1. Over the years many people have contributed to my understanding of the history of Upper Zambezi societies, especially Mose Kaputungu Sangambo and all the Luvale, Luchazi and Lunda who volunteered their time and knowledge. I owe a special debt to Dr Wim van Binsbergen, with whom I discussed much of what is written here, and to Leroy Vail, who made many important suggestions for changes in the initial draft, not all of which I was able to incorporate.

I also wish to express my gratitude to the Fulbright-Hays Dissertation Fellowship Program; the National Science Foundation, Washington, D. C.; the Ford Foundation through the Center for African Studies of the University of California, Los Angeles; and WOTRO, the Netherlands Institute for Tropical Research, for supporting the field work upon which this article is based.

2. A broader description of (especially) Luvale history can be found in Robert Joseph Papstein. 'The Upper Zambezi: a history of the Luvale people, 1000-1900', unpublished Ph. D. dissertation, University of California, Los Angeles, 1978.

3. See C.M.N. White, 'Clan, chieftainship and slavery in Luvale political organization', Africa, 27, 1 (Jan. 1957), pp.59-75, and 'Luvale political organization and the Luvale lineage', Proceedings of the Thirteenth Conference of the Rhodes-Livingstone Institute (1959), pp. 113-120. In the Luvale language there is no distinction between 'clan' and 'tribe', the word muyachi being used for both.

4. For a description of the vitally important Chavuma area, see A. Hansen, 'When the running stops: the social and economic incorporation of Angolan refugees into Zambian border villages', unpublished Ph. D. dissertation, Cornell University, 1976. Also, C.M.N. White, 'Land tenure and village structure among the Luvale tribe', Man, 60 (1960). pp.103-112.

5. During our many field trips in Zambia and Zaire this was always the protocol used when we arrived as 'strangers' in a new village.

6. This is discussed in White, 'Clan, chieftainship and slavery', pp.59-75, and in Papstein, 'History of the Luvale people', pp. 14-15, 129-59.

7. These are dates arrived at by means of archaeological and lexicostatistical methods combined with oral data. See Papstein, 'History of the Luvale people', pp.69-76, 120.

8. C.M.N. White, 'The material culture of the Lunda-Lovale peoples', Occasional Papers of the Rhodes-Livingstone Institute, No.3 (1948), p.5.

9. See Papstein, 'History of the Luvale people'. Appendix 3, 120 word Diagnostic List: Cognition Percentages, pp.286-91.

10. See A. Delille, 'Besnijdenis bij de Aluunda's en de Aluena's in de streek ten Zuiden van Belgish Kongo (grensstreek Belgish Kongo-Angola)', Anthropos, 25, 5-6 (1930), pp. 881-8. Also White, 'Material culture', passim .

11. Testimony of Mose Kaputungu Sangambo, Mize Capital, 14 June 1973.

12. There is no historical evidence for a conflict between the Luvale and Lunda systems of production prior to their fairly recent competition for Chavuma. This competition appears to be largely based on the possibility of producing for the increased market in food and the largely unrealized potential for 'exporting' food to the Copperbelt.

13. C. M. N. White, "The role of hunting and fishing in Luvale society', African Studies, 15, 2 (1956), p.85.

14. Papstein, 'History of the Luvale people', p. 121; and R.M. Derricourt and R.J. Papstein, 'Lukolwe and the Mbwela of North-western Zambia', Azania, 11 (1977), pp.169-76.

15. Mose Kaputungu Sangambo, The History of the Luvale People and their Chieftainship, eds., A. Hansen and R. J. Papstein (Los Angeles, 1979), pp.35ff.

16. Slave trading caravans were still operating in the Upper Zambezi area as late as 1906. Zambia National Archives (ZNA), A/1/1/10, 'BSAC In-Letters', Selborne to High continue

Commission Office, Johannesburg, 15 Nov. 1909. See also D. Wheeler, 'The Portuguese in Angola, 1836-1891: a study in expansion and administration', unpublished Ph.D. dissertation, Boston University, 1963.

17. Testimony of Mose Kaputungu Sangambo, Mize Capital, 2 July 1973.

18. C.M.N. White, 'An outline of Luvale social and political organization', RhodesLivingstone Paper No. 30 (1960). This was confirmed by my own field work.

