1. For the area of Interlacustrine Africa, see, for example, Cohen, D. W., “A Survey of Interlacustrine Chronology,” JAH 11 (1970), 177–201CrossRefGoogle Scholar; Webster, J. B., ed., Chronology, Migration, and Drought in Interlacustrine Africa (New York, 1979)Google Scholar; Henige, David, The Chronology of Oral Tradition: Quest f or a Chimera (London, 1974)Google Scholar; idem., “Oral Tradition and Chronology,” JAH 12(1971) 371-89; idem., “Reflections on Early Interlacustrine Chronology: An Essay in Source Criticism,” JAH 15 (1974), 27-46. The comprehensive bibliography on method formerly produced annually in History in Africa includes discussions on issues relating to comparative chronology.

2. The starting point for the chronology of Rwandan dynastic history is Kagame, A., La Notion de génération appliquée la généalogie dynastique et à l'histoire du Rwanda dès Xè-XIè siècles à nos jours (Brussels, 1959).Google Scholar For elaboration, see his Inganji Karinga (Butare; 1959 [1943; 1947]); Un abrégé de l'ethno-histoire du Rwanda (Butare, 1972)Google Scholar; “La chronologie au Burundi dans les genres littéraires de l'ancien Rwanda,” Etudes Rwandaises 12 (1979), 1–30.Google Scholar The most important reassessments of Kagame's proposals are Vansina, Jan, L'évolution du royaume Rwanda dès origines à 1900 (Brussels, 1962)Google Scholar; Rennie, Keith, “The Precolonial kingdom of Rwanda: A Reinterpretation,” in Transafrican Journal of History 2 (1972), 11–53Google Scholar; and Nkurikiyimfura, J.-N., “La revision d'une chronologie: le cas du royaume du Rwanda,” in Sources orales de l'histoire de l'Afrique, ed. Perrot, C.-H., Connin, G., and Nahimana, F. (Paris, 1989), 149–80.Google Scholar See also van Noten, F. L., Les tombes du roi Cyirima Rujugira et de la reine-mère Nyirayuhi Kanjogera: Description archéologique (Tervuren, 1972)Google Scholar, and the review of this work by Vansina, Jan in IJAHS 7(1975), 528–31.Google Scholar

3. The best known of these was Pagès, A., Au Ruanda, sur les bords du lac Kivu (Congo Belge). Un royaume hamite au centre de l'Afrique (Brussels, 1933).Google Scholar But such observers were not in a position to have access to deep knowledge of Rwandan historical sources, and this seems to be the thrust of Kagame's implicit critique of such works. In fact Kagame seldom explicitly referred to the earlier works on Rwandan history by Europeans; the two exceptions are Notion de génération, 92-112; and “La documentation du Rwanda sur l'Afrique Interlacustre des temps anciens” in Centre de Civilisation Burundaise, La Civlisation ancienne des peuples des Grands Lacs (Paris/Bujumbura; 1981), 300–30.Google Scholar Nonetheless, his own proposals leave certain questions unresolved, partly because his technique also obviated the analysis of his primary sources—their divergences and ambiguities, their social provenance, and the conditions of their recording. His technique was often to give a short extract from the data, and promise a subsequent, and more complete, presentation of these data. Lacking such elements, we will probably never entirely resolve these issues. But that does not mean the debate is closed; this inquiry touches on a few of these issues.

4. On Kagame see Vidal, C., “Alexis Kagame entre mémoire et histoire,” HA 15 (1988), 493–504Google Scholar; and Chrétien, J.P., “Confronting the Unequal Exchange between the Oral and the Written” in African Historiographies, ed. Jewsiewicki, Bogumil and Newbury, David (Beverly Hills, 1986), 84–87.Google Scholar

5. A. Pagès, Royaume hamite, and an unnamed official writing in the Territorial archives of Kibungu, both cited in d'Arianoff, A., Histoire des Bagesera, souverains du Gisaka (Brussels, 1952), 38.Google Scholar The Territorial Archives were transferred to Kigali in 1956, and apparently never returned to Kibungu; at any rate, they are not accessible at present. Luc de Heusch accepts d'Arianoff's kinglist, but notes the chronological incompatibility with Rwandan claims to contemporaneous status with Gisaka kings, as asserted by Kagame and d'Arianoff: de Heusch, , Le Rwanda et la civilisation Interlacustre (Brussels, 1964), 46Google Scholar; see also ibid., 100-01, 111-12.

