² For the African side, the standard authorities are Vansina, Jan, Kingdoms of the Savanna (Madison, 1966), 129–30, 134–8, 142–6Google Scholar, and Birmingham, David, Trade and Conflict in Angola (Oxford, 1966), 78–132Google Scholar. Delgado, Ralph, História de Angola, 4 vols. (Lobito, 1948–1955)Google Scholar, is the most extensive example of the Portuguese viewpoint. Nzinga's role as leader of Mbundu resistance appears, for example, in Rodney, Walter, ‘European Activity and African Reaction in Angola’, in Ranger, Terence O., ed., Aspects of Central African History (London, 1968), 56–9Google Scholar, Wheeler, Douglas and Pélissier, René, Angola (London, 1970), 39Google Scholar, and Davidson, Basil, In the Eye of the Storm (New York, 1973), 86, 91–2.Google Scholar

³ The primary sources for Nzinga present a number of difficult technical problems, as there is virtually no direct evidence on some of the most crucial episodes in her career. Official Portuguese correspondence touches on only the most formal aspects of her relations with Luanda, Lisbon, and Rome, and is usually quite uninformative about the questions raised here; much of this documentation has been published in Brásio, António, ed., Monumenta Missionária Africana—Africa Ocidental (Lisboa, 1952–1957, 11 vols)Google Scholar. A number of seventeenth and eighteenth century Portuguese chronicles deal with Nzinga, but most of their data can be traced back to the accounts of CapuchinE missionaries who attended her in Matamba from 1654 to 1663; their version was first presented in Montecuccolo, Giovanni Antonio Cavazzi da, Istorica Descrizione de'tre Regni Congo, Matamba, et Angola … (Bologne, 1687)Google Scholar, and is now available together with selected contemporary documents in the excellent edition by de Faria, P. F. Leite and de Leguzzano, P. Graciano Maria, Descrição histórica dos tres reinos do Congo, Matamba, e Angola (2 vols) (Lisboa, 1965)Google Scholar. The Capuchine perspective was strongly coloured by Nzinga at a time when she was seeking close relatins with the missionaries and with the Portuguese, some forty years after many of the events described. The veracity of this account may be doubted in several important respects.

⁴ The discussion is necessarily general, owing to the unavailability of specific details. The background material on Mbundu traditions and social and political structures was obtained in Angola during 1969 and 1970. This research was supported by a fellowship awarded by the Foreign Area Fellowship Programme of New York. The Programme has no responsibility for any of the conclusions expressed herein. A fuller treatment both of the dynamics of Mbundu political development and of the internal history of the ngola a kiluanje state will appear in my Kings and Kinsmen: Early Mbundu States in Angola (Clarendon Press, forthcoming).

⁵ Cuvelier, J., Koningen Nzinga van Matamba (Brugge, 1957), 42.Google Scholar

⁶ Brásio, , Monumenta Missionária Africana, VII, 417–20, 426–9.Google Scholar

⁷ Cavazzi, , Descrição histórica dos tres reinos, II, 64–70Google Scholar, for Nzinga's version of the story. No evidence adequately defines her pre-1622 position at the Ngola court; it seems she must have already been an important and ambitious personage, though without the status conferred by a position in the hierarchy of state titles.

⁸ The first evidence of Nzinga's alleged visit to Luanda dates from her old age, and the first colourful Portuguese descriptions appeared in print even later (not until the 1680s). See, for the latter, de Cadornega, António de Oliveira, História Geral das Guerras Angolanas (ed. Delgado, José Matias, 3 vols) (Lisboa, 1940–1942, reprinted 1972), I, 112–16Google Scholar. Birmingham, , Trade and Conflict, p. 89Google Scholar, implicitly notes the suspicious lack of contemporary documentation. A later Portuguese chronicler, Corrêa, Elias Alexandre da Silva, Historia de Angola (Lisboa, 1937, 2 vols), I, 231–2Google Scholar, betrayed his uneasiness with her version of events, owing to inconsistencies in her story that at once justified her hatred for her brother, the mbande a ngola, by recalling that the king had murdered Nzinga's son, but simultaneously claimed that he had entrusted her with so critical a mission as negotiating peace terms with the Portuguese.

