Fingerprints on the Terracotta Warriors Show Women and Teenagers Built Them Too

October 24 — According to a Xinhua report yesterday, a study by the Emperor Qinshihuang’s Mausoleum Site Museum and the School of Archaeology at Shandong University has provided biological evidence, drawn from the figurines themselves, of who made the Terracotta Warriors.

The team extracted 79 independent fingerprints from 23 figurines in trench T23 of Pit No. 1 and used a Bayesian mixture model to infer age and sex at the population level. The results suggest that adult men made up about 45.0%, adult women about 32.7% and juveniles under 15 about 22.3%. It is the first time scholarship has used quantitative biological evidence to show that terracotta production was dominated by adult men but was not carried out by adult men alone.

For a long time, knowledge of the makers came mainly from written records and inscriptions on objects; the figurines themselves had offered no direct demographic evidence. The study could “read” fingerprints because ridge breadth is statistically linked to age and sex: adult men’s ridges are usually further apart than those of women and juveniles. But traditional methods have clear limits — the ridge-width distributions of adult women and of 12- to 16-year-olds overlap heavily, so a single fingerprint is hard to classify, and shrinkage during firing also disturbs measurements.

The team therefore adopted the Bayesian mixture model proposed by Burchill and colleagues in 2023. They first used ultra-depth-of-field 3D photogrammetry to capture 91 candidate prints, removed repeated impressions on the same figurine whose ridges deviated by less than 0.02 mm, kept 79 independent samples, and then calibrated against modern reference data to estimate the population structure.

Where the fingerprints sit on the figurines offers clues about which steps women and juveniles took part in. Of the 79 prints, 28 are on heads, 27 on armour and 16 on arms — together more than 90%. By how they formed, 73 are surface prints left by pressing directly on wet clay during shaping and trimming, and 6 are relief prints from gripping when appliqués such as armour studs and belts were attached. This distribution corroborates the “moulding plus assembly” production model: heads, arms and armour were handled repeatedly at the prefabrication stage, and the artisans who left prints did not necessarily work on a whole figure.

The study also shows that the peak age for adult male fingerprints was about 20, that female fingerprints likewise clustered in young adulthood, and that the share of juveniles matches Qin-era records of apprentices assisting artisans.

The researchers stress that these proportions are population-level statistical estimates and cannot be mapped onto any single artisan; trench T23 is only a local area, and both the sample and the reference data have limits, so the results should be read as estimates based on the available sample. The work lays a foundation for larger fingerprint-demographic surveys in Pit No. 2 and other burial pits.

The paper’s first author is Li Xiaoxi, a curator at the Emperor Qinshihuang’s Mausoleum Site Museum, and the corresponding author is Shen Jingyi of Shandong University’s School of Archaeology, with co-authors including Lan Desheng, Wang Jingyi, Yu Yang, Cai Jingye, Xia Yin and Zhou Ping.