Indian Ocean coast. For centuries, people from the west of Lake Tanganyika had migrated to the northern and eastern shores of the lake in search for better fishing possibilities or related to the salt trade. Since the 1870s, they came in larger numbers as caravan porters, independent traders or war refugees (emphasized in Manyema memories), slaves (emphasized by missionaries and colonial officials), and to some extent as slave traders (emphasized by expedition leaders).^{Footnote 23} While histories of Manyema coming to Tanzania do differ considerably, it is certain that people from Umanyema were scattered widely all over western Tanzania and up to the coast at the onset of colonial rule.^{Footnote 24}

In early colonial files, there are some incidents recorded on the land border between Belgian Congo and German Urundi, including caravan trade, meetings of officials, surrenders, and requests for information, for example on a Manyema Menschenräuber (kidnapper).^{Footnote 25} However, the lake border is rarely mentioned. An important exception is a correspondence of officials of the Congo Free State with German officials from 1896 onwards on a large number of rebelling soldiers of the Force Publique.^{Footnote 26} Congo State officials repeatedly complained that rebels together with their families had crossed Lake Tanganyika or the Ruzizi River on its northern tip to seek refuge under German protection. In 1900, this involved up to 1,500 persons alone, and the Congo Officials asked for the surrender of the people and their weapons.^{Footnote 27} In the German files are no corresponding documents on the arrival of these Wahuni, the Swahili term for “rebels.” However, German Officials were eager to recruit experienced Manyema soldiers for their Schutztruppe and two leaders of the Ituri rebels, Mulamba and Kandolo, are cited in a British summary report on the Wahuni as “outstanding” figures in Ujiji.^{Footnote 28} Moreover, many of my Manyema interview partners proudly claimed Wahuni origin and, as the old quarter Kwa Wahuni in rural Ujiji may further evidence, we can assume that most of these Wahuni rebels stayed and settled in German East Africa.^{Footnote 29}

This episode is generally seen as the last wave of large-scale Manyema migration to Tanzania. It provides a good example of how historical research bridging two regions can add to reconstructions of histories no longer discernible if one consulted only the archives of one “side.” The loss or unknown whereabouts of German archival material caused by the First World War makes this approach especially relevant.^{Footnote 30} Moreover, combining oral with written sources from either side allows situating the Wahuni in perspective. Disloyal rebel soldiers in the eyes of the Congo Free State, who received little support by the local rural population in eastern Congo and had to flee across the border, became highly valued members of the German colonial army and were proudly remembered for their boldness among Manyema migrants in Tanzania.

Crossing the Belgian-British Colonial Border

From the 1920s onwards, Manyema border-crossing mobility seems to have changed in nature and motive. The majority of first-generation Manyema migrants, especially former slaves and refugees, tended to settle permanently on the German side of the lake. Written evidence for border-crossers in the late German colonial period is rare, both in German and Belgian records. Beginning from the mid–1920s, however, some of the first generation migrants and more frequently their children traveled back and forth to the other side of the lake and beyond. Political or safety considerations did play a minor role for migration after the end of the colonial conquest and the hardships caused by the First World War. In line with broader colonial developments, trade and working opportunities, but also existing ties with family and friends and religious education increasingly inspired circular migration and border-crossing mobility.

Colonial investment in infrastructure, with service provided by the Liemba ferry, several government steamers, and a regular dhow (traditional sailing vessel) service, facilitated mobility across Lake Tanganyika. Missionaries welcomed this enhanced communication for themselves but feared increased prostitution and separation of African families.^{Footnote 31} The British administration in Kigoma observed in 1924 that a “certain amount of movement of natives back and forward across the Anglo-Belgian border takes place annually and does not seriously affect the number of Natives in the district.”^{Footnote 32} The unpopularity of local leaders, tax pressures by the Belgian Government as well as intermarriage, social and business connections “between the people of the Township [Ujiji] and their friends in the Congo” were given as reasons for this border-crossing mobility.^{Footnote 33} The British administration opposed Belgian requests to force their subjects to return: “Choice of residence being apparently no crime, no power exists to turn such people out of the Territory, nor to demand of such few of our native subjects who for one reason or another prefer to reside under the Belgian Flag.”^{Footnote 34}

