Paths in the Rainforests: Toward a History of Political Tradition in Equatorial Africa

Notes

Notes Chapter One: Voids and Blinders, Words and Things 1. R. Cornevin, Histoire de {'Afrique 2:26: "Thus appeared to us a first province which we call central equatorial Africa and which could be called the ahistorical region of black Africa, for no important state developed there." 2. Basic ethnographic summaries can be found in H. Baumann, ed., Die Volker Afrikas und ihre traditionellen Kulturen 1:685-784, and H. Baumann and D. Westermann, Les peuples et les civilisations de {'Afrique, 191-214, which posits two culture circles: pygmy and forest people. G. P. Murdock, Africa: Its Peoples and Their Culture History, instead distinguishes several cultural historical groups, namely: pygmies (48-51), northwestern Bantu (271-78), equatorial Bantu (278-84), and "Mongo and Luba" (284-90). The number of ethnic groups is derived from these sources; about 150 distinct languages are spoken in the area. Total population is estimated from recent figures for Gabon, Congo, Equatorial Guinea, the relevant regions of Zaire, and estimates for southern and southwestern Cameroon. 3. R. Cornevin, Histoire de {'Afrique 2:29 (quote). For the hostile milieu argument, see 28-29, where the horrors of Stanley's expedition to rescue Emin Pasha are cited as evidence; for the result, see 1:11: "They have preserved prehistoric civilisations until this very day." See idem, Histoire du Congo Leo, 29, for explicit environmental determinism, the assertion that no society "going beyond" clan structure can exist there, and the pronouncement, "These societies ... have remained outside history." A similar but Marxist view is 1. Suret Canale's in Afrique noire 1:44-45: "As long as mankind has not reached the stage that it can destroy or manage the forest . . . , that forest . . . acts as a brake on social evolution." 4. For example, 1. Cornet, "Art Pygmee," 97-99. The fascination with pygmies, largely motivated by such views, has not abated in recent years, despite the abundance of publications. By 1970 the first specialized bibliographies appeared : cf. M. Liniger Goumaz, Pygmees et autres races de petite taille: Bibliographie generale; F. Plisnier-Ladame, Les Pygmees. In the 1980s the rate of research and publication is actually increasing, and at least one doctoral dissertation on the subject appears each year. 5. H. Deschamps, Traditions orales et archives au Gabon, 13. English authors use the expression "segmentary societies." 6. A. de Calonne-Beaufaict, Les Ababua, 60-65 (metaphor ofthe amoeba, 64); G. Hulstaert, Les Mongo: Aper<;u general, 36-37. 303 304 Notes to Pages 5-10 7. As is explicit in 1. Maes, "Les Warumbi," 628: "They do not know anything about the history of the tribe, ignore its origin, its migrations and the period when it arrived in the region where it is now settled." That was also all there was to history for the three most influential colonial syntheses: R. Avelot, "Recherches sur I'histoire des migrations dans Ie bassin de I'Ogooue et la region adjacente"; G. Van der Kerken, L'Ethnie Mongo; A. Moeller, Les grandes !ignes des migrations des Bantous de la province orientale du Congo Beige. 8. H. Baumann and D. Westermann, Les peuples: "cercle congolais du nord"; H. Baumann, ed., Die Volker Afrikas, 685-91. For the seeming uniformity of the forest, see M. Kingsley, Travels in West Africa, 101. 9. A very widespread cliche. See, for instance, a 1959 citation in M. C. Reed, "An Ethnohistorical Study of the Political Economy of Ndjole, Gabon," 93 n. 20: "the broken and degenerate races, driven into the forests and malarial swamps by the war tribes," a view he still heard in Libreville in 1984. 10. These include approximately 60 professionals from 1875 to 1985, many fewer than anywhere else in Africa. The figure includes approximately 25 historians and 35 anthropologists, and excludes another 30 anthropologists working on pygmies. 11. See map 1.4 and chapter 2. 12. A. Kroeber, Anthropology, 253-54, 411; 1. Steward, Theory of Culture Change, 37; G. R. Willey, Archaeological Theories and Interpretations, 373-74. 13. For the methodology concerning the written sources see B. Heintze and A. Jones, eds., "European Sources for Sub-Saharan Africa." For the methodology of oral tradition see 1. Vansina, Oral Tradition as History. The statement about their chronological depth is based on the examination of the traditions that are extant in writing. For the archaeological situation see F. Van Noten, ed., The Archaeology of Central Africa. Since then (1982) intensive research has been underway in Gabon and surveys undertaken in Cameroon, Congo, Gabon, and parts of Zaire. Cf. articles in Muntu, Nsi (also current bibliography), and Nyame Akuma. 14. H. Brunschwig, "Une histoire de l'Afrique noire est-elle possible?" 2:85. 15. In general, see C. A. Schmitz, ed., Historische VOikerkunde. For the central European school, see F. Grabner, Methode der Ethnologie; W. Schmidt, The Culture Historical Method of Ethnology. For the American school, see A. Kroeber, Anthropology, and more recently the journal Ethnohistory (1953-). 16. Recent manuals include R. Anttila, An Introduction to Historical and Comparative Linguistics; T. Bynon, Historical Linguistics; H. H. Hock, Principles of Historical Linguistics; W. Lehmann, Historical Linguistics: An Introduction ; and the multiple book review "Archaeology and Language," Current Anthropology 29 (1988): esp. 445-47 (P. Baldi) and 449-53 (R. Coleman). The classic application to culture history remains E. Sapir, Time Perspective in Aboriginal American Culture: A Study in Method. Notes to Pages 11-15 305 17. W. H. 1. Bleek established the language family in 1851 and baptized it Bantu in 1869. C. Meinhof, "Grundriss einer Lautlehre der Bantusprachen," gave phonetic proof in 1899. In 1906 he published the first basic grammar of "common Bantu," Grundzuge einer vergleichenden Grammatik der Bantusprachen. 18. B. Heine, H. Hoff, and R. Vossen, "Neuere Ergebnisse zur Territorialgeschichte der Bantu"; Y. Bastin, A. Coupez, and B. de Halleux, "Classification lexicostatistique des langues bantoues"; Y. Bastin and A. Coupez, pers. com. 1985. 19. M. Swadesh, "Lexicostatistical Dating of Prehistoric Ethnic Contacts"; idem, "Towards Greater Accuracy in Lexicostatistical Dating." 20. D. H. Hymes, "Lexicostatistics So Far"; C. D. Chretien, "The Mathematical Models of Glottochronology"; see also n. 16 above. 21. R. Anttila, Historical and Comparative Linguistics: in general, 133-53; for Worter und Sachen (words and things) as a combination of linguistic and ethnographic data, 291-92. See also T. Bynon, Historical Linguistics, 62-63. 22. M. Guthrie, Comparative Bantu, to be checked with E. Meeussen, Bantu Lexical Reconstructions. Further discussion of Guthrie can be found in 1. Vansina, "Deep Down Time." 23. "CS" stands for "comparative series" and "ps" for "partial series." Guthrie arranged his entries by CS and the number refers to the entry. But his "ps" are not always arranged under a number. Hence a page reference is given here. The vowel renderings for the CS in the text do not include the cedilla under i and u, which distinguishes vowels of first and second degree. These are therefore not differentiated in the text. For an exact rendering, see Guthrie's Comparative Bantu. 24. See discussion in R. Coleman, "Archaeology and Language," Current Anthropology (1988): 437-68, contra C. Renfrew. Starred items are as much "hard data" as archaeological sites are. 25. 1. Vansina, The Children of Woot, 293,297 (cf. Lele ngal). 26. Another example is the Tio fishing technique called unteku. It is a loanword because Tio never has a k as a second consonant of the root (-teku). Upriver the Likwala call the technique muteku, normal form in their language. The Tio borrowed it from them or from one of the other related groups of fishermen on the Zaire River. The anomaly of the form indicates that the loan is fairly recent. Otherwise the Tio speakers would have nativized it by dropping k. 1. Vansina, The Tio Kingdom of the Middle Congo: 1880-1892,136. 27. See chapter 5 and the discussion of kazi katsi (map 5.5; pp. 153-55). 28. M. McMaster, "Patterns of Interaction," chap. 7, e.g., fig. 7.11. 29. D. Cordell, "Throwing Knives in Equatorial Africa"; P. McNaughton, "The Throwing Knife in African History." The throwing knife is also found in seventeenth-century Loango: S. Brun, Schiffarten, 11. Compare also osere, osele, "throwing knife" in the Ivindo Valley, and oshele in the middle Lukenie Valley over 1,000 km away. 306 Notes to Pages 15-19 30. 1. Vansina, "Deep Down Time," map showing -kaya "tobacco"; and idem, Art History in Africa, 168-73. 31. Skewing is the preservation of a borrowed form which is otherwise irregular in language. Skewed forms indicate borrowing rather than common descent. They are so indicated by M. Guthrie. The contrary does not rule out borrowing at all, for there are strong tendencies to adjust loans to usual patterns of "phonological nativization." Cf. H. H. Hock, Principles ofHistorical Linguistics , 90-97. 32. M. Swadesh, "Towards Greater Accuracy," 123, using a 85.4 percent retention rate per millennium as the constant (130). 33. R. Anttila, Historical and Comparative Linguistics, 396-98; T. Bynon Historical Linguistics, 266-72. For a strong negative reaction, cf. R. Coleman, "Archaeology and Language," 450. 34. A. Jones and B. Heintze, eds., "European Sources for Sub-Saharan Africa before 1900: Use and Abuse," esp. 1-17; 1. Koponen, "Written Ethnographic Sources and Some Problems Concerned with Their Use in African Historiography ," 55-69; R. Thornton, "Narrative Ethnography in Africa, 18501920 ," 503-20. 35. J. Vansina, "The Ethnographic Account as a Genre in Central Africa," 439, for examples of plagiarism by administrators. For the case of Trilles' work on pygmies, see K. Piskaty "1st das Pygmaenwerk von Henri Trilles ein zuverlassige Quelle?" 36. Affonso Mvemba Nzinga and later kings of Kongo are inside sources but they are too far south. Cf. L. Jadin and M. Dicorato, Correspondance de Dom Afonso roi du Congo: 1508-1543. 37. 1. Vansina, "Ethnographic Account." 38. E. Bassani has made a special study of ancient objects and iconography. See, for instance, his "I disegni dei Manoscritti Araldi del Padre Giovanni Antonio Cavazzi da Montecucolo." 39. C. Geary, "Photographs as Materials for African History." 40. F. Van Noten, Archaeology of Central Africa, 11. 41. E. Meeussen, "Het aanleggen van kollekties met geluidsopnemingen," 19. 42. W. 1. Samarin, "Bondjo Ethnicity and Colonial Imagination," 345-65. 43. The discussion shows the administrative importance of "tribes." This case affected two provinces and the "ethnic" justification for their boundaries. Cf. 1. Maes, "Note sur les populations Lalia et Yasayama du territoire des Dzalia," 175 (citing Stryckmans and Mazy); Van de Capelle, "Populations de Mondombe et Yolombo"; idem, "Note and Map"; A. Moeller, Grands lignes, 194-95; versus the authorities cited by M. De Ryck, Les Lalia-Ngolu, 21-22, and G. Van der Kerken, L'Ethnie Mongo, 54-55,64,75; These include Catholic missionary leaders fostering a "greater Mongo" consciousness. By 1935 the issue was administratively decided in favor of an inclusion in the Mongo group. Protes- Notes to Pages 20-21 307 tant missionaries in the region maintained a separate Ngando "tribe" well after that date. 44. P. Van Leynseele, "Les Libinza de la Ngiri," iv, 30-35. 45. R. Harms, "Oral Tradition and Ethnicity." 46. T. E. Bowdich, Mission from Cape Coast Castle to Ashantee, 422-34; G. A. Robertson, Notes on Africa, 340; S. Koelle, Polyglotta Africana: itineraries . For a dicussion of ethnicity in Gabon, see L. Biffot, "Contribution a la connaissance et comprehension des popUlations rurales du nord-est du Gabon," 16-59. 47. G. Brausch, " 'Polyandrie' et 'mariage classique' chez les Bashi Lele"; M. Douglas, "A Form of Polyandry among the Lele of Kasai." 48. Field data for Kuba groups; 1. Mertens, Les Ba Dzing de la Kamtsha 1:207-8. For cases involving objects, see 1. Vansina, Art History in Africa, 3133 , 165. 49. P. Laburthe Tolra, Les seigneurs de la foret, 422 n. 6 (a village near Yaounde moved 42 km at once in 1891); R. Letouzey, Contribution de la botanique au probleme d'une eventuelle langue pygmee, 31 n. 15 (Baka hunters in eastern Cameroon moved over a 50-km radius); C. Amat, "Ngovayang II. Une village du sud-Cameroun," 61, 92, map 6 (200-km movement of a village in a generation); G. Hulstaert, Les Mongo, 16 (marriage up to 50 km away from birthplace), taking local variability in spatial mobility into account. For instance, in areas of very low population densities people moved farther and in areas of high popUlation density distances were shorter. 50. Goebel, "Carte ethnographique de la zone Uere-Bili." Thus a set of maps of Uele already indicated ethnicity village by village in November and December 1908. 51. Uninhabited (Unbewohnt) or even Tote Zone ("dead area") occurs often in German reports and maps of southeastern Cameroon. Cf. F. Von Stein, "Expedition des Freiherrn von Stein," 185-86, and map by M. Moisel, "Skizze," 42-43. For population distributions in Gabon and Congo see G. Sautter, De l'Atlantique au fieuve Congo, vol. 2: in fine; for Zaire, see P. Gourou, "Notice de la carte de densite de la population au Congo Beige et au Ruanda-Urundi." 52. For the Independent Congo State, Congo, portions of Gabon, and Cameroon see routes on the map of 1. Du Fief, "Carte de l'Etat Independant du Congo." All the rivers navigable by small boat were also reconnoitered; crosschecking information of the 50-km radius grid with reported ethnic groups shows that I could not find data specifically referring to the Babole of the lower Likouala-aux-herbes (but there are descriptions of the villages along the river) or to the Mbuli, Langa, and Kuti villages of the upper Tshuapa-Lomami area. Each "ethnic group" involved fewer than 1,000 inhabitants. 53. Occasionally one finds out-and-out fakes. Thus postcards from the 1930s showed nude Kuba women. In this area, however, men and women were always dressed. 308 Notes to Pages 21-26 54. E. Pechuel Loesche, Volkskunde von Loango; W. de Mahieu, Structures et symboles; idem, Qui a obstrue la cascade? 55. The major ones are M. Denis, Histoire militaire de l'Afrique equatoriale fram;aise; A. Lejeune-Choquet, Histoire militaire du Congo (Zaire); La Force Publique de sa naissance a1914 (Zaire). For German Kamerun see reports in Deutsches Kolonialblatt. Dates of colonial conquest are also often cited in local documents. The political register of each post in the Congo Independent State and in the Belgian Congo was kept from the date when the post was established and colonial rule began. Many of these registers still exist. 56. W. H. Sheppard, Presbyterian Pioneers; E. Torday and T. A. Joyce, Notes ethnographiques ... Les Bushongo; H. Bucher, "The Mpongwe of the Gabon Estuary: A History to 1860." 57. G. Le Testu, "Les coutumes indigenes de la circonscription de la Nyanga (Gabon)"; idem, "La soumission des Bawanji"; C. Coquery Vidrovitch, Le Congo au temps des grandes compagnies concessionnaires, 1898-1930, 212-19. 58. 1. Vansina, Geschiedenis van de Kuba van ongeveer 1500 tot 1904, 35053 , for an inquiry on the spot about the informants and information given to E. Torday for his Bushongo. 59. E. Zorzi, Al Congo con Brazza. 60. Among others, M. Douglas (Zaire) and M.-C. Dupre (Congo) both reported this. Still, Dupre was a mother of twins, recognized as such, and initiated in the arcana of twin lore. Douglas was an "honorary man," yet her gender prevented her from learning much about the senior men's society. It is clear that in fact gender roles in such circumstances become negotiable. 61. G. Van Buick, Les recherches linguistiques au Congo Beige; Y. Bastin, Bibliographie bantoue selective. After an initial period of trial and error, missions usually adopted a single language for the whole mission area rather than attempting to preach and teach in all the languages. Hence good records exist only for a minority of languages. On this question, ct. G. Van Buick, Les deux cartes linguistiques du Congo Belge, and the rebuttal by G. Hulstaert, Au sujet de deux cartes linguistiques du Congo beige. 62. Cf. H. Scheub, African Oral Narratives, Proverbs, Riddles, Poetry and Song: An Annotated Bibliography. For an example, see T. Obenga, Litterature traditionelle des Mbochi. 63. E. Van de Woude, "Documents pour servir ala connaissance des populations du Congo Beige, 6-62, esp. 20-45; B. Jewsiewicki, "Etude analytique des archives administratives zairoises." 64. G. Van Buick, Deux cartes linguistiques, and Y. Bastin, Bibliographie bantoue selective, are now supplemented by K. Kadima et aI., Atlas linguistique du Zaire: 1nventaire preliminaire. This is part of the projected linguistic atlas of all central Africa (Alae). The first systematic survey in equatorial Africa was the linguistic survey of the northern Bantu borderland (1949-52), but the whereabouts of the hundreds of standard general vocabularies gathered by Van Buick are unknown. See G. Van Buick, Mission linguistique, 1949-1951,17-60. From Notes to Pages 26-38 309 the 1970s onward such surveys have been undertaken again III Cameroon, Gabon, and Zaire. 65. H. A. Johnston, A Comparative Study of the Bantu and Semi-Bantu Languages, and others in his bibliography, 785-815. Other vocabularies have been published since, and a few earlier ones escaped even Johnston's attention. 66. Cultural vocabularies comprising from 110 to 500 items are now being compiled for some 150 languages in the area, sometimes from the most heterogeneous sources. Only the items dealing with the main social and political institutions have been used for this book. 67. These are A. Moeller, Grandes /ignes, and G. Van der Kerken, L'Ethnie Mongo. On the latter, see G. Hulstaert, "Une lecture critique de L'Ethnie Mongo de G. Van der Kerken." 68. See 1. Jacobs and 1. Vansina, "Nshoong atoot: Het koninklijk epos der Bushoong," for a conscious historical tradition. See also A. De Rop and E. Boelaert, Versions et fragments de l'epopee Mongo: Nsong'a Lianja (and previously published versions); D. Biebuyck and K. Mateene, The Mwindo Epic from the Banyanga (Congo Republic); H. Pepper and P. De Wolf, Un mvet de Zwe Nguema, for epics. 69. A list appears in E. Van de Woude, Documents, 13-17, 53-62; 1. Vansina, "Ethnographic Account," 434-35. 70. Cf. n. 4. Les racines du ciel is the title of a novel dealing with pygmies. But even a serious account such as N. Ballif's came with the title Les danseurs de Dieu: Chez les pygmees de la Sangha ("The Dancers of God: Among the Pygmies of the Sangha"), an allusion to the dwarf of Pharao Pepi II of the Sixth Dynasty. 71. L. L. Cavalli-Sforza, African Pygmies, contains good examples of such preconceptions. Some modern studies take their cue more from the methodology of primate behavioral study than from the humanities, as in the title "The Relation between Exploration and Mating Range in Aka Pygmies," which is a contribution to Cavalli-Sforza's book. 72. W. N. Fenton coined the term. See "Ethnohistory and Its Problems." Chapter Two: The Land and Its Settlement 1. P. T. White, "Tropical Rainforests," 2-46, on the vulnerability of these milieux. 2. Cf. FAa, UNESCO, Soil Map ofthe World; G. Laclavere, ed., Atlas de La repubLique Unie du Cameroun, 25-27; R. Walter and 1. Barret, eds., Geographie et cartographie du Gabon, 30-33; P. Vennetier, Atlas de La Repub/ique popuLaire du Congo, 18-19. 3. G. Laclavere, Atlas, 16-19; R. Walter and 1. Barret, Geographie et cartographie, 22-25; P. Vennetier, Atlas, 10-15. 4. See n. 3 and F. Bultot, "Notice de la carte des zones climatiques du Congo Beige et du Ruanda-Urundi. " 310 Notes to Pages 39-41 5. F. Bultot, "Notice"; P. Vennetier, Atlas, 14 (Owando); G. Ladavere, Atlas, 19; R. Walter and J. Barret, Geographie et cartographie, 25. 6. The date varies from year to year. For Zaire, F. Bultot, "Notice," 7 (maps); idem, "Saisons et periodes seches et pluvieuses au Congo beIge et au Ruanda-Urundi." Bultot lists the usual dates for the onset of the dry season as well as the dates at which dry seasons began in abnormal years. 7. For the common view of pygmies as living fossils in the 1950s: R. Hartweg, La vie secrete des pygmees, 27: "They are a living prehistoric museum "; also 117. In the 1980s K. Duffy, Children of the Forest, vii-viii, 176-77: "a prehistoric group of hunters and gatherers who survived into the twentieth century" and "prestone age" (viii); "... the living spirits of our not so distant hunting and gathering ancestors" (177) and frontispiece: "An Egyptian dwarf about 2000 B.C." "Children" and "live in the present" (176), "harmony with the ecosystem" (176), and "innocent age" (177) round off the stereotype, unchanged since the Hellenistic pygmies in the land of the lotus eaters. 8. Sources cited in M. Reed, "An Ethnohistorical Study of the Political Economy of Ndjole, Gabon," 96. 9. The earliest exposition of the stereotype of the forest in H. M. Stanley, Through the Dark Continent, 408-9, begins with: "On the 6th of November [1876] we drew nearer to the dreaded black and chill forest called Mitamba, and at last, bidding farewell to sunshine and brightness entered it." For the full elaboration, see idem, In Darkest Africa, 134-292, esp. 138; T. H. Parke, My Personal Experiences in Equatorial Africa, 71-358; and other sources listed in R. Jones, The Rescue of Emin Pasha. It usually escapes readers that the terrors of the forest were described in order to excuse major errors of judgment made during the expedition. 10. P. Richards, "The Tropical Rain Forest"; K. A. Longman and 1. Jenik, Tropical Forest and Its Environment; F. Fournier and A. Sasson, Ecosystemes forestiers tropicaux d'Afrique, 92; W. Robyns, Contribution Ii ['etude des formations herbeuses du district forestier central du Congo beige; 1. Vansina, "Peoples of the Forest," 77 n. 1 (further references). For maps, see F. White, The Vegetation ofAfrica; G. Ladavere; Atlas, 20-24; R. Walter and 1. Barret, Geographie et cartographie, 34-37; P. Vennetier, Atlas, 16-17; and L. Peeters, Les limites foret-savane dans Ie nord du Congo en relation avec Ie milieu geographique. For a general discussion of human-habitat interaction, see 1. Vansina, "L'homme, les fon:!ts et Ie passe en Afrique." 11. R. Harako, "The Mbuti as Hunters," and map by F. White, Vegetation of Africa, versus C. Turnbull, "Forest Hunters and Gatherers: The Mbuti Pygmies ," 41. For further discussion see T. B. Hart, "The Ecology of a Single Species-Dominant Forest and Mixed Forest in Zaire." 12. Evidence for high human activity of long ago is discussed in chapters 5 and 6. Indicators for higher population densities followed by lower densities are the frequency of certain trees such as Lophira alata and the Aucoumea. At the Notes to Pages 41-47 311 other extreme, stands of Gilbertiodendron dewevrei indicate low population densities for the last half millennium or more. 13. G. Laclavere, Atlas, 22-23. 14. G. Sautter, De ['Atlantique aufleuve Congo, 2:947-76. P. Gourou, Les pays tropicaux, 10-19, is less sanguine. 15. The "pygmies" are the celebrated case in this connection. Cf. 1. Hiernaux, The People ofAfrica, 113-49; idem, "Long-term Biological Effects of Human Migration from the African Savanna to the Equatorial Forest." But more recently R. C. Bailey, "The Socioecology of Efe Pygmy Men in the Ituri Forest," 26-27, disputes this stand and points out that the present habitat of pygmies in the Ituri was savanna until c. 900 B.C. or possibly as late as A.D. 1200! The question needs further study. 16. P. Gourou, La densite de la population rurale au Congo belge, 116. 17. 1. Maley, "Fragmentation de la foret dense," 310, for a map and the link between amount of rainfall and the Benguela current; F. Van Noten, ed., The Archaeology of Central Africa, 22-25; 1. Flenley, The Equatorial Rain Forest : A Geological History, 29-54, and human influence, 116-18. 18. F. Van Noten, Archaeology, 25. Elsewhere the forest was receding then. In the same area cf. M. C. Van Grunderbeek, E. Roche, and H. Doutrelepont, "L'age du fer ancien au Rwanda et au Burundi," 46-48. 19. L. Demesse, Techniques et economie des pygmees babinga, 7, 28. 20. G. Laclavere, Atlas, 20-24; R. Walter and 1. Barret, Geographie et cartographie, 34-37; P. Vennetier, Atlas 36-37. For savannas in the Late Stone Age, see R. Deschamps, R. Lanfranchi, A. Le Cocq, and D. Schwartz, "Reconstitution d'environnements quaternaires par I'etude de macrorestes vegetaux (Pays Bateke, R. P. du Congo)." 21. F. Van Noten, ed., The Archaeology of Central Africa; F. Van Noten et aI., "The Prehistory of Central Africa, Part 2"; R. Bayle des Hermens, "The Prehistory of Central Africa, Part 1" and references there. Archaeological activity has increased since then, however, and readers should consult the journals Nyame Akuma, Nsi, Muntu, and the African Archaeological Review. 22. F. Van Noten, ed., Archaeology, 29-30; L. Fiedler and 1. Preuss, "Stone Tools from the Inner Zaire Basin"; B. Peyrot and R. Olisly, "Paleoenvironnement et archeologie au Gabon," 14-15; R. Lanfranchi, "Recherches prehistoriques en Republique populaire du Congo," 7. 23. But see R. C. Bailey, "Socioecology," 26-27, for some implicit doubts. 24. 1. Hiernaux, The People ofAfrica, fig. 19, table 25. It follows from the palaeogeography that if Hiernaux's view on biological adaptation is accepted, then the clusters probably represent pygmy groups issuing from the different forest refuges, as indicated in J. Maley, "Fragmentation," 310. 25. On survivals of "pygmy" languages, see S. Seitz, Die zentralafrikanischen Wildbeuterkulturen, 19-23; 1. M. C. Thomas, "Emprunt ou parente?" 16061 ; R. Letouzey, "Contribution de la botanique au probleme d'une eventuelle 312 Notes to Pages 49-52 langue pygmee"; and A. Vorbichler, "Die Sprachliche Beziehungen zwischen den Waldnegern und Pygmiien in der Republik Kongo-Leo"; idem, "Zu dem Problem der Klasseneinteilung in Lebendiges und Lebloses in den Pygmiien- und Waldnegerdialekten des Ituri, Congo." 26. This already before the Neolithic: L. Fiedler and 1. Preuss, "Stone Tools," 179, 182. B. Clist, "Un nouvel ensemble neolithique en Afrique centrale : Le groupe d'Okala au Gabon," 45, shows that stone axes in the Ogooue Delta came from about 300-350 km upstream. 27. A Gabonese informant cited by H. Deschamps, Traditions orales et archives au Gabon, 25. 