Eighth century B.C. The Medes have thrown off Assyrian rule and, determined never to be ruled that way again, refuse to give any one man ultimate power. No monarchy. No leader.
The country immediately falls apart — small kingdoms, village fighting village.1
In one village there is a man named Deioces who becomes known for fair dealing and for settling disputes well. He does it so successfully that every legal conflict in the area comes to him. The courts elsewhere are corrupt and nobody trusts them, so villages far and wide start bringing him their cases. He becomes the sole arbiter of justice in the land.2
And at the height of it, he stops.
He will no longer sit in the chair of judgment. He complains that other people's problems have consumed so much of his time that he has neglected his own affairs, and he retires.3
The country descends into chaos again — worse than before, because now there had been something and it was gone. Crime rises. Contempt for the law is "never greater." The Medes hold a meeting of all the villages and conclude they cannot live this way. They will have to appoint a king.4
The man they want, of course, is Deioces.
He is hard to convince. He wants nothing more to do with their infighting and bickering. They beg. He finally agrees — and then attaches terms.5
An enormous palace, to be constructed for him. Bodyguards. A capital city built from which to rule.
All of it is done. And then he sets the terms of his rule: "Admission to his presence was forbidden. Communication with the king was only possible through messengers. No one in the royal court could see him more than once a week, and then only by permission."6
He ruled fifty-three years, extended the empire, and laid the foundation for what became the Persian empire under his great-great-grandson Cyrus. During the reign, "the people's respect for him gradually turned into a form of worship: He was not a mere mortal, they believed, but the son of a god."7
Greene quotes the source directly, and it is worth having exactly because it is the clearest statement of the law anywhere in the chapter:
"There was a risk that if they saw him habitually, it might lead to jealousy and resentment, and plots would follow; but if nobody saw him, the legend would grow that he was a being of a different order from mere men."8
That is not a claim about scarcity increasing value. It is a claim about what fills the gap.
Seen habitually, a ruler is a man — and a man in a chair everyone can see is a man whose chair other people can imagine occupying. Not seen, the same ruler is whatever the population needs him to be, and what they supply is not skepticism. It is divinity.
Notice the two named failure modes: jealousy and resentment. Both require comparison, and comparison requires a visible object. Absence does not make people admire you more. It removes the material they would otherwise use to measure you against themselves.
The case is usually read as one move. It is two, and they do completely different jobs.
The first withdrawal — quitting the judgment chair — was a demonstration. Greene's interpretation: by serving so many clients, Deioces "had become too noticeable, too available, and had lost the respect he had earlier enjoyed. People were taking his services for granted."9 The retirement made the country experience life without him. It converted an assumption into a felt absence.
That withdrawal is reversible by design. He came back. It was a proof, and proofs have to be observed to work.
The walls, the messengers, one audience a week — this is a different thing entirely. It is permanent, and it is not a demonstration of anything. Nobody is being shown what life without Deioces is like; he is ruling the whole time. What is being prevented is observation.
The first withdrawal manufactures demand. The second prevents evaluation. Greene runs them together under one law, and they have opposite relationships to the audience: one requires the audience to notice and conclude something, the other requires them never to see enough to conclude anything.
Which also explains why the second one had to come after the first. You cannot become unobservable before anyone has decided they need you — that is not mystery, it is absence, and it is the failure the law's own Reversal warns about.
The chapter treats this as a technique available to a reader. Two features make it much narrower than that.
He had actually been good. The reputation was earned over years of correct judgments in a country where every alternative court was corrupt. The withdrawal worked because there was a genuine capability whose loss was genuinely felt. Remove that and the retirement is just a man who stopped showing up.
He had a monopoly. The Medes had no other arbiter. The demonstration works only if the thing withdrawn has no substitute — which is a fact about the market, not about the technique. In a country with three good judges, Deioces retires and the cases go to the other two.
So the case demonstrates the law under conditions that make almost any withdrawal work, and the chapter does not say so.
Greene's interpretation ends with the palace as the law's "ultimate realization."10 It stops there.
But the record he is summarising continues, and the ending is available in the same source: Deioces founded a dynasty that produced the Persian empire, and it produced a court in which the ruler was structurally unable to be seen. That is the same architecture Law 18 spends a chapter identifying as fatal — Shih Huang Ti in his 270 pavilions, sleeping in a different room each night, dying so isolated that almost nobody knew.
Two laws, two chapters apart, recommending and condemning the same building. Law 16 calls the walls the ultimate realization of a principle; Law 18 calls a fortress "a symbol of power's isolation" and "an easy target."
Greene does not cross-reference them. The variable that would sort them is available in his own two cases — Deioces withdrew from a position of manufactured demand, Shih Huang Ti withdrew from fear — but neither chapter names it.
You are the person everyone comes to.
Not formally. There is no title attached to it. You are just the one who knows how the thing actually works, and so the questions route to you — from your team, from the next team, from people who do not report to you and never will. You are good at it, and being needed feels like standing.
Notice what is actually happening to the price. Every answer you give for free establishes that answers are free. The volume that feels like influence is the mechanism by which the influence is being spent, and there is no moment at which anyone will notice and stop.
Deioces' move was not to complain about the volume, and this is the part people get wrong. He did not announce he was overloaded, negotiate for recognition, or ask for a title. He stopped, gave a mild and non-accusatory reason, and let the absence do the arguing.
