Hindi and Tamil differ in far more than just vocabulary. Languages differ in how they are organised, and these differences are not spread evenly around the world. In some regions, neighbouring languages like Basque and Spanish differ dramatically from one another. In others, they share many of the same features, like Tamil and Kannada.
A new study by researchers from Europe and Japan has found human history itself gives shape to the pattern. After analysing thousands of languages alongside genetic data from populations around the world, the researchers found that places that were relatively isolated for a long time tended to harbour more linguistic features. On the other hand, regions shaped by migration and sustained contact contained languages that were more alike.
“I was struck by the clarity of the signal,” Anna Graff, lead author of the study and a biological anthropologist at the University of Zurich, said. “In studies of human history and diversity, such clear global patterns are often difficult to find.”
A feature checklist
Crucially, the researchers did not count the number of languages in a region. Instead, they measured how different neighbouring languages were from one another. Some languages placed verbs near the beginning of a sentence while others placed them at the end. Some distinguished between a hand and a finger with separate words while others used the same term for both.
A region where languages had more such differences was considered more linguistically diverse than one where languages shared many of the same traits — even if both regions had the same number of languages.
To investigate, the researchers assembled one of the largest datasets of its kind, combining information from more than 4,200 languages with genetic data from over 5,700 individuals representing 650 populations around the world. Then they divided the world into hundreds of roughly 500-km-wide hexagonal cells and calculated two measures for each: diversity of linguistic features found among languages spoken there and the genetic diversity of the local populations.
Thus they found that regions whose populations had been relatively isolated over long periods tended to have more diverse linguistic features.
The effect was modest. Imagine describing each language using a checklist of 333 characteristics. The difference associated with isolation was roughly equivalent to making about 11 items on the checklist vary substantially across the languages spoken in a region.
Yet the pattern proved remarkably persistent, appearing again and again, across statistical tests.
‘Accumulation zones’
The researchers also tested if the pattern could be explained by other obvious factors. Languages spoken close together often influence one another while related languages can resemble one another because they inherited features from a common ancestor. Yet neither explanation fully accounted for the trend.
At first glance, the findings also seem counterintuitive. One might expect places with a long history of migration and population mixing to have more language patterns. But Dr. Graff and her colleagues argued that contact often has the opposite effect. As people move, trade, and interact, languages can borrow words, sounds, and grammatical features from one another, gradually becoming more alike.
Isolation allows languages to evolve independently, increasing the chances that they will grow more different over time.
They described these differences by classifying a place as a “spread zone” and or an “accumulation zone”. Spread zones are repeatedly reshaped by large population movements associated with the spread of farming, state expansion, empire-building, and, more recently, colonialism.
New Guinea, home to more than 800 languages, is a good example of an accumulation zone. The island’s languages are remarkably varied in their linguistic features. According to the authors, its long history of relative isolation allowed these differences to accumulate rather than become smoothed out by repeated migration and contact.
Linguistic fossils
The significance of such regions goes beyond explaining present-day diversity. The authors argued that linguistic hotspots may offer rare glimpses of how human languages became more varied before large migrations and population movements reshaped much of the world.
In that sense, places such as New Guinea may preserve traces of a linguistic world that existed before waves of expansion and migration reduced diversity across regions. That is, having been less affected by forces that make languages more alike, they may preserve ways of organising language that have disappeared elsewhere.
That makes the loss of small languages especially consequential. Languages spoken by small and relatively isolated communities are “important windows into the diversity and breadth of linguistic structures we humans are able to learn, transmit and process,” Dr. Graff said. “At the very least, documenting them is essential if we want to preserve a record of the full range of linguistic variation.”
Signatures on languages
“Patterns in genetic diversity certainly do capture history very well at both broad and fine scales,” Iain Mathieson, professor of genetics at the University of Pennsylvania Perelman School of Medicine, said. “Very naturally, one would expect those patterns also to correlate with features of linguistic and cultural diversity at least on average, and this interesting study shows exactly that.”
The findings do not imply, however, that genes determine language. The historical movement of people can leave traces in both genetic and linguistic patterns but the two do not always travel together.
The study’s central argument also resonates with observations from the field. Patience Epps, professor of linguistics at the University of Texas at Austin who studies indigenous languages of Amazonia, said the findings align with observations from her own field.
In the Upper Rio Negro region, where Dr. Epps conducts fieldwork, communities often intermarry and maintain extensive contact across language groups. She noted that Amazonia still lacks the detailed genetic datasets required for a systematic test of the study’s conclusions.
The authors also argued that such local processes, repeated across centuries and larger regions, can leave broader signatures on patterns of language diversity around the world.
(Anirban Mukhopadhyay is a geneticist by training and science communicator from New Delhi.)
Published – July 14, 2026 09:15 am IST
