Phylogenetic analyses for the origin of sortal classifiers in Mongolic, Tungusic, and Turkic languages

Abstract

AbstractNumeral classifiers are one of the most common types of nominal classification systems. Their geographical distribution worldwide is concentrated in Asia, which infers a scheme of diffusion from a linguistic innovation. This study investigates the origin of classifier systems in the Mongolic, Tungusic, and Turkic languages in the Altaic region with a phylogenetic analysis based on data from 55 languages. The Single Origin Hypothesis suggests that Sinitic is the most probable original source of classifier systems found in Asia (Her & Li 2023). Under this hypothesis, classifiers are unlikely to be an indigenous feature of the Altaic region, and indeed their phylogenetic signal turns out to be weak. We also conduct a qualitative analysis on the classifier inventory of the studied languages to assess the robustness of phylogenetic methods. The results also indicate that classifiers are most likely a borrowed feature in the Mongolic, Tungusic, and Turkic languages.