Statistical evidence for the Proto-Indo-European-Euskarian hypothesis

Based on a new reconstruction of Proto-Basque, and regular sound correspondences between this Proto-Basque and Proto-Indo-European as standardly reconstructed, Blevins (2018) argues that Proto-Basque and Proto-Indo-European have a common ancestor that pre-dates the two proto-languages. Part of this argument is based on proposed Proto-Indo-European/Proto-Basque cognate sets that include basic vocabulary items. In this study we offer statistical support for Blevins’ conclusions by using a Monte Carlo simulation that allows us to estimate the probability that the proposed lexical correspondences could have arisen by chance. The method makes use of phonotactic language models to generate possible words in a pair of languages, and then attempts to discover consistent correspondences between the words, producing a list of possible “cognates”. The method differs from some previous approaches by considering matches between all segments in the word pairs. By running such a simulation a large number of times, we can estimate the probability that two languages with the given phonotactics could have produced the number of cognate pairs observed in the actual data. The method is independently assessed by comparing wordlists from 100 pairs of languages, related and unrelated, where relations are known. Our conclusion is that the proposed correspondences are unlikely to have arisen by chance, supporting a distant relationship between Proto-Basque as reconstructed by Blevins (2018) and Proto-Indo-European.