28. Y. Bastin, A. Coupez, and B. de Halleux, "Classification lexicostatistique des langues bantoues (214 releves)" and Bastin's 1983 communication about a larger set of western Bantu languages; P. B. Bennett and 1. P. Sterk, "South Central Niger-Congo: A Reclassification"; and more recently T. C. Schadeberg, "The Lexicostatistic Base of Bennett and Sterk's Reclassification of Niger-Congo with Particular Reference to the Cohesion of Bantu." The group of Y. Bastin and A. Coupez is still at work on providing a complete genetic tree of the Bantu languages. 29. M. Guthrie, Comparative Bantu: index, 2:177,178; 1. Vansina, "Western Bantu Expansion," 132, 138, map 139. For details see chapter 3. 30. F. Nsuka and P. De Maret, "History of Bantu Metallurgy: Some Linguistic Aspects." 31. Exactly why some splinter groups did not expand farther promises to be a rewarding topic for future study. The scissions of the Yambasa-Nen, of the Myene-Tsogo, and of the Aka-Mbati small groups are of this type. The expected situation is exemplified by the split between the northern Zairian languages and the southwestern group. According to Hock, Principles ofHistorical Linguistics, 449, the presence of "flat tree diagrams" with mainly binary branchings are indicative of the spread of innovations in a dialect continuum in large ensembles of dialects; furthermore, the group average method used to find splits favors only two-way splits. Finally, historians will remember that a split represents a time period, not a moment, and a region, not a spot, and that the dates refer to language differences which took hold after separation. 32. According to the principle of Ockham's razor. The suggested routes are based on placing the locale of dispersals at the boundary of the resulting groups. The more branches disperse at once, the easier it is to fix such a locale. In this situation, however, only two-way splits occur but for one case. Specific environmental barriers such as the great swamps, the arid Bateke plateaus, and the inhospitable Du Chaillu massif have been taken into account as well. In the end, however, only archaeology will provide more than general directions of expansion. 33. M. McMaster, "Patterns of Interaction," chap. 2 for Buan. The label Soan is suggested here for the neighboring branch. The name is derived by Notes to Pages 53-57 313 analogy from the "So" and "Eso" languages, which, among others, belong to this group. 34. P. De Maret, "Resultats des premieres fouilles dans les abris de Shum Laka et d'Abeke au nord-ouest du Cameroun: Belgian archaeological project in Cameroon" (Obobogo); idem, "Recent Archaeological Research and Dates from Central Africa," 134-35 (Obobogo); idem, "The Ngovo Group: An Industry with Polished Stone Tools and Pottery in Lower Zaire," 127 (dates for Ngovo and Sakuzi); J. Denbow, A. Manima-Moubouha and N. Sanviti, "Archaeological Excavations along the Loango Coast Congo," 37-38 (Tchissanga); M. Eggert, "Imbonga and Batalimo: Ceramic Evidence for Early Settlement of the Equatorial Rain Forest," 133 (Imbonga; rejecting the Hanover dates still leaves samples 3, 7, and 9 and dates 310,340, and 320 B.C.), 139-41 (Batalimo-Maluba; one date is at least 190 B.C. and two others 40 B.C. and A.D. 20); idem, "Archaologische Forschungen im zentralafrikanischen Regenwald," map 3223 (Imbonga and Batalimo-Maluba). 35. B. Clist, "Travaux archeologiques recents en Republique du Gabon, 1985-1986," 10, and pers. com. 1988. Ibid., 9, also gives 2920 +/-90 B.C. and 2450 +/-70 B.C. for different and earlier ceramics in this area, which have so far not been confirmed. If they are indeed associated with ceramics one would have to dissociate the spread of ceramics (earlier) from that of a food-producing economy, the latter being associated only with the expansion of the western Bantu languages. 36. M. McMaster, "Patterns of Interaction," chap. 2 for internal Buan divisions. This relates to the Zaire bend and the northeast, not to the Ubangi Valley where the Batalimo-Maluba types date to the last two centuries B.C. only. 37. The first case is calculated on carbon dates, the others are based on estimated loci of dispersal and glottochronological estimates. The first two cases should be compared with the speed of movement between the same Congo panhandle and the Niari Valley where the Kongo-Congo-Gabon groups split. That yields 600 km (1,200 double) in a mere 170 years, or 70 km/decade, which is suspiciously high. Despite the vagueness of dating estimates and of localities inherent in the calculation, the results remain suggestive of considerable variability in the rates of migration. In the examples above, the seceding northern Zaire speakers went slowly at first because they had to master the watery and marshy environments of the great swamps. Once at home on the waters, however, they raced. 38. S. Seitz, ZentralaJrikanischen Wildbeuterkulturen, 22; W. Dupre, "Die Babinga-Pygmaen," 29-38. It is likely that the original term for pygmy specifically was *-yaka. 39. For example, S. Bahuchet and H. Guillaume, "Relations entre chasseurs-collecteurs pygmees et agriculteurs de la foret du nord-ouest du bassin congolais," 111-17. 40. This stance conveniently denied rights from first occupation to the hunters and gatherers by essentially claiming that they were not fully human. 314 Notes to Pages 57-61 41. On fishing, see chapter 3. The rarity of suitable settlement sites along many rivers made clashes even more likely. 42. Future biological research would help to frame such questions better. Meanwhile, analogies with the decimation by disease of the autochthons in the New World should be distrusted. This was no invasion by organisms from another continent, but probably more a question of the relative prevalence of certain diseases in certain habitats. 43. D. Calvocoressi and N. David, "A New Survey of Radiocarbon and Thermoluminescence Dates for West Africa," 10 (Taruga), gives a ninth-century date for Taruga; L. Digombe et a!., "Early Iron Age Prehistory in Gabon," 183. See also M. C. Van Grunderbeek, E. Roche, and H. Doutrelepont, Le premier age du fer au Rwanda et au Burundi, 48-50 (great lakes). 44. In Gabon the situation is evolving rapidly with an active archaeological program. Cf. L. Digombe et a!., "Early Iron Age Prehistory"; B. Clist, "Early Bantu Settlements in West-Central Africa," 380-82. Contrast this with the older data by P. De Maret, "Ngovo Group," 129-48, and more recent information in B. Clist, "La fin de l'age de la pierre et les debuts de la metallurgie du fer au Gabon," 24-28, 29; L. Digombe et a!., "Recherches archeologiques au Gabon: Annee academique 1986-1987." Most recently B. Clist, "Un nouvel ensemble," 47-48, urges caution, which is reflected on map 2.10. 45. M. Eggert, "Archiiologische Forschungen," 3234-35; idem, "Imbonga and Batalimo," 134, 141. 46. F. Van Noten, "The Early Iron Age in the Interlacustrine Region," 62, 65,67. 47. M. McMaster, "Patterns of Interaction," chap. 4 and table 4.3. 48. F. Nsuka-Nkutsi and P. De Maret, "Etude comparative de quelques terme metallurgiques dans les langues bantoues," 731-42; M. Guthrie, Comparative Bantu, 1:132, 138-40 (see Appendix items 97-103); C. P. Blakney, "On 'Banana' and 'Iron'," 85-118. Distinctions between western and eastern Bantu must be added to these sources. 49. Such as C. Wrigley, "Speculations on the Economic Prehistory of Africa ," 201-2 (iron spear). Copper and lead in lower Congo were probably smelted as early as iron. Cf. 1. Denbow, A. Manima-Moubouha, N. Sanviti, "Archaeological Excavation," 39, who locate such smelting activity at MadingoKayes in the second or third century A.D. 50. E. Perrot and E. Vogt, Les poisons defleches et les poisons de t'epreuve des indigenes de {'Afrique, 58-80. On the inferiority of wooden or horned tip spears, Bosekonsombo (Nkumonsombo) correspondence to Hulstaert: supplement to his 19 February 1971. 51. A prehistoric ore center at Lebombi in Haut Ogooue Province lasted for half a millennium in a region where iron smelting was still a specialty in 1900; see P. R. Schmidt, L. Digombe et a!., 17. The region of Abala was a major ironsmelting center as well, probably before the eighteenth century. M.-D. Dupre, "Pour une histoire des productions de la metallurgie du fer chez les Teke, Notes to Pages 61-64 315 Ngungulu, Tio, Tsayi," 201-7; and R. Lanfranchi, "Esquisse archeologique des regions Teke," 83. 52. On the botanical characteristics, see G. Rossel, "Gewasinnovaties in Gabon," 25-49; E. De Langhe, "Bananas Musa spp.," 53-78; N. W. Simmonds, Bananas; R. M. Eggert, Das Wirtschaftssystem der Mongo am Vorabend der Kolonisation, 7-11; P. T. Perrault, "Banana-Manioc Farming Systems of the Tropical Forest," 6-28, 116-142; A. Raponda-Walker and R. Sillans, Les plantes utiles du Gabon, 447-54. 53. E. De Langhe, "La taxonomie du bananier plantain en Afrique equatoriale ," 417-49; G. Rossel, "Gewasinnovaties," 28, 29; N. W. Simmonds, Bananas. 54. D. N. McMaster, "Speculations on Coming of the Banana to Uganda," 57-69, favors the East African route, perhaps via the Zambezi. 1. Barrau, as cited by 1. Bouquiaux and 1. Thomas, "Le peuplement oubanguien," 3:816, favors the upper Nile. For E. De Langhe's views as interpreted by P. T. Perrault, see "Banana-Manioc Farming Systems," 121-22, and his "La taxonomie," 447-49. Argument based on varieties in 1. Vansina, "Esquisse historique de l'agriculture en milieu forestier," 21-24 nn. 32-35. 55. G. Rossel, "Gewasinnovaties," 31-38; C. P. Blakney, "On 'Banana' and 'Iron'," 54-78; W. 1. Mahlig, "Lehnwortforschung und Ethnohistorie," 720 , map 18; E. Bylin, Basakata: Le peuple du pays de l'entre-fleuves LukenyeKasai , 98. -bugu and bugugu: Bu- is not a prefix class 14, which in Tsogo is bo-, and even if it were, bugugu could still be a reflex of -bugu; see A. RapondaWalker and R. Sillans, Plantes utiles, 305. 56. For -toto, see C. P. Blakney "On 'Banana' and 'Iron'," 78. For Gabon, see A. Raponda-Walker and R. Sillans, Plantes utiles, 305; and for Bekwil, Gabon, and Cameroon, see G. Rossel, "Gewasinnovaties," 32; -pokolpokul bogo as "banana tree" or "banana grove" occurs in Huku (Amba), Nyanga, Lega, Boa and related languages, Bira, Komo, Bali, So, Heso, Enya, Lengola, Metoko, Mongo and closely related languages, Sengele, Benga, and probably elsewhere. Other generic terms for "banana," such as fondo, ndoo, -kube, -bo, gbedelkede in the north from the Ubangi to Douala and the Cross River, as well as -koma and -doso in southern Maniema, indicate that only the major outlines of a linguistic study are known so far and that the full history of the banana in the area will be much more complex than had been imagined. 57. G. Rossel, "Gewasinnovaties," 26-31; G. A. Wainwright, "The Coming of the Banana to Uganda," 145-47; S. Munro Hay, "The Foreign Trade of the Aksumite Port of Adulis," 107-25. The banana was not such an exotic plant for the Axumites, who cultivated ensete, a relative of the banana; M. McMaster, "Patterns of Interaction," fig. 5.3. Using semantic evidence McMaster concluded that bananas spread among the Buans then living near the middle Bomokandi after A.D. 170 by glottochronological estimate, but ancient diffusion may well be indistinguishable from common origin here. The evidence assembled by M. Watson, Agricultural Innovation in the Early Islamic World: The 316 Notes to Pages 64-66 Diffusion of Crops and Farming Techniques, 700-1100, who argues for a late Muslim introduction, would best fit the high-yielding AAA cultivars. 58. Certainly before A.D. 1000. The introduction of the common fowl parallels the diffusion of the banana. To date, chicken bones are attested in Rwanda and southern Shaba by A.D. 1000. 1. Vansina, "Esquisse historique de l'agriculture," 25. 59. See n. 25. 60. L. Bouquiaux and 1. M. C. Thomas, "Le peuplement oubanguien," 807-24, includes Gbaya; D. E. Saxon, "Linguistic Evidence for the Eastward Spread of Ubangian Peoples," 66-77, excludes Gbaya as P. Bennett (ms. communication ) does also. Thomas and Bouquiaux are unreliable on chronology and on postulated migrations. The dates are not based on any calculations. Tentative glottochronological dating is so early that the proposed Ngbaka and Gbanzili migration from Uele, rather than from farther northwest in the Central African Republic, is doubtful. The account of their migration on maps 3, 4, 6, and 7 and text 810-12 is incorrect in postulating late migrations up and down the lowermost Ubangi as far as Lake Tumba. Local oral traditions have a depth of three to four centuries here and do not remember anything but Bantu speakers in these areas. 61. The terminology for this group is still variable. The terms Moru-Madi (Tucker and Bryan) are still used, though Ehret labels the group east-central Sudanic. However, south-central Sudanic agrees best with the overall linguistic classifications. For the group, cf. A. N. Tucker and M. A. Bryan, The NonBantu Languages ofNorth-Eastern Africa, 1-9; C. Ehret et aI., "Some Thoughts on the Early history of the Nile-Congo Watershed," 85-112; A. Vorbichler, "Linguistische Bemerkungen zur Herkunft der Mamvu-Balese," 1145-54; idem, Die Phonologie und Morphologie des Balese, 9-12; idem, Die Sprache der Mamvu, 29-31, whose late chronology (after A.D. 1000) cannot be accepted given M. McMaster, "Patterns of Interaction," chap. 2, in which she proposes "before 440 B.C." for the proto-Mamvu subgroup in Uele; N. David, "Prehistory and Historical Linguistics in Central Africa," 80-81, would see their area of origin as far as 100 north rather than the 40 north proposed by Ehret. 62. These are the Nyanga and Tembo, who both speak languages of the great lakes. Western Bantu farmers began to occupy northern Maniema only in the last three centuries B.C. (M. McMaster, "Patterns of Interaction," table 1.4: 380-180 B.C. for the Bira split), so that a spread of Nyanga by an estimated c. A.D. 220 (split between Nyanga and other great lakes) is entirely plausible. The Tembo followed one or more centuries later. 63. See 1. Vansina, "Western Bantu Expansion," 139-40. Although the agricultural innovations are evident, the extent of social and political elements remains unclear. Further research in the eastern vocabulary present in the western Bantu savanna languages is needed. 64. The Ngbandi movement is based on M. McMaster, "Patterns of Interaction ," chap. 5, which according to her follow the Mba and Mondunga migrations . See her table 5.3 for the earliest Ngbandi influences in Uele. Notes to Pages 66-68 317 65. For Ngbandi some evidence of mutual influence is shown in G. Hulstaert, "Lomongo en Ngbandi." Ngbandi influence is strong in various groups of Apagibeti, where it mingles with influences of Mondunga-Mba on a Bantu base. See, for instance, A. Van Houteghem, "Overzicht der Bantu dialekten van het gewest Lisala," 42-43, 49; H. Nzenze, "Note sur les Pagabete," 137. The languages of the Boa group also show some influence from this group (e.g., the use of -Ie suffix). 66. For the late date of the Balese split, see A. Vorbichler, Sprache der Mamvu, 29, for a 93 percent correspondence between Mamvu and Balese, which gives an A.D. 1720 date. For southern central Sudanic influences on Bantu, see M. McMaster, "Patterns of Interaction," chap. 4, tables 4.1,4.2, and 4.3, for Bali. The influence of this Mamvu group on Komo or Bira is as evident as on Bali. These languages even abandoned their prefix classes. Southern central Sudanic influence on phonetics and on vocabulary are seen throughout the northeastern Bantu languages. 67. The only available chronological estimate places the onset of the Ngbandi migration along the Uele at A.D. 280. M. McMaster, "Patterns of Interaction ," chap. 5, table 5.3, shows the earliest Ngbandi influence from c. A.D. 170 or shortly thereafter. 68. J. M. C. Thomas and L. Gouquiaux, "Le peuple oubanguien," 816, citing Barrau, merely state that the diffusion of the banana paralleled the migration of the Ngbaka Mabo. As to the generic terms, fondo in Ngbandi and the related ndoo in the Ngbaka Mabo-Gbanzili group do not seem to have congeners farther eastward. In the equatorial subgroup poongo- (Ie) in Mondunga and bo- (Ie) in Mba differ from each other. Yet M. McMaster, "Patterns of Interaction," chap. 5, argues that at the time of migration from the east, bananas there were already called -bugu. 69. By A.D. 500 (63 percent correspondence) we are at the level of minor linguistic subgroups and separate just under two-thirds of all languages listed for the area in the Tervuren survey. By A.D. 1000 (74 percent correspondence) we are at the level of individual languages with nearly three-fourths of all languages listed. With the exception of the Sanaga-Ntem area no major disjunctions, indicative of late massive migrations, occur on the linguistic maps. For the SanagaNtem expansion, see chapter 5. 70. In his survey work, M. Eggert, "Imbonga and Batalimo," 3234-35, recognized a new ceramic, labeled Bondongo in the inner Zaire basin, starting from the thirteenth century. The distribution suggests a spread of fishermen from the lower Ubangi to the Ruki up the Tshuapa and the Maringa, but site excavations are needed. 71. For population densities, see F. Goffart, Le Congo, 94; E. V. Thevoz, "Kamerun Eisenbahn-Erkundungs-Expedition Bevolkerungsdichte"; G. Sautter , De ['Atlantique, vol. 2, endmaps; P. Gourou, Pays tropicaux; A. AnnaertBruder "La densite de la population du Congo-Kinshasa," 161; 1. Vansina, "Peoples of the Forest," 110. Vegetation can reflect low or high densities in the past, but not in the very 318 Notes to Pages 69-74 remote past: P. Foury, "Indications donnees par l'etat actuel de la vegetation sur la repartition ancienne des groupements humains," gives examples of vegetation following a high level of farming activity some three to four centuries ago. R. Letouzey, Etude phytogeographique du Cameroun, 142-45, confirms for Lophira alata. Stands of Gilbertiodendron dewevrei (ibid., 252), an evergreen forest climax habitat, point to undisturbed conditions for the last half millennium or more. But vegetation can probably not attest to the situation a millennium ago, and, at present, estimates about the time a given habitat needs to establish itself remain approximate. Still for the last half millennium forested habitats could tell us much about the distribution and the relative intensity of farming and farmers. 72. W. De Craemer, 1. Vansina, and R. Fox, "Religious Movements in Central Africa," 463-65. Chapter Three: Tradition: Ancient and Common 1. "A" stands for Appendix: Comparative Lexical Data. "CS" stands for Comparative Series and the reference number in M. Guthrie's Comparative Bantu, vols. 3 and 4. For the form of the CS and the evidence where no CS form is given, see Appendix. 2. Villages disappeared only among the Nen, Fia, Saa, and Mvumbo (Ngumba) peoples in Cameroon. The hamlet also was the normal settlement near the middle Uele, but there mostly among non-Bantu speakers. The district very rarely became coterminous with the village. On the upper Ngiri there were no districts in the area in precolonial days. Libinza is a modern ethnic name derived from the name for a village (P. Van Leynseele, "Les Libinza de la Ngiri," 32-33). I do not know of any recorded disappearance anywhere of the House. 3. 1. A. Barnes, "African Models in the New Guinea Highlands,"; P. Van Leynseele, "Les Libinza," additional thesis no. 1. For general descriptions, d. A. Cureau, Les societes primitives de l'Afrique equatoriale, 324-38, or 1. Dugast, Monographie de la tribu des Ndiki, 2:451-53. For a portrait of a "big man" (Pando of the Gunabemb in southeastern Cameroon), see C. W. H. Koch, Das Lied des Landes, 78-82, and 72-74, 91-93; Kota and Nen are languages in which the term for leader literally means "big man" (-nen). 4. H. Koch, Magie et chasse dans la forb camerounaise, 46. 5. Ibid., 48. 6. "Rich," "famous," kum, reflex of CS 1263-64 derived from CS 1265 "big man." For a distribution map of CS 165, see 1. Vansina, "Deep Down Time." 7. This was true in the later common culture. The trophies of these animals belonged to leaders who wore them. To appropriate a trophy was a formal declaration of insurgence. Leopard and python flesh were consumed only by a few persons in formal communion rites presided over by the leader. A special term for such "animals of leadership" exists in most languages. This custom does not imply that at one time these people were primarily hunters. It derives from Notes to Pages 74-75 319 the notion of "living being," which includes only organisms that move. The "kings" of several classes of animals are the animals of leadership. Funeral rites for slain leopards were especially linked to leadership. Compare, for instance, Loango c. 1640-1668 in O. Dapper, Naukeurige Beschrijvinge der Afrikaensche gewesten, 542-43; and c. 1700 in N. Uring, The Voyages and Travels of Captain Nathaniel Uring, 36-37, who mentions "the gallows," as do A. Raponda-Walker and R. Sillans, Rites et croyances des peuples du Gabon, 235-36, for an Ivea village in 1907, which in turn recalls similar rites all over northern Zaire. 8. W. MacGaffey, "The Religious Commissions of the BaKongo." On the egalitarian ethic embedded in witchcraft generally, see 1. Vansina, The Tio Kingdom , 240-41. Ethnographic comments on the notion of "luck" also fit this type of leadership. Because the notion "luck" is not recorded in most dictionaries, the potential proto-western Bantu is still unknown. See, for instance, G. TeBman, Die Bubi auf Fernando Po, 130-39. On the pangolin, cf. M. Douglas, "Animals in Lele Religious Symbolism," 50-56; D. Biebuyck, "Repartition et droits du pangolin chez les Balega"; W. de Mahieu, Structures et symboles, 38-40. Its unusual behavior and the taxonomic difficulties it presents have made it the emblem of this type of authority. 9. T. Irstam, The King of Ganda. For a list of "traits" of "divine kingship ," see 1. Vansina, Le royaume kuba, 98-103; F. Hagenbucher Sacripanti, Les fondements spirituels du pouvoir au royaume de Loango, 78-87; A. Aymemi, Los Bubis en Fernando Poo, 57-69. 10. Only larger-scale excavations will allow archaeologists to find and estimate respective sizes of House residential quarters. 11. House space was indicated by ditches across the village street (Foma: B. Crine, La structure sociale des Foma, 19), separate plazas (Boa: Vedy, "Les Ababuas," 200-202), open spaces between the "hamlets" corresponding to the quarters (Tio: 1. Vansina, Tio Kingdom, 506), or by communal houses for each House in the village street (Bekwil: A. Cureau, Les societes primitives, 242). 12. To use a felicitous adjective of C. Levi-Strauss without evolutionary implications. 13. For instance, by the Myene forms: Mpongwe ob6ta, "parent"; Tsogo ob6ta, "clan"; and northwestern Bantu as Kosi eb6a, "House plaza," and Eso liotsi, "House." The form is universal with the meaning "House" in the inner basin block of languages (Mongo). 14. 1. Vansina, "Lignage, ideologie et histoire en Afrique equatoriale," 140-48. The etymology of terms for "House" does not usually refer to gender, as one would expect in unilineal descent situations. The etymologies found refer to: "house," "bed," "hearth," "hearthstones," "plaza of authority," "offspring of," "grandchildren of," "species," "root," "shield," and "communal shed." Contrast this with "belly/female sex" in matrilineal groups of lower Zaire and the lower Kasai areas, and "sister" among matrilineal Myene and present-day patrilineal Mpongwe. Contrast also with "male" in the now strongly patrilineal groups of south-central Cameroon and "penis"/"vagina" around Lake Mai Ndombe, where there is now a double descent system. 320 Notes to Pages 75-78 15. The meaning of some reflexes given by M. Guthrie needs correction, e.g., G. Hulstaert, "Notes sur la langue des Baf6to," 129, mb6a: "road," "at home," "home village." Also add ps 482 to CS 447. In this connection see CS 694 "to name" and ps 192 (3:187) "namesake." Namesakes were a way to integrate outsiders among some hunting and gathering groups with high mobility, including the Mbuti of Ituri. Cf. R. Me. Netting, Cultural Ecology, 15. The form is ancient but not western Bantu. Still, it includes more western reflexes than Guthrie has, for instance, for Bira. 16. G. Laumanns, Verwandtschaftsnamen und Verwandtschaftsordnungen im Bantugebiet; update with M. Guthrie, Comparative Bantu, 2:178d. Because kinship terms form a coherent system, two different comparisons are involved: one for the form and meaning of each term, one for the whole system. For an instance, see 1. Vansina, Children of Woot, 105-6, 108-9, 291-95. 17. Although the original for "grandparent" may well have been CS 1204 in form, ps 309, cited with CS 1204, is probably the same. It is not just eastern Bantu as claimed by Guthrie, because many reflexes are found in Gabon and in the inner basin. CS 992 is probably proto-western Bantu and an innovation with respect to CS 1204. A variety of other terms also occurs, some with shared etymologies for "close ancestor." Note the reduplicated shape of the form as for "my" father and mother, forms due to baby talk. 18. G. P. Murdock, Africa: Its Peoples and Their Culture History, 29-30, after idem, Social Structure, 184-259, where the supposedly evolutionary development is given. 19. 1. Vansina, "Peoples of the Forest," 88 n. 46 for Fang, Mongo, and Kama, respectively with four, eight, and three main forms of marriage. P. Laburthe Taira, Les seigneurs de la foret, 239-45, for the Bane, also with eight or nine primary forms. Cicisbeism was particularly widespread, since it was a mechanism to attract young men to one's House. As G. Dupre, Un ordre et sa destruction, 148-49, has argued, in general various forms of marriage were products of political decisions and not unworldly systems exchanging women automatically. The error in anthropological theory stems from the fact that for Europeans there could be only one valid marriage form at a time. Early on in the colonial period lawyers, administrators, and missionaries chose the form which they felt was "the only valid one." 20. E. Meeussen, Bantu Lexical Reconstructions, 5: 3. 21. Village size for 111 villages in southeastern Cameroon (1910-1912) yields 104.8 on average with 49.16 for the smallest average per ethnic group and 325 for the highest (on the margins of the grasslands in a war-torn situation). Extremes go from 5 or, if discounted, 12, to 787 or, if discounted, 462, with a clustering around 100; see C. W. H. Koch, "Die Stamme des Bezirks Molundu," 286-304. Early impressionistic figures elsewhere, or numbers for single villages elsewhere in the area, tend to confirm these results. 