Paris, 1651. There is a rebellion running against the French monarchy, and one of the men running it is Cardinal de Retz.
The young king, Louis XIV, and his whole court are in the city. This is a problem. Greene calls it "a tremendous burden on the revolutionaries" — the crown sitting physically inside the revolution, watching, close.1
Then the court leaves. Packs up and establishes itself in a palace outside the capital.
They breathed a sigh of relief. This later proved their downfall, however, since the court's absence from Paris gave it much more room to maneuver.1
The thing they were most frightened of was the thing keeping them safe. The relief was the defeat, arriving early and disguised as good news.
De Retz, who "prided himself on his insights into human schemes and why they mostly fail,"1 wrote it up afterward:
The most ordinary cause of people's mistakes is their being too much frightened at the present danger, and not enough so at that which is remote.2
Nineteen words, and note what they do not say. This is not "people are shortsighted." It is a claim about a ratio. The fear is not absent from the distance; it is misallocated across it. There is a fixed budget of alarm and proximity spends nearly all of it.
That framing matters because it predicts a specific error the shortsightedness framing does not: the danger that recedes will feel like a danger that shrank. Not one you forgot — one you survived. The revolutionaries did not fail to think about the king. They watched the king get further away and read distance as diminishment.
Here is the whole mechanism, unglamorously.
You have one instrument for measuring threat, and it works by how much fear it produces. Fear is generated by things you can see, hear, and act on now. A danger three months out generates almost none, so the instrument reports a small number.
The instrument is not broken. It is reporting fear accurately. The error is treating its output as a measure of size. It measures nearness and you read it as importance, and there is no internal signal to tell you which quantity you are looking at.
Which is why the correction cannot be worry more about the future. You cannot manufacture fear about a thing that is not producing any. The correction has to be a procedure that bypasses the instrument entirely — writing the danger down, assigning it a number, asking somebody who is not standing where you are standing.
One thing about the attribution is worth holding, because Greene passes over it and it changes how the sentence reads.
De Retz is not an observer of the Fronde. He "spearheaded" it.1 The rebellion he is explaining the failure of is his own, and the sigh of relief he describes was his.
So "the most ordinary cause of people's mistakes" is a man writing my in the third person. He generalises his defeat into a law about people, which is what memoirists do, and Greene quotes the general form.
This is worth more than a biographical footnote for one reason: it is the only evidence anywhere in the chapter that the error is detectable afterward. De Retz saw it. He could name the exact moment, describe the feeling, and identify the mistake — years too late, and only from outside the position. Nobody in this law is shown catching it in time.
Which sharpens the practical claim considerably. The remote-danger error appears to be reliably diagnosable in retrospect and almost never in the moment, and a manual that quotes the retrospective diagnosis as though it were a prospective instruction has skipped the hard part.
Watch the two ledgers move in opposite directions.
The rebels' ledger. The king was near. Nearness generated pressure. The king left. Pressure ended. Entry recorded: situation improved.
The crown's ledger. In Paris, the court was inside a hostile city, observable, constrained in every move by an audience that included its enemies. Outside Paris it had, in Greene's phrase, "much more room to maneuver."1 Entry recorded: capability restored.
Same event. Both ledgers correct. And the rebels' entry is the one that reads as good news, because their instrument was calibrated to the thing that had gone away.
The generalisable form: when an opponent removes a source of pressure on you, that is not information about your position. It is information about theirs. A withdrawal is a move. It costs something and is therefore being paid for by something, and the payment is almost never your comfort. The relief you feel is a fact about your nervous system, not about the board.
The difficult client goes quiet.
For six weeks they have been on the phone twice a week, and now nothing, and the first thing you notice is how much easier the mornings are. You get work done. Somebody asks how it's going with them and you say fine, actually, it's calmed right down.
Stop there. That sentence is the sigh of relief and you should treat it as an alarm.
Ask one question: what did they gain by going quiet? Not what did you gain — you already know that, it's the mornings. What did they get? They got out of a channel where you could answer them. They got room to talk to people you cannot hear. If they are building a case, they are building it now, unobserved, and the quiet is the condition it requires.
Write it on paper, because you will not be able to feel it. Under the difficult-client heading, one line: silence began 4 March, reason unknown, exposure if adverse: the renewal. You are not trying to worry. You are trying to keep a number somewhere your fear cannot round it down.
Then check the line in three weeks, when the mornings are still easy.
De Retz supplies a marginal to Law 28 as well, forty-eight lines before this one. There it is fear again, working the other direction:
Fear, which always magnifies objects… so that fearful persons seldom fail to fall into real inconveniences, occasioned by imaginary dangers… being in that like unto hares.3
Two adjacent laws, one source, two opposite errors. At Law 28, fear inflates a near danger and you hesitate over nothing. At Law 29, fear deflates a distant one and you walk into something real.
