JOHANNA NICHOLS is professor in the Slavic Department at UC Berkeley. Her research interests include the Slavic languages, the linguistic prehistory of northern Eurasia, language typology, ancient linguistic prehistory, and languages of the Caucasus, chiefly Chechen and Ingush. She is the author of 'Linguistic Diversity in Space and Time" (Chicago, 1992).
Tracing slow-changing features of grammatical and lexical structure across a sizable sample of languages representing the world's language families and all geographical areas has proven to be the only means of tracing linguistic prehistory back over 10,000 years. Some of the successes of this method will be demonstrated; they are beginning to give us a spare but reliable picture of the composition and distribution of the world's linguistic population beginning before the time, roughly 50,000 bp, when humans (and their languages) expanded out of the Old World tropics to colonize the Pacific, the New World, and the high latitudes, and they give us a more detailed picture of origins and migrations going back to the waning of the last glaciation (c. 20,000 bp).
This method is only as good as the language sample one can draw, and several examples will be given where extinction or inadequate description of little-known languages creates a (potential or actual) gap of major proportions in our overall understanding of the origin and dispersal of human language. In addition, the method is only as good as the structural features surveyed and the definitions making it possible to code them consistently for diverse and diversely described languages; and this in turn is only as good as linguistic theory and descriptive practice. Several of the most reliable and most revealing structural comparanda could not have been surveyed before theoretical linguistic advances of the last few years, and there is every reason to anticipate comparably important theoretical discoveries in future years. For many languages that were superbly described decades ago and have since gone extinct or undergone massive anglicization, russification, etc., needed information is irretrievably lost. An outline program will be presented for fostering informed prescriptivism that will enable speakers of minority languages to maintain their languages in the natural state (including natural evolution and change) without sacrificing access to (and even bilingualism in) major world languages like English; and informed descriptivism among linguists that will hasten the pace of descriptively useful theoretical advances.