13—
From Ethnic Identity to Tribalism: The Upper Zambezi Region of Zambia, 1830–1981

Robert Papstein

Introduction

In December 1980 I arrived in Zambia's Zambezi District for a short period of field work in Chavuma.^[1] When I presented my credentials to the District Governor, he politely, but firmly, told me that not only would I not be allowed to do field work but that I was to be confined to the boma under the direct supervision of the CID and that I would be returned to Lusaka on the next weekly airflight. He explained that the district was, after months of negotiations, finally becoming calm again after the publication of a new Luvale history book which I had helped to edit and which challenged many long-held and hotly contested historical views of the Lunda relating to the authenticity and antiquity of their Senior Chief, Ishinde. In the preceding months Luvale-Lunda conflict had resulted in blocked roads, government services being suspended, house burnings, beatings, and a resurrection of ethnic animosity in the district unknown since the 1950s. In 1981 tribalism dominated Zambezi District.

This essay is about the development of this 'tribalism' in the Upper Zambezi region of Zambia among the Luvale and Lunda speaking peoples between around 1830 and 1980.^[2] While ethnic differentiation, based on differences of language, or at least dialect, historical traditions, small differences in material culture and cosmology, did exist objectively in the past in the Upper Zambezi, these differences have in the last hundred years been transformed into rigid and self-conscious 'tribal' markers. The region's people sought to adapt to and influence changes initiated during the colonial period which continue to dominate local politics and local relationships with the Zambian national state.

'Ethnicity' and 'tribalism' are highly charged words in contemporary Africa. The terms are often regarded as an auto-explanation for contemporary political conflict, but their ubiquitous use belies their vagueness. I use the term ethnic awareness to describe the result of a long-term, historical process in which the particularism of early Bantu-speaking segmentary lineages evolved into a view of an enlarged social field with loyalties defined in terms of similar languages and culture and with primary social and economic allegiances directed largely towards the lineages and clans (which crosscut contemporary 'tribes'). This was overlaid by a genealogically linked chiefly political structure which functioned on a very limited basis and almost entirely at the village or micro-regional level. The Luvale, Lunda, Luchazi and Mbunda-speaking peoples of the Upper

Zambezi area certainly had a developed ethnic self-awareness prior to either their contact with mercantile capital through participation in the trans-Atlantic slave trade in the early nineteenth century, or their experience as colonized peoples between 1906 and 1964.

What changed so dramatically from the mid-nineteenth century onwards was that a previously slowly evolving, fluid ethnic self-awareness was transformed into a new, harder 'tribal' structure to the extent that 'tribalism' was stronger and more politically relevant in 1981 than it was in 1881. Among the most important early reasons for this transformation was the fear created by the slave trade which encouraged small, lineage-based villages to come together into large, stockaded villages controlled by increasingly powerful chiefs. The emphasis upon ethnic identity was a potentially protective element in avoiding enslavement, as enslaving the follower of a chief struck the base of his newly enhanced authority. Of critical importance after 1906 were the administrative ambitions of the colonial state which sought to graft on 'traditional' structures a modern 'tribal' administration which enlarged, as well as replaced to varying degrees, many of the functions of the earlier polities. Under these pressures Upper Zambezian peoples evolved into 'tribes' or, more accurately, weak proto-states, the sort of polities which colonial administrations encouraged in many parts of Africa and elsewhere: strong enough to carry out policy and maintain order but not sufficiently politicized to serve as an organizational focus against colonial rule. In the Upper Zambezi the hallmark of the modern tribe was a severely curtailed hierarchy of 'recognized' 'Senior' and 'Sub' Chiefs (and, after the 1930s, 'Native Authorities'), with new types of control over their populations through the expanded power of chiefs' courts, tax collection, implementation of colonial education, health, public works, and agricultural policies.

Having suggested a historical transition from the primacy of lineage/clan to the primacy of 'tribe', it is necessary to indicate why the clans did not retain their former attraction and therefore why tribal structures developed as they did. Perhaps most damaging—and most difficult to document—was the gradual change in individual perceptions that the village, lineage, or clan, was no longer the most secure protector and that access to land, fishing and hunting rights, healing, social recognition and economic advancement was increasingly regulated by larger polities such as the chief and tribe which were supported by the expanding colonial administration.

With the exception of the Lunda-speakers, all the region's 'tribes' possess the same clans. The matrilineal clans have no formal leadership or organization yet they have played an essential role in Upper Zambezi political, economic, religious and social life. As late as the 1940s they were still regarded as more important, in personal relationships, than tribal affiliation.^[3] It was by ignoring the clans and emphasizing the 'tribe', symbolized by the new, appointed chiefly hierarchy, that the fundamental ideological restructuring of Upper Zambezian societies began.

It is not necessary to enter into the current debate over the utility of such concepts as 'modes of production' except to say that all Upper Zambezian societies in the early nineteenth century fell within some form of what is generally described as lineage/domestic mode of production whereby the means of production was regulated through indirect (clan) and direct (genetic) relationships. When a group of men passed through the male initiation ceremony, mukanda, this not only granted them social recognition as adults but also entitled them to a part of the means of production in anticipation of their marriage. Although villages in this agriculturally marginal zone are generally small, in the better growing areas, such as Chavuma, they can be quite large, with hundreds of people living in a

single village. While it is possible that a village might be entirely Luvale-speaking, the pragmatism with which the Luvale and others practise a mixture of uxorilocality and virilocality within a mostly matrilineal system does not guarantee this. A village might also be linguistically heterogeneous, as many seem to be now, and certainly, because of exogamy, clan heterogeneous. In a village the apportionment of land and the other adult rights to fishing grounds and hunting time as well as other shared responsibilities would, in the first place, be the responsibility of one's immediate clansmen, tempered and expressed through the will of the headman and his advisers. In a difficult case consultation and approval could move even higher, requiring sanction at the level of the chiefs court.^[4] The trust of another person because he was a member of one's clan rather than one's tribe suggests the significance of a pre-colonial pan-tribal layer of allegiance totally ignored by later colonial organizational policies.

The importance of clan as against tribe not only was a constant of the past but is an essential element of the present. When travelling long distances or entering unfamiliar villages one always performs an obsequy, invoking one's clan formula, at the muyombo tree found in every Upper Zambezian village. Upon hearing the formula—a largely fictive genealogy which links the reciter to the clan founders—one is taken to one's 'relatives', who bear the responsibility for hospitality and social introductions.^[5]

It is not possible to explain here how these clans came into existence and why they have retained their influence except to say that a developed Luvale ethnic awareness was well advanced by the mid-nineteenth century. It was based in part on the military force of the Luvale NamaKungu royal clan and the historical, religious and technological innovations which they introduced.^[6] It was into this ancient system, already under stress because of the individual economic and social opportunities offered by linkages with the Atlantic slave trade, that British colonial administration, supported by Lozi expansionist aspirations, intruded at the turn of the century.

