07_Kiviksgraven_Hugh_McColl_Eske_Willerslev_Photo_Nicolas_Delgado_Clausen

Rock art in the King’s Grave at Kivik (Kiviksgraven) in southern Sweden, an exceptionally large Bronze Age burial monument dating to c. 1500 BCE, coinciding with the expansion of East Scandinavian Bronze Age ancestry. From left: Hugh McColl, lead author; Eske Willerslev, senior author. Credit: Nicolas Delgado Clausen

Europe’s Language Map Was Redrawn by Migrations, Ancient DNA Suggests

In A Nutshell

  • Two Nature studies used DNA from 4,587 ancient people to trace where Germanic and Celtic languages came from.
  • A Baltic-linked group entering Sweden about 4,000 years ago likely carried early Germanic, about 800 years later than traditionally assumed.
  • Ancestry from eastern central Europe moving west about 3,200 years ago best fits the spread of Celtic.
  • Both are best-fit answers, not proof, since DNA cannot show what language people spoke.

English and Irish seem to have little in common, but two new studies in Nature suggest the roots of both languages may trace back to ancient migrations. Using DNA from 4,587 ancient Europeans, researchers found a Baltic-linked population entering Sweden about 4,000 years ago, a likely carrier of early Germanic, and a separate wave out of eastern central Europe that best fits the spread of Celtic.

Each answer comes with a catch. “DNA sequences do not tell us directly which language people spoke,” said senior author Eske Willerslev of the University of Copenhagen in a university release. That makes these best-fit explanations, not proof.

Read together, the studies describe a language map redrawn over thousands of years. Ancestry linked to Celtic’s spread moved west, and much later, Germanic-linked ancestry moved south into regions presumed to speak Celtic, including present-day Germany, Poland and the Netherlands.

Ancient DNA From 4,587 People Traced Germanic and Celtic Languages

An international team led by the University of Copenhagen and Leiden University published both studies together. Researchers sequenced 712 ancient individuals and added 4,009 genomes that other teams had already published. After removing close relatives and low-quality samples, the main dataset held 578 new individuals plus the 4,009 published ones, 4,587 in all, spanning roughly 5,000 years from the late Stone Age through the Viking Age. The Celtic study reused that dataset.

Long shared stretches of DNA, inherited from common ancestors, did the sorting. People who share many such stretches fall into clusters, which let researchers tell closely related populations apart and watch ancestry shift across the map through time.

Ales stenar
Ales stenar (Ale’s Stones) in southern Sweden, a monumental Iron Age stone ship-setting overlooking the Baltic Sea, dating to c. 600–1000 CE and broadly coinciding with the period of back-migration into the region. From left: Eske Willerslev, senior author; Hugh McColl, lead author. Credit: Nicolas Delgado Clausen

Germanic Languages Likely Arrived With a Baltic-Linked Population

Before about 2,800 years ago, Scandinavia held three genetic groups. One, labeled Eastern Scandinavian, carried extra ancestry from hunter-gatherers whose closest match was in Latvia and Lithuania, and it likely formed somewhere between Finland and northeast Sweden. It shows up in Sweden by about 4,000 years ago and was widespread across Scandinavia by about 3,500 years ago, alongside new burial customs and the first lasting bronze trade networks.

Word borrowing with Celtic and with early Finnish- and Sami-related languages indicates early Germanic was spoken between northern Germany and the eastern Baltic by the Late Bronze Age. By then, the other two Scandinavian groups had shrunk. That makes the Baltic-linked group the simplest carrier, and it places Germanic’s arrival in Scandinavia some 800 years later than traditionally assumed.

Later migrations went in several directions. Anglo-Saxons in England carried ancestry that best matched northern Germany. Then, as Old Norse took shape between roughly 1,575 and 1,200 years ago, northern German and southern Jutland ancestry moved into Denmark and southern Sweden, and on the Danish islands local ancestry had essentially vanished by about 1,250 years ago.

A Westward Migration Best Fits Celtic Languages

Celtic once stretched from Spain to the Balkans, yet scholars disagree on where it started. Four theories compete: Bell Beaker groups of western and central Europe, Atlantic sea traders, inland France, and eastern central Europe. Each was scored on direction, territory and timing, and the best theory needs the fewest extra assumptions.

Two earlier migrations fit less well. Around 4,400 years ago, ancestry from the grasslands of eastern Europe spread through Bell Beaker groups across the region where Celtic was later spoken. Between 3,800 and 3,200 years ago, people from southwestern Europe moved into Britain, a fit for the sea-trade theory. Both came before most estimates for when Proto-Celtic, Celtic’s unwritten shared ancestor, formed, which cluster between 3,600 and 2,800 years ago.

A third migration fits best. Ancestry tied to the Urnfield culture of eastern central Europe spread west from about 3,200 to 2,800 years ago, reaching France, the British Isles, northeastern Spain and Italy, and it lands in the peak window for Proto-Celtic’s formation, about 3,200 to 3,000 years ago. Cremation means few genomes survive from that period, so the dates lean on the few burials that exist plus later samples.

One complication remains. In Iron Age England, the largest share of ancestry tended to trace to older Bell Beaker groups, so Celtic would have had to spread with a small genetic footprint in Britain.