According to oral evidence, African traders continued to cross Lake Tanganyika as in former times by canoe rather than with official government steamers in order to circumvent customs duties. This “first beginning of smuggling,” that is the use of the border as a resource, led to conflicts with the Belgian colonial regime.^{Footnote 35} In general, the Belgian colonial administration took stronger efforts to control cross-border movements than their British counterparts, who gave travel permits more readily.^{Footnote 36} Massive Belgian investment in government-owned and private mining companies as well as in transport and harbor infrastructure to export raw material in the early 1920s required an immense labor force. Therefore, officials tried to keep subjects within the colony and to distribute them according to demand, for example to the Katanga mines. While the possible loss of tax revenue was a frequent issue in reports on the uncontrolled movement of “natives” within the Belgian colony, reports on transborder migration were more concerned with the loss of men power.^{Footnote 37} In addition to the effort to keep their working population in the colony, Belgian Officials tried to recruit laborers from neighboring countries such as Tanganyika Territory.^{Footnote 38} At the same time, Belgian officials tried to control individual immigration. The widely discussed immigration regulations from 1933 required for non-subjects knowledge of a European language, regular account keeping, a passport, a certificate of good conduct, and a health certificate.^{Footnote 39} In fact, the spread of sleeping sickness and other diseases in the Western Region in the 1920s alarmed also British health officials. Suspecting African migrants workers from neighboring countries to be the main vectors, the colonial administration tried to better control immigration and insisted on medical exams for recruited laborers.^{Footnote 40}

More than the British, Belgian officials as well as Catholic missionaries in eastern Congo were worried about the growing influence of Islam in the region from the 1920s onwards and saw religiously motivated transborder mobility as the main reason.^{Footnote 41} While Islam was part of the Swahili coastal culture since the ninth century, the large-scale spread of Islam beyond the coast began in the late nineteenth century and increased rapidly around the First World War. The Sufi brotherhoods (Turuq, sg. Tariqa) Qadiriyya and Shadhiliyya played a prominent role in the regional surge of Islam, especially in urban contexts.^{Footnote 42} These Islamic movements were established in Ujiji probably in the 1920s and revived older social and commercial connections among Muslims in the region. Sheiks and disciples of the Qadiriyya, referred to as Mulidi (from murīd, a follower of a brotherhood) in Belgian Congo, and Coran School pupils extensively traveled across the lake – at least, as long as Belgian Colonial officers let them do so. In the late 1930s, Belgian colonial policy increasingly aimed at intercepting Islamic connections across the border through close observation of the border stations, entry bans, the relegation of Sheikhs, or confiscation of Islamic newspapers.^{Footnote 43}

A closer look at these Islamic networks reveals the prominence of Manyema in this respect. Akida Kangala of Bangubangu origin was relegated by the British from Tabora to the Belgian Congo due to accusations by Tabora’s religious elite. He brought the Mulidi brotherhood with him and kept close connections with Tabora, from where he received religious instructions.^{Footnote 44} Sheikh Abdallah Kitenge, another Manyema Qadiriyya leader, regularly traveled from Ujiji to Belgian Congo in the early 1930s.^{Footnote 45} The children of the late Sheikh Abdul Muhsin Kitumba remember his life history in detail, which provides a deep insight into the close Muslim networks between Ujiji and eastern Congo.^{Footnote 46} Sheikh Abdul Muhsin’s parents were born in Kisangani when the area stood under “African-Arab” rule, and they were already Muslim when they fled the Wahuni war to Ujiji. They fostered the religious education of their son, who later went several times to eastern Congo to teach for a few months, partly together with his Zanzibar based tutor Sheikh Hassan bin Ameir.^{Footnote 47} While he received the travel permit from the British authorities without much difficulty, the following anecdote told by his sons illustrates that the border-crossing still had its challenges:

Sheikh Abdul Muhsin went to the Belgian Congo with a group of followers by boat, bringing the flag of the Qadiriyya with him. When he arrived, locals informed the government about a group with a flag who apparently wanted to undermine the government. After an interrogation about his travel intentions, the Belgian officials suggested a test to prove his purely religious aims: the Sheikhs’ group and a group of local Christians were asked to do a canoe competition and were allowed to paddle with their hands only. In the third round, the canoe of the Muslim group “went as driven by a machine” and they won the competition. Accordingly, the Sheikh was allowed to “enter and do religious work.” The flag, already known to the British officials in Ujiji, was accepted as a purely religious sign.^{Footnote 48}

Whether officials were happy to welcome Muslim preachers or not, the anecdote illustrates that the Belgian administration feared the political influence of incoming strangers, especially when connected to Muslim brotherhoods.

