The inspiration for this dissertation comes from the variability in a sample of Late Pleistocene fossil human crania found in the Southern Levant, in present day Israel. This sample (n=13) is comprised of both Neandertals from Tabun and Amud, and the remains from Skhul and Qafzeh that are considered predecessors of modern humans. Many authors describe the heterogeneity of the Levantine sample as unusual, greater than what modern populations could be expected to exhibit, and many believe they represent groups of different human species.

The focus of this study is on whether the magnitude of Levantine variation is really unusual—because this has significance for understanding the pattern of human evolution and identifying past species. The issue is addressed in a statistical, comparative context: 113 crania at 6 museums and laboratories in 5 countries were examined. To demonstrate the Levantine sample variation is really caused by taxonomic differences, at the minimum the Levantines should exhibit a greater magnitude of variation than expected in a comparable modern population of mixed ancestry from a confined geographic area and limited time span. This would indicate that the Levantine variation is not from species mixture; it involved a mixture of human populations.

This null hypothesis is tested by comparing the Levant sample's non-metric cranial variation with the cranial variation in 100 mixed-ancestry medieval crania from the Pannonian Plain (fourth to eighth century AD.). Dichotomous, binomial responses were collected for 67 traits. The mean and variance of ratios of the scores describe the heterogeneity versus homogeneity. One thousand samples of n=13 were randomly drawn with replacement from the Pannonian data. The fossil Levantine sample was compared to the resampled Pannonian distribution to test whether the magnitude and variance of the Levant data are expected within it.

The results show that the magnitude of the variation in the Levantine sample is not unusual when compared to the Pannonian sample, and the null hypothesis of admixture of different populations cannot be refuted. These results are strongly supported by the recent ongoing advances in the study of ancient and modern nDNA, showing significant Neandertal mixture in the Levant.