I see tribalism replacing ethnic awareness in the Upper Zambezi area in four overlapping phases. The first, from around 1830 to 1907, is the period of the slave trade and of Luvale domination of firearms. The second, from 1906 through the 1930s, was shaped by the impact of the early administrative policies of the colonial government. The third, beginning in the 1920s, saw the evolution of the colonial moral and political economies. Through their schools the Plymouth Brethren missions created a small group of locally educated culture brokers who reinterpreted Lunda and Luvale history both to the colonial authorities and to their own people and who articulated local dissatisfaction against one another and against the colonial administration, often using historical arguments. The participation of the Luvale and Lunda in labour migration to Zaire, Angola, Northern and Southern Rhodesia and, through the Witwatersrand Native Labour Association (WNLA), to South Africa in the 1930s and 1940s heìghtened awareness of the Upper Zambezi area's disadvantaged position vis-à-vis the rest of Northern Rhodesia's peoples. The colonial administrative and education policies mandated by the British government between 1941 and 1963 have remained largely unchanged in independent Zambia's Zambezi District down to the present. Finally, there is the fourth period, dominated by the impact of the continuing Luvale History Project, which began in 1969 and which is still continuing in the 1980s.

The Luvale and the Lunda

Luvale and Lunda speakers have occupied the Upper Zambezi since the

seventeenth century.^[7] The area is remote from the line of rail, urban/industrial centres, markets and state services. People of the area have thus had the opportunity to maintain a vigorous commitment to their own culture while, at the same time, engaging in the labour migration common to rural communities in the whole of southern Africa, with all of the contact and changes this has made possible. In 1948, C.M.N. White, a British civil servant who lived in the region for many years and who wrote extensively about Lunda and, especially, Luvale language and customs, estimated the population at about 30,000 Lunda (including the Lunda-Ndembu of Mwinilunga District), 60,000 Luvale (whom White called Lwena), 40,000 Luchazi and 20,000 Chokwe.^[8] The population has increased significantly since the late 1940s, but White's figures, although estimates, give an informed idea of the scale of Upper Zambezian societies at the peak of the colonial period.

There is a remarkable structural symmetry between the Luvale and Lunda (who are the focus of this essay) which makes them, on very important levels, societies which could, and did, coexist amiably. They correctly believe that they share different aspects of a common historical tradition, while linguistic similarities make communication relatively easy between all the societies of the Upper Zambezi.^[9] My own observations confirm White's often stated view that, with relatively minor variations, a highly similar material culture exists throughout the Upper Zambezi area. Lunda and Luvale cosmologies are also remarkably similar.^[10] Both groups freely intermarry, although intermarriage on the present scale is said to be recent. This has led to the irony that the two octogenarian antagonistic spokesmen of the Lunda and Luvale tribes, the Lunda Senior Chief, Ishinde, and Mose Kaputungu Sangambo are, in fact, in Sangambo's words, 'brothers' since they had the same father but Lunda and Luvale mothers.^[11] There is, moreover, certainly no social, political, or structural feature which should necessarily bring Luvale and Lunda polities into conflict.

The economies of the Luvale and Lunda are also complementary.^[12] The Luvale tend to favour the grasslands and thin cryptosepalum woodlands of the predominantly Kalahari sands west of the Zambezi river. An exception to this general rule is the zone of extremely fertile soils of the Chavuma area which straddles the Zambezi where it flows from Angola into Zambia. The Luvale are superior fishermen who harvest tons of mud barbel from the annual streams and ponds which appear during the rainy season. As early as the 1940s the Luvale were exporting 200 tons of dried fish to the Copperbelt.^[13]

As hunters, with little if any interest in fishing, the Lunda-speakers preferred to settle in the denser, once game-rich forests found on the eastern bank of the Zambezi. Lunda-speaking farmers are, however, also attracted to the fertility of the Chavuma area, as are the Luchazi and Chokwe speakers, recent immigrants from Angola. Chavuma is the most agriculturally productive part of Zambezi District—possibly of the entire Northwestern Province with the exception of the Mwinilunga area. An ethnically heterogeneous population has inhabited Chavuma since the late eighteenth century.^[14] Luvale and Lunda oral traditions do not mention any political conflicts between the two groups until the late nineteenth century when a series of 'wars' broke out.^[15] In fact, the traditions of both Luvale and Lunda emphasize their peaceful coexistence in the Upper Zambezi while at the same time setting forth somewhat different interpretations of the origins of political authority and the antiquity of Luvale and Lunda political titles, and vastly different interpretations of modern settlement patterns.

Yet local politics are now dominated, to their last detail, by Luvale and Lunda

tribal strife, a tribalism which reached such intensity that States of Emergency had to be declared in Zambezi District in the 1940s and 1950s.

The Upper Zambezi Slave Trade, 1830–1907

Participation in the Atlantic slave trade reached the Upper Zambezi in the mid-eighteenth century, attained its peak in the 1830s and 1840s, and slowly died out between then and the turn of the century.^[16] In 1907 slaves in that part of the Upper Zambezi under British administration were officially freed by the new colonial administration but a system of debt slavery continued on a limited scale for decades.^[17] To understand the significance of this for Luvale ethnicity and tribal identity, as well as the importance of changes made in the chiefly system by later British administrators, one must understand the essentials of Luvale political succession. The Luvale are matrilineal and uxorilocal. Chieftainship is restricted to a single clan among some thirteen clans, the NamaKungu. All children of female chiefs are therefore chiefs (vamwangana ). A child of a male chief is called Mwana Uta or 'child of the bow'. He can never become a mwangana . This means that, depending on the number and fecundity of female chiefs, it was possible to have hundreds of Luvale chiefs at any one time. With very few exceptions, chiefly genealogies tend to be very shallow for obvious reasons. During the slave trade certain vamwangana were able to create important new chieftainships. These coexisted with older titles, and with the Kakenge, whose ancient chieftainship provided the necessary legitimizing historical links which each chief required to be accepted as a mwangana .^[18] This proliferation of chiefs with vastly varying degrees of actual authority was to confront the early colonial administration with the 'need' to create a clear hierarchy of political power and one which was small enough in number to be 'manageable'.

Because of their advantageous location on the plains which fall away from the Angolan highlands to the west, the Luvale were the first in the Upper Zambezi to receive Ovimbundu traders in search of export slaves. As a general rule the Ovimbundu were not interested in taking the slaves themselves, but preferred instead to buy them for guns, cloth and jewellery. I have discussed elsewhere the response of the Luvale NamaKungu chiefs to the opportunities offered them by the slave trade and the links between Luvale expansion, guns and slaves.^[19] By the mid-nineteenth century, when we have travellers' accounts describing the region, virtually all of the major chiefs were also important slave traders. Given the nature of Luvale chiefly succession, it is clear that those who were able to control the slave trade and the economic power and access to firearms which it represented were those who became some of the most important chiefs.

The idea of chiefs as entrepreneurs is certainly not original to this essay. But the opportunities of the time gave to Luvale chiefs, and, quite conceivably, chiefly pretenders, the possibility of establishing a unique economic/military position of unprecedented strength in their competition for lands and followers. It is clear that the system of domestic production was being augmented by elements of a new mercantile economy in ways which strengthened chiefs and created 'big men' able to take advantage of international trade. In terms of Luvale-Lunda relations, the relative Luvale monopoly of firearms and the aggressive, expansionist policy which Luvale chiefs were following meant that any defenceless group was subject to enslavement. Luvale traditions are quite explicit in stating that many Luvale—in addition to the Lunda—were enslaved, and sometimes by their own chiefs. The systematic and large-scale enslavement of Lunda people by Luvale chiefs and 'big men' was less an indication of some ancient ethnic animosity as it was an

acknowledgment, in a new situation moulded by mercantile capitalism, of the capabilities of the powerful over the powerless. Beginning in the 1890s, Luvale slaving parties, usually led by local chiefs or their agents, carried out a vigorous series of attacks against the Lunda which came to be known as the Wars of Ulamba.^[20] In an unprecedented request, the Lunda Chief Ishinde appealed to the Lozi Paramount, Lewanika, for help against the marauding Luvale. Lewanika, who undoubtedly saw an opportunity for expanding his influence, sent a military contingent against the offending Luvale chiefs which was defeated in 1892, Luvale military prowess finding an ally in disease among the Lozi.

