The canon
本草纲目
The Bencao Gangmu, printed in 1596, is Li Shizhen’s survey of everything then known about what can be eaten or used as medicine: 1,892 entries and 11,096 formulas, gathered over twenty-seven years and sorted into sixteen divisions and sixty categories. It is the classical source for 药食同源 — food and medicine share one origin.
The compiler
Li Shizhen was born in 1518 in Qizhou, in what is now Hubei, into a family of physicians who wanted him to be an official instead. He sat the provincial examination three times and failed it three times. Some time in his twenties he stopped trying and took up his father’s work, and by 1552 he had begun the book.
His complaint about the existing surveys of materia medica was not that they were old. It was that they were sloppy: the same plant filed twice under two names, two different plants filed once under one, quantities copied wrong, and errors repeated faithfully from edition to edition because nobody had gone and looked. He went and looked. He walked, he asked farmers and fishermen and woodcutters, he grew things, and he read some eight hundred earlier works and named them.
He finished in 1578, aged sixty, and then spent years failing to find a printer willing to take on fifty-two scrolls. The scholar Wang Shizhen supplied a preface in 1590 that finally moved it. Li Shizhen died in 1593. The first edition came off the blocks in Nanjing in 1596, and he never saw it.
How it is built
The title is the method. 纲目, gāngmù, is the net-rope and the mesh: a coarse classification that holds a fine one. Sixteen divisions hold sixty categories, the categories hold the entries, and every entry is written under the same eight headings in the same order — which is what makes it possible to compare two things four hundred years later.
The filing system
The sixteen divisions are not alphabetical and they are not arbitrary. They run water, fire, earth, metals and stones, herbs, grains, vegetables, fruits, trees, clothing and utensils, insects, scaled creatures, shelled creatures, birds, beasts, and human. Li Shizhen says plainly why: he is going from the plain to the elaborate, from the mineral to the living, and putting people at the end because that is where they come in the order of things.
Two hundred years before Linnaeus, this is a working classification. It has none of Linnaeus’s machinery and it is no theory of descent, but it does the one thing a classification must: it puts like next to like, so a reader who knows one entry can guess at its neighbours. Joseph Needham called it the greatest scientific achievement of the Ming.
Four of the sixteen divisions — grains, vegetables, fruits and the domestic animals — are simply food. That is not an aside in the book’s architecture. It is roughly a quarter of the whole survey, sitting in the same sequence and written under the same eight headings as everything else, with no line drawn between what you eat and what a physician reaches for.
One origin
The idea is older than the Bencao Gangmu and the phrase is younger, but this is the book that makes it undeniable, simply by filing a pear and a mineral under the same set of headings. Yàoshí tóngyuán says that the kitchen and the medicine chest are two ends of one shelf, and that a great many things sit in the middle where the distinction stops being useful.
The clearest statement of it is earlier. Sun Simiao gave a chapter of his Beiji Qianjin Yaofang, of 652 CE, to ordering the diet, and told the physician to find the source of a disorder and answer it with food first, turning to medicinals only when food had not answered. That is a priority, not a platitude: food is the first instrument, and the one an ordinary person can pick up alone.
It is now written into Chinese law. The health authorities maintain a published catalogue of substances counted as both food and medicinal — ginger, jujube, goji, tangerine peel, lotus seed, job’s tears and their neighbours — which the 2024 revision brought to 106 items. Everything our food notes mention sits on that list, and nothing outside it does.
Where the earlier books disagree I have set the disagreement down, and where I have found them mistaken I have said so under its own heading.
Two entries
Goji sits in the tree division and jujube in the fruit division, several hundred pages apart, each with its names, its disagreements, its nature and flavour and its attached formulas. You can also buy either one in a supermarket in Ohio, which is the whole point of the phrase above.
In a reading
A Bencao House report suggests food. It can do that honestly because the tradition it draws on never separated food from the rest, and because the specific things it names — a slice of ginger in hot water, jujube in a congee, chrysanthemum steeped and left to cool — are groceries in China and groceries here. The Bencao gives each of them a nature, a flavour and four centuries of notes, and the food notes in a report are a plain English summary of those.
We stay inside a small part of the book on purpose. The overwhelming majority of the 1,892 entries are not food, and nothing we write will ever mention them as something for you. We send no herbs, we suggest no herbs, and we do not name a formula for you to make. The line is the same one Sun Simiao drew in 652 CE, and we stop at the first half of it.
What that leaves is genuinely useful and genuinely modest: a short list of ordinary foods that suit the pattern your answers describe, a shorter list of things to have less of for a while, and the reason for each in one sentence. How a reading is made shows where in the process that list is written.
Questions
A physician from Qizhou in Hubei, 1518 to 1593, the son and grandson of physicians. He sat the provincial examinations three times, failed, and gave up the official career for medicine. He spent twenty-seven years on the Bencao Gangmu, finished it in 1578, and died in 1593, three years before it was printed.
Fifty-two scrolls. 1,892 entries, of which 374 had never appeared in a bencao before. 11,096 attached formulas. 1,109 drawings, made by his sons. Everything is filed into sixteen divisions and sixty categories, running from water and fire through plants and animals to human material at the end.
Yàoshí tóngyuán — food and medicine share one origin. It is the principle that the kitchen and the medicine chest are the two ends of one shelf, and that many of the same plants sit in both. It is why a Chinese physician asks what you eat before suggesting anything else, and it is the whole basis of what this house is allowed to suggest.
He cited a Chinese encyclopaedia of 1596 in The Variation of Animals and Plants under Domestication, published in 1868, for its account of varieties in domestic fowl and goldfish. The date and the description point to this book, and it is usually identified as such, though Darwin worked from a translated extract rather than the original.
Only foods. We stay inside the everyday kitchen half of the Bencao — ginger, jujube, goji, tangerine peel, lotus seed — and inside China’s own published list of substances counted as both food and medicinal. We send nothing, we suggest no herbs, and we do not touch the rest of the book.
The reading
A reading names the pattern your answers make and then says what to eat and what to ease off, drawn from the everyday end of the Bencao and written in plain English.