A merchant is going travelling and leaves his iron with a friend for safekeeping — "a great number of bars of iron, which were a principal part of his wealth."1
He comes back broke. He asks for the iron.
The friend, who owed money and sold it, says: "there was a rat in the room which ate it all up."1
Now watch what the merchant does, because the obvious moves are all available and he takes none of them.
He does not call the man a liar. He does not go to a magistrate. He does not point out that rats do not eat iron.
He agrees. "It is a terrible misfortune to me indeed; but I know of old that rats love iron extremely; I have suffered by them many times before in the same manner."1
The friend is delighted — "extremely pleased" — and, to smooth things over, invites him to dinner the next day.
Between those two moments, the merchant meets one of the friend's children in the street, takes him home, and locks him in a room.
The chapter's Moral Effect is "you teach others a lesson by giving them a taste of their own medicine" — "you mirror what other people have done to you."2
Take that literally and the merchant should have stolen something of the friend's and sold it.
He did not. Nothing was stolen and nothing was sold. What he mirrored was not the theft.
He mirrored the demand.
The friend's actual offence was not taking the iron; it was requiring the merchant to accept a physically impossible story and let it stand between them. So the merchant constructed the one thing structurally identical to that: a loss, explained by an impossibility, presented with a straight face and an invitation to believe it. "Yesterday in the evening, as I parted from hence, I saw an owl in the air with a child in his claws."3
That is a much more precise instrument than do to them what they did to you, and it is worth naming because it is what makes the fable work where Ivan's two years of national humiliation only half-worked. The Moral Effect as Greene teaches it points at the behaviour. Pilpay's merchant points at the epistemics — not you stole from me but you asked me to believe something that cannot be true, and here is what that feels like from where I was standing.
You can mirror a physical act only if you are willing and able to commit it. You can mirror an insult to someone's intelligence with nothing but a sentence.
The decisive moment is the friend's own outburst, and it repays slowing down.
*"Why, you most foolish and absurd creature!" replied the friend, "are you not ashamed to tell such an egregious lie? An owl, that weighs at most not above two or three pounds, can he carry a boy that weighs above fifty?"*3
The friend has just done the merchant's entire work for him.
He has stated the principle — mass constrains what a small animal can do to a large object — with numbers, indignantly, in his own words, while believing the subject is owls. He has also supplied the moral vocabulary: ashamed, egregious lie, foolish and absurd.
Every term the merchant needs is now on the table, and none of it came from the merchant.
All that is left is one sentence of arithmetic: "As if in a country where one rat can eat a hundred tons' weight of iron, it were such a wonder for an owl to carry a child that weighs not over fifty pounds in all!"3
Notice he still does not accuse. He does not say you lied about the iron. He simply sets the two quantities beside each other and stops. The friend does the last step himself, and the fable says so: he "found that the merchant was no such fool as he took him to be."3
A conclusion the target reaches in their own voice is not available to be argued with, because arguing means contradicting the sentence they just shouted. And the reason this fable ends in restitution while the chapter's other Moral Effect cases end in humiliation or diffuse compliance is that the merchant engineered the target's own words into the trap and then got out of the way.
Now the second thing the merchant does not do, which is the harder discipline.
He has won. The friend is exposed, in his own house, having just called someone else a shameless liar seconds before discovering he is one. This is the moment to say it.
The merchant says nothing more.
And the fable's outcome runs through that silence. The friend "begged his pardon for the cheat which he designed to have put upon him, restored him the value of his iron, and so had his son again."3
He confesses. Voluntarily. Without being accused.
That is the whole difference between this fable and the one Greene prints on the facing pages, where La Fontaine's fox is served meat in a narrow-necked urn, gets nothing, and goes home "empty-bellied, tail dragging, ears drooping" — having learned exactly nothing.