The reason mattered. My own affairs have suffered — nobody's fault, no grievance, nothing for anyone to defend against. A withdrawal framed as a complaint invites a negotiation about whether the complaint is fair. A withdrawal framed as unremarkable invites nothing, and leaves the room to discover the cost on its own schedule.
Then the two conditions, honestly. Were you actually good, and is there a substitute? If the answers are yes and no, this works. If there is a substitute, you will not be missed — the cases go to the other two judges, and you will have taught the room that you are optional. That is not a reason to keep answering for free. It is a reason to solve the substitutability problem first, which is a different and slower piece of work than withdrawing.
The strongest thing here is the source. Greene quotes Herodotus directly and by name, and the quoted sentence contains the mechanism rather than merely the outcome — which is unusual for this book and puts the case on firmer ground than most of the corpus.
Tension: the case contains two different withdrawals and the law has one category. The demonstration (reversible, requires an audience to notice) and the concealment (permanent, requires an audience never to see) do opposite work and are presented as one technique.
Tension: Law 16 and Law 18 recommend and condemn the same architecture. Deioces' walls are the law's "ultimate realization"; Shih Huang Ti's are a fatal error two chapters later. Neither chapter references the other. Documented, not resolved — the deciding variable appears to be whether the withdrawal follows manufactured demand or follows fear, and Greene names it in neither place.
Tension: the preconditions are invisible in the chapter. Real capability and no substitute. Both are doing enormous work and neither is stated as a condition.
🚩 SECONDARY WITHOUT PRIMARY does not fully apply — Herodotus is named. But no book, section or translator is given, and the surrounding narrative in the Histories (including how far Herodotus himself is reporting a Median tradition rather than a documented reign) is not available from this source. [POPULAR SOURCE] on Greene's framing; [PLAUSIBLE — needs corroboration] on the fifty-three years and the sequence of events.
Open questions. Was the retirement planned as a manoeuvre, or genuine and then exploited when it worked? Greene asserts intention — "He determined early on that the country needed a strong ruler, and that he was the man for the job"11 — with nothing behind it, and the case reads equally well as an opportunist recognising a gift. And: does the second withdrawal have a cost that a fifty-three-year reign conceals? A ruler nobody can see is a ruler who cannot hear, which is Law 18's entire argument.
Within Law 16 the sharpest pairing is with Guillaume de Balaun, the chapter's other long case, and the two are near-perfect inverses.
Deioces withdrew deliberately, from a correct reading of his position, and got a kingdom. Guillaume withdrew deliberately too — but for the wrong reason, expecting to produce anger and reconciliation, and got devotion he did not want and then a genuine rupture he could not undo.
Both men engineered an absence and only one of them understood what absence does. Which produces the finding the chapter has no room for: the technique's outcome does not depend on the operator's intent or skill at withdrawing. It depends on what the other party's imagination supplies while you are gone, and that is not a thing the operator controls. Deioces got divinity. Guillaume got a woman kneeling in a courtyard begging forgiveness for a slight that never happened — a failure, from his point of view, produced by the technique working.
Against the vault's existing Absence and Withdrawal as Active Strategy, this case supplies the evidence that page was built without. It also corrects it: that page treats absence as a tactical insertion, and Deioces' second withdrawal is a permanent architecture, which is a different instrument with a different failure mode.
Behavioral mechanics — The Air of Mystery
Law 6's material on unresolvability: a figure who cannot be finished reading holds attention, because the mind keeps returning to an unclosed loop.
Deioces is that mechanism given a wall and fifty-three years, and the pair produces something neither has alone.
Law 6's mystery is a property of the person — you behave unresolvably, and people cannot settle you. Deioces' is a property of the architecture, and the difference is that it does not require him to do anything at all. Once admission is forbidden and communication runs through messengers, the mystery is produced by the building whether the king is brilliant, ordinary or asleep. Which means the second withdrawal converts a performance into an infrastructure, and infrastructure does not have off days.
That is the real upgrade and it is also the trap. A performed mystery can be adjusted when it stops working; you can appear, correct an impression, show something. An architectural mystery cannot be selectively suspended, because the moment the king is seen the whole apparatus that made him a being of a different order is spent in one afternoon. Herodotus's sentence is usually read as an argument for concealment. It is equally an account of a position with no exit.
Business — Esoteric Interests as an Irreplaceability Moat
That page's argument is about substitutability: the durable position is not being better at a common thing but being the only person at a specific intersection, so there is no comparison set.
Deioces' case looks like a scarcity story and is actually a substitutability story, which is the correction the pair produces.
Greene reads the withdrawal as creating value by removing supply. But removing supply only creates value if demand cannot go anywhere else — and the Medes' judges were corrupt, so it could not. The withdrawal did not manufacture Deioces' irreplaceability; it revealed an irreplaceability that already existed and was being given away for free.
That distinction has a sharp practical edge the chapter obscures. Withdrawal is a test before it is a technique, and it is a test you cannot fail safely — if a substitute exists, the absence teaches everyone that you are optional, permanently, and you cannot un-run it. Which means the sequence the business page implies is the correct one and Greene's is backwards: build the position with no comparison set first, and the scarcity move becomes almost unnecessary, because a genuinely unsubstitutable person does not need to disappear to be valued. Deioces' retirement was cheap for him precisely because he was never at risk of the answer coming back wrong.
Sharpest implication. What absence does is not raise your value — it removes the material people use to measure you. Herodotus names the two things habitual visibility produces, jealousy and resentment, and both require a comparison an unseen man does not supply. Which means the payoff is not admiration but incomparability, and incomparability is a position with no way back.
Generative questions.