LANGUAGES Tanganyika, the river Chambezi, and the districts round the north Upper Congo, is seemingly related to the Manyema, but is less, The Lower Kasai would appear to.be occupied by (23) end of Lake Nyasa. Another language of typical Bantu construc- corrupted. group which may be called, from one of its dialects, the Kibuma. tion is the Ki-yanzi or Ki-bangi tongue, spoken on a section of the aThese languages march with the Kongo dialects to the west. Hie Central Congo and on the Lower Ubangi river ; another is the wellpart of the Central Congo, on both banks, and the Lower known Oci-herero of Damaraland. Ki-makonde, on the east coast greater are occupied by peoples speaking (24) the Ki-bangi (or of Africa, is archaic ; so in some features is the celebrated Ki- Ubangi language. West ol that, and about Stanley Pool and swahili of Zanzibar, and the Ci-bodzo of the Zambezi delta; while Ki-yanzi) interior of French Congo, is (25) the Kiteke group, which, with the celebrated Zulu-Kafir dialects, though they exhibit several the allied languages, marches on the west with the Kongo dialects/ marked changes in vocabulary and phonetics (both probably of its on the north-west with (26) the Gabun or Mpongwe family. recent date), preserve nevertheless so many of the characteristics of and of the Mpongwe group (which ends about the Gabun river) the hypothetical mother tongue that, until the languages of the North (27) the Cameroon group, which includes all the coast dialects threat lakes came to be known, Zulu-Kafir was regarded as the San- is to the Rio del Rey (where the Bantu territory ends), and also skrit of the Bantu. Perhaps it is so, as Sanskrit is now viewed in up dialects of the island of Fernando Po. In the interior of relation to the original Aryan tongue. In that case the lake the and of the Gabun there is a wide tract of country indialects would correspond to Lithuanian (only “more so,” that is Cameroon by tribes such as the Fan, which are evidently recent to say, even nearer to the original Bantu than are the Lithuanian habited and can only be regarded as semi-Bantu. Ihe general dialects to the hypothetical Aryan parent speech), the Oci-herero intruders, structure of these languages {Mbudikum, Mfut, Bayoh, Bamom, to Greek, and Ki-yanzi to Gothic. PahhwC‘{Y‘A), Papiax, Param, Penin, Balu, Nki, Melon, Momenya, With our present inadequate knowledge of the affinities of the Juka, Kamuku, Basa, Juku, even the Mici south of the existing Bantu tongues it is difficult and premature to divide them Bagba, Benue) is essentially Bantu, but of a very corrupt kind. But accurately into groups and subdivisions, the enumeration of which although the shattered prefixes are of obviously Bantu origin, only also in an article like the present becomes too much of a wearisome a very small part of tli6 vocabulary can be called Eantu, and in string of names ; but there may be mentioned severally (1) the many important root-words the Bantu connexion is not at all Uganda group of languages, occupying the area between Albert obvious. These languages seem to be of a degraded type, and Kyanza and the White Nile, the north and west coasts. of possibly are very remotely descended from cognate dialects of the Victoria Nyanza, the north and north-west coasts of Tanganyika original mother language. Eastward of this intrusive wedge of semiand the Congo forests (a clearly-marked congeries of kindred dia- Bantu we again come to Bantu languages to the north of the Congo,, lects) ; (2) the Ki-emba or Ci-wemba, between Tanganyika, Bang- within the limits of the Congo basin, the Sanga group (28) of diaweulu, and Nvasa ; (3) the Nyamwezi group, south of Victoria lects spoken by the Bonjo, the Baya, the Ndere, and the. a.ngere, Nyanza ; (4) the Swahili; (5) the Nyika, and (6) the Kikamba in the basin of the river Sanga and its affluents. Immediately east groups, on the east and north-east; (7) the Ci-hehe group, south of this Sanga group, although the people would appear to be Bantu, of the Nyamwezi and north of the Ruvuma river ; (8) the well- we know" too little about their dialects to be able to classify them marked Makonde- Yao, to the south of the Ruvuma, and south and in groups. Farther east still, in the region between the Welle and east of the Yao ; (9) the remarkable Makua tongue (which though Albert Nyanza, we come to (29) the Bavira, the Wazamboni, very archaic in the roots of its vocabulary is, like the Secuana, and other tribes encountered by Stanley in his march to the reliet singularly corrupt in phonetics). Southern and Eastern Nyasaland of Emin Pasha. These inhabitants of the north-easternmost corner and Central Zamhezia are filled (10) with various Ci-nyanja or of the Congo forests and the adjoining grass country up to the Manama dialects. Immediately south of the Central and Lower southern half of the western shore of Albert Nyanza bring us Zambezi (putting aside the intrusive Zulu) is (11) the Makaraka to the Uganda group of languages from wffiich we started, and group. South and west of that again is (12) the Secuana,1 spoken back this group the dialects last referred to would seem to be related. over a wide area in closely-allied dialects. . (The Secuana, like the to This summary of the geographical grouping of the Bantu lanMakua, though truly Bantu in its roots, is. extremely corrupt in guages may concluded by again directing attention to. the7 strikits phonetics.) It has not, however, except in