Who produced Emperor Qin Shihuang’s Terracotta Army, China? A Bayesian mixture model approach to fingerprint age and sex determination

The demographic composition of the Terracotta Army's producers has long been debated. Fingerprints preserved on the clay figurines offer a critical source of direct biological evidence to address this question. However, traditional fingerprint analyses, which rely on linear regression approaches, suffer from inherent methodological flaws that limit their inferential power. This study presents the first systematic demographic reconstruction of Terracotta Army producers using 79 well-preserved fingerprints recovered from Unit T23, Pit 1 of the Qin Shi Huang Mausoleum. We analyze mean ridge breadth (MRB) measurements using a Bayesian mixture model, which addresses the critical limitation of traditional frameworks that incorrectly treat MRB and ridge density (RD) as independent predictors. At the population level, the model estimates the assemblage comprises 45.0% adult males, 32.7% adult females, and 22.3% juveniles (<15 years) -an age threshold inferred from Qin dynasty legal texts defining age-based legal responsibility. These findings challenge the long-held assumption that Terracotta Army production relied exclusively on adult male labor, providing the first quantitative biological evidence for a diverse workforce including women and children. This study demonstrates that population-level Bayesian fingerprint analysis offers a robust methodological framework for investigating ancient craft labor organization. However, it should be emphasized that these proportions reflect fingerprint frequencies rather than counts of individual fingerprints, and behavioral biases plus reliance on a modern European reference dataset may introduce potential overestimation of female and juvenile representation.