Islam was not simply “brought” from the coast to the hinterland. As Becker describes for rural Lindi in Tanzania, also people from Belgian Congo and Urundi actively sought to gain religious education for themselves or their children.^{Footnote 49} They took the hardships of crossing the border to Ujiji where Sheikh Abdul Muhsin or Sheikh Khalfani Kiumbe are, among others, still remembered today for their institutions of Islamic learning.^{Footnote 50} Sheiks from Kasongo, Nyangwe or Albertville (today Kalemie) stayed temporarily in Ujiji, Dar es Salaam, or Zanzibar for higher learning before they returned home.^{Footnote 51} Young children from eastern Congo were frequently housed by Manyema families in Ujiji.^{Footnote 52} Well-to-do individuals like the trader Yusufu Lusangi from Kisangani helped to spread Islam by transporting children between towns in Belgian Congo and Tanganyika Territory, financially assisting their education, or by donating land for a mosque in Albertville.^{Footnote 53} The growth of the towns on both sides of the border in the 1930s combined with budget shortages after the Great Depression led to a growing consciousness and will of parts of the African urban population to advance their own causes.^{Footnote 54}

Belgian officials, not aware of the financial solidarity among urban Muslims across borders, wondered why a “simple Muslim peasant (…) has enough income in order to send his six-year-old son Mukole to a Coran school in Kigoma.”^{Footnote 55} Transborder movement of Muslims was no elite phenomenon but formed part of the routine experience of numerous Manyema being at home in the wider region albeit transborder mobility slightly decreased following the Belgian relegations and bans from 1936 onwards. After the Second World War, however, a new wave of proselytism began and religious networks from the Indian Ocean to the eastern Belgian Congo intensified again. Ujiji was labeled the “headquarters for the Congo-Moslems and their different tariqa” by a Catholic missionary and was closely connected with Albertville across the lake.^{Footnote 56} Restriction on these movements was only partially successful. “The Muslim advance” as Belgian Officials stated in 1955, “has no reason to stop at the Congolese borders or at those of the Mandated Territories.”^{Footnote 57}

Besides religion, my Tanzanian based interviewees remembered many other reasons for mobility across the border for the 1950s: (partly illegitimate) trade with parrots, gold, vitenge cloth, dagaa (Lake Tanganyika’s famous dried fish), work, learning of local language and custom, celebrations, visits to (partly hitherto unknown) relatives, and, finally, curiosity to see the homeland that the elders were speaking about so often.^{Footnote 58} Travel permits did not constitute a major problem and although yearly about thousand people crossed the lake between Ujiji and Albertville, the authorities were not alarmed, as people “always had had the habitude to visit and receive their parents who were settled in the Kigoma region.”^{Footnote 59}

In the run-up to independence, political events like the BELBASE strike in 1957, TANU activity in the wider region and the Katanga struggle in 1961 led to a close observation of border-crossings by officials.^{Footnote 60} Safety concerns severely limited back and forth mobility by the Manyema and other people in the region. Cross-border mobility became increasingly unidirectional with people fleeing the post-independence outbreaks of violence in DRC towards Tanzania. These refugees, however, were labeled Kongomani by Tanzanians (including those of Manyema origin) and did not form part of the Manyema community.^{Footnote 61}

Identifications between Congo and Tanzania

Manyema mobility across or around Lake Tanganyika was, as I have demonstrated above, a process not interrupted, albeit in parts regulated by the colonial powers involved. Colonial perspectives on African migrants in their colony went beyond the abstract discussion about who belonged to a colony or not. The question, whether someone was regarded as “alien native” or “subject” had very real and concrete consequences for the people concerned. In this part of the article, I therefore want to outline colonial views on “alien natives” between “East” and “Central” Africa in Tanganyika Territory and the Zanzibar Protectorate and discuss the concrete implication of these views for the Manyema. In a second step, zooming in on the Manyema communities in Dar es Salaam and Zanzibar town, I will trace the divisions within local Manyema communities, namely between Manyema oriented towards the Sultanate of Zanzibar and Manyema oriented towards the Belgian Congo. However, these demarcations among Manyema were not fixed, as individual Manyema or whole groups situationally crossed these boundaries to the “other side,” with concrete implications for colonial jurisdiction.