The corpus has already recorded the variable as an exit, and this fable shows what an exit is made of. The merchant left his friend a version of events he could concede to: I was caught being clever, by someone cleverer, and the sensible thing now is to settle. That is a survivable story. What he withheld was the other version — you are a thief who robbed a friend and I have proved it — which is true, and which nobody can accept out loud.
A mirror that forces a confession of character produces a defence. A mirror that permits a confession of tactics produces a settlement. The merchant wanted his iron, not an apology, and he built the exit that got it.
The fable is charming and it should not be allowed to stay charming, because of what is sitting in the middle of it.
He took a child off the street and locked him in a room. Overnight.
And the father's night is described. He had the boy "cried by sound of trumpet" through the city — the public alarm, the whole town told a child is missing — and by morning he is telling his guest "I have lost one of my children… I know not what is become of him."3 That is a parent who has spent a night believing his son might be dead.
The fable frames this as ingenuity. Greene reprints it in a chapter recommending the technique, in a margin, uncommented — beside a passage whose only technical advice about the Moral Effect is "do not be afraid to add a touch of caricature and exaggeration."4
There is no upper bound anywhere in that instruction. The chapter never asks how far the reflection may go, and this marginal is the case that makes the omission conspicuous, because the answer it silently gives is as far as a hostage.
The proportionality question is not a squeamish one. It is the operational question the Moral Effect most needs and never gets: the technique's power comes from making the lesson felt, and felt scales with harm, so a method with no ceiling has a built-in drift toward whatever produces the strongest sensation. Ivan reached two years of national ridicule. The merchant reached a child in a locked room. Both are described as successes.
There is also a plainer reading available, and holding both is more honest than picking one.
The merchant did not merely stage a demonstration. He took collateral and held it until he was paid.
The sequence is a hostage exchange with a lesson wrapped around it: seize the asset, let the counterparty feel the loss, name your price by analogy, release on payment. "Restored him the value of his iron, and so had his son again"3 — the two clauses are joined, in that order, and the and so is doing the work of a receipt.
Which means the fable proves less about the Moral Effect than it appears to. Would the friend have paid if the merchant had merely made the owl joke without holding the boy? The fable cannot say, because it never separates the demonstration from the leverage.
Greene's chapter credits the first mechanism and the second one is what actually guarantees the outcome. That is the same structure the build has logged elsewhere in this book — a technique demonstrated by a case whose result was secured by something the chapter does not count.
One more detail worth rescuing, because the fable supplies it and then does nothing with it.
The friend sold the iron because he "owed several sums of money, having sold the iron to pay his own debts."1
Not greed. Debt.
The story hands us a man in financial trouble who took what was in his house and made up something absurd because he had nothing better, and then treats him purely as a cheat to be outwitted. His circumstances change no part of the fable's operation and are never mentioned again.
It is a small thing and it points at a real limit of the whole Moral Effect family. The technique works on behaviour and is blind to cause. A lesson delivered by mirror teaches the target what their action felt like from the other side; it teaches the teacher nothing about why the action happened. The merchant got his money back from a man who needed money, and neither of them ever discussed that.
Someone has given you an explanation that cannot be true.
Not a lie you can prove — the useful kind, the kind with no evidence either way. The file was definitely sent. The message must have gone to spam. The delay was on their supplier's end. You know it is not true and you cannot show it, and the thing that stings is not the loss, it is being expected to accept it and move on.
Your two obvious options are both bad. Accept it, and you have taught them that this works on you. Challenge it, and you are in an argument about evidence you do not have, which you lose while looking petty.
So take the merchant's route. Accept it warmly and completely. Of course, spam filters are terrible at the moment. Give them nothing to defend against — the whole play depends on them believing they got away with it, because a person who thinks the danger has passed is a person who will speak freely.
Then, shortly after, hand them a structurally identical impossibility of your own. Not a punishment, not a costly one — the same shape. Your report is late and your document management system appears to have eaten it. Deliver it exactly as they delivered theirs: straight-faced, mildly regretful, and with the same expectation of being believed.
