On the Origin of the Indo-European Feminine Gender Category

It has been approximately a century since Brugrnann (1897) presented in detail his theory of the origin of the feminine gender category in Indo-European. In shorty he argues that one or more original abstract and collective nouns which came "to be used to denote ... [an] individual person or thing" (26) happened to denote an object with natural female sex. The formal property which characterized these nouns specifically the suffixes *>"£ * *7, and especially *-tf became morphologically reanalyzed as exponents of female sex. Subsequent analogical processes were responsible for the generalization of feminine markers to words in concord with the noun and for the ultimate grammatical, riot natural, manifestation of the^category in the individual dialects. Today this theory still retains modest acceptance. For example, Bomhard (1982: 253) characterizes his own view of the origin of the feminine as "not far differentfrom that proposed long ago by Brugmann," while Lehniann (1958,1993: 152-153) and Miranda (1975) endorse his views outright. Of course, Brugmann's proposal is now widely viewed in the context of Meillet's description (1965: 211-229) of gender in the proto-language whereby "das idg. System der drei Geschlechter muss aus einem Zweiklassensystem entstanden sein. Darauf weist allein schon die Tatsache, däss in altertümlichen Flexionsklassen das M und Fern, sich in der Flexion nicht unterscheiden, dagegen sich gemeinsam vom Neutrum abheben; vgl. z.B. pater^ meter" (Szemerenyi 1980: 143). As Burrow (1973: 202) observes, "This state of affairs is faithfully reflected in Hittite, which is .--distinguished from all other IE languages by the absence of a*special feminine gender.'* Still, Brugmann's hypothesis is clearly not without its problems, Fodor (1957: 19) points out that Brugmann's sketchy analysis of the extension of feminine gender markers to the variable word classes in agreement with nouns represents a "serious shortcoming," "for had the motion of variable word classes not come about, then those [purely nominal] suffixes would in themselves have been insufficient to bring about the category of [feminine] gender. ... The clue to the problem [of the origin of the feininiiie gender] lies hidden not