The Neanderthal "chignon": variation, integration, and homology.

The occipital bun ("chignon") is cited widely as a Neanderthal derived trait. It encompasses the posterior projection/convexity of the occipital squama and is associated with lambdoid flattening on the parietal. A 'hemibun' in some Upper Paleolithic Europeans is thought by some authors to indicate interbreeding between Neanderthals and early modern Europeans. However, 'bunning' is difficult to measure, and the term has been applied to a range of morphological patterns. Furthermore, its usefulness in phylogenetic reconstruction and its homologous status across modern and fossil humans have been disputed. We present a geometric morphometric study that quantitatively evaluates the chignon, assesses its usefulness in separating Neanderthals from modern humans, and its degree of similarity to Upper Paleolithic 'hemibuns.' We measured the three-dimensional coordinates of closely spaced points along the midsagittal plane from bregma to inion and of anatomical landmarks in a large series of recent human crania and several Middle and Late Pleistocene European and African fossils. These coordinate data were processed using the techniques of geometric morphometrics and analyzed with relative warps, canonical variates, and singular warps. Our results show no separation between Neanderthals and modern humans, including early modern Europeans, when the shape of the occipital plane midsagittal-profile is considered alone. On the other hand, Neanderthals are well separated from both recent and fossil modern humans when information about the occipital's relative position and relative size are also included. Furthermore, the occurrence of a highly convex and posteriorly projecting midline occipital profile (interpreted as the occipital bun) is highly correlated (>0.8) with a flat parietal midsagittal profile and with antero-superiorly positioned temporal bones across both our recent and our fossil human samples. We conclude that the shape of the occipital profile alone should not be considered an independent trait, as it is very tightly integrated with braincase shape. Our analysis does not support differences in integration of the posterior midsagittal profile and the cranial base in Pleistocene and recent humans.