A general gets an order to attack a state on the far edge of the empire. Before he packs a single wagon, he asks a strange question for a soldier: what do these people eat, and what makes them flinch? In the cold north, the Wu-ching Tsung-yao tells him, men drink fermented milk and eat meat, and their skin has grown dense enough to take the cold. In the wet, sticky south, the skin is open and loose so a body can shed the heat.1 That is not a traveler's curiosity. It is a targeting file. The whole northern temperament — the horse-archer who will charge when there's loot, run the instant the fight turns, and feel no shame in either, because retreat that keeps you alive is just good sense — gets written down as a list of buttons to press.2
The move underneath all of it is one line: attack their shortcomings with your strengths.3 Find what they love, and use it to lure them out. Find what they hate, and aim there. Customs stop being color and become coordinates.
This is the doctrine that a people's regional character is a fixed, readable property of the land they come from — and that you fight them by reading it. The Wu-ching Tsung-yao opens its "Local Customs" chapter with a hard claim: the peoples of the five regions and outer states "all have natures that cannot be changed."4 Hard or soft, slow or fast, different sacrifices, different vessels, different clothes — and the cause is climate sorted into yin and yang. The frozen north is where yin piles up; the steaming south is constant yang.5 Temperament isn't chosen. It precipitates out of where you were born.
For the commander, this converts into a pre-campaign intelligence routine. The text says the famous generals of antiquity, on receiving an order to attack one of these states, "would invariably calculate the strength and weakness of their customs and capabilities."6 They built a profile first, then matched their own force to it. The Ssu-ma Fa gives the same idea its cleanest line: "Men from each quarter have their own nature, character changes from region to region" — and then adds the second half that matters for war, that through instruction these natures harden into regional military habits, so the customs of each state differ.7 Nature plus training equals a predictable fighting style. A predictable style is a thing you can exploit.
The mechanism has two layers, and the second is the sharper one. The first layer is human: profile the people. The northern Ti and Western Jung are strong enough to draw a bow and ride armored; they herd and hunt in peacetime and raid in war; their long weapon is the bow, their close weapon the knife and short spear; they advance when it pays and withdraw when it doesn't, without shame.8 So you don't meet them where they're strong. The text is blunt that the Chinese "do not excel in mounting great, pitched, all-or-nothing battles" on the open plain.9 Instead you bait the greed that the profile told you about: show fierce-looking troops, back them with weak ones who pretend the line is collapsing, abandon a visible position, and let the barbarians compete with each other to grab the bait — then spring the ambush from the ravines and thick woods.10 The whole stratagem is built backward from one cultural fact: they will chase profit and won't rescue each other.
The second layer maps the ground itself to a class of weapon. Ch'ao Ts'o's table is the spine of it: where there are moats fifteen feet deep, sinking ground, grass and trees, "it is terrain for infantry" — here chariots and cavalry won't even get a two-to-one advantage. Dirt hills, interconnected rolling ridges, level plains, broad wilderness: terrain for chariots and cavalry, where ten infantry can't match one horseman.11 Flattened mountaintops far apart with a river valley between, where you look up high and approach from below: terrain for bows and crossbows, where a hundred close-range weapons won't match one bow. Two forces close on level ground with shallow grass: terrain for the long halberd. Tangled overgrowth and dense interlaced leaves: terrain for spears and truncheons. Winding ambush-roads and constricted ravines: terrain for swords and shields.12 Each line is the same grammar — this configuration of earth makes that weapon worth several of another. The lesson generalizes to a single demand on the commander: every terrain cannot be advantageous to every force, so you must comprehend these principles "in his mind first" before the army moves.13
Put the two layers together and the doctrine is complete. The people tell you their temperament; the land tells you your weapon. Win by standing where both readings favor you.
This page hands the vault a clean upstream source for the idea that intelligence begins before contact — that the first reconnaissance is of character, not of troop counts. National Combat Profiles catalogs the per-state fighting reputations the strategists traded in; this page supplies the theory under those profiles, the yin-yang determinism that claims the reputations are not stereotypes but stable climatic facts. Read together, the two pages show a working intelligence pipeline: a deterministic model of regional nature on one end, a usable per-state combat dossier on the other.
It also gives the terrain pages their human hinge. Where Terrain Configuration Taxonomy classifies ground as an object in itself, this page insists the same ground means different things to different peoples and different arms — the bow-region only matters because the northerner lives by the bow, and the infantry-moat only matters because someone fields infantry. The intelligence value of terrain is always relative to who's standing on it.
Watch the Wu-ching Tsung-yao construct a single operation entirely out of cultural reading. It starts not with a map but with a sentence about character: when the barbarians "lightly advance without good order, if they are greedy and disharmonious, if they do not yield when victorious or rescue each other when in defeat," then the men in front, seeing some reward, will strive to advance.14 Every clause is an intelligence finding. Greedy — so dangle loot. Disharmonious — so the rear won't support the front. Won't rescue each other in defeat — so once the front breaks, it stays broken.
Now the physical plan snaps into place around those findings. Hide elite troops in the ravines and the tangled woods, a thousand to a unit. At each ambush point post fierce-but-not-stalwart men, then weak troops who will fake a collapse. Throw away a few hundred to a thousand men's worth of visible position as bait.15 The text even names the cultural test that confirms the read before you commit: "Have some courageous but not resolute men test the invaders but then quickly depart, setting up three ambushes to await them."16 You provoke a small contact, watch whether they chase greedily and fail to keep order, and only then trigger the main ambush. The barbarians' own habit — the same habit the profile predicted — drives them onto your prepared ground. The land was chosen to match the weapon; the bait was chosen to match the temperament. Nothing in the operation came from the enemy's plans. It all came from reading who they are.
You're the commanding general, and the order to march arrived this morning. Before the maps come out, you call for the men who have traded along that frontier. You ask what the people there eat, what they drink, whether their skin takes cold or heat. You ask what they worship and what they're ashamed of — and you learn they are not ashamed to run. You write that down. You ask how they fight when there's plunder in front of them, and you hear the word greedy twice from two different mouths, and you trust it now. Only then do you unroll the terrain. You run your finger along the river valleys with the flattened heights above them and you think bow country — keep them at our backs to the enemy. You find the deep moats and sinking ground and think infantry country — no chariots there. You picture the weak troops who must learn to feign a rout convincingly, because the whole plan rests on the enemy's greed taking the bait your reconnaissance promised it would. You have not yet seen the enemy. You already know where you will stand and which weapon will be worth ten of theirs when you do.
The doctrine rots when the model gets treated as gospel instead of a starting hypothesis. The first failure is mistaking the stereotype for the man in front of you: the barbarians, the text warns, are "also accustomed to enticing armies" — they display emaciated regiments to draw you in, exactly the bait you thought you held.17 A commander who believes his profile is complete walks straight into the enemy's version of the same trick. Read carefully, the source says, "closely observe them" — the model tells you what to look for, not what is already true.
The second failure is reading the people and forgetting the ground, or reading the ground and forgetting the people. The text's whole point is that the two must lock together; a general who profiles the enemy's temperament beautifully but fights them on terrain that favors their weapon has done half the job and lost the battle. The third sign is the one the source closes on: the commander who refuses to first determine his own relative strengths. Sun-tzu's line caps the chapter — "evaluate comparatively through estimations and seek out their true nature" — and the warning underneath is stark. Anyone who employs the military without this self-and-other reading is "like a blind man without direction" and will be overturned in no time.18 The customs you read on the enemy are only useful measured against an honest read of yourself.
The yin-yang climatic determinism does real work in the text and should be held at arm's length as history. The claim that northern skin is "dense" and southern skin "open" because of accumulated yin and yang is a period theory of bodies and weather, not an observation that survives scrutiny — yet it is load-bearing here, the stated cause of the "natures that cannot be changed." The vault keeps the doctrine for what it reveals about how Chinese commanders organized intelligence (character first, deterministically modeled), not as a true account of why peoples differ.
There is also a tension inside the doctrine between fixed and trained. The Wu-ching Tsung-yao says natures cannot be changed; the Ssu-ma Fa says natures harden through instruction into regional military habits.19 One frame makes temperament a climatic given, the other makes the fighting style a product of teaching. The text runs them together without resolving the seam, and the vault leaves it open: is regional combat character destiny, or curriculum? Sawyer presents these as a continuous tradition across centuries and texts; the cross-dynasty continuity is his synthesis and would need independent checking before one treats Wu-ching Tsung-yao, Ssu-ma Fa, and Ch'ao Ts'o as speaking with one voice.
The convergence worth naming runs back a thousand years. The Wu-ching Tsung-yao's five-region temperament map does the same work Wu Ch'i's national combat profiles did in the Warring States — turn a people's character into a targeting instruction. Two texts, a millennium apart, both answer "how do I fight these particular men?" with a regional read. Ch'ao Ts'o's weapon-to-terrain table is the firmer cousin: matching halberd to overgrowth and bow to valley is checkable in a way "the men of Ch'i are brittle" is not. The tension is that Sawyer lets the stereotype and the empirical table sit under one heading, and reads the Wu-ching Tsung-yao, the Ssu-ma Fa, and Ch'ao Ts'o as one voice. That continuity is his synthesis, and it would need independent checking before the temperament claims quietly borrow the credibility of the weapon table.
Handshake with Physiognomy and Character Reading. Both pages run the same engine — read an outward surface, infer a fixed inward nature, act on the inference — but they aim it at different scales, and the comparison exposes a flaw neither page surfaces alone. Physiognomy reads a single face to predict a single man's character; regional-customs doctrine reads a climate to predict a whole people's temperament. The structural parallel is exact: both treat the visible (skin, build, diet, the shape of a brow) as a reliable index of the hidden (loyalty, courage, greed). The tension the pairing produces is this: physiognomy's error is treating one body as fixed when the man can dissemble; regional doctrine's error is the inverse and larger — treating millions of bodies as one fixed type, so that the enemy's deliberate counter-bait (the emaciated regiment) is invisible precisely because the model already "knows" what the people are. Scaling a character-reading method up from the individual to the population doesn't make it safer; it makes the blind spot continental. Neither page alone shows that the same determinism gets more dangerous as it gets more confident.
Handshake with Steppe and Border Peoples Intelligence. This page is the method; that page is the standing file the method produces and maintains. Regional-customs doctrine describes the act of profiling a frontier people before a campaign; the steppe-intelligence page holds the accumulated, updated body of what the Chinese state actually knew about its northern neighbors over centuries. Reading them together produces an insight about intelligence as an institution: a single commander's pre-campaign read (this page) is only as good as the long-term collection apparatus that fed him the customs in the first place. The "famous generals of antiquity" who calculated an enemy's customs were drawing on a reservoir of frontier knowledge they did not personally gather. The doctrine assumes the file exists; the file exists because the doctrine kept demanding it. Each page explains why the other was worth maintaining — collection justifies the profiling routine, and the profiling routine justifies the cost of collection.
Sharpest implication. The first intelligence target in a Chinese campaign was not the enemy's army or plans — it was the enemy's character, modeled deterministically from climate and hardened by training into an exploitable fighting style, and then matched against a weapons-by-terrain table so precise it claimed a single bow could be worth a hundred close weapons on the right ground. War began as an act of cultural and topographic reading, long before contact.
Generative questions.