The models are sophisticated statistical constructions, and they produce estimates of the timing and nature of population events — divergence, isolation, dispersal — with associated uncertainty ranges.

But genetic models of population history cannot, by themselves, tell us where the people were.

A model that estimates a period of Beringian isolation lasting roughly 10,000 years before the dispersal into the Americas tells us that the founding population was somewhere — probably somewhere in Beringia, based on the geographic logic of the situation — but it does not tell us specifically where the people were living, what they were eating, or what direct evidence of their presence might be found.

The genetic model constrains the history; it does not locate the evidence.