The chapter's biggest claim gets its shortest treatment.
Finally, the reason you strike at the shepherd is because such an action will dishearten the sheep beyond any rational measure. When Hernando Cortés and Francisco Pizarro led their tiny forces against the Aztec and Incan empires, they did not make the mistake of fighting on several fronts, nor were they intimidated by the numbers arrayed against them; they captured the kings, Moctezuma and Atahualpa. Vast empires fell into their hands. With the leader gone the center of gravity is gone; there is nothing to revolve around and everything falls apart.1
⚠ Two of the most studied events in world history, in one sentence each, with the causation stated as obvious.
⚠ Greene concedes the mechanism is irrational and it is the most interesting word in the passage.
Dishearten the sheep beyond any rational measure.
⚠ If the reaction is disproportionate to the actual loss, then the technique is not working on the leader's contribution. It is working on a belief about the leader — and beliefs can be false, which means the technique's effect is decoupled from whether the target was load-bearing at all.
⚠ And that puts this passage in direct conflict with the Prescott marginal printed in the same chapter. Prescott's Inca is "the keystone of the political fabric which must fall to pieces by its own weight when that was withdrawn" — a structural claim, in which the collapse is proportionate and mechanical.
⚠ Two theories of the same event, both in Law 42: disproportionate psychological collapse and genuine structural dependency. ⚠ They make opposite predictions about a group whose leader was not actually essential — and Greene never notices he is holding both.
⚠ The metaphor returns and it is the same one as the Keys' opening.
There is nothing to revolve around and everything falls apart.
⚠ Planets orbiting a sun, restated as consequence. And the corpus recorded on the earlier page that the image imports physical necessity into a contingent arrangement — here it does the work of skipping the two years between Moctezuma's capture and the fall of Tenochtitlan.
⚠ Because vast empires fell into their hands is compressing a war. The Aztec resistance continued after Moctezuma's death; Tenochtitlan fell in 1521 after a siege. ⚠ A capture that requires a subsequent siege to convert into a conquest is not the mechanism the sentence describes.
🚩 ⚠ The two cases are the chapter's most consequential evidence and the account omits the factors most historians treat as decisive.
Epidemic disease. ⚠ Smallpox reached central Mexico during the campaign and killed on a scale no battle did. It is not mentioned.
Indigenous alliances. ⚠ Tiny forces describes the Spanish contingent and not the army in the field. Cortés fought with tens of thousands of Tlaxcalteca and other allies who had their own reasons to want Tenochtitlan destroyed. They are not mentioned.
A civil war already in progress. ⚠ Atahualpa had just defeated his brother Huáscar when Pizarro arrived. The Inca state Pizarro decapitated was freshly fractured by a succession war — which is a large part of why the decapitation took. It is not mentioned.
⚠ The corpus should be exact about what this does and does not establish. It does not show that striking the leader was ineffective — Prescott's account of Cajamarca describes a genuine immediate rout, and it is good evidence. ⚠ What it shows is that the chapter's flagship demonstration of its most dramatic claim is the case where the causal story is most overdetermined, and the alternatives are removed rather than argued against.
🚩 SECONDARY WITHOUT PRIMARY — no dates, no source, no episode. ⚠ Prescott is cited in full in this chapter's margin for exactly one of these events, and the Keys' version of it names no source at all.
⚠ The passage's other justification is different from the psychological one and is not developed.
They did not make the mistake of fighting on several fronts, nor were they intimidated by the numbers arrayed against them.
⚠ This is a resource allocation argument, not a claim about morale. A small force cannot win a distributed war and can win a single decisive action — so concentration on one target is correct arithmetic regardless of what the target's removal does psychologically.
⚠ And it is the more defensible of the two arguments in the paragraph, because it does not require any claim about how the enemy will react. Concentrating scarce force is sound whether or not the sheep scatter.
⚠ The corpus should record that Law 42 has three separable justifications and treats them as one: the sheep collapse irrationally, the structure genuinely depends on the keystone, and a small force must concentrate. ⚠ Only the third is available to a reader who cannot verify the other two in advance.
⚠ The two theories in this paragraph differ on a single observable, and the chapter's cases cannot distinguish them.
If the collapse is structural — Prescott's keystone — removal produces disintegration proportionate to what the person was actually holding, and a group with distributed functions survives.
If the collapse is psychological — beyond any rational measure — removal produces disintegration proportionate to what the group believed the person was holding, which can be much larger, and can occur even where the functions were distributed.
⚠ The distinguishing observation is a group whose leader was not essential. Structural theory predicts continuity; psychological theory predicts collapse anyway.
⚠ And Law 42 contains no such case, because every example is selected on the outcome. ⚠ The Reversal's Grant is the closest thing to one and it is filed as a lesson about revenge rather than as evidence about the mechanism.
⚠ Which leaves the reader with an unhelpful position: the technique's two proposed engines require completely different preparation — the structural one requires you to identify a genuine dependency, the psychological one requires you to identify a belief — and the chapter's advice, find the one head that matters, is silent about which you are looking for.
⚠ The corpus should record the practical consequence. Where the two theories disagree, the psychological one is the cheaper bet and the more fragile result: a group that collapses because it believed it depended on someone can recover when the belief updates, and a group that collapses because it actually depended on them cannot.
Separate the three arguments before acting on any of them. ⚠ Concentration of scarce force is sound arithmetic and needs no theory of the enemy. Structural dependency requires you to identify what actually runs through the person. Disproportionate collapse requires you to know what the group believes. They call for different preparation and the chapter runs them together.