6. d'Arianoff, , Bagesera, 36–37Google Scholar: “Les notables banyagisaka [donnent] libre cours à leur fantaisie, se contredire les uns les autres, et se contredire eux-màmes…Si nous gardens quelque espoir …c'est surtout grace aux précieux moyens de controle qui nous a été fournis par la récente mis à jour de la chronologie dynastique rwandaise.” This assumption was restated in Kagame, , Notion de génération, 98.Google Scholar

7. d'Arianoff, , Bagesera, 38.Google Scholar

8. Ibid., 36, with emphasis added.

9. The most elaborate exegeses on this episode are Kagame, Inganji Karinga II, Section 6; and Kagame, A., Abrégé, 57–61.Google Scholar Other versions appear in Pagès, , Royaume hamite, 114–20Google Scholar, and d'Arianoff, , Bagesera, 50–53.Google Scholar

10. Ibid., 54-56. It is worth pointing out that both are suspect: the tie-in to Mutabazi would correspond to the original Kimenyi were Kagame's two additions to be deleted; the second tie-in (that of Kigeri Nyamuheshera), also to a Kimenyi, is suspect on internal grounds, to be discussed below.

11. On the importance of Kagame's influence, see Vidal, “Alexis Kagame,” and Chrétien, , “Unequal Exchange,” 84–87.Google Scholar

12. d'Arianoff, , Bagesera, 15.Google Scholar

13. Ibid., 16.

14. Freedman, James, “Ritual and History: The Case of Nyabingi,” Cahiers d'Etudes Africaines 14 (1974), 174CrossRefGoogle Scholar; note that of the cases cited, only Rwanda had a “repeating cycle” of names. Freedman, “Three Muraris, Three Gahayas, and the Four Phases of Nyabingi” in Webster, , Chronology, 175–85Google Scholar, develops new dimensions to this theme, but adds little to the interests that engage us here.

15. Freedman, , “Ritual and History,” 174.Google Scholar But note two aspects to this argument: that names are assumed to recur in all (“the,” not “some”) interlacustrine kingdoms; and that this feature of repeating dynastic names becomes transformed to a “cycle” of kingship names; in fact, very few kingdoms adhered rigidly to such a cyclic pattern.

16. Ibid.

17. Ibid. And in most cases, where there are gaps in chronology or content these are filled in by reference to Kagame's kinglist (ibid., 175-78), even to the point of denying local traditions to the contrary.

18. Vansina, Jan, “Note sur la chronologie du Burundi,” Bulletin d'ARSOM 38 (1967), 431.Google Scholar This pattern and its emergence is also considered in Chrétien, J.-P., “Du hirsute au hamite: les variations du cycle de Ntare Rushatsi, fondateur du royaume du Burundi,” HA 8 (1981), 3–41Google Scholar; idem., “Les traditionnistes lettrés du Burundi à l'école des bibliothèques missionnaires (1940-1960),” HA 15 (1988), 407-30.

19. Kagame, Notion de génération; idem. Abrégé; and idem., La poésie dynastique du Rwanda (Brussels, 1951) are the most relevant sources.

20. Kagame, A., “La Chronologie du Burundi dans les genres litteraires de l'ancien Rwanda,” Etudes Rwandaises 12(1979), 1–30Google Scholar; Vansina, Jan, La légende du passé (Tervuren, 1972), 192.Google Scholar On tie-ins see Mworoha, E., ed., Histoire du Burundi: dès origines à la fin du XIXè siècle (Paris, 1987), 131–34Google Scholar; Vansina, , “Notes sur la chronologie,” 432, 434Google Scholar; Kagame, , Notion de génération, 46n2.Google Scholar

21. On Baranyanka see Chrétien, , “Du hirsute au hamite,” 16–18Google Scholar; Chrétien, , “Unequal Exchange,” 83–84Google Scholar; idem., “Nouvelles hypothèses sur les origines du Burundi: les traditions du nord” in L'arbre-mémoire: traditions orales du Burundi, ed. L. Ndoricimpa et C. Guillet (Paris/Bujumbura; 1984), 12-15; and idem., “Traditionnistes lettrés,” 410, 416, 421, 424.

