The limits of a face, and of a question
We once knew this ancient human population mainly through traces left in genomes. Ancient proteins from two Xujiayao fossils now place pieces of their bodies on firmer ground. A skull fragment and a mandible carry the COL1A2 collagen substitution identified in two laboratories. More than supplying one face for a vanished population, the result changes which physical features can be linked to that population through a direct molecular test.[1]
Placing a fossil in the old human family tree involves two jobs: describing its shape and testing its identity against an independent trace. Protein sequences did the latter at Xujiayao; anatomical comparisons enriched the former. The suggestion that other fossils resemble this population offers a valuable line of inquiry, but it does not mean that all were molecularly identified. Unresolved finer branches within the protein family tree show why that distinction matters.[1]
Variation instead of a straight line
The fossils combine traits reminiscent of different earlier and later human groups. It is tempting to place them on a straight line running from a primitive beginning to us; that line hides the variation the evidence actually shows. A low cranial profile, a pronounced brow or a jaw shape does not reconstruct an entire living person. Cranial volume does not measure intelligence. When we turn a measured resemblance into a life story, habit takes the place of curiosity.[1]
Reading molecular identity alongside anatomical variation offers a way to stop treating ancient human populations as incomplete versions of one another. Yet only two fossils here were molecularly identified; their number alone cannot define the appearance of every member of the population. A proposal about wider variation depends on comparisons with more fossils and remains open to testing. Giving up a single uniform portrait does not require replacing it with another portrait unsupported by evidence.[1]
The space an exhibit can leave
A face beside a bone in a museum case is an understandable invitation to visitors. But an image can hide uncertainty if it does not distinguish what a protein identified, what bone shape supports and what an artist supplied. Xujiayao provides a useful example for such a label: the Denisovan identity of two bones has strong support, while the resemblance of other fossils and the individuals’ complete appearance are known with less certainty. This is a proposal for presenting evidence, not a study of what museums currently do.[1]
The evidence about ancient relatives does not establish that they asked our questions or lived our lives. Proteins do not preserve their language, intentions or memories. But the boundary we draw while naming two bones exposes a present choice: we choose between telling the past as a finished family tree and presenting it as an inquiry open to new measurements. Leaving uncertainty visible does not diminish the past. It shows the strength of a finding alongside the questions it cannot answer.[1]