Colonial Discussions about “Alien Natives” and Manyema “Nationality”

Most Manyema migrants and their descendants, even if highly mobile, settled within the borders of what is now known as Tanzania, be it on the mainland or on the Zanzibar Archipelago. Despite the challenges of alleged or real slave origin, these Swahili speaking Muslims got in general well integrated into the urban communities. The Manyema were partly involved in urban and national politics and gained Tanzanian citizenship after independence. British colonial officials, on the other hand, usually regarded them as “alien natives,” defined as “not a British subject, a British Protected Person nor a naturalized British Subject.”^{Footnote 62} This led, under colonial rule, to discussions on the status of the Manyema in the colony and hence touched on issues such as financial support for repatriation, access to land, tax and labor duties or administration of estates.

A Bembe migrant, who was formerly working for a British employer in Tanganyika Territory and wanted to go back to the Belgian Congo; a young Manyema widow who was stuck with her children in Aden without any financial means; and an elderly Manyema who had worked as clove picker on Pemba island and sought help to travel back to the Tanganyikan mainland – these are examples of Manyema life histories, which contributed to a debate on repatriation requests by “alien natives.”^{Footnote 63} In the latter case, Tanganyika Territory officials stated that “as the Manyema tribe, to which he is said to belong, is not native to the Territory it is regretted that this Government is unable to accept his return without further evidence that he is in fact normally domiciled here.”^{Footnote 64} However, proof of the applicants’ former occupations on the mainland was finally accepted – and an “alien native” was in fact “repatriated.”^{Footnote 65} Various comments in the file made clear, that officials were well aware of a large number of Manyema living in their territory as well as of transborder mobility around Lake Tanganyika. Still, they were not willing to accept the Manyema as “native tribe” of Tanganyika Territory in general.^{Footnote 66}

Manyema migrants were “alien natives” in a double sense. First, they came from a neighboring colony and hence did not qualify as Tanganyikan “natives” and second, they did not have a territorial home or “tribal area,” British colonial policy habitually assigned to a “tribe.”^{Footnote 67} Instead, the Manyema settled within the broader Kigoma region and in towns all along the central railway line as well as in the Tanga area and in Iringa. This posed problems to the “tribal” distribution of land and led to an interesting view on transborder mobility by a District Officer in Lindi in 1936. He argued that

This was not only true for borderlands, as he went on to say, as for example

Conscious of the multiple connections and movements of ethnic groups living in the border regions and beyond, and appreciating the economic contributions of immigrants, including the Manyema, this District Officer suggested not to take territorial borders as all-encompassing divide. Thus, he strongly opposed a “discriminative listing” of “alien natives” for land registration and tax issues.

The Ministry of the Colonies in Brussels was seemingly not even aware of early Manyema migration up to Zanzibar Island, as an exchange from 1926 reveals. Officials confirmed the existing practice of granting passports to Congolese subjects in Zanzibar according to testimony by a Manyema intermediary, probably motivated by the constant labor shortage in Belgian Congo. At the same time, they wondered why Congolese subjects without passports were living on Zanzibar at all and raised many questions about their motives for migration, their occupation, religion, language and so forth.^{Footnote 70} In 1950, the Colonial Secretary in Zanzibar was again asked to provide a list of all “Belgian nationals” residing in Zanzibar including their place of birth and date of arrival. This list finally provided in 1954, revealed that all “Belgian nationals” on Zanzibar were “believed to be of the Manyema tribe from the west side of Lake Tanganyika.”^{Footnote 71}

The following examples, notably addressing the context of inheritance and emigration in Zanzibar, show the difficulties the colonial authorities faced by establishing the territorial belonging of individual Manyema. Ethnic origin, relation to the African-Arab rulers in eastern Congo, time of leaving and arrival on the islands of Pemba and Zanzibar, status as to free or unfree, careers pursued after migration, treaties between the colonial powers, local administrative procedures – all these could have an impact on Manyema classification as Belgian Congolese or Sultan’s subject.