Wait for the sentence you actually need, which is theirs. Systems don't just eat files. That is the whole point of the exercise. You have now got the principle stated, out loud, in their voice, about a case they think is yours.
Then do the minimum. Set the two beside each other, lightly, and stop talking. No — they don't, do they. Do not complete it. Do not say so about that email. The step you leave undone is the step that makes it theirs.
And then leave the door open, which is the discipline the whole thing turns on. They need a version of this they can walk out of — we both fudged something, let's not — and you have to let that version stand even though the true version is that they lied and you did not. Say nothing about character. Take the fixed behaviour and let it go.
Two hard limits, both from the fable's own worse half. Match the shape, never the stakes — the merchant's version worked and it worked by taking a child, and the technique contains no ceiling of its own, so you supply one. And if what you actually need is the thing back rather than the behaviour changed, then this is not your tool: the merchant got paid because he was holding the boy, not because he made a joke about owls.
Strongest evidence is the friend's outburst, because it is the mechanism visible in the text rather than inferred. He states the disproof of his own lie, unprompted, with numbers, in a case he believes is about somebody else's dishonesty.3
Tension one — the fable does not separate the lesson from the leverage. The merchant was holding a hostage when the restitution was made. Whether the demonstration alone would have produced payment is unrecoverable, and the chapter credits only the demonstration. Preserved, not resolved.
Tension two — the two Moral Effect marginals in this chapter give opposite outcomes and the body of the chapter uses neither. Pilpay's works and produces restitution; La Fontaine's fails and teaches nobody. Printed pages apart, neither cited.
⚠ Proportionality is absent from the fable and from the chapter. The Moral Effect's only technical instruction is to add caricature and exaggeration,4 with no ceiling stated anywhere in Law 44, and this marginal's ceiling turns out to be a child in a locked room. Recorded as an omission in the source, not a claim about what the source intends.
🚩 SECONDARY WITHOUT PRIMARY. Attributed as "FABLES, PILPAY. INDIA. FOURTH CENTURY."5 No translator, no edition, no translation date. The Pilpay corpus reaches English through a long chain — Sanskrit to Middle Persian to Arabic to Persian to European vernaculars — and the chain is invisible here.
🚩 TRANSLATION DISTORTION. The English is visibly early-modern ("in regard that he was not very wealthy," "had him cried by sound of trumpet"), which places this rendering many centuries and several languages from any fourth-century Indian original. [LOW CONFIDENCE] on any claim about what the story meant in its own setting; the structural reading above is a reading of the English text as printed.
Open question — is the epistemic mirror generally more available than the behavioural one? Mirroring a demand to be believed costs a sentence; mirroring an act requires committing it. If that holds, the Moral Effect's usable range is much wider than the chapter's taste of their own medicine framing suggests — but this is one case.
Open question — what does the exit cost? The merchant bought restitution by never naming the theft. The friend keeps his self-account intact and remains a man who will do this again to someone else. Nothing in the fable prices that.
Within Law 44 this marginal is one half of a pair, and the pair is more informative than either piece.
The Fox and the Stork is the same technique failing. The stork serves the fox in a long-necked urn, the fox goes home hungry and humiliated, and nothing is repaid, admitted or repaired. The two fables sit a few pages apart in the same chapter, in the same margin, illustrating the same effect, with opposite results — and Greene draws no conclusion from either.
The variable is the exit. Pilpay's merchant leaves a story his friend can concede to; La Fontaine's stork leaves nothing but the humiliation. The chapter's own margins contain a controlled experiment on the Moral Effect's success condition and the chapter does not run it.
Against Ivan and the Tartar Czar, the Observance this effect is officially built on, the merchant's version is better in every respect the chapter cares about. It is aimed at one person who definitely remembers the original — no old enough to remember problem. It produces a specific, measurable outcome. It repairs the relationship rather than ending it. And it costs two days rather than two years and a country's standing. The best case for the Teacher's Mirror in Law 44 is a fourth-century fable in the margin, not the Observance the chapter chose.
