A rebel army is hammering a city. The king of that city begs for rescue. When that fails, he goes over the general's head — and the emperor himself, the man who can have the general's family executed, sends a direct written order: go save the city. The general reads it, and does nothing. He keeps building his walls higher. That night his own soldiers panic and start fighting each other right outside his tent, and he is asleep through it and does not get up. The whole court expects him to ride out in a blaze of rescue. Everyone is watching for the dramatic charge. And the surprise — the genuinely shocking move, the thing no one saw coming — is that he refuses to move at all.
That is Chou Ya-fu in 154 BCE, and it is the strangest entry in the entire Chinese deception canon, because the deceptive, unexpected, war-winning maneuver is the absence of a maneuver.1 He won by disobeying an order to act.
The Revolt of the Seven States was a coordinated rebellion — Wu, Ch'u, and five other kingdoms against Emperor Ching of the Former Han.1 Chou Ya-fu, appointed Defender-in-Chief, was sent east to crush it. What he actually did, and what makes this page worth writing, is a strategy with three components that each violate the obvious:
And the paradox that gives the page its spine: the Hundred Unorthodox Strategies and Li Ch'üan disagreed about what this even was. Li Ch'üan cited it as an example of refusing the unorthodox. Sawyer argues it is itself the deepest kind of unorthodox — because the unorthodox is defined against expectation, and here everything dramatic was what was expected, so the patient, disobedient, sacrificial defense became the surprise.1
The whole page turns on one idea, and it is the same idea that runs the entire orthodox/unorthodox tradition: ch'i (the unorthodox) is not a type of action. It is a relationship to expectation. A flanking column is unorthodox only when the enemy expects a frontal assault. If the enemy expects the flank, the frontal assault becomes the unorthodox move.
Now apply that to Chou Ya-fu. Standard reading: the unorthodox is the bold, the deviant, the surprise attack. So defense — patient, walled-up, refusing battle — looks like the most orthodox thing imaginable, the opposite of ch'i. That is exactly how Li Ch'üan read it: a refusal of the unorthodox.1
Sawyer flips it. What was expected in 154 BCE? Everyone — the king of Liang, the court, the emperor, the whole instinct of the age — expected aggressive rescue. The dramatic relief of a besieged ally was the orthodox move, the one the situation seemed to demand. So the surprising move, the one the enemy and the court could not predict and could not counter, was the refusal: hold, starve them, sacrifice Liang, defy the emperor. "Chou's actions came to furnish a prime illustration of adopting an attritional, defensive posture that can similarly be deemed unorthodox because far more dramatic actions were ordered and expected."1
This is the load-bearing insight of the page and it generalizes hard: when the entire room expects action, inaction is the unorthodox move. The deviant is always defined against the expectation, and sometimes the expectation is the charge.
This page is the clearest test case in the vault for the proposition that the unorthodox is a relationship, not a tactic — so it bolts directly onto the foundational duality.
The campaign reads as a series of refusals, each more startling than the last.
It opens with Chou Ya-fu's plan, stated to the emperor before he marched: "Ch'u's soldiers are nimble and light so it will be difficult to engage their front in battle. I would like to employ the minor state of Liang as a lynch pin. If we then sever the enemy's supply lines, we can gain control over them."1 Note what is in the plan from the first sentence: Liang is named as a "lynch pin" — a piece to be spent. The sacrifice is not a battlefield accident. It is doctrine, declared in advance, and the emperor assented to it.
Chou proceeded to the assembly point at Jung-yang. Wu had already attacked Liang, and the king of Liang asked for rescue. Chou marched northeast — but not to Liang. He went to Ch'ang-yi, built solid fortifications, and set his defenses.1 The king of Liang sent an emissary. Chou "simply maintained his defenses and made appropriate improvements without going out to rescue them."1
The king of Liang then escalated, submitting a memorial directly to Emperor Ching, who in turn ordered Chou to rescue Liang. Here is the line that makes the episode famous: "Chou did not obey the mandate, but instead strengthened his walls without going forth."1 He defied his sovereign. His justification was the principle that a field commander retains independence — a doctrine the Hundred Unorthodox Strategies and Sun-tzu both license, but which is a capital risk to actually invoke against a Han emperor who could erase your bloodline.
What he did do was send the marquis of Kung-kao and others with light cavalry to sever the supply routes behind Wu and Ch'u.1 Soon the rebel armies lacked provisions, grew hungry, and wanted to withdraw. They tried several times to provoke battle. Chou always refused.1
Then two small, telling episodes. First, the night panic: Chou's own soldiers were frightened and disordered, fighting each other outside his tent. Chou was sleeping soundly and did not get up — and the disorder "shortly settled down by itself."1 A commander's stillness as a stabilizing force; his refusal to react to internal alarm was itself a kind of discipline broadcast downward.
