Crosswords Sudoku and Comics
Science

Isolated Populations Develop More Diverse Languages as Their Gene Pools Shrink

A University of Zurich-led study published in PNAS analyzed global genetic and linguistic datasets to reveal an inverse relationship between the two measures of human diversity.

Slides for "Non-dominant language Wikipedias: Lessons from the Russian Knot" presentation at Celtic Knot 2020 Conference
Slides for "Non-dominant language Wikipedias: Les…      Papua New Guinea Tribal Languages Map    Frhdkazan / Wikimedia Commons (CC BY-SA 4.0)
By Free News Press Editorial Team
Published May 5, 2026 at 8:15 PM PDT

Regions where people have mixed freely across generations tend to be genetically diverse. Their languages, it turns out, tend to cluster closer together. The relationship flips in isolated populations, where limited gene flow has narrowed the genetic pool while languages have been free to diverge in wildly different directions.

That counterintuitive pattern is the central finding of a new international study led by researchers at the University of Zurich, published in the Proceedings of the National Academy of Sciences.

The research team combined large-scale genetic datasets with structural linguistic data, analyzing how genetic variation among individuals correlates with variation in language structure across regions worldwide. They controlled for a range of confounding factors including geographic proximity, population density, environmental conditions, and deep population history such as the timing when humans first settled different continents.

After all those controls, the inverse relationship held across the globe.

"We were struck by how robust this inverse relationship is across the globe," said Anna Graff, the study's lead author and a linguist at the University of Zurich. "Places where people have mixed more tend to be genetically diverse, but their languages are structurally more similar. In contrast, places with long-term isolation show less genetic diversity, yet much greater diversity in how languages are structured."

The explanation, according to the researchers, lies in what contact and isolation each do to genes versus language. When populations mix, genetic material combines and diversifies. But linguistic contact tends to smooth out structural differences, as speakers borrow features from neighboring languages and over time converge toward more shared forms. Isolation cuts both processes off, but in opposite directions: the gene pool narrows while each language community is left to evolve on its own trajectory, accumulating distinct features with no outside influence pulling it toward similarity.

"Contact increases genetic diversity, but it also promotes the spread of linguistic features, making languages more similar. Isolation, by contrast, limits genetic diversity while allowing languages to evolve independently," said Chiara Barbieri, senior author and population geneticist at the University of Cagliari.

The findings help explain why certain regions stand out as hotspots of linguistic diversity. New Guinea and the Himalayan region are both relatively isolated genetically, and both host extraordinary concentrations of structurally distinct languages. In New Guinea alone, hundreds of languages are spoken across a relatively small geographic area, many of them unrelated to one another in structure or vocabulary.

Senior author Balthasar Bickel, also of the University of Zurich, described these hotspots as windows into what human language can become when it develops without outside interference.

The study suggests that the same demographic forces that shaped the genetic makeup of human populations across deep history also shaped the structure of the languages those populations spoke, but running in opposite directions depending on whether those populations were meeting or separating.

Official Journal of the European Union - L_202302424 of 31 October 2023 - English edition
Official Journal of the European Union - L_202302…      Papua New Guinea Tribal Languages Map    Publications Office of the European Union / Wikimedia Commons (Public domain)