one or two dialects, ing contrastbewhich exists, from a linguistic point of view , north borrowed any “ clicks ” from the Hottentots. south of the Bantu border line—a border line which, comTo the south and east of the Secuana group lies (13) the im- and on the west coast of Africa close to the Niger delta, portant Zulu-Kafir section, which with cognate dialects stretches mencing following more or less closely the limits of the Northern alon02* the coast regions from Port Elizabeth in the south to Inham- and watershed to Albert Nyanza, the White Nile, and Victoria bane in the north. The dialects in the coast region between In- Congo thence continues in an irregular line to the vicinity ot hambane and the Zambezi share their alliance between the Zulu, Nyanza, North of this line there is great linguistic confusion, Makaraka, and Nyanja groups. The word “Kafir” has been so Zanzibar. an almost uncounted number of separate linguistic families long misapplied to the most southerly of the Bantu peoples that it and exists, each differing from the other as they differ in turn from is difficult to ignore it. It is simply the cant term of “infidel This is the condition of affairs throughout all N egro Africa applied by the Arabs of Sofala to the heathen Bantu round them, Bantu Abyssinia and the Atlantic. Once the Bantu border line and communicated by the Arabs to the earliest Portuguese navi- between crossed, however, there is but one family of languages —the gators who visited these shores. The Portuguese applied the teim isBantu—prevailing, with the possible exception of relics oi a 1 yguiy Cafres to all the east coast negroes from Natal to Quillimane. language in the Congo forests, of Hottentot or Nilotic fragments in North-west of the Secuana is (14) the Oci-herero group of Damara- the Zanzibar Hinterland, and, of course, the Hottentot and Bushland, Ovamboland, and Portuguese West Africa as far north as man dialects in South-West The general homogeneity ot Benguela and the highlands of Bihe. Herero is by far the most the Bantu languages is greaterAfrica. that of the existing condition archaic of this section of the Bantu languages, the dialects related of Aryan tongues, though it is than great exaggeration to say that to it in the north and north-east being much more corrupt. Ihe the various Bantu dialects differafrom Central Zambezi region, west of the great Nyanja congeries of dia- do the existing Romance languages. one another no more than , lects, is mainly occupied by (15) the Tonga or Toka family, and The following propositions may be laid down to define tne west of this again is (16) the Kiloi language-group of Barotseland, peculiar features of the Bantu languages
- —
, , . which is related to (17) the Kiboko section of the Upper Zambezi 1. They are agglutinative in their construction, their syntax being and Lower Kasai. Then to the west conies (18) the great Bunda formed adding prefixes and suffixes to the root, but no infixes (that group, including most of the dialects of Angola and ot the Lower is to say,by no syllable incorporated into the root-word). Kwango. These shade by insensible gradations into (19) the Kongo 2. The root {excepting its terminal vowel) is unchanging to alt family° on the Lower Congo from Stanley Pool to the sea, and ex- intents and purposes, though its first or penultimate vowel or contending<rnorth along the coast to Loango. The southern basin of sonant may be modified in pronunciation by the preceding prefix or the Con o is mainly occupied by one great family.(20), the Rua or succeeding suffix. one exception there is no inflexion; that Lunda dialects, which stretch eastwards to the vicinity of Tanga- exception (scarcely inWith origin a true one) is in the preterite tense of the nyika, and even seem to include the Ki-lungu and Ki-manibwe of verb in certain languages where the root changes m its termination, the Nyasa-Tanganyika Plateau. It is not known yet whether the by the absorption of a suffix. . ? lan"ua(re of the Tu-shilange and Batetela is related to the Kuo. probably Z No two consonants come together without an intervening vowel On°the Upper Congo, from the vicinity of West Tanganyika (except where one of them is a nasal, a labial, or a semi-vowel); no and Nyangwe, down past Stanley Pool to the confluence of the consonant is doubled (except by the accidental juxtaposition of two Aruwimi there seems to be one dominating form of the Bantu ms or n’s, one of which represents an abbreviated particle); no wora language which may be classified as (21) the Manyema group, and ever ends in a consonant, except in rare instances where the terminawhich exhibits very great corruption. The Balolo group (22) to tion through contraction and the dropping of a vowel becomes a nasa the west, which occupies the regions immediately south ot the b °. Substantives are divided into many classes or genders indicated 1 In accordance with the phonetic system adapted from Lepsius, in by the pronominal particle prefixed to the root. Some 0fthes*W' which Bantu languages are transcribed, c stands for the English ch. fixes are used in a plural sense, others in the singular No migular Secuana is the language spoken by the Bechuana can be used as a plural, nor can a plural prefix be employed in 2 This is the Portuguese town and district, which should really be prefix the singular number. There is a certain degree of correspondence pronounced Inyanibane—Portuguese nh ny.
140
BANTU
