In recent literature on the social structure of African peoples much attention has been paid to the role of the lineage in village and local organization. Among one Central African people at least–the Mwinilunga Ndembu of Northern Rhodesia–the spatial separation of adjacent genealogical generations is a structural principle of equal importance. Other anthropologists working in Angola and western and central Northern Rhodesia have obtained data which suggest that this principle is widespread in Central Africa.

Among the different tribes of the Kuba group, three forms of initiation rites are practised: the eastern tribes (Ngende, Ngongo, and the Shoowa) follow a common form, the Kete of the Luebo region have their own, and the Bushong of the south have yet another form.

I was initiated according to the Bushong rites at the village of Mapey in the south from 23 October to 9 November 1953. Initiation rites at the capital have fallen into disuse and, apart from the capital, may only be performed in villages which pay a special tribute to the king and are known as ‘royal villages’. The people of Mapey allowed me to be initiated because the king himself vouched for me, and because they hoped that I would refute the allegations of immorality brought by the missionaries.

I Propose to examine in this short paper, in the very limited time at my disposal, some of the ways in which a study of the Semitic languages of Ethiopia, ancient as well as modern, can and should be of use to our understanding of general Semitic linguistic phenomena and to a clearer appreciation of inter-Semitic relationships. At present we shall be concerned with the aid extended by Ethiopics to other languages rather than with the Ethiopian tongues at the receiving end, so to speak; with their active rather than their passive capacity, though I should not wish to pretend that it is always possible, or even desirable, to disentangle the two. In a minority of cases only can we detect a direct ‘influence’ of Ethiopic on other Semitic languages, i.e. impact at a time when these tongues were already fully differentiated from each other. Generally, the value of Ethiopian languages lies in their having preserved a feature lost elsewhere, or a root that has otherwise disappeared, or in revealing a parallel that may be significant to an assessment of general Semitic affinities.

Persons accustomed to dealing with languages of known and proven ancestry may have difficulty in believing that language classification can be problematic. They may be unaware that, whereas many modern languages may be classified according to historical evidence, those without a recorded past must be grouped in the light of present-day similarity or dissimilarity. None of the languages investigated by the western section of the Northern Bantu Border Survey has a history which goes back beyond last century, and many of them had never been previously recorded. Oral tradition, usually so informative on ethnic events, is generally vague, if not absolutely silent, on linguistic matters.

La nouvelle édition du dictionnaire de Kitching et Blackledge par Mulira et Ndawula fournit une base solide à l'étude comparative du Ganda. L'objet du présent article est d'exposer les règles de correspondance qui résultent de la comparaison du Ganda avec les formules du bantou commun posées par Meinhof, Homburger, Bourquin, et d'autres, et surtout de dormer des listes provisoires d'irrégularités réelles ou apparentes (voir la partie D). Celles-ci permettront à ceux qui étudient d'autres langues bantoues de rectifier certaines reconstructions, d'établir de nouvelles variantes, et de faire meilleure part aux emprunts et autres faits dus à des facteurs d'ordre géographique.