22. For instance, P. De Maret, "The Ngovo Group," 120-21, or for the sites near Loango, 1. Denbow et aI., "Archaeological Excavations," 37-42. Notes to Pages 78-80 321 23. A. Cureau, Societes primitives, 218-44; 1. Annaert, Contribution a l'etude georaphique de {'habitat et de {'habitation indigenes en milieu rural dans les provinces orientale et du Kivu. 24. But in some languages such as lofe or Foto the term means "path," clearly outside the village, and yet also "our village." Cf. n. 15. 25. Contrary to A. Cureau, Societes primitives, 214, who defines the village as "created by one man, for one man." 26. G. Dupre, Un ordre et sa destruction, 117-22. 27. From CS 1003 such terms as nkaan, "initiation," in southern Zaire, including some Kongo and Kuba. From CS 1709,1711: *-teende, "an initiate." The distribution includes Maniema and portions of the north and the northwest. Also *-yende, "an initiand" (cf. Bushong -byeen, Kete -bende, Nen -yindi, Duala -enda, "to circumcise," and bwende, "initiation"). 28. Boy's initiations have often been described. For example, d. W. de Mahieu, Qui a obstrue la cascade?; A. Droogers, The Dangerous Journey; L. Perrois, "La circoncision Bakota (Gabon)" (children of high status); H. Van Geluwe, Les Bali et les peuplades apparentees, 64-67 (no circumcision then); 1. Vansina, "Initiation Rituals of the Bushong" (also without circumcision); H. Van Geluwe, Mamvu-Mangutu et Balese-Mvuba, 70-72, for the middle Bomokandi, and 158-60 for Ituri. Unlike the practice of all other peoples, neither initiation nor circumcision was common among most peoples (Mongo group) of the inner Congo Basin. 29. Mumbanza mwa Bamwele, "Histoire des peuples riverains de I'entre Zaire-Ubangi," 163 and 155-73; A. Wolfe, In the Ngombe Tradition,S!. The term used is kola, a reflex of CS 1194. The Duala and Kpe age groups used another term (E. Ardener, Coastal Bantu of the Cameroons, 70-71); the Lele and Mongo term was derived from CS 664. "row." For the Lele, see M. Douglas, The Lele of Kasai, 68-84; for the Mongo, see R. Philippe, Inongo: Les classes d'dge en region de la Lwafa (Tshuapa). 30. See discussions in chapters 5 (Kongo, Loango, Cameroon, and Gabon) and 6 (Maniema). For minor associations, e.g., of elephant hunters along the Lokenye. see 1. Cornet, "La societe des chasseurs d'elephants chez les Ipanga," also known among the Kuba. 31. P. Laburthe Tolra, Seigneurs de la foret, 318-51; G. Hulstaert, Les Mongo, 43-46; 1. Vansina, "Vers une histoire des societes Mongo," 18-20 and references; idem, The Tio Kingdom, 359-65. Quarrels within a village were even more restricted. Fighting of this sort went by a special term (CS 675), and usually no weapons more lethal than sticks were allowed. 32. CS 1872 may not have been quite proto-western Bantu according to Guthrie. But in addition to his references one can add many from the language groups he mentions or others close by (e.g., A80, C 30, C 60). One should add Basaa -tumb, "to hit strongly," and tumbna, "to hit each other ferociously," which not only expands the distribution of the term considerably but also pro- 322 Notes to Pages 80-82 vides good reason for assuming that the term derived from the verb "to burn" at a very early date in western Bantu history. 33. For instance, F. Autenrieth, Ins Inner-Hochland von Kamerun, 105-9, for a minor trade war; E. Sulzmann, "La soumission des Ekonda par les Bombomba," for the war of conquest by Ikenge north of Lake Mai Ndombe or the famous war of Pupu in northern Gabon c. 1880. Cf. A. Meyamm, "Zog Djo: Un puissant sorcier, un chef et une divinite bakwele," 17-21. Pupu is mentioned by Pecile as an ethnic group in 1885: E. Zorzi, Al Congo can Brazza, 563, and 557-63 for local wars. See also L. Perrois, Chroniques du pays kota, 33-48; R. Deschamps, Traditions orales et archives au Gabon, 66, 75. 34. The term is found from the Gabon coast (Nkomi) and the Kongo coast to the Kuba in the southeast and as far as the Mongo and Ngombe languages in the northeast (J. Vansina, Children of Woot, 314 n. 7, for part of the distribution ). "Village charm" is the meaning everywhere, except in Nkomi, where it is "hero, ancestor." Were it not that the distribution is continuous, and that so far no Cameroonian reflexes have been found, one would conclude proto-western status. This does not seem to be the case. An innovation in the southwestern block of languages after their split from northern Zairian languages, followed by diffusion to the inner basin, fits the known data well. 35. See Appendix. 36. H. M. Stanley, In Darkest Africa 1:141, for the quote. For "dead zones" in southeastern Cameroon, for instance, see B. Forster, "Aus dem Siidostwinkel Kameruns," 158; A. Schulze, "German Congo and South Cameroons ," 129, 132, 141-42, 157-61, and the 1902 map by M. Moisel ("Unbewohnt" "Tote Zone"). 37. L. Fievez, "Le district de l'Equateur" (each district has its tattoos); A. Engels, Les Wangata, 17-18; A. Verstraeten, "Les tatouages." Administrative inquiries for the purpose of identification were made as late as 1930; see 10rissen , "Documentation ethnique: Tatouages," 1-2. 38. Although Guthrie has no CS, *-cenge (village plaza," "capital") is undoubtedly of hoary antiquity. It is common from the Ogooue Delta to the Zaire bend and beyond. Aka and Bobangi have "village plaza." Saa nseng, "plaine, plateau, terrain plat," has the most general sense. The form occurs in all the branches of western Bantu in the equatorial region, except for northwestern Cameroon, the Nen-Yambasa group, and Bioko. The form is probably protowestern Bantu, but the meaning "central village" is an innovation. Because it occurs in Myene, in the southwestern group of languages, in the western part of the northern Zaire languages, and in the inner basin, this may well have arisen before the secession of the northern Zaire languages. 39. There are many more reflexes of CS 714 in western Bantu. CS 714 is formally linked to CS 709 "to join by tying" and CS 711b "to go straight." CS 709 may well have been the source item for the others. CS 665 has the same form as CS 664 meaning "line of objects" and often, in the central Zaire basin, "age grade," but it may not have the same origin. The form of CS 665 is thought to derive from CS 714 through a vowel change. This new form corresponds to that of a set of terms meaning "to arrange" (CS 657), Notes to Pages 82-86 323 "to heap up" (CS 658), which cannot be taken as the root of the etymology for "clan," although the existence of this allied form helps to understand the vowel change in CS 665, since the semantic fields are so close. The item listed by Guthrie as "clan" (CS 779) derives from "village quarter ." Guthrie's claim that there is no "synonymous CS" for "clan" is spurious given the vague character of the terms "clan," "tribe," and even "race" and "family." Terms for "clan" are also often used to designate "species." 40. J. Vansina, "Lignage, ideologie et histoire," 138-40, and chapter 5 (southern Gabon) for cases of obvious clan expansion and contraction in the nineteenth and twentieth centuries. Clan solidarity is expressed by the same terminology used for membership in a district. 41. 1. Vansina, "Esquisse historique de I'agriculture en milieu forestier." 42. The earliest of these is Obobogo; see P. De Maret, "Recent Archaeological Research," 134. 43. Cucumerops edulis was found everywhere in the area as an important secondary crop. One term for it (*-gondo) is perhaps even proto-western Bantu. For archaeological evidence of greens such as Vernonia, see 1. Vansina, "Esquisse historique," 17. 44. P. De Maret, "Recent Archaeological Research," 134. Once again Obobogo is the earliest site. No western Bantu terms have been found as yet for any of these trees, although some terms are regionally widespread. Thus nsafu for Pachylobus is found from the Kongo coast to the lower Lomami at least. For other cases, see A. Bouquet and A. Jacquot, "Essai de geographie linguistique sur quelques plantes medicinales du Congo Brazzaville." Also see n. 57 below. The complex terminology for raffia palms and their products may well yield more proto-western Bantu terms. 45. M. Miracle, Agriculture in the Congo Basin, 31-175; 1. Vansina, "Esquisse historique," 7-10. On the shifting of villages, G. Dupre, Un ordre, 108-13, asserts that in most cases this occurred before soil exhaustion would have been the main cause. 46. On falga, see L. Biffot, "Contribution," 69-74. There are very few actual descriptions of dawn gardens; see G. Le Testu, "Notes sur les cultures indigenes dans l'interieur du Gabon," 542-44. 47. CS 2176 proto-western Bantu. On the southern margins, reflexes of this term meant "sea salt," and "salt from grasses" was designated by a reflex of CS 1031, itself derived from CS 1030, meaning "to filter," because salt was obtained by filtration of the ashes. For an example of the technique, see the illustration in E. Torday and T. A. Joyce, Notes ethnographiques . . . Les Bushongo, 22, 134, plate 17; for a case of restricted access to the famous salt pans of Odzala, see E. Zorzi, Al Congo con Brazzo, 567-69. 48. Comparative material with other rainforest areas in the world supports this conclusion. For instance, the Hmong of Thailand could cut all vegetation and burn it during the dry season without any collective labor, because the season was long enough. See R. Mischung, "SeBhaftigkeit und Intensivierung beim Brandrodungsfeldbau," 241-55. 324 Notes to Pages 86-89 49. M. Miracle, Agriculture, 43-61; P. T. Perrault, "Banana-Manioc Farming Systems of the Tropical Forest," 236-40. Burning was not practiced in latenineteenth -century Bioko (G. TeBmann; Die Bubi auf Fernando Poo, 46-49) or among the Nyanga of Maniema (D. Biebuyck, Rights in Land and Its Resources among the Nyanga, 21-25). 50. Cf. chapter 2; also M. Watson, Agricultural Innovation; 51-54. 51. M. Watson, Agricultural Innovation: "sugar cane," 24-30; "citrus fruits," 42-50; "eggplant," 70-71. For eggplant in the crop rotation, see G. Dupre, Un ordre, 77-78; S. Jean, Les jacheres en Afrique tropicale, 52. For sugar cane wine, see M. Miracle, Agriculture, 203-5, compared with 201-3 (palm wine). 52. M. Watson, Agricultural Innovation, 66-69; 1. Vansina, "Esquisse historique," 25; the zanj name in southern Arabia suggests an introduction from East Africa, not the reverse, before A.D. 950. 53. M. Miracle, Agriculture, 137-42 (lower Congo, lower Zaire); T. Obenga, La cuvette congolaise, 78-79 (Likouala Basin); 1. Dugast, Monographie de la tribu des Ndiki 1:114-21 (Nen country); G. TeBmann, Die Bubi, (green compost on Bioko); G. Sautter, De {'Atlantique aufieuve Congo, 262-64 (Alima-Sangha); P. Van Leynseele, "Les Libinza," 55-56; Mumbanza mwa Bamwele, "Histoire," 219-22, 254-58, 312-14, 319-26 (Ngiri). For an older account, see S. Brun, Schiffarten, 15 (Loang6: knee-deep ditching). Much of this intensification was due to new commercial conditions accompanying the growth of the Atlantic trade and to the introduction of cassava in that context. 54. M. Miracle, Agriculture, 61-72,74-87 (two field systems in the south), 87-103,104-7 (Sudanese agriculture). 55. G. Hulstaert, "Nordkongo: Der zentrale Teil," 1:724, about the northwestern Mongo of the Ruki area. 56. S. Bahuchet, Les pygmees aka de la foret centrafricaine, 189-226; idem, Etude ecologique d'un campement de pygmees babinga," 526-32, 547-49; G. Bibeau, "De la maladie ala guerison: Essai d'analyse systemique de la medecine des Angbandi du Zaire," 96-97. The Ekonda know some 50 different edible mushrooms, according to D. Van Groenweghe, Bobongo: 260 n. 5. ''The nutritional state of the forest dwellers is in general at least as good as that of the savanna dwellers, this being especially the case for the Pygmies and Pygmoids" (1. Hiernaux, "Long-term Biological Effects of Human Migration from the African Savanna to the Equatorial Forest," 214). 57. A. Chevalier, "Les rapports des Noirs avec la nature." In general, cf. A. Raponda-Walker and R. Sillans, Les plantes utiles du Gabon (includes comparative terminology); G. Hulstaert, Notes de botanique mongo (includes terminology ); A. Bouquet, Feticheurs et medecines traditionelles du Congo; A. Bouquet and A. Jacquot, "Essai de geographie linguistique" (comparative terminology); E. Motte, Les plantes chez les pygmees aka et les Monzombo de la Lobaye. 58. See 1. Vansina, "Habitat, Economy, and Society in Equatorial Africa," for the reasons why. Notes to Pages 90-91 325 59. Especially S. Bahuchet, 'Etude ecologique," 529-30, 547-49, for substitution ; and G. Le Testu, "Notes sur les cultures indigenes dans l'interieur du Gabon," 547-48; P. L. Martrou, "Le nomadisme des Fangs," 512-13; and M. De Ryck, "La chasse chez les 'Lolia-Ngolu,' " 233, for seasons. 60. Reflexes of -tamb-, "to set a trap," and -tambo, "trap" (CS 1659), which Guthrie holds to be the common Bantu form, are quite common in the portion of the area west of the Zaire, including Kongo and Ubangi. The later eastern Bantu form -teg-, "to set a trap" (CS 1699), also occurs in Kongo and is also reported in Nzabi, Metuku, and Tembo. The northern and central portions of the area use several other and fairly old terms, showing something of the dynamics in the technology of trapping. On traps, see G. Lindblom, lakt- och Fangstmetoder bland afrikanska folk. 61. For some common terms for fauna, see M. Guthrie, Comparative Bantu, 2:176-77. Common terms for game such as warthogs, water antelopes, bongo, monkeys, elephants, and even leopards are reflexes of proto-western Bantu and stress the importance of hunting among them. In addition, the common western Bantu term meaning "to chase game" (CS 129) refers to the drive hunt (with or without nets). Yet it is possible that some other terms are ancient loans from an autochthonous language. Because these languages are extinct, this can be proved only by showing (a) that the terms are innovations in western Bantu and, perhaps, (b) that their shape shows some features which are aberrant . The "A" reflexes of CS 904 "to hunt" would be such a possible form. 62. Descriptions in H. Koch, Magie et chasse, 177-229 (Djue); G. Dupre, Un ordre, 49-53 (Nzabi), M. De Ryck "La chasse," 233-36 (Lalia). For terms for spear and hunting net, see 1. Vansina, "Do Pygmies Have a History?" 43842 ; for "bow" and "crossbow," see idem, "Deep Down Time." On hunting as a barometer, see e.g., M. Douglas, Lele of Kasai, 207-8,215-18,251-56. 63. G. Sautter, "Le plateau de Mbe," 27-30. 64. The Kele of Gabon were perhaps such a group in the nineteenth century , although they were also passionate traders (P. du Chaillu, Voyages et aventures dans l'Afrique equatoriale, 433-40; also R. Avelot, "Notice historique sur les Ba-Kale," despite its unsubstantiated conjectures). For peoples in northern Maniema, see W. de Mahieu, Qui a obstrue la cascade? 12-15 and index "chasse"; D. Biebuyck, Rights in Land, 14-26. 65. CS 1949 "dug out canoe," CS 735 "to paddle," CS 790 and 1014 "paddle ," CS 638 "to fish with a line," and the derived CS 640 "fish hook." CS 333, 427,429 and 1858, all interconnected forms, "fish." CS 731 meaning "to dip" in the western part of the area may have a connection with CS 732 "to fish with a basket or net," which for Guthrie is proto-Bantu. But more attestations of the form are needed. As the vocabulary for fishing and its tools is further explored more terms may appear. For this potential, see Y. Ankei, "Connaissance populaire du poisson chez les Songola et les Bwari." But most of the similarities in terminology or in technology (e.g., the brasero) are probably due to multiple diffusion, because fishermen were also highly mobile traders at all periods. 326 Notes to Pages 92-94 66. For instance, in the Soan group the splits between So (fishermen) and Mbesa (landlubbers) and Lokele (fishermen) and Eso (landlubbers), or in the "Ngombe group" the split between the "water people" and the "landlubbers." 67. This history remains unresearched, or rather the detailed research done by social science scholars is generally not deemed of enough interest to be published at length. General monographs of the type of A. Goffin, Les pecheries et les poissons du Congo, are incomplete and old. For the potential see for example: 1. P. Gosse, "Les methodes et engins de peche des Lokele," or Y. Ankei, "Connaissance populaire du poisson." 68. 1. Vansina, "Esquisse historique," 25 n. 46. 69. For the terminology of sheep, see C. Ehret, "Sheep and Central Sudanic Peoples in Southern Africa," but also 1. Vansina, "Esquisse historique," 19 n. 28. All the terms in forest areas are also common in neighboring savanna areas, and none stretch right across the rainforests. Nineteenth-century evidence for trade showed trade in sheep from the margins of the savanna into the forest areas. As to pigs, CS 887-88 refer to "warthog" here and "pig" there. The meaning "pig" is clearly an innovation. By 1900 pigs were found only in the northwestern part ,of the area studied and on its southwestern and southern margins. 70. See the Appendix for terminology. Ceramics are directly attested on most sites. For specific industries and technologies, see E. Coart and A. De Hauleville, La ceramique; E. Coart, Les nattes; H. Loir, Le tissage du raphia au Congo beige. Published inventories of collections held by museums are also a rich source of information, especially on carved items and on metalwork. 71. N. Van der Merwe, "The Advent of Iron in Africa," 485-501. For distributions, see E. Cline, Mining and Metallurgy in Negro Africa; L. Frobenius and L. Ritter van Wilm, Atlas africanus; E. Maquet, Outils de forge du Congo, du Rwanda et du Burundi. 72. H. Koch, Magie et chasse, 164-7l. 73. For a fourth form meaning "to sell" (CS 1697), an innovation from CS 1698-99 "to set a trap"(!) in the lower and middle Zaire area, see chapter 5 (p. 147). 74. J. Denbow et aI., "Archaeological Excavations," 39 (Loango); B. Ciist, "Un nouvel ensemble," 45 (Gabon); L. Fiedler and 1. Preuss, "Stone Tools from the Inner Zaire Basin," 179-82 (Lake Tumba). 75. P. De Maret, "The Ngovo Group," 116. 76. See L. Sundstrom, The Exchange Economy of Pre-Colonial Africa, for a general ethnographic overview. See also A. Thonnar, Essai sur Ie systeme economique des primitifs d'apres les populations de l'Etat Independent du Congo, 82-114. 77. H. Tegnaeus, Blood-brothers, 103-18, 132-33. 78. M. Eggert and M. Kanimba, "Recherches archeologiques et ethnographiques dans les regions de l'Equateur (Zaire), de la Cuvette de la Sangha et de la Likouala (Congo)" (Sangha, Djah, and Likouala-aux-herbes); R. Lan- Notes to Pages 95-98 327 franchi and B. Pinc;on, "Resultats preliminaires des prospections archeologiques recentes sur les plateaux et collines Teke en R.P. du Congo," 25-27 (Malebo Pool to Alima). 79. W. MacGaffey, Religion and Society in Central Africa, 1-18; L. MallartGuimera , Ni dos, ni ventre, 23-101,217-19, among others argues, however, for the existence of a single "symbolic system." This may well exist, but then it does so at an unconscious, undogmatic level. 80. C. Turnbull's description of monotheistic pygmies venerating "the forest " as Molimo by playing on trumpets is to be treated with caution (Wayward Servants, 76-80, 251-67; idem, "The Molimo: A Men's Religious Association among the Ituri Bambuti"). The term is a common reflex of CS 619. It may well have been used by one or another Christian group to signify "God." It was used as a borrowed term in that sense by Ngbandi in the 1970s for the name of a new religious movement (G. Bibeau, "De la maladie," 169-74). 81. Linked to CS 621 "to get lost" and CS 620 *-kudimba, "to forget." 82. The use of the term nkisilngesi in the forests of Gabon, southern Congo, and portions of its southern rim in Zaire, as well as in all the southern savannas, is typical. It refers to a charm and to the entity activating it: in the forest, a nature spirit; in the savanna, a hero. 83. Supplement Guthrie with A. Kagame, La philosophie Bantu comparee, map 3. 84. As 1. F. Thiel, Ahnen, Geister, Hochste Wesen, 169-77; for a discussion of "Nzambi," see ibid., 38-58. 85. A. Kagame Philosophie Bantu comparee 149-50, E. Andersson, Contribution a l'ethnographie des Kuta 2:23-31; J. Ittmann, Volkskundliche und religiOse Begriffe im Nordlichen Waldland von Kamerun-Afrika und Uebersee, 30-32; E. Ardener, Coastal Bantu, 92, for "sky" and "God" in Kpe and Duala; P. Wurm, Die Religion der Kiistenstiimme in Kamerun, 32-34. 86. W. MacGaffey, "The Religious Commissions of the BaKongo." His observations seem to hold throughout equatorial Africa. 87. Distribution of poison ordeals is in H. Baumann, Die Volker Afrikas 1:633, map 34; and W. Schilde, Orakel und Gottesurteile in Afrika, 229-61, esp. 247-58. The common names in the area are *-kasa and *-bundu in the western half and often together, *-ipomi in the south, and *-avi in the southeast and east. The trees used for the poison in most of the area were Strychnos species and Erythrophleum guineense; E. Perrot and E. Vogt, Poisons des fieches, 81-96, 157. The arrow poisons used, on the other hand, were often Strophantes spp. or cadaveric fluids. 88. "Abnormality" is not often included in dictionaries. Nor is "misfortune " or "good fortune." Good fortune corresponds to "strength" (CS 840). See also CS 909 and ps 249 (3:239). 89. Because the specific terminology for the tools and practice of medicine has not yet been gathered in enough detail, it is still too early to distinguish clearly between ancestral, later common, and still later innovative practices. 328 Notes to Pages 98-103 Even a discussion of the notions of health, illness, personhood, and individuality , although germane here, has been reluctantly omitted for the sake of concision . The notion of health implies "wholeness," and illness was a deficiency upon which the patient could act. The theory of personhood was complex. But to date none of these notions has been documented enough from different parts of the area to conclude anything further. Terms such as "dancing" or "drum," which are proto-western Bantu and were often used as shorthand for "performance ," religious or not, should not be considered as specific evidence for the practice of medicine. 90. W. De Craemer et aI., Religious Movements, 472-74. 91. The importance of an "exit" option and hence "empty" space has been brought home to me by M. Schatzberg. For example, see his The Dialectics of Oppression in Zaire, 137-38. 92. On 4 per km2 , ct. 1. Vansina, "Le regime foncier dans la societe kuba"; idem, "The Peoples of the Forest," 79, for an estimate of 3.75 per km2 c. 1880, and 110 for map. Chapter Four: The Trail of the Leopard in the Inner Basin 1. See chapter 3, n. 66. A common term, elinga or balinga, designates fishermen south of the great bend of the Zaire from the confluence of the Ubangi to the middle Lomami. By 1900 the hunter gatherers had disappeared from the region north of the Zaire River bend and for the most part had been absorbed by the farmer trappers in most forest areas on dry soils of the eastern half of the inner basin. 2. M. Eggert, "Archaologische Forschungen," 3223-25; idem, "Remarks on Exploring Archaeologically Unknown Rain Forest Territory," 286-91 and 320, for the few dates available. Although the eponymous site of Bondongo itself lies inland well east of Lake Tumba, its ceramics stemmed from the clay banks of the Ruki or the Momboyo; idem, "Der Keramikfund von BondongoLosombo "; E. Sulzmann, "Zentralafrikanische Keramik aus voreuropai"scher Zeit." 3. M. Eggert, "Remarks," 312. For iron, see idem, "Archaologische Forschungen," 3235-37. 4. See chapter 3, n. 78. 5. The evidence about the western Mongo is the richest of all data in the inner basin. For lack of data and research data, see map 1.3; the amount of data needed for upstreaming becomes greater and needs to be more precise the further one moves back through time. With future intensive local research on "words and things" one can hope to obtain a complete picture of the situation c. A.D. 1000 and even earlier. 6. For detail, see 1. Vansina, "Vers une histoire des societes mongo," 1225 , for the situation in the nineteenth century and earlier history. Etukd strictly meant "an anthill used as a stone for the hearth." The term lokutu, "patrimony," Notes to Pages 104-7 329 is derived from nkutu, "corpse," but is itself the source of bokutu (dialect: bokutsu) , "rich person," from which many ethnonyms are derived all over the basin of the Lukenie, including okusu on the middle Lomami. 7. On the terminology itself, see E. Boelaert, "Terminologie classificatoire des Nkund6"; for Mongo explanations of their nk6l6!nkita ("mistress"! "profit") system and the Crow terminology, see G. Hulstaert, Le mariage des Nkundo, 164-72; on the importance of valuables and bridewealth, see ibid., 105-230. Nsomi was later borrowed by the Mpama, in Ngenge, and in Lingala as "important person, free person" and as ncomi, "free man," in Bobangi. 8. However, iron smelting was practiced in the area as late as the eighteenth century by the "Losakanyi" (near Ingende and later west of Lake Tumba); see n. 3. The term konga, for copper bangles which formed the core of matrimonial compensation, in recent times refers to the Kongo people. Copper came from the mines of lower Congo, perhaps for many centuries. The dating of more hoards of valuables, as well as of the appearance of copper in the area, should strengthen this reconstruction. 9. See chapter 3, n. 7. Apart from the common noun for the leopard, the animal often was respectfully referred to in a circumlocutory way by a praise name. North of the Zaire bend and east of the Lualaba this practice became so usual that the common western Bantu noun was lost. 10. As the Belgian colonial administration realized. Early instructions to its political agents exhorted them to follow this trail to establish the hierarchy of "traditional" chiefs. Especially in the Oriental Province it was quite common to justify the proposed structure of a political hierarchy by referring to the trail of the spoils of the leopard. See, for instance, A. Bertrand, "Quelques notes sur la vie politique, Ie developpement, la decadence des petites societes du bassin central du Congo," 81-82. 11. See, for instance, N. Rood, Ngombe-Nederlands-Frans woordenboek, 350. 12. And indeed in 1877 Stanley fought a large fleet at Basoko. H. M. Stanley, Through the Dark Continent, 497-501. He gives several other versions elsewhere, but they all extol the size of the enemy fleet and his own bravery. 