Raiding and Morphological Preferences: A Hypothesis for Sex-Biased Admixture Between Neanderthals, Denisovans, and Anatomically Modern Humans

Recent genomic evidence from Platt et al. (2026) demonstrates a strong sex bias in Neanderthal–modern human admixture, favoring Neanderthal males mating with anatomically modern human (AMH) females. While mate preference is proposed as the parsimonious explanation, the underlying mechanisms remain unspecified. Here, I hypothesize that this bias arose from inter-group raiding dynamics operating in concert with a hardwired neoteny-recognition system. Critically, the asymmetry in admixture was driven not only by Neanderthal males preferentially selecting AMH females—a pattern the raiding model explains—but also by the reciprocal absence of selection by AMH males on Neanderthal females. I ground this dual asymmetry in a mechanism more fundamental than cross-cultural mate-choice studies: the phylogenetically ancient cognitive system for neoteny recognition that underlies infant bonding in all higher primates. Because neoteny recognition is prerequisite to parental investment, it was present and operational in Neanderthals, Denisovans, and AMH alike. AMH females, retaining neotenous craniofacial features into adulthood to a markedly greater degree than archaic females, would have triggered this system preferentially in Neanderthal males, driving their coercive acquisition of AMH partners through raiding. Simultaneously, Neanderthal females—sharing the robust cranial and skeletal phenotype of their males—failed to trigger this system equivalently in AMH males, meaning modern human males did not select archaic females as mates with comparable frequency. The resulting asymmetry is thus twofold: Neanderthal males actively sought AMH females, while AMH males did not reciprocally select Neanderthal females. This framework explains why admixture was overwhelmingly unidirectional despite both groups possessing the same neoteny-recognition circuitry. I extend this model to Denisovan admixture, where parallel genomic sex-bias signals provide independent confirmation. The model integrates behavioral ecology, evolutionary psychology, and paleoanthropology with the 2026 genetic data, predicting asymmetric gene flow without invoking hybrid incompatibilities. I outline testable implications for future archaeological and genomic studies.