⁹ A fact intimated by Nzinga herself to Cavazzi, her missionary confidant and adviser; Cavazzi, , Descric¸āo histórica dos tres reinos, I, 259, II, 70.Google Scholar

¹⁰ Brásio, , Monumenta Missionária Africana, VII, 248–50.Google Scholar

¹¹ Miller, Joseph C., ‘The Imbangala and the Chronology of Early Central African History’, J. Afr. Hist. XIII, 4 (1972), 549–74.CrossRefGoogle Scholar

¹² Miller, , Kings and Kinsmen, chapter VIIIGoogle Scholar, provides documentation for the structure and ideology of the Imbangala war bands.

¹³ The data on Nzinga and the Imbangala are scattered through Cadornega, , História Geral, I, 142–66 and 207 ffGoogle Scholar. Also see documents in Brásio, , Monumenta Missionária Africana, VII, 549–50, 526–9, and VIII, 91–108.Google Scholar

¹⁴ The historical account of ancient female rulers in Cavazzi, , Descrição histórica dos tres reinos, I, 22, and II, 79Google Scholar, might be suspected on grounds of having come from Nzinga, but it receives independent confirmation from sixteenth-century records in Brásio, , Monumenta Missionária Africana, II, 477–81.Google Scholar

¹⁵ Cadornega, , História Geral, I, 194–211Google Scholar; Cavazzi, , Descrição histórica dos tres reinos, I, 21, and II, 79–81.Google Scholar

¹⁶ Pieter Mortamer, report published in Naber, S. P. I'Honore, ‘Nota van Pieter Mortamer over het gewest Angola, 1643’, Bijdragen en Medeelingen van het Historisch Genootschap gevestigd te Utrecht, LIV, (1933), 1–42.Google Scholar

¹⁷ The treaty is published in Cavazzi, , Descrição histórica dos tres reinos, II, 332–3Google Scholar, together with several related documents. For repeated inferences that discord existed between Nzinga and her followers by 1648–50, see the background in Cadornega, , História Geral, II, 129, and 499–501Google Scholar (note), and Corrêa, Silva, Historia de Angola, I, 266.Google Scholar

¹⁸ The terms of the 1656 treaty were not unusual in seventeenth-century Angola, but this fact does not diminish the force of the argument, since the typical treaty introduced just the sort of Portuguese influence in African polities that Nzinga seemed to welcome.

¹⁹ A crucial aspect of the negotiations preceding the 1656 treaty may have been her offer to name as heir a sister, Dona Barbara, who had lived as a Catholic with the Portuguese for more than three decades; her proposal explicitly excluded from consideration her mercenary following (termed ‘slaves’). Cadornega, , História Geral, II, 502 n.Google Scholar

²⁰ Repeated attempts to locate an oral record of Nzinga in the Mbundu traditions I recorded in 1969 failed utterly, despite easy identification of the ngola a kiluanje, the mbande a ngola, the kaza, the kazanje, and other contemporaries with whom she is known to have interacted.

²¹ For evidence that the nzinga a mona had been the leader of a strong anti-Nzinga party even before her death, consult a revealing instance of a political defeat suffered by Nzinga in the early 1660s; Cadornega, , História Geral, II, 188–90.Google Scholar

²² Cadornega, , História Geral, II, 219–23, 254–8, 295–7, and 354–8.Google Scholar

²³ Intermittent Portuguese victories over Nzinga's successors in 1681, between 1709 and 1713, and about 1738 only temporarily slowed the growing isolation of these kings; see Cadornega, , História Geral, II, 402–4, and 555–7 nGoogle Scholar. and Corrêa, Silva, Historia de Angola, I, 335, 362–5.Google Scholar

²⁴ The absence of direct evidence inevitably lends a speculative quality to some of the arguments offered here, and so their validity depends on the success of the main hypothesis in explaining the known facts about Nzinga's life. I am, of course, aware that some readers may view this technique as a shortcoming and would urge them not to lose sight of the tentative nature of my conclusions.