Quick Answer
The Shanhaijing is the most important surviving source for early Chinese mythology, and it was compiled as a practical inventory of terrain, minerals, useful animals and the sacrifices owed to local gods. The myths are incidental to the filing system, which is exactly why they survived.
Key Takeaways
- It reads like a gazetteer, not a story collection. Over five hundred and fifty mountains and three hundred waterways, entry after entry, in a fixed format.
- Hundreds of entries end with a use. Eat this and you will not suffer that, which makes the text an ancestor of Chinese materia medica.
- The Queen Mother of the West appears here first, with a leopard's tail and tiger's teeth, presiding over catastrophe. The serene goddess came much later.
- Chinese bibliographers could never place it. Filed as geography, later reclassified as fiction, and still argued over.
- The famous illustrations are Ming. They postdate the text by well over a thousand years.
🕑 14 min read
What It Is
The Shanhaijing, usually translated as the Classic of Mountains and Seas, survives in eighteen sections covering more than five hundred and fifty mountains and around three hundred waterways.
It is composite. The earliest layers go back to roughly the fourth century BC and the received form was not settled until the early Han. Liu Xin edited and presented it at the Han court, attributing it to Yu the Great, the legendary flood-controller, and his minister Boyi. Guo Pu, in the fourth century AD, wrote the commentary that shaped how the text has been read ever since and supported the old attribution.
Nobody now accepts that attribution. What we have is an anonymous accumulation, assembled over several centuries, by people whose purposes we can partly infer from the format and cannot recover directly.
The division that matters: the first five sections, the Classics of the Mountains, are the systematic core and the most consistent. The remaining thirteen, dealing with regions beyond, are looser, later in parts, and considerably more fabulous. A reader who finds the far sections wild should know that the near ones are much more sober.
How It Actually Reads
Anyone approaching this book expecting a mythology will be startled by the first page, so it is worth setting expectations precisely.
The format is a survey. It moves along a mountain range in a stated direction, and for each peak it gives roughly the same information in roughly the same order: how far it is and in which direction from the previous one, what river rises there and where that river goes, what trees and plants grow on it, what minerals and jade are found, what animals live there and what they look like, and what happens if you encounter or consume them.
Then it moves to the next mountain and does it again. And again, for hundreds of entries.
The animals are where it turns strange, and the strangeness is delivered in the same flat register as the geography. A creature like an ox with four horns, whose appearance means a flood. A bird with three heads and six eyes. A fish with the face of a human that makes a sound like a mandarin duck. These are not presented as marvels. They are presented as local fauna, in the same sentence structure as the timber.
That flatness is the most interesting thing about the book and the hardest to convey in summary. Nothing in the text separates the plausible from the impossible, because the compilers were not sorting on that axis. A mountain has iron ore, a certain herb, a bird whose call sounds like its own name, and a nine-tailed fox, and all four are recorded in the same voice for the same reason: someone travelling there should know.
The Eat-It Formula
The single most repeated element in the text is a statement of use, and it appears in hundreds of entries.
Eat this fish and you will not suffer from boils. Wear this plant at your waist and you will not become confused. Eat this bird and you will not have nightmares. Eat this creature and you will not be jealous. Carry this and you will not fear thunder.
The formula is remarkably consistent, and it points at what the text was for. This is not wonder-literature. It is closer to a field manual: an inventory of what a stretch of country holds, including its medicinal and apotropaic ones, organised so that the information can be found by location.
Read that way the Shanhaijing sits at the head of a long Chinese tradition of pharmacological cataloguing that runs through the classical materia medica literature and eventually to Li Shizhen's great compendium in the sixteenth century. Our guide to Chinese herbal basics covers where that tradition ended up.
Some of the effects are physical, some psychological, and some social. Eating a certain creature to avoid jealousy is a claim about character rather than the body, and the text draws no line between the categories.
What You Owe the Mountain
Each group of mountains closes with a passage that modern readers skip and that is probably the key to the whole work.
The compilers state the ritual owed to the gods of that range. What the gods of these mountains look like, often as composites with human faces and animal bodies. What is to be offered: which animal, of what colour, whether a cock or a pig or a sheep. What jade is used, whether it is buried, whether rice is offered, whether the offering is made whole or divided.