22. The best examples of this are found in Vansina, La légende du passé; Chrétien, “Du hirsute au hamite;” and idem., “Nouvelles hypothèses.”

23. Vansina, , Légende du passé, 192Google Scholar; Chrétien, “Traditionnistes lettrés.”

24. Kagame, , “Chronologie du Burundi,” 24–25Google Scholar, also notes that one of the most influential of the early accounts on the Rundi royal dynasty—that of Mgr. Gorju, Julien, Face au royaume hamite du Ruanda. Le royaume frère de l'Urundi (Brussels, 1938)Google Scholar—was premised on the assumption that the line descended from the Bacwezi, the rulers of an extensive domain some 200 years ago; this excluded a four-cycle regnal history. I concur that Gorju's assumptions were erroneous, but this in itself does not invalidate the possibility of eight reigns in this dynasty. Ironically, Kagame adopts a similar logic—that external tie-ins set the framework for internal analysis and that “fit” becomes the most essential criterion in historical reconstruction; he just uses a different, though no less dogmatic, criterion for setting up an external grid within which Rundi history is assumed to have operated. For a wider view of regional chronology see Kagame, “Documentation du Rwanda;” and Webster, J. B., “Noi! Noi! Famines as an Aid to Interlacustrine Chronology” in Webster, , Chronology, 1–39.Google Scholar

25. Kagame, , “Chronologie du Burundi,” 19.Google Scholar

26. Curiously, Kagame justifies this silence in the Rundi kinglist by reference to a Rwandan tradition, by which he maintains that genealogical data before Cyilima Rujugira (whom he cites as the contemporary of the two successors to the third Ntare) were no longer retained: ibid., 22; idem.,“Code ésotérique de la dynastie du Rwanda,” Zaire 1 (1947), 377-78; d'Hertefelt, M. and Coupez, A., La royauté sacrée de l'ancien rwanda (Terveruen, 1964), 94–154.Google Scholar

27. Kagame, , Notion de génération, 46n1, wim emphasis added.Google Scholar

28. Kagame, “Documentation du Rwanda.”

29. Here Kagame seems to overlook his own argument that the recent cyclic pattern in Rwanda was set in place by a personal decision of Mutara Semugeshi, a king who arbitrarily abolished three dynastic names: that of his grandfather (Ndahiro), that of his father (Ruganzu, the epic hero of the Rwandan dynasty), and even his own dynastic name (Nsoro). It seems curious that he would thus cast doubt on his own legitimacy by removing the names of his two predecessors—not to mention his own royal persona—from the pool of dynastic names. Kagame, , “Code ésotérique,” 377, 377nn23, 24.Google Scholar Also idem., Notion de génération, 11n2; idem., “Chronologie du Burundi,” 18; idem., Abrégé, I, 113-14; Vansina, , Evolution, 69–70Google Scholar, attributes this to the reign of Rujugira. The principle is summarized in Hertefelt, /Coupez, , Royauté sacrée, 478Google Scholar (“Noms Dynastiques”). Curiously, by 1900 the Rwanda kinglist recorded only a maximum of three cycles of dynastic names (and one of those is less than certain); yet Kagame insists on four cycles—invariably—for Burundi. But if Rujugira was a usurper, as Vansina maintains, there remain—ironically—only two complete cycles for Rwanda back from 1900: the very number Kagame rejects for Burundi by reference to Rwanda!

30. Vansina, “Notes sur la chronologie;” Mworoha, , Histoire du Burundi, 131–33Google Scholar; Chrétien, “Du hirsute au hamite;” idem., “Nouvelles hypothèses;” idem., “Traditionnistes lettrés.”

31. Kagame, , Notion de génération, 98n1.Google Scholar

32. Such transformations bear close parallels with those portrayed elsewhere in the region: in Bunyoro, Sir Tito Winyi, hitherto known by the dynastic name of Winyi II, in 1934 suddenly became transformed, “in a trice,” to Winyi IV; Henige, , Chronology, 105–14.Google Scholar See also Henige, “Reflections.”