13. G. Sautter, De l'Atlantique, 244; 259-60 and 261 for Bonga as the "Venice of Africa." In this sector Lukolela was on higher ground but less favorably located to control traffic from the affluents. Some other sites controlling confluences were: Bolobo (Alima), Irebu (Lake Tumba), Mbandaka (Ruki), Mobeka (Mongala). But note that Lisala, a major town in the nineteenth century , had only the advantage of lying on a headland. Iboko, where Stanley also fought a major battle in 1877, was not near any confluence. 14. The term is found in Aka, Lingala, Ngombe, Doko, Mbudza, Bati, Boa, and Komo in Maniema. It does not occur in Soan. 15. 1. Maes, "Les sabres et massues des populations du Congo Beige," for the distribution of sabers; H. A. Johnston, George Grenfell and the Congo 2:766, 775, for pictures. Perhaps at first throwing knives had been used for this purpose 330 Notes to Pages 107-10 in the forest areas. On the throwing knife, d. 1. Maes, "Armes de jet des populations du Congo BeIge"; 1. Vansina, Art History in Africa, 168-70, and add Loango c. 1610 to the map there. 16. See, for example, A. Lejeune-Choquet, Histoire militaire, 191-200, 209-23. G. Burrows, The Curse of Central Africa, 241-44; P. Salmon, "Les carnets de campagne de Louis Leclercq," 268-69, for a comparison of nineteenthcentury warfare north of the Zaire bend and in Mongo country. 17. G. Burrows, Curse of Central Africa, 236-37, describes Bango military training performed as a war dance. 18. For instance, among the Bangba; M. Fernandez-Fernandez, "Le rite et I'experience," 47-69, and esp. 69-71. In addition they had a dependent of "brother-in-law." The ancestor was supposed to have come to live in the village as a dependent of his wife. 19. For a description of the Omaha system, see L. De Sousberghe and 1. Ndembe, "La parente chez les Lokele," 734-40; B. Crine, La structure sociale des Foma, 23-35. 20. See F. Heritier, L'exercice de la parente, for a structuralist analysis of Omaha, Crow, and related systems of kinship. She stresses the links between such nomenclatures and patterns of marriages. 21. G. P. Murdock, Social Structure, 239-41, shows a 100 percent correlation between Omaha and "patrilineal descent" but not the reverse. F. Heritier, L'Exercise de La parente, 30 and 68 n. 7, correctly concludes that Omaha is therefore not a necessary terminology for patrilinearity. Moreover, Crow and Omaha terminologies do exist in undifferentiated or bilateral descent systems. 22. 1. Vansina, "Vers une histoire," 18,29; F. Requier, "Rapport d'enquete: Chefferie des Nkole nki Yamba," 20-22, for rich details on the various types of gibbets; M. Fernandez-Fernandez, "Le rite et I'Experience," 57, for a symbolic use of such a gibbet. 23. For the python (nguma), for instance, d. 1. Soupart, "Les tatouages chez les Budja," 323; M. Fernandez-Fernandez, "Le rite et I'experience," 48. 24. The most senior ones were usually in the center or brought up the rear. See, e.g., 1. Vansina, "Vers une histoire," 18, 28-29; M. Fernandez-Fernandez, "Le rite et I'experience," 41, 42, 69; 1. Requier, "Rapport d'enquete," 18-20. 25. 1. Vansina, "Vers une histoire," 18. Colonial officials streamlined such assemblies and their hierarchies, but did not create them. 26. M. Fernandez-Fernandez, "Le rite et I'experience," 14; 1. Requier, "Rapport d'enquete," 18. 27. It is a linguistic innovation in the Bantu languages of this area. Although it occurs in a number of Ubangian languages, it would be rash to claim that the Bantu speakers borrowed it from there. The closest of the Ubangian languages to the region is Ngbandi, where ya-occurs with the same meaning, but not as a prefix. Monzombo (of the Gbanziri group), now spoken on the Ubangi below Bangui, uses it in exactly the same way as the Bantu speakers do. See 1. Vansina, "Vers une histoire," 27-28, 50-51, nn. 100 and 101. Notes to Pages 111-14 331 28. For the divergence of opinion as to whether the Lokal6 hunters and gatherers were pgymy or not, see Bosek6nsombo, letter 19 February 1971, 5; G. Hulstaert, "Petite monographie des Bondombe," 37-45, 84-85; idem, Elements pour l'histoire Mongo ancienne: 40; G. Hulstaert and Nakasa Bosek6nsombo, "Encore Bondombe" 196-97, 198-200,209-10,212-17. 29. 1. Vansina, "Vers une histoire," 26, 30, and nn. 84, 89, 115, 116; A. Bertrand, "Le probleme Mongandu." 30. 1. Vansina, "Vers une histoire," 31 and n. 122. Along the Tshuapa River near Bondombe (Mondombe) and farther west, the spearmen were stopped by efficient archery. 31. 1. Requier, "Rapport d'enquete," 22, for nongolndongo as "oldest lineage ." Mpama and Ntomba of Mai Ndombe have ndongo, "capital" (the meaning of the modern town Inongo), and Bushoong ndweengy, "harem." G. Hulstaert, Dictionnaire lomongo-fram;ais, has ndongo, "harem" and later "political capital ." On sabers and on the hand piano, see 1. Maes, "Sabres et massues," 358-60, 364. 32. 1. Vansina, "Vers une histoire," 27. This date is derived from "dead reckoning" backward in Kuba history. Kuba chiefdoms began to appear a century or more before the unification of the state c. 1625. The battle occurred before the first Kuba villages crossed the Sankuru. An "at the latest 1500" date fits well with estimates given for the onset of the nkumu expansion. 33. F. E. Dhanis, "Le district d'Upoto et la fondation du camp de I'Arouwimi," 15-17,26,27-28; D. Wijnant, "Het Doko volk in hun handel en wandel," 593; idem, "De Doko's," 422. 34. D. Wijnant, "De Doko's," 410-15 (sacred kingship and officials); idem, "Het Doko volk," 211-13; F. E. Dhanis, "District d'Upoto," 31 (harem and dependents). 35. D. Wijnant, "De Doko's," 419. 36. D. Wijnant, "Het Doko volk," 206-7 (the six main matrilineages); idem, "De Doko's" 410 (Bondongo dynasty); idem, "Eene bladzijde uit de geschiedenis van Boela," 605-8 (the dynamics of succession and secession). 37. This spread was assisted in part by the dynamics of marriage. Married women continued to follow their husbands to the latters' places of residence, and despite the prevailing endogamy eventually lineages spread beyond each town; see D. Wijnant, "Het Doko volk," 210. For districts, see F. E. Dhanis, "District d'Upoto," 15-17,26; D. Wijnant, "Eene bladzijde," 605-8. 38. Mumbanza mwa Bamwele, "Histoire," 105-16, notes that the lando, Bamwe, and Ndolo remember the adoption of the matrilineal norms (107) which diffused (107-9). He erroneously thinks that the phenomenon was an original innovation stemming from the economic importance of women and the general poverty that did not allow for marriages with matrimonial compensation (11016 ). Some of these groups bordered on Doko before the Ngombe expansion of the eighteenth century (45, and indeed the Bolondo were of Doko orgin). Hence it is most likely that matrilinearity spread from the Doko into the area during the eighteenth century. 332 Notes to Pages 114-16 39. D. Wijnant, "Het Doko volk," 207-8; idem, "Het Doko volk in hun handel," 584 n. 1. Doko crossed to the left bank before the Ngombe did. They were matrilineal, but later adopted the Ngombe language and by 1920 were "losing" their matrilinearity. But for matrilinearity in residence and in bridewealth as late as 1960, see N. Rood, "Lidoko et Mowea," 126. For the northern Mongo groups, see G. Van der Kerken, L'ethnie Mongo, 569-71, 573; De Coster, "Coutume mongo," 89/1 (division leopard, succession: matrilineal after "brothers"); L. De Sousberghe, Don et contredon de la vie, 125-26, 138 (bilateral for descent but matrilineal for succession after brothers). 40. 1. Vansina, Art History, 32; E. Sulzmann, "La soumission des Ekonda par les Bombomba," 1-5. The organization of these peoples shows a few features that recall institutions of the patrilineal spearmen farther east. To resolve the question, more in-depth study is needed of the Mongo populations of the Maringa and Lopori. 41. The process described in this paragraph is suggested by the distribution of patrilineal succession by brothers as opposed to patrilineal succession by sons. Primogeniture was a known Ngbandi practice of succession; see B. Tanghe, De Ngbandi naar het leven geschetst, iii-iv, 82. The ambiguous meaning of "oldest" was and still is common in many patrilineal systems in $eneral. Localization of the process derived from the tradition that the first Abandia chiefs spoke Ngbandi. Suggested chronology is arrived at by "dead reckoning" backward from the Nzakara and Abandia chronologies (n. 42). For the Ngbandi practices in the nineteenth century, see n. 45. 42. E. de Dampierre, Un ancien royaume bandia du Haut-Oubangui, 15781 ; A. Hutereau, Histoire des peuplades de l'Uele et de l'Ubangi, 61-137. 43. A. de Calonne-Beaufaict, Les Azande; A. Hutereau, Histoire, 140246 ; H. Vanden Plas, "La langue des Azande," 1:38-60; E. E. Evans Pritchard, The Azande, 267-83. 44. B. Tanghe, De Ngbandi naar het leven, 87-88 (historian), 107-10, 13952 (war); idem, De Ngbandi: Geschiedkundige bijdragen (history); G. Bibeau, "De la maladie," 82-88, 92-93, 98-102. 45. On leopards, see B. Tanghe, De Ngbandi naar het leven, 6-7, 8-9, 67-69, 73, 131, 141, 154, 248-55. On Ngbandi prestige, see G. Hulstaert, "Nordkongo-Der zentrale Teil," 738; L. A. Almquist, "Symbolic Consensus in Ritual Practice," 74-75. 46. Various groups should be distinguished. Those of Bosobolo in the far north have been strongly influenced by the Ngbaka (A. Wolfe, In the Ngombe Tradition, 135-43; R. Mortier, "Ubangi onder Iinguistisch opzicht," 108-9). The Ngombe-Mbati of the lower Ubangi are acculturated by the Monzombo (R. Mortier, "Ubangi," 108-9). Many of the Bobo (Ngombe of Budjala) were clients of the Ngbandi (A. Verdcourt, "Organisation coutumiere des juridictions indigenes dans Ie territoire de Bomboma," 17; M. Francart, "Note sur les institutions primitives des indigenes du territoire de Budjala," and farther south some were strongly influenced by the peoples of the upper Ngiri (A. Verdcourt, "Organisation coutumiere," 19). The two other blocks of the Likime-Bango and south of the Zaire bend dominated their areas and seem to Notes to Page 117 333 represent best the original institutions on the Likame (A. Wolfe, Ngombe Tradition, 135-43). The "Ngombe" are a telling example of the creation of a "tribe" in colonial times, and show that no ethnographer or historian should trust "ethnic groups." See Mumbanza mwa Bamwele, "Les Ngombe de I'Equateur; Historique d'une identite." 47. The term was libota (A. Wolfe, Ngombe Tradition, 23; N. Rood, "Lidoko et Moweo"; lib6ta, birth, lib6ti, "birthplace" and "family in the wider sense," which stresses bilaterality and territoriality). A. Wolfe, 'The Dynamics of the Ngombe Segmentary System," 169, states that only this term is used, because in the Ngombe mind all the different types of lineages-whether ethnic, major, village, quarter, extended, or restricted family-are only stages of the same entity. Other evidence including the naming of all groups by boso-X, "we of X," which includes descendants and dependents, points to the same conclusion . However, evidence for strict unilineal ideas and practices is meager. Kinship terminology and rules of exogamy were bilateral, as was the distribution of spiritual power over kin; residence with the mother's brother was not rare even in the 1950s; and a cult for the father and the father's father may be patrifilial rather than expressing a concept of patrilinearity. Hence Wolfe's analysis in terms of segmentary lineages does not hold up, especially before c. 1920. 48. N. Rood, "Lidoko et Mowea"; "age group"; A. Wolfe, In the Ngombe Tradition, 35, for the difficulty of seceding from a village. A new village could obtain a chief (kumu) only with the blessing of the chief of its parent villages and the chiefs of other villages in the area. 49. A. Wolfe, In the Ngombe Tradition, 32-34, for the status and role of chiefs and for village government; terms in N. Rood, "Lidoko et Mowea"; elombe, "leader," and mowe, "speaker"; cf. also F. E. Dhanis, "District d'Upoto," 16, 26-27 (elombe); Colle, "Les Gombe de I'Equateur," 155-57 (chief's status and investiture) and 149 (limited polygamy of chiefs). On the whole the "Ngombe" chief was more comparable to his Doko counterpart than to the Ngbandi chief. 50. A. Verdcourt, "Organisation coutumiere," 171: "Their social and political constitution was much more anarchic than that of the Mongwandi and did not allow them to resist the Mongwandi." Nor were any people in the whole area a pushover. The obstinacy of their resistance to aggression is evident in the development of extensive fortifications in the whole area; see M. Eggert, "Archaologische Forschungen," 3238-40, for remnants of northern Mongo ditches and walls. 51. Colle, "Les Gombe de l'Equateur," 141-47, 157-60, 162-69. South of the Zaire the Ngombe sold slaves to traders coming to the main towns on the Lulonga and Lopori. The dominant place of war in the nineteenth century is attested by numerous personal names alluding to circumstances surrounding birth such as "war" and "civil war"; special expressions for "civil war," "trench warfare," "fight to the finish," and "flight from war" were frequent (M. Guilmin, "Proverbes des Ngombe"). Gruesome tales about training with throwing knives (Colle, "Les Gombe de I'Equateur," 144-45) also show how much war permeated Ngombe culture in the nineteenth century. No wonder that the 334 Notes to Pages 118-20 riverine people called them "the warriors" in 1889 (F. E. Dhanis, "District d'Upoto," 26-27). 52. On the mono charm, see E. Canisius, "A Campaign amongst Cannibals ," 120-21. 53. The most detailed data concerning the Bobango are in the works of 1. FraBle. The conceptual underpinning of patrilinearity in the ideas about personality is detailed in his Negerpsyche im Urwald am Lohali, 14-17,52-96,141-44. For the Mbuja, H. Soupart, "Les coutumes budja," is the major source of information. For their manpower; see A. Lejeune-Choquet, Histoire militaire, map 192, text 191-200. In one battle no fewer than 1,800 shields were gathered on the battlefield, which presumes well over 2,000 combatants (192). 54. These include the Bobango, Mbesa, and all "Bangelima" in general. Cf., for instance, M. Fernandez-Fernandez, "Le rite et l'experience," 32-36, 40-41,287, for the Angba; V. Rouvroy, "Historique des Bobango et de quelques tribu voisines," 2. 55. M. Fernandez-Fernandez, "Le rite et I'experience," 42-45, 51-58, for the Angba; 1. FraBle, Negerpsyche, 129, 134. 56. 1. FraBle, Meiner Urwaldneger Denken und Handeln, 51,54; idem, Funf Jahre als Missionar im Herzen Afrikas, 16 ("150 years ago" and Moyimba was chief when Stanley passed through the area in 1877. Moyimba died in 1911); Hemotin, "Historique des Basoo," 3-4, and genealogy annexed; De Bock, "Note concernant la constitution politique des Basoo," 4, dates the earliest chief to 1725 (six generations before Moyimba at 25 years each yields 150 years subtracted from c. 1875, when Moyimba became chief). On succession, see FraBle, Negerpsyche, 134-36. 57. E. Torday, "Der Tofoke," 189-90, erroneously reports that all Eso recognized one supreme chief and that all village headmen were his kin. His remarks apply only to the Kombe group (perhaps 30 villages by 1883): L. Appermans, "Chefferie Kombe," 62-74, esp. 66, 67, 69, and 73. Another chiefdom in the making was that of the Bambelota. A. Bertrand, "Quelques notes," deals with the Yawembe chiefdom. 58. See n. 53 and E. Canisius, Campaign amongst Cannibals, 90, 97-98, 120-21, 403. This is not surprising at a time when the large trading towns on the Zaire River numbered as many as 1O,00G-including perhaps Yambinga, a Mbuja settlement-and at a time when large military forces from the north raided the Itimbiri Valley. 59. Ekopo means "skin" in the languages around Mai Ndombe, and "majestic authority" among the Nkundo. The form occurs without final vowel as ekub or ekop along the lower Kasai and among the Kuba. The emblems were lumps of kaolin everywhere, earth from an anthill in the Mai Ndombe area, a gigantic three-foot double bell, and a whole treasure of other objects kept by muyum among the Kuba. 60. E. Sulzmann, "Die bokopo-Herrschaft der Bolia"; E. W. Muller, "Das Furstentum bei den Sudwest-Mongo"; N. Van Everbroeck, Mbomb'Ipoku: Le Notes to Pages 120-23 335 seigneur al'abfme, 133-60; L. Gilliard, "Les Bolia: Mort et intronisation d'un grand chef"; A. Scohy, "A propos des Nkumu du lac Tumba"; H. D. Brown, 'The Nkumu of the Tumba"; 1. B. Stas, "Le nkumu chez les Ntomba de Bikoro"; Mpase Nselenge Mpeti, L'evolution de La soLidarite traditionelle en milieu rural et urbain du Zaire, 104-12; Cordemans, "Les Pama," 6-41; A. Windels, "Chefferie des Pama-Bakutu," 32-70; 1. Vansina, "Vers une histoire," 47-48 n. 85. 61. E. Sulzmann, "Orale Tradition und Chronologie," 553-58, 561-68. 62. Since their spirit Mbomb'Ipoku was considered to be the most powerful and senior "national" spirit; E. Sulzmann, "Die bokopo-Herrschaft," 404-5. 63. Or mbangala, a large savanna antelope. The place was called Mondombe. Most commentators have identified it with Bondombe on the middle Tshuapa River, but (a) there may be other "Mondombe" and (b) it is not clear that this name really belongs to the traditions. 64. N. Van Everbroeck, Mbomb'Ipoku, 6-50, who summarizes archival sources dating from the 1920s; E. Sulzmann, "Orale Tradition," 555-58, and citation of crucial earlier manuscript sources, 561-63; Vermassen, "La chefferie des Basengere," 1-3. 65. The usual localization of Mondombe at Bondombe on the Tshuapa will not fit. The nearest intercalary savanna lies north of the Zaire near the mouth of the Itimbiri. Savanna antelopes may well have roamed the vast savanna at Ila. 66. As no authors have made the point hitherto, it is worth stressing that these traditions do not refer to mass migration, but to the movements of leaders and their followers who established their rule over or allied themselves to aborigines. 67. Contra 1. Vansina, "Peoples of the Forest," 94, which gave undue prominence to Mpama tradition. 68. E. Sulzmann, "Orale Tradition," 527-28, n. 4, and comment on the map by Vossius (1666) showing the Boma kingdom. Pages 561-63 are a discussion of the chronology of Van Everbroeck based on generations and yielding 1550 for the arrival of the first king in Bolia country. She accepts the notion that Bolia inhabited the area by c. 1300, because Bondongo ceramics are found in this area. With a list of 40 remembered names of kings a deep time-depth can be expected. 69. 1. Vansina, Children of Woot, 245, 123-26. 70. E. W. Muller, "Das Furstentum." 71. N. Van Everbroeck, Mbomp'Ipoku, 149-53; E. Suizmann, "Die Bokopo-Herrschaft," 404-6; L. Gillard, "Les Bolia," 229-35; 1. Vansina, Children of Woot, 49-51. 72. See n. 64; A. Windels, "Chefferie des Pama-Bakutu," 2-5; Cordemans, "Les Pama," 10-11. 336 Notes to Pages 123-29 73. E. W. Miiller, "Das Fiirstentum," (Ekonda); G. Hulstaert, Dictionnaire: nkum (Mongo); E. Sulzmann, "Orale Tradition," 554 (Bobangi), 557 (Ekonda, Iyembe). 74. Bokapa ekopo occurs in traditions from the Mpama to the Kuba; Bolongo Mpo from the Ngwi kingdoms at least to the Kuba. For titles, see, e.g., N. Van Everbroeck, Mbomp'lpoku, 27 (iyeli) , 32 (etati), 1. Vansina, Geschiedenis van de Kuba, 161-62. lyeli was the title of the Ngong6 nkum of Idanga. 75. E. Sulzmann, "Orale Tradition," 578 n. 51 (Bosanga which was also the name of their first "capital" in the area); 1. Vansina, Children of Woot, 41-63; 113-15,123-25 (Kuba). 76. E. Sulzmann, "Orale Tradition," 574-75 n. 26; 1. Cornet "A propos des statues ndengese"; G. Van der Kerken, L'ethnie Mongo 1:341-44, 2:658-74. On Bolongo Itoko, see note 29 and 1. Vansina, Geschiedenis, 169. 77. S. et 1.Comhaire-Sylvain, "Les populations de Mai Ndombe," 24-26; G. Van der Kerken L'ethnie Mongo, 653-54; 1. Vansina, Geschiedenis, 48 nn. 86-87; Ndaywell, "Organisation sociale et histoire," 158-92 (Ngwi political organization ), 325-50 (Ngwi history), 375 (Ngwi chronology). The estimated date for their crossing of the Kasai is 1650-1742. This chronology rests on reign lengths and an uncertain Kuba tie-in seems particularly precarious. 78. G. Brausch, "La societe N'Kutschu," 50-59; L. De Heusch, "Elements de potlatch chez les Hamba," 337-48 (nkumi okanda); Dimandja Luhaka, "Le pays de Katako-Kombe al'epoque coloniale," 25-28. 79. L. De Heusch, "Un systeme de parente insolite," 1012-27 (Jonga in upper Tshuapa); A. Moeller, Les grandes lignes des migrations, 540-41, 567 index: nkumi indicates a possible diffusion among the Mituku, as does the Lega form ngumi. See also chapter 6. 80. 1. Vansina, "Vers une histoire," 26-27, and nn. 88, 94; G. Hulstaert, "Over de volkstammen van de Lomela," 21, 29, 33, 39, 42; G. Van der Kerken, L'Ethnie Mongo, index: ekofo. For the complex case, see A. Bengala, "Le noble des Booli," 105-11. The title was kokakoka, an intensive form of "elder." 81. These internal histories have so far been worked out for only some societies. See Ndaywell, "Organisation" (Ngwi), and R. Tonnoir, Giribuma, but also see the critique by E. Sulzmann "Orale Tradition," and 1. Vansina, Children of Woot. Much additional research is sorely needed. For instance, the processes by which the title system was developed into a system of territorial control by the societies around Lake Mai Ndombe and the process by which a system of double descent for succession was developed among the Ntomba and Sengele still need to be studied. Even the detailed organization of the nkumu system among the populations of the Lukenie and of the ekopo south of the Tshuapa is still unknown. Chapter Five: Between Ocean and Rivers 1. Thus early settlement occurred both by sea and over land. See map 2.8. The spread of iron smelting seems to have occurred on the plateau first and then Notes to Page 130 337 to have reached the coast later via the Ogooue Valley; the distribution of a variety of features indicates communication over the plateaus as well as along the rivers. Cf., for instance, the diffusion of throwing knives diagonally from the great Ubangi bend to the Gabon Estuary and to Loango. 2. See map 2.10. 3. K. Born, "Nordkongo--Der Westen," 712-13; for connections with adjacent West Africa, see P. A. Talbot, The Peoples of Southern Nigeria, vols. 2-3. (e.g., 3:754-801, "Societies and Clubs"); w. Hirschberg, "Das CrossfluBgebiet und das Kameruner Grasland," 2:355-72. 4. As an introduction to associations, see, for Gabon, A. RapondaWalker and R. Sillans, Rites et croyances, 171-293. The earliest reference is A. Battell, The Strange Adventures of Andrew Battel!, 56-57, 82-83; the tattoo he describes on p. 56 corresponds to a mwiri tattoo in the same area mentioned by Raponda-Walker and Sillans, p. 230. For northern Gabon and southern Cameroon , see R. H. Nassau, Fetichism in West Africa, 138-55. For Cameroon, see H. Nicod, La Vie mysterieuse de l'Afrique noire, 91-122; C. Dikoume, "Etude concrete d'une societe traditionnelle: Les Elog-Mpoo," 113-32; 1. Rusillon, "Le role social d'une societe secrete d'Afrique equatoriale"; 1. Perono, "Les Basa," 103-5.; P. Wurm, Die Religion der Kiistenstamme in Kamerun, 12-53; P. Valentin, Jujus of the Forest Areas of West Cameroon. For an in-depth study (among the Kosi), see H. Balz, Where the Faith Has to Live, 184-318. 5. *-kuda. For Bubi, see n. 48. In Abo country the adolescent age group was politically represented by a leader at village level: F. Autenrieth, Ins InnerHochland von Kamerun. For Elog Mpoo, see C. Dikoume, "Les Elog-Mpoo," 133-34. Only coastal age groups under Duala or Malimba influence formed an association. For Duala, see M. Bekombo-Priso, "Les classes d'age chez les Dwala," 289-307. 6. H. Balz, Where the Faith Has to Live, 37-43, discusses the differences between the Kosi-Mbo group and the Kpe group, where according to him villages were much less compact, although they were still fenced in. 7. 1. Dugast, "Banen, Bafia, Balom," 141-42,160,166-67 (the reference to "village" is to a French colonial administrative unit); 1. Champaud, Mom: terroir Bassa, 15. Still, L. M.Pouka, "Les Bassa du Cameroun," 158, states that the size of a village varied according to the size of the families settled there; E. Von Skopp, "Sitten und Gebrauche der Bakoko in Kamerun," 485, who gives a size from 3 to over 50 houses, substantiates this; see also G. Zenker, "Die Mabea," 2-3. However, the neighboring Ngumba lived in villages; cf. L. Conradt , "Die Ngumba in Siidkamerun," 333. 8. The A 70 speakers form a coherent group called "Pahouin" or even "Fang" in the literature. But P. Laburthe Tolra, Les seigneurs de laforet, 18-19, rightly denounces the terminology because it unduly generalized from one observation in one ethnic group to all the others. To avoid a dubious ethnic name as well as the clumsy A 70 designation, derived from the partly outmoded classification by M. Guthrie, the group is labeled here as the Sanaga-Ntem group. 338 Notes to Pages 130-32 9. Their languages are closely related to a number of others, such as Ngumba and Mabea in the west, So in enclaves along the Nyong, and a group of peoples in southeastern Cameroon. Guthrie called them A 80. Here they are referred to as the southern Cameroon group. Data on some Njem groups are particularly weak. Still, the reports in Deutsches Kolonialblatt of the German expeditions in the area allow one to conclude prudently that they were on the whole quite similar to the Djue, Makaa, and Njem from Congo, about whom monographs have been written. 10. For the meaning of this form, see C. Gregoire, "Le champ semantique du theme bantou *-banja." She still misses some reflexes, among them the type baa, abeny, banja, banjo in the languages of Cameroon, which there designate the communal buildings for men. South of the Ogooue the reflex mbanza designates a village temple, and farther south and east the reflexes mbanza (Kongo and Mboshi) and mbee (Teke peoples) refer to "capital." 11. For southeastern Cameroon, see C. Koch, "Die Stiimme des Bezirks Molundu," 286-304. High averages of inhabitants per village obtained there in 1910-1912 for the Kaka (325), Bokari (194), Bumbom (129.6), and Gunabemb (118.82). For the Bekwil villages, see Didier, "Le nomadisme des Sangassangas ," 592 (150-200 people, but up to 1,000 people), and E. Zorzi, Al Congo con Brazza, 543 (the Bekwil village was the biggest village so far seen in Africa); their Bokiba (Kota) and Mboko neighbors to the south also had sizable villages, ibid., 525-27, 530, 573. 12. Reflexes occur in the Sanaga-Ntem, the southern Cameroon, the Kaka group of languages, and Sekyani (subgroup by itself), all class 5/6. Note that, within the southern Cameroon group, both to the west (Ngumba and Djieli) and to the east (in the Njem group of languages) of the present extent of the SanagaNtem group, this form is preceded by *kwa- meaning "at the place of So-andSo ." This particular innovation presumes that the toponyms were derived from the name of leaders. It is itself older than the expansion of the Sanaga-Ntem group, hence pre-fifteenth century. Among the Sanaga-Ntem speakers and part of the Kaka group a second form refers to villages as *boga. 13. CS 665 *-dongo: e.g., -long (Saa), bo-nong (Nen), ke-Iong (Kpa?), kolong (Kpa?), joong (Yambasa), ayong (Sanaga-Ntem group), and others. Cf. A1. 14. *-dom-. All have Omaha cousin terminology. The precise system of terminology in Makaa remains unknown, however, and present knowledge of the terminology in the Njem set of languages remains rudimentary. Furthermore languages within the Saa group share a set of terms relating to patrilineal groups, and languages within the Sanaga-Ntem group another set. At least two other terms are shared by the Nen north of the Sanaga and the western languages of the Njem set. 15. The etymology of the innovations ndog, "patrilineage," and log, "patrilineage," in Nen and in the Saa group remains unclear. They are not related (tonal opposition). Log is not yet accepted by all Nen groups and may well be younger than ndog. Notes to Pages 132-34 339 16. As seen on the maps in E. V. Thevoz, "Kamerun Eisenbahn," map 3 (present densities); R. Letouzey, "Vegetation," 22-23 (savannas wooded or not, semideciduous forest recolonizing savannas abandoned in the late 1800s, and islands of semideciduous forests). 17. The innovation may well have diffused from the neighboring populations of the grasslands to the speakers of Nen and the Basaa group or to the Sanaga-Ntem group rather than the reverse, but conclusive evidence is not yet available. 18. 1. M. W. Wognou, Les Basaa du Cameroun, 8-12, passim (patrilineal structures); S. B. Mandeng, "Traditional Healing of a 'Non-Ordinary' Disease (Kon) among the Basaa of Cameroun," 74-77 (Ngok Lituba), 77-80, 85-101 (social structure); C. Dikoume, "Les Elog-Mpoo," 20-23 (Ngok Lituba), 83-88 (terminology). Saa genealogies do not prove migration; they only claim it. Moreover they suffer from both telescoping and eponymic additions, as is shown by the absence of branching in most generations, and cannot be used to date a migration. The Saa occupied the site of Douala before the Duala settled there, in c. 1660 at the latest. Cf. E. Ardener, "Documentary and Linguistic Evidence for the Rise of the Trading Polities between Rio del Rey and Cameroons: 15001650 ," 81-116 (Van Leers and first evidence for the Duala language); idem, Coastal Bantu of the Cameroons, 17 (the toponym Ambos, known by c. 1507, could reflect the name Abo, a Saa related group) and 19 (six generations in the Duala ruling lineage before 1792). 19. R. Letouzey, Etude phytogeographique, 141-45 (Lophira alata); P. Foury, "Indications," 7-13. 20. S. B. Mandeng, "Tradional Healing," 87 (the naming of children by father is more rigid than among the Bamileke), and 1. M. W. Wognou, Les Basaa, 40 (the kinship system was 8 generations deep, 10 with a "founder" and "a closure of patrilineal kinship"). On Nge associations, see n. 4, and 1. M. W. Wognou, Les Basaa, 5,6; C. Tastevin, "Societe secrete du Ge chez les Ba Koko du Cameroun," 891-901; idem, "Societe secrete feminine chez les Ba Koko," 901-6. On mbog and related concepts, see S. Epea, "Message chretien et visage de l'homme chez les Basa, Bantie du sud Cameroun," 152-63; 163-79 (mbombog). On segmentary terminology, 1. M. W. Wognou, Les Basaa, 24, gives 15 terms of which 9 might refer to segments of different size. No other Basaa author does so. P. Lemb and F. de Gastines, Dictionnaire Basaa-Fram,;ais, do not corroborate his claims either. Each of the terms cited means either a social group and is applied from family to ethnic group or has other meanings ("group," "company," "tomb," "camp," "patrimony," "father's domain"). Dju, "monogamous family," mbai, "courtyard (group)," and sas, "phratry," are not found in the dictionary. On balance it seems clear that there were neither clearly delineated segments nor a specific terminology to designate them. 21. M. Guthrie, The Bantu Languages of Western Equatorial Africa, 40-44, is the only general overview extant. Of the 10 languages recognized, 7 were spoken north of the Nyong, 5 of them around Nanga Eboko, 2 north of the Sanaga. The remnant speakers north of the Sanaga and west of Nanga Eboko 340 Notes to Pages 135-36 are all F6k and have always lived there. See 1. C. Barbier, Mimboo, reine d'Asem, 15-17 and map. 22. P. Alexandre and 1. Binet, Le groupe dit Pahouin, 40-49,62-69, which is confirmed by monographs for all the (seven) ethnic groups studied. 23. P. Alexandre, "Proto-histoire du groupe Beti-Bulu-Fang," remains the best known of these; for a partial critique, see C. Chamberlin, "The Migration of the Fang into Central Gabon during the Nineteenth Century." 24. See n. 2l. For the southern Cameroon languages, see M. Guthrie, Bantu Languages, 45-49 (the "northern Gabonese languages" are the group labeled B 20); ibid., 59-63; A. Jaquot, Les classes nominales dans des langues bantoues des groupes BIO, B20, B30, 69-75,295-97 (lexicostatistics show that Sekyani forms a group by itself). Speakers of these languages probably occupied most of northern Gabon before the first Fang expansion. They were split only by the second Fang migration, which began c. 1840. 25. Density figures in P. Alexandre and 1. Binet, Groupe dit Pahouin, 10-11; 1. C. Pauvert and 1. L. Lancrey-Javal, "Le groupement d'Evodowa (Cameroun)," 11-12; G. Sautter, De l'Atlantique, endmaps. 26. A. Panyella, "Esquema de etnologia de los Fang Ntumu de la Guinea Equatorial desde el punto de vista etnoI6gico," 15-17, map 1. Half of northeastern Rio Munf was no longer covered by primary forest in the 1950s. He erroneously concludes from this that the Ntumu were not adapted to life in the rainforests! 27. 1. L. Wilson, Western Africa, 302 (scouts); F. Autenrieth, Ins Inner Hochland von Kamerun, 126-27,138 (pioneer villages during expansion). 28. H. Bucher, "The Mpongwe of the Gabon Estuary," 128-29 (Mpongwe of the late 1700s); P. Alexandre and 1. Binet, Groupe dit Pahouin, 14-15; P. Alexandre, "Proto-histoire," 532-35 (Fang after 1840s). Even in the rapid migrations after 1840 the Fang expanded continuously by drift of individual settlements over short distances rather than suddenly by large groups over long distances. 29. 1. Vansina, "Peoples of the Forest," 80-81, for numerical evidence from Gabon and Cameroon. In these cases the movements had a purpose: to reach trading stations on the coast. That may have speeded up their march. The Sanaga-Ntem expansion may have been somewhat slower for lack of a specific objective. But the iron-producing district south of the Ntem may well have been known to the Ntumu. In addition the calculations assume a linear advance, not sideways movements, yet an eastern and the western movements occurred over a wide front. The total advance may have been slower than the 20-30 km attested for the 1800s. 30. On absorption, see P. Laburthe-Tolra, Seigneurs de la foret, 101-2, 120-22 (most later Fang may well be "absorbed" populations; Ngovayang later was a stable migratory unit combining Ngumba and Fang). On large villages see n. 11. 31. Glottochronological dating yields: 78 percent or A.D. 1160 for FangEton split and the total depth of all known Sanaga-Ntem languages; 83 percent or A.D. 1360 for Ntumu-Fang and Ewondo-Fang splits; 85 percent or A.D. 1440 Notes to Pages 137-39 341 for Ntumu-Ewondo and Ewondo-Eton splits; 87 percent or A.D. 1510 for BuluEwondo split; and 90 percent or A.D. 1620 for Fang-Bulu split. P. LaburtheTolra 's maps (seventeenth century), 108-9, in Seigneurs de laforet, are too late because his calculations were based on genealogies which suffered from telescoping . See B. Ciist, "Travaux archeologiques," 12 (sites A.D. 1590 +1-60, A.D. 1600 +1-70, and A.D. 1680 +1-60). Future ceramic studies may well show whether these were sites of new immigrants. 32. For distribution of crossbows (always named CS45 *bano ("arrowshaft"] in eastern Bantu), see 1. Vansina, "Deep Down Time," map "bow." For illustrations , see, e.g., G. TeBmann, Die Pangwe, 1:141; H. Koch, Magie et chasse, 127. For xylophone and harps, see, e.g., G. TeBmann, Die Pangwe, 2:322,330. 33. The Ntumu even dropped the use of the term mvog, "lineage," for etunga used by their western neighbors. One should not forget that there was also resistance to change: e.g., the Sanaga-Ntem concept of evuslevur witch substance was not borrowed by autochthons. And some original features were abandoned by the emigrants, e.g., divination through the use of a spider common in the Saa group and among the Ewondo but not practiced by the Ntumu (G. TeBmann, Die Pangwe, 2:196). See A127. 34. P. Alexandre, "Proto-histoire," 535-37, cites some of these. For a recent example, see M. Ropivia, "Les Fangs dans les Grands Lacs et la vallee du Nil"; idem, "Migration Bantu et Tradition orale des Fang." 35. M. de Teran, Sintesis geograjica de Fernando Poo, 11,58, 39, and in general for the natural background. The earliest full-fledged ethnographies about the Bubi are O. Baumann, Eine Afrikanische Tropen-Insel: Fernando Poo und die Bube (data from 1886-1887), A. Aymemi, Los Bubis en Fernando Poo (data from 1894-1939), and G. TeBmann, Die Bubi auf Fernando Poo (data 1915-1916). Among these Aymemi is by far the most reliable. In contrast TeBmann 's data on social life are often unsound. Bubi institutions are often misrepresented in summary ethnographies. Contrary to the typical description, the Bubi were not organized in segmentary patrilineages, did not practice matrilineal succession, had no secret societies, no formal age-grading system, and did not live just in compact villages. 36. On unfavorable winds and currents, cf. P. de Marees, Beschryvinghe ende historische verhael van het Gout Koninckrijck van Gunea, 242; T. Boteler, Narrative of a Voyage of Discovery to Africa and Arabia, 412; M. de Teran, Sintesis geograjica, 9-26. On early Bubi-European relations, see C. Crespo GilDelgado , Notas para un estudio antropologico y etnologico del Bubi de Fernando Poo, 17,170-72. 37. 1. M. Rurangwa, "Nota sobre la investigaci6n linguistica del Bubi," 141, argues that it is a single language. But T. R. H. Thomson, "The Bubis, or Edeeyah of Fernando Po," 106, claimed that the extreme dialects were mutually unintelligible in 1841. Bubi may have been a language cluster rather than a single language. 38. L. Silveira, Descripcion de la Isla de Fernando Poo en visperas del tratado de San Idelfonso, 29,30,36, for the situation in 1772. The documentation from 1819 onward is rich enough to follow the breakdown of isolation step 342 Notes to Pages 139-40 by step. On trade and imports by 1846, see 1. Clarke, Introduction to the Fernandian Tongue, v, and 1. M. Usera y Alarcon, Memoria de la Isla de Fernando Poo, 32-33. By 1857 T. 1. Hutchinson, Impressions of Western Africa, 192, reported on the last-known hoard of stone axes on the island. 39. A. Martin del Molino, "Secuencia Cultural en el Neolftico de Fernando Poo"; idem, Etapas de la cultura Carboneras de Fernando Poo en el primer milenio de nuestra Era; P. De Maret, "The Neolithic Problem in the West and South," 60-62; B. Ciist, "1985 Fieldwork in Gabon," 7 (sites of similar ceramics: 930 B.P. on Bioko, 1150 +/-60 B.P. at Libreville). 40. T. 1. Hutchinson, Impressions, 201, already reports Bubi origins on the Santa Isabel peak. Subsequent mountain climbers also report this, as does O. Baumann, Eine Afrikanische Tropen-Insel, 74. Traditions about the great migration are reported in some detail by A. Aymemi, Los Bubis, 13-22, and briefly by G. TeBmann, Die Bubi, 11-12. All authors and later descriptions cite various emblems, palladia, and the sacred fires. 41. Speculation as to a place of origin in Gabon is old (0. Baumann, Eine Afrikanischen Tropen-Insel, 73; G. TeBmann, Die Bubi, 12). The ceramic link found by Ciist (n. 39) is still too tenuous as reported to build on. Spanish authors rather favored the coasts of Rio Muni or even the southern Cameroons. Further archaeological work and more systematic study of loaning between the mainland languages and southern Bubi speech will one day settle this issue. The date for the earliest arrivals could, just possibly, correspond to the eleventh century, when inland settlement began in earnest. All the authors mentioned that if the migration had occurred after 1500 the earliest arrivals would have been reported by Europeans, especially on Sao Tome. But this argument from silence is suspect, if only because of the isolation of the island. Indeed a trickle of seventeenth- and eighteenth-century fugitives from the Portuguese islands to the south of Bioko was not reported before c. 1780 (0. Baumann , Eine Afrikanische Tropen-Insel, 74). 42. The reconstruction of institutions before the Great Migration is based on linguistic and ethnographic data. The reasoning used is the following: Innovations in vocabulary, common to all dialects but not shared by mainland languages , indicate presence in this early period. So does continuity with basic western Bantu terminology. The latter, however, is not foolproof. Some western Bantu terms died out and were later reimported from the coast. For instance, boribo ("spirit"), a synonym for mmo, is derived from CS 619. Reflexes of this root occur in all the mainland languages opposite the island. The fact that it is a duplicate for mmo alerts one to the likelihood that the term had been replaced by mmo in the deep past and was later reintroduced, presumably during the Great Migration. Second, major original Bubi institutions-such as the opposition between married and unmarried men and the patterns of settlement-that are unknown anywhere on the mainland, show an original vocabulary in form, and are not linked to later institutional developments resulting from the Great Migration are presumed to antedate it. The data come from A. Aymemi, were compared with Notes to Pages 142-44 343 the information of G. TeBmann, and have been checked against earlier information from the 1840s and 1850s. In this summary it is not possible to cite the detailed evidence on each point. 43. The relevant traditions justified the position of the aristocracy and provided a social charter documenting their relative rights by conquest. The supposed order of arrival closely reflected their relative hierarchy in the 1880s. 44. For Tete and Bokoko, see A. Aymemi, Los Bubis, 14-18. On iron, note that Bubi bojele, "iron," derives from CS 800 *-geda (and see Guthrie 1:138, topogram 25). Reflexes from this root occur among all the languages on the mainland from Mt. Cameroon to Rio Muni. The existence of the term reminds one that, although the Bubi did not use, work, or smelt iron, they knew of its existence. Whether the term was anterior to the arrival of the main body of immigrants or was introduced by them, it is a nice reminder that the immigrants arrived with iron weapons and tools. 45. The reconstruction of institutions in this period is based on the reasoning outlined in n. 42. Linguistic innovations, common on the mainland and present either in all Bubi dialects or in the main southern dialects only, have been used as evidence. Duplication in form sometimes suggests borrowing. Thus the terms bese, "central village," and nse, "district," share the same root, but in different classes. Nse classes 9/10 was the usual term on the whole coast and seems to have been imported to designate the new districts. Duplication of meaning (synonyms) may also indicate borrowing (see n. 47). 46. A. Aymemi, Los Bubis, 53-70 (social stratification and "government "), 84-85 (caste). 47. The term botuku, "lord," a true synonym of the earlier etakio, "ruler," both opposed to bataki, "dependent" or "servant" (Ayememi, Los Bubis, 53), is more frequent than the other term. One suspects it to be a loan in Bubi. It may well have been the designation for "leader"/"chief" among the immigrants. But the suspicion remains unconfirmed so far, because the term has not yet been found in any of the mainland languages. 48. A. Aymemi, Los Bubis, 37-50, 167, 173-74 (on marriage); ibid., 68, and C. Crespo Gil-Delgado, Notas, 155 (on the attempts of chiefs to control all virgin girls). 49. A. Aymemi, Los Bubis, 124-25 (priests of God), 57-63, 86-90, 14248 , 152-56 (major collective rituals), 165-66 (royal high priests). 50. O. Baumann, Eine Afrikanischen Tropen-Insel, 105-6; A. Aymemi, Los Bubis, 32, 115-22; C. Crespo Gil-Delgado, Notas, 156, 180. 51. Generally authors estimate his succession as c. 1850, which may be too late. The date suggested here rests on the following considerations. Moka died in 1899 reputedly 105 years old. He became first chief of Riabba as the senior of his generation, and only later achieved kingship. Given the existing practice of succession he was probably 30 or more when he became chief. Yet it seems unlikely that a person over 50 would succeed in unifying the island. Hence the 344 Notes to Pages 145-46 opinion that he unified the kingdom in his 40s, some five years after he became chief. 52. On historical laws, C. Crespo Gil-Delgado, Notas, 156, speaks of "the regular cycle common to the history of all peoples" which led to the rise of confederations and later of a kingdom. On p. 180 he says that this is "the normal development of the historical phenomenon [of centralization]." On previous kings, ct. A. Martin del Molino, "La familia real," 37-40 (a previous dynasty at Ureka, but O. Baumann, Eine Afrikanischen Tropen-Insel, 32-33, gives an alternative explanation); A. Aymemi, Los Bubis, 162 (the case of the Lombe Lagoon); O. Baumann, Eine Afrikanischen Tropen-Insel, 113 (a king's role in the ouster of the Spanish at Concepci6n Bay. But this event was followed by a Duala slave raid there against which the Bubi seemed helpless). On wealthy chiefs near Clarence measured by the size of their harems, see T. R. H. Thomson, Bubis, 112 (the richest had 200 wives and many guns; Moka had only some 60 wives, late in his life). On guns in the north, see C. Crespo Gil-Delgado, Notes, 174. 53. For the main events and publications, see A. de Unzueta y Yuste, Historia geografica de la isla de Fernando Poo. 54. C. Crespo Gil-Delgado, Notas, 17,170. 55. 1. Vansina, "Deep Down Time" map 3: *-pika. The map shows the extent of the meaning "slave," most of it stemming from Kongo as a result of the slave trade. Bobangi's skewed form is a good example. But some terms cannot be explained in this way. Thus the extension into Kasai (Luba group, Songye, and Kusu) and Shaba is not derived from Kongo. Neither are bofika, "slave woman," in Mongo and mpika in eastern Mongo (Mompono), nor are many Gabonese reflexes such as nyeka (Kota), moyeka (Ndaza, Shake), oyiha (Mbede, Ndumu), oheka (Mbosi, Koyo)-all designating "slave." The distribution suggests a term common to the whole southwestern block and the savanna languages derived from it, with a meaning such as "servant," "dependent." See 1. Thornton, The Kingdom of Kongo: 21, P. P. Rey, Colonialisme, neocolonialisme et transition au capitalisme, 86-87. As a result of the slave trade the meaning changed in Kongo and diffused from there. 56. 1. Vansina, "Antecedents des royaumes bateke (tio) et kongo," n. 12. The earliest reference to them is A. Battel, Strange Adventures, 59 (they already "pay tribute"). On dependency in the nineteenth century, see P. P. Rey, Colonialisme , 96-101. 57. See Appendix. Also see, e.g., 1. Vansina, The Tio Kingdom, 47. 58. The Portuguese word palavra, "word," translates the ancestral western Bantu expression in African languages, CS 771 *-gambo (CS 1919 is the same stem; ct. E. Meeussen, Bantu Lexical Constructions, 17): e.g., Kongo mambu, Lingala likambo. It refers to any "affair," court case, legislation, or discussion. See, e.g., 1. Vansina, Children of Woot, 19,252 n. 2, 305 n. 22. 59. W. MacGaffey, "The Religious Commissions." 60. On nkinda, see 1. Vansina, Children ofWoot, n. 15; the precise meaning "village charm" dates to the common southwestern language. Notes to Pages 147-51 345 61. Based on cereals in the savannas and the AAB banana in the forests. Unfortunately no dated remains in the quadrant between A.D. 350 and 1300 have been found: B. Ciist, "Pour une archeologie du royaume kongo." Still the distribution of the term njale, nzadi, covering the whole of this area is of interest here because it refers to the major routes of communication: H. A. Johnston, A Comparative Study of the Bantu and Semi-Bantu Languages, vol. 1. However, the area covered excludes the Ogooue and its affiuents. Its distribution crosscuts the genetic divisions and forms a single block, so it spread partly by borrowing , well before the age of the slave trade (here c. 1525-1600), because it extends far beyond the area of this trade. The date and place of its appearance have not yet been ascertained. 62. This scenario is based only on analogy provided by later ethnographic descriptions. 63. See its distribution in 1. Vansina, "Antecedents," n. 24, and idem, Tio Kingdom, 585. 64. 1. Vansina, "Antecedents" nn. 18,26, for the etymologies of "capital," *-pata (CS 1455). 65. The term mwene (CS 1971) is derived from an early Bantu form *-enye, "one-self" (CS 1970; E. Meeussen, Bantu Lexical Constructions, 13). Cf. also 1. Vansina, "Antecedents," n. 27. 66. In all Kongo and Teke languages mpu designates both "hat" and "authority ." The term *-badi (CS 28) with the meaning "court" is found in Tio, Mfinu, Boma, Kongo, and Mbosi. The etymology of the terms cannot be fixed with any certainty, and hence it is still unknown whether they are Teke or Kongo in origin. For nlunga, see 1. Vansina, "Antecedents," n. 28; the origin of this term is Kongo. In Kongo only: *-yad-, "to rule" (CS 1890), derives from CS 873 and 874, with the first meaning "leopard skin" (CS 1102) rather than "carpet." Kongo and Teke had terms of different origin for "to succeed": ibid., n. 29. The Teke term was shared by Bembe (northeastern Kongo), by the Mongo languages , and by those of the lower Kasai. 67. Ngtiti- also occurs in Kongo, especially in northeastern Kongo, which seems to have borrowed it from the Teke group. On titles, see P. P. Rey, Colonialisme , 253-60, comparing the Tio and Loango kingdoms. Known Kongo and Loango titles date from the sixteenth and seventeenth centuries, known Tio titles, with one exception, only from the nineteenth century. 68. 1. Daeleman, "A Comparison of Some Zone B Languages in Bantu"; map 5.5 is based on the extant vocabularies. See also 1. Vansina, "Antecedents," n. 32. The phenomenon had already struck Europeans in the sixteenth century; F. Pigafetta and D. Lopes, Description du Royaume de Congo et des Contrees environnantes, 35. 69. J. Vansina, "Antecedents," n. 33. 70. 1. Vansina, "The Bells of Kings." The linguistic terminology *-kungu, for the twinned bells, and *-pam-, for the single bell, is found in the Teke group, in Mbede-Obamba, and in the lower Kasai area (Yans, Mpur, Nswo, Ding, and Mbala). Only Boma, Saa, and Burna in the area of the linguistic phenomenon 346 Notes to Pages 151-54 have another form for twinned bell, *-Ionja, which is of Mongo origin and a later intrusion here, probably due to the influence of the nkumu courts. 71. The later-known principalities (precursors of the kingdoms of Loango, Kongo, and neighbors) lay in Mayombe or on the adjacent coast with two exceptions only. This was also the area with the richest ecological diversity (forests, coastal and mountain savannas, coasts, the Zaire Estuary, and a diversity of mineral deposits). 72. For the chronology, see 1. Vansina, "The Bells of Kings." The technology of these bells diffused from West Africa and reached known archaeological sites in Shaba and Zambia before 1400. At a diffusion speed of 11.75 km per year (which is quite fast) they were in the first Teke principalities by 1221. The speed of the diffusion is probably too high, and the archaeological dates are probably too late, because small single bells already appear in Shaba between A.D. 1100 and 1300. Hence the conclusion that they were in the Teke principalities before 1200 (probably in the eleventh or twelfth century). 73. 1. Vansina, Tio Kingdom, 313-23, for the lower level squires c. 1880. The term antsaana, "plebeians," is also found in the lower Kasai area as well as among the Saa (banshian). In Burna the term means "orphan," clearly an older meaning. 74. CS 1175 *-kued, "to marry," has Kongo and Teke reflexes with several Kongo derivations (nkwezi, "wife's kin of her generation"; kinkwezi, "affine"; ko, "parent-in-law"). For the connection with bridewealth, see the discussion in chapter 3: p. 77 and n. 20. 75. 1. Vansina, "Antecedents," n. 44 (Crow) and 45 (ngwa nkhazi). 76. Lineages make sense in terms of succession and/or inheritance: to acquire and defend property or position in society. They tend to appear as common , centralizing, overarching institutions appear and to grow as centralization grows, at least in social formations with "representative government," which was the Kongo as well as, for instance, the Yoruba case. Cf. P. C. Lloyd, "The Political Structure of African Kingdoms," 84, 98-102; idem, Political Development of Yoruba Kingdoms, pp. 1-9. It is not surprising, then, that Yoruba lineages are found only in centralized kingdoms and that undifferentiated groups occur where kingdoms do not exist. Whether or not lineage structures later break down depends on the further transformations occurring in the centralized body politic. 77. 1. Vansina, "Antecedents," n. 46; n. 49 for the term *-bida (luvila in Kongo) of doubtful status as a designation for "social group" before c. 1900. For these terms, see also A2. 78. F. Hagenbucher Sacripanti, Les fondements spirituels du pouvoir au royaume de Loango, 22-28, 62; P. P. Rey, Colonialisme, 251, 260. A. Hilton, The Kingdom of Kongo, 22-23, 29, tentatively proposes a similar scenario for Kongo. 79. 1. Miller, Kings and Kinsmen: Early Mbundu States in Angola, 43-54, 59. Mother's brothers were called lemba after the mulemba fig tree, which stood Notes to Pages 155-58 347 for the power of the heroes of the lineage. For Mai Ndombe, see Mpase Nselenge Mpeti, L'evoLution de La solidarite, 89-95. 80. J. Vansina, "Antecedents," 51. Innovation included the adoption of throwing knives as well as body armor and battle axes, and the differentiation of spears into throwing and stabbing spears. 81. Nkoli (CS 1110) is western Bantu, not common Bantu as Guthrie has it. See also A23. The meaning "pawn" predominates over "captive," given by Guthrie. Maybe the widespread Mongo ethnic nickname nkole, which has a pejorative meaning, is another reflex? Cf. G. Van der Kerken, L'ethnie Mongo 2:710-11, 715-16, but G. Hulstaert does not have the term in his dictionary. 82. For terms relating to markets and the week, see 1. Vansina, "Probing the Past of the Lower Kwilu," 349-53. The origin of most terms lies north of the lower Zaire and to the immediate south of the copper belt. 83. B. Clist, "Archeologie du royaume Kongo," map 3. 84. 1. Vansina, "Antecedents," nn. 55 and 66. 85. For a list of the later provinces of Loango, F. Hagenbucher Sacripanti, Fondements spirituels, 64-70 (Chilongo, Mampili, Loandjili, NgaKanu, Mayombe, Makangu, and Makunyi); P. P. Rey, Colonialisme, 265-68; and P. Martin, The External Trade of the Loango Coast, 10-19 (Loangiri, Loangomongo , Pili, and Chilongo were the inner provinces). One must add Mayumba to the north and the states ofNgoi, Kakongo, and Bungu farther south. A number of other principalities which later became five of the provinces of Kongo bordered on this core to the south and east. For these, see F. Pigafetta and D. Lopes, Description, 68-69. 86. F. Bontinck, ed., Histoire du royaume du Congo (c. 1624), 60,84-87; F. Pigafetta and D. Lopes, Description, 69,85-89. 87. For Loango: F. Hagenbucher Sacripanti, Fondements spirituels, 22-23; P. P. Rey, Colonialisme, 251, 264-65; P. Martin, External Trade, 5-9; o. Dapper , Naukeurige Beschrijvinge, 143-44, for an origin in Zarri in Kakongo, not far from Bungu. For the Tio state: 1. Vansina, Tio Kingdom, 440 and map; F. Pigafetta and D. Lopes, Description, 168-70 n. 87 and 200 n. 275, for the old term Anziko and the surrounding lands, Anzicana; Ebiatsa-Hopiel-Obiele "Les Teke," 34-47, for an etymology of Anziko as asi + Nkoo, "those of the Nkoo district." It is most probable that the name must be read asi + 6koo, "those of the king" (i.e., "the kingdom"). The first mention of Teke or rather mundiquetes in F. Pigafetta and D. Lopes, Description, 32,65,66,84. 88. For tentative dates for Loango and Kongo, see P. Martin, External Trade, 9. For a review of opinions about Kongo, see 1. Vansina, "Antecedents," n. 10. Dates vary from indeterminate but quite old, to 1300, to the latest proposed date "in the early fifteenth century," A. Hilton, Kingdom of Kongo, 31, which is barely tenable; O. Dapper, Naukeurige Beschrijvinge, 219, for the claim that the Tio kingdom was the oldest. 89. This broad comparison is based for the most part on the cited works of P. Martin, A. Hilton, 1. K. Thornton, F. Hagenbucher Sacripanti, P. P. Rey, and 348 Notes to Pages 158-59 1. Vansina (Tio Kingdom), as well as on W. G. L. Randles, L'ancien royaume du Congo des origines ii la fin du XIXe siecle; Y.-N. Gambeg, "Pouvoir politique et societe en pays teke"; and F. Ewani, "Recul et stabilisation teke." 90. On associations, see n. 4. 91. See n. 10. The most famous of these are the Tsogo temples; cf. illustration in A. Raponda-Walker, and R. Sillans, Rites et croyances, 196, showing a construction similar to the public building for men elsewhere. 92. For the distribution of this bell, see map 5.7; the earliest references to it date from the Gabon Estuary in 1593 (1. H. Van Linschoten, Itinerario, 3: 11-12) and Rio Muni in 1603 (A. Jones, German Sources for West African History, 15991669 ; 27 [English translation], 345 [original German; compare with W. Crecelius, "Josua Ulsheimer Reisen nach Guinea und Beschreibung des Landes," 104-5]). See illustration in T. Griffon du Bellay, "Le Gabon," 2:316. 93. The languages of the Ngounie province (Sira, Punu, Lumbu, and Sango) have kazilkatsi for both mother's brother and wife, except for Tsogo, which borrowed it for mother's brother but not for wife. See G. Dupre, Un ordre et sa destruction, 151-67, for a description of matrilineages and clans in one population. 94. For South Gabon and Congo, cf. G. Dupre, Un ordre, 163; P. P. Rey, Colonialisme. For Myene, Fang, and southern Gabon, see H. Bucher, "The Mpongwe of the Gabon Estuary," 28-30. Generally, see H. Deschamps, Traditions orales, 32,34,45,52,68-69,88-89, 100. 95. For vocabulary, see G. Gaulme, Le pays de Cama, 102-14; for politics , ibid., 146-48; for the official Maramba association extending from Mayombe to Cape Lopez, see A. Battell, Strange Adventures, 56-58. In later centuries a "dance," ivanga, among the Myene and even all along the coast of Rio Muni reproduced part of a Loango-type titulature and appropriate roles; C. Gonzalez Echegaray, Estudios Guineos, 183-99, and 1. A. Avaro, Un peuple gabonais ii l'aube de la colonisation, 107-14. Among the Myene the dance belonged to the Abulia and Apandji clans (109). The latter has the same name as the Buvandji clan which ruled Loango. 96. F. Pigafetta and D. Lopes, Description, 31, for the mention of the extension of Loango to Cape Lopez. See discussion in G. Gaulme, Le pays de Cama, 92-119. See also A. Battell, Strange Adventures, 58, who mentions lords of Sete and of Kesock, east and north of Mayumba, but does not make clear whether they still belonged to the realm of Loango. Given the political significance and extent of Maramba, it seems likely that they were dependent on Mayumba. 97. G. Gaulme, Le pays de Cama, 101-2, 114-16, 119; 1. A. Avaro, Un peuple gabonais, 45-47. 98. These are the confederations of N. Metegue N'nah, Economies et societes au Gabon dans la premiere moitie du XIxe siecle, 18-20. 99. See n. 95 and K. D. Patterson, The Northern Gabon Coast to 1875,2023 ; G. Rossel, "Gewasinnovaties in Gabon" 1:125-26 (diffusion of crops), 2:238-59 maps. Notes to Pages 161-64 349 100. M. Alihanga, Structures communautaires traditionelles et perspectives cooperatives dans la societe altogoveenne, 89-156 (Mbede government), 58-72 (wars with Mbosi, Tege, and others); T. Obenga, La cuvette congolaise, 33-70 (government); 1. Ollandet, "Les contacts Teke-Mbosi," 1:140-89,226-67 (Mbosi government); 1. Laffier, "Beitriige zur Ethnologie der Tege," 77-146 (Tege government ); Y.-N. Gambeg, "Pouvoir politique et societe en pays teke," 315-24 (Tege-Mbosi-Mbede wars); F. Ewani, "Recul et stabilisation teke," 230-36 (Tege-Mbosi wars). 101. H. Deschamps, Traditions orales, 74 (Shake); L. Perrois, La circoncision Bakota, 17-18, (Kota wars), 39-46 (circumcision), 72-81 (associations), n. 11 (villages); E. Darre, "Notes sur la tribu des Bomitaba," 314-16; E. Darre and Le Bourhis, "Notes sur la tribu Bomitaba," 25-27 (government, titles, emblems); E. Darre, "Notes sur les Kakas de la circonscription de l'IbengaLikouala ," 16-17 (officials). 102. 1. Vansina, "Probing the Past," 345-53, and the use of terms for "clan" related to Kongo kanda. For religious vocabulary, see 1. F. Thiel, Ahnen, Geister, Hochste Wesen, 38-58,92-98,113-15,122-23,144-48,154-56. 103. Ndaywell, "Organisation sociale," 253-57, 268-75, 318-409; 1. Vansina, "Probing the Past," 340-45. 104. See n. 70. 105. As appears from the following preliminary statistic. Out of 21 known terms with a political meaning in Saa, the affiliation of 6 remains undetermined although most of these seem to be innovations. Of the 15 others: 3 are western Bantu, all with innovation in meaning; 4 relate to Teke languages (but 2 of these have the same root) with innovation in meaning, and 1 of these is shared with Burna; 3 relate to other lower Kasai languages; 2 are related and probably derived from Kongo, 2 are innovations; and 1 is derived from Mongo. In Jia, njuuwele ("paramount chief") corresponds to ndjuu (Saa) + wele (Boma). Out of 14 Boma terms, 4 remain undetermined, all with original meanings. Of the others, 3 relate to Mongo; 2 to lower Kasai; 1 each to Teke, Yans, and Saa; 1 is derived from Burna; and 1 related to Burna and Kongo. Out of 22 Yans terms: 6 remain undetermined, mostly original; 6 are related western Bantu forms with identical meanings in Kongo or Tio; 3 to lower Kasai generally; 2 to Boma (one of which is also related to Mongo); 1 each to Kongo, Teke, Mongo, Burna, and western Bantu. Out of 11 Burna terms: 3 remain undetermined; 2 are western Bantu; 2 are related to lower Kasai generally (one of which, however, is also related to Kongo); 2 to Teke (one of these is also related to Saa); 1 to Boma and Kongo; 1 to Yans. 106. R. Tonnoir, Giribuma, 261-312. 107. M. Storme, Ngankabe, and H. Hemeni, "Historique de la tribu des Banunu," 2-6. The earliest references to Mwene Mushie are on maps only. 108. R. Tonnoir, Girubuma, 103-217; E. Sulzmann, "Orale Tradition." 109. See n. 105. 350 Notes to Pages 164-69 110. In contradiction to Tonnoir, E. Sulzmann, "Orale Tradition," 543-45, 569, claims an origin for the Boma rulers among the Bolia nkumu. This may be so, yet Boma political terminology contains several basic elements which are shared with other lower Kasai groups. For example, nkese, "political domain," is in fact derived from Burna ke-se. It seems as if a Mongo group founded a kingdom over existing principalities of the lower Kasai type. Later Mongo influences in the region are evident, especially among the Jia, whose dynasty came from Mai Ndombe (ibid., 546-47) and from Mongo-Boma wars, R. Tonnoir, Giribuma, 221-24. The use of the term nkum in the Saa expression kumemembe, "chief of police," is also a Mongo loan (mbe being probably identical with Tio mbee, "capital"). Toponyms such as Mushie or Mosenge are derived from Mongo bosenge, "capital"; 1. Vansina, "Probing the Past," 353. Finally one should not forget the implantation of the Ngwi nkumu farther upstream on the southern side of the lower Kasai, from where they influenced their neighbors. 111. The Ding, the Lwer, and the Mpur claim this. See Ndaywell, "Organisation sociale," 353-65, and esp. 368-69; 1. Mertens, Les Ba Dzing de La Kamtsha, 9-13. Chapter Six: The Eastern Uplands 1. F. Goffart, Le Congo, 94; M. Robert, "Considerations suggerees par l'etude du milieu physique centre-africain," population maps; P. Gourou, "Notice de la carte de densite"; C. Delhaise, Les Warega, 24, 39-40. The higher densities in eastern Maniema seem to be a demographic spillover from the very dense populations farther east, although culturally they do not belong to the great lakes. 2. E. Meeussen, "De talen van Maniema," and the lexicostatistic surveys. Nyanga and Tembo are languages of the great lakes group. Southern central Sudanic so strongly influenced the whole Komo group (Komo, Bira, Amba, Huku) and Bali that most ofthese were labeled "Sub Bantu" by M. Guthrie, The Classification of the Bantu languages, 18-19,40, 83, 85 (Amba, Angba, Bali, Bira, Bua, Huku). For language contacts in these areas, see A. Vorbichler, "Sprachkontakte am Beispiel einiger Sprachen und Dialekte in Nordost-Zaire," 433-39. 3. 1. Vansina, "The Past." The section has benefited from C. Keirn's commentary and information, for which I am very grateful. 4. On language families see chapter 2. Besides these 28 languages another 4 at least had been spoken in the area before. Such a polyglot situation was preserved wherever extended households with their retainers were the basic ethnic referents, the usual situation among the non-Bantu speakers. Language then was part of the household identity. See B. Costermans, Mosaique bangba. Dominant languages spread and were replaced, but the speech of the home was clung to tenaciously. As a result the linguistic history of the area is remarkable for its unique detail. One day this will allow for the elaboration of a detailed institutional and cultural history based on relative "fine" linguistic chronology. Notes to Pages 169-73 351 5. On the Uele Neolithic, cf. F. Van Noten, The Uelian; F. and E. Van Noten, "Het ijzersmelten bij de Madi"; A. de Calonne-Beaufaict, Les Azande, 135-38,145. F. Van Noten, ed., The Archaeology, 58-59, holds the controversial view that the whole Uele Neolithic is of Iron Age date. The data do not mean that the smelting of iron was adopted only in the seventeenth century. The fabrication of hematite axes and the use of iron tools seem to have coexisted for almost two millennia. 6. Provisionally C. Ehret et aI., "Some Thoughts on the Early History of the Nile-Congo Watershed," 87-90; N. David, "Prehistory and Historical Linguistics in Central Africa," 80-81; M. McMaster, "Patterns of Interaction." 1. Goeyvaert's ongoing linguistic research is expected to clarify greatly the early history of the southern central Sudanic group. 7. Iron was introduced first from the west, later from the east, according to M. McMaster, "Patterns of Interaction," chap. 3, sec. 2. 8. On fishermen, M. McMaster, "Patterns of Interaction," chap. 5, sec. 2, argues that some Ngbandians (Sango) were fishermen on the Uele and the Bomokandi, and were later replaced on these rivers by Kango. On the Abarambo, cf. Bruggen, "Rapport constituant la chefferie Kiravungu," 1; I. Czekanowski, Forschungen im Nil- Kongo-Zwischengebiet 6 (2): 224. 9. D. E. Saxon, "Linguistic Evidence," 70-73, 75-76; N. David, "Prehistory ," 88-91. 10. C. Ehret et aI., "Early History of the Nile-Congo Watershed," 89-90, 99-102; N. David, "Prehistory," 81-82. 11. M. McMaster "Patterns of Interaction," chap. 3, on Buan. There is a vast literature on pygmies in Ituri, but it tends to be ahistorical; see 1. Vansina, "Do Pygmies Have a History?" 12. M. McMaster, "Patterns of Interaction," chap. 6. On material culture, see C. Van Overberghe and E. De Jonghe, Les Mangbetu, 241-89. 13. M. McMaster, "Patterns of Interaction," chap. 7, secs. 2-4. On terminology in Mangbetu, cf. M. I. Hubbard, "In Search of the Mangbetu"; M. McMaster, "Patterns of Interaction," chap. 7, sec. 3. On Mangbetu expansion, cf. De Maeyer, "Etude generale sur la tribu des Makere," 1; N. Chaltin, "Exploration de la Lulu et de l'Aruwimi," 105-8 (Popoi); P. Schebesta, Vollblutneger und Halbzwerge, 44,80,152 (Abelu, Lombi, Meje); H. Van Geluwe, Les Bira et les peuplades limitrophes, 11, 12 (Lombi). 14. M. McMaster, "Patterns of Interaction," chap. 7, sec. 4, and illus. 7.7 and 7.8; C. Van Overberghe and E. de Jonghe, Les Mangbetu, 530-41; I. Czekanowski, Forschungen, 6 (2): 567. 15. On the lower social strata, see E. De Jonghe, Les formes d'asservissement dans les societes indigenes du Congo Beige, 79,81,83-85,89,98 n. 1, 105; on "leading fighters" and "mercenary warriors" (mando in Mangbetu), cf. Van Ermingen, "De i'organisation des Makere du territoire de Zobia," 1; "Les Bekeni," 4; H. Hackars, "Inspection du territoire d'Avakubi," 4; "Etudes sur 352 Notes to Pages 174-75 les pygmees: Territoire des Babali/Barumbi," 1,5 (leading fighters among Bali, Lika, Bodo, and Mangbetu). 16. Despite the appearance of an Omaha system of kinship terminology, after marriage by delayed exchange with transfer of matrimonial goods became the preferred form of marriage over the original sister exchange (lengbe) in the Mamvu tradition a lineage structure still did not develop. 17. On legitimacy by the use of affinal links, see C. Keirn, "Precolonial Mangbetu Rule," 41. On Manziga, see p. 176. 18. M. Siffer, "Note generale sur les Mabodu," 1-3; A. Winckelmans, "Histoire generale de la tribu Maha de la peuplade Mabudu"; H. Van Geluwe, Les Bali et les peuplades apparenUJes, 13-15. The emigration may perhaps be connected to repercussions in the upper Ituri region, caused by the expansion of the kingdom of Bunyoro in western Uganda. Bodo letakaletaka, "land," is related to butaka with the same meaning in the languages of the great lakes. 19. M. Siffer, "Note generale," 4; H. Hackars, "Inspection du territoire d'Avakubi," 4-5; A. Bertrand, "Note sur les Mabudu-Madimbisa (dits Wasumbi)," 7; H. Van Geluwe, Les Bali, 78-80. There just might be a relation between Mamvu-Balese embaa and Mangbetu ne + mba, "tribute" or "metal." But tones, vowel length, and vowel quality are too poorly rendered in the sources to conclude. 20. I. Czekanowski, Forschungen 6 (2): 567. M. Siffer, "Note generale," 5, 7-9, underlines the importance of manpower and a temporary unified command in war (gama, "chief"). 21. Bouccin, "Les Babali," 695. By 1900 the mambela chains (seven) did not yet encompass all Bali villages, and one chain used a rite different from the six others; see P. E. Joset, Les societes secretes des hommes-leopards en Afrique, 98-106. H. Van Geluwe, Les Bali, 64-67, and 67 n. 1 gives a bibliography. 22. The term mukama for a political position also occurs among the Pere (D. Biebuyck, The Arts of Zaire, 2:247-48), perhaps among the Mbo and Ndaka, and among the Nande and Toro. On poisoned arrows in the region, see E. Perrot and E. Vogt, Les poisons de fieches, 100-103, and references to T. H. Parke and H. M. Stanley. 23. The earliest reference here to leopard men dates to 1895; see P. Salmon "Les earnets," 260; H. Van Geluwe, Les Bali, 85-88. Leopard men were known elsewhere in the region of Beni and in southern Maniema; d. P. E. Joset, Societes secretes, 17-91, 106-17 (Pakombe, Bali, and neighbors), who claims that Aniota (or Anioto) is actually a Bodo term (55 n. 1); E. Cordelia, "Appunti sulla zona del Maniema," 976 (Lega). 24. C. Keirn, "Precolonial Mangbetu Rule," 108-11. The Mabodo now coined the bava- neologism for the name of lineages after mava- coined by the Meje. Their ethnic name is Ma-bodo, rather than Ba-bodo, because its present form is a nineteenth-century Mangbetu name. By then such was the prestige of the Mangbetu that the Mabodo adopted the name Mabodo for themselves. Notes to Pages 175-78 353 25. 1. Czekanowski, Forschungen 6 (2): 413; H. Seidel, "Das Uele-Gebiet," 185. On the religious significance of the rocky outcrops, see B. Costermans, "Tore, God en Geesten bij de Mamvu en hun dwergen," 540-41, 546. 26. Van Ermingen, "Organisation," 1; A. Landeghem, "Etude preparatoire sur les Babua," 1-4; A. Hutereau, Histoire des peuplades de l'Uele et de l'Ubangi, 45-46; A. de Calonne-Beaufaict, Les Azande, 125-27. On Makere-Buan relations , see chapter 4 for developments on the Likati. 27. A. de Calonne-Beaufaict, Les Azande, 96-98, 137-39; A. Hutereau, Histoire, 246-48; F. Nys, Chez les Abarambo. 28. On the traditions collected since the 1860s, cf. C. Keirn, "Precolonial Mangbetu Rule," 38-40; R. Bertrand, Notes pour servir a l'etude des Mangbetu, 27-80; P. Denis, Histoire des Mangbetu et des Matschaga jusqu'a l'arrivee des Belges, 7-20. 29. C. Keirn, "Precolonial Mangbetu Rule," 51. 30. Ibid., 50-52, 60; R. Bertrand, Notes, 81-114; P. Denis, Histoire des Mangbetu, 21-46. 31. C. Keirn, "Precolonial Mangbetu Rule," 72-77 (politics of kinship), 81-90 (ideology, magic, emblems), 97-101 (advisors). 32. Ibid., 93-97 (military organization), 119-59 (economic institutions). 33. Ibid., 52-63, 230-40; R. Bertrand, Notes, 115-26; P. Denis, Histoire des Mangbetu, 47-109. 34. On nebeli, cf. C. Keirn, "Precolonial Mangbetu Rule," 90-92; L. Vincart, "Notes pour servir a I'histoire des peuplades environnant Ie poste de Massidjadet," 546; G. Casati, Zehn Jahre in Aequatoria 1:102; 1. Czekanowski, Forschungen, 164-67; C. Delhaise-Arnould, "Les associations secretes au Congo: Le Nebili ou Negbo," 284. For events after 1873, see C. Keirn, "Precolonial Mangbetu Rule," 240-96; P. Denis, Histoire des Mangbetu, 111-28; R. Bertrand, Notes, 126-27. 35. D. Biebuyck, Arts of Zaire, 266-67, for quotations. 36. The earliest general overview is A. Moeller, Les grandes lignes des migrations des Bantous de la province orientale du Congo Beige. Governor Moeller summarized the traditions of origin and the general features of social and political structure of the populations as reported by his administrators. Readers must know administrative goals, thought, and practice of the period and have access to at least some of the original reports, as well as to earlier documents, to use the volume intelligently. 37. D. Biebuyck, Arts of Zaire, 267-68 (quote from 287). 38. Along the Lualaba, fishermen and traders speaking languages of this group went downstream and occupied the present site of Kisangani by or before c. 1750. Cf. A. Droogers, The Dangerous Journey, 31-33; idem, "Les Wagenia de Kisangani entre Ie fteuve et la ville," 155-57; T. Papadopoullos, Cases of Tribal Differentiation, 17-21, who documents an Enya settlement even farther among the Mba on the Lindi (41-42). By 1955 they had abandoned their language. 354 Notes to Pages 178-81 39. Mainly the people now called Nyanga and Tembo. For estimated dates, see map 2.8. 40. For associations in southeastern Zaire, cf. G. Wauters, L'esoterie des noirs devoilee, 97-111 (Songye); P. Colle, Les Baluba, 2:527-627, and T. Reefe, The Rainbow and the Kings, 13-14, 46-48 (Luba); A. F. Roberts, "Social and Historical Contexts of Tabwa Art," 35-36 ("Butwa" association). One could argue that they existed by 800 on the grounds that shell currency made sense only for social payments (see n. 43). Even if that was so, and the connection remains weak, that still dates the rise of associations more than half a millennium after people from the southeast settled in the forests of Maniema. 41. 1. Druart, "Carte succincte du District du Maniema," notes on "Wagenia" and n. 38. The language was Enya and the common name was Genya north of Kindu; farther south the fishermen were called "Baluba." 42. Products traded included copper, palm products, iron products, salt, raffia cloth, raffia threads (to Kivu), red earth, red powder from the Pterocarpus soyauxii tree, lion teeth, bark cloth, beans, and fish (near the rivers). Major trade currents involved iron (mostly from Lega country and Manyara southeast of Kasongo), salt (Bushi near Lake Kivu, Micici, Kirundu, Nyangwe, the Malela district west of Nyangwe), raffia cloth (of Songye origin from the Lomami and beyond), palm products (areas near the confluence of the Elila and west of the Lualaba farther downstream from the confluences of the Ulindi and the Lowa), and copper. But no author has paid much attention to precolonial trade, and references are extremely scattered. In general, see E. Cordelia, "Appl.mti," 963-72. For details about the Lega, cf. C. Delhaise, "Chez les Wassongola du sud ou Babili," 165-66; L. Lietard, "Les Warega," 133-34; G. Ainti, "Les Mituku," 1-2. On quartz crystals, see D. Biebuyck, Arts ofZaire, 206-7, and R. Packard, Chiefship and Cosmology, 68-69. 43. D. Biebuyck, "La monnaie musanga des Balega," 676-84; 1. Bettendorf , "Die Angst der Mituku vor den Mizimu under deren Vetretern, den raffinierten Mekota," 119; A. Moeller, Grandes !ignes, 404, 427, 434; P. De Maret, Fouilles archeologiques dans la vallee du Haut-Lualaba, Zaire 1:165-66, 260-63 (A.D. 800 and later). De Maret mentions a possible connection with social payments to be made on entering an association. 44. D. Biebuyck, Lega Culture, 7-12,82-84; idem, Arts of Zaire: 209-20, diagram 9. Yogolelo Tambwe ya Kasimba, introduction ii l'histoire des Lega, 26, is very specific: only the high dignitaries of the bwami knew the history and genealogies of the Lega. 45. W. de Mahieu, Qui a obstrue fa cascade? 31-42; idem, Structures et symbofes: 1,56-58, 115 n. 5; idem, "A l'intersection du temps et de l'espace, du mythe et de I'histoire, les genealogies," 415-37. 46. The ethnography of southern Maniema remains poorly known with the exception of the Lega and Bembe peoples studied by D. Biebuyck and Mulyumba. However, these constitute the two largest population groups of the area, and many different groups have evidently been incorporated under the Notes to Pages 181-83 355 label Bembe and even Lega. The uncertainties of ethnic nomenclature are described by D. Biebuyck, Lega Culture, 3-8, and idem, Statuary from the PreBembe Hunters, as well as by Mulyumba wa Mamba Itongwa, "La structure sociale des Balega Basile," 13-22. 47. A. Moeller, Grandes !ignes, 421,433,443,446,450-64; D. Biebuyck, Arts of Zaire: 210-11; T. Papadopoullos, Cases of Tribal Differentiation, 22-25. 48. M. Piscicelli, Nel Paese dei Bangobango, 156-58 (Bembe: "La casta dei bwami e fra i bakombe, potentissima"); E. Cordelia, Verso rEWa, 41, 103-4; C. Delhaise, Les Warega, 337-45; D. Biebuyck, Lega Culture, 46-50, 85. On the privileges of bakota dignitaries among the Mituku, see 1. Bettendorf, Die Angst der Mituku, 145-46. Along the Lualaba the Zanzibari called the chiefs they installed mokota, the title of the high dignitaries in the overarching associations, thus recognizing them as the previous government. 49. D. Biebuyck, Lega Culture, 7-8 and n. 12, 11-12,82, and idem, Arts of Zaire, 205-11, for a tradition of Lega origins from the Lualaba to the north; C. Delhaise, Les Warega, 47-48, refers not to Lega origins (from the south) but to the Museme epic, as does A. S. Clarke, "The Warega," 66. But, given the genetic relationships of their language, the bulk of the Lega people clearly do not come from the northwest but from the south. Hence what do the traditions about the Lualaba refer to? The direction might be a logical cosmological conclusion , if the creation was in the absolute downstream (D. Biebuyck, Lega Culture , 11 n. 14). But there seems to be more than this. The tradition seems to tell of the origin of the lower bwami ranks, despite the aphorism, "Bwami has no inventor; it is the fruit that came from above." Given the distribution of related associations this makes eminent sense. 50. D. Biebuyck, Lega Culture, 83-84, relates an opinion that bwami was invented to do away with war. 51. C. Van Overberghe, Les Basonge, 475-76 (the groves were still recognizable in 1906; see end map); T. Reefe, Rainbow and the Kings, 130-31,230 n. 10; A. Verbeken, "Accession au pouvoir chez certaines tribus du Congo par systeme electif," 654-57 (Lwaba, ehata); idem, "Institutions politiques indigenes," 1-3. The rotation was announced by a new circumcision cycle; see C. Wauters, L'esoterie, 125-26, 159, 164-65. Eata is a reflex of *-pata (like Kongo libata). 52. A. Moeller, Grandes lignes, 151-53, 508-11; T. Reefe, Rainbow and the Kings, 130-32, 149-50. Eata flourished before 1820 among the Hemba, Songye, and Kusu. Moeller gives the seven predecessors of Lusuna, which covers most of the eighteenth century (153). Eata antedates Luba influences on Songye political institutions which began c. 1700. A date for its emergence in the fifteenth and sixteenth centuries is derived from the date of arrival of bwami's block of upper ranks among the Sileo 53. See n. 47. The term mokota may perhaps also be due to eata influence. 54. A. Moeller, Grandes !ignes, 428. 55. Ibid., 435-50, 453-54; C. Delhaise, "Chez les Wasongola," 120-21, 134-35, 195-99; D. Biebuyck, Arts of Zaire, 235-38, 243-46; E. Cordelia, 356 Notes to Pages 183-84 "Appunti," 971; Cdt. Borms, "Reconnaissance du pays Bango-bango et d'une partie de I'Uzimba," 256b. 56. A. Moeller, Grandes /ignes, 421-35; D. Biebuyck, Arts of Zaire, 23840 ; 1. Jak, "Eenige ethnographica over deWalengola-Babira," 48-50; Van Belle, "Territoire des Walengola-Wasongola-Mituku," 4-9; 1. Bettendorf, Die Angst der Mituku, 118-19,145-47; H. Marmitte, "Baleka-Mituku," 3-8; 1. Jak, "De bakota en hun grafhutten in Belgisch Congo," 37-39. 57. On the ngodzi and his lieutenant moganda, see A. Moeller, Grandes lignes, 433; Van Belle, "Territoire," 4. Note the stress on wealth and the acquisition of clients by giving them women to marry. 58. D. Biebuyck, Arts of Zaire, 241-43; K. Kudi, Le Lilwakoy des Mbole du Lomami, 4-32; V. Rouvroy, "Le'Lilwa'," 783-98; 1. Tachelet, "De Montfortaanse missie in Belgisch Kongo," 44-49; R. V. Abbeloos, "De sekte lilwa," 311-31; R. Bouccin, "Au sujet du Lilwa"; 1. F. Carrington, "Lilwaakoi-a Congo Secret Society," 237-44. For the lilwa as a boy's initiation, see n. 82; M. De Rijck, "Une societe secrete chez les Lalia Ngolu," 2-7. 59. R. Philippe, Inongo, 8-108. 60. D. Biebuyck, Lega Culture, 82-84; ibid., 205, 209-16, 281-83 (origin of the Babongolo near Kirundu; "kaluba" sold the hat to the Babongolo); Yogolelo Tambwe ya Kasimba, Introduction, 22 (Kimbimbi). Mulyumba wa Mamba Itongwa, "Structure sociale," 309-19, shows that although in the decentralized western area, studied by Biebuyck, almost every head of a House held the highest rank (one man per 41.8; p. 318), in the east centralization was far greater, as there were four main kingdoms by the 1880s and hence there was one man of the highest rank for about 20,000 people! 61. Mwami and bwami, reflexes of *yama, an ancient root in the languages of the great lakes, are found in almost every language of the group, including all the languages adjacent to the forest. 62. Bishikwabo Chubaka, "Note sur l'origine de l'institution du 'bwami' et fondements de pouvoir politique au Kivu oriental," 6 (the Nyindu were then Lega) , 5-13 (the acquisition of bwami as told by the Nyindu and their neighbors ), 13-17 (structure of the Nyindu bwami), and idem, "Histoire d'un etat shi en Afrique des grands lacs: Kaziba au Zaire," 243-49; Mulyumba wa Mamba Itongwa, Structure sociale, 319-29 (structure of Sile bwami), 8-11 and 329-43 (bwami wa lusembe), 343-47 (Nyindu and Shi traditions), 331 (succession list of the 'Alenga dynasty at 25 years per generation, starting in the third, yields c. 1670 for its onset; Shi calculations also yield a date well before 1700, despite Bishikwabo, "Note sur l'origine" 16). By rough reckoning a century would elapse before the bwami reached this area from the Lualaba, and another century for the spread of the eata from the Lomami, which would then have started in the fifteenth century or c. 1500. Trees in the sacred groves of the Sile dynasty and in the eata groves can eventually be dated and yield a better chronology. Mugaruka bin-Mubibi, "Histoire clanique et evolution des etats dans la region sud-ouest du lac Kivu," 425, gives c. 1240-1260 for the elaboration of bwami on the Ulindi. Even corrected to 1300-1350 (leaving the Ulindi, cf. p. 400), the Notes to Pages 184-86 357 dating seems early. It is ultimately derived from a spurious early chronology for Rwanda beginning in A.D. 906 (165-67, 371, 405-7). 63. R. Sigwalt, "The Early History of Bushi," 79-142, 187-339; Mugaruka bin-Mubibi, "Histoire clanique," 329-434, annexes. Bishikwabo Chubaka, "Histoire"; R. Sigwalt, pers. com. (1985), 17, now estimates that a large part of Maniema was occupied by speakers of great lakes languages before the Lega immigrated. This is untenable, except perhaps for the easternmost Lega groups (Sile, Mwenda, Nyindu). 64. For example, the mpunju, and bubake rituals; cf. D. Biebuyck, Lega Culture, 82, 85 n. 7; idem, "Organisation politique des Nyanga," 313-15 (bubake); idem, "La societe kumu face au Kitawala," 33-34 (mpunju). 65. Although D. Biebuyck, Arts of Zaire, 6, says: "The bwami association ... dates from the beginnings of Lega society" and therefore gives no priority. On kinship, cf. D. Biebuyck, "Maternal Uncles and Sororal Nephews among the Lega; idem, Les Mitamba: Sysemes de mariages enchaines chez les Bembe; and idem, "De vorming van fiktieve patrilineaire verwantschapsgroepen bij de Balega" (unrelated groups attached as segments to patrilineages). For the territorial organization, cf. C. Delhaise, Les Warega, 339-47 (chiefs); D. Biebuyck, Lega Culture, 46-50. For gumi: cf. E. Burke, A Small Handbook of Kilega, 36 (bugumi), 62 (mugumi). 66. The considerable differences between the accounts of D. Biebuyck, Mulyumba, and Bishikwabo relate to these differences in size. Biebuyck studied the western, decentralized Lega, the others the statelike social formations of the Sile and Nyindu. 67. D. Biebuyck, Arts of Zaire, 246-57, his "Organisation politique," "De Mumbo-instelling bij de Banyanga," and Rights in Land and Its Resources among the Nyanga, 8-11, 26-33; R. Hoffman, "Rapport sur la chefferie Bapere," 1-4; A. Moeller, Grandes !ignes, 296 (Pere borrowed political titles from the Nande). So far no publications exist about the Tembo. 68. From the Hunde and Nande specifically. For the latter, see R. Packard, Chiefship and Cosmology, 28-52,67-71, 125-26. 69. D. Biebuyck, "Mumbira" 42 (Pere, Mbali, Ndaka, Mbo, Mbuti, and hence Bira share aspects of this initiation); L. de Sousberghe, "Mumbira et Limombo des Nyanga," 42 n. 21. 70. Mbuntsu, mambela, mondaa, and dokpo. The latter two occur in W. de Mahieu, Qui a obstrue la cascade? 32. Of the nsindi initiation among the Pere, only the name is known. For the Nyanga and Pere, see D. Biebuyck, Arts ofZaire, 246-51; his "Nyanga Circumcision Masks and Costumes," "Mumbira," and "Sculpture from the Eastern Zaire Forest Regions"; R. Hoffman, "Croyances et coutumes des Bapere," 1-8. From the Pere we know only that they had healing associations "like the Komo" (Hoffman, "Croyances et coutumes," 3), although one also found elements typical for the nkumbi initiation of Ituri such as occurred, e.g., among the Bira, cf. C. Turnbull, Wayward Servants, 63-65; H. Van Geluwe, Les Bira, 82-94. Almost every Nyanga House had its own rituals. D. Biebuyck, "Mumbira," 42, notes the names of mbuntsu, mangwe, mpande mukuki, ima 358 Notes to Pages 186-88 nekukuya, ukanga, kakoka, kasindi, bubira, mwari, bundia, nyamosoku, and mumbira, kasumba, lusumba, and esomba, several of which are found in Ituri, northern or southern Maniema. Common vocabulary items related to brotherhoods abound in northern Maniema and Ituri in particular. 71. As seen from the retention of such terms as kumu ("chief"), etuka ("House"), noko ("mother's brother"), and gandja ("boy's initiation"). 72. Because they have been the object of outstanding studies by W. de Mahieu, the Komo are privileged in this section. On their atomization, see W. de Mahieu, Structures et symboles, 5-17 and 17 n. 1; and F. Stradiot, "Rapport concernant l'organisation des populations Barumbi et Bakumu des territoires de Makale-Wandi, Lubutu et Ponthierville," whose results suggest that each clan counted on average perhaps 300 members and each House in it on average fewer than 40 (cf. W. de Mahieu, Structures et symboles, 8). Yet the House, coinciding with the village, was the only unit in physical reality. As a result of the atomization , it is nearly impossible to recover much history before 1876, apart from migrations and cosmology; see ibid., 6 (failure of the administration to regroup segments by common genealogies), 8 (genealogical memory was five to seven generations deep), and 15 (collective memory was mainly maintained by settlements over hunting grounds). 73. Although this description is derived from Komo data it applies also to the Mbuti (cf. C. Turnbull, Wayward Servants, and his The Mbuti Pygmies, 17986 ) and to the Bira of the forest (cf. e.g., 1. Czekanowski, Forschungen, 348-55; M. Siffer, "Note generale sur les Babira," 3; P. E. Joset, "Vie juridique et politique des Babombi," 4 [a-bJ). After 1876 great changes occurred with the Zanzibari domination. 74. Apart from recently borrowed terms there was apparently no terminology for dependent statuses. 75. W. de Mahieu, Structures et symboles, 7. Houses split easily. Hence the technical name for segments, osuk6, i.e., "place where the generation of the grandchildren of X live"; W. de Mahieu, "Les Komo," 103 n. 10. 76. The Bira and Balese of Ituri were less mobile because farming was much more important to them than to the Komo; cf. H. Van Geluwe, Les Bira, 51-52, and idem, Mamvu-Mangutu et Balese-Mvuba, 129-30. The available data on the Bira and Balese, the "pygmy hosts," remain quite scanty, mostly byproducts of the research on hunters and gatherers. It may be briefly noted that chiefs here as in Komo land were a Zanzibari innovation, that the Bira advanced with the Zanzibari and later the Belgians into Balese lands, and were also quite mobile at that time. See M. Baltus, "Les Balese," 117-21, for the recent history of the farmers of Ituri. 77. W. de Mahieu, Structures et symboles, 55,63,73-74. 78. Ibid., 77, for the lack of congruence between conceptual and physical realities as the engine behind these dynamics of innovation. De Mahieu gives the example of kasea, introduced from the Lengola to replace komp6mbo, which was felt to be ineffective. On the introduction of umbd, see W. de Mahieu, Structures et symboles, 86-88. Its rituals were inspired by the esomba rites, but Notes to Pages 189-98 359 in turn the gandja borrowed certain prerogatives from umba; see W. de Mahieu, Qui a obstrue La cascade? 9. In this area yaba diffused in the early 1930s from Komo land (where it was a ritual sequence) as far as the Uele; cf. W. de Mahieu, Qui a obstrue La cascade? 207, 310-13; E. Bock, "Trois notes sur la secte Biba"; Five, Etudes Bamanga; S. Kreutz, "Extrait du rapport trimestriel." 79. See nn. 70, 85. 80. Gandja, ganza, kandza, mu-kanda, etc., "initiation for boys," are all reflexes of *-ganja found from Ubangi to Ituri and northern Maniema, as well as from the middle Kwango to the Lualaba and the middle Zambezi. W. de Mahieu, Qui a obstrue fa cascade? 27-31, is the only author to give a wider semantic field, including "sperm" and "madness." These may be local derivations of meaning. 81. Cf. map 6.4. See n. 69 (diffusion of the mambefa style with a bird sequence); W. de Mahieu, Qui a obstrue La cascade? 31-34, for aLutu and kentende. 82. LiLwa is best known from the Lokele, Eso, and Mbole regions. The overall practices and rituals of initiation for boys found among the Bali, and there named mambela, were also common among their neighbors to the west as far as Basoko under the name of lilwa. Cf. H. Sutton Smith, Yakusu, the Very Heart of Africa, 8,9,28,63-68; W. Millmann, "The Tribal Initiation Ceremony of the Lokele," 364-80; Van Dieren, "La circoncision chez les Bamanga"; 1. FraBle, Meiner UrwaLdneger, 55 (So); idem, Negerpsyche, 85 (the name LiLwa). 83. The term moame may have been introduced here only after c. 1876; see W. de Mahieu, Structures, 73. His Qui a obstrue La cascade? 141-378, details the initiation and the cycle of the mena gandja. See also D. Biebuyck, Arts of Zaire, 233-35, but note that he gives Umba fathers as the highest rung (234) and perhaps other details are of "colonial" vintage. See n. 78. 84. W. de Mahieu, Qui a obstrue La cascade? 382-412. 85. The diffusion of two successive patterns of circumcision, the bwadi and the -tende, over areas stretching from the middle Kasai to the Lowa River and from Shaba to the Lindi River, respectively, both probably moving from south to north, confirms a complex and very dynamic history. Interchange from neighbor to neighbor affected enormous areas, and diffusions certainly took place in both directions, not just from south to north. 86. Seepage is indicated by the diffusion of emblematic objects such as bells and throwing knives, or by the still little-known diffusion of the lesser Asiatic crops such as taro and sugar cane. But, as the examples show, these kinds of innovations exerted at best a minor influence on the attitudes of peoples in the area and on their institutional developments. Chapter Seven: Challenge from the Atlantic 1. A fully detailed history of the area has not yet been written. For the written sources see chapter 1. Over most of the affected area reliable oral traditions allow one to upstream well into the eighteenth century, and in some 360 Notes to Pages 198-202 cases into the seventeenth. After 1800 the oral record becomes especially detailed . Written sources for the nineteenth century have by no means been exhausted for the coastal areas, and there still are discoveries to be made even for the eighteenth century. 2. The exemplary systematic study of the Atlantic trading system is 1. Miller, Way of Death. It deals with the Portuguese slave trade in the eighteenth century, is centered on Angola, and touches only incidentally on the forest areas. Still its discussion of the overall organization, the commercial mechanisms , and the institutions of the trade is invaluable and fully applies to the harbors on the Loango coast. 3. 1. Vansina, Kingdoms of the Savanna, 41-46,41-46; W. G. L. Randles, L'ancien royaume du Congo, 87-96. 4. R. Garfield, "A History of Sao Tome Island," 1-32. Slaves from the mainland were imported after 1493. The settlers acquired some 930 slaves between 1494 and 1499 (14). 5. J. Vansina, Kingdoms of the Savanna, 46-54; W. G. L. Randles, L'ancien royaume du Congo, 130-38; A. Hilton, Kingdom of Kongo, 50-68. On nzimbu, see F. Pigafetta and D. Lopes, Description, 157-58 n. 37. On the volume of slave exports, see P. Lovejoy, Transformations in Slavery, 38. B. Clist, "Pour une archeologie," map 3, indicates ceramic links between the Malebo Pool area and the lower Zaire in the sixteenth century (group 1 and 5 wares) and also the river upstream (group X ware found from Mai Ndombe to the banks of the lower Kwango and at sites in Kinshasa, including Ngombela). Ngombela has been dated in one instance to A.D. 1645; see D. Cahen, "Contribution a la chronologie de l'age du fer dans la region de Kinshasa," 135, but clearly more dating is needed. On this site, see H. Van Moorsel, Atlas de prehistoire de la plaine de Kinshasa, 224-77, and the nzimbu treasure (256 bottom). 6. A. Hilton, Kingdom of Kongo, 69-141. The controversy among historians about the origin of these Jaga continues. For a caravan consisting of multiple parties, see A. Cavazzi de Montecliccolo, Descrifa6 hist6rica dos tres reinos, Congo, Matamba e Angola 1:112 (ch. 1, sec. 229),245-46 (ch. 2, sec. 245). 7. P. Martin, External Trade, 33-42. 8. Ibid., 43-68. 9. Ibid., 70, 130-31; O. Dapper, Naukeurige Beschrijvinge, 146, 157-59. The Ivili of Gabon now settled on the lower Ngounie are probably descendants of such Viii traders. Marriage with a close relative of a trading partner was the preferred strategy. It is often referred to, from Gabon to the Malebo Pool, from early times on and was still practiced in the nineteenth century. Bloodbrotherhood was the preferred link in the inner basin by the 1880s and was then spreading far and wide, but the practice may be quite old nevertheless; see H. Tegnaeus, Bloodbrothers, 104-18. For these and other ties among the Tio, see 1. Vansina, Tio Kingdom, 87,263-64. 10. An Ibare "kingdom" appears in texts of c. 1612 and 1620: A. Brasio, Monumenta Missionaria Africana 6: 104, 438. Ibare refers to the name ebale for the Zaire River, common in languages of water people north of the Alima, and Notes to Pages 203-4 361 its "kingdom" was the stretch of river upstream of the Alima confluence. The name Moenhemuge (Mwene Mushie) already appears on a map of 1561. This refers to Mushie on the Kwa and its lands inhabited by the Nunu; see E. Sulzmann, "Orale Tradition," 527. 11. O. Dapper, Naukeurige Beschrijvinge, 202, mentions cloth, metalware (especially copper basins), and shells or beads in the Kongo trade. Guns were in demand, but the Portuguese did not sell them. Textiles and guns, along with brass pans and other metal items, were to become the staples of the trade. By 1700 the composition of the packet for a slave already included the three categories of commodities described; see 1. Barbot, Jr., and 1. Casseneuve, "An Abstract of a Voyage to the Congo River or the Zaire and to Cabinde in the Year 1700," 509-12. Luxury textiles had been a sign and an instrument of social stratification before the Europeans arrived, and the role of foreign raiment and cloth as items of prestige and status was but a continuation of this habit (0. Dapper, Naukeurige Beschrijvinge, 199, for the wearing of imported cloth by the great). Leaders also wanted European guns, first as a symbol of power, then to arm elite guards, and again this affected local balances of power with the same results as described. Because gunpowder could not be made locally, the acquisition of guns created perforce a need for gunpowder. Meanwhile near the coast the importance of textiles or sundries as political weapons lessened as they became more common when the volume of such imports grew. The rulers on the coast constantly demanded different novelties. Hence such eccentric gifts as the two peacocks, the two white dogs, and the Dutch drum for the king of Loango in 1612 (S. Brun, Schiffarten, 10). Meanwhile the mass of earlier imports remained in demand by people of lesser status as they strove to emulate their lords. Finally textiles and guns gradually became necessities rather than luxuries on the Loango coast. 12. The figures given are estimates derived from data in P. Lovejoy, Transformations in Slavery, 52-54; P. D. Curtin, The Atlantic Slave Trade: A Census, 224-25 fig. 16, 266 fig. 26, tables 33, 34, 36, 63-67; P. Martin, External Trade, 73-92, 124, 137-38; K. D. Patterson, The Northern Gabon Coast to 1875,32-38; R. A. Austen and K. Jacob, "Dutch Trading Voyages to Cameroun: 17211759 ," 10-13. Lovejoy and Curtin give totals only for the Angolan and Loango coast trade combined. 1. Miller, Way of Death, 233, has slightly higher estimates than are given here for the northern coasts between 1650 and 1830. 13. K. D. Patterson, Northern Gabon Coast, 8, 14 (buying slaves), 32-38; H. Bucher, "The Mpongwe," 120; R. A. Austen and K. Jacob, "Dutch Trading Voyages." 14. The outer limits are described after S. W. Koelle, Polyglotta Africana; P. D. Curtin, Atlantic Slave Trade, 255-56 (maps), 289-90, 295-96, 298 (dates); T. E. Bowdich, Mission from Cape Coast Castle to Ashantee, 426-30; and the first European observers of the 1880s. Summary for Congo and southern Gabon, P. Martin, External Trade, 124-29; for Cameroon, see A. Wirz, Vom Sklavenhandel zum Kolonialen Handel, 82-91. 362 Notes to Pages 204-9 15. P. D. Curtin, Atlantic Slave Trade, 224-25 fig. 16,226 fig. 26, tables 66, 69,74; P. Lovejoy, Transformations in Slavery, 141, 144; 1. Miller, Way of Death, 233; P. Martin External Trade, 138-42. After 1830 the data are quite deficient. 16. P. Martin, External Trade, 93-115. By 1700 this system was already firmly established. Cf. 1. Barbot, Jr., and 1. Casseneuve, "Voyage to the Congo River," 509-12. 17. P. Martin, External Trade, 102; P. de Marees, Beschryvinghe, 242-43 (fort on Corisco in 1600); P. Van den Broecke, Reizen naar West Afrika, 26 (temporary shop at Cape Lopez), 57 (a house for trade in Loango by 1610); L. Degrandpre, Voyage ala cote occidentale d'Afrique fait dans les annees 1786 et 1787, 65-67. 18. P. Martin, External Trade, 70 (charms for protection), 118-22 (caravans ). On canoes and flotillas, cf. R. M. Eggert, Das Wirtschaftssystem, 70-72, 118-20; C. Coquilhat, Sur Ie Haut-Congo, 85. 19. R. Harms, "Oral Tradition and Ethnicity." 20. For descriptions of currencies, see A. Mahieu, Numismatique du Congo, 1484-1924 (Zaire); T. Obenga, La cuvette congolaise, 105-9, 1.-F. Vincent , "Dot et monnaie de fer chez les Bakwele et les Djem" 283-91; 1. Guyer, "Indigenous Currencies and the History of Marriage Payments"; 1. Guyer, "The Iron Currencies of Southern Cameroon"; A. Cureau, Les societl!s primitives, 298-301. A study of the relevant terminologies and of their relative spatial extension has not yet been undertaken. For the "money market" at the Malebo Pool, see 1. Vansina, Tio Kingdom, 282-88. 21. The notion of wealth has been little studied. But see P. Van Leynseele, "Les Libinza," 96-137, esp. 96-98 (mosolo, "wealth"). 22. For slave status and terminologies, see E. de Jonghe, Les formes d'asservissement. See also n. 5. On the gradation of unequal statuses as reflected in terminology, see R. Austen, "Slavery among Coastal Middlemen," 307-20. Household slaves were employed on plantations by their owners in the seventeenth-century Kongo capital and in both the Gabon Estuary and the hinterland of Douala by the nineteenth centuries. On legal condemnation to slavery, see P. Martin, External Trade, 167-68. 23. For this subsection, see 1.-L. Vellut, "Notes sur l'economie internationale des cotes de Guinee inferieure au XIXe siecle," 15-16,22 (multiplication of posts), 39-49 (growth of trade between c. 1845 and 1890); A. Pinto, Angola e Congo, 234-37, 352-66, 370-400 (the Loango coast in 1882); A. Wirz, Vom Sklavenhandel, 96-99 (Kribi); C. Duparquet, "Etat commercial de la cote du Loango et du Congo," map of commercial stations from Sete to Ambriz in 1875; and the commentary published by F. Bontinck, "Etat commercial du littoral ... Document inedit," 165-83. 24. P. Martin, External Trade, 149-57; 1.-L. Vellut, "Notes sur l'economie," 22-24,28 (ivory imports in London); A. Wirz, Vom Sklavenhandel, 60,63-66; G. Dupre, "Le commerce entre societes lignageres," 625-26, 630-32 (Nzabi), 642 (Punu slave plantations); see n. 34 for special groundnut plantations in Gabon. Notes to Pages 209-15 363 25. P. Du Chaillu, Voyages et aventures dans l'Afrique equatoriale, 44-50. 26. P. D. Curtin, Atlantic Slave Trade, 252, table 75, for the lower figure, which comes from ships known to the British Foreign Office; 30,000 is an estimate by Dutch traders based on the tonnage of slave ships leaving the Zaire Estuary (see J.-L. Vellut, "Notes sur I'economie," 40). No thorough quantitative research has yet been undertaken on these coasts in this period. For the general situation, see P. Martin, External Trade, 143-49. 27. 1.-L. Vellut, "Notes sur I'economie," 23-24; A. Wirz, Vorn Sklavenhandel , 63. 28. See pp. 233-35. 29. On technology, see D. R. Headrick, "The Tools of Empire," 83-149, 165-79. On trade, travelers, and treaties, see N. Metegue N'nah, L'implantation coloniale au Gabon, 15-29,32-35. 30. 1. Vansina, "Esquisse historique" 26-32; G. Rossel, "Gewasinnovaties," 123-27; W. G. L. Randles, L'ancien royaume du Congo 66-67, for domestic animals. Pigs are first mentioned on the mainland c. 1575 (F. Pigafetta and D. Lopes, Description, 76) but were already kept on Sao Tome by c. 1500 (R. Garfield , "History of Sao Tome Island," 22). Later they were common only in the Kongo area. They may have been introduced from Europe. Sheep antedated the Europeans but disseminated more in the eighteenth and nineteenth centuries. Ducks were introduced by Europeans well after 1600. 31. R. M. Eggert, Das Wirtschaftssystem 11-17 (cassava in Equator Province ), 16 (yields of about 20 tn.lha., slightly higher than about 15-18 tn.lha. for ABB plantains). Cassava bread was made at the Malebo Pool by 1698; tobacco was in great demand there, but still not cultivated (see F. Bontinck, Diaire congolais, 127). For the labor involved in preparing such bread, see 1. Vansina, Tio Kingdom, 148-50. 32. R. Harms, River of Wealth, River of Sorrow, 53-54; G. Sautter, De l'Atlantique 1:272-74. See M. Miracle, Agriculture, 1-5, "A. Manioc," for the situation by 1950. 33. M. Miracle, Agriculture, map 1-5; idem, Maize in Tropical Africa, 82, 94-95; J. Vansina, Children of Woot, 176-78, for the advantages of maize over other cereals among the Kuba. 34. A. Raponda-Walker, "Enquete sur l'Agriculture noire au Gabon et sur certaines techniques utilisant des produits vegetaux," 725-28, has the term -boga, "field for groundnuts," in the whole Ngounie area and among the Seki near Rio Muni. Among several peoples (Punu, Gisir, Wungu, and Tsogho) the term was used for a large field of groundnuts, smaller fields having special names. This situation testifies to the importance of groundnuts as a commercial crop in the nineteenth century. G. Le Testu, "Notes sur les cultures indigenes dans I'interieur du Gabon," 547, confirms. He also mentions that only two species of beans were grown on the coast, and there they were planted on small surfaces only. Yet beans were important in the slave-trading era as food for 364 Notes to Pages 215-19 slaves on board ship; cf. N. Uring, The Voyages and Travels, 41,47. Beans were also a major crop in the inner basin during the nineteenth century. 35. 1. Vansina, Tio Kingdom, 266-310, 450; M. d'Atri, "Relations sur Ie royaume du Congo, 1690-1700," 64 (Malebo Pool); R. Harms, River of Wealth, (middle Zaire River); G. Le Testu, "Notes sur les cultures indigenes," 542 (grown in falga), 543 (Masango). 36. For examples of intensive agriculture, see G. Sautter, De l'Atlantique 1:262-64 (Bonga), 500-510 (Kongo), 606-11 and 617-18 (Kamba); and V. Drachousoff, Essai sur l'agriculture indigene au Bas-Congo, 127-55. For kitemo, working parties, and then associations, see G. Sautter, De l'Atlantique, 561-63; A. Thonnar, Essai sur Ie systeme economique des primitifs, 64-65; 1. Vansina, Tio Kingdom, 81, 114. 37. P. Van Leynseele, "Les Libinza," 55-56; Mumbanza mwa Bamwele, "Histoire," 219-22, 254-58, 282-88, 312-14, 319-26. 38. G. Sautter, De l'Atlantique 1:259-64; P. Geisler, "Allgemeiner Bericht tiber die Kongo-Lobaje Expedition," 106. 39. See map 7.2 and also, e.g., P. Van Leynseele, "Les Libinza," endmap. 40. E. Zorzi. Al Congo con Brazza, 371,376. 41. R. Harms, River of Wealth, 48-70, discusses the growing regional economy along the Zaire upstream of the Malebo Pool. See also 1. Vansina, Tio Kingdom, 277-81. 42. P. D. Curtin, Atlantic Slave Trade, 116-26,205-73, and summary fig. 26 (266); P. Lovejoy, Transformations in Slavery, 44-54, 144-46. 43. 1. Miller, Way of Death, 151, 153, table 5.2 (165). 