The specificity is administrative. This is not devotional writing. It is the sort of information a person responsible for performing sacrifices at the correct place in the correct manner would need to have written down.
Which suggests a purpose for the whole compilation. A state or a priesthood maintaining ritual relations with the gods of a large territory would need exactly this: where the mountains are, who lives on them, what they are called, and what they require. The natural history and the mythology are the surrounding context for a ritual gazetteer.
The Mythology Nobody Meant to Preserve
Embedded in the inventory is the earliest surviving version of a great deal of Chinese mythology, and the accidental preservation is worth appreciating.
Kua Fu chased the sun. Overtaking it, he was seized by thirst, drank the Yellow River and the Wei dry, set out north for a great lake and died on the way. His abandoned staff became a forest of peach trees. The story appears in a few lines, without moral or explanation.
Jingwei was the daughter of the Fiery Emperor, drowned while playing in the Eastern Sea. She became a bird, and the bird carries twigs and stones from the western mountains and drops them into the sea, attempting to fill it. The phrase for hopeless persistence in Chinese still refers to her.
The ten suns bathed in a hot spring and perched in a mulberry tree, one going out each day in turn.
Gun and Yu and the great flood: Gun stole the self-renewing soil from the high god to dam the waters, was executed for it, and his son Yu completed the work by channelling the water instead of blocking it.
None of these is told as a story. Each is a note attached to a place or a creature, in the same voice as the mineral deposits. Later Chinese literature took these fragments and developed them enormously, and this is where the record begins.
The Queen Mother Before She Was Beautiful
The single most valuable entry in the book, for anyone interested in how deities change, concerns Xiwangmu, the Queen Mother of the West.
In later Chinese religion she is a magnificent and benevolent goddess, the highest female deity of the Daoist pantheon, keeper of the peaches of immortality, receiving emperors and immortals in her paradise on Mount Kunlun. She appears constantly in poetry, painting and popular religion in that form.
In the Shanhaijing she has a human shape with a leopard's tail and tiger's teeth. She is good at whistling or howling. Her hair is wild and she wears a headdress. And her office is stated plainly: she presides over catastrophes sent from the sky and over the forces of destruction.
That is a plague and disaster deity, and she is frightening. The transformation from this figure to the serene Queen Mother of later tradition took centuries, and it is documented almost entirely because a gazetteer recorded what she looked like before anyone had a reason to improve her.
This is why the Shanhaijing is worth more to a student of religion than a polished mythology would be. A myth collection preserves the version its compiler found meaningful. An inventory preserves what was there, including the parts that later tradition would rather have edited. The unrevised, unattractive, administratively recorded Queen Mother is a piece of evidence no devotional text would have handed down.
The Far Countries
The later sections leave the surveyed mountains and describe regions beyond, and the further out they go the stranger the inhabitants become.
There is a country of people with one arm and one eye. A country whose people have a hole through the chest, through which a pole can be passed to carry them. A country of people with very long arms and another of people with very long legs, who work together at fishing. A country of women, who conceive by bathing in a certain river. People whose heads have been struck off and who fight on with faces in their torsos.
Read as ethnography this is worthless. Read as a record of what the compilers of a Chinese text believed lay beyond the known world, it is a valuable document about the edges of a mental map, and it belongs beside the comparable material in Greek and Roman writers about the margins of their own world. The pattern of placing the monstrous at the far edge of the known is close to universal, and this is one of the fullest surviving examples.
Reading the Animals as Signals
One feature of the creature entries is easy to miss and changes how the whole book works.
A great many animals are recorded not for what they are but for what their appearance means. This bird, when it appears, means the region will suffer drought. This beast, when seen, means there will be a great flood. This creature appearing in a state means that state will experience war, or a plague, or an abundance of grain.
So the catalogue is doing double duty. It is a record of what lives where, and it is a key for reading events, in the way that later Chinese omen literature and the enormous apparatus of portent interpretation in the Han dynasty would develop at length.
This connects the text to a way of thinking that runs through early Chinese thought and is quite alien to modern habits. The natural world is not a neutral background against which human affairs happen. It is legible, and an unusual animal is information about the state of the realm in the same way a comet or a strange birth would be.
