In India the Brahmins of the middle class often have one chief wife, and at the same time several domestic concubines.[487]

We have seen that in Homeric Greece the concubinate was a general practice, and in no way censured. In later times, when Greece was more civilised, the primitive domestic concubinate disappeared, but there always remained to alleviate the ennui of monogamic marriage what we call concubinage, or hetaïrism, which was openly practised by Socrates and Pericles. “If,” says Lecky on this subject, “we could imagine a Bossuet or a Fénelon figuring among the followers of Ninon de Lenclos, and publicly giving her counsel on the subject of her professional duties and the means of securing adorers, this would be hardly less strange than the relation which really existed between Socrates and the courtesan Theodota.”[488]

All societies which have had any legal form of marriage have adopted the concubinate, either free or more or less regulated, but it has nowhere been so precisely legalised as in ancient Rome. I shall say a few words about it, not that I intend to walk in the steps of our legists, but in order to show what assistance ethnographical sociology could be to the science of written law. By its means alone can the legal texts, which have been a hundred times studied, commented on, and criticised in an isolated manner, as if they related to sociological facts without analogy in the world, be connected with the general evolution of customs and institutions.

At the bottom, the Roman concubinate is essentially similar to the others; it has merely been legalised with more care, and transformed into an institution as regular as marriage proper. It was, besides, indispensable in a country where the right of marriage, the jus connubii, was restricted. The leges Julia and Papia Poppæa also expressly authorise it.

In short, the Roman concubinate was a free union between a man and a woman not wishing, or not being able, to marry.[489] It was lawful to have as concubine a woman with whom marriage was forbidden—an adulteress, an actress, a woman of bad life, or a freed slave. This last case was the most frequent, most moral, and the most protected by the laws.

The intention of the parties, revealed either by a formal declaration, or by the inequality of conditions, determined between marriage and the concubinate. The dowry was one of the signs which served to distinguish marriage from the concubinate.

The Roman concubinate was only, in fact, a marriage of inferior degree.[490] Thus a married man could not take a concubine. A bachelor could not have several at the same time.[491]

The concubinate implied paternity. The child was considered as a natural child of the father (naturalis, non vulgo conceptus), though he did not enter the father’s family or become his heir, but followed the status of his mother.[492]

The institution of the Roman concubinate evolved naturally, and its conditions were more and more ameliorated.

Under Constantine, the legitimation of children born from a concubinate was permitted in a general way by marriage between the father and the woman who had been his concubine up to the day of marriage. It was necessary, however, that the man should not have at the time a legitimate child. But Justinian authorised the legitimation even in this last case; he granted also the benefit of legitimation to the children of an enfranchised slave marrying her master, provided that the latter had not then any legitimate children.[493]