Anterior sphenoid in modern humans

Abstract

Lieberman has proposed¹ that reduced midfacial projection (MFP), in which most of the face lies beneath the neurocranium, is a major unique, derived character of anatomically modern Homo sapiens, and that this reduction is largely a consequence of reduced anterior sphenoid length (ASL). Lieberman's conclusions were based on comparisons of a small sample of archaic Homocrania with those of Holocene and Pleistocene anatomically modern H. sapiens. We have made new measurements of ASL and MFP, and find that ASL was incorrectly estimated in those archaic fossil crania in which these landmarks are unambiguously preserved. It turns out that the anterior sphenoid in modern humans is no shorter than in archaic Homo.

Main

The new measurements were taken from better-quality radiographs and computed tomography scans², ³ and from the original specimens of Gibraltar 1 and Broken Hill (courtesy of C. Stringer, T. Molleson and F. Zonneveld). ASL values in Holocene and Pleistocene modern humans are 19.9 mm (s.d. 2.0) and 20.0 mm (s.d. 1.8), respectively¹, not significantly different (P>0.05, Scheffé's F) from those of archaic Homo (Gibraltar 1, 17.2 mm; Monte Circeo, 16.9 mm; Broken Hill, 17.2 mm). Figure 1a confirms that reduced MFP in anatomically modern humans is not associated with a shorter ASL.

To assess the spatial relationships of ASL and MFP in relative terms, we did a geometric morphometric analysis comparing Holocene modern human crania with the three archaic Homo fossils (Fig. 1b,c). The transformation grid indicates that, relative to the size of the landmark configuration, MFP is shortened and ASL is lengthened in Holocene H. sapiens. The factors underlying these changes may include facial reduction, increased basicranial flexion, and expansion of the temporal lobes in the middle cranial fossae. The comparison also suggests that the pharyngeal area between the palate and the foramen magnum is anteroposteriorly constricted in Holocene modern humans, as was inferred by Lieberman¹, but that this is unrelated to ASL.

Figure 1: Analysis of measurements.

a, Plot of ASL and MFP (after Fig. 3a of ref. 1), with new measurements of Broken Hill, Gibraltar 1 and Monte Circeo. MFP is the distance from nasion to foramen caecum perpendicular to the posterior maxillary plane. ASL is the minimum distance from sella to the posterior maxillary plane. b, c, Geometric morphometric comparison of eight facial, neurocranial and basicranial landmarks from 28 combined-sex Holocene H. sapiensskulls (Indian subcontinent), Monte Circeo, Gibraltar 1 and Broken Hill. Principal components analysis of tangent coordinates following generalized Procrustes analysis yields 12 non-zero eigenvectors⁴. Only principal component I (PC I; 37% total shape variance) separates fossil from extant specimens. The shapes represent means for the fossils (b) and Holocene H. sapiens (c), both on PC I and each rescaled to their respective mean size. A cartesian transformation grid (thin plate spline⁵) from fossil to extant means is superimposed. Numbers: 1, nasion; 2, prosthion; 3, maxillary tuberosity; 4, laterally projected average intersection between greater wings and planum sphenoideum; 5, anteriormost point of cranial cavity; 6, foramen caecum; 7, sella; 8, basion (estimated in Neanderthals); 3 to 4, posterior maxillary plane.

We conclude that, although ASL is intraspecifically correlated with MFP in modern humans and chimpanzees¹, it does not account for the unique form of the modern human cranium. Our analysis highlights the need for research that integrates comparative morphometric analyses with developmental studies of cranial growth in human and non-human primates.