Phylogenetic Methods in Historical Linguistics: Greek as a Case Study

Abstract

AbstractWe review and assess the different ways in which research in evolutionary-theory-inspired biology has influenced research in historical linguistics, and then focus on an evolutionary-theory inspired claim for language change made by Pagel et al. (2007). They report that the more Swadesh-list lexemes are used, the less likely they are to change across 87 Indo-European languages, and posit that frequency-of-use of a lexical item is a separate and general mechanism of language change. We test a corollary of this conclusion, namely that current frequency-of-use should predict the amount of change within individual languages through time. We devise a scale of lexical change that recognizes sound change, analogical change and lexical replacement and apply it to cognate pairs on the Swadesh list between Homeric and Modern Greek. Current frequency-of-use only weakly predicts the amount of change within the history of Greek, but amount of change does predict the number of forms across Indo-European. Given that current frequency-of-use and past frequency-of-use may be only weakly correlated for many Swadesh-list lexemes, and given previous research that shows that frequency-of-use can both hinder and facilitate lexical change, we conclude that it is premature to claim that a new mechanism of language change has been discovered. However, we call for more in-depth comparative study of general mechanisms of language change, including further tests of the frequency-of-use hypothesis.