For the development of grammatical sex the Badauye language affords important indications. That the "feminine" is really a things class and not connected with sex appears from the following remarkable facts. [Reference will first be made to Reinisch, who discusses the question and says:]

"The distinction of the two sexes certainly originated in sex, but since in the Bedauye, as in all Hamitic languages, there are no neuter nouns but all substantives are of either the masculine or the feminine gender, the masculine, in the present state of the language, expresses size, distinction, energy, in addition to sex, while the feminine expresses mainly smallness, weakness, and passiveness. Thus, for example sa, ’cow,’ is masculine because, as is well known, it is the main support of the household economy, while sa, ’meat,’ is feminine because in comparison with sa, ’cow,’ it is of less importance."

In my opinion this shows, on the contrary, that the distinction is not based on sex, for in that case the cow because of her milk, that is, a specific female quality, would not have been of the masculine gender. She is, however, as Reinisch emphasizes, the basis of existence among a shepherd people, and just as in the Fulde language important animals are sometimes placed in the human class so here this animal is placed in the human class, that is, the "masculine." The flesh of this animal is a thing and naturally goes with the things class. This agress with the fact that ando, "the excrement of animals," is masculine when the reference is to ox, horse, camel, but feminine when the reference is to small animals.¹

The Hamitic languages in their present condition are therefore peculiar in that presumably there has been a historical change, an original animate-inanimate shifting to a masculine-feminine classification through an identification of certain qualities with male and female. It is noticeable also that where (as in American languages, with few exceptions) animate-inanimate are the basis of categories a discrimination of superior or inferior qualities may promote or demote objects from one class to the other. In Algonkian "a rigid classification of the objective world into things animate and things inanimate underlies the whole structure of the language. . . . Every verb and every noun must fall into one or the other class." Plants fall in the inanimate class but a number of important ones, such as corn, tobacco, and apples, are classified as animate, while little animals are often placed in the class of lifeless things.¹ "The Iroquois distinguish strictly between nouns designating men and other nouns. The latter may again be subdivided into a definite and indefinite group. The Uchee distinguish between members of the tribe and other human beings."² In Bantu the prefix iki denotes a class of things but it is used also to place the blind, the deaf, cripples, and simpletons in the things class.³ In Hupa, an Athapascan dialect, one prefix is used when speaking of adult Hupa, another when speaking of Hupa children, and sometimes of very old people, of the members of other tribes and races, and of animals.⁴

In Bantu a distinction is made between persons being or acting in their own capacity and those having a subordinate or representative rol– e. Umu prefixed to a verb designates such conditions as messenger, secondary wife, widow, prisoner, unmarried man, etc.⁵

¹Meinhof, C.n/an/an/an/an/a, , 139–140 (L. Friederichsen and Co. By permission).

¹ Jones, op. cit., 761.

² Boas, Handbook . . . , 1: 36.

³ Meinhof, C. Grudzüge einer vergleichenden Grammatik der Bantusprachen, 13.

⁴ Goddard, op. cit., 1: 117.

⁵ Meinhof, Grundzüge . . . , 6.