Both are the same finding stated once: fear scales with proximity, not with magnitude. De Retz has the general law and Greene splits it across two chapters as two separate warnings, connecting them nowhere.
⚠ And there is a lift. Greene's Image for Law 28 is "The Lion and the Hare" — printed unattributed, forty lines after de Retz's "like unto hares."3 The Image comes from the marginal.
The sentence names an error and gives no correction, and the gap is worth stating plainly rather than filling in on Greene's behalf.
Be more frightened of remote things is not executable — that is the instrument problem above. And the maxim contains no scale. How remote is remote enough to warrant attention? Every decision has an infinite tail of distant consequences and no rule here for where to stop scanning. Applied without a stopping rule, the maxim produces paralysis, which is the failure mode the previous law was written against.
Greene does not notice the gap because he reads de Retz as illustration rather than as instruction. But a reader who takes it as instruction gets an imperative with no bound on it, sitting immediately after a chapter that says hesitation is fatal.
Strongest evidence. The 1651 case is a good one because the rebels' error is not a failure of intelligence — they knew where the king was — and because the mistake announces itself as its own opposite. A mechanism that makes the error feel like relief is much better supported by a case where relief is what the participants report.
Tension — the maxim and the previous law point opposite ways. Law 28 says hesitation is fatal and boldness closes gaps. This maxim says the danger you are not frightened of is the one that gets you. Neither chapter cites the other, and de Retz is quoted in both.
Tension — no bound on the scan. Documented above.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no edition, no page, no date is given for the de Retz quotation, and the Fronde narrative around it is uncited.
Open questions. Is the ratio claim right — a fixed budget of alarm redistributed by distance — or is distant fear simply not produced at all, in which case the two framings imply different interventions? And does the effect run on temporal distance, physical distance, or social distance, given that the 1651 case confounds all three?
The vault's Greene corpus contains the counter-case, and it is the strongest one available: Marengo and the Plan with Branches — Napoleon on the floor of his office for several days, surrounded by maps and note cards, having "calculated the Austrians' reactions to the feints he was planning."
That is a man overriding the instrument by brute procedure. Not by feeling the remote danger more, but by writing every branch of it on a card and laying the cards out on the floor where they could be seen. The correction de Retz's maxim does not supply, Napoleon's method supplies — and it is entirely external. Paper, floor space, days.
And then Napoleon in Russia and the Lure of the Void shows the same man losing ten thousand horses in a week to a danger nobody had to hide from him. Same operator, best-in-class procedure, and the remote danger still arrived — which suggests the procedure buys resolution rather than immunity, and Greene's corpus never says so because it sorts these two by outcome and files them under different laws.
Behavioural economics — the same curve, measured. Present Bias and Hyperbolic Discounting gives the experimental version. Offered $100 today or $110 next week, most people take the hundred. Offered $100 in fifty-one weeks or $110 in fifty-two, the same people wait. Identical premium, identical delay, opposite choice — the only variable is whether one end is now.
De Retz describes this for threats; the discounting literature measures it for rewards. Putting them together produces something neither states: the distortion is symmetrical, and that is bad news for the obvious defence. If only threats were discounted you could correct with a standing rule to inflate distant risks. But distant rewards shrink by the same curve, so a person applying that correction would systematically overweight remote dangers against remote gains and end up unable to act at all — which is the paralysis the maxim has no bound against. The correct correction is not to inflate one side. It is to move both sides into a register where the curve does not operate: numbers on paper, compared to each other, away from the moment.
Sales technique — the operator's version, run on someone else. Consequence Questions — NEPQ Stage 5 is de Retz's maxim weaponised. The technique asks the prospect what happens if nothing changes, and its whole design principle is that the prospect articulates the consequence themselves — the operator never claims urgency, because a claimed urgency is discounted and a self-generated one is not.
That is a precise fit with the mechanism here, and it exposes something the strategic frame cannot see on its own. De Retz treats the discounting of remote danger as a private cognitive defect. The sales frame treats it as a durable, exploitable property of every buyer — which means the remote danger you eventually come to feel may well have been installed by someone whose interests are served by your feeling it. The maxim tells you to fear the distance more. It gives you no way to distinguish a distant danger you surveyed from a distant danger somebody walked you into articulating, and those two produce identical feelings and opposite correct responses.
Sharpest implication. When an opponent stops pressing you, the relief is data about your nervous system and nothing else. A withdrawal is a purchase — somebody paid something for the room it bought them, and the price is almost never your comfort. The rebels of 1651 did not miss the king leaving. They watched it happen and filed it under good news.
Generative questions.