Talleyrand did not like to work hard.
That is Greene's opening claim, and it is not an insult — it is the whole method:
He never liked to work too hard, so he made others do the work for him—the spying, the research, the detailed analyses. With all this labor at his disposal, he himself never seemed to strain.1
And then the effect it produced:
*When his spies revealed that a certain event was about to take place, he would talk in social conversation as if he sensed its imminence. The result was that people thought he was clairvoyant.1
A man at a party makes a light remark about how things are going in Vienna. Three weeks later the thing he mentioned happens. Nobody in the room saw the courier.
The detail that makes this more than "he had good intelligence" is the form the knowledge came out in.
His short pithy statements and witticisms always seemed to summarize a situation perfectly, but they were based on much research and thought.1
Not a briefing. Not a memorandum. A witticism, in social conversation, apparently unprepared.
That form is doing three things at once, and only the third is the one Greene names.
It compresses. A sentence that summarises a situation perfectly is a compression of a great deal of material, and compression is expensive — it costs more work than the long version, not less.
It arrives in a register where nobody checks sources. A briefing invites how do you know. A remark at dinner does not. The setting does the epistemological work: nothing said lightly gets audited.
It conceals the labour. Which is the reason Greene puts it in this chapter.
The first two are the more interesting and he does not discuss them. The conversational register is not merely where Talleyrand hid the work — it is where the claim became unfalsifiable in the moment and impressive in retrospect.
Here is the problem with this case that the chapter does not raise.
Talleyrand's reputation was for prescience — an inexplicable feel for events. That is a reputation with a specific structural weakness: it cannot survive a miss, and it will be tested constantly, because a man believed to see the future gets asked about the future.
An intelligence network produces accurate anticipation at some rate below one hundred percent. The clairvoyance reading of that performance assigns the successes to a faculty and has no account of the failures except that the faculty lapsed.
Compare the alternative reputation available from the same facts: this man is exceptionally well-informed. That version survives a miss easily — good sources are sometimes wrong — and it is also true.
He took the more flattering and more fragile of the two. Greene reads that as the effect of the law working. It is at least equally readable as the law's cost: concealment does not merely hide the mechanism, it selects the audience's replacement explanation, and you do not get to choose which one they pick. Houdini got occult powers. Talleyrand got second sight. Neither is what the man had, and neither can be defended.
Look at what Talleyrand is buying with the delegation, because it is not time.
He could have done research himself. He was capable of it. What the spies purchase is the separation of the labour from the person — a structure where the work happens in one place and the knowledge appears in another, with nothing linking them that an observer can see.
This is a genuinely different technique from the rest of Law 30's cases, and the chapter files them together.
Rikyū, Houdini and Yorinobu all did their own hidden work and then hid it. Their concealment is an act of will, maintained continuously, vulnerable to a single observer at the wrong moment. Talleyrand's concealment is architectural. He does not have to hide his labour, because there is no labour of his to hide. The work is genuinely elsewhere, done by people who are not him.
That is much more robust. Nobody can catch Talleyrand working late, because Talleyrand is not working late. It is also more expensive, and it requires an apparatus — which returns us to the standing finding of this build: the instruction is conditional on a position the book never teaches you to acquire. You cannot delegate the spying if you have no spies.
To those in government, including Napoleon himself, Talleyrand gave the impression of immense power—an effect entirely dependent on the apparent ease with which he accomplished his feats.2
Worth pausing on, because it changes the risk profile.
Napoleon was not an ordinary mark. He ran his own intelligence, he was not credulous about court reputations, and he had every reason to know how foreign ministries work. That he received the impression anyway is the strongest single piece of evidence in this chapter for the law's reach — the effect operated on someone who had the means to see through it and the motive to try.
And it is also where the story ends badly, in a direction Greene does not pursue here. The vault's other Talleyrand pages carry the sequel: the relationship with Napoleon ends in open rupture, and Talleyrand outlives the empire. The reputation for effortless omniscience is precisely what makes a subordinate intolerable to a ruler who cannot locate the source of it. An unfathomable minister is an asset until the master starts wondering what else he cannot see.
You have information other people do not have. Real information, obtained by ordinary means — you read the thing nobody reads, you talk to someone in the other department, you keep notes.
The instinct is to present it. A summary, a deck, a here's what I found and here's how I found it.
That gets you credited as diligent, which is a lower-value reputation than it sounds, and it also gets your method copied within a quarter.
The Talleyrand move is to let it out sideways, in the register where nobody audits: a short remark, apparently offhand, in a conversation about something else. I'd be surprised if that supplier holds past March. Then stop. Do not supply the reasoning. Do not append the evidence.