19. Papstein, 'History of the Luvale People', pp.237ff.

20. Sangambo, History, pp.35ff.

21. British South Africa Company claims to the territory of west central Africa were based on a series of agreements signed with Lewanika, King of the Lozi. It was therefore in Lewanika's interest to claim that his territories were far wider than was actually the case. The BSAC encouraged these extravagant claims for they gave the company the widest possible concession area.

22. This system of corvée labour, already existing in Bulozi itself, was never actually implemented in Balovale. For the Luvale and Lunda, however, it presaged Lozi administrative intentions.

23. This is described by Robert E. Schecter, 'History and historiography on a frontier of Lunda expansion: the origins and early development of the Kanongesha', unpublished Ph.D. dissertation, University of Wisconsion-Madison, 1976.

24. I have already indicated how Luvale chiefly titles proliferate in so far as all the children of female chiefs are entitled to the chiefly title, mwangana . The British colonial administration ignored the custom and did not end it. The Luvale responded by seeking out their most able vamwangana, moving them from chieftainship to chieftainship to give themselves the greatest leverage with the British. In one case a 'senior' chief was never actually installed and yet he functioned in that capacity because of his effectiveness in representing Luvale interests to the colonial government.

25. It is unclear why Bruce-Miller decided to use the river as a formal 'tribal' boundary. It probably stemmed from a misreading or hearsay report of the views of Livingstone, who met the Lunda chief Ishinde on the eastern bank of the Zambezi in 1854 and who was told that the Luvale lived on the other side. In any event, Bruce-Miller's attempt to 'tidy up' the ethnic boundaries of the district was the beginning of the Chavuma land dispute which has continued to the present day.

26. For the area in general, see S. Shaloff, "The Kasempa salient: the tangled web of British-Lozi relations', African Social Research, 5 (1972); C.M.N. White, 'Notes on the political organization of the Kabompo district', African Studies, 9 (1950), pp.185-93; Kusum Datta, 'The policy of indirect rule in Zambia (Northern Rhodesia), 1924-1953', unpublished Ph.D. dissertation, University of London, 1976; and B.C. Kakoma, 'Colonial administration in Northern Rhodesia: a case study of administration in the Mwinilunga district, 1900-1939', unpublished M.A. thesis, University of Auckland, N.Z., 1971.

27. White, 'Kabompo', pp. 185-93.

28. Many Luvale remember this time of Luchazi immigration into Northern Rhodesia. I was frequently shown 'Luvale' villagers who were, in fact, of Luchazi origin or who were known as 'children of the village', a form of debt peonage for having been taken in as refugees.

29. Admitting to being Mbwela—thus not of Luvale royal tradition—was not an admission easily made. However, very important Luvale identified themselves to me in this way as a means of clarifying for me the different socio-ethnic elements which today are subsumed under 'Luvale'.

30. There is actually a Ngangela Dictionary, by Emil Pearson (Morales, Mexico, 1970) in which the term includes five dialects of Luchazi, two of Mbunda, Niemba (or Western Ngangela) and Nkangala.

31. M. Gluckman, 'The role of the sexes in Wiko circumcision ceremonies', in M. Fortes, continue

ed., Social Structure (London, 1949), pp.145-67. See also Gluckman, 'Circumcision rites of the Balovale tribes', African Studies, 13, 2 (1954), pp.89-92; C.M.N. White, 'The Balovale peoples and their historical background', Rhodes-Livingstone Journal, 8 (1949), pp.26-42.

32. Papstein, 'History of the Luvale people', pp.286-8.

33. Most of the shops which still exist in Zambezi town, except for the parastatals, are owned by mission-educated Zambians.

34. I am reluctant to indicate specific names here for fear of imputing inaccurate views of participation in specific events for which I do not have conclusive evidence. However some names do appear in the text and indicate the process I am seeking to indicate.

35. Mose Kaputungu Sangambo, Muwema Toloshi Paciencia and Thomas Chinyama are such figures. They realized that if the uniqueness of Lunda and Luvale history could not be articulated it would be relatively easy for the combined interests of the Lozi and colonial officials to sub-infeudate them to the Lozi. The most recent discussion of this can be found in Mabel C. Milimo, 'Relations between the Lozi and the subject tribes and the colonial administration, 1890-1941', unpublished Ph.D. dissertation, Oxford University, 1981.

36. Virtually all former migrants I spoke with indicated that Angola and Zaire were destinations of the last resort; that South Africa offered the best salaries; and that the Copperbelt and Southern Rhodesia were second best.

