In 1502 a huge piece of marble was sitting in a dark hall of Santa Maria del Fiore in Florence, gathering dust.
It had been magnificent raw stone. Then an unskilled sculptor bored a hole straight through it where a figure's legs should have been, and generally mutilated it. Piero Soderini, Florence's mayor, had thought about commissioning Leonardo or some other master to rescue it, and had given up — everyone agreed the stone was ruined.1
Michelangelo's friends wrote to him in Rome. He came, looked at it, and concluded he could get a fine figure out of it by adapting the pose to the way the rock had been mutilated.2
Soderini said it was a waste of time, that nobody could salvage such a disaster, and eventually let him try. Michelangelo decided on a young David with a sling.2
Hold on to the opening detail, because the whole page turns on it: the artist's first act was to accept a constraint as given and design around it rather than argue it away.
Weeks later, with the statue nearly finished, Soderini came into the studio.
Greene's phrase is exact: "Fancying himself a bit of a connoisseur, he studied the huge work, and told Michelangelo that while he thought it was magnificent, the nose, he judged, was too big."3
Two things are true at once and both matter.
The judgement is wrong — Michelangelo saw immediately that Soderini was standing directly under the giant figure and didn't have the proper perspective.3
The judgement is also binding. This is the mayor, and Greene notes what arguing would have cost: "To offend such a man by arguing would not only gain Michelangelo nothing, it would put future commissions in jeopardy."4
So the artist faces the exact situation the engineer faced at the start of this law — a powerful man, confidently wrong, about a technical matter the artist knows better. And Michelangelo's own risk is bigger than one nose: "by changing the shape of the nose he might ruin the entire sculpture."4
He says nothing. He gestures for Soderini to follow him up.
At the nose he picks up his chisel — and also "a bit of marble dust that lay on the planks."3
With Soderini a few feet below him on the scaffolding, he taps lightly with the chisel and lets the dust fall from his hand, little by little.
He actually did nothing to change the nose, but gave every appearance of working on it. After a few minutes of this charade he stood aside: "Look at it now." "I like it better," replied Soderini, "you've made it come alive."3
Nothing on the statue. Not a grain of it.
Two things about Soderini changed, and it is worth separating them because Greene doesn't.
His position. He was now several feet up a scaffold, close to the nose, instead of standing underneath a giant figure looking up. The proportion he had objected to was a product of where he had been standing, and he was no longer standing there.
His belief. He now thought a correction had been made.
Both are load-bearing, and only one of them is honest.
Here is the thing this page exists to record, because the chapter does not.
Greene's interpretation names one cause: "His solution was to change Soderini's perspective (literally bringing him closer to the nose) without making him realize that this was the cause of his misperception."4
On that reading, this is a genuine demonstration. Soderini's eyes were given better data. The nose really does look right from up there. Nothing false was needed.
But the anecdote does not stop at moving him. Michelangelo also collected dust in advance, tapped a chisel on nothing, performed a charade, and said "look at it now" — which is an assertion that something has been done.
And Soderini's reply settles it. He does not say I was standing badly. He says "you've made it come alive." He is crediting an alteration that did not occur.
So the case contains two mechanisms and Greene's interpretation names only the first while the punchline requires the second.
If the perspective change were doing the work alone, the dust would be unnecessary. Michelangelo brought the dust because he did not trust the perspective change to be enough — he needed Soderini to not notice that his objection had evaporated on its own, since noticing that is noticing he had been wrong.
The dust exists to protect the patron from learning anything.
This matters beyond one anecdote, because Law 9's whole promise rests on it.
The judgment is "get others to agree with you through your actions."5 Greene's summary of this case is bolder still: "No one is offended, and your point is proven."4
Michelangelo's point was not proven. Soderini left the studio believing the nose had been too big and had been fixed. He was wrong when he arrived and wrong when he left, about both facts. What he acquired was not agreement but the sensation of having been right and heeded.
Set that against the chapter's other cases and the law splits cleanly in two.
Real demonstrations — the audience genuinely comes to know something. Khrushchev's silent hall learns what facing Stalin felt like. Amasis's Egyptians learn how reverence attaches to things. The vizier's sultan learns what thirty years of care should produce.
Placations — the other party learns nothing and is left in a comfortable false belief. Michelangelo's nose. Wren's dummy columns, where the mayor never discovers the columns are fake and only "posterity" is expected to understand.6
Greene runs both under one heading and grades them identically, because his success criterion is did the operator get what he wanted without giving offence. On that criterion they are the same. On the criterion the law's own language uses — proven, agree — they are opposites.
Placation has a bill and it arrives later.
Soderini remains a man who believes he has a good eye and that his interventions improve statues. Nothing corrected him; in fact he received strong evidence that he was right. The next commission will bring the next judgement, delivered with more confidence.
The technique protected one statue and reinforced the condition that endangered it.
That is a real cost and this book has no line for it, because Greene's unit of analysis is the individual encounter. Fortunately for posterity, Michelangelo found a way to keep the perfection of the statue intact4 — and the statue is what survives, so the ledger looks clean.
A client with authority and a mistaken eye has told you to damage the work.
First, check whether they're wrong about the thing or wrong about the vantage. Michelangelo's whole opening is diagnostic: he identified that Soderini's error was positional. Many client notes are like this — a reaction to how the work was encountered rather than to the work.
