Li Shizhen and the Bencao Gangmu (本草纲目) Explained: 27 Years Fixing a Library

In 1596, three years after its author died, a printing house in Nanjing finally issued a book that had taken most of a working life to write. Its title was 《本草纲目》 — Běncǎo Gāngmù, literally "Roots and Herbs, Outline and Details." In English it is almost always the Compendium of Materia Medica, which is a reasonable translation and a slightly misleading one, because it makes the book sound like a drug list.

It is not a drug list. It is a reclassification project, and that is the thing worth understanding about it.

---

The man, briefly, without the halo

李时珍 (Lǐ Shízhēn) was born on 3 July 1518 in Qizhou, Qichun County, in what is now Hubei province, and died in 1593 (Wikipedia: Li Shizhen). His father, 李言闻 (Lǐ Yánwén), was a well-regarded local physician. In Ming China that was a respectable trade and a socially unimpressive one, which is why his father pushed him toward the examination system instead.

He passed the first level, becoming a 秀才 (xiùcái) at fourteen, and then failed the provincial examination three times, which ended any realistic path into the bureaucracy (Chinese Wikipedia: 李时珍). He became a doctor instead — not as a calling, initially, but as what was left.

He was good at it. At thirty-eight he was hired by the Prince of Chu's household in Wuchang as 奉祠正 (fèngcí zhèng), with charge of the princely medical office, and in 1556 a recommendation carried him to the 太医院 (Tàiyīyuàn), the Imperial Medical Academy in Beijing. He lasted about a year and resigned (Chinese Wikipedia).

That one year matters more than the résumé line suggests. The Imperial Medical Academy held the empire's best pharmacological library. He read it, and what he concluded was that it was broken.

---

The actual problem he set out to solve

This is the part that gets skipped in favour of the heroic version, and it is the part that makes him a scientist rather than a saint.

China had a continuous 本草 (běncǎo) tradition — pharmacopoeia literature — running back well over a thousand years, layer upon layer, each compiler adding to the last without ever pruning. Li Shizhen's complaint about the inherited corpus, as Chinese sources record it, was a cataloguing complaint:

品数既烦，名称多杂。或一物析为二三，或二物混为一品。 "The items are already too numerous and the names a tangle. Sometimes one thing is split into two or three; sometimes two things are merged into one."

That is a taxonomy failure, stated in taxonomy language. The same plant appearing under four names in four books gets counted four times and prescribed four different ways. Two unrelated substances sharing a regional name get treated as interchangeable. And — the point he seems to have found genuinely alarming — poisonous minerals inherited from the alchemical tradition were sitting in the pharmacopoeia labelled as life-extending (Chinese Wikipedia).

So the project was not "collect more herbs." It was: take a thousand years of accumulated, self-contradictory text, check it against physical specimens and against each other, and impose a structure that makes the contradictions visible.

He consulted over 800 earlier books and travelled to collect and examine material himself.

---

What "27 years" actually means, and why sources fight about it

The figure you will see everywhere is 27 years, with three complete revisions. English Wikipedia gives exactly that (Li Shizhen).

But the numbers do not agree across sources, and pretending otherwise would be the easy lie:

• English Wikipedia's article on the book itself says he "worked for more than three decades" (Bencao Gangmu).
• Chinese Wikipedia's account is different again: 30-odd years to complete it, followed by a further 12 years spent revising it three times (Chinese Wikipedia: 本草纲目).

What every version agrees on is the shape: a draft finished in 1578, and an author who then kept going back into it until he died. The "27 years" is a clean number attached to a messy process, and it is worth knowing it is a rounded figure rather than a documented start-and-stop date. If you want the Chinese phrase for the quality involved, it is 持之以恒 (chí zhī yǐ héng), "hold to it with constancy" — or more brutally, 呕心沥血 (ǒu xīn lì xuè), "vomiting up the heart and draining the blood." The tradition has a whole vocabulary for this; we collect more of it in Chinese proverbs about patience.

---

What is actually inside it

Here the sources are tighter, and the numbers are remarkable.

• 1,892 entries — the gāng (纲), the "outline" headings of the title. Both English and Chinese sources agree on this figure. Of these, 374 had not appeared in earlier pharmacopoeia (Li Shizhen).
• 11,096 attached prescriptions (Bencao Gangmu; Chinese Wikipedia) — roughly 8,000 of them not widely circulated before.
• Around two million Chinese characters of text (Bencao Gangmu).
• 16 divisions (部, bù) subdivided into 60 classes (类, lèi), spanning plants, animals and minerals. Both language editions agree on 16 and 60.

Two figures genuinely conflict and you should know which:

• Length. Chinese Wikipedia says 52 juan (卷, fascicles); English Wikipedia says 53. 52 is the figure standard in Chinese scholarship.
• Illustrations. Chinese Wikipedia gives 1,109 drawings; English Wikipedia gives 1,160. Li Shizhen's son 李建元 (Lǐ Jiànyuán) did the illustrating.

