The Empirical Paper Confirms the Premise but Muddies the Picture
2026-07-21 · AI-generated insight
Winegard and Carl offer a tidy theoretical scaffold for race: humans dispersed, faced divergent selection pressures, and were kept apart by oceans, deserts, and mountains long enough to drift and adapt apart. It's a clean model — isolation plus differential selection equals divergence. The ancient DNA paper supplies exactly the kind of hard evidence the model predicts: not just drift, but pervasive directional selection, hundreds of alleles pushed by Darwinian forces, including trait-relevant changes in body fat, cognition, and disease risk.
But read closely, the empirical work complicates the philosophers' neat story in two ways.
First, timing. Winegard and Carl locate the action 75,000 years ago, in the great dispersal. The genomic time-series locates the strongest, most pervasive selection in the past ten millennia — long after the initial split, coinciding with agriculture and dense settlement. Selection isn't mainly the ancient sorting that produced continental groups; it's ongoing, recent, and local. The forces shaping populations kept working after the map was drawn.
Second, and more pointedly, the whole methodological burden of the ancient DNA paper is distinguishing directional selection from "migrations, population structure, or non-adaptive" change. That is precisely the confound the isolation model waves away. West Eurasia, the paper reminds us, is not a sealed compartment behind a mountain range — it's a palimpsest of repeated migration and admixture, and separating adaptation from gene flow required 15,000 genomes and careful statistics.
So the two notes rhyme and grate at once. The empirical paper vindicates the premise that populations diverge under selection, giving Winegard and Carl more than they asked for. Yet it also shows the process is messier, more recent, and more entangled with movement than a story built on stable barriers can accommodate. If selection is this vigorous and this recent even within one region, the boundaries the isolation model treats as fixed are better read as momentary snapshots of a system that never stopped churning.