Classic of Mountains and Seas

The eighteen surviving books fall into groups. The first five, the Classic of the Mountains, take up roughly half the text and work through mountain ranges in five directions. The format is relentless and formulaic: so many leagues in a given direction lies such a mountain, on it grows a certain tree or is found a certain jade or ore, in its waters live such fish, and the god of these mountains has a particular form and takes a particular sacrifice.

The remaining books cover the seas and the outer regions: Classics of the Seas Beyond and Within, of the Great Wilds, and a final book on the lands within the seas. Here the register shifts. The formulae loosen and the entries carry narrative, so most of the famous mythological material sits in this second half.

The contents are not sorted into the credible and the fabulous, which is a large part of the point. A mountain bearing recognisable iron and cinnabar is described in the same voice as a country of people with holes through their chests who are carried about by poles, or a beast whose appearance portends drought. Many entries end with a practical note: eat this fish and you will not have boils, wear this plant and you will not be confused, this creature's cry announces war.

Estimates of what it holds run to hundreds of mountains, hundreds of named waterways, and several hundred creatures and deities. No narrative thread connects any of it, and no figure is introduced or followed through. The book is a list, and its myths arrive as incidental detail inside a gazetteer.

There is no author. The text is a compilation, and its sections differ in language, in the geography they assume and in the states they treat as contemporary, which places their composition across a long span, roughly the fourth century BCE through the Han. The mountain books are generally taken to be earliest, the Great Wilds books latest. Whether it originated as a shamanic manual, a ritual geography for imperial administration, or a text written to accompany pictures is disputed and not resolvable from the evidence.

The first fixed form is Han. Liu Xin, working in the imperial library at the turn of the era, presented an edited version and argued for its antiquity, attributing it to Yu the Great and his minister Boyi after the taming of the flood, a claim no modern scholar accepts. The text he passed on was in eighteen books, which is the number that survives, though earlier bibliographic notices give different counts.

The commentary that made the book usable is by Guo Pu, who lived from 276 to 324 CE and was also a poet, diviner and lexicographer. His annotations gloss the vocabulary, identify creatures where he could, and preserve variant readings and lore that would otherwise be lost; every later edition rests on his. Later scholars, notably Bi Yuan in the eighteenth century and Hao Yixing in the nineteenth, produced the critical editions still cited.

Guo Pu also wrote verses on the illustrations, and the received text refers to pictures at several points. Those original images are lost. The elaborate woodcut illustrations found in Ming and Qing editions, which supply most modern visual impressions of the book, were made more than a thousand years later.

Chinese readers have never agreed what kind of book this is. The Han bibliography classed it under forms and prognostication. The great Tang and Song catalogues moved it about, and the imperial Siku quanshu of the eighteenth century eventually filed it under fiction, having decided it was not reliable geography. Sima Qian, writing his history around 100 BCE, said plainly that he would not venture to speak of the things in it. That verdict, from the most authoritative of early historians, tells you the credibility problem was recognised at once.

Yet it was used. Physicians and alchemists mined it for materia medica, since so many entries specify what a plant, mineral or animal does to the body. Officials and geographers took its mountain sections seriously enough to attempt identifications of real ranges and ore deposits, and some of those identifications hold. Poets and painters took it as a storehouse of images, and it stands behind a long line of Chinese strange-tales literature. Modern readers have added their own uses, from claims that its distant lands describe real foreign territories to its wholesale absorption into video games, film and fantasy.

Its indispensable role is different from all of these. Early China produced no systematic mythography, and most of what would have circulated orally is gone. The Shan Hai Jing is where the surviving fragments happen to be recorded: Xiwangmu, the Queen Mother of the West, who appears here in her earliest form with tiger's teeth and a leopard's tail, an epidemic-dealing power quite unlike the benign goddess she later became; Kuafu, who chased the sun until he died of thirst; Nuwa; and Jingwei, the drowned girl transformed into a bird that carries twigs and pebbles to fill in the sea that killed her. Without this compilation, most of that would be lost.