Reliable classification of intentional cranial vault modification and nonsynostotic deformational plagiocephaly using 3D geometric morphometrics.

The style and degree of cranial vault modification (CVM) can have a significant impact on the reliability of assigning crania into operational categories and correlates. These issues can, in turn, influence the interpretation of cultural practices in the archaeological record. If we intend to deploy head shape as a meaningful unit of analysis to address issues of anthropological significance, we need a reliable way to demarcate discrete categories. Here we employ 3D geometric morphometrics to study head shape variation and improve classification accuracy. The present study rendered 3D models of 52 crania (8 unmodified, 44 modified annular) affiliated with the prehispanic Chanka Culture (AD 1000-1400) of south-central Peru. A combination of landmarks and semi-landmarks captured the shape and curves of sampled crania. Principal components analysis (PCA) revealed clusters of shape categories, and agglomerative hierarchical cluster analysis quantified the spectrum of variation. Our methods found four categories of circumferential head shape: annular erect, annular oblique, unmodified, and a new operational category termed slight. This last category was most prone to misclassification through gross inspection. We argue that differences in head shape are due to diverse modes of babywearing and childcare in the late prehistoric Andes. This study provides 3D geometric morphometric techniques that can help alleviate some of the classification errors that have long confounded practicing bioarchaeologists.