The classification scheme is the intellectual achievement, and it is easy to undersell because it looks obvious now. Li Shizhen threw out the old three-tier ranking of drugs by moral grade — superior, middle, inferior — and reorganised everything by what the substance is: waters, fires, earths, metals and stones, then herbs by growth habit and habitat, then grains, vegetables, fruits, trees, then insects, scaled animals, shelled animals, birds, beasts. It runs, broadly, from the inanimate to the human.

That is a natural-kinds hierarchy. Each entry is then broken into fixed fields — name and its variants, a critical review of what previous authorities claimed, description of appearance, method of preparation, flavour and nature, applications, attached prescriptions. Standard fields, consistently applied, with the earlier literature explicitly quoted and then adjudicated.

The reason it is a landmark in the history of science is not that it catalogued more plants than anyone else. It is that it is structured, cross-referenced, and argues with its own sources in the open. This is a work built on 实事求是 (shí shì qiú shì), "seek truth from facts," and on 集腋成裘 (jí yè chéng qiú) — "a fur coat from a thousand fox armpits," the idiom for a great thing assembled from tiny accumulations.

---

Now the honest part

The Bencao Gangmu also contains material that is not medicine by any modern standard, and treating it as an unblemished scientific monument does it no favours.

English Wikipedia flags two plain errors of fact: the work states that quicksilver (mercury) and lead are not toxic, and that otters are exclusively male (Bencao Gangmu). The mercury one is not a trivial slip — it is precisely the inherited alchemical claim he elsewhere set out to correct, surviving into his own text.

Chinese Wikipedia's account is blunter still, listing entries such as 孝子衫 (xiàozǐ shān), a filial son's mourning garment, and 寡妇床头尘土 (guǎfu chuángtóu chéntǔ), dust from the head of a widow's bed, presented as materials with therapeutic effect (Chinese Wikipedia: 本草纲目). These are sympathetic magic. They are in the book.

You can hold both facts at once, and you should. Li Shizhen was doing something recognisably scientific — systematic collection, direct observation, explicit source criticism, consistent classification — inside a sixteenth-century framework that had no way to distinguish a pharmacological mechanism from a symbolic one. He could tell you that two books contradicted each other. He could not run a controlled trial. So when the tradition handed him an item with no observable effect but a coherent symbolic logic, he generally kept it and recorded what the sources said.

The book is a monument to method, not to modern pharmacology. Reading it as the second thing is how you end up with the tiresome internet genre of "ancient Chinese medicine already knew everything," which is both false and, oddly, less impressive than the truth.

---

The ending, which is not triumphant

He did not see it printed. The draft was done in 1578; printing began around 1593, the year he died; the first edition — the 金陵本 (Jīnlíng běn), the Nanjing edition — appeared in 1596 (Bencao Gangmu; Chinese Wikipedia).

His son Li Jianyuan presented a copy to the throne. The emperor, per English Wikipedia's summary, "saw it but did not pay much attention" (Bencao Gangmu). Thirty years of a man's life, received as a filing.

It was the reprints that did the work — 1603, 1606, 1640, and onward without interruption into the present century. Li Shizhen also left the 《濒湖脉学》 (Bīnhú Màixué, "Lakeside Master's Study of the Pulse") and 《奇经八脉考》 (Qíjīng Bāmài Kǎo, "Examination of the Eight Extraordinary Vessels"), both of which survive; most of his other titles do not.

On UNESCO: the Ben Cao Gang Mu is inscribed on the Memory of the World Register, and the register listing gives the year as 2011, with the copy held by the Library of the China Academy of Chinese Medical Sciences — the same year as the Huangdi Neijing (《黄帝内经》), the Yellow Emperor's Inner Canon (Memory of the World Register – Asia and the Pacific). That is the sourcing available to us; UNESCO's own entry page for the item could not be retrieved at the time of writing, so treat 2011 as the register listing's figure rather than something we confirmed at the primary source.

---

Why it is worth your attention

Strip away the nationalist framing on one side and the reflexive dismissal on the other, and what is left is a specific and durable achievement: one man decided that a thousand years of inherited text had become unusable through duplication and drift, and spent the rest of his life imposing an order on it that let a reader see where the sources disagreed.

That is the same job a modern reference work does. He did it alone, on foot, by hand, with no institution behind him after he walked out of the Imperial Medical Academy, and with no reasonable expectation of seeing it published.

The book's own title tells you what he thought he was doing. 纲 (gāng) is the main rope of a net; 目 (mù) is the mesh. Pull the rope and the whole net comes up in order. He was not adding herbs. He was building the rope.

---

More on the people and texts behind Chinese cultural history: the real Tang dynasty astronomy and Chang'an, the history behind Journey to the West, and Chinese proverbs about patience.

Explore our 1,000+ Chinese idioms with English explanations, or start with idioms about Wisdom & Learning.