After the retreat of the Lozi, the Luvale continued raiding the Lunda who fled ever deeper into the forests. It is likely that, had the demand for slaves continued, the Luvale would have decimated the Lunda, but the closure of the market, for which the Luvale were dependent upon the Ovimbundu, ended the Wars of Ulamba. It is still common, however, in the heat of modern politics, for Luvale partisans to recall the Wars of Ulamba as 'proof of their 'superiority' over the Lunda. It was these changes in the patterns of the Upper Zambezi's history which cast relations between Luvale and Lunda in terms of ethnic or tribal politics. However, the coming of colonial administration created even more serious—or at least more immediate—problems for both groups, and while the Wars of Ulamba helped to form each group's view of the other, opposition to certain British administrative policies required a temporary common front and cooperation.

The Early Administration of Balovale Sub-District, C. 1907-C. 1930

When Balovale boma was opened in 1907 it was a sub-district of Barotseland. This was because the agreements Lewanika had signed with the British South Africa Company (BSAC) gave the Company the right to administer all of Bulozi and its dependencies. Lewanika had convinced the BSAC, which was anxious to counter possible Portuguese claims, that the Upper Zambezi was a part of the Lozi domain—a claim, supported, in the Lozi view, by their intervention in the Wars of Ulamba. Because it suited both BSAC and Lozi interests, Balovale, as it was then called, was regarded as a part of Bulozi.^[21]

The Lunda and, especially, the Luvale were totally opposed to direct or indirect Lozi rule and complained vigorously to a succession of District Commissioners that the historical justification used for Lozi overlordship was mistaken. Nevertheless colonial administrators continued to assert Lozi rule and each 'recognized' Lunda and Luvale chief was placed under the nominal control of a Lozi induna . The language of local administration was Lozi. All major decisions were referred to the Barotse Province headquarters in Mongu. And the Lozi were given an essentially free hand to 'bring administrative order' into Balovale Sub-District. To add to the injustice of having autonomous peoples under their domination, the Lozi sought to indenture the local population by instituting a system of corvée labour, presumably for public works and the extension of royal gardens, and a royal tribute from the rich fishing grounds.^[22] Luvale and Lunda resisted Lozi sub-imperialism, presenting their cases to the local authorities through missionaries of the Christian Mission in Many Lands (Plymouth Brethren) and a cadre of newly literate Luvale and Lunda mission-educated teachers and evangelists.

While the Luvale and Lunda were cooperating to resist Lozi encroachments, they became aware that the British, Portuguese and Belgian governments had

reached agreements concerning colonial borders that resulted in both groups being 'legally' divided between Northern Rhodesia, Angola and the Congo Free State. The Luvale refused to accept this division, continuing to regard the Kakenge chieftainship in Angola as their most important political title, as they do today. The Lunda of Mwinilunga opted for another solution by creating a second Kanongesha, their senior title, in Mwinilunga, leaving the original Kanongesha in the Portuguese territory he ruled.^[23] The Lunda of Balovale, with the help of local colonial civil servants, brought the Ishinde chieftainship from Angola into Northern Rhodesia, establishing there a Lunda Senior Chief against the claims of the existing Lunda chief, Mpidi. The Luvale attempted to assert the primacy of the Chinyama Litapi chieftainship, but this was denied by administrators who were not prepared to walk the sixty kilometres to Litapi. Instead the Ndungu chieftainship was moved from the Chavuma area to opposite the Balovale boma and declared 'senior.'

The division of the Upper Zambezi between three colonial powers and the subsequent restructuring of the hierarchy of local chieftainships, when combined with 'recognition' of a very few chieftainships, meant that the Lunda and, especially, the Luvale were given a political structure that was both almost wholly new and without significant customary power. Not only was the structure pyramidal to an unprecedented degree, but the recognition of a limited number of 'official' chiefs meant that the titles would remain permanent.^[24] In effect, the British created a form of positional succession.

In 1923, in an attempt to bring administrative 'order' into a district regarded by both the British and the Lozi as 'wild' and ungovernable. District Commissioner Bruce-Miller decided the Zambezi river would be the dividing line between the Luvale and Lunda 'tribes'.^[25] The use of the Zambezi as an administrative border, though it reflected a wholly erroneous understanding of culture, ethnicity, politics, and existing settlement patterns, was so compulsively appealing that virtually all District Commissioners attempted to employ it. The use of the river as a tribal boundary would have resulted in the bulk of the best arable land in Chavuma falling under Lunda authority when, by all accounts—then and now—Chavuma was a predominantly Luvale area under the Ndungu chieftainship. Bruce-Miller proposed not only that the newly arrived Lunda chief, Ishinde, take over Chavuma, but that the Luvale population be resettled on the eastern bank.

The Luvale at Chavuma and elsewhere resisted every effort to resettle them, and violence soon broke out with the Lunda, who supported the plan. Bruce-Miller pursued this foolish and unnecessary policy until he was replaced. Even though the forced resettlement policy was never actually attempted again, it became an article of faith among subsequent District Commissioners that the Luvale belonged 'properly' on the Zambezi's west bank and the Lunda on its east bank. Commitment to this point of view, reflected in the formulation of subsequent policies, has been the single most important stimulus to tribal strife between Lunda and Luvale. Every local political decision was—and still is—evaluated in terms of whether it will further or diminish each side's claim to Chavuma, the area's best agricultural land.

As Luvale and Lunda struggled against Lozi sub-imperialism and between one another over Chavuma, they were slowly subjected to a process of bureaucratization. We actually know very little about the local, internal effects of the creation of bureaucratic administrative structures in rural Africa.^[26] But in Balovale each person was required to declare for the district register his/her 'tribe', chief, headman and village. Never were people asked about their clan which, when combined with allegiance to a chief and headman, was the actual

avenue of access to land, fishing and hunting rights, and social acceptance. At the same time, the positions of the chiefs themselves came under scrutiny with a view to limiting the number of chiefs and creating a hierarchy of chiefs for each 'tribe'. Clearly, the chiefs with the largest land areas and populations under their sovereignty were going to be 'recognized', and there was considerable difficulty among both the Luvale and Lunda chiefs when Ndungu and Ishinde were moved to lands traditionally held by other chiefs. Thus the chiefs and headmen, who also faced the problem of 'recognition', were naturally eager to have the greatest number of persons possible inscribed in their 'book'.

