Law 39's Reversal opens with a warning — "Study the enemy beforehand: Some fish are best left at the bottom of the pond" — and gives Tyre.1
First bait. The city's leaders, "confident they could withstand Alexander," send ambassadors saying they will recognise him as emperor but will not admit him or his forces. ⚠ "This of course enraged him, and he immediately mounted a siege."2
Four months pass. ⚠ He decides the struggle is not worth it and offers terms.
Second bait. ⚠ *"But they, feeling that they had already baited Alexander and gotten away with it, and confident that they could withstand him, refused to negotiate—in fact they killed his messengers."*3
This pushed Alexander over the edge. Now it did not matter to him how long the siege lasted or how large an army it needed; he had the resources, and would do whatever it took. He remounted his assault so strenuously that he captured Tyre within days, burned it to the ground, and sold its people into slavery.4
⚠ This is the detail that makes the case worth a page, and Greene's own text contains it.
Four months of siege, and Alexander offered terms. ⚠ The Tyrians' assessment was correct — they had baited him, they had withstood him, and he was preparing to leave.
They were winning when they killed the messengers.
⚠ So the Reversal's lesson is not never provoke the strong. Tyre provoked the strong and it worked. The lesson is narrower and much more useful: they could not tell the difference between a provocation that had succeeded and a licence to repeat it.
⚠ Now it did not matter to him how long the siege lasted or how large an army it needed.
Greene's sentence is precise and the corpus should hold it exactly.
Alexander was already angry at month one — this of course enraged him — and he still ran a cost-benefit calculation and concluded the siege was not worth it. ⚠ Anger did not remove his judgement in the first round.
What the second provocation removed was the budget. ⚠ The question stopped being is Tyre worth this and became is my capacity sufficient — and it was.
⚠ Which is the actual danger of provoking someone stronger: not that they become irrational, but that they stop applying a proportionality test you were relying on. You are not protected by their strength. You are protected by their calculation, and that is the thing anger removes.
⚠ Greene's rule: *"Find the gap in their strength. If there is no gap—if they are impossibly strong—you have nothing to gain and everything to lose."*5
But Tyre had found the gap. ⚠ The city was on water, the siege was expensive, and Alexander's own conduct at month four confirms the assessment.
The rule as stated would not have saved them, because they passed it.
⚠ The rule that would have saved them is not about the enemy's strength at all: a provocation that has worked has spent itself. The information Tyre had — we baited him and got away with it — was true and had no predictive value, because the second provocation was a different act arriving at a different party.
⚠ 🚩 SECONDARY WITHOUT PRIMARY · 🚩 SINGLE SOURCE — no date (the siege is 332 BCE), no ancient source, no historian. Arrian, Diodorus and Curtius all cover it and none is named.
🚩 ⚠ The compression is severe and one element is wrong in a way that matters. "He captured Tyre within days" — the standard accounts have the siege lasting roughly seven months in total, with the final assault following the completion of a causeway built out to the island. ⚠ Greene's account has no causeway, which is the single most famous feature of the siege and the reason the assault eventually succeeded. [CONTESTED].
🚩 ⚠ Without the causeway, the story reads as anger supplied the will — with it, the story is engineering supplied the access. The chapter needs the first.
🚩 ⚠ Burned it to the ground, and sold its people into slavery is reported as the technique's consequence and is the destruction of a city. The corpus should record the mechanism and mark the register.
⚠ Tyre's error is legible and it is not arrogance.
They asked can we withstand him — ⚠ a question about themselves, which they had four months of good evidence for.
They did not ask what does refusing his terms do to him — ⚠ a question about him, for which their evidence was actively misleading, because the thing they had learned was that he could be endured, not that he could be insulted.
⚠ And killing the messengers is the tell. It is not a defensive act; it converts a refusal into an atrocity, and it is the kind of thing you do when you have concluded the other party's responses are no longer your problem.
⚠ The corpus recorded the mirror of this at the Ditch High Priest: each step locally reasonable, the sequence invisible from inside. Tyre's second refusal was locally reasonable too.
You provoked somebody stronger, it worked, and you are considering doing it again.
Treat the success as expenditure, not as evidence. ⚠ This is the case's whole content. We baited him and got away with it is a true statement about the past with no bearing on the next attempt — the first provocation consumed whatever tolerance was there.
