A minister has been ordered to kill himself. Before he does it, he says this to his king:
"Tear out my eyes, oh King, and fix them on the gate of Wu, so that I may see the triumphant entry of Yueh."1
Wu Tzu-hsiu had spent years telling the king of Wu that the war in the north was going badly and that the southern state of Yueh was watching. The king laughed at him and then, sensing betrayal, ordered his death.
Within a few years a Yueh army came through that gate.
The setup is worth having exactly, because the motive is not territory.
Early sixth century BC. Wu is a growing power, but it "lacked the great history and civilization of the Middle Kingdom, for centuries the center of Chinese culture."2
Greene's statement of the king's reasoning is one line: "By defeating the Middle Kingdom, the king of Wu would instantly raise his status."2
Not resources. Not security. Standing.
And the war behaved the way such wars do: "It began with great fanfare and several victories, but it soon bogged down. A victory on one front would leave the Wu armies vulnerable on another."2
That sentence is the mechanism of the whole law. Every gain purchased an exposure elsewhere, so the campaign could win indefinitely without ever arriving anywhere.
The break is more interesting than a disagreement about strategy.
In 490 BC, Wu Tzu-hsiu sent his son away to safety in the kingdom of Ch'i.1
That is not an argument. It is a hedge, and everyone could read it: "In doing so he sent the king a signal that he disapproved of the war, and that he believed the king's selfish ambition was leading Wu to ruin."1
So the minister did not merely predict the collapse. He acted on his own prediction in a way that made it public, and the king was correct that this was a kind of defection — a chief adviser who has evacuated his family has withdrawn from the enterprise while still holding office.
The king's response — "accusing him of a lack of loyalty and, in a fit of anger" ordering his suicide — is the reaction of a man who has understood the signal precisely.
The ending is the part that turns the case from a strategic lesson into something stranger.
When the Yueh army surrounded the palace, "the king remembered his minister's last words—and felt the dead man's disembodied eyes watching his disgrace." Unable to bear the shame, he killed himself, "covering his face so that he would not have to meet the reproachful gaze of his minister in the next world."3
The war is not what ends him. He is not killed by Yueh; he dies of being watched.
And the covering of the face is a claim about the afterlife — he expects the gaze to continue, and is arranging for it not to reach him.
Wu Tzu-hsiu's curse was designed for exactly this. He did not ask for revenge or rescue. He asked to be a witness, and it worked on a man who had already killed him.
Greene immediately widens the case:
"The story of Wu is a paradigm of all the empires that have come to ruin by overreaching… This is what happened to ancient Athens, which lusted for the faraway island of Sicily and ended up losing its empire. The Romans stretched the boundaries of their empire to encompass vast territories; in doing so they increased their vulnerability."4
The Athens comparison is apt and it is also the same Sicilian expedition Greene uses elsewhere in this book to prove a different law — a corpus-level duplication where one event is made to demonstrate two incompatible lessons in separate chapters.
And the Rome claim is where the generalisation gets loose. "Their useless expansion led their empire into oblivion" compresses several centuries into a single cause, and the chapter offers nothing to support it.
Then the law is stated as physics:
"It is almost a physical law: What is bloated beyond its proportions inevitably collapses… What is concentrated, coherent, and connected to its past has power. What is dissipated, divided, and distended rots and falls to the ground."4
🚩 The selection problem is total here. Overreaching empires that collapsed are the ones we have names for. Empires that expanded and consolidated successfully are simply not in the sample, and "almost a physical law" is doing the work that evidence would.
Greene supplies one operational criterion, quoted rather than argued:
"If you are not in danger, do not fight."4
That is a genuine decision rule, and it is aimed exactly at Wu's failure. The king was not in danger. He was in pursuit of status, which the rule classifies as an insufficient reason.
It also implies something the chapter does not draw out: wars of ambition and wars of necessity have different failure modes. A war you must fight ends when the threat does. A war fought for standing has no natural terminus, because standing is comparative and there is always more of it — which is why Wu's campaign could produce "several victories" and never approach a conclusion.
You are being pulled toward an expansion — a new market, a second product, a bigger role — and the case for it is that it would raise your standing.
Run Sun-tzu's test first, and run it honestly: are you in danger if you don't?
If yes, the expansion is defensive and the question is only whether you can afford it.
If no — if the real answer is that it would make you more impressive — then you are the king of Wu, and you should know that the campaign you are considering has no stopping condition, because the thing you are pursuing is comparative and cannot be finished.
Then look for the structural tell, which is the most checkable thing in this case: does each win create an exposure somewhere else? "A victory on one front would leave the Wu armies vulnerable on another" is not bad luck; it is what overextension feels like from inside while it is still going well. Winning and thinning are the same motion.
