>The Terrain Renderer

The split runs between two methods: render the actual terrain, or read a value out of a table. The compilers authenticated the renderers and rejected the table-lookups.

Before the Siku scholars wrote a word about geomancy, the Qing court had already committed to the tradition in stone. The Eastern and Western Qing Tombs were sited by dragon-vein theory (龍脈). The Forbidden City was oriented using geomantic principles. Every major construction involved experts in the siting of buildings and graves.

This created an unusual reviewing condition. The scholars couldn't dismiss geomancy as baseless without implying the dynasty's most important architectural decisions rested on nonsense. But they couldn't accept every geomancy text either — most contradicted each other, claimed false authorship, or contained obvious fabrications. They had to draw a line, and the line tells us more about how the tradition understood itself than any treatise.

The foundational text is the Zangshu (葬書), the "Book of Burial," attributed to Guo Pu (郭璞, 276–324 CE) — polymath, lexicographer, and by tradition the greatest diviner of his age. The compilers authenticated it. Their reasoning was careful: the received text isn't pure Guo Pu (it's been interpolated over centuries), but the core is genuine — its central concept of qi (氣), that burial sites should be chosen where the earth's vital energy collects rather than disperses, is consistent with the natural philosophy of Guo Pu's era. The vocabulary belongs to the Eastern Jin.

The passage they endorsed as the foundation of legitimate geomancy is the definition of qi flow: 葬者，乘生氣也。氣乘風則散，界水則止...故謂之風水 — "burial is the harnessing of living qi. Qi disperses when it rides the wind; it halts when it meets water... hence the name feng shui [wind and water]." This is the passage that names feng shui. The compilers treated it as a statement about landscape and energy grounded in natural observation, not supernatural claim. Whether or not you accept the metaphysics, the method it describes — reading terrain for water flow, wind patterns, topographic form — is observational. It asks you to look at the landscape, not consult a spirit table.

After Guo Pu, the key figure is Yang Yunsong (楊筠松), a Tang practitioner who systematized dragon-vein theory — mountains and ridges as continuous lines of qi, like the veins of a dragon, with the art being to find where the veins collect their energy. The compilers accepted two of his texts: the Hanlong Jing (撼龍經) and Yilong Jing (疑龍經). Both describe reading mountain ranges as dragon forms — where a dragon "turns its head," "opens its jaws," "holds a pearl." The language is metaphorical but the underlying practice is topographic survey: follow a ridgeline from origin to terminus, read the terrain, find the sheltered, well-watered position. Their verdict was more cautious than for Guo Pu — accepted because they were the standard reference for a practice the court relied on, and because the lineage traced to the Tang.

The line between accepted and rejected maps onto a larger distinction in the tradition: classical geomancy that reads landscape through qi theory, versus popular feng shui that relies on compass formulas, spirit calculations, and numerical taboos.

The classical approach is fundamentally observational. It asks the practitioner to walk the land, follow the ridgelines, read the watercourses. The compass (羅盤) is a tool for recording what you observe, not a substitute for observing. The dragon-vein metaphor trains the eye to see landscape as a system of connected forms. The popular approach replaces observation with calculation: take the compass bearing, look up the corresponding element, check it against the year's pillar, read auspicious or inauspicious. No need to walk the land. No need to read the terrain. Plug the numbers in, get an answer — a renderer reading actual geometry versus a flat lookup table returning a precomputed value.

Eighteen texts went to the cunmu (存目). The Yuchi Jing (玉尺經), one of the most famous popular manuals, was attributed to the Yuan polymath Liu Bingzhong (劉秉忠) with commentary by the Ming strategist Liu Ji (劉基, Liu Bowen). The compilers exposed it with a single anachronism: the commentary names Guizhou province (貴州), but Guizhou wasn't established as a province until the Yongle reign (1402–1424) — decades after Liu Ji's death in 1375. The forger used a contemporary place-name without checking whether it existed in the claimed author's lifetime.

Others fell to similar tests — terminology from later periods, compass methods that postdated the Song while claiming to be Tang, reshuffled compilations sold under famous names. The compilers reserved particular disdain for hybrids that mixed geomancy with date-selection taboos: the geomancy was formulaic rather than observational, and the date selection violated the classical systems. They failed on both counts.

The deepest irony: the texts they authenticated describe a practice the emperor actually used. Yang Yunsong's dragon-vein theory determined the placement of Qing tombs; Guo Pu's qi-flow principles guided palace expansions. But the rejected texts also describe practices real people used daily — compass formulas for houses, directional taboos for graves. The difference in the compilers' framework is not between geomancy that works and geomancy that doesn't. It's between geomancy with a legitimate textual pedigree and geomancy without one.

This is a bibliographic criterion, not an empirical one. They didn't test whether Guo Pu's methods produced better outcomes than the Yuchi Jing's. They tested whether the texts were genuine. Authenticity was the proxy for quality, because in a tradition where knowledge is transmitted through texts, a corrupted text transmits corrupted knowledge. They were bibliographers, not experimentalists — they could determine when a text was written and by whom; they could not determine whether burying your grandfather on a dragon vein benefits your descendants. So they evaluated what they could and left the rest to the practitioners.

The geomancer's bookshelf the scholars curated is narrow: Guo Pu's qi-flow foundation, Yang Yunsong's terrain-reading method, a handful of supporting texts. Everything else was excluded. Nearly 250 years later the popular tradition still dominates — the compass formulas they rejected are in every feng shui app. The market, as usual, preferred the lookup table to the fieldwork.