Languages, evolution and statistics: human sound systems were shaped by changes in bite configuration.Response to Tarasov & Uyeda (2020)

Abstract

In Blasi et al. (2019) we have shown, through a series of statistical analyses and models, that human sound systems have been affected by a transition in bite configuration starting from the Neolithic. Tarasov and Uyeda (2020) (henceforth T&U) raise a number of observations in relation to our article. We appreciate T&U's engagement with our work and their sharing of the code and data of the analyses reported. In brief, their technical comment involves five analyses: Binomial Causal Models (BCM) 2. Linear Regression of across-area variation in labiodentals and subsistence 3. Predictive Posterior Simulations (PPS) 4. Poisson Linear Regression (PLR): model comparison 5. Phylogenetic AnalysesIn what follows, we show that the discrepancies they report between our findings and theirs are due mostly to ill-specified models, weak (or missing) statistical evidence, and a misinterpretation of our results. After these issues are addressed, we conclude that T&U's claims do not hold.