A hydraulic engineer named Cheng Kuo walks into the court of Ch'in carrying a map and a plan. He is good — genuinely good — at moving water. He tells the king of Ch'in that he should chisel a canal three hundred kilometers long, from the Ching River around the northern mountains toward the Luo, to irrigate the fields and grow the realm rich.1 It is an attractive pitch. Ch'in's ruler "liked to initiate major undertakings."2 What the king does not know is that Cheng Kuo was sent by the rival state of Han for one purpose: to bury Ch'in's manpower and treasure in an enormous public-works project so exhausting that Ch'in would have no strength left to march east and swallow Han.3
So Ch'in digs. And digs. And then, mid-construction, the king learns the truth and prepares to execute his engineer. Cheng Kuo says the one thing that saves his life — and it is true: "Initially I was a secret agent, but completing this canal will truly be to Ch'in's advantage!"4 The king thinks about it, decides he is right, and lets him finish. The canal opens. It converts millions of acres of salt-laden ground into productive rice fields. Ch'in "never again suffered from famine and was able to grow rich and strong," and then went on to conquer everyone — including Han.5 They named the canal after the man who was sent to ruin them.
The stratagem is the betrayal that doubles back into a gift. Han aimed a weapon at Ch'in's economy and accidentally handed it a breadbasket.
The canal stratagem is a covert economic-exhaustion program: an enemy state plants a clandestine agent — here a credentialed technical specialist — to lure the target into a vast, draining public undertaking, on the theory that a state pouring its labor and revenue into monumental construction cannot also field armies of conquest. It belongs to the same family Sawyer traces back to Wen Chung's effort to undermine Wu at the end of the Spring and Autumn period: "the policy of exhausting a state through massive public undertakings was already part of current knowledge."6 Han "essentially mimicked Su Ch'in's mid-fourth-century example by trying to generate unrest among the workers impressed for a massive, arduous canal project."7
The distinctive feature here is the backfire. Most exhaustion stratagems in the corpus are judged by whether they drained the enemy. This one is remembered because it inverted — the drain became a dividend. Sawyer's word for it is exact: "Ironically, Han's subversive efforts significantly augmented Ch'in's economic base and thereby substantially contributed to their own demise."8
The exhaustion stratagem rests on a hidden assumption, and the assumption is where it breaks. It assumes that a giant project is net-negative for the builder — pure cost, pure drain. That holds when the project is a tomb, a wall built nowhere useful, or a palace. It fails catastrophically when the project is productive infrastructure.
A canal is not a tomb. A canal is a machine that, once running, pays back its construction cost in flood after flood of irrigated grain. Han weaponized the construction phase and forgot the operating phase. The labor and gold Ch'in spent were real costs — but they were front-loaded, and they purchased a permanent increase in carrying capacity. Cheng Kuo's saving line is not a clever escape; it is an accurate engineering judgment. He tells the king the literal truth about the asset he was sent to weaponize: the exhaustion is temporary, the advantage is permanent.
The deeper mechanism is that Cheng Kuo had two motives and only one of them was Han's. Sawyer notes he was apparently also motivated "to enrich himself as its director."9 An agent who genuinely wants the project to succeed — because his own fortune rides on it — is a poor saboteur. His self-interest aligned with the target, not the sender. The stratagem was undone partly by its own instrument.
This page hands the covert-programs corpus its sharpest cautionary tale: the boomerang case. Every covert economic program assumes it can predict whether a given drain weakens or strengthens the target. Cheng Kuo proves the prediction can invert completely. It gives Systematic Covert Programs a failure mode that pure deception programs do not have — the program can succeed operationally (the canal got built, Ch'in spent the labor) while failing strategically (Ch'in got richer).
It also pairs with its sibling Systematic Bribery as Strategic Policy to bracket the two halves of Ch'in's century. Bribery dissolved the alliances that might have stopped Ch'in; the canal fed the armies that finished the job. Han tried to starve Ch'in and instead helped fatten it for the conquest that ended Han.
The pivot is a single exchange, and it deserves slow reading.
The king of Ch'in "realized Cheng Kuo's true intentions and therefore wanted to execute him."10 This is the moment the stratagem is supposed to die — the agent exposed, the project halted, the manpower recovered. In almost any other espionage anecdote, exposure ends the operation. Here it does not, and the reason is that the asset under construction has acquired a logic independent of its sender's intent.
Cheng Kuo does not plead, deny, or grovel. He concedes the espionage outright — "Initially I was a secret agent" — and then makes a claim about the future the king cannot easily refute: "completing this canal will truly be to Ch'in's advantage."11 The king, "feeling this to be true, eventually had him complete it." Sawyer adds the context that makes the king's decision rational rather than gullible: against a background of "burgeoning economic prosperity stimulated by canal building, Cheng Kuo's advice would certainly have been plausible."12 The king is not fooled twice. He simply recognizes that a half-finished canal is the worst of all outcomes — all the cost, none of the yield — and that finishing it converts a sunk sabotage into a strategic asset.