37. Unfortunately, we do not have a clear idea of WNLA's operations in the Upper Zambezi. However, the organization scoured the region for workers, sending barges up the Lungwebungu river into Angola. It maintained permanent 'out-stations' on the western bank of the Zambezi together with cattle herds to feed recruits prior to trans-shipment. And there are still former WNLA buildings in Zambezi boma today. Recruits were taken by barge to Mongu or by road to the Copperbelt after 1941. In later years WNLA transported migrants by air. Interview with Johnston Ngona, WNLA recruiter, 6 Feb. 1973.

38. C. Mitchell, 'Distribution of African labour by area of origin on the copper mines of Northern Rhodesia', Rhodes-Livingstone Journal, 14 (1954), pp.30-42. In 1975, while teaching at the University of Zambia in Lusaka, I was told that Lunda and Luvale, with very few exceptions, could not aspire to what were regarded as better jobs because they were ethnically stigmatized. See also the essay in this volume by Brian Siegel for a discussion of the process of ethnic stigmatization.

39. Northern Rhodesia, Report of the Commission Appointed to Examine and Report upon the Whole Question of the Past and Present Relations of the Paramount Chief of the Barotse Nation and the Chiefs Resident in the Balovale District both East and West of the Zambezi River, with Special Reference to the Ownership of the Land and the Methods by which the Tribes have been Governed, and to Make Recommendations for the Future (MacDonnel Commission) (Lusaka, 1939). The full testimony can be found in ZNA ZP 5/1/A, ZP 5/1/5. See also ZP 1/1A/1B, ZP 5/2-4 and KDE 2/3/1-12.

40. Testimony of Mose Kaputungu Sangambo, 26 Feb. 1980.

41. Thomas Chinyama, 'The early history of the Balovale Lunda', typescript, n.d. (c. l941), pp.6-7.

42. The restriction of the offices of chief and headman to a few recognized political leaders and the policy of Indirect Rule augmented earlier political prerogatives by granting the right to influence and even decide who would benefit from the new opportunities offered by the emergent economy.

43. According to local testimony, there was chronic, small-scale, inter-tribal violence in Chavuma from the 1930s onwards, reaching a peak in the late 1940s and 1950s.

44. See Kakoma, 'Mwinilunga' and Datta, 'Indirect Rule', passim .

45. The influence of western-style primary education upon the creation of tribal consciousness is a main theme of David Wilkin, 'To the bottom of the heap: continue

educational deprivation and its social implications in the Northwestern Province of Zambia, 1906-1945', unpublished Ph.D. dissertation, Syracuse University, 1983.

46. This too was a recurring theme when discussing the Chavuma issue. How widespread this phenomenon was I cannot say. It was, however, one of the major domestic problems mentioned in connection with the dispute over the Chavuma area, especially for those people actually resident there.

47. See, for example, T.O. Ranger and E. Hobsbawm, eds., The Invention of Tradition (London, 1983), passim .

48. While the Luvale and the Lunda recognize that that it is no longer practically possible to move the capital at the death of a chief, Mize Capital is a testimony to how they managed to hold to the essence of the custom without wholly abandoning it. Formerly chiefs' houses were simply sealed up; later, only the room in which the chief had died was sealed. Most recently, the chief has been removed from the house prior to death so that the substantial investment a modern chief's house represents can be inherited by his heir.

49. Ultimately, the chiefs authority rests on the consent of the headmen and the population. Once both Luvale and Lunda decided that Luweji would not choose theirside and that she favoured the rapidly increasing Chokwe population, she was forced to abdicate.

50. I am now told that the Lunda have responded with a 'traditional' ceremony of their own—to demonstrate Lunda crafts and dancing, and to serve as a focal point of orientation for Lunda dwelling in the cities.

51. My role was originally to 'help' with the project while conducting research on a regional scale. But for reasons which are by now obvious, it proved impossible to work with both the Lunda and the Luvale.

52. Sangambo, History, p.32.

53. Testimony of the District Governor, Zambezi District, 20 Jan. 1981.

54. When I suggested to Sangambo that he might wish to delete the section on the Ishinde chieftainship (p.32) in the interests of local harmony, he declined.

55. I understand that Dr G. Kubik, of Doz University, Vienna, has recently produced a Luchazi history based on oral interviews, but I have not yet seen it.

56. For example, Welsh, Friesan and Occitan ethnic consciousness and languages coexist, occasionally abrasively, with the British, Dutch and French national states.

57. Sangambo, History, p.91. break