The Terracotta Makers Left Their Marks

A child listens; a parent reads. This page answers the questions that come after the story. The script is available from the script button during playback.

What this is about

These are the words the story stops to explain. The Terracotta Army is a set of big clay statues, shaped like people and baked hard, about two thousand two hundred years ago. A workshop is a place where makers work side by side. A crossbow is a bow set on a wooden frame, and pulling the trigger sends the arrow. A standard is an agreement to make many things the same size and shape. Rustproofing means putting a thin coating over metal so it does not rust. An archaeologist works out the past from things dug up out of the ground.

There is one question here — who made these? The Terracotta Army carries one man's name, but it was made across several workshops, and the marks those makers left are what we read today. This is neither a great-man story about the emperor nor its mirror image, a tyrant story — and the evidence says nothing in support of the second. What it describes is organisation, standards, and marks.

What actually happened

• In the third century BCE, thousands of clay figures and more than forty thousand bronze weapons were buried beside the tomb of the first emperor of Qin. The study that measured them treats the site as a question about mass production and the organisation of labour (DOI: 10.30861/9781407316901).
• The clay paste is not the same across the figures. Some carry the stamped names Gong (宫) and Xianyang (咸阳), and the split in composition follows those names. The study reads this as what appear to be the products of two workshops (DOI: 10.1111/arcm.12613).
• Two crossbow triggers from Pit 1 carry writing in ink. Multispectral imaging, a digital microscope, and Raman analysis found it, and the ink turned out to be soot-based. The study describes an office culture built on small acts of accountancy in which objects and people were marked so their movements could be tracked, their quality checked, and their numbers counted (DOI: 10.1186/s40494-018-0239-5).
• Measuring the triggers and mapping where they lay shows they were produced in batches, that the batches were possibly held in an arsenal, and that they ended up spread by sector within Pit 1 (DOI: 10.1017/s0003598x00050262). A study of all forty thousand weapons — by type, size, microscope, chemistry, and position — reads out work divided into production cells and steps for checking quality (DOI: 10.5334/ai.1316).
• In 1998, more than two hundred sets of ceremonial stone armour came from the same site. Each set is more than six hundred limestone plates laced with bronze wire. The craftsmen sawed and split thousands of plates out of the raw rock and polished and drilled them, and the tools and the tool marks document that laborious work (DOI: 10.1179/sic.2006.51.supplement-2.170).
• The figures were all originally painted — sculptures, chariots, and weapons alike — and the few colours that survived the centuries were lost during excavation (DOI: 10.1179/sic.2006.51.supplement-2.177). After more than two thousand two hundred years in wet soil the ground layer beneath the paint had broken down, which is what makes keeping the paint such a hard problem (DOI: 10.1002/anie.200301633).

Questions a child may ask

"Wasn't it the First Emperor who made them?" Setting a job and making a thing are two different acts. What the evidence shows is an organisation split across workshops. There is also a study that goes after the "one man did it all at once" story. Re-reading the earliest sources for the claim that writing was unified in 221 BCE, it argues there was a political and ideological agenda behind ascribing the process to a single person. The dug-up material shows no such sudden shift; even the edict plates on the new standard weights were not written in one script, and the weights themselves ran off their face value. The script changed over decades and even centuries, and the idea of a single unification was, on this argument, a Han-era creation (DOI: 10.1556/aorient.57.2004.2.2). Making standards was likewise gradual refinement rather than abrupt innovation (DOI: 10.1163/15685322-10803006).

"How many soldiers are there?" ⚠️ Our evidence does not give a count. Eight thousand is the number that circulates, but the paper we read says only "thousands" (DOI: 10.30861/9781407316901). The script therefore gives no number. The number of makers is not in the evidence either.

"Why were they buried at all?" They were made to guard the tomb. The study that measured the faces describes them as a real army intended to protect the emperor in the afterlife (DOI: 10.1186/s40494-022-00662-0). The script states the tomb once, as background, and does not describe death.

"Is it true every face is different?" One study measured the faces with high-precision 3D point clouds and set them beside twenty-nine ethnic groups in China today; it concludes the figures were crafted from real portraits. But the paper itself says first that this is still debated — were the models based on individual people, or made from standardized models (DOI: 10.1186/s40494-022-00662-0)? A study of building 3D models from ordinary photographs calls this material "a famous and contentious set of archaeological objects" (DOI: 10.1016/j.jas.2014.05.014).

"Didn't they use chromium plating two thousand years ago?" No. That was the accepted view for forty years, and it was overturned in 2019. Testing weapons, lacquer, and soil together, and re-running the process in the lab, the team found that surface chromium goes with the type of object and not with how well it kept. The lacquer was the chromium-rich material, and the chromium on the metal came over from nearby lacquer after burial. The authors write that the theory should be abandoned. The popular name is chromium plating, but the paper tests a chromate conversion coating, not a metal layer. The good preservation is probably down to the moderately alkaline pH and very small particle size of the burial soil, plus the composition of the bronze (DOI: 10.1038/s41598-019-40613-7).

"Who works all this out now?" Since the find of 1974, five decades of digging have challenged earlier ideas about art, technology, and belief, done by teams from many fields and countries. The results require the written sources to be read again — this period used to be known mainly from the Shiji, written in the first century BCE (DOI: 10.1093/oxfordhb/9780197800324.013.0009). The objects mostly come out in fragments, and sorting them is still done by hand (DOI: 10.1109/access.2019.2962791).

What is evidence and what is reading

Established — the 1974 discovery, the forty thousand-plus bronze weapons, the two hundred sets of stone armour and six hundred plates from 1998, the original paint, the ink writing on the triggers.

Held loosely on purpose — ⓐ that Gong and Xianyang mean two workshops is what the study says seems to be so. ⓑ that the batches sat in an arsenal is a possibility. ⓒ the chromium paper's reason for good preservation is probable, not settled. ⓓ the face study's conclusion arrives with the debate still open. The script asserts none of the four.

The grain of the evidence — most citations here are measurement and analysis at a single site (clay chemistry, trigger sizes, facial point clouds, surface chromium). With one site there is no second place to check against. Two are exceptions: the papers on unified writing and on manufacturers' labels look at other bodies of material, so the script never states their findings as facts about the Terracotta Army and says "another study reads it this way" instead. The abstract of the weapons monograph is a publisher's description.

⚠️ What we could not include — the name of a single maker. Editions for children usually fill four kinds of material about a person: the scene of the work, a choice the person made, what it changes for a child today, and honour and what came after. This one could not fill "a choice the person made." No craftsman is named anywhere in the abstracts we read, and without a name there is no choice to tell. We did fill the scene of the work (soot ink, the marks of saw and grindstone and drill), what touches a child today (we read those marks now), and honour (the credit went to one man's name). Choices made by the organisation — batch work, laying pieces out by sector — are a system, not a person's choice, so they do not count here.

Left out because the subject is different — the elixir of life, the rivers of mercury, the burning of books. Our evidence carries none of them. The Great Wall turns up only in an unreviewed paper filed twice, so it is not cited. The well-known story that a farmer found the site while digging a well is also not in the evidence, and the script says so out loud rather than dropping it quietly.

Things to try together

• Turn over a bowl or a toy at home. There is a maker's name or a number stamped underneath. People were doing the same thing two thousand two hundred years ago, and that is what this edition is about.
• Make several of the same shape together — clay balls will do. Finding how hard it is to make them alike is what standard means.