Ask what else was already happening. ⚠ The chapter's two flagship cases were both preceded by conditions that did most of the work — a succession war on one side of the world and an epidemic on the other. A decapitation that lands in an already-fracturing structure is a very different operation from one that lands in a healthy one, and only the first is well evidenced here.
If you are betting on belief, expect it to be recoverable. ⚠ A group that scattered because it thought it could not function will discover otherwise, and the window is short. Prescott's Cajamarca rout happened in an afternoon; Tenochtitlan took two years and a siege. Plan for the second and treat the first as a bonus.
⚠ And on the defensive side, the reading is unusually clear. The disproportionate collapse is a belief about indispensability, which means it is addressable in advance and cheaply — by making it known, before anything happens, who does what and who can act without whom. A group that has rehearsed its own continuity does not scatter, because the belief the technique exploits was never installed.
Strongest evidence. The concentration argument is sound independent of everything else in the passage, and Prescott's Cajamarca account — printed in the same chapter — is genuine evidence of an immediate rout following a capture.
Tension — beyond any rational measure and Prescott's keystone are two incompatible theories of the same event, both in Law 42, unreconciled.
Tension — vast empires fell into their hands compresses a two-year war, including a siege, into a capture.
🚩 [POPULAR SOURCE]. 🚩 SECONDARY WITHOUT PRIMARY — no dates, no source, no episode, for two of the most documented events in world history. 🚩 ⚠ Epidemic disease, indigenous allies numbering in the tens of thousands, and an Inca succession war are all absent, and the alternatives are removed rather than argued against. ⚠ Tiny forces describes the Spanish contingent, not the army in the field.
Open questions. ⚠ Would either decapitation have worked against an intact state? Both landed on structures already under extraordinary stress, which is the one variable the chapter cannot vary. ⚠ And how long does the disheartening last? The passage implies permanence and one of its own two cases refutes it.
⚠ This paragraph and the Prescott marginal describe the same event with different mechanisms, on facing pages. ⚠ Prescott's is better — keystone, load-bearing, proportionate — and it is the one in the margin. The corpus has now recorded this shape repeatedly across Laws 40–42: the sourced marginal carries the stronger version of the chapter's own argument.
⚠ Against the go passage: go has no king and this paragraph is pure chess. The chapter's two strategic models are eight lines apart and only one of them is compatible with this case.
⚠ And against the three functions: Atahualpa is the corpus's clearest instance of load, and Moctezuma may be the clearest instance of the model failing — a captured king whose capture did not end the war. ⚠ The chapter presents them as a matched pair.
Behavioural mechanics — the opponent with nothing to capture. The Abstract Center of Gravity opens on the situation this paragraph cannot handle: some opponents do not have a capital city — no seat, no king, no single object whose seizure ends anything.
With the leader gone the center of gravity is gone; there is nothing to revolve around and everything falls apart.
The insight neither produces alone: ⚠ Greene treats the centre of gravity as though it were always a person and always findable, and the abstract-centre frame supplies the category he is missing: a centre of gravity can be an idea, a supply line, a source of legitimacy, a population's willingness — none of which can be captured in an afternoon. ⚠ Which reframes the Cortés and Pizarro cases as the easy end of a spectrum rather than as the paradigm: they faced opponents whose centre of gravity happened to be a person who happened to be reachable. ⚠ And it explains the chapter's silence about everything that has happened in war since. The whole modern difficulty — insurgency, networked opposition, movements with no address — is the case where the centre of gravity is abstract, and Law 42's instruction, aim at the leaders, has nothing to say to it except to keep looking for a leader. ⚠ The corpus should record the diagnostic: before striking, ask what the group would still have if the person were gone tomorrow. Where the answer includes a shared belief, a grievance, or an income, the person is not the centre.
Psychology — the threshold where responses invert. Every Dog Has a Breaking Point records Pavlov's finding that a nervous system pushed hard enough does not degrade smoothly — at some threshold it flips into a qualitatively different mode, and the responses reverse. The bold animal goes meek; stress stops adding up and starts inverting.
Dishearten the sheep beyond any rational measure.
What the pairing produces: ⚠ Greene's beyond any rational measure is offered as a curiosity and the collapse literature makes it a mechanism with a shape. Disproportion is exactly what a threshold effect looks like from outside: ⚠ nothing happens, nothing happens, and then everything happens at once, and the size of the final input tells you nothing about the size of the response. Which is why a capture that changes little materially can end a war — and equally why an identical capture elsewhere changes nothing. The variable is not the blow; it is how close the system already was to its line. ⚠ And that reframes the chapter's missing context as the decisive context rather than as pedantry. A state in the middle of a succession war and a city in the middle of an epidemic are systems already near threshold, and the capture is the last increment rather than the cause. ⚠ The corpus should record the operational form for both sides: the question is never how hard is this blow but how loaded is this system — and the defensive corollary is that reducing existing load matters more than defending the leader.
Sharpest implication. ⚠ Beyond any rational measure concedes the effect is disproportionate — so the technique works on a belief about the leader rather than on the leader's contribution, which puts it in direct conflict with Prescott's keystone in the same chapter. ⚠ The paragraph holds three separable justifications and treats them as one: irrational collapse, genuine structural dependency, and the arithmetic of concentrating a small force. Only the third is available to a reader who cannot verify the others in advance. ⚠ And vast empires fell into their hands compresses a two-year war and a siege into a capture, while a succession war, an epidemic and tens of thousands of allies go unmentioned.
Generative questions.