How people’s genes are revealing why some languages are so unique
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
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.
The BriefWire