Explaining the linguistic diversity of Sahul using population models

Area of Science:

• Linguistics
• Population Genetics
• Computational Linguistics

Background:

• The Sahul continent (Australia and New Guinea) exhibits extreme linguistic diversity, with numerous unrelated language families.
• Existing classifications struggle to resolve deep relationships between these families, often exceeding the limits of standard linguistic methods.

Purpose of the Study:

• To infer the historical structural characteristics of Sahul languages using an admixture model.
• To identify ancestral language populations and reconstruct early dispersals and contact events.

Main Methods:

• Applied a Bayesian algorithm, originally developed for genetic marker analysis, to model linguistic admixture.
• Analyzed structural characteristics of languages across the Sahul region.

Main Results:

• Identified 10 distinct ancestral language populations, with some correlating to known phylogenetic groups.
• Detected evidence of early dispersals, including hypothesized links between Australian and Papuan languages.
• Observed systematic language contact effects and a significant underlying phylogenetic signal in many languages.

Conclusions:

• The admixture model successfully inferred deep linguistic history and population structure in Sahul.
• The findings provide novel evidence for ancient linguistic connections and dispersals within the region.
• Despite contact and admixture, a strong phylogenetic signal persists in many Sahul languages.