Baize Is Not in the Classic of Mountains and Seas

Most people meet the name Baize attached to the Classic of Mountains and Seas: an ancient beast that could speak and knew everything. The association is so widespread that few ever go back to check which book it actually comes from.

Here is a fact that tends to surprise people: there is no Baize in the received text of the Classic of Mountains and Seas.

The claim that Baize comes from that classic can be traced to the Yuanjian Leihan, an encyclopaedia compiled under the Kangxi emperor. Under its entry for Baize it marks the content as quoted from the Classic of Mountains and Seas — but read the received classic from cover to cover and the passage is not there. A mislabelled cross-reference is the likeliest explanation, rather than something the classic ever contained. An encyclopaedia mistypes a source once, and three centuries of readers copy it faithfully.

The earliest reliable record of Baize is in the Baopuzi by Ge Hong (Eastern Jin, fourth century), and it is one sentence: the Yellow Emperor, having run the spirits and their wiles to ground, “recorded the words of Baize.” Those few characters fix its position. Baize does not step forward to foretell. Baize records what is hard to tell apart.

The fullest account is in the Yunji Qiqian (Song dynasty), quoting the Chronicle of Xuanyuan: the Yellow Emperor, touring east to the sea, met the beast Baize at the shore. It could speak and understood the nature of all things. Asked about the spirits of the world, Baize named 11,520 kinds — every being formed from essence or transformed from a wandering soul — and the emperor had them drawn as pictures and shown to the realm.

Set the key sources side by side and they point to one sentence:

Source	Text	What it establishes
Yunji Qiqian, quoting the Chronicle of Xuanyuan (Song)	“It could speak, and understood the nature of all things”	The most-cited definition
Song shu, Treatise on Auspicious Portents (Liang)	“It could speak and knew the essence of all things, to warn the people, to remove harm in their time”	Its function made explicit: to warn, to remove harm
The Auspicious Portents Chart (quoted in the Ming Jili, ch. 43)	“It arrives when the ruler sees far”	The condition of its appearance: it comes to those who see far
The Ruiyingtu (quoted in the Tang Kaiyuan Zhanjing, ch. 116)	“It appears when a worthy ruler’s virtue reaches the remote”	A second record of the same condition
Baopuzi, “Dengshe” (Ge Hong, Jin)	“Know the names of all spirits … and the spirits withdraw of themselves”	The doctrine: to know the name is to know the answer
Ming Huidian / History of the Ming, treatise on carriages and dress	“Dukes, marquises, imperial sons-in-law and earls wear qilin and baize”	The Ming rank-badge regulations

The Baopuzi states the principle most plainly: “know the names of all spirits … and the spirits withdraw of themselves.” To know the name is already to know the response. The core of the whole Baize tradition is recognition, not prediction.

One thing gets exaggerated, so let us be precise: what Baize knows is the name and nature of a thing, not the shape of the future. What people asked of it was “what is this thing in front of me,” not “what will happen to me next year.” Those two questions are often merged in discussions of Chinese divination; they are not the same question.

The pictures the emperor had drawn were compiled into the Baize Chart, also called the Baize Jingguai Tu. It is not a book about fortune. It works like a reference manual:

Something strange happens that you cannot name → open the chart → find its name → know what to do

In medieval China nearly every household kept one, hung on a wall or pasted by the door. A line preserved in Chan Buddhist recorded sayings says exactly this: “with a Baize chart at home, anomalies dissolve on their own.” What the chart does is turn the state of “not knowing how to react” into an action you can take: recognise it first.

Most of the original is lost. What survives is a Dunhuang fragment from the ninth or tenth century, now split between the Bibliothèque nationale de France and the British Library. The same usage travelled to Japan and lasted longer there: the Hakutaku Hikai Zu, hung in Edo-period homes, lists eighteen household anomalies in two columns, each with a name, with the image of Baize at the foot. The method never changes: speak the name.

Two objects show where Baize stood later on — a banner and a robe, both reserved for the emperor and the titled nobility.

The banner. The baize banner was on the imperial establishment from the Tang. The New History of the Tang is blunt: “the vanguard squad bears two baize banners, each carried by one man wearing a horizontal sword.” The vanguard is the unit that clears the road for the emperor, and the baize banner rode at its head. The Ming entered it in the Grand Procession, as chapter 182 of the Collected Statutes sets out; the Qing kept it, and the Draft History of the Qing lists the baize banner among the banners of the emperor’s Grand Procession, used for the three great sacrifices. The emperor’s own, only on the heaviest occasions, and at the very front.

The robe. Under Ming regulations the baize insignia was not ranked among the garments of civil or military officials; it belonged to dukes, marquises, imperial sons-in-law and earls. In 1977 the tomb of Xu Fu, sixth Duke of Wei, yielded a silk robe with a woven baize insignia. It is a Grade-One national artefact, held by the Nanjing Museum Administration. Among formally published archaeological finds it is the only known surviving baize rank badge; qilin badges, by contrast, survive in several examples.

Xu Fu held his title for fifty-two years. His great-great-grandfather was Xu Da, the founding general of the dynasty, and the Dukedom of Wei ran from Xu Da in 1370 to Xu Wenjue, eleven generations, ending only with the dynasty in 1644.

What is rare is not the auspicious beast. It is the eye that recognises it. That is the layer of the name we took.

To us, “Baize” is the name of a sequence of moves: recognise first, decide second. To see what that sequence looks like on you, all you need is a birth time — register and the chart generates directly, with no questionnaire to answer first.

Questions you can paste straight to Baize:

• In your system, which step does “Baize” actually correspond to — the chart, the question, or the review?
• I do not want “what happens to me this year.” I want to see my own structure first. Where would you start reading?
• How do you draw the line between recognition and prediction?

It answers from your chart, not from a generic FAQ.

To see that line drawn in full, read the method and boundaries. First time on Baize and not sure which step comes first? The full order — chart, question, review, coming back — is in First Time on Baize: Where to Start.

Does Baize really come from the Classic of Mountains and Seas?

No. The received classic contains no Baize. The claim can be traced to a cross-reference in the Qing encyclopaedia Yuanjian Leihan, but the text it points to is not in the classic. The earliest record is the Baopuzi by Ge Hong.

Can Baize predict the future?

In the tradition it recognises: it names what is in front of you, its nature, and how to respond. It does not foretell. The classical function given to it is “to warn the people, to remove harm,” which lands on the present.

Is the Baize Chart still available?

Most of the original is lost. A Dunhuang fragment survives, split between the Bibliothèque nationale de France and the British Library. Japan preserved the usage more completely; the Edo-period Hakutaku Hikai Zu is still to be seen.

How is Baize told apart from the qilin and the xiezhi?

They are often confused. In Ming rank-badge regulations, dukes, marquises, imperial sons-in-law and earls wore qilin and baize, while censors and investigating officials wore the xiezhi. Descriptions of Baize vary; the rule most often used is that the qilin has hooves and the baize has claws.

What does the name Baize Mingdao have to do with Baize?

It takes the action, not the myth. Baize recognises first and decides second, which is exactly how we want the tool used: read your structure first, then test a decision against it.

By borrowing the name, are you claiming to see more clearly than others?

No. We borrow a method, not a myth. We do not promise predictions; what we publish is the boundary — what this tool answers well, and where it fails.