Ancient DNA Rewrites Germanic Origins
English and Irish may trace back to ancient migrations, according to two new studies of DNA from thousands of ancient Europeans. (Image by StudyFinds)

Migration Alone Does Not Explain Germanic and Celtic Languages

Both studies end on the same caution. Norway kept a largely continuous population yet still ended up speaking Old Norse, so the authors credit a mix of population movement and cultural change. Goths show it too: burials in Poland matched Scandinavian ancestry, but later Goths in Ukraine and Spain mostly carried local ancestry, which fits locals adopting Gothic language and identity. For Celtic, “language shifts need not track genetic change.” If related dialects were already spoken across the region, newcomers and locals in Britain may have understood each other, letting Celtic words and grammar spread by borrowing instead of replacement.

Europe’s language map was drawn partly by people on the move and partly by neighbors borrowing from neighbors. Until more genomes fill the gaps, a Baltic-linked arrival for Germanic and an eastern central European one for Celtic remain the leading explanations, not settled ones.


Paper Notes

Limitations

Germanic study: Cremation practices left few preserved genomes from roughly 1,575 to 1,200 years ago, so the researchers leaned on archaeological, historical and climate evidence to estimate when the population shift in southern Scandinavia took hold. No ancient genomes exist from the region and period where the Eastern Scandinavian group is thought to have formed, which leaves its exact origin unresolved. The authors note that genetic ancestry and language do not always travel together and cannot rule out Paleo-Germanic spreading through cultural contact with no genetic footprint. Gaps in sampling between northern Jutland and northern Germany limit how precisely some source regions can be pinned down, and the paper says additional sampling is needed to confirm the homelands of the Angles, Saxons and Jutes. Quality filtering cut the newly sequenced set from 712 genomes to 578 individuals in the main analysis.

Celtic study: Few genomes exist from the cremation-dominated Urnfield period (3,200 to 2,800 years ago), and samples from 3,700 to 3,300 years ago are also scarce, so which populations formed the eastern central European ancestry remains unclear. Limited sampling between the Netherlands, western Germany and northern France means the source of a smaller northern ancestry signal in France cannot be pinned down. Choosing the simplest fit is a way to compare theories, not a measure of what actually happened, and only four predefined theories were tested. The westward migration left a smaller genetic footprint in Britain and Iberia than in central regions, which the eastern central Europe theory has to explain. The authors stress that language change is not driven only by population movement, and that “spread” and “population movements” are descriptive terms that say nothing about the social mechanism behind them.

Funding and Disclosures

Germanic study: The paper acknowledges grants from the Lundbeck Foundation, the Novo Nordisk Foundation, the Wellcome Trust, the Carlsberg Foundation, the Danish National Research Foundation, the University of Copenhagen and Ferring Pharmaceuticals A/S, which support the Lundbeck Foundation GeoGenetics Centre. Individual authors also received support from the European Research Council, VILLUM FONDEN, Riksbankens Jubileumsfond, the Emil Aaltonen Foundation and the Faculty of Health and Life Sciences of Linnaeus University in Sweden, among others. The authors declare no competing interests.

Celtic study: Hugh McColl is supported by an ERC Starting Grant (NEOSEA). Guus Kroonen, John Koch, Johan Ling and Martin Sikora are supported by Riksbankens Jubileumsfond through the MARITIME ENCOUNTERS program. Kristian Kristiansen is supported by Riksbankens Jubileumsfond and by the European Research Council Synergy project COREX. The Lundbeck Foundation GeoGenetics Centre is supported by the same funders listed above. The authors thank Illumina for collaboration and declare no competing interests.

Publication Details

Germanic study: “Tracing the spread of Germanic languages using ancient genomics” was published open access in the journal Nature on November 4, 2026 (received March 6, 2024; accepted April 24, 2025). Hugh McColl of the University of Copenhagen, Guus Kroonen of Leiden University and the University of Copenhagen, Kristian Kristiansen of the University of Gothenburg and the University of Copenhagen, and Eske Willerslev of the University of Copenhagen are the corresponding authors. McColl and Willerslev led the study, which has a large international team of co-authors. DOI: https://doi.org/10.1038/s41586-025-09064-1.

Celtic study: “Tracing the spread of Celtic languages using ancient genomics” was published open access in the journal Nature on November 4, 2026 (received February 16, 2025; accepted September 22, 2026). Authors are Hugh McColl and Guus Kroonen, who contributed equally, along with Thomaz Pinotti, William Barrie, John Koch, Johan Ling, Jean-Paul Demoule, Kristian Kristiansen, Martin Sikora and Eske Willerslev. McColl, Kroonen, Kristiansen and Willerslev are the corresponding authors. Affiliations include the University of Copenhagen, Leiden University, the University of Gothenburg, the University of Cambridge, the University of Wales Centre for Advanced Welsh and Celtic Studies, and Panthéon-Sorbonne University. DOI: https://doi.org/10.1038/s41586-026-11198-9.

Sequence data for the 578 newly sequenced genomes are deposited at the European Nucleotide Archive under accession PRJEB87274.

About StudyFinds Analysis

Called "brilliant," "fantastic," and "spot on" by scientists and researchers, our acclaimed StudyFinds Analysis articles are created using an exclusive AI-based model with complete human oversight by the StudyFinds Editorial Team. For these articles, we use an unparalleled LLM process across multiple systems to analyze entire journal papers, extract data, and create accurate, accessible content. Our writing and editing team proofreads and polishes each and every article before publishing. With recent studies showing that artificial intelligence can interpret scientific research as well as (or even better) than field experts and specialists, StudyFinds was among the earliest to adopt and test this technology before approving its widespread use on our site. We stand by our practice and continuously update our processes to ensure the very highest level of accuracy. Read our AI Policy (link below) for more information.

Our Editorial Process

StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate. We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves. All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.

Our Editorial Team

Steve Fink

Editor-in-Chief

John Anderer

Associate Editor