And against The Moral Effect itself, this fable exposes the definition's imprecision. Mirror what other people have done to you would have had the merchant stealing; what he actually mirrored was the demand to accept an impossibility. The corpus should record the Moral Effect as having at least two forms — behavioural and epistemic — which the chapter runs together under one name.
Self-Persuasion Principle — The Prospect Persuades Themselves — the observation that the same fact carries different weight depending on the channel it arrives through: told that a restaurant is good, you nod; discovering it yourself, you go.
The merchant's fable is the adversarial edition of that principle, and putting them together sharpens both.
The sales version is about conviction — a conclusion the buyer reaches themselves is acted on, where the same conclusion supplied by a seller is merely noted. The mechanism is that self-generated conclusions are not filed as claims made by an interested party, so they never get discounted.
Pilpay's merchant needs something stronger than conviction. He needs the friend to hold a position he cannot retreat from — and self-generated conclusions have a second property the sales frame does not need and does not name: they are hard to abandon, because abandoning one means contradicting yourself out loud. The friend cannot walk back an owl cannot carry a fifty-pound boy without becoming, in that instant, a man arguing for the physically absurd. He is pinned by his own sentence, in his own house, in front of the person he stole from.
So the pairing yields a distinction the business page has no occasion to draw. Self-persuasion in a sale is valuable because it motivates. Self-persuasion in a confrontation is valuable because it commits. And the second use has a design requirement the first does not: the target must state the principle while believing the case is about someone else — which is why the merchant needed an owl and a different child, and why simply asking do you think rats eat iron? would have produced nothing but a shrug and a shift.
The Hostage Guard: Kinship as Control — Genghis Khan's practice of selecting the most promising young warriors from conquered noble families into his personal guard: genuine honour, genuine advancement, and simultaneously a standing hostage arrangement that guarantees the families' behaviour.
The merchant's child in a locked room is the same instrument, run for two days instead of a generation, and the comparison is what makes the fable's second mechanism visible.
Both cases hold a relative rather than a possession, and both are structured so the holding is never named as a threat. Khan's guard is an honour that would be an insult to refuse; the merchant's child is missing, not taken — the father is never told his son is held, and the merchant never once mentions a condition. In both, the leverage does its work while formally not existing, which is precisely why neither party has to acknowledge it and why the arrangement can end without anyone being defeated.
What the pairing produces is a correction to how the fable is usually read, this chapter included. It is filed as a case of a lesson taught — cleverness, wit, the pleasure of the reversal. The child is not part of the lesson. He contributes nothing to the argument about owls and rats; the analogy would work identically with a hypothetical child, or an invented one. He is there to make non-payment expensive.
Which means the fable is running two instruments at once and crediting one: an epistemic mirror that makes the friend unable to maintain the lie, and a hostage that makes him unwilling to try. Khan's case, where the pedagogy is absent and only the leverage remains, is the control condition — and it worked fine without any lesson at all.
Sharpest implication. The Moral Effect's success condition is not the accuracy of the mirror but the exit — whether the target is left a version of events they can concede to without confessing to a character. The merchant never accused his friend of anything; he set two impossibilities side by side and stopped talking, and got his money, his relationship and an apology he had not asked for. La Fontaine's stork, holding a technically perfect mirror with no door in it, got a hungry fox who learned nothing. And the fable's uncomfortable second half is that none of this is why the merchant was paid: he was holding the man's son, and the chapter that recommends this technique has no ceiling on it anywhere.
Generative questions.
If mirroring an epistemic demand costs a sentence while mirroring an act requires committing it, why does the chapter — and everyone else — teach the expensive version?
The exit works by letting the target keep a self-account that is false. What is the cumulative cost of settlements bought that way, for the person granting them and for whoever the target does it to next?
The fable gives the thief a motive — debt — and then ignores it entirely. Is a technique that operates on behaviour and is structurally blind to cause a tool, or just a way of winning exchanges with people whose situation you have declined to learn?