LINGUISTIC RELATIVITY REVISITED AND RENEWED

During the first half of the 20th century, linguistic relativity—not an explicit term but a common assumption in language research—prevailed well in American linguistics. It was the main characteristic of Boas’s works on American-Indian languages, Sapir’s writings on language and culture, and even Bloomfield’s “scientific linguistics.” It is easy to understand relativism in the Boasian and Sapirean traditions, since Boas and Sapir were both linguist anthropologists. Their expertise in anthropology, or their deep concern with local cultures, no doubt gave strong influence on their linguistic research.[1] Bloomfield, as best exposed by his Language (1933), was different. When he claimed that linguistics was a “scientific” discipline, he referred to “natural sciences,” not in terms of content but in terms of methodology. Bloomfield’s methodology was rigorous: the approach must be inductive, the object must be observable (language is primarily speech), and the generalizations must be empirically verifiable. The end-results of Bloomfieldian research were meticulous, detailed descriptions of individual languages. One important generalization was that “languages are different.” Relativism is indeed the hallmark of pre-Chomskyan linguistics. [1] More on Boas and Sapir will be presented in section 2 of this paper