And attend to your Wu Tzu-hsiu. Somebody near you has said the unwelcome version, probably more than once. Watch what they do rather than what they say — the minister's argument was ignored and his son leaving the country was the real message. When people who understand the situation start quietly hedging, that is your data, and it arrives before any of the numbers do.
The strongest material is the structural observation — each victory generating a new vulnerability — which describes a mechanism rather than asserting a moral.
Tension: the generalisation is built entirely on survivorship. Wu, Athens and Rome are collapsed empires; successful expansions are absent from the argument, and "almost a physical law" is asserted rather than shown.
Tension: the Sicilian expedition does double duty in this book. Greene uses it here for overreach and elsewhere for a different lesson, without cross-reference.
Tension: the case's most vivid element is not about concentration at all. The eyes on the gate, the curse, the king covering his face — the material that makes this memorable is about shame and witness, and the law it illustrates is about resource allocation.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no chronicle, text or translator is named for a sixth-century BC episode; the Sun-tzu quotation is given without a work or edition; the closing quotation about covering his face is set in quotation marks with no attributed source at all. [POPULAR SOURCE]
Open questions. Was Wu Tzu-hsiu right for the right reasons, or right about a danger any adviser would have named — the case treats his accuracy as vindication of concentration doctrine and it may simply be an accurate read of one neighbour. And: why does the king believe the gaze continues after death? The chapter reports the belief and never touches it, though it is the reason the case ends the way it does.
Against the law it opens, this transgression demonstrates a narrower claim than the law makes. It shows that expansion without consolidation creates exposure. It does not show that depth beats breadth, that a rich mine outperforms several shallow ones, or that intellect is intensive — the case is about military overextension and the law is mostly about attention.
Set beside the Rothschilds the chapter's own two flagship cases sit awkwardly together. Wu is punished for expanding to five fronts; the Rothschilds are celebrated for expanding to five capitals. Greene's stated distinction is cohesion — the brothers held together and Wu did not — which means the operative variable is not concentration at all but whether the expanding entity keeps a shared interior, and neither the law nor its title says so.
And against the Reversal, Wu is the case that makes the Reversal legible. Wu was the stronger party overextending; the Reversal's guerrillas are the weaker party dispersing deliberately. Concentration is advice for whoever has enough to lose, and the chapter never states the precondition.
Behavioral mechanics — Goal Dilution Effect
That page documents a specific finding: adding benefits to an offer weakens its perceived effectiveness. A product claiming to do one thing reads as better at that thing than an identical product claiming to do it plus four others.
Wu's war is the same effect operating on a state's capability rather than on a consumer's perception.
What the pair produces: dilution has a perceptual half and a material half, and they compound in the same direction, which is why overreach is so hard to arrest from inside. The material half is Greene's — each front genuinely thins the army. The perceptual half is that a power fighting everywhere reads as strong nowhere, which is precisely the judgement Yueh made when it "began to notice the kingdom of Wu's problems."
So an overextended party is attacked earlier than its actual weakness warrants, because rivals price the dilution signal before the material weakness matures. Which reverses the intuition about when overreach becomes dangerous. The king assumed he was safe until the armies were genuinely stretched. The exposure opened the moment the stretch became visible — and visibility precedes vulnerability by a long way.
Business — When Nothing's Certain, Go to the Stories You Know You Can Tell
That page tracks a career making the same move three times across three decades: after the debut, after a critical drubbing, after an era ran out — each time returning to songwriting, the competence underneath everything else.
It is the manoeuvre the king of Wu never made, and the trigger is what makes the pair useful.
What the pair produces: the return to core competence is triggered by uncertainty, and Wu's failure was that his uncertainty never arrived. In the business case the pivot is prompted by not knowing what comes next — the doubt is the signal to contract.
The king had the opposite condition: he was certain. "One more big victory and the great Middle Kingdom would be his." Confidence is what suppressed the contraction, and Wu Tzu-hsiu's function in the story is to supply the missing doubt from outside — which is why removing him removed the last mechanism that could have stopped the campaign.
So the practical failure mode is not overreach but the absence of a doubt-holder. An expanding operation needs somebody whose job is uncertainty, and it needs them to be listened to precisely when things are going well. Wu's king had one, correctly identified him as disloyal, and was right — the doubt-holder had in fact defected, which is what happens when the role goes unprotected for long enough.
Sharpest implication. A war fought for standing has no stopping condition, because standing is comparative and there is always more.
That is why Wu could win repeatedly and never approach an ending — and why Sun-tzu's "if you are not in danger, do not fight" is a rule about terminability rather than about caution.
Generative questions.