Second, the feint test: Wu's forces raced toward the southeast corner of the walls. Chou ordered his troops to prepare the northwest instead. When they had finished, Wu's soldiers indeed raced to the northwest — and failed to gain entrance.1 Chou had read the feint as a feint. The visible southeast rush was the zheng meant to draw his defense; the real ch'i was coming northwest; Chou allocated his strength to the unorthodox direction before the enemy committed to it. This single paragraph is the orthodox/unorthodox duality played out in miniature inside a defensive posture that Li Ch'üan called a refusal of the unorthodox — which is precisely Sawyer's irony.
Then the resolution: famished, Wu and Ch'u withdrew. Chou dispatched his elite units to pursue and "extensively destroy them."1 The king of Wu abandoned his army, fled south of the Yangtze to Tan-t'u with a few thousand stalwarts. Chou issued an edict — a thousand catties of gold for the king of Wu's head — and in slightly more than a month a native of Yüeh beheaded him.1 The whole defense and pursuit lasted seven months. Wu and Ch'u were completely pacified.
The case is so clean because the enemy chose his own defeat. Sawyer notes the rebels had rejected better plans: T'ien Po-lü urged a secondary unorthodox force; a young general urged a segmented strike at Luoyang to seize the empire's "lynch pin" while avoiding mismatched battles between their infantry and the Han's chariot-and-cavalry forces.1 Both plans were rejected as too rash, the king's son fearing that fragmenting their power would let a subcommander grow too strong.1 So the rebels marched west en masse — the orthodox, expected, undeceptive approach — straight into a wall that would not fight them and a cavalry that ate their supply. The side that refused the unorthodox was, in the end, the rebels.
The whole world is telling you to act. The crisis is loud, your ally is screaming for help, and your boss — the one who can end you — has put it in writing: do something dramatic, now. Every instinct in the room, including yours, says charge.
You sit inside your fortifications and you ask one cold question: do I yet have the strategic power to win the fight they want me to start? You know yourself. You know your infantry is mismatched against their strength in the open, or your numbers are thin, or the ground favors them. The honest answer is no. So you decide the thing that feels like cowardice and is actually the move no one has prepared for: you do not go out.
Your ally is being hammered. You named him a lynch pin from the start — a piece you would spend — and you spend him. You let the enemy break his teeth against your ally's walls while you build your own higher and send light, fast units around the back to cut the one thing the enemy cannot live without: his food. You are not idle. You are doing the invisible work while refusing the visible work.
The order comes down from the very top: rescue him. You read it. You do not obey it. You invoke the field commander's independence and you keep building. This is the moment that takes nerve, because if you are wrong you are not just defeated, you are a traitor. You bet your life on the judgment that the dramatic action everyone demands is exactly the action that loses.
That night your own people panic — alarm inside your own camp, fighting outside your tent. You do not get up. You let your stillness be the message: if the commander is asleep, the danger is not real. It settles by itself.
Then the enemy, starving, makes his feint — a loud rush at the southeast. You do not send your strength to meet the noise. You send it to the northwest, where the real blow is coming, because you have read which hand is the tap and which is the lift. The feint fails. The enemy, hungry and humiliated, turns to leave. Now you move — for the first time — and you send your best troops to destroy a beaten army on the march. You waited seven months to act for one of them. That is the whole craft.
This is one of the most dangerous strategies to run because it looks identical to cowardice and incompetence right up until it works. The failure modes:
The narrative comes through Sawyer from the standard histories and the military encyclopedias, and the interpretive tension is unusually sharp here because it is preserved inside the source material itself.
The genuine contradiction: Li Ch'üan, an actual classical commentator on Sun-tzu, cited this episode as an example of failing to employ the unorthodox — a refusal of ch'i.1 The Hundred Unorthodox Strategies and Sawyer read the same events as a prime illustration of the unorthodox.1 These are not two phrasings of one view; they are opposite classifications of the identical campaign. This page does not resolve the contradiction and should not. It marks the live seam in the tradition: is the unorthodox a kind of action (bold, deviant, surprising-by-nature — Li Ch'üan's implicit position) or a relationship to expectation (whatever the room didn't predict — Sawyer's position)? Chou Ya-fu is the case that forces the choice, and two serious readers chose opposite sides.
Sawyer's bias, flagged. Sawyer's project throughout The Tao of Deception is to demonstrate the Chinese tradition's distinctive sophistication about deception and the unorthodox. That bias is doing visible work here: reading a successful defensive attrition campaign as a deep instance of ch'i is exactly the move that makes the Chinese corpus look uniquely self-aware. A skeptic has a clean counter: "win by starving a stronger enemy who attacks your fortified position" is Fabian strategy, attested in Roman, and many other traditions, with no Chinese theoretical apparatus required. Quintus Fabius Maximus ran the same playbook against Hannibal without a single mention of ch'i. So the honest question is whether Sawyer has found a distinctively Chinese doctrine or has draped a universal weak-side tactic in the tradition's vocabulary. The behavior is universal. What may be distinctively Chinese is the explicit, two-sided argument about whether the behavior counts as unorthodox — that meta-debate, not the attrition, is the real evidence for the tradition's self-awareness.
