Europe's skin lightening happened fastest in a narrow window, and then nearly stopped

David Reich, the geneticist whose work on ancient DNA has repeatedly redrawn the population history of Europe, makes a claim about skin pigmentation that is sharper and more temporally precise than most of what appears in public discussion of the topic. The period of strongest depigmentation in Europe, he argues, falls between roughly 4,000 and 2,000 years ago. After that window closes, the pace drops substantially.

That framing carries real weight, because it locates the bulk of the change not in deep prehistory but in a relatively recent and archaeologically legible period. The Bronze Age and early Iron Age fall squarely within those bounds. So does a great deal of documented population movement, including the expansion of steppe-ancestry groups into Europe that Reich’s own research has done much to characterize. Whatever selective pressures were operating, they appear to have been operating with unusual force during that span and then to have eased.

The conventional picture of skin lightening in European populations is often presented as a gradual drift toward lower melanin production over tens of thousands of years, shaped by reduced ultraviolet radiation at higher latitudes and the body’s need to synthesize sufficient vitamin D. That picture is not wrong on its face, but Reich’s framing complicates the timeline considerably. If the steepest part of the curve runs from 4,000 to 2,000 years ago, then whatever was driving the change was not operating at a constant rate across the full span of anatomically modern human habitation in Europe. Something accelerated it, and something else, or the exhaustion of the selectable variation, slowed it.

The period of strongest depigmentation is between about 4,000 to 2,000 years ago, and then after that it's much less. David Reich

Pinning the acceleration to that specific window opens questions that are easier to ask than to answer. Population replacement events, dietary shifts, changes in subsistence strategy, and alterations in UV exposure from changing land use have all been proposed as contributors to rapid pigmentation change in ancient European populations. Reich does not specify in this claim which mechanism or combination of mechanisms accounts for the pattern. The claim is about the shape of the curve, not its cause.

That restraint is itself informative. Reich is characterizing a finding from the genetic record, not advancing a causal theory. The data, as he reads it, shows a concentrated period of change followed by a marked deceleration. The mechanism remains an open inference. A reader who wants to assign the acceleration to a single cause is going further than the claim licenses.

What the claim does license is a narrower but still significant point: the assumption that depigmentation in Europe was a slow background process, spreading diffusely across millennia at a roughly uniform rate, does not fit the pattern Reich describes. The change was concentrated. It was strong enough in that two-thousand-year window to register as the dominant signal. And it was followed by something that looks, on the genetic evidence, like a plateau rather than a continuation.

For researchers working on selection coefficients, population structure, or the timing of phenotypic change in ancient Europe, that temporal specificity matters. It constrains where to look and what to try to explain. A finding about the rate and timing of a selective sweep is a different kind of evidence than a finding about the sweep’s direction alone. Reich is pointing at the rate. That is the part of the claim worth sitting with.