"When you have won a victory, tighten the strings of your helmet."1
Read the instruction literally, because the literal reading is the whole content. It is not rest. It is not stop. The helmet stays on, and the buckle gets pulled tighter than it was during the fighting.
That is a different claim from anything else in Law 47. The rest of the chapter treats the moment after victory as a moment to halt — step back, take a mental pause, refrain from advancing. The proverb treats it as a moment of increased exposure requiring increased protection. You are not being told the danger has passed and you may relax; you are being told the danger has changed shape and your existing guard is now loose.
Which is more precise about what actually happens. Nothing about your situation becomes safer when you win. What becomes safer is how it feels — and the helmet strings are exactly the thing that gets loosened by feeling safe, without anyone deciding to loosen them.
The scene: Hideyoshi has engineered a victory for his master; the master "wanted to go further, to take on and crush yet another powerful enemy."2 Standard Law 47 setup, and the expected advice is don't.
That is not the advice. "For Hideyoshi this was the moment for the shogun to switch from force to cunning and indirection, setting his enemies against one another through a series of deceptive alliances."3
The instruction is change instrument, not cease activity.
This matters because it dissolves the problem the chapter creates a page earlier — that visible restraint "will make you look weak and small-minded" and "seeming to fail to follow up on a victory can lessen your power."4 A switch does not look like restraint. It looks like a different campaign. The master remains active, the momentum is preserved, and the counterreaction that a second frontal advance would have produced never assembles, because there is nothing frontal to react to.
So Law 47 has two remedies and only one of them is stated as the law. The halt is the version for someone who has nothing else to do. The switch is the version for someone with a second instrument.
Greene gives the principle in one sentence and it is more rigorous than it looks: "Too much force creates a counterreaction; too much cunning, no matter how cunning it is, becomes predictable."5
Two failure modes, and they fail by different physics.
Force decays through reaction. It is visible by construction — that is what makes it work — and visibility lets the people it lands on find each other and combine. The failure is external and social: they organise.
Cunning decays through legibility. It works by being unobserved, and every use leaves a trace. The failure is informational and internal to the record: nothing organises against you, but the pattern becomes readable, and a readable indirection is worse than a direct move because you are still paying its costs and no longer getting its benefit.
Now notice the fit. Force's problem is that people can see it; the cure is to become unobserved. Cunning's problem is that its traces accumulate; the cure is to do something with no trace to read. Each mode's characteristic failure is the other mode's operating condition. Alternation is not variety for its own sake. It is running one clock down while the other resets.
Which also tells you when to switch: not on a schedule, and not when the current mode stops working, but when its specific decay signal appears. For force, that is opponents talking to each other. For cunning, that is someone anticipating you correctly.
The specific cunning Hideyoshi recommends is worth separating from the general principle: "setting his enemies against one another through a series of deceptive alliances. In this way he would avoid stirring up needless opposition by appearing overly aggressive."3
Direct advance spends your resources to reduce an enemy. This spends their resources, and reduces two at once. The arithmetic is obviously better, and the interesting question is why anyone advances directly at all.
The answer is that the play has a narrow window, and Law 47 is describing exactly when it opens. Deceptive alliances require that other parties be actively recalculating — willing to entertain a new arrangement, unsure who they should be worried about. Parties in a settled configuration have no reason to move. A recent victory is what unsettles them, because it changes the relative sizes on the board and forces everyone to redo the sum.
So the moment after a win is not merely the moment of maximum danger. It is the moment of maximum fluidity, and the same fact produces both. Everyone is reconsidering their position, including the people who were content yesterday.
That reframes the alternation as opportunistic rather than defensive. The switch to indirection is not a retreat into safety while things cool down. It is a move onto ground that only exists for as long as the recalculation lasts.
Look at how Hideyoshi delivers it. He is a subordinate telling his master to abandon a plan the master has already announced enthusiasm for. Every part of that is expensive.
He does not say you are being arrogant. He does not offer his own judgment at all. He "reminded him of the old Japanese saying."1
The proverb is doing something structural. It converts a personal challenge into a reminder of common knowledge — the content arrives with no author, so there is nobody to be defied and nothing to lose face against. The master is not being corrected by his general; he is being reminded of something he already knew, which is the only form in which the correction is free to accept.
This is a general instrument and worth naming. When you must brake a superior, supply an impersonal authority rather than an argument. A proverb, a precedent, a rule that predates both of you. The stronger your evidence and the more clearly it is yours, the more the exchange becomes about whose judgment prevails — and that is a contest a subordinate wins only at a cost that exceeds the value of being right.
Greene's closing instruction names the target of the manoeuvre: "When you are victorious, then, lie low, and lull the enemy into inaction."6
Lull is the operative word. The purpose of the switch is not primarily to protect you from your own overreach — it is to change what the other side believes about your trajectory.
An enemy watching a run of force extrapolates the run. They prepare, they seek allies, they act early because acting later will be worse. An enemy watching that run stop draws a different inference: that you are consolidating, or spent, or content. Both inferences reduce their urgency, and reduced urgency is the resource the indirection phase spends.
So the two halves are not independent. The deceptive alliances Hideyoshi recommends only work on parties who are not currently expecting to be manoeuvred against — and the visible cessation of force is what produces that state. The halt is the setup for the cunning, not an alternative to it.
🚩 The historical framing of this passage is unreliable and the vault should record it rather than pass it on.
Greene calls Oda Nobunaga "the shogun."2 [CONTESTED] — Nobunaga never held the title. He deposed the last Ashikaga shogun, Yoshiaki, in 1573 and pointedly did not take the office.