33. For further discussion of this prevalent assumption see Newbury, David, “Bushi and the Historians: Historiographical Themes in Eastern Kivu”, HA 5 (1978), 131–51.Google Scholar

34. Kagame, , Inganji Karinga VII, #61-68Google Scholar; idem., Poésie dynastique, 37-38; idem., Les milices du Rwanda précolonial (Brussels, 1963), 24; idem., Abrégé, 87-91; Vansina, , Evolution, 86–87.Google Scholar

35. Ibid., 87.

36. For an analysis of these traditions see Newbury, David, Kings and Clans (Madison, 1991), 126–42.Google Scholar

37. Kagame, , Inganji Karina, II, #68Google Scholar; Delmas, L., Les généalogies de la noblesse (Batutsi) du Rwanda (Kabgayi, 1950), 28–29Google Scholar; Kagame, , Milices, 23–24Google Scholar; idem., Abrégé, 91; idem., Poésie, 37-38.

38. Delmas, , Généalogies, 47–49Google Scholar; there are may other instances in Delmas that could be drawn on to make the same point—that commoner genealogies are almost invariably shorter than the royal genealogy.

39. Viane, L., “Histoirque des Bahunde,” (Bukavu, typescript; n.d.).Google Scholar To be sure, the Bunyungu kinglist is suspect. But what is significant here is that even were the list fabricated by reference to other kinglists as then available, it speaks to a shorter Rwandan list at the time of its creation, since the tie-in to Rwanda is precise.

40. Kagame explains the discrepancy of commoner to royal genealogies by reference to the mechanism noted above (note 29, and “Code ésotérique,” 386), by which commoner ancestors before Rujugira were no longer remembered because their spirits could no longer harm those descendants who failed to honor their memory. But it is important to note that this practice was effective only by the decree of a Rwandan monarch. Therefore, while an interesting argument for those families close to the royal line, it is unlikely to apply to those areas incorporated into Rwandan state structures only after the time of Rujugira. It certainly does not apply to those areas outside Rwanda, such as Buhavu and Buhunde.

41. [R. P. Feys], “Histoke des Bahavu” (unpublished manuscript: White Fathers Archives, Bukavu; n.d.).

42. The nearest mission to Buhavu was at Mwanda, near Karhana, south of Buhavu. From this mission the priests visited Mpinga, the heartland of the Buhavu kingdom.

43. On the Ijwi kings see Newbury, , Kings and Clans, esp. 178–99Google Scholar; idem., “Kamo and Lubambo: Dual Genesis Traditions on Iwji Island (Zaire)” (M.A. thesis, University of Wisconsin-Madison, 1977).

44. Newbury, , Kings and Clans, 166–77Google Scholar; Feys, “Histoire des Bahavu.”

45. Newbury, Catharine, The Cohesion of Oppression (New York, 1988)Google Scholar, was based on this research, although the interview data cover greater historical breadth than what appears in the published work.

46. I consider this unlikely since none of these names are remembered as former Havu kings in the oral sources extant today on Ijwi or in Mpinga, the current center of the senior Havu line. It remains possible that these names are drawn from the dynastic line of one of the other small polities west of the Nile-Zaire Divide, a social formation no longer extant in the published soruces.

47. Vansina, Rennie, and Nkurikiyimfura would date this to a seventeendi-eighteenth century cycle. Their dates are as follows: Kagame: Ruganzu (1510-1543), Rujugira (1675-1708); Vansina: Ruganzu (1600-1624), Rujugira (1744-1768); Rennie: Ruganzu (1603-1630), Rujugira (1756-1765); Nkirikiyimfura: Ruganzu (1600-1623), Rujugira (1731-1769).

48. Kagame, Poésie dynastique; idem., Abrégé, 118-22; idem., Milices.

49. Vansina, , Evolution, 87.Google Scholar

50. Emmanuel, Ntezimana, “Coutumes et traditions des royaumes Hutu du Bukunzi et du Busozo,” Eludes rwandaises 13 (1980), 15–39Google Scholar; idem., “L'arrivée des Européens au Kinyaga et la fin des royaume Hutu du Bukunzi et du Busozo,” Etudes rwandaises 13(1980), 1-29; de Lacger, L., Ruanda (Kabgayi 1961 [1939]), 82–89Google Scholar; Newbury, , Cohesion of Oppression, 39.Google Scholar The ritual autonomy of Bukunzi—including a ritual attack by the Rwandan court—is discussed in d'Hertefelt, /Coupez, , Royauté sacrée, 151, 277-79, 449–50.Google Scholar