In the case of the death of a Manyema freed slave in Zanzibar in 1915, Belgian authorities had to prove that the deceased had in fact been a Belgian subject, in order to be able to administer his estate.^{Footnote 72} The question of who was a Manyema in Zanzibar and what territorial identity he or she had, whether he was a subject of the Belgian Congo or Sultan of Zanzibar, was thus directly linked to concrete financial issues for the colonial administration as well as for the heirs. There were about one thousand Manyema living in Zanzibar town at the time, with an estimated property of 100,000 to 150,000 Rupees. Therefore, the case stimulated a lively debate between Belgian and British authorities from 1915 onwards. The Sultan regarded former slaves on Zanzibar as his local subjects and so did the British. Most Manyema accepted this status even if they declined to have come as slaves to Zanzibar. The first idea to tackle the problem of territorial belonging was to introduce a time limit, from when on all Manyema migrants to Zanzibar were considered Belgian subjects and no longer Sultan’s subjects. Seen from a Tanganyika mainland perspective, this was problematic for those Manyema who had crossed to the eastern side of Lake Tanganyika before the establishment of colonial rule and had proceeded to Zanzibar only much later. Besides this, there was the difficulty for the administration to prove an individual’s claimed year of arrival on the island. Hence, the British asked the Belgians to acknowledge that all the Manyema in Zanzibar were subjects of the Sultan, “unless the Consular Authorities are in a position to prove that the individual in question was a free man at the time of his entry into the protectorate.”^{Footnote 73} A certificate of liberation issued by the Belgian authorities, for example, was considered valid proof. The discussion went on with an official in 1918 not seeing any “connexion between status (free or not free) and nationality.”^{Footnote 74} According to the Nationality and Naturalization Decree 1911, someone born in Belgian Congo could not be a Sultans subject unless his father was a Zanzibari.^{Footnote 75} In 1920, the British authorities asked their Belgian colleagues to comprehend, “that natives of foreign nationality, whose families have for generations been domiciled in the Protectorate, may well be unaware of their exact legal status in this respect and are thereby led to make declarations by which officers of the administration may be misguided.”^{Footnote 76} Whether Manyema, as alleged, were really “unaware of their exact legal status” or whether they rather consciously aimed at altering it for specific reasons, members of the Manyema communities of Dar es Salaam as well as of Zanzibar town stressed their affiliation either to the Belgian Congo or to the Sultanate of Zanzibar.

Divergent Belongings: Belgian Congolese versus Sultan’s Subjects

It would be too easy to state that all Manyema were, of course, more aware of connections rather than of an (alleged) divide between Central and East Africa, i.e. DRC and Tanzania. Today, Manyema often connect this topic with discussions of their own national belonging. Many Manyema would clearly say that they are Tanzanian nationals, having been naturalized after independence.^{Footnote 77} At the same time, they would insist to be Tanzanians with Congolese roots, or as one interview partner put it, “(…) we were born here and became Tanzanians but I could say that I’m not Tanzanian.”^{Footnote 78} While national identity did not yet play an important role in the 1930s, a similar vagueness in defining one’s regional identity and orientation towards Central Africa (Belgian Congo) or East Africa (Tanganyika and the Sultan’s Zanzibar) can also be discerned for earlier times, as following examples from Dar es Salaam and Zanzibar reveal.

Many Manyema men and women came to Dar es Salaam since its foundation and, attracted by the work opportunities in the young colonial capital, became the town’s second largest group of immigrants.^{Footnote 79} The colonial capital constituted also an important temporary meeting site for Manyema from all over the colony. A group of Manyema from Iringa, for example, walked in a three months march by foot to Dar es Salaam. The aim of the trip was to meet fellow Manyema from all over the colony, to perform “ngoma ya kimanyema” (“Manyema dance”) and to look for future wives preferably from one’s own kabila (ethnic group) or ancestor’s home villages among the town’s diverse Manyema population.^{Footnote 80} However, a long-lasting split within Dar es Salaam’s Manyema community demonstrates that there were different ways to cope with one’s Congolese backgrounds.