The result is recorded with brutal economy: muddy water from the Ching River, millions of acres of salt land turned to rice, "several bushels per acre," no more famine, and unification of the realm.13 The state that sent the saboteur was among the realm Ch'in then absorbed.
You are the king of Ch'in, and you have just been handed proof that the engineer who has had your peasants digging for years was sent by Han to break your back. Your first instinct is the sword. Then you walk to the edge of the half-cut channel and look at it — kilometers of raw earthwork, your gold already in the ground, your men already worn down. You do the arithmetic the saboteur is counting on you to do. Kill him now and fill in the ditch, and you have paid the entire cost of the sabotage for nothing. Let him finish, and the same ditch becomes the thing that ends your famines forever. The agent is standing in front of you telling you the truth about your own infrastructure. You sheathe the blade. You tell him to finish it. And years later, when the rice is coming in and your armies are marching east well-fed, you let the canal keep his name — Cheng Kuo — because the joke is now entirely on Han.
You have chosen the wrong project to weaponize when the project produces more than it consumes. A canal, a road, a harbor — anything that compounds — is a gift dressed as a drain. Aim your exhaustion at tombs and walls, not at machines that pay back.
You have chosen the wrong agent when his fortune rises with the project's success. Cheng Kuo wanted the canal to work because he was its director; a saboteur who profits from completion will quietly do excellent work for your enemy.
You are running it wrong when you cannot tell front-loaded cost from permanent cost. The whole stratagem mistook a one-time construction bill for an ongoing burden. If the target can amortize the drain into capacity, your weapon becomes their endowment.
And the deepest sign of failure: when exposure makes the target better off by forcing him to finish what you started. A sabotage that improves on discovery was never sabotage; it was investment you funded for your enemy.
The chapter Sawyer draws on is the Shih Chi's treatment of canals and the Yellow River, which "coincidentally notes this covert attempt to exhaust Ch'in."14 That word coincidentally matters: the episode survives not as a celebrated spy story but as a footnote inside a hydrology chapter. The espionage framing may be the historian's later gloss on what was, at the time, simply a famous canal with an odd origin story. Whether Han genuinely conceived the project as exhaustion-warfare, or whether that motive was read backward once Ch'in's enrichment made the irony irresistible, is an open question Sawyer does not settle.
The numbers — three hundred kilometers, millions of acres, several bushels per acre — carry the same uncertainty as all Warring States figures. They establish the scale of the irony more than they establish precise fact.
Sawyer wants Cheng Kuo to be a clean example of the exhaustion stratagem, but his own telling keeps undercutting the category. He has to admit the agent had a private profit motive that aligned him with the target; he has to admit the source is a hydrology chapter, not an espionage record; and he has to lead with the irony rather than the tradecraft. Where Sawyer is on firm ground is the lineage — he convincingly threads Wen Chung's exhaustion of Wu, Su Ch'in's mid-fourth-century version, and Han's canal into a single recognizable family of statecraft.15 Where he reaches is in treating the canal as a fully intentional, well-run operation; the evidence reads more like a sabotage that was half-hearted, self-undermined by its own agent, and then retroactively dignified by the magnitude of its backfire. The convergence and the tension sit in the same sentence: the stratagem was real and it barely functioned as designed.
To Confirmation Bias as Ancient Problem. Han's plan only works if you assume Ch'in's ruler is a man who cannot resist a grand project — and the histories say exactly that of him: he "liked to initiate major undertakings."16 Han read a real trait correctly and built a weapon on it. But the same trait that made the king take the bait also made him finish the canal when he learned the truth — a man who loves great works does not abandon the greatest one half-dug. The insight neither domain reaches alone: a behavioral profile sharp enough to start a manipulation can be the very thing that reverses it. Han's read on the king was accurate and exploitable; it was also the reason the exploit failed. Profiling an enemy's compulsion tells you how to hook him, but not which direction the hook will pull once reality enters the room.
To Contrary Employment Doctrine. Sun-tzu's contrary-employment principle turns an enemy's resources back against him — but it assumes the resource is fungible and the direction reversible. The canal is contrary employment that seized up: Han tried to turn Ch'in's appetite for construction into a self-inflicted wound, and the appetite simply ran in its own favor. The handshake exposes the boundary of the doctrine: you can turn an enemy's agent (a fan chien) or an enemy's belief against him cleanly, because those have no independent productive value. You cannot reliably turn an enemy's productive capacity against him, because capacity, once built, serves whoever owns it. Cheng Kuo marks the line where contrary employment stops working — the moment the thing you weaponize starts generating value on its own.
Sharpest implication: Sabotage aimed at a productive asset is indistinguishable from investment, because the asset does not care who paid for it. Han built Ch'in a breadbasket and called it a wound.
Generative questions: