Africa. Its Peoples and their culture history/Part Eight/Expansion of the Bantu/Northwestern Bantu
George Peter Murdock
1897 — 1985
Professor of Anthropology, Yale University
Africa. Its Peoples and Their Culture History
New York. McGraw-Hill. 1959. 456 p.
Explore also (a) the SemanticAfrica Peoples Vocabulary
(b) the Northwestern Bantu mind-mapping diagram
Expansion of the Bantu
— 35 —
The Bantu peoples occupy today approximately one-third of the African continent (see Map 14). As has been repeatedly noted, however, practically all of this vast territory was, until after the time of Christ, preempted by other peoples—mainly hunters and gatherers of the Pygmoid race in the west and of the Bushmanoid race in the east and south. During the past 2,000 years the Bantu have revealed a capacity for explosive expansion paralleled, among all the other peoples of the world since the dawn of recorded history, only by the Arabs after Mohammed, the Chinese, and the European nations since the Discoveries Period. It now becomes our task to locate the place of origin of the Bantu and to date, trace, and account for their phenomenal dispersion.
Linguistic evidence clearly indicates the region from which they came. As Greenberg (1949) has demonstrated, the Bantu languages as a group, despite their wide distribution, constitute but one of seven branches of the Macro-Bantu subdivision of the Batoid subfamily of the Nigritic stock. The other six branches of Macro-Bantu, each of which is strictly coordinate with Bantu as a whole, are confined to a very small area near the Cameroon-Nigerian border. On linguistic grounds it is impossible that the Bantu can have come from anywhere else.
The original homeland of the Bantu must have lain immediately adjacent to the territory of the speakers of the other Macro-Bantu languages, i.e., in the Cameroon highlands and a thin lowland strip connecting the latter with the coast opposite the island of Fernando Po. It must have covered an extremely restricted area, for surviving remnants of Bongo hunters provide telltale evidence of former Pygmy occupation only a few miles to the south and east.
Map 14. Culture Provinces of Bantu Africa
(1—Cameroon Highlanders, 2—Northwestern Bantu, 3—Equatorial Bantu, 4—Mongo, 5—Luba , 6—Central Bantu, 7—Northeast Coastal Bantu, 8—Kenya Highland Bantu, 9—Interlacustrine Bantu, 10—Tanganyika Bantu, 11—Middle Zambesi Bantu, 12—Southwestern Bantu, 13—Shona, 14—Thonga, 15—Nguni, 16—Sotho)
The Eastern Nigritic peoples (see Chapter 29) have already illustrated the means which first enabled the Negroes to penetrate the tropical forest. Sudanic agriculture developed in savanna country in more northerly latitudes, and its plants were adapted to these geographical conditions and not to the equatorial rainforest. The Malaysian banana, taro, and yam, on the other hand, had originated in a nearly identical environment in outheast Asia and were thus ideally suited to the habitat of the Pygmies. Once the ancestral Bantu had borrowed these crops, their expansion to the south and east became inevitable.
This conclusion rests not on speculation but on a solid induction from the lists of food plants reported in ethnographic sources. The writer has compiled such lists, containing from eight to more than twenty cultigens each, for nine tribes of Northwestern Bantu, the subject of the present chapter. Bananas, taro, and yams appear in every one of the nine, and in nearly every instance as staples, whereas no crop of the Sudanic complex except the oil palm occurs in more than one list, even including four additional shorter ones. One can scarcely conceive of stronger proof that the Northwetern Bantu could not have entered their present habitat until they received the Malaysian food plants-unless, of course, they had reverted to a hunting and gathering economy, a sacrifice which tillers throughout history have invariably refused to make.
In the absence of direct historical and archeological radiocarbon evidence, we possess only two methods of arriving at an approximate date for the beginning of the Bantu penetration of the tropical forest. One comes from linguistics. The close genetic relationship among the various Bantu languages indicates the relative recency of their differentiation from a single ancestral speech community. Olmsted (1957) has applied modem glottochronological techniques to this problem, comparing ten maximally diverse Bantu languages with one another. Though his estimates of the time of separation vary with the particular pairs compared, and especially with differences in the purely technical treatment of doubtful cognates and use frequency, his results on the whole suggest an elapsed period of about three millennia. If one allows for a thousand years of linguistic differentiation in the Bantu homeland before the dispersion began, this estimate would accord perfectly with the conclusions from the alternative line of reasoning.
The second method depends upon inferences from known historical facts. We have already examined the evidence indicating that the Malaysian food plants were established on the Azanian coast of East Africa by A.D. 60, the approximate date of the visit by the author of the Periplus of the Erythraean Sea (see Chapter 26). They might have been introduced several centuries earlier, but not much more than that if we accept the archeological evidence dating the southern expansion of the legalithic Cushites. Radiocarbon dating places Bantu at Zimbabwe in Southern Rhodesia in the seventh century. We thus have a time span of roughly 600 to 900 years within which the Malaysian crops diffused across the continent to the Cameroon-Nigerian border and the Bantu subsequently traversed the equatorial forest into East Africa. Since the Yam Belt was occupied by full-fledged agricultural peoples capable of borrowing new crops with ease, the westward diffusion presumably required considerably less time than the eastward migracion, with its necessary successive adaptations to new geographical conditions. We cannot be far wrong, then, if we fix the first century as an estimated date for the beginning of the Bantu expansion.
One obvious route into the tropical forest ran southward along the Atlantic coast, and this seems to have been followed at an early period by emigrants from the lowland portion of the Bantu homeland. This chapter deals with the present inhabitants of this region. They need not, of course, be the direct descendants of the first emigrants, for the latter doubtless pressed forward into new territory, to be followed by later waves from the homeland. One such wave, indeed, originating apparently in the Cameroon hig hlands and pearheaded by the Fang nation, has pushed southward in fairly recent times into the hinterland of the northwestern Bantu, confining the latter to a narrow coastal strip in the north, and has even reached the Atlantic in Gabon, on the equator, dividing its predecessors into a northern and a southern cluster. The component tribes of the two clusters are classified below.
The peoples of this cluster, with the possible exception of the Bubi, are doubtless the latest to leave the original homeland and the most strongly influenced by recent historical developments there.
- Bubi (Adija, Boobe, Ediye). These people, numbering about 10,000, inhabit tht offshore island of Fernando Po. Presumably because of their isolation, they are linguistically somewhat divergent.
- Duala (Diwala, Douala, Dwela), with the Bodiman, Limba (Lemba, Malimba). Mungo (Mongo), Oli (Buli, Ewodi, Eyarra, Wouri, Wuri), and Pongo. They number about 70,000.
- Koko (Bagoko, Bakoko, Betjek, Edea, Mwele), with the Bassa (Basa), Bimbi (Babimbi), Bongkeng, Dibum, Mbang, and Ndogpenda. They number abour 180,000.
- Kossi (Bakosi, Kosi, Kossi), with the Abo (Abaw, Bankon, Bo), Babong, Baffe (Bafaw), Bakaka, Bakem, Balong (Balon, Balung) , Balondo, Baneka, Bareko Flong, Manchas, Mbo, Miamilo, Musmenam, Ninong, and Sossi (Basosi, Sosi). They number about 60,000.
- Kpe (Bakwedi, Bakwele, Bakwiri, Kpeli, Kweli, Kwiri, Wakweli), wiyh the Isuwe (Bimbia, Mbia, Subu, Usuwu), Kole (Bakolle), Mboko (Bamboko, Bambuku, Boko, Womboko), and Wovea (Bobe, Bobea, Bora). They number about 50,000.
- Kundu (Bakundu) , with the Baji (Bao, Bavo, Bayi, Bolo, Bulu), Balue (Ballue, Barue), Efu, Issangili, Kogo (Bakogo) , Kombe (Ekombe, Ewuni, Kumbe), Lundu, (Balongo, Balundu, Barondo, Murundu), Mbonge (Barombi), and Ngolo. They probably number about 50,000.
- Ngumba (Mgoumba, Mvumba), with the Mabea (Bujeba, Ibea, Kaschua, Magbea). They number about 12,000.
- Puku (Bapuku, Buku, Puko), with the Benga (Benge, Mabenga, Mbenga), Beundo, Kumbe (Kombe, Ngumbi), Toko (Banaha, Banoh, Banoko, Noho, Noo), Nyong, Tanga (Batanga), and Yasa. They number about 10,000.
- Seke (Baseke, Basheke, Basiki, Museki, Segami, Sekiani, Shake, Sheekan, Sheke, Shekiani), with the Lengi (Balenge, Molingi). They number a few thousand.
In this cluster an unusually large number of Pygmies of the Bongo division live interspersed among the dominant Bantu in a symbiotic relationship with them. The ethnographic sources rarely give population data and are unfortunately deficient in many other respects.
- Duma (Aduma, Badouma, Baduma, Douma, Maduma, dumu), with the Changi (Batchangui), Kanike (Bakanike), Mbete (Ambete, Bambete, Oumbete, Umbete), Ndumbo (Andumbo, Bandumbo, Mindoumbo), and Njawi (Bandjabi, Ndjavi, Njabi).
- Lumbo (Balombo, Baloumbo, Balumbu, Lumba), with the Bwissi (Babouissi, Babwissi), gove, Rama (Barama, Bavarama, Varama), Shango (Asango, Ashango, Assongo, Machango, Masango, Massange), Shira (Echira, Eshira), Vungo (Bavoungo, Bavungo), and Yaka (Bayaka, Yakka).
- Mpongwe (Bayugu, Empoongwe, Pongo, Pongoue), with the Galoa (Galloa, Galoi, Galua, Igulua, Ngaloi), Ininga (Anenga, Enenga), Jumba (Adjumba), Nkomi (Camma, Commi, Kama), and Orungu (Oroungou, Rungu).
- Shogo (Ashogo, Ishogo, Issogo, Mitchogo), with the Kanda (Okanda, Okande), Pubi (Bapoubi, Pobi, Poubi, Pove, Powi), Puno (Apingi, Apinyi, Apono, Bapindji, Bapouno, Bapuno), and Simba (Asimba, Cimba). The Puno alone number about 40,000.
- Teke (Bateke), with the Atyo, Fumu (Bamfono, Bamfumungu, Banfumu, Banfunuka, Mfumungu, Wamfumu), Lali (Balali), Ngangulu (Bangangoulou), Ntere, Nunu (Banunu), Shikuya (Achikouya), Tegue (Ategue, Bategue), Tende, and Tsaya (Mursaya).
The Northwestern Bantu subsist primarily by agriculture, which they conduct by the swidden, or slash-and-burn, technique. We have already noted the utter insignificance of the Sudanic plants and the predominance of the Malaysian ones. Bananas constitute the staple crop in four of the fourteen societies of the province, taro in three, yams in two, and combinations of them in two others. Besides the Malaysian complex, only the American is represented other than sporadically—by beans, cucurbits, maize, malanga, manioc, peanuts, peppers, and sweet potatoes. Of these, maize, malanga, peanuts, and sweet potatoes have achieved modest importance in several societies, and manioc has attained the status of a staple or co-staple among the Mpongwe, Ngumba, and Teke.
In addition to agriculture. the Northwestern Bantu depend substanrially upon fishing, hunting, animal husba ndry, and gathering, in order of decreasing importance. Most groups keep at least a few goats, sheep, dogs, and chickens as well as occasional cats and Muscovy ducks. Some of the tribes of the northern cluster also possess pigs and cattle, but the latter are not used for milk. Though regular markets are reported only for the Duala, intertribal trade assumes modest proportions everywhere. Most tribes recognize some conventional medium of exchange, commonly bars of iron. Men do all the hunting and most of the gathering. Although they also clear the land, women perform all other agricultural operations. Both sexes engage in fishing, bur female participation consists chiefly in the catching of small fry with baskets and hand nets.
The Koko, the Teke, and the Mabea subtribe of the Ngumba live in neighborhoods of dispersed family homesteads, but the prevailing settlement pattern is a fairly compact village consisting of dwellings aligned on either side of a single street. Most villages have a “palaver house” at one end or else several such, one for each lineage, located at intervals along the middle of the street. Dwellings conform to one uniform type—a rectangular hut with a gable roof thatched with palm leaves and low walls of wood, bark, palm-leaf mats, or wattle and daub. Land is collectively owned, sometimes by the village but more commonly by its component lineages, and is parceled out to individuals in heritable usufruct without right of sale.
The Bubi, Duala, Koko, Koi, and Puku have petty paramount chiefs. Elsewhere, however, each village enjoys complete political autonomy under a headman and a council of elders. Secret societies control the government in several tribes of the northern cluster, but age-grades are reported only for the Duala. Intervillage feuding and petty warfare are endemic, but only the Koko, Kundu, Shogo, and Teke practice cannibalism, presumably borrowed in most instances from the adjacent Equatorial Bantu. Although circumcision is nearly universal, being absent only among the Bubi, girls are not subjected to clitoridectomy. All the N orthwestern Bantu keep slaves, recruited by purchase, by capture in war, or by enslavement for debt or crime. Only the Bubi, however, recognize a hereditary aristocracy, and status differences among freemen based on relative wealth are reported only for the Duala and Kpe.
Marriage invariably involves a substantial bride-price, but no other form of consideration. The Bubi and Koko favor unions with crosscousins; the Kpe and Puku forbid them. All tribes practice extensive polygyny, which, for the Kpe at least, is always nonsororal. Each wife has a separate hut, and the husband visits each in regular rotation. The household normally comprises an independent polygynous family, extended families being highly exceptional. Among the Koko and Kpe a man has two status wives, the first supplied by his father and the second normally by his mother’s brother. He divides his household between them into two sections, to which he assigns alternately the additional wives whom he obtains through his own efforts. Preferential secondary marriages of the levirate type appear to be universal, and, except for the Kpe, most tribes also practice the sororate.
The sources, though rarely satisfactory on social organization, demonstrate clearly that the area is basically matrilineal. They specifically attest matrilineal descent for all the tribes of the southern cluster, with the possible exception of the Duma. In the northern cluster, however, the influence of the adjacent Equatorial Bantu, Cameroon Highlanders, and Southern Nigerians has frequently wrought a transition to patrilineal descent. Even here, however, evidences of an ear lier matrilineate still survive. The Bubi follow the matrilineal rule in regard to succession. The Kundu and Puku show traces of matrilineal inheritance. The Duala, though patrilineal today, still adhered to matrilineal descent when first studied by German ethnographers. The Kpe retain exogamous matrilineages to the present day, along with agamous patrilineal sibs and lineages, and are thus characterized by double descent. The same might well prove to be true of the patrili neal Koko, Kundu, and Puku, or of some of them, if we possessed more detailed information about their social systems. Cousin terminology is reported only for the Duala and Kpe, who follow the Iroquois pattern.
Residence in marriage is always virilocal from the first. Although it is certainly patrilocal in all patrilineal societies, it is at least sometimes avunculocal in the matri lineal tribes, and this rule may well predominate in those groups that are reported to have localized lineages and matrilineal inheritance, e.g., the Lumbo, Mpongwe, and Shogo. On the whole, the weight of the evidence strongly suggests the derivation of the social organization from the matrilineal and avunculocal system which we have already reconstructed for the ancestral Macro-Bantu peoples of the southern igerian plateau (see Chapter 13).