Once you notice the omen entries, the mixture of ore, herb, beast and god stops looking miscellaneous. Everything in the entry is something a responsible person needs to know about that place: what it will give you, what will hurt you, what it foretells, and who must be paid.
Practice: write your own entry
Take one place you know well and write it up in the Shanhaijing format. Distance and direction from the last place. What water rises there and where it goes. What grows. What can be taken from the ground. What lives there and what it looks like. What it is good for. What appears there when something is about to change. And what, if anything, is owed to it. The exercise takes twenty minutes and it will show you which of those six questions you have never asked about somewhere you have lived for years.
A Book Nobody Could File
Chinese bibliographers have struggled with this text for two thousand years, and the struggle is itself informative.
Early catalogues placed it under geography, which is what its format claims to be. Later, the compilers of the great eighteenth-century imperial library, the Complete Library of the Four Treasuries, moved it into fiction, presumably on the reasonable ground that geography is not supposed to contain nine-tailed foxes.
Modern scholarship divides three ways, reading the text primarily as mythology, primarily as fiction, or primarily as geography, with the geographical reading currently the most common. Serious efforts have been made to identify its mountains and rivers with real features, with partial success in the core sections.
The classification problem is not a defect in the text. It is a consequence of applying categories the compilers did not have. The distinction between a natural history and a mythology assumes a prior distinction between the natural and the supernatural, and the Shanhaijing was assembled by people who were not sorting the world that way.
Which Translation to Read
Anne Birrell, Penguin Classics (2001). The complete text in an affordable edition with introduction and notes. Some sinologists have criticised particular renderings, especially of creature names, and it remains the standard general-reader complete version.
Richard Strassberg, A Chinese Bestiary (2002). A selection rather than the whole, focused on the creatures, reproducing the Ming woodblock illustrations, with annotation of a much higher order. If you want to understand what you are reading, this is the better book, and it is not complete.
Sun, Chen and Goldblatt (2021). A recent complete translation, with Howard Goldblatt's involvement bringing considerable experience of Chinese into English.
On the cheap reprints: inexpensive print-on-demand volumes under this title are usually old partial translations out of copyright, often without notes. This is a text where the annotation carries most of the value, since an unglossed entry about a beast that resembles a badger with one eye and three tails is close to meaningless without a note explaining what is known and what is guessed. Buy Birrell or Strassberg.
A note on the pictures, since they sell many editions: the famous images come from Ming dynasty woodblock printings and later Qing colour versions. They are more than a thousand years younger than the text and they are illustrations of it, not evidence about it.
Who Should Read It
Read it if you want Chinese mythology at its source rather than in retellings.
Read it if you are interested in how ancient people organised knowledge, because the organising principle here is genuinely foreign and genuinely coherent.
Read Strassberg first, then Birrell if you want the whole thing.
Do not read it front to back. It is a reference work and it defeats sequential reading by design. Do not use it to argue that ancient Chinese explorers reached America, a claim associated with Henriette Mertz that requires very selective reading and has no support among specialists.
Our position: an inventory that accidentally became the foundation of a mythology, and is more valuable for being an inventory. The flat administrative voice recording ore, timber, a healing fish and a disaster goddess in the same breath is a window into a way of organising the world that no later text preserves as well. Buy Strassberg for the annotated selection, keep Birrell for completeness, read the sacrifice passages that everyone skips, and look up the Queen Mother of the West before anyone made her beautiful.
For adjacent reading, see the Inner Chapters of Zhuangzi for a very different early Chinese imagination, the history of the I Ching for the other great composite Chinese classic, and the Enuma Elish for a text whose purpose is equally unlike what modern readers expect.
Five hundred and fifty mountains, and what each one owes.
Affiliate disclosure: Thalira earns a small commission on purchases made through the link above, at no additional cost to you. We recommend books we have read and think are worth your time, and we say plainly when we think parts of them are weak.
Frequently Asked Questions
What is the Classic of Mountains and Seas?
The Shanhaijing, a Chinese text in eighteen sections describing more than five hundred and fifty mountains and three hundred waterways, together with their minerals, plants, animals, spirits and required sacrifices. It was assembled by many hands between roughly the fourth century BC and the early Han dynasty.
