Graphemic Analysis and the Spoken Language Bias.

Abstract

In the history of linguistics, the investigation of written language got off on the wrong foot. For a long time, writing was regarded as a secondary medium, its raison d'etre being the recording of spoken language. Spoken language in turn was regarded as the primary object of linguistic investigation. Consequently, there was no interest in an unbiased analysis of writing; if writing was analyzed at all, it was seen through the eyes of phonology. It took decades for things to improve, but written language was eventually accepted as a linguistic object in its own right. Yet the old spoken-language-bias is still influential, for example in the call for papers to this research topic. It is (as mostly nowadays) stated implicitly, which makes it harder to tackle. The topic editors state the importance of prosody for spoken language and the lack of its explicit marking in written language. On this basis they suggest that 'the informativeness of written text may seem astonishing.' This line of reasoning could be called the phonocentric fallacy: What is important in spoken language must be important in written language; moreover, there cannot be independent structures and relations in writing mediating between graphemic form and meaning, i.e., bypassing the phonological route.