51. Lacger, , Ruanda, 109Google Scholar; Kagame, , Poésie dynastique, 41, 121.Google Scholar

52. Sigwalt suggests that instead of direct Rwandan involvement, Shi pretenders usurped the throne of Ngweshe, perhaps with tacit Rwandan support. Sigwalt, R. D., “The Early History of Bushi: An Essay in the Historical Use of Genesis Traditions” (Ph.D., University of Wisconsin, 1975), 65–66.Google Scholar But this interpretation differs from Rwandan conquest, as claimed in the Rwanda royal sources: Kagame, , Poésie dynastique, 41.Google Scholar Furthermore, claims in Rwandan sources to Nyamuheshera's attacks on the kings on Ijwi are clearly specious: see Newbury, , Kings and Clans, 126–42.Google Scholar This casts doubt on the validity of similar claims to Nyamuheshera's exploits elsewhere in the region. By contrast, Sigwalt suggests considerable Rwandan influence in this area during Gahindiro's reign; Sigwalt, , “Early History of Bushi,” 72.Google Scholar See also Masson, P., Trois siécles chez les Bashi (Bukavu, 1970)Google Scholar

53. Kagame, , Poésie dynastique, 41, 121.Google Scholar

54. Newbury, , Kings and Clans, 65-80, 156–65.Google Scholar The anachronistic nature of this claim is shown by the fact that it reappears as a rationale for Rwabugiri's attack of the late nineteenth century; for a discussion of this episode see Newbury, , “Rwabugiri and Ijwi,” Etudes d'Histoire Africaine 7 (1975), 155–73.Google Scholar

55. Kagame, , Poésie dynastique, 129–30Google Scholar; 92 and 121 add little; 102 refers to his western conquests, but Kagame's reference to Nyamizi is clearly an anachronism which applies to Kigeli Rwabugiri. On Rwabugiri's campaigns in this area see Newbury, D., “Les campagnes du Rwabugiri”, Cahiers d'Etudes Africaines 14 (1974), 181–91CrossRefGoogle Scholar; idem., “Rwabugiri and Ijwi.”

56. For an account of Rwabugiri's battles on Ijwi see ibid. On the Rwandan traditions of Nyamuheshera see Kagame, , Poésie dynastique, 41, 121Google Scholar; Nyamuheshera's attacks on Ijwi are not mentioned in Kagame, Abrégé. For fuller consideration of these traditions see Newbury, , Kings and Clans, 126–42.Google Scholar

57. For another example of this see Kagame, , Le code des institutions politiques du Rwanda précolonial (Brussels, 1952), 59–60n34.Google Scholar

58. Kagame, , Abrégé, 121Google Scholar, where he notes that Bwanacyambwe was retained by Rwanda from these wars. Vansina, suggesting that this area was seized during a war between Gisaka and Ndorwa rather than as a result of a specific Rwandan campaign, uses this as an example of the general tendency to inflate military claims in Rwandan official dynastic traditions, and he refers specifically to the Ndorwa “wars” of Kigeli Nyamuheshera as illustrative of this tendency; Vansina, , Evolution, 75-76, 87.Google Scholar

59. Kagame, , Milices, 66.Google Scholar Similarly Pages includes no information on Nyamuheshera or his successor, Mibambwe Gisanura, though he includes them both on his kinglist; Royaume hamite, 136. Nonetheless, it is interesting that Pagès notes Kigeri Ndabarasa, four reigns later, as Kigeri II, not, as Kagame asserts, Kigeri III; ibid., 143.

60. Kagame, , Milices, 66.Google Scholar

61. Kagame, , Abrégé, 122.Google Scholar

62. Ibid., 121.

63. Delmas, , Généalogies, 54–56.Google Scholar

64. Ibid., 55-56.

65. In Poésie dynastique, 40, Kagame does mention some minor campaigns, however.

66. Delmas, , Généalogies, 55–56.Google Scholar This is not impossible; after all, Nzuki did descend from Ruganzu according to this genealogy. But this group also descends from a more recent king, Semugeshi. The implication here is that Nzuki was not a lineal descendant, but a junior brother, of Semugeshi—and that the same process, counting intra-generational seniority as inter-generational lineality, could be extended to others. In addition to its chronological implications, this reference of course also brings into question the claim of Semugeshi to the status of king. Rather, he may have been an eminent authority associated with kingship over a long period, and only later elevated to that status to complete the cycle. We have already mentioned the bizarre assertion that he abolished the royal names of Ruganzu, Ndahiro, and of his own reign from the cycle of dynastic names.

