Nous n'avons certes pas la prétention de donner ici une solution définitive ou complète au problème des origines de nos tribus malgaches. Madagascar, comme chacun sait, abonde en énigmes, et il faudra sans doute encore bien du temps, bien des recherches, bien des tâtonnements pour y répondre. Notre intention n'est que, pour ainsi parler, de faire le point soit des connaissances déjà acquises, soit des probabilités déjà entrevues au sujet du peuplement de la Grande-Ile.

Until comparatively recently the problem of the use of an objective methodology in field research did not sufficiently exercise the minds of research men in the social sciences. The field man usually gathered what he thought he observed, and generalized his collected facts into preconceived categories. Many of the recorded data we depend upon at the present time are of this quality. To-day the exercise of a technique is too frequently an implicit and uncritical part of the theoretical concepts of the researcher. Obviously a methodology must have an ideological context which motivates it, but it is highly necessary to separate the theoretical concepts from it, if we are to obtain the objectivity of the natural sciences.

Sometimes I have an uneasy feeling after attending meetings or conferences connected with well-meaning, but not always wisely directed, efforts to better the lot of the African child. We good-intentioned folk would be well advised to find out something more about the system and ideas which African parents themselves practise or hold regarding this subject before we rush in to criticize, interfere with, or offer advice to those who in the long run are mainly responsible—namely, the childrens’ own parents. Too often, I fear, our attitude is to treat these parents as being themselves but poor, benighted, heathen children.

1. Eima kiandeni emanyifwa ni mofi.

Lit. He who digs in the hollow is known (i.e. betrayed) by the smoke, i.e. A man may dig quietly for a month without any one knowing much about it, but sooner or later he will have to burn his weeds, then the smoke will rise and fill the valley, and the secret be out. ‘Things done in secret shall be revealed openly.’

In his recent writings on the subject of marriage and kinship, Malinowski has repeatedly emphasized what he terms the ‘principle legitimacy’. By this he means the rule, found in all human societies, that a woman has to be married before she is allowed legitimately to conceive. ‘Roughly speaking, an unmarried mother is under a ban, a fatherless child is a bastard. This is by no means only a European or Christian prejudice; it is the attitude found amongst most barbarous and savage peoples as well.’ Where prenuptial intercourse is regarded as illicit and immoral, marriage is obviously the essential prelude to the birth of legitimate children, i.e. children having full social status in the community. But even where prenuptial intercourse is tolerated, this tolerance does not extend to liberty of conception. The unmarried boys and girls may indulge freely in sex, but there must be no issue. An unmarried mother will be subjected to punishment and become the object of scorn, her child possibly killed or aborted, while often the putative father is also penalized unless he marries the girl. Almost universally, a child born out of wedlock has a different status from the legitimate offspring, usually very much to his disadvantage. Facts such as these show that the group of mother and child is considered incomplete in the eyes of the community, and that the sociological position of husband and father is everywhere felt to be indispensable.

La petite tribu des Batchopi (phon: Bat∫opi) occupe le territoire de l'Afrique orientale portugaise compris entre le 34e et le 35e degrés de longitude est (Greenwich) et le 24e et 25e degrés de latitude sud. Autrefois beaucoup plus nombreuse, établie sur un beaucoup plus vaste pays, cette tribu a été décimée par les guerres d'extermination que les potentats Bangoni, Mouzila et Goungounyana entreprirent contre elle. En outre, les Batchopi se sont peu à peu différenciés des Vatonga (sans ‘h’) d'Inhambane ou Vakoka, leurs frères par le sang et la langue. La proche parenté de ces deux groupes, on peut même dire leur commune origine, est facile à prouver: en effet, ils possèdent maintenant encore un grand nombre de noms de famille communs (Nyamu∫we, Nyaphosi, Nkome, Nkumbe, etc.), et le vocabulaire des deux langages respectifs montre un fond commun considérable (je ne note ici qu'un fait particulier: le mot ‘pierre’ se trouve être le même, en tchopi t∫iwindi, en gitonga giwindi, et ce mot est tout à fait différent de celui que l'on rencontre dans les autres langues bantoues).