Manyema community leaders in 1927 asked the local Commissioner to mediate between two factions which competed in ngoma (dance performance) and mali (property), and to give an order called “kuondoa fitna” to rule out factionalism. Anybody trying to split the community should be “(…) made to leave the town and brought back home.”^{Footnote 81} Accordingly, an agreement was signed by leaders of various ethnic groups belonging to the Manyema community as well as the competing watoto wa Kimanyema (the Manyema children), the younger male and female leaders of the ngoma groups “Arabian Congo” and “Belgian Congo.” Factionalism was inherent to ngoma competitions all over eastern Africa during this time, often mimicking historical hostility. These dynamics are generally seen as a means to integrate newcomers in the different towns by generating enormous solidarity among the moieties.^{Footnote 82} The severe reaction of community elders in Dar es Salaam as to this praxis hints at the complicated process of Manyema identification, as well as their fear about a possible violent split as it had happened in Ujiji.^{Footnote 83} Despite the intervention by colonial officials, this factionalism was revived in Dar es Salaam in 1933. After unsuccessful attempts to solve the problem with the help of a Manyema intermediary from Tabora, the Commissioner of Eastern Province was approached again in 1938.^{Footnote 84} A year later, the Congo Union Association was founded, which united all “Eighteenth [sic!] tribes of Wamanyema.”^{Footnote 85} The association’s English title refers to the common Congolese origin as the unifying feature of Manyema identity. Its Swahili name, Umoja wa Wamanyema, Manyema Union, underlines their aim to foster “our Unity here, for the whole of East Africa.”^{Footnote 86}

Was eastern Congo by Manyema in Dar es Salaam then simply seen as part of eastern Africa? A definite answer to this question is difficult, as the terms “East Africa” or “Central Africa” rarely came up in my interviews. This may be because most Manyema migrated to what is today Tanzania at a time when the demarcations of East Africa and Central Africa were not yet common sense among the African population. The thinking in factions, anyhow, went on, and Manyema stressing the historical relation of eastern Congo with the “Arabs” or with the “Belgians” were competing for the community’s leadership in Dar es Salaam up to the 1950s.^{Footnote 87}

It is not clear from these sources, what made individuals in Dar es Salaam refer to one or the other in the first place. Did it have to do with their geographical or ethnic origin like in Ujiji, with differences that already existed in their home areas before they left, or had the time and mode of leaving have any bearing? Similar divides within the Manyema community in interwar Zanzibar may add to our understanding. Jonathon Glassman, working on the development of racial thought in Zanzibar, interprets Manyema self-identification as “Belgian Congolese” as a rejection of being a Sultans subject which had been until then a highly relevant feature of citizenship in Zanzibar. Condemning the ferocities of African-Arab rule they themselves or their ancestors had suffered in their eastern Congolese places of origin, some Manyema started to express their loyalty rather to the Belgian colonial power by waving the Belgian flag and founding a Belgian Congo Club, the later Manyema Union.^{Footnote 88} Thus, it was not only Colonial Officials but also the Manyema who actively decided whose subject they wished to be. Moreover, individual Manyema crossed internal community boundaries back and forth for various reasons applying for or rejecting their territorial identities.^{Footnote 89}

