The Concept of Evolution in Linguistics

Abstract

Language has been viewed for a long time as a field of evolution where new things originate through variation. In the early nineteenth century, important insights were reached within this still pre-Darwinian tradition by historical-comparative linguistics. The laws of sound change and vowel shift discovered by Grimm and others are essentially valid today, although development by then was largely considered as loss and deprivation. A strictly Darwinian view of language change was advocated in 1863 by Schleicher, who considered languages as organisms to which biological principles of adaptation and survival were to be applied directly. The difficulty to reconcile insights of language development with principles of biological evolution was the requirement to construe languages as species, the genetic information of which is subject to random variation and completely independent of reproduction of their members. On closer inspection, the range of possible language variation is provided by the human language faculty, and the domain of competition for reproduction is the actual use of language. This leads to two competing views of language change. A Lamarckian view considers variants as the product of functional needs, subsequently inherited by the language used, while a strictly Darwinian view considers variants as random products of the language faculty which do survive or not according to conditions of language use. Although in many, notably lexical, cases the functional origin of variants is beyond doubt, the Darwinian nature of much language change is still attested by properties that cannot be due to functional requirements, but emerge from traits of the language capacity and are not suppressed by functional needs. Language change must thus be considered as evolution of mixed nature. A short epilogue raises the question whether this type of mixed evolution could be the general nature of human history and cultural development. Skepticism is indicated for two reasons. First, the language faculty is domain specific, its conditions cannot be generalized to arbitrary fields. Second, there is no reason to assume a universal cultural capacity from which action patterns randomly arise.