Now the part he got wrong, and it is the part that matters. Choose the explanation you want them to reach, because they will reach one, and if you leave it open they will pick the most flattering available and it will be the one you cannot defend.
So when the March thing happens and someone says how did you know — answer. Say the boring true thing: I've been watching their delivery times since October. You keep the credit, you lose the magic, and you have swapped an unfalsifiable reputation for a durable one.
Talleyrand kept the magic. It was worth a great deal, right up until it was the reason he had to be removed.
Strongest evidence. The mechanism is fully specified — delegated research, conversational delivery, retrospective confirmation — and the Napoleon detail supplies a hard test case. It is also internally consistent with the rest of the vault's Talleyrand corpus, which is unusual for this book's recurring figures.
Tension — the reputation acquired is fragile and Greene treats it as the prize. Clairvoyance cannot survive a miss; well-informed can. Documented above.
Tension — this case is architectural, not behavioural, and is filed with behavioural ones. Talleyrand has nothing to conceal because the labour is genuinely other people's. That is a different technique with different prerequisites.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no dated episode, no named prediction, no memoir citation. The claim "people thought he was clairvoyant" is offered with no instance attached: we are told the reputation existed and given no example of a remark that produced it. This is the chapter's weakest-sourced case and its most-cited figure.
Open questions. Is there a recorded Talleyrand remark that functioned this way, or is the clairvoyance reputation itself a nineteenth-century legend Greene has inherited? And: is the architectural version of Law 30 available below the level of a foreign ministry — does anyone without staff get to do this?
Talleyrand is the vault's most-used Greene figure after Bismarck, and this is his sixth page. Read across them, the corpus contains something no single chapter states.
Talleyrand as Conversational Spy shows him extracting in conversation. This page shows him emitting in conversation. Talleyrand and the Baiting of Napoleon and Talleyrand, Napoleon and the Market Pigs show the same channel used to provoke and to absorb. Talleyrand at the Window, 1830 shows him reading a regime change from a sound.
Six laws, one man, and the same instrument every time: the social register, used for intelligence in both directions. Greene distributes him across chapters as an illustration of six different principles. What the pages together show is not six techniques but one, applied repeatedly — and the book's chapter structure is what prevents that from being visible to a reader going in order.
Against the Houdini material in this same chapter, the distinction is clean: Houdini's concealment is personal and effortful, Talleyrand's is structural and cheap. The chapter recommends both under one instruction and only one of them is available to a reader without an apparatus.
Behavioural mechanics — the same channel, run in the other direction. Talleyrand as Conversational Spy documents him using social conversation as an intake — the setting where people say things they would never write down, because nothing said at dinner feels like disclosure.
This page is the same property exploited for output. Nothing said at dinner feels like a claim, either.
The insight neither page produces alone: the conversational register has a single underlying feature — it suspends the accounting. Statements made in it are not filed, not attributed, not audited, and not held against the speaker. That is why it works for extraction and for emission, and it explains why one operator would use it for both rather than needing two techniques. It also predicts the vulnerability: the whole apparatus depends on the setting being believed informal. A man known to be working the room is in a room that has started keeping records, and both directions close at once.
Psychology — why the audience supplies the wrong cause. The Question You Actually Answered describes the mind swapping an unanswerable question for an answerable one and returning the substitute with full confidence, the swap never surfacing.
Talleyrand's listeners face an unanswerable question — how does this man know things before they happen? — with no access to couriers, ministries, or payments. So they answer an available one: what kind of man is this? And the answer, from the evidence in front of them, is an uncanny one.
What the pairing produces is the correction to Greene's whole framing of concealment. He treats hiding the mechanism as creating a gap in the audience's understanding — a mystery, which is valuable. The substitution frame says there is no gap. The audience does not experience an absence and sit with it; they close it immediately, with the most cognitively available cause, and then hold that cause with the confidence of a perception. Which means concealment is not the withholding of an explanation. It is the involuntary commissioning of a replacement one, written by the audience, in your name, and you are then responsible for defending it. Houdini got the occult. Talleyrand got second sight. Neither chose, and neither could recant.
Sharpest implication. Concealing your mechanism does not leave a mystery — it commissions a replacement explanation you did not write and cannot recant. The audience closes the gap instantly with whatever cause is most available, and the flattering ones are systematically the fragile ones: clairvoyance cannot survive a miss, and well-informed can. The operative decision is therefore not whether to hide the work but which explanation to leave lying around, and Greene's law is silent on the only part that determines whether the reputation is defensible.
Generative questions.