Second, move them rather than telling them. If the objection is a product of context, change the context. This part is honest, costs nothing, and is the half most people skip because it feels indirect.
Third, decide deliberately whether you are demonstrating or placating. They are different acts with different bills. Ask: do I need this person to understand, or do I only need them to stop? Michelangelo only needed him to stop.
Fourth, if you are placating, keep the work intact. The single non-negotiable is that nothing was actually done to the nose. A placation that damages the work to buy peace is the worst of both.
Fifth, price the repeat. You will work with this person again, and a placation leaves their judgement untouched and slightly emboldened. If the relationship is long, at some point the cheaper move is to let them be visibly wrong once.
And the honest note: what Michelangelo did was a deception, performed on a man who trusted him, to keep a commission. It was proportionate, it harmed nobody, and it saved the David. It was still a charade, and Greene's word for it is charade too.
Strongest support. The scene is completely specified — the position, the dust gathered in advance, the light tapping, the line delivered — and every element is mechanically necessary. It's also a rare case where the source gives the other party's exact words, which is what makes the analysis above possible at all.
Tension — Greene's interpretation contradicts his own anecdote. He attributes the result to the perspective change; Soderini's reply credits an alteration that never happened. Documented above, not resolved.
Tension — the flagship case of demonstrate, don't explicate contains no demonstration. Nothing was proven to anyone. This is the law's Observance and it does not do what the law says.
Tension — against the vault's psychology corpus on feedback. The technique works by ensuring the other party never receives corrective information about their own judgement. That is exactly what most of the vault's material on skill acquisition and self-knowledge identifies as the mechanism by which bad judgement persists.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — the anecdote is traditionally attributed to Vasari, and Greene names no source. Michelangelo's David, the flawed block and Soderini's mayoralty are all matters of record; the studio scene and the reported dialogue are not corroborated here. [POPULAR SOURCE]
Open questions.
Greene stages this against The Arguer — the engineer who explained three times and was flogged to death — as observance against transgression.
The pairing is doing less work than it appears to, and the gap is the useful part.
The engineer needed the consul to know the smaller mast was better, because a ram that fails is a siege that fails. His problem was not that he wanted agreement instead of compliance; it was that the false belief had physical consequences he could not design around.
Michelangelo needed nothing of the kind. He needed one afternoon and one man's mood, and the false belief was harmless.
So the two cases are not the wrong and right ways to handle the same problem. They are two different problems, and only Michelangelo's is solvable by the method the law recommends. Greene's own opening detail proves it — Michelangelo took the ruined block and adapted the pose to the mutilation. The whole case is built on a man for whom the constraint could be absorbed by the work.
Which is why the law reads so persuasively and transfers so poorly. Its showcase is drawn from a domain where the client's belief and the object's integrity can be separated. In the engineer's domain they cannot, and the book files his death under his own poor technique.
→ history · Amasis and the Golden Footbath
Amasis, despised for humble origins, melts a gold footbath into a statue of a god, lets the Egyptians revere it for a while, then convenes them and reveals what it used to be — a vessel they had washed their feet and vomited in — and says his own case is the same.
Both are demonstrations by a person who cannot afford to argue. But they run in opposite directions and the contrast defines the whole law.
Michelangelo's demonstration ends with the audience not knowing. The dust exists to conceal that the objection dissolved on its own.
Amasis's demonstration ends with the audience being told outright. He builds the experience, then names it, out loud, and even adds an explicit verbal argument on top: so just as they had come to revere the transformed footbath, they had better pay honour and respect to him.
The insight neither page yields alone: Amasis breaks Law 9's stated rule — he explicates, at length, after demonstrating — and it is the more durable result. Michelangelo bought an afternoon; Amasis converted a hostile population permanently.
Which suggests the actual technique is sequence, not substitution. Demonstration first, so the claim is felt rather than asserted; explanation second, so the audience knows what they have learned and can carry it. Demonstrate, do not explicate forbids the second half, and the chapter's own most successful case does it anyway.
→ behavioral-mechanics · Count Victor Lustig and the Blank Telegrams
Lustig received a steady stream of telegrams and tore them up unread in front of people. They were blank, and the full effect ran anyway.
Michelangelo's chisel is the same instrument in a respectable setting. The tapping conveyed nothing; the dust came from the planks, not the statue; the words look at it now asserted a change that did not exist.
The pair names a category the vault has been circling: the empty prop. A physical object or action that carries no content, whose entire function is to license an inference the observer then makes and owns.
The insight neither produces alone is about why it's so hard to resent. A person who is lied to can identify a false statement. Soderini was never told anything untrue about the statue — he was shown a man working, which is not a proposition, and he supplied the nose has been improved himself. That is why the technique leaves no offence behind, and it is also why Greene's promise of no one is offended is delivered on a technicality: nobody is offended because nobody can locate what was done to them.
Sharpest implication. The Observance of demonstrate, do not explicate is not a demonstration. Michelangelo proved nothing, changed nothing, and sent his patron away holding two false beliefs — that the nose had been wrong, and that it had been fixed. The law's real content, visible once its own case is read closely, is not "show instead of tell" but "let them keep being right" — and those are different instructions with different costs, only the first of which Greene priced.
Generative questions.