He calls Hideyoshi the "future emperor."2 This is plainly wrong. Toyotomi Hideyoshi never became emperor, and never became shogun either — his birth barred the shogunal title, which is why he took kampaku (imperial regent) in 1585. The distinction is not pedantic; the whole shape of his career turns on it.
He credits Hideyoshi with engineering "a stunning victory over the army of the formidable General Yoshimoto."2 Imagawa Yoshimoto was killed at Okehazama in 1560, and that was Nobunaga's victory. Hideyoshi was a very junior retainer at the time.
The underlying advice survives all three, because the proverb and the force/cunning alternation do not depend on who said them to whom. But the anecdote as told is not history, and 🚩 SECONDARY WITHOUT PRIMARY applies to all of it — Greene cites nothing.
You have just won and the obvious move is to do it again to the next one.
Before deciding whether to advance, check the guard rather than the map. What did you let slip during the push — the thing you stopped checking, the person you stopped calling, the process you suspended because it was slowing you down? Those are the helmet strings, and they loosened without a decision. Tighten them today, while it still feels unnecessary.
Then, instead of choosing between advancing and stopping, ask which instrument you have been using. If it has been force — direct, visible, unambiguous — check whether the people it landed on have started talking to each other. That is the decay signal, and it appears before any of them acts.
If it has, do not stop. Switch. Find the thing you can do that no one experiences as an attack: an alignment between two parties who both have a reason to want something you want, a favour placed where it will be remembered, a position taken quietly. Activity continues, nothing frontal presents itself, and the coalition that was assembling loses its object.
And if you are the one who has to say this to someone above you, do not bring your own judgment. Bring a rule that predates you both. The advice they can accept is the advice that nobody appears to be giving.
Evidence quality: [POPULAR SOURCE], 🚩 SECONDARY WITHOUT PRIMARY. Three factual errors in the framing, documented above — [CONTESTED] for Nobunaga's title and the Yoshimoto attribution, plainly incorrect for "future emperor." The proverb itself is genuine and widely attested in Japanese usage; Greene gives no source for it either.
Tension: the chapter's stated law is stop; this passage's advice is switch. Greene presents them as the same instruction. They are not, and the switch version is the only one that answers his own objection that visible restraint reads as weakness.
Open questions: the alternation model implies decay signals for each mode — coalition formation for force, successful anticipation for cunning. Greene names neither. If the switch is supposed to be timed rather than scheduled, what is the trigger?
The force/cunning alternation is one of Greene's most developed ideas and it belongs more to The 33 Strategies of War than here, where it arrives compressed into a paragraph. In the later book the alternation is treated as a strategic primitive with its own logic; in Law 47 it is a remedy for an emotional condition. The compression costs the idea its mechanism — the reader gets vary your approach rather than each mode's failure mode is the other's operating condition.
The corpus tension is with the laws that recommend total commitment to a single mode. Law 15's annihilation instruction and the chapters on boldness both argue that half-measures are the error and that the powerful commit. This passage says commitment to a mode is itself the error, because modes decay. Greene does not place them together, and the reconciliation available in the text is the same one Law 47's Reversal offers elsewhere: commit totally within an engagement, alternate between them.
The convergence worth marking is with the corpus's courtier material. Hideyoshi's proverb technique — impersonal authority in place of personal judgment — is precisely what Greene recommends for advising a superior throughout the earlier laws. What is unusual is seeing it used to restrain rather than to flatter or redirect, which is the harder case.
The Zen–Bujutsu Relationship — the martial-contemplative material supplies the state the proverb is actually prescribing, which Greene's strategic reading flattens into a metaphor about caution. In the bujutsu framing, the danger after a decisive strike is not that the enemy recovers but that the practitioner's zanshin — the continued alertness that outlasts the technique — collapses at the moment the technique succeeds. The lapse is not a judgment error and cannot be corrected by deciding to be careful, because it is a physiological release: the body reads completion and stands down. That gives the proverb a mechanism the strategic account lacks entirely. Greene's chapter attributes post-victory failure to arrogance, an attitude one could in principle choose against; the contemplative frame says the failure is an involuntary discharge that arrives before any attitude forms. The insight neither reaches alone: if the lapse is pre-cognitive, then every remedy in Law 47 that works by reflection — halt, reconsider, examine the role of luck — is aimed at a layer beneath which the damage has already occurred. What the tradition prescribes instead is a trained continuation, a physical habit of maintaining the guard through the completion, which is why the instruction is about a buckle rather than about a state of mind.
Alternate Through the Current — the creative-practice frame treats alternation between modes as a way of managing a resource that depletes unevenly, and it names the variable Greene's version omits: recovery time. In the creative case, switching between generative and evaluative modes works because the depleted faculty restores while the other is running, and switching too fast gives neither the interval it needs. Apply that to force and cunning and the alternation model gains a constraint it badly wants. Greene implies you can switch at will, whenever the decay signal appears. The recovery framing says each mode also has a minimum off-period — force needs long enough for the assembling coalition to disperse and find other business, cunning needs long enough for the accumulated traces to lose salience — and a switch made before that period elapses restores nothing. The insight the pairing produces: alternation has a floor as well as a trigger, and an operator switching rapidly under pressure gets the costs of both modes and the reset of neither. That predicts a specific failure state Greene never describes — the leader who is constantly changing approach, visibly, and finds that opponents are simultaneously organised against them and able to anticipate them.
The transferable claim is the complementary-decay model: that force fails through coalition formation and cunning through legibility, and that each mode's off-period cures the other's accumulated damage. That is checkable against any extended contest with a documented record of approach changes, and it predicts that successful long-run operators alternate on decay signals rather than on schedules.
The open edge is the floor raised in the alternation handshake — whether there is a minimum interval below which switching stops working. Greene has no timing model at all, and the two-mode structure implies one is needed.