The best way to examine this first problem—in what manner the ‘family complex’ is influenced and modified by the constitution of the family in a given society—is to enter concretely into the matter, to follow up the formation of the complex in the course of typical family life, and to do it comparatively in the case of different civilizations. I do not propose here to survey all forms of human family, but shall compare in detail two types, known to me from personal observation: the patrilineal family of modern civilization, and the matrilineal family of certain island communities in North-Western Melanesia. These two cases, however, represent perhaps the two most radically different types of family known to sociological observation, and will thus serve our purpose well. A few words will be necessary to introduce the Trobriand Islanders of North-Eastern New Guinea (or North-Western Melanesia) who will form the other term of our comparison, besides our own culture.

These natives are matrilineal, that is, they live in a social order in which kinship is reckoned through the mother only, and succession and inheritance descend in the female line. This means that the boy or girl belongs to the mother's family, clan and community: the boy succeeds to the dignities and social position of the mother's brother, and it is not from the father but from the maternal uncle or maternal aunt, respectively, that a child inherits its possessions.

Every man and woman in the Trobriands settles down eventually to matrimony, after a period of sexual play in childhood, followed by general licence in adolescence, and later by a time when the lovers live together in a more permanent intrigue, sharing with two or three other couples a communal ‘bachelor's house’. Matrimony, which is usually monogamous, except with chiefs, who have several wives, is a permanent union, involving sexual exclusiveness, a common economic existence, and an independent household. At first glance it might appear to a superficial observer to be the exact pattern of marriage among ourselves. In reality, however, it is entirely different. To begin with, the husband is not regarded as the father of the children in the sense in which we use this word; physiologically he has nothing to do with their birth, according to the ideas of the natives, who are ignorant of physical fatherhood. Children, in native belief, are inserted into the mother's womb as tiny spirits, generally by the agency of the spirit of a deceased kinswoman of the mother.[2] Her husband has then to protect and cherish the children, to ‘receive them in his arms’ when they are born, but they are not ‘his’ in the sense that he has had a share in their procreation.

The father is thus a beloved, benevolent friend, but not a recognized kinsman of the children. He is a stranger, having authority through his personal relations to the child, but not through his sociological position in the lineage. Real kinship, that is identity of substance, ‘same body’, exists only through the mother. The authority over the children is vested in the mother's brother. Now this person, owing to the strict taboo which prevents all friendly relations between brothers and sisters, can never be intimate with the mother, or therefore with her household. She recognizes his authority, and bends before him as a commoner before a chief, but there can never be tender relations between them. Her children are, however, his only heirs and successors, and he wields over them the directpotestas. At his death his worldly goods pass into their keeping, and during his lifetime he has to hand ov