I have already noted that the population of the district in 1948 contained, in addition to Luvale and Lunda, significant numbers of Luchazi and Chokwe and some Mbunda. Large numbers of Luchazi had entered Northern Rhodesia after the failure of their revolt against the Portuguese in 1916–1917 and the brutal repression which followed.^[27] They originally entered Bulozi, but because of Lewanika's objections to such large concentrations of 'foreigners', some were resettled at Kabompo. Neither Bulozi, where they felt they were treated as slaves (vandungo ), nor the forest lands of Kabompo were attractive, and many Luchazi migrated into Balovale grasslands and settled among the Luvale either in their own villages or as resident 'foreigners' in Luvale villages.^[28] No doubt this was also a time of significant intermarriage among Luchazi and the Luvale, who regarded the Luchazi as 'relatives' sharing the same clans as well as the same historical and social traditions.

The use of intermarriage to blur and redefine ethnic affiliations is a major theme of Luvale and Lunda history, extending for centuries into the past. Luvale traditions speak of how the Luvale chiefs of the NamaKungu clan occupied the Upper Zambezi with a mixture of force and monopolies on cultural and technical innovations and consolidated these with the autochthones, the Mbwela, through intermarriage. Even today it is common for Luvale to admit privately that even though they are Luvale, they are also 'really' Mbwela. The function of intermarriage, especially in times of crisis, has played an important role in allowing crossing of ethnic lines.^[29]

From the scanty data extant, it is difficult to gain a firm idea of the scale of Luchazi (as well as Mbunda) migration into Balovale. The 1920s and early 1930s was a period of important ethnic movement and redefinition, with the 'safest' ethnic identity in terms of rights of residence in the sub-district being either Luvale or Lunda. The choice of ethnicity was related to area of residence: non-Lunda residing in Chavuma would, if forced, choose a Lunda identity to protect their farms and rights of residence after the relocation policy was announced in 1923. Non-Luvale residing on west bank lands would have to choose Luvale ethnicity to protect against their resettlement to the east bank or their return to Kabompo and possibly Bulozi.

But how were these kinds of alignments possible when, in the case of Chavuma and the better lands and fishing/hunting areas, a wholesale incorporation of 'new' Luvale or Lunda would be resisted by those threatened by the increased competition for resources? The answer, however tentative, requires some discussion of local language, clan structures, cultural taxonomies, material culture and historical traditions. We are so accustomed to identifying the differences between people that often we fail to establish the continuities.

Our written records and the oral testimonies of the peoples themselves suggest some confusion concerning the meaning of ethnicity and, especially, 'tribe'. Portuguese records tend to lump all of the peoples of eastern, savannah Angola and western Zambia (except the Lozi) under the pejorative term 'ngangela' .^[30] For

the Portuguese, the ngangela was the vast plain which reached from the central Angolan highlands to the Zambezi. In this area they saw no significant cultural differences between the inhabitants. This term includes the people we know as Luvale, Luchazi and Mbunda, as well as other, smaller societies which view themselves as distinct from their neighbours. Gluckman clustered the same basic group under the Lozi term 'wiko', meaning 'peoples of the west'—again assuming that there existed little in their political, social or material cultures to differentiate them as separate groups. White wrote about the same people as the 'Balovale' and later as the 'Lwena'.^[31]

There are five indigenous 'languages' spoken in the Upper Zambezi, plus two imported languages, Lozi and English. Lunda, one of the five, is actually two mutually intelligible dialect clusters, Lunda and Lunda-Ndembu. The other four are the dialects of Luvale: the vakaKasavi (language spoken by the people along the Kasavi/Kasai river); vakaMbunda (spoken were the earth is red (mbunda )); vaka Yambeji (spoken along the Zambezi); and vakaMbalango (spoken in the plains area between the Lungevungu river and Bulozi). The Balovale are peoples who speak one of these dialects and live where the mavale plant grows. C.M.N. White preferred to use the term 'Lwena' instead of 'Luvale', but this term, which has connotations of venereal disease, was usually only applied to Luvale dialect speakers in the northern areas of Luvale country near the Lwena river in Angola but not in Zambia.

Lexicostatistical linguistic analysis of each language reveals that certain groups—the Luvale, Luchazi and Mbunda—are as close linguistically as they are culturally; that the Chokwe are similar both linguistically and culturally; and that the Lunda-speakers represent a somewhat different language and culture, and are related to a later stage of the historical traditions shared by each group.^[32] Locally, similarities are overwhelmingly acknowledged in opposition to differences—except where these might be exploited for some personal gain. In the past an attack along tribal lines was weak although 'tribal' differences today offer greater latitude for definition and manipulation.

The Evolution of the Colonial Political Economy 1920–1960

The arrival of missionaries of the fundamentalist Christian Mission in Many Lands (Plymouth Brethren) was of critical importance in the development of tribalism amongst the Luvale and Lunda. The Brethren reached the Upper Zambezi after creating mission stations across the interior of north-central Angola at the same time as the British colonial government was asserting an effective administration over Balovale and Mwinilunga. In addition to their new and powerful religious message, they offered the only access to western medical services and education, and they 'demonstrated' the utility of their teachings by becoming some of the most important traders and building contractors in Northwestern Rhodesia.^[33] Both Luvale and Lunda recognized, in the years between 1907 and 1923, that they lacked the organizational and literary skills to resist the injustices arising from Lozi sub-imperialism and the land resettlement plan. The Lozi demand for corvée labour and the Chavuma resettlement programme demonstrated how easily they could be manipulated within a system they neither understood nor could effectively influence except by temporarily effective passive resistance.

In 1922 George Suckling founded the Chitokoloki Mission Station in Balovale

Sub-District, an offshoot of the original Brethren station at Mwinilunga. Suckling was an outspoken, gregarious man who quickly became identified as the spokesman for Lunda and Luvale grievances, and found himself besieged by young men seeking religious training and the literate skills which accompanied it. Over the years, beginning in the 1920s, this group of mission-educated Luvale and Lunda emerged as the most forceful opinion makers in Balovale. After Zambia's independence in 1964, many of them rose to positions of importance in the Zambian civil service where they continue to have an important influence.^[34]

Without this group of educated, literate men, a development which was taking place on a small scale throughout Central Africa, the Luvale and Lunda would not have been able to resist Lozi pretensions as effectively as they did. Nor would they have been able to influence and moderate the implementation of colonial policies which they found either misguided or unjust. These forceful and articulate men carried an enormous burden during the colonial years, and, indeed, many carry similar responsibilities today. They not only responded to the political imperatives which propelled them into positions of responsibility and authority, but also began to use their newly acquired skills to give form and detail to the new 'tribal' world views which were emerging out of the tensions and structures imposed by British colonial administrative policies. While the immediate concern with history and custom can be interpreted as reflecting backward-looking conservatism, it became the most important and virtually the only tool over which the Luvale and Lunda had control and which they could use to influence policy. While officials in London or Lusaka might regard the destruction of certain local traditions and customs as one of the aims of British administration, local District Commissioners, were they to be effective, had to take them into account if the district was to run 'smoothly'. If local intellectuals could somehow link the solution of local problems to an understanding of local history and custom, they would gain a powerful lever in affecting the outcome. The Luvale and Lunda were successful in doing this although the long term implications for the development of tribalism were hardly comprehended.