Manyema men, for example, refused to act as clove pickers or porters under the Labour Decree in 1916, proving that they were Belgian subjects by showing certificates issued by the Consulate.^{Footnote 90} Two Manyema women, born in Dar es Salaam and living on Zanzibar asked the DC for permission to travel to Saadani on the Tanganyikan mainland in 1920. When he refused, they went to the French Consul, declared that they were Belgian subjects and finally got their passports.^{Footnote 91} In the same year, leaders of eight different Manyema ethnic subgroups passed a declaration to the British authorities and expressed their wish to be registered as Sultan’s subjects again. They complained that when a Manyema died, Manyema of the Belgian Congo faction reported to the French Consulate and hence, the Belgian Government tried to take over the administration of estates.^{Footnote 92} In 1927 again, a group of Manyema women applied for registration as Zanzibar subjects and presented themselves as “forerunners of a large number of similar applicants.”^{Footnote 93} To avoid further conflicts within Zanzibar’s Manyema community, Manyema elders planned to assemble all Manyema to record their status. Colonial officials were hoping that this “very vexed question may be settled once and for all and these unfortunate Manyemas know how they stand.”^{Footnote 94} However, numerous cases from the 1930s show that competition as well as switching between the still existing “Belgian” and “Arab” faction of the community had not ended. Community leadership included the right to verify an applicant’s national identity for official registration and was thus a powerful, and at times abused, position.^{Footnote 95} The Manyema (Belgian subjects) Union Zanzibar-Pemba did not unite all the Manyema of the Archipelago. Instead, the “Manyema Union is just another name for Belgian Congo natives who do not wish to recognize the Sultan as their head,” concluded the Provincial Commissioner in 1937.^{Footnote 96}

The quote above underlines Manyema agency in identifying with the Belgian Congolese or the Sultan’s faction. Being a member of one of the two factions was not only a powerful means for coping with former experiences in eastern Congo. It was important for social cohesion and demarcation within local communities and constituted yet another possible layer of identity between the formal super-ethnic category Manyema and one’s ethnic subgroup, clan or village of origin. Finally, the territorial dimension connected with the respective faction had very concrete practical implications and some Manyema actively changed from one faction to the other and asked for official recognition of this. These boundaries within the Manyema communities in Tanganyika Territory and Zanzibar were not only relevant for the first generation of migrants, but also for their offspring. Thus, also far away from the border region and long after initial migration, the Manyema continued to live with and across boundaries through shifting identifications with one or the other side of the border.

Epilogue: Tracing the Manyema Connection

In tracing the Manyema connection between colonial Central and East Africa, I integrated data from interview partners and archival material from both sides of the alleged “divide” as well as corresponding archives in Belgium, Germany, and Great Britain. Interlocutors on both sides of Lake Tanganyika embraced the opportunity that their history is recovered – a history, which is doubly marginal in separated epistemologies, but stands out as the link of the wider region once we epistemologically overcome the East-Central divide. Most Manyema have never stopped to bridge this divide between eastern Congo and Tanzania in both their daily practices and identifications.

While one of my Manyema interview partners stated that the Manyema as a group were not nostalgic and others complained that their children were only choking about the whole “Manyema issue,” most of my mostly male elderly interview partners were happy to share their ideas and assist me in tracing their history. People asked me to print life histories and myths, record local languages and songs, or to send a camera for documentation purposes. They issued letters of introduction, offered to help me with my visa to DRC, and allowed their daughters to work for me as research assistants. And, naturally aware that mobile histories require mobile research, one research assistant even accompanied me on a cargo boat to Kalemie on the Congolese bank of Lake Tanganyika, where her grandmother once traded and part of the family still lived. In a similar vein, mosque elders in Kalemie, well connected to Tanzania and within the DRC, even tried to organize a car trip up to Kasongo for me, to the heart of the old Manyema region, which unfortunately turned out to be too expensive and dangerous for me.

It seems natural, that migrant elders who remark that the old languages, culture, and histories are fading away, are highly interested in keeping memories alive and that they stress on-going connections to their (partly mystic) home regions. At the same time, interviewees also proudly emphasized Manyema contributions to their new country, such as Islamic scholarship or political activism towards independence and nation-building, which was, according to Mohamed Said, long time forgotten.^{Footnote 97} However, as I have shown for the colonial period, the Manyema did in fact constitute an important group of actors (among others), who even after the colonial division of Central and East Africa, continued to cross multiple boundaries. They maintained transborder connections of various kinds, despite attempts by (especially Belgian) colonial officials to restrict such movements. Furthermore, the discussions on Manyema nationality and community identity in the last section demonstrate that orientations towards Central Africa or East Africa and the wider Indian Ocean or even political divides were never clear-cut and definite. Besides colonial ideas of who should belong to which colony, they involved concrete efforts by individual Manyema to navigate between the two spheres. The Manyema, being connected to and at home on both sides of a newly established border, therefore exemplify a group who, despite all obstacles, managed to thwart this divide by bridging it through routine practices and ideas of belonging.