Abstract

Facial representation from the skull is an interdisciplinary practice that seeks to generate a plausible likeness of a deceased individual by integrating cranial morphology with soft tissue depth data. Despite methodological and technological advances, its scientific validity, reproducibility, and the boundary between objective inference and subjective interpretation remain contested. This narrative-critical review surveys literature from PubMed, Scopus, Web of Science, and Google Scholar, without a lower date limit and using combinations of facial reconstruction, facial approximation, and validation keywords, tracing the field’s evolution from the plastered Neolithic skulls of Jericho to contemporary genetic phenotyping. It examines methodological traditions, validation studies, and epistemological challenges, highlighting the gap between scientific aspirations for precision and the inherent uncertainty of soft tissue prediction. The evolution of the Russian, American, and Manchester schools reflects advances in measurement and standardisation. Although digital tools have improved efficiency and consistency, they have not eliminated informed artistic judgement or interpretative uncertainty. The terminological debate over “reconstruction,” “approximation,” and “representation” frames expectations regarding accuracy and scientific authority. Validation studies report variable levels of anatomical and perceptual accuracy, depending on evaluation criteria, reporting bias, the shortage of independent blinded tests, and perceptual factors in human recognition. The review distinguishes the epistemological demands of forensic, archaeological, and palaeoanthropological applications and examines the risks of misrepresentation, from reinforcing racial stereotypes to the requirements of forensic identification. Facial representation is best understood as an evidence-based process with inherent degrees of uncertainty rather than a definitive reconstruction. Strengthening scientific credibility requires population-specific soft tissue databases, standardised validation protocols, and transparent reporting of limitations. By integrating historical, methodological, and applied perspectives, the review presents an overview of current knowledge and contributes to critical discussion of the “scientific art” of facial representation.