Teeth and Tongues; Walkin' and Talkin'
Paul Newman

The bedrock of Greenberg's African linguistic classification -- and the key to its success -- was the notion that such a classification must be based exclusively on linguistic evidence and that one must disavow recourse to extraneous factors relating to race and culture. By contrast, in marshaling support for his American Indian classification, Greenberg often cited corroborating evidence provided by physical anthropologists regarding human dentition and by archeologists regarding the settlement of the New World. At first sight, it would appear that Greenberg had abandoned his ideological principles for opportunistic purposes. But on closer analysis, one finds that there is no contradiction between his earlier and later positions. Rather, Greenberg always had an anthropological perspective in his approach to language and he was always looking for ways in which language classification and language contact could tell us about the history and movement of peoples and about prehistoric contacts between populations. In this paper, I shall review Greenberg's ideas regarding the Bantu expansion (in Africa) and the Amerind migration (in the New World) and show how legitimate historical inferences can be drawn from linguistic data while avoiding the ever-present dangers inherent in forsaking the fundamental Boasian postulate regarding the independence of language, race, and culture.