Official and independent public pages used here include UNESCO’s 1987 Mausoleum listing (https://whc.unesco.org/en/list/441/ (opens in a new tab)) and Britannica’s Qin Shi Huang biography (https://www.britannica.com/biography/Qin-Shi-Huang (opens in a new tab)).
The first fact about Qin Shi Huang is a source limit. The richest narrative is Han, written by people who had reason to make Qin a warning. The richest objects are underground and were unseen until 1974. A dossier that treats Shiji as a photograph is naive. A dossier that treats Qin as unknowable is evasive.
The received harms are book control, specialist killings of uncertain scale, corvée on walls, palaces and the tomb, convict labour, household registration, succession intrigue, and a collapse within years of his death. Scientific reporting on mercury in the mound is a low-confidence toxic addendum. 1970s and post-2012 political praise of the unifier is a modern polarising use, not a terracotta fact.
The offsets are real in the record: ending interstate war among the seven, commanderies instead of some fiefs, script and metrology, roads and canals, a written legal culture now visible on bamboo, and — much later — the knowledge explosion after 1974 and a World Heritage listing. Those gains do not rebate the dead workers. They are separate paths.
What the score must not say: that he is Xi Jinping; that 460 scholars is a Factrail count; that the tumulus is opened; that UNESCO undid corvée; that a banliang coin is a welfare state. Complex contribution is the honest type. Confidence is high on 1974–1987 archaeology and listing, medium on Shiji’s numbers, low on slogans about buried Confucians as a headcount.
Li Si, Zhao Gao, Jing Ke, Sima Qian, and modern PRC speakers are specific related roles, not a pad. The formula writes the score when evidence on these rows is public and existing drivers already have indicator links.
1974 is not 221. 1987 is not corvée. A Shiji sentence is a Han indictment, which is why confidence on burnings and killings stays medium or low. Bamboo slips from Shuihudi are a different kind of evidence: they show written penal administration without requiring us to believe every Han anecdote. Nature’s mercury reporting is another kind: a possible chemical signature, not an opened chamber.
The terracotta figures lost paint in the first excavations. That information loss is stored separately from the discovery gain. Tourism in Lintong is stored separately from convict labour. Official stelae of self-praise are stored separately from canal-building. If those distinctions feel pedantic, they are the alternative to a statue and the alternative to a cartoon.
What would change the picture: a securely dated Qin statute of book control independent of Shiji; an opened tumulus with a documented workforce ledger; a replicated mercury map with a published method. Until then, the dossier refuses fake headcounts and refuses fake eternity.
Related people are court, assassin, historian, and later political users. They will be created as identities where missing, then later enriched if they receive their own dossiers. That is how a live graph grows without inventing meetings. Off-family indicator padding — a health series nailed to a stele without a mechanism — is refused.
The live score, if evidence is published and driver-indicator links already in production can see these paths, will not look like a museum caption. It will look like a sum of extractive labour, administrative standards, knowledge loss and later knowledge gain. If it looks too neat, the confidence fields were ignored. A closed tumulus is not a confession and not a tourist insult; it is a conservation (and political) choice stored as its own row.
