Five thousand years ago, marauding waves of nomadic horsemen swept out of the bleak Caspian steppes across Eurasia. The Kurgans, as they’re now called, were a warring culture who imposed their leadership and language on large swathes of people. From its violent origins, their tongue would give rise to the world’s most successful language family—at least according to one theory among linguists.

Proto-Indo-European was the ancestral language that, over millennia, branched into hundreds of languages spoken by 45 percent of the world’s population, including English, Hindi, Russian, and Urdu. The location of its birth has been the subject of fiery debate for centuries.

A fascinating and contentious new study in Science supports an origins theory that couldn’t be more different from the Kurgan hypothesis. In this alternate scenario, the proto-language spread and evolved not through conquest, but rather with the gradual, peaceful expansion of agriculture out of present-day Turkey.

The research team reached this conclusion through the innovative use of computational models borrowed from evolutionary biology. Languages, like DNA, mutate at measurable rates. If you can trace the evolution of similar words across different languages, you should be able to project backwards to identify their points of divergence and ultimate origins.

The researchers analyzed a list of 200 words common to 103 languages—words like sun (Sonne in German, sól in Old Norse, sure in Kashmiri) and mother (màthair in Gaelic, mati in Slovenian, ma in Bengali). In every model they ran, the evidence pointed back to what is now Turkey, where farmers some 8,000 years ago may have radiated outward in search of new farmland.

Despite the intriguing evidence, many specialists question the methodology. Whether our linguistic predecessors were warriors or farmers may never be certain, but creative models have brought new life to an age-old question.