Phylogenetic Linguistics

Computational phylogenetic linguistics borrows the statistical machinery developed in evolutionary biology — Bayesian inference, maximum likelihood, and distance-based network methods — and applies it to coded linguistic data, chiefly cognate-judged basic vocabulary, to infer language family trees and estimate when branches diverged. By treating linguistic characters like the molecular characters in a gene alignment and modelling their change probabilistically along a tree, the approach produces classifications with explicit measures of uncertainty and, when calibrated, dated phylogenies. Its best-known applications are the Gray and Atkinson and Bouckaert et al. analyses of Indo-European origins.