Is it a book of mythology?
It is the single most important source for early Chinese mythology and it does not present itself as one. The format is a gazetteer, and the myths appear inside entries about terrain, useful minerals and local gods. Nobody set out to write a mythology; the mythology survived because it was catalogued.
Who wrote it?
Nobody knows. Liu Xin, editing it in the Han, attributed it to Yu the Great and his minister Boyi, and Guo Pu supported that attribution in his fourth century commentary. Later scholarship rejects it. The text is composite, assembled over several centuries by anonymous hands.
What is the eat-it formula?
A recurring pattern in which an entry names a plant or animal and then states its effect: eat this and you will not suffer from a particular complaint, or wear it and you will not become confused. Hundreds of entries end this way, which makes the text an early ancestor of Chinese materia medica.
What myths does it preserve?
Kua Fu chasing the sun and dying of thirst, the bird Jingwei carrying stones to fill the sea that drowned her, the ten suns, Gun and Yu and the great flood, and the earliest description of the Queen Mother of the West, who appears here with a leopard's tail and tiger's teeth rather than as the later serene goddess.
How did Chinese scholars classify it?
With difficulty. Early bibliographies listed it under geography. The compilers of the Complete Library of the Four Treasuries in the eighteenth century moved it to fiction. Modern scholarship is split three ways between reading it as myth, as fiction and as geography, with geography currently prevailing.
Are the illustrations ancient?
No. The famous images of nine-tailed foxes and headless torsos with faces on their chests come from Ming dynasty woodblock editions and later Qing colour versions, printed well over a thousand years after the text was compiled. They are illustrations of the text rather than part of it.
Does it describe America?
No. The claim, argued most prominently by Henriette Mertz in the twentieth century, maps the text's eastern lands onto North American geography. It requires selective reading and it is not accepted by sinologists or archaeologists. The text's distant regions become steadily more fabulous the further out they go.
Which translation should I read?
Anne Birrell's Penguin Classics version of 2001 for the complete text with introduction and notes. Richard Strassberg's A Chinese Bestiary of 2002 for a selection with the Ming illustrations and superb annotation. A more recent complete translation by Sun, Chen and Goldblatt appeared in 2021.
Is it difficult to read?
It is repetitive rather than difficult. The entries follow a formula and the cumulative effect over hundreds of them is closer to reading a catalogue than a narrative. Most readers do best dipping into it rather than working through in order.
Why does it matter for Chinese culture?
Because so many figures that appear later in Chinese poetry, painting, religion and now film and games make their first surviving appearance here. Later Daoist and popular tradition developed them enormously, and this is where the record starts.
What are the sacrifices in it?
Each group of mountains closes with instructions for the offerings owed to the gods of that range: which animals, what colour, what jade, whether buried or offered whole. These passages are the clearest evidence that the text had a practical religious purpose rather than a literary one.
Read one mountain properly
Open the book anywhere in the first five sections and read a single entry with full attention: the distance, the river, the trees, the ore, the creature, the use, the offering. Notice that you have just been handed a complete practical relationship with one piece of ground, covering what it yields, what lives there, what it can do to you and what you owe it. Very little modern writing about place attempts all four at once, and this text does it five hundred and fifty times.
Sources & References
- Birrell, A., trans. (2001). The Classic of Mountains and Seas. Penguin Classics.
- Strassberg, R. E., ed. and trans. (2002). A Chinese Bestiary: Strange Creatures from the Guideways Through Mountains and Seas. University of California Press.
- Sun, J., Chen, S. and Goldblatt, H., trans. (2021). Complete translation of the Shanhaijing.
- Guo Pu (fourth century AD). Commentary on the Shanhaijing, the basis of the received text.
- Liu Xin (Western Han). Editorial preface presenting the text at court.
- Siku Quanshu (Complete Library of the Four Treasuries), eighteenth century. Bibliographic classification of the text as fiction.
- Chinese Text Project. Text and editions of the Shanhaijing. ctext.org.
- Birrell, A. (1993). Chinese Mythology: An Introduction. Johns Hopkins University Press.