44. H. Klein, "The Portuguese Slave Trade from Angola in the Eighteenth Century," 914 n. 41; 1. Miller, Way of Death, 159-64. Fewer than 10 percent of the slaves exported were children; see H. Klein, "Portuguese Slave Trade," 9035 . The loss of the reproductive capacity of 4,525.9 women (all presumed fertile) not compensated by the higher fertility of polygynous households amounts to 2,444,000 -:- 6 = 407,333.33 = 0.11 percent per year. 45. L. Degrandpre, Voyage 2:x, 25, 37 (1786-87 and earlier). He counted a sixth (16.6 percent) as from the Zaire River, a fourth (25 percent) as, or acquired by, Teke, and the rest (58.4 percent) as "Mayombe," including southern Gabonese. Slaves bought by Kongo, almost all from central Africa, were sold in Kabinda; at Malemba, the third harbor, both "Mayombe" and "Kongo" were sold. Equal weight in numbers is given here to the -three harbors. In reality Loango had a smaller volume of slave exports then than the two others and hence Kongo are certainly underestimated. A third of the Kabinda exports is credited to Mondonge and two-thirds to Kongo and Sonho, which may again be an underestimate. There is something wrong with S. Koelle's sample of slaves c. 1830-1840 (P. D. Curtin, Atlantic Slave Trade, 295-98): there are no slaves at all from upstream on the Zaire farther than the confluence of the Kwa. On the other hand the relative numbers given for lower Zaire, lower Congo, Gabon, and Cameroon correspond to expectations. Notes to Pages 220-25 365 46. Too many variables are involved here to allow for any useful calculation . Women's work increased during the period of heavy slave trading because they took over a portion of male tasks. Moreover the processing of cassava added greatly to the work load. For unfree women the work load may have been heavy enough to increase the chance of miscarriages. And as R. Harms, River of Wealth, 182-84, found among the Bobangi under these circumstances, many women resorted to abortion. 47. 1. Vansina, Kingdoms ofthe Savanna, 48-50. King Affonso rejected the gist of the regimento of 1512, which proposed to model Kongo's administration, titulature, and laws after those of Portugal. A. Hilton, Kingdom of Kongo, passim, discusses the integration of foreign religion and foreign ideas into the social structure of Kongo before 1665. But her reconstruction of Kongo religion and of some social institutions before 1483 goes too far beyond what the evidence warrants (8-31). 48. A. Hilton, Kingdom of Kongo, gives perhaps the best-balanced account despite a tendency to overrate economic factors and underrate institutional innovation in Houses and matrilineal groups. See also 1. Thornton, "Early Kongo-Portuguese Relations: A New Interpretation," 183-204. 49. 1. Thornton, The Kingdom of Kongo, 85-121 (general), 117-19 (new oral tradition); 1. Janzen, Lemba, 1650-1930,49-51,61-70. 50. Contrast the evidence in P. Martin, "Family Strategies in NineteenthCentury Cabinda" with the idealized views in 1. Martins Vaz, No mundo dos Cabindas 1:171-77 (kinship), 2:11-59, 151-75 (state of Ngoy), and 2:175-83 (dynastic history of Ngoy). On ndunga, see ibid., 1:50-58; Z. Volavka, "Le Ndunga." 51. P. Martin, External Trade, 159-74. There was a regency already in 1701; d. N. Uring, Voyages, 29; F. Hagenbucher Sacripanti, Fondements spirituels, 61-64, 70, 72, 75-76, 78-99. P. P. Rey, Colonialisme, 255-59; 1. Janzen, Lemba, 46-49. 52. P. Martin, External Trade, 168-69, P. P. Rey, Colonialisme, 197-99, 266,514. 53. 1. Janzen, Lemba, 5 (map of trading routes); P. P. Rey, Colonialisme, 220-24; P. Martin, External Trade, 18 (Bungu destroyed in 1623 and "Jaga" in 1642). 54. 1. Janzen, Lemba, 51-61,70-79 (centrallemba region), 95-272 (variants ) . 55. 1. Vansina, Tio Kingdom, 455 (map 16), 456-63; P. Bonnafe, Histoire sociale d'un peuple congolais, 47-51; M.-C. Dupre, "Naissances et renaissances du masque kidumu," 27-45; 1. Loftier, "Beitrage," 78-79; Y.-N. Gambeg, "Pouvoir politique," 464-94; J. Ollandet, "Les contacts Teke-Mbosi," 197-201; T. Obenga, La cuvette congolaise, 41. The contents of major Tio nkobi are said to stem in part from Loango and symbolize the ocean, wealth, peaceful relations , and power. 366 Notes to Pages 225-29 56. M. Alihanga, Structures, 151-53 (warlords), 240-41 (ancestral sanction ), 259 (nkobi); L. Perrois et al., Gabon: Culture et techniques, 77-79 (carvings surmounting Mbede reliquaries); E. Andersson, Contribution al'ethnographie des Kuta 1:169-71 (warlords). 57. R. Harms, River of Wealth, 30, 71-85, 126-42; 1. Vansina, Tio Kingdom , 449; F. E. Dhanis, "District d'Upoto," 26, 27 (slave origins); 28 (trade language as far as Yalulema). 58. R. Harms, River of Wealth, 119-121, 143-59, 163-74 (older House), 175-96 (newer firm), 183-84 (population control); Mumbanza mwa Bamwele, "Histoire," 462-72. A. Thonnar, Essai sur Ie systeme economique, 17-27,30-37 (both lower Zaire and inner basin); Mpase Nselenge Mpeti, L'evolution; 16, 94, 159-60 (the mbwayoyi leader of a firm among the Ntomba of Mai Ndombe and firms). 59. 1. Vansina, Tio Kingdom, 74. 60. A. Thonnar, Essai sur Ie systeme economique, 37-39 (women's work and hierarchy. What he says of a lemba wife in lower Congo also applied to first wives in the firms of the inner basin). On social mobility, see R. Harms, River of Wealth, 157-58 (in Bolobo the slave Mobombo succeeded the chief Ebaka); M. Froment, "Trois affluents franc;ais du Congo," 467 (at Bonga the slave Ndombi succeeded to the paramount Mpakama, probably in the 1870s or earlier, but lost the paramountcy); 1. Vansina, Tio Kingdom, 408-9 (Ngaliema); and other cases are known (Mumbanza mwa Bamwele "Histoire," 472-74). Ngankabe, "queen of the Nunu," was an example of a successful woman leader in the 1870s and 1880s; see M. Storme, Ngankabe, 66-76. In 1886 a woman, Combabeka, also exercised considerable influence in Bonga as an ally of now "old" chief Ndombi; see C. Coquery Vidrovitch, Brazza et la prise de possession du Congo, 18831885 ,467. 61. For ezo and likundu, see R. Harms, River of Wealth, 200-203. The belief in likundu witch substance and the autopsy after death spread from the lower Congo and Gabon along the trade routes as far as Uele; see H. Baumann, "Likundu: Die Sektion der Zauberkraft." 62. G. Hulstaert, Le mariage des nkundo, 348-54, and a story of how a usurper died on the spot (348-49, bolumbu). 63. L. Siroto, "Masks and Social Organisation among the Bakwele People of Western Equatorial Africa," 294-314. 64. C. Coquilhat, Sur Ie Haut-Congo, 189-94, 198-212 (Mata Bwike); A. Thonnar, Essai sur Ie systeme economique, 26. 65. Mumbanza mwa Bamwele, "Histoire," 457-62; 1. Vansina, Tio Kingdom , 74,255-56. All figures given by contemporary visitors were estimates only, but they were internally consistent. 66. For Nkasa slave raiding on Momboyo-Tshuapa, Lulonga, cf. G. Van der Kerken, L'Ethnie Mongo, 469-71, and index "Nkasa" (1100); G. Hulstaert, Elements, 70-71. For traditions in the Lobaye area: L. Bouquiaux and 1. N. C. Thomas, "Le peuplement oubanguien," 810-12; S. Bahuchet, "Notes pour Notes to Pages 230-33 367 l'histoire de la region de Bagandou," 54-55 (Ngando); 1. M. C. Thomas, Les Ngbaka de la Lobaye, 255-58. 67. E. Sulzmann, "La soumission des Ekonda par les Bombomba," 5-11, 17; Colle, "Les Gombe," 162-67. E. 1. Glave, Six Years ofAdventure in Congoland , 188-205. 68. Almost every source comments on fortifications in these areas. For example, E. Froment, "Un voyage dans I'Oubangui," 211 (Monzombo and Lobala); Colle, "Les Gombe," 165-66, 167; E. 1. Glave, Six Years ofAdventure, 189; and remnants seen by M. Eggert, "Archaologische Forschungen," 3238-40 (Ngombe-Mongo); F. Thonner, Dans la grande foret de l'Afrique centrale, 45, 64, 68; 1. De Wilde, "Dans la Mongala," 187; S. Bahuchet, "Notes pour I'histoire," 54-55 (Ngando). However, the area completely escaped the attention of P. Briart, "Les fortifications indigenes au Congo." 69. H. Nicolai, Le Kwilu, 117-29; G. De Plaen, Les structures d'autorite chez les Bayansi, 3-12; H. Van Roy, Les Byaambvu du Moyen Kwango, 80128 ; M. Plancquaert, Les Yaka, 110-18, 176; R. Tonnoir, Giribuma, 218-24; Ndaywell, "Organisation," 399-408. 70. 1. Vansina, Children of Woot, 127-96. 71. J. H. Van Linschoten, Itinerario, 4-13; P. de Marees, Beschryvinghe, 243-45; K. D. Patterson, Northern Gabon Coast, 1-25; H. Bucher, "Mpongwe Origins," 60-61,64-70; A. Raponda Walker, Notes d'histoire du Gabon, 49-59; E. Mbokolo, Noirs et Blancs en Afrique Equatoriale, 12-28. See also chapter 5, nn. 95 and 98. 72. P. de Marees, Beschryvinghe, 244-51; 1. A. Avaro, Un peuple gabonais, 97-114, 143-46; G. Gaulme, Le pays de Cama, 89-175. See also chapter 5, nn. 95-97. 73. K. D. Patterson, Northern Gabon Coast, 26-67; H. Bucher, "The Mpongwe," 117-39; idem, "The Settlement of the Mpongwe Clans in the Gabon Estuary," 149-75; T. E. Bowdich, Mission from Cape Coast Castle to Ashantee, 424-39, for the hinterland of Gabon as it was then known to traders. 74. 1. A. Avaro, Un peuple gabonais, 146-54 (Orungu), 219-29 (Galwa); A. Raponda Walker, Notes d'histoire du Gabons, 60-89; K. D. Patterson, Northern Gabon Coast, 68-89; Bodinga wa Bodinga, Tradition orale de la race Eviya, 10; H. Deschamps, Traditions orales, 21 n. 1,40 n. 19,42 n. 22 (kidnaping bird); M. Koumba-Manfoumbi, "Les Punu du Gabon, des origines a1899," 241-50. 75. E. Ardener, "Rise of Trading Polities between Rio del Rey and Cameroons : 1500-1650"; J. Bouchaud, La cote du Cameroun dans l'histoire et la cartographie, 35-90. In 1603 Ulsheimer traded for ivory in the Cameroon Estuary . There was no major town, no trade language was known, and trade occurred on the river. Indeed the inhabitants first tried to storm the ship. Despite there not being a major town, there must have been an above-average density of population, given that they attacked with some 60 canoes; see W. Crecelius, "Josua Ulsheimers Reisen nach Guinea und Beschreibung des Landes," 102-3 (translated in A. Jones, German Sources, 25,343-44). 368 Notes to Pages 233-37 76. R. Austen and K. Jakob, Dutch Trading Voyages, 24; 1. Bouchaud, La cote du Cameroun, 91-123; L. Z. Elango, "Britain and Bimbia in the 19th Century (1833-1878)," 11-17. See E. Ardener, Coastal Bantu of the Cameroons , 27, for the expUlsion of a major Duala leader, Bile, from Douala to Bimbia well before 1826 as the last break off. Earlier Duala leaders had left to try their luck as leaders in Kpe country or formed fishing settlements (Kole) west of Mt. Fako. G. Balandier, "Economie, societe et pouvoir chez les Duala anciens," 361-79, stresses the role of the associations, but misses the genealogic irregularities which clearly betray that the Duala were, by 1800, a collection of Houses not strictly organized into lineages. 77. P. Laburthe-Tolra, Les seigneurs de la foret, 127-96; P. Alexandre, "Protohistoire," 531-35, 557 (Bulu, Fang); 1. F. Vincent, "Traditions historiques chez les Djem de Souanke," 65-67 (Bulu war); idem, "Dot et monnaie de fer," 276-78 (Djem and Bekwil). 78. C. Chamberlin, "The Migration of the Fang," 429-56; H. Deschamps, Traditions orales, 65-66 (Kota), 72 (Shake), 75-76 (Bekwil), 78-80 (Bichiwa), 83-86 (Fang), 91-92 (Ntumu), 103 (Fang Betsi). See chapter 5, n. 101. 79. G. Dupre, Le commerce, 645-57; E. Andersson, Ethnographie, 26-30 (30-35 is pure speCUlation). 80. E. Mbokolo, Noirs et blancs, 29-147; idem, "Le roi Denis," 73-95; H. Bucher, "The Mpongwe," 223-359. Ntoko was the best-known merchant prince, well in evidence by 1836, and the one who helped the Presbyterians to start a mission in 1842; see T. Omboni, "Viaggi nell'Africa Occidentale," 217-31 (219-20 for portrait of Songhey and Ntoko, "esteemed traders on these coasts entirely trustworthy in matters of business"). 81. K. D. Patterson, Northern Gabon Coast, 108-49; 1. A. Avaro, Un peuple gabonais, 152-73 (Orungu), 231-36 (Galwa); G. Gaulme, Le pays de Cama, 189-96; O. Lenz, Skizzen aus West Afrika, 193-205; idem, "Reise auf dem Okande in West Afrika," 250-52. 82. A. Wirz, Vom Sklavenhandel, 45-52; idem, "La riviere du Cameroun," 178-94; R. Austen, "Slavery among Coastal Middlemen," 305-33; E. Mveng, Histoire du Cameroun, 171-81; 1. Bouchaud, La cote du Cameroun, 125-57; L. Z. Elango, "Britain and Bimbia," 25-33 and passim; C. Dikoume, "'Les Elog Mpoo," 114 n. 6, 211-12. 83. A. Wirz, "La riviere de Cameroun," 176-78, for missionary influences in the matter. 84. 1. Jacobs and 1. Vansina, "Nshoong atoot: het koninklijk epos van de Kuba," 32-35. 85. Corrosive effects are not an inevitable law of nature. In many cases the first political impact of a world economy was to help produce a spatial scale of political organization to match the economic expansion. The Lunda state in central Africa grew into a slave-raiding empire after 1700 and began to feel "corroding effects" only by the 1850s. In West Africa the Asante and Danxome states established themselves in an area and era of intensive participation in the Atlantic trade. Rather than corroding or wilting away under the onslaught of Notes to Pages 239-44 369 ever more massive trade, a series of states from Egypt to Madagascar blossomed in the nineteenth century as imperial expansions by economic proxy. Chapter Eight: Death of a Tradition 1. This estimate was given in 1919 by the commission for the protection of the autochthons for the Belgian Congo (0. Louwers and A. Hoornaert, La question sociale au Congi, 23) and may well turn out to be an underestimate. The cases reported in note 13 may appear to be dramatic extremes. but there are many parallels especially from lower Zaire, lower Congo. Equateur province, and north Kasai, which explains why the commission gave its estimates. Equally dramatic losses were reported from north Congo (especially the lower Likouala, Alima, and Sangha areas). the lower Lobaye, and from most of Gabon. Major epidemics, a primary cause of mortality, tapered off only after the swine flu epidemic of 1918-1919, and when measures in the struggle against sleeping sickness became effective, also after 1920. 2. C. Keirn, "Precolonial Mangbetu Rule," 232, 240-310; P. Ceulemans, La question arabe et Ie Congo (1883-1892); 326-31; R. D. Mohun, "Consular Dispatch from Basoko Camp," 36; P. Salmon, La derniere insurrection de Mopoie Bangezegino. 3. 1 Vansina, Kingdoms of the Savanna, 235-41; P. Ceulemans, La question arabe, passim; F. Bontinck, L'autobiographie de Hamed ben Mohammed elMurjebi Tippo Tib (ca. 1840-1905). On Zanzibari auxiliaries, cf. P. Salmon, "Les carnets de campagne," 272-73; R. D. Mohun, Consular Dispatch, 41. On Zanzibari influence just after their defeat (Fall 1894), see E. 1 Glave. "New Conditions in Central Africa" 31:906-15. 32:699-705; and by May 1920, 1 Druart, "Carte succincte," legend "arabises." 4. 1. Abemba. Pouvoir politique traditionnel et Islam au Congo oriental, 914 ; W. de Mahieu, "Les Komo," 99-100. On mokota chiefs, Zanzibari rule, and chiefs in lower Lomami, see L. Appermans, "Chefferie Kombe," 67, 69, 73. 5. N. Metegue N'nah, L'implantation coloniale au Gabon, 15-65; E. Mbokolo, Noirs et blancs, 41-218, esp. 151-75; O. Baumann, Eine Afrikanische Tropen-insel: 104-5. 6. A. Lejeune Choquet, Histoire militaire; La Force Publique: 1 Meyers, Le prix d'un empire. On the 1897 mutiny, cf. P. Salmon, La revolte des Batetela de {'expedition du Haut-fturi; l-L. Vellut, "La violence armee dans I'Etat Independant du Congo." 7. Recent studies are D. Van Groenweghe, Du sang sur les lianes; l-L. Vellut and D. Van Groenweghe, "Le rapport Casement"; R. Harms, "The World Abir Made." The military histories (see n. 6) carry reports of only the few more sizable military operations. 8. M. Denis, Histoire militaire de {'Afrique equatoriale franr;aise. For the rubber regime, see C. Coquery Vidrovitch, Le Congo, esp. 71-220. 9. H. Rudin, Germans in the Cameroons, 193-97; A. Ruger, "Die Aufstand der Polizeisoldaten"; individual reports on military expeditions mainly 370 Notes to Pages 244-45 in the issues of Deutsches Kolonialblatt, e.g., Scheunemann, "Die Unruhen im Sudbezirk von Kamerun 1904 bis 1906" 18:347-52, 391-99. For the rubber regime, see A. Wirz, Yom Sklavenhandel, 107-47; and for a specific situation by 1914, see L. Haase, Durch unbekanntes Kamerun, 93-96. 10. For military operations against the Bekwil, cf. M. Denis Histoire militaire; Treichel, "Die Lage im Postenbereich Sembe von 1 Oktober 1912 bis 30 September 1913"; Heym, "Die Gefechte gegen die aufstandige Ebaleute im Bezirk Sembe"; G. TeBmann, "Ethnologisches aus dem sud6stlichen Kamerun," 91; C. Robineau, "Contribution it l'histoire du Congo," 302-13. 11. P. Salmon, "Les carnets de campagne," 244. The area was at war from 1886-1887 to 1906 with successive Zanzibari and Independent Congo State occupations, including two year-long uprisings (1894-1895 and 1906) against the rubber regime. 12. M.-C. Dupre, "La dualite du pouvoir politique chez les Teke de l'ouest," 80-90, esp. 88-89. The last campaign lasted from 1912 to 1916. The data for the three remaining villages are from 1933. 13. In general, see R. Headrick, "The Impact of Colonialism on Health in French Aequatorial Africa, 1880-1934." On sleeping sickness, see M. Lyons, "The Colonial Disease." On the lower Kwa, see A. De Backer, "Brief," 460-61 (Ebeke, arrival and destruction caused by the arrival of the Ponthier column); E. Costermans, "Le district du Stanley Pool," 26; idem, "Reconnaissance du pays entre Ie Congo, Ie Kassai et Ie Kwango," 40-42, 53-54 (still many people in 1897; Tua had 7,000 inhabitants); L. Anckaer, De Evangelizatiemetode van de Missionarissen van Scheut in Kongo, 110 (the mission was abolished in 1900 because the population was perishing as a result of sleeping sickness); L. Hamerlinck, "Un voyage chez les Bamfunuka," 17 (in 1932 sleeping sickness still ravaged the country and Tua counted only a few hundred inhabitants-maybe 5 percent of its former population). For a case in lower Zaire, see H. A. Johnston, George Grenfell 2:547-53. 14. See A. De Backer in n. 13. 15. R. H. Nassau, Fetichism in West Africa, 141-55 (cases in Corisco, Rio Mun!, Batanga, and near Ndjole from the 1850s to 1879). Ibia j'Ikenge, Customs of the Benga and Neighboring Tribes, with a motto on the title page: "The customs of the people are vain," quoting the prophet Jeremiah. 16. For the Lilwti, see H. Sutton Smith, Yakusu, 61-68; W. Millman, "The Tribal Initiation Ceremony of the Lokele," 365-68. The mambela of the Bali in the same area was eventually outlawed as a "front" for terrorist attacks, after missionaries such as 1. Christen and Kawaters ("Reifezeremonien") pressed the government to "exile all authorities responsible for mambela" (1. Christen, Mambela et Anyoto, 54). Kuba initiation was strongly opposed by all missionaries and finally forbidden in the early years of the depression, because it lowered productivity. Although associations often continued in abbreviated and anemic forms in many cases, very few survived in their original richness, as happened among the Tsogo. For an introductory summary of their bwiti, see A. RapondaWalker and R. Sillans, Rites et croyances, 171-214. R. Sillans and O. Gollnhofer have published many works about this bwiti. Notes to Pages 246-48 371 17. A sharp increase in the rise of specific religious movements to cope with this repression occurred after 1920 as a last major reaction stemming from adherence to the principles of the old tradition. For examples of such movements , see G. Dupre, Un ordre et sa destruction, 354-85; J.-F. Vincent, "Le mouvement Croix Koma"; J. Vansina, "Lukoshi, Lupambula." These should not be confused with the rise of African churches or the Fang bwiti, which are attempts at the construction of a new tradition. 18. A. F. Roberts, "Social and Historical Contexts of Tabwa Art," 21 (Monsignor Roelens, bishop of upper Zaire). Such statements also occur in administrative reports. The curse of Cham was a virtual cliche among all missionaries, especially before 1940. 19. J. Vansina, "Ethnographic Account," 434-35. 20. A. de Calonne-Beaufaict, Les Ababua. The model was quickly adopted by the jurist F. Cattier and the administrators A. Bertrand ("Quelques notes sur la vie politique," 75-89) and G. Van der Kerken (Les societes bantoues, ii, 7172 , 78-79 [Ababua], 81-86), who were able to impose it as "the organization" to be found everywhere before 1920. For the Ngombe case, see A. Wolfe, In the Ngombe Tradition. "Patriarch" was a term invented by G. Van der Kerken (see Les societes bantoues, 81), according to A. Bertrand. It was given legal substance as the "paternate" a legal neologism coined by E. Possoz, Elements du droit coutumier negre, 22-24, 65-75, 97-101. 21. The direct lines of the administrative imposition of the new rural structures remain to be traced for the French and German colonies. For French Equatorial Africa, see A. Cureau (Les societes primitives) and G. Bruel (L'Afrique equatoriale franf,:aise, 186-99, and La France equatoriale africaine, 167-210), who shared the same "administrative" view of equatorial Africans as families living in villages. But how this view developed and who imposed it on the administrators remains unclear. 22. For Tio descent, see J. Vansina, Tio Kingdom, 29-40; J. Druart, "Carte succincte," "Basonge," still mentioned "a rather special system of an aristocratic republic with temporary sovereignty or presidency" in 1920. But then G. Van der Kerken, Les societes bantoues, 89, labeled this eata system as "a decadent conception, born from the loosening of tribal ties and, probably, of the disappearance in the clan or the tribe, of the family of the chiefs by right of blood." And since his book was authoritative, that was that. In the same way, he had also dismissed the existence of leaders of firms as "disaggregation" and not "the old organization" (84 n. 1). 23. A. Sohier, Le mariage en droit coutumier congolais. For lengbe, see R. Cleire, "Les bases socio-culturelles du mariage traditionnel au Congo," 60-62. On customary laws in general, see E. Possoz, Elements du droit coutumier negre. 24. J. Vansina, "The Bushong Poison Oracle," 245-60. 372 Notes to Pages 250-57 Chapter Nine: On History and Tradition 1. For instance, the largely arbitrary shapes of throwing knives allow one to attribute confidently a common origin to similar shapes in different localities. One such particular shape is the "hornbill head," now found in northeastern Gabon and neighboring areas of Congo, and c. 1610 also in Loango. That shape must have a single common origin. Mapping the names for throwing knives yields independent evidence. The "hornbill head" knives are all called -sele, usually osele. This makes the case for diffusion from a common origin practically a certainty. But shapes and names do not always match. Thus along the Lukenie River woshele indicates a throwing knife of a very different design, a design which itself extends, under various names, from the Ubangi River bend to the Kuba kingdom. In general, see 1. Vansina, Art History, 168-73. 2. Such as guns adapted to shoot spears (1. Vansina, Tio Kingdom, 123) or Cabinda shipwrights building ships seaworthy enough for sailing to Brazil despite the lack of appropriate tools (J.-L. Vellut, "Notes sur I'economie," 11 n. 2). 3. The Tio Kingdom is an extreme example of decentralization. It was a kingdom only because the inhabitants held it to be one. There was no central administration, judicial system, army. or tribute system. Senior lords were practically autonomous, but there was only one king and no duplication of senior titles. 4. As the existence of "dead zones" shows. See chapter 3, n. 36. 5. See T. K. Earle. "Chiefdoms in Archaeological and Ethnohistorical Perspective," 279-308, for a recent bibliography and survey of social evolutionary theory. Africanists have always been allergic to its appeal; see, for example, I. Kopytoff, "The African Frontier," 77-78. 6. R. Harms, Games against Nature, 97-113. 7. G. Hulstaert, Notes de botanique mongo; A. Raponda-Walker and R. Sillans, Les plantes utiles; and to a lesser degree of completeness, E. De Wildeman, Sur des plantes medicinales utiles du Mayumbe (Congo beige), and A. Bouquet, Feticheurs et medecines traditionelles. I know of the inadequacy of anthropologists from personal experience among the Tio. I simply could not follow their instructions in the botany of their grasslands. 8. L. Fiedler and 1. Preuss "Stone Tools from the Inner Zaire Basin," 182 (Ruki-Kwa stone tools). In general, see 1. Vansina, "L'homme, les fon~ts et Ie passe en Afrique." 9. M. Sahlins, Stone Age Economics, 33-34 (hunter gatherers in general), R. Harms, River of Wealth, 182-83 (Bobangi). Among the Ntomba and Sengele of Lake Mai Ndombe, a woman was to abstain from intercourse after parturition until her child was old enough to walk (about two years of age). Then she presented the child to the leader of her House or firm and received formal permission to resume intercourse. 10. See map 2.4b, after a portion of R. Letouzey's map of Cameroon. 11. Thus the correlation between the length of the dry season and the size of the working party of men needed for slash-and-burn clearing is clear. But that Notes to Pages 258-62 373 does not exclude a range of options to choose from. For variations, see L. E. Sponsel, "Amazon Ecology and Adaptation"; R. Mischung, "Se13haftigkeit und Intensivierung beim Brandrodungsfeldbau" (Hmong). 12. E. Hobsbawm and T. Ranger, eds., The Invention of Tradition; D. Lowenthal, The Past Is a Foreign Country. 13. In African historiography, moreover, "tradition"-and even more, "traditional"-in colonial times meant a static precolonial situation and signaled a complete denial or disregard of precolonial history. Hence the terms became anathema to most later historians. Some nationalist authors, however, continued to used the terms in the same meaning as their colonialist opponents had done, but reversed their emotional polarity. They now implied African originality unsullied by foreign accretions. 14. P. Bellwood, The Polynesians: Prehistory of an Island People; I. Goldman , Ancient Polynesian Society; P. V. Kirch, The Evolution of Polynesian Chiefdoms ; P. V. Kirch and R. C. Green, "History, Phylogeny, and Evolution in Polynesia"; F. Lichtenberg, "Leadership in Proto-Oceanic Society: Linguistic Evidence. " IS. As is evident from the detailed studies in E. Rogers and F. Shoemaker, Communication of Innovations, and H. Barnett, Innovation. 16. E. Shils, Tradition. 17. I. Kopytoff. "African Frontier," 33-34. 18. I. Kopytoff, "African Frontier," 10, IS, for the citations. For a listing of the major precolonial traditions in Africa, see 1. Devisse and 1. Vansina, "Africa from the Seventh to the Eleventh Centuries," 750-93. 19. I. Kopytoff, "African Frontier," 22. ...