Open question on disobedience. The episode treats Chou's defiance of the emperor as vindicated brilliance. But the histories also record that field-commander independence got many lesser men killed. We see the survivor. How many Chou Ya-fus defied the throne, lost, and were erased from the manuals as traitors rather than canonized as exemplars? The teaching tradition's survivorship bias is invisible inside the case and worth holding.
The clash here is internal to the sources Sawyer marshals, and it is the cleanest author-tension in the unorthodox corpus: Li Ch'üan versus the Hundred Unorthodox Strategies. Both are canonical, both are commenting on the same Chou Ya-fu campaign, and they classify it as opposites — refusal of the unorthodox versus prime example of it. Sawyer sides with the latter, and his siding is itself the argument: he reads the disagreement as evidence that the category of unorthodox is relational, not substantive, because two expert readers could look at one campaign and disagree about whether it qualified. The tension reveals something neither commentator states outright — that ch'i is unstable as a label precisely because it is defined against expectation, and expectation is not a fixed thing two readers will share. Li Ch'üan, expecting the unorthodox to look bold, saw a refusal. The Hundred Unorthodox Strategies author, attending to what the court and enemy expected in 154 BCE, saw the surprise. They disagree because they are measuring against different expectations — and that, exactly, is why the unorthodox can never be pinned to a type of action. The convergence underneath the disagreement: both grant that Chou won, and that he won by doing the opposite of what was demanded. They only differ on what to call it.
The deepest move in this campaign — sacrifice a piece to bait the enemy into exhausting himself against it, then strike the depleted force — is not a fact about Han logistics. It is a structural pattern that appears anywhere a weaker party must convert a stronger party's aggression into self-harm. Two non-history traditions name the exact machinery.
Cross-domain (Japanese strategy) — Sutemi — Sacrifice Strategy. Sutemi is the deliberate sacrifice of position, of a piece, even of one's own apparent safety, to draw the opponent into a committing action that the sacrifice has prepared the ground to punish. Chou Ya-fu's spending of Liang is sutemi at the scale of states. He named Liang a "lynch pin" — a thing to be thrown away to fix the enemy in place — exactly as a sutemi practitioner offers a target the opponent cannot resist attacking. The Japanese frame adds a precision the bare history lacks: in sutemi the sacrifice is only valid if it commits the opponent — locks him into a line of action he cannot reverse. Liang did this. Wu and Ch'u "blunted themselves in repeated assaults" on it, committing their strength and their time and, fatally, eating their supply while they did. The handshake produces an insight neither field supplies alone: the test of a good sacrifice is not whether it survives but whether it removes the enemy's optionality — Liang's value was that attacking it felt mandatory to Wu, and the moment it felt mandatory, Wu's freedom was gone. A sacrifice the enemy can ignore is just a loss; a sacrifice the enemy must attack is a trap.
Behavioral-mechanics — Deceptive Positioning. Chou Ya-fu positioned himself so that the enemy's own hunger and the court's own expectations did his fighting for him. He never had to win an engagement; he arranged the board so the rebels defeated themselves through commitment to the wrong line, and so his disobedience read, in hindsight, as genius rather than treason. Deceptive positioning is the art of arranging a situation so the target's predictable response produces the outcome you want. The military case sharpens the principle with the northwest-wall episode: when Wu feinted southeast, Chou positioned his defense northwest — he positioned against the enemy's real intention rather than his displayed one, reading which signal was the lift and which was the tap. The handshake yields what the influence literature usually omits: deceptive positioning has a reading prerequisite. You cannot position correctly against a feint until you can distinguish the feint from the real move, and that distinction is a separate, prior competence from the positioning itself. Chou had both — the read and the placement — and the campaign shows they are two skills, not one.
A third handshake runs to Makura-osae — Holding the Pillow: Chou's refusal to be drawn — denying the enemy the engagement he kept trying to provoke, smothering every attempted initiative before it could form — is the campaign-scale version of pressing down the opponent's intention at the instant it rises, so the rebels' "sharp spirit" was never permitted to land anywhere.
The Sharpest Implication The unorthodox is not a menu of clever tricks. It is whatever the room in front of you does not expect — and the corollary is brutal: if everyone expects boldness, the bravest available move is restraint, and it will look exactly like cowardice until the instant it wins. This means you cannot pre-load "the unorthodox move" in training, because the unorthodox is defined entirely by the expectations of a specific moment, and those expectations include the expectations of your own side — your ally, your boss, the throne. Chou Ya-fu's deepest disobedience was not to the enemy. It was to the consensus of his own command structure about what a good general does. The unorthodox stroke, at its purest, is the one your own people are ordering you not to make.
Generative Questions