67. Kagame, , Abrégé, 117, 118Google Scholar; idem., Milices, 61ff. Others note that this position was most likely first introduced under Cyilima Rujugira: Delmas, , Généalogies, 67, 93Google Scholar; Vansina, , Evolution, 70.Google Scholar

68. Kagame, , Abrégé, 106, 117Google Scholar; de Heusch, Rwanda.

69. Kagame, , Milices, 62.Google Scholar

70. Ibid., 61.

71. Ibid., 56, 61; idem., Abrégé, 118.

72. The concept of the umutabazi, or liberator-hero, is an institution in the Rwandan traditions, designating a warrior who is martyred for the cause of Rwanda; his death, according to the augurs, assures the victory of Rwanda in war: Kagame, , “Code ésotérique,” 368n8.Google Scholar

73. Delmas, , Généalogies, 108.Google Scholar

74. Kagame, , Poésie dynastique, 157, 158Google Scholar; Kagame informs us that Poem 49 “fait allusion à un retour en armes;” emphasis added.

75. Kagame, Milices; the histories of those armies formed under Ruganzu are found on 54-60.

76. Delmas, , Généalogies, 185.Google Scholar

77. Kagame, , Milices, 56.Google Scholar

78. Kagame, , Notion de génération, 47Google Scholar: “L'air des monarchies situees dans le Nord-Est [du Rwanda] (Nkole, Buganda, etc.) admettent l'intronisation des frères, des oncles et des cousins du monarque décédé…Les dispositions du ‘Code ésotérique’ nous ont épargné cette éventualité en ce qui concerne le Rwanda. Nos monarques ont regné de père en fils, d'une manière ininterompues.” On Rwaka Karemera, compare Kagame, , Abrégé, 133–34Google Scholar, with Vansina, , Evolution, 51.Google Scholar This situation recalls the manner in which Rutalindwa was removed from the royal genealogy when overthrown by his uncles, who then invested his half-brother, Yuhi Musinga, in the succession struggle to follow Rwabugiri's death in 1895. Rutarindwa does not appear on Kagame's list of kings, even though he was enthroned before the death of Rwabugiri as co-ruler. Kagame, , Milices, 87, 89.Google Scholar

79. See Vansina, Jan, “The Power of Systematic Doubt in Historical Enquiry,” HA I (1974), 109–27.Google Scholar

80. It has been argued—most forcefully in Vansina, JanDe la tradition orale (Tervuren, 1961)Google Scholar and in many works by Kagame—that the controlled tramission of Rwandan royal traditions made them more reliable than other oral sources. However, with increasing awareness of the powerful role of cultural hegemony and the role of ideology in historical understanding, and with greater sensitivity to how the power of knowledge accompanies the knowledge of power, these assumptions have been increasingly challenged. Greater control over the transmission of esoteric traditions may serve to reinforce ideological assertions of the past rather than to conserve the “facts” in some pure positivistic form. In fact, removing such historical debates from the public realm may be one of the essential techniques of defining history according to the perspectives of the ruling elite. This appears to be less frequent in societies without a written tradition—or one which abjures the use of writing in political contexts; see Ewald, Janet, “Speaking, Writing, and Authority: Explorations in and from the Kingdom of Taqali,” Comparative Studies in Society and History 30 (1988), 199–227.CrossRefGoogle Scholar But it is not unknown: Rwanda provides an excellent example of how esoteric—ideological—history is captured through the use of esoteric traditions. The claim that controlled transmission assures historical fidelity is manifestly false; it only makes for easier manipulation of sources, whether through deliberate distortion or (more likely) unconscious alteration in the name of ideolgical/hegemonic coherence. The cases of Gisaka, Ndorwa, and Burundi, discussed above, provide sufficient illustration of the rationalizations for manipulation.