I have already noted that British policy curtailed and structured hierarchically the polities of the Upper Zambezi, and that the Luvale and Lunda responded to this in differing ways. In both cases one antique chiefly title was regarded as senior to all others, although this functioned on a genealogical-legitimizing level rather than as a basis for regional administration. British policy converted the theory of Luvale and Lunda chieftainship into reality. One might have expected a greater resistance to this change, but this did not happen because the creation of a political hierarchy was not threatening to the new chiefs or the new intellectuals, and because they perceived that the implementation of a more centralized political model gave both the Luvale and Lunda greater claims to political existences separate from the Lozi. The most forceful means of doing this, in 'traditional' terms, was not only by emphasizing the cardinal importance of their history but also by developing, through historical research, their connections with their ancient origins, the powerful Ruund state in Zaire. People began to regard themselves less as members of a clan with chiefs who shared elements of a common historical tradition and more and more as Luvale and Lunda 'tribesmen', as British policy encouraged them to do, especially after the creation of Native Authorities. It was pre-eminently the group of Luvale and Lunda intellectuals educated at Chitokoloki Mission who gave form and content to these new feelings and, in turn, stimulated these growing commitments further by stressing the ethnic and historical uniqueness of each 'tribe'.^[35]

We know very little about labour migration from the Upper Zambezi. Most of

what appears here is impressionistic, and my purpose in discussing it is merely to suggest the role of migrancy as an element in the development of Luvale and Lunda tribalism. The Upper Zambezi is remote from the urban centres of Central and Southern Africa. A road wasn't built linking Balovale to the Copperbelt until 1941. Before that, and for some years after, the Zambezi river was the major artery of transportation for the region. Although Luvale and Lunda responded to the meagre markets available to them in Angola and Northern Rhodesia, such marketing opportunities were restricted by isolation, the small scale of production, and the lack of efficient transport. Thus migration to labour markets was, for most men, the only source of cash.

Few Upper Zambezian people migrated to Angola or Zaire because they regarded the pay and working conditions there as inferior to those of the Copperbelt, Southern Rhodesia or South Africa.^[36] The Witwatersrand Native Labour Association (WNLA) recruited systematically in the 1940s and 1950s, using permanent local agents, a system of barges which penetrated all of the major rivers of the region, and out-stations where workers were housed until they could be brought into the boma for transportation. At the boma WNLA maintained its own gardens and cattle herds as well as substantial hostels.^[37]

Labour migrancy gave the Luvale and Lunda their first urban experience, and those who migrated to the Copperbelt were shocked to learn that they were regarded as social inferiors by the more numerous Bemba and, of course, the Lozi. The Balovale people spoke unfamiliar and difficult languages, they remained fiercely committed to customs which others found 'bizarre', as, for example, the mukanda circumcision ceremony, and their reputation for herbal and magical expertise often made others spiteful or fearful of them. Their lack of education and urban experience, and their relatively small numbers, made it easy for others, in the bitter competition for work, to view them as rustics fit only for the worst and lowest paid jobs. The Luvale in particular soon developed an urban reputation as night soil carriers and menial workers. In general, 'Lunda' and 'Luvale' became very lowly ethnic identifications in town.^[38] There is little doubt, however, that ethnic identification was merely an idiom used in the broader political economy of the urban centres in the competition for better jobs.

At the same time that perceived low ethnic status was a hindrance to finding jobs, it also laid the foundations for even more enduring social problems for Upper Zambezian peoples. By 1964, the time of Zambian independence, certain ethnically defined groups had come to dominate the choice positions within certain sectors of the economy. The Bemba, for example, had achieved this in the mining sector. And the fact that for decades such preferences were 'justified' or accepted meant not only that at independence most Upper Zambezian people held relatively unskilled positions, but also that the possibilities for rising in the system were limited. This situation was certainly not limited to any single group, but there is no question that within contemporary Zambian society certain ethnic groups or 'tribes' have larger representation in certain government departments or parastatals than their absolute numbers would permit in a random selection of qualified personnel.

The political economy of the colonial state had encouraged the creation of tribal groups in the rural areas and consequently these identities applied within the urban industrial economy as well, reinforcing the rural perceptions. This not only meant that rural peoples had to identify themselves with a 'tribe' in order to 'fit in' and enjoy official legal recognition in their local district, but also ensured that tribe, through the Pass System which allowed only so many tribesmen to migrate to town, was used to identify and tacitly to separate workers according to

their presumed abilities into such categories as the 'clever' Bemba or Lozi and the 'backward and wild' Luvale or Lamba.

The official colonial policy which regarded the Luvale and Lunda as subjects of the Lozi plagued every District Commissioner from the day he arrived in Balovale boma until the day he left. The dispute consumed so much time and so inhibited the administration of the area that it was finally decided in the late 1930s to hold a Commission of Enquiry into the issue. The Luvale and Lunda rejoiced at the prospect, partly because they were certain of victory and partly because the colonial administration had at last accepted their claims that only through an investigation of their history and customs could the matter be equitably adjudicated. The MacDonnel Commission thus took testimony in Bulozi, Balovale, the Copperbelt and Lusaka between 1938 and 1939.^[39]

The MacDonnel Commission is the epic event of modern Luvale and Lunda history. By 1938 both groups had a cadre of literate, experienced intellectuals who, in cooperation with local missionaries, orchestrated their testimonies to the Commission and who, in the preparation of masses of written materials—almost all of which was historical in nature—sought to demonstrate the separate and independent origins and development of the Lunda and Luvale tribes and their autonomy from the Lozi. In doing this they presented, consciously and unconsciously, a picture of ancient and centralized tribal polities which neatly fitted British preconceptions but which, in fact, they had only recently created. Both Luvale and Lunda set down, on paper and for the first time, universalist rather than local views of their histories. They understood that not only was this immediately important for their claims to autonomy from the Lozi, but that ultimately they would have to make similar presentations concerning their own conflict over Chavuma. Therefore the Luvale and Lunda testimonies asserted their independence from the Lozi but differed dramatically in their interpretation of their origins in the Congo, the 'migration' into the Upper Zambezi and, of course, the antiquity of each group's presence in the Chavuma area.

The MacDonnel Commission issued its report in 1941. It sustained both Luvale and Lunda claims to autonomy. The government responded by creating Balovale District and separate Luvale and Lunda Native Authorities. As the Luvale and Lunda saw it, they had saved themselves from Lozi overlordship after a generation of struggle. As I was repeatedly told by one of the Luvale who played a central role in presenting the Luvale view to the Commission, it was the finest moment of his life when the decision to create Balovale District was revealed.^[40]

In 1941 Thomas Chinyama reworked some of the testimony given to the Commission and published, in both English and Lunda, The Early History of the Balovale Lunda . This appeared in the Longman Central African Literature series, which had been created to give 'tribes' an outlet for publication of 'tribal' histories. The booklet was a bombshell in district politics because it asserted in bold and uncompromising terms Lunda claims to be the original settlers of Chavuma and thus the area's 'proper' occupants. Chinyama went further, claiming that Senior Chief Ishinde's title was of greater antiquity than that of Luvale Senior Chief Ndungu. The Luvale were deeply affronted, for they not only considered their 'Ndungu' title older and more prestigious than that of 'Shinde' but also felt that they, the Luvale, were the rightful owners of Chavuma.^[41] With the Lozi question barely behind them, the Luvale and the Lunda became locked in an administratively vicious and sometimes physically dangerous struggle over which tribe 'controlled' Chavuma. The full implications of the exact ways in which this control would be important remains somewhat speculative. But one certainty is that the new administrative powers of the chiefs and their important

headmen would have given them greater control over land, trading licences and, later, after the beginning of the MPLA nationalist struggle in Angola, control over refugees who could easily be converted into underpaid farm labourers.^[42] For the majority of Chavuma's population, who by this time had seen the new power of the chiefs and Native Authorities, the thought of a Lunda chief pursuing Lunda interests in a predominantly Luvale area was unthinkable.