Ask what the provocation did to their cost ceiling. ⚠ Not are they angry — anger is compatible with sound judgement, and Alexander proved it at month four. Ask whether the matter has stopped being one they price. The signal is a shift from is this worth it to can I do it, and once it has happened there is no ceiling to appeal to.
Do not close the exit. ⚠ Alexander offered terms and Tyre refused them. A stronger party looking for a way out is the best outcome a provocation can produce, and the only way to lose from there is to take the exit away.
⚠ And notice which question you are answering. Tyre asked can we hold and needed to ask what does this do to him. Evidence about your own resilience tells you nothing about the other party's threshold, and it feels like it does.
Strongest evidence. A two-round structure in which the first provocation demonstrably worked, followed by a repetition that inverted the outcome — with the mechanism stated precisely (it did not matter how long the siege lasted) as a change in the cost ceiling rather than in the anger.
Tension — the Reversal's stated rule (find the gap in their strength) would not have saved Tyre, because Tyre had found it.
Tension — the causeway is missing, and it is the reason the final assault succeeded; without it the case reads as a story about will.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no date, no ancient source, no historian, for a heavily documented siege. 🚩 [CONTESTED] — "captured Tyre within days" against a siege of roughly seven months with a constructed causeway.
Open questions. Was the first bait a provocation at all? ⚠ Tyre's opening position — recognise him as emperor, refuse entry — is a negotiating stance, and a fairly conventional one. Greene calls it baiting because of the outcome, and the corpus should note that the chapter's framing requires Tyre to have been playing the law's game when they may simply have been declining to be occupied. If so, the Reversal's lesson is not about provocation at all.
⚠ This is a fifth Reversal shape — boundary plus a separate positive technique — and the corpus should record the count: character-prescription (Laws 31–34), refusal (35, 37), boundary+case+instruction (36), boundary+three-outcomes (38), and boundary+new-technique (39).
Against the chapter's own Observance the pair is exact and unremarked: ⚠ Selassie built the entire trap before provoking, and Tyre provoked with no trap at all. The Reversal's real content is the corpus's finding on the Observance page — the provocation is the last step, and Tyre made it the first and then the second.
⚠ And against Law 29 the convergence is direct. That law's finding was that success is the dangerous moment, and Tyre is the clean case: the danger arrived at month four, when they had won. ⚠ Ten laws apart, and Law 39's Reversal never mentions that it is describing Law 29.
Decision science — the threshold that stops being applied. Loss Aversion establishes that a shift in reference point changes which outcomes are even weighed, and that the shift is not a failure of reasoning within the frame but a change of frame.
Now it did not matter to him how long the siege lasted.
The insight neither produces alone: the chapter reads Alexander's second response as rage overwhelming judgement. ⚠ The framing account is more precise and more frightening: his judgement was intact both times and the question changed. At month four he was solving is Tyre worth this; after the messengers he was solving is my capacity sufficient — and both are rational problems with correct answers. ⚠ Which identifies what a weaker party is actually relying on when they provoke: not the stronger party's restraint but their continued use of a proportionality test. You are safe while the matter is being priced, and a provocation's real risk is that it moves the matter out of the priced category entirely — at which point strength, which was previously irrelevant, becomes the only variable.
Negotiation — the exit you must not close. The Illusion of Choice treats the operative move as shaping the other party's option set so that their free decision arrives where you need it.
Alexander offered terms and Tyre removed the option.
What the pairing produces: the choice-architecture frame is usually about constructing options for someone else. ⚠ Tyre's case is the destructive inverse and it is the more common error: they eliminated the only branch in which they survived. A stronger opponent who is looking for a way out has, for that moment, a very small option set — accept terms or continue an expensive siege — and Tyre deleted the first. ⚠ Which yields the rule the Reversal gestures at and does not state: when a stronger party offers you an exit, the exit is the whole negotiation. Everything you have achieved is realised at that moment or not at all, and killing the messengers is the maximally efficient way to convert a favourable position into a terminal one.
Sharpest implication. ⚠ The first provocation worked — four months of siege and Alexander offered terms — and Tyre could not tell the difference between a provocation that had succeeded and a licence to repeat it. And what the second one removed was the cost ceiling, not the judgement: it did not matter how long the siege lasted. ⚠ You are not protected by a stronger party's restraint but by their continued use of a proportionality test — and the danger is that the matter moves out of the priced category, at which point their strength becomes the only variable.
Generative questions.