Both Luvale and Lunda constantly petitioned government to decide in their favour regarding Chavuma, presenting elaborate historical documents largely derived from materials presented to the MacDonnel Commission and over the years supplemented by newly collected and interpreted information to support their claims. Both sides illegally attempted to place their own chiefs in Chavuma. In response the District Commissioner declared it a chiefless area, which it remains today. Nonetheless violence frequently broke out between Lunda and Luvale groups.^[43]

As the Chavuma dispute festered, its influence was felt at every level of district administration, with both sides tending to see every decision and every policy as somehow related to the issue of Chavuma. Not only were the Luvale and Lunda involved but the by now substantial and increasing number of Chokwe and lesser groups of Luchazi and Mbunda, none of whom had recognized chiefs in the area, 'sat the fence' as their rights of residence and to land were thus bound to the fates of competing Luvale and Lunda. Since Chavuma was declared 'chiefless', with real local power being taken up by a few very strong headmen, it was less pressing, and in fact could ultimately become dangerous, to choose a new ethnic identity.

Some of the critical institutions in the 1940s were the Lunda and Luvale Native Authorities. There can be little doubt that the creation of Native Authorities, which gave in theory but rarely in practice the 'tribes' a 'modern' administrative structure, was moderately useful in carrying out Indirect Rule.^[44] Yet the term coined was something of a misnomer; perhaps 'Native Responsibilities' would be more descriptive of their functions as they never had serious authority in making policy. While their creators saw the Native Authorities as institutions of modernization, their creation encouraged, or indeed forced, people to seek the solutions to local problems through the newly formulated tribal structure.

The Native Authorities were not merely the organizations through which disputes such as the Chavuma issue could be presented. They also offered to chiefs, headmen and the newly established Chiefs' Courts unprecedented power in legal matters, especially the right to collect fines in cash and kind. They thereby provided a bureaucratic 'class' or group with a new source of wealth and control. Ironically, the tribe became the very ideal of modernism, representing as it did 'modern' administration, and, through its Native Authority, access to the clear benefits to be had from western-style trading, agricultural improvements, transportation, medicine, and, most important of all, education.

Ethnic politics reached their fullest development in 1948 when government announced the beginnings of government-sponsored primary school education.^[45] Schools in Luvale areas were to teach in the Luvale language; schools in Lunda areas in the Lunda language. Mission schools had generally followed this pattern although there were important exceptions such as Chitokoloki which, while in a Lunda chiefs area, taught in Luvale. The government commitment to establish widespread primary education meant that schools would be established in Chavuma and therefore, following government policy, a single language had to be chosen. Since Chavuma, even though it was chiefless, was still administratively regarded as 'Lunda' following the long-standing policy of using the Zambezi river as the 'tribal' boundary, Lunda was chosen as the sole language of primary instruction.

The majority Luvale population believed that this decision would ultimately mean the installation of a Lunda chief as well. In 1949 violence between Lunda and Luvale in the form of the burning of houses and crops and the assaulting of people reached the point where the District Commissioner in Balovale declared a State of Emergency in Chavuma and summoned troops to re-establish order. Both Luvale and Lunda informants today agree that in the late 1940s it was not safe for a Luvale to use the road along the east bank of the Zambezi from Balovale to Chavuma.

It is difficult to imagine the importance local people attached to primary education in the 1940s. Even the most rudimentary education was crucial in a British system which gave enormous emphasis to the 'educated' and which required 'certificates' for virtually every employment in the new administrative sector. Those who became local teachers, school inspectors and teachers' assistants became the 'new men' who controlled access to education. This group, which had first gained education, became the opinion makers and even the heroes of Luvale and Lunda society, coexisting with the chiefs and hunter heroes of the past. Western education, acquired formally or informally, was also one of the gateways to the emerging capitalist economy. Of course it was possible to participate in the new economy as a producer, especially a small-scale producer. But the goal of the economically ambitious was to be a trader, and for this one needed literacy and a knowledge of the details of trade.

I am not certain how the school language problem was actually negotiated, but a compromise was reached which provided both Lunda and Luvale language schools. This solution avoided the question of when the Chokwe and Luchazi would also have their own schools. The language policies of the government merged in the general problem of 'control' over Chavuma, and since there was no easy, mutually acceptable solution at hand, it procrastinated. By the mid-1950s the Luvale, frustrated by broken promises and delays, actually installed a chief in Chavuma. The Lunda threatened to go to war if the government did not remove the illegal chief and again a State of Emergency was declared.

By the 1950s Luvale and Lunda tribalism was fully developed. It reached into every aspect of life and into every corner of the district and beyond. Lunda boycotted Luvale traders and vice versa. Travel in one another's territory was unsafe and the same people who had sat together to prepare evidence for the MacDonnel Commission and who had celebrated their joint victory over Lozi pretensions now no longer spoke to each other. Couples who had married across ethnic lines found themselves ridiculed by both sides and, at times, even forced by their families to divorce.^[46]

The government realized that without some kind of settlement at Chavuma, periodic outbreaks of violence would recur. In 1956 it decided, therefore, to seek a clever solution to the Chavuma problem by having the issue arbitrated by the Mwaant Yaav, the Paramount Chief of the Ruund in Zaire, the historical homeland claimed by both Luvale and Lunda.

Luvale and Lunda contact with the Ruund state of the Katanga region of southwestern Zaire had profound effects on the self-image of both groups. The ancient, highly centralized Ruund state was a perfect foil to the claims of the Lozi and others that Upper Zambezian societies were without 'proper histories and chiefs'. Here was a historical tradition to which both Luvale and Lunda had ancient claims even if their actual knowledge of the Ruund was piecemeal and rudimentary. In 1956 Luvale and Lunda delegations visited Musumba, the capital of the Mwaant Yaav, as a part of the arbitration process for Chavuma, and they later entertained his representatives in Balovale and Chavuma.

For his part the Mwaant Yaav was more than willing to cooperate with Belgian and British officials, not because he wished to lend an arbitrating hand to the Chavuma matter, but because he also was attempting to foster in the middle and late 1950s a pan-Ruund movement, a 'gathering-in of all of the peoples of the Ruund tradition. In this way Mwaant Yaav Ditend (Tschombe) hoped to lay foundations for his own expansive political ambitions of the 1960s.

Mose Sangambo, perhaps more than any other Luvale or Lunda, was intensely interested in history and perceived the political use to which a new kind of history could be put. Sangambo was elated by his visit to the Congo, first to Musumba and later to Inkalanyi, the remains of the ancient polities which predated the rise of the Ruund Empire and to which the earliest Luvale traditions are linked. There was a potential contradiction in Luvale identity with the Ruund: for the Ruund hegemony meant the suppression of the Inkalanyi polities. And it is to the Ruund that the earliest Lunda traditions are connected. But Sangambo and the Lunda representatives both preferred, though for somewhat different reasons, to see the totality of the historical traditions as common to both groups.

Sangambo and other Luvale opinion makers wished to amalgamate the prestige of the Ruund into their own current traditions and to adopt the symbols of Ruund statehood and political authority to elaborate the new political structures and realities within Balovale District. The Lunda, who are heirs only to later parts of the historical traditions—the period after the rise of Uruund—wished to be associated with the earlier parts of the tradition which would give them a kind of 'parity of antiquity' with the Luvale and thus preempt Luvale claims that their chieftainship was older than that of the Lunda and therefore, making a logical leap, that they had necessarily entered Chavuma first as the rightful 'owners of the land'. In both cases, each appropriated historical traditions which actually belonged to the other to manipulate the land dispute at Chavuma and to establish the basis for settling, on historical grounds, other potential conflicts in the district.

I do not wish, however, to represent all histoncal interests of the Luvale and Lunda, and especially those of Mose Sangambo, as being centred on Chavuma. Sangambo in particular has been collecting historical information for fifty years and, while he recognizes its utility in putting forward Luvale claims and views, he has also developed into a professional historian, as have many of his colleagues who, through their literacy and their exposure to modern historical writing, have attempted an unprecedented synthesis of Luvale history for future generations. This has necessarily involved interpretation, extrapolation from limited data and the informed judgments which all historians are forced to make. It has also meant that Sangambo and his colleagues have reconstructed traditions which they feel best represent their present and historical culture, and have also elaborated certain elements which respond to external or internally felt requirements. I hasten to note that the 'invention of tradition' is hardly restricted to the Upper Zambezi and that it plays an important role in articulating and directing opinion in most societies.^[47] Nor need I elaborate on the conflict of historical interpretation which generates the great arguments within the historical profession. What Sangambo and his contemporaries have written is unique only in the sense that we are unaccustomed to finding such research and dedication in rural Africa. Out of these concerns Sangambo has given the Luvale people their first comprehensive, albeit contentious, history—a history which elaborately sets out the origins of his tribe at Inkalanyi, the emergence of the Mwaant Yaavs and the rise of the Ruund Empire, and the subsequent evolution of Luvale history down to the present day. Thomas Chinyama attempted, in far shorter form and with far less research, to do the same for the Lunda.

To enhance the tribal identities which their histories described and to assert the vitality and modernity, as well as the antiquity, of their tribal structures the Luvale and Lunda adopted, or—as they see it—readopted in the 1950s new forms of ancient political symbols from the Ruund which they had lost or which had developed at Musumba since their migration from Uruund. These included the royal crowns (michama ), the royal executioner's sword (mukwale ), and the elaborate royal fences (lilapa ) which today surround the houses of the Senior Chiefs. The stability of titles which the colonial imperium imposed encouraged the building of permanent 'capitals'—in violation of ancient custom which decreed that the chiefs capital was destroyed at his death and moved to a 'clean' place.^[48]

While the visit to Musumba gave the Luvale and Lunda the opportunity to enhance their historical knowledge and to rediscover their origins, it provided no satisfactory solution to the Chavuma problem. Mwaant Yaav Ditend found his own 'pan-Ruund solution' by appointing his 'daughter' Luweji as chief at Chavuma. While this solution disappointed both Luvale and Lunda, it was nonetheless acceptable to the Northern Rhodesian government. Luweji was given Lunda and Luvale advisers to help her administration, but both sides eventually came to believe that she was actually administering Chavuma in the interests of her Chokwe husband.^[49] During the late 1950s and early 1960s this led to a three way struggle for power in Chavuma between the Chokwe, Luvale and Lunda, and to the ultimate breakdown of authority and the return to house burnings, school boycotts and petitions to the government. In 1963, with independence approaching, Luweji was deposed and returned to Zaire. Luvale and Lunda partisans both attempted to place a chief of their own in Chavuma. Both failed. The government again declared Chavuma a chiefless area and turned the problem over to the new Zambian government which today still faces the same dilemma as its colonial predecessor and has had the matter 'under study' for twenty years.

During this period there have been many attempts by both sides to install a chief. Their tenacity is not the atavistic response of tradition-bound people. More correctly, no one in the Chavuma area regards their interests as adequately protected by constantly changing and often indifferent civil servants. While the lack of a chief is still a very important local issue, it is essentially an idiom which masks the anxious desire of the local population for material improvements. Luvale and Lunda intellectuals and civil servants, some of whom now hold important national positions, realize that a major contributing factor to the lack of 'development' in this potentially rich agricultural area is the tribal strife and antagonism which lead the central government to doubt the wisdom of investing limited resources in an area where localism is likely to hinder the success of any project. At the same time the Upper Zambezi is at the end of the road—at least it was for decades during the war of independence in Angola and the civil war which followed. But the major locally perceived reason why the issue must be settled is that this would be a first step towards economic development, better school and medical services, and the creation of an infrastructure which would allow local farmers and fishermen greater participation in the national economy.

The Luvale History Project, 1938–1981

The Luvale History Project is something of a misnomer in that it had its roots in the 1930s but did not reach its current state of organization until the late 1960s. The original impetus to write a synthesis of Luvale oral traditions—or more correctly, the Luvale political traditions—came, as I have shown, out of the

struggle with the Lozi, the creation of a functioning political and administrative hierarchy and the beginnings of the Chavuma dispute. It received an enormous stimulus from the appearance in 1941 of Thomas Chinyama's Early History of the Balovale Lunda, from the social stigma that Luvale felt in the towns and the consequent lack of economic opportunity, and especially from those Luvale civil servants who after independence felt that the national Luvale social and economic position could be improved if the whole nation was as aware of their history as it was of the history of other ethnic groups.

For those Upper Zambezian people long resident in town during the colonial period, and perhaps even more strongly after independence, improving their 'tribal' image by making available to their fellow citizens books about their customs and their history—written almost always in English, for writing in Luvale would have been to lose their most important audience—was one method to create local pride and to counteract existing prejudices. The continuation of earlier stereotypes has stimulated—indeed commanded—Luvale intellectuals to reinterpret Luvale custom and history and thus give the concept of a Luvale or a Lunda tribe a modern, as well as an ancient, significance. It has also created the intellectual dilemma of attempting to show the vitality and variety of their 'tribal' life without threatening the often enunciated anti-tribal policies of the Zambian state.

This process has involved many individuals. During a long and difficult beginning, Senior Chief Ndungu Sakavungu and Mose Sangambo led the Luvale through the harrowing years of the Lozi and Chavuma disputes. Today it involves the Mwondela brothers, John and Willie, who began their careers as mission teachers and who are still important figures in both national and local politics. The Mwondelas became the driving force of the Luvale History Project and the source of funds and inspiration for the revival and elaboration of Luvale customs. John Mwondela is largely responsible for the revival of the Likumbi lyaMize (Mize Days), a ceremony held each August which recalls Luvale history and custom and demonstrates many of the old crafts and skills now rarely practised.^[50] Willie Mwondela has, from his positions as Zambian Ambassador to the People's Republic of China and currently the Republic of Kenya, written extensively on Luvale custom and tradition. James Chinjavata, originally a research assistant to C.M.N. White, participated in White's research into Luvale language, history and customs. These are but a few examples of the Luvale who have used their literacy to conserve and explain Luvale history to a broader world.

I have far less information about the Lunda, but they also seized the opportunities offered by the Plymouth Brethren mission schools and they too, like the Luvale, sent some of their 'best and brightest' to Chitokoloki, Chavuma and Sakeji schools. Among the currently important Lunda who attended Brethren schools are Samuel Mbilishi, who has served on the Central Committee of the United Independence Party (UNIP), and Dawson Muhongo, who worked first as a boma messenger and then for many years as a hotel waiter in Salisbury, and who was installed as Senior Chief Ishinde in the 1950s because he was considered an educated man who could 'deal' with Europeans. He forcefully directs Lunda affairs today as he has for the last thirty years. Then there is Thomas Chinyama, a Lunda with a prestigious Luvale name, who wrote the first Lunda history after the end of the Lozi dispute.

Sangambo and his colleagues, notably Nelson Cikomo, supported by the administrative and financial resources of local patriots and townsmen, had worked on the History Project since its beginning and collected substantial, if at times uneven, amounts of data. By the late 1960s what was needed was someone

to help rationalize the data and give it a publishable form. They initially found this aid in the person of Dr Arthur Hansen who, with his wife Dr Anita Spring, was then conducting anthropological research in the Chavuma area. Dr Hansen helped Sangambo to collect data among Luvale chiefs and provided an outline for the book. On the departure of Hansen and Spring in 1972, 1 arrived in Zambia to conduct historical research among the Luvale and was immediately enlisted into the project.^[51] My main contribution was providing Sangambo with transportation as I went about my own research; I also made a visit, with him, to the Ruund and to Inkalanyi in 1973. In 1979 Hansen and I edited Sangambo's material into a book and 500 copies were printed as Mose Sangambo's A History of the Luvale People and Their Chieftainship, with Hansen and myself as editors. The books arrived in Zambezi District in 1980 and were sold out in three weeks. Sangambo's history book had been, like Chinyama's forty years earlier, a bombshell in local politics.

In keeping with the gentle pace of local affairs in Zambezi, Sangambo's book was perceived in the first instance as a rebuttal to Thomas Chinyama's of 1941, which had elevated the Lunda over the Luvale chieftainship.^[52] Sangambo also made a vigorous historical assertion of Luvale claims to Chavuma. But it also transcends local disputes and sets out Luvale political history for all Zambians to read, hence his insistence that it be published in English. This brings to national attention rural and urban Luvale wishes for higher ethnic status, a greater role in Zambian affairs, and easier access to employment. And, perhaps most importantly, it is the gift of Sangambo and all of the people who worked with him over decades to future generations of Luvale and Zambians.

The Lunda in Zambezi District regard the book as the newest and perhaps most dangerous of all Luvale attempts to control Chavuma. They are not surprised by this, but they are profoundly offended by Sangambo's assertion that their Senior Chief Ishinde was only a headman when he 'left' Uruund, not a chief, and that he became a chief years later through his own initiative and not through actual genealogical connections to the Ruund royal family. In this view, he is seen as a jumped-up latecorner, brought into Northern Rhodesia at the time of the British restructuring of chieftainships, when compared with the antiquity of Luvale chieftainship and settlement in Chavuma.

Lunda reactions to these claims were direct and immediate. Years of accumulated ethnic separatism and tribalizadon surfaced at the appearance of Sangambo's book. Once again Luvale and Lunda dared not travel through each other's territory. The shops and 'tea carts' in the boma, now entirely locally owned, were strictly patronized by one or the other tribe, reflecting a hardening of long standing practices. Violence broke out again in Chavuma as Luvale and Lunda partisans especially identified with tribal politics were the victims of crop and house burnings. The roads leading to Chitokoloki Mission hospital were temporarily blocked as Lunda protesters sought to prevent the Luvale Senior Chief from attending a ceremony dedicating a new wing of the hospital.^[53]

At the provincial and national level the Lunda sought to have the book banned, confiscated and burned. The Luvale reacted with equal vigour, threatening to 'march to the President' if the book was banned and if my research, part of which was directed towards preparing a new, expanded edition, was prohibited. After weeks of meetings between the District Governor, an Mbunda man of inexhaustible patience and moderation, Luvale and Lunda delegations and myself, it was unanimously decided to permit my research, on a limited scale and in a very restricted part of Chavuma. The Lunda are still seeking a banning of the book; Sangambo remains adamant that what he writes is historically true—that it is only

a coincidence that it supports Luvale claims—and that he will not retract it publicly or in further editions.^[54]

The Lunda say they are beginning new historical research to give a 'clearer' idea of their past and to challenge elements in Sangambo's interpretation. This was one of the compromises reached in our meetings; that the proper answer to the Luvale claims was a new Lunda book. The Luchazi, whose history I had originally attempted to include in my own research and who had refused to be lumped together in a 'Luvale' book reproached me for their exclusion but have embarked on a book of their own.^[55] The Kaonde and Mbunda have also begun local history projects.

Conclusion

What I have tried to describe in this essay are the outlines of how two 'tribes', the Luvale and the Lunda, have arrived at their contemporary political structures and self-identities. Not only have the Luvale and Lunda taken up more centralized political structures with new symbols of power and authority and expanded a historical tradition from a local to a more universal level of interacting 'tribes' which explains and justifies these innovations, but unfortunately they have also come—at least at the local level—to regard these new forms of tribal identity both as exclusive and as their only effective way of asserting influence on local, provincial and national affairs. There is no question but that there are Luvale and Lunda tribes today and that tribalism—however that may be defined—is a central factor in local and national politics. But it is instructive to note that regionalism, ethnic separatism and movements that would be described in Africa as 'tribal' are widespread in some of the oldest western states, and that there is no necessary conflict which cannot be overcome in reconciling local cultural, linguistic and historical differences within the structure of the national state.^[56] This resolution certainly cannot be found by ignoring tribal differences; rather, it is essential to understand their historical evolution and meaning. One of the most important intellectual unifiers for the national state is its sense of common history. This does not mean that each ethnic group or region has shared the same experience but that for a multiplicity of historical reasons they now share borders and institutions which serve their citizenry. Once schools began to teach Central African or Zambian history the issue of ethnic representation became a crucial one. If the Luvale and Lunda, and the other peoples of the Upper Zambezi, find no place in their national school books and are required to learn the histories of other peoples whom they regard as competitors in the search for jobs, status, and economic opportunity, the idea of the 'History of the Nation' has little local meaning.

The Luvale and Lunda recognize to some degree that they have been tribalized in the negative sense. They face the dilemma of wanting to know and be proud of their local history, and to show both to succeeding generations and to the world at large how they have evolved as a society. At the same time they recognize that history cannot be solely concerned with local issues and that their 'tribal' consciousness as Luvale and Lunda must also make way for a national Zambian consciousness. The colonial period redefined the tribe and created for it what were called, at the time, modern institutions. If tribalism is the idiom through which local societies are still forced to participate in the modern political economy, Sangambo writes hopefully: 'We [Luvale, Lunda, Luchazi, Mbunda and Chokwe] were once brothers at Inkalanyi; we have separated to found different tribes but now we are coming back together again in our new Zambian nations.'^[57]