Early seventeenth-century France. An ambitious cleric wants to rise.
The obvious target is Louis XIII, who is the king.
Richelieu attached himself to the king's mother instead — and "catapulted through the ranks of the courtiers, all the way to the top."1
Greene's statement of what he had noticed: "it was not King Louis XIII who decided things, it was the king's mother."1
The general claim is short and it is the useful part of this law:
"Power itself always exists in concentrated forms. In any organization it is inevitable for a small group to hold the strings. And often it is not those with the titles."1
Two separate assertions there, and they are worth pulling apart.
Decision-making concentrates. However large the organisation and however elaborate its consultation, the number of people who actually settle things is small. This is a structural claim and it is broadly right.
And the concentration is not visible from the org chart. Titles record formal authority, which is a different thing from the capacity to decide — and the gap between them is not an anomaly to be corrected but the normal condition.
So the targeting problem is real: the loudest signal about who matters is systematically unreliable, and it is the only signal most people ever consult.
Greene's framing: "In the game of power, only the fool flails about without fixing his target. You must find out who controls the operations, who is the real director behind the scenes."1
The connection is efficiency of aim. The chapter's arrow image — "You cannot hit two targets with one arrow" — assumes the target has been identified. This passage is the only place in the law that addresses identification at all.
And it makes concentration much more dangerous than the law admits. Spreading your effort across twelve people is inefficient and forgiving — one of them is probably right. Concentrating everything on one person is efficient and unforgiving: if you have picked the figurehead, you have spent your whole campaign on someone who cannot help you.
The higher the concentration, the higher the cost of a targeting error — and the chapter recommends maximum concentration without ever raising this.
Worth pausing on, because it explains why the org chart failed here specifically.
Marie de' Medici's power was not delegated and not formal. She held it through relation, proximity, and the king's disposition toward her — which means it appeared in no document, could not be removed by any procedure, and would have been invisible to anyone studying the structure of the French court.
That is the general shape of undocumented authority: it runs on access and disposition rather than on grant. Which gives you the place to look. Not who has been given power, but who has continuous, informal access to the person who has, and whose view that person cannot easily hold out against.
Greene never generalises the mechanism, but the case contains it.
There is an accident in this corpus worth recording.
Louis XIII appears twice in two consecutive chapters. At Law 21 he is "the prince of France who later became Louis XIII," who "played the fool" while those above him suspected his ambitions.2 Here he is the king who does not decide.
Both appearances describe a man who is not exercising the power his position implies — in one case strategically, in the other passively — and Greene does not connect them.
Which raises a question neither chapter asks. A figurehead and a man deliberately performing harmlessness look identical from outside. Richelieu read the court correctly by the outcome; the Law 21 reading would have him misreading a deliberate performance.
So the diagnostic problem is worse than the chapter allows. "Who actually decides" assumes the answer is stable and discoverable. Someone playing the fool is generating exactly the evidence that would make you route around them — and the whole point of that tactic, per Law 21, is that the routing is a mistake.
You need a decision from an organisation, and you have identified the person whose title says they make it.
Before you spend anything, run the observations rather than asking anyone. Asking who decides produces the org chart, because that is the honest answer to the question people think you asked.
Watch what gets reversed. The clearest signal is a decision that was made and then quietly changed. Whoever changed it decides; whoever made it does not.
Watch who is consulted before rather than informed after. The sequence of a conversation carries more information than the seniority of its participants.
Watch for the person whose objection ends things. In most rooms one person's discomfort stops a proposal and everyone else's slows it down. That asymmetry is usually invisible in the moment and obvious in the pattern.
And watch access rather than authority. Richelieu's insight was about proximity — who has unstructured time with the decider, whose view arrives before the meeting rather than in it.
Then, before concentrating, price the error. If you are wrong about the target, concentration means you lose everything you spent. So the correct sequence is diffuse observation first, concentrated effort second — which is the opposite order to the one the chapter implies.
The strongest material is the structural claim: decision-making concentrates in small groups, and those groups do not correspond to titles. That is checkable and broadly supported by how organisations actually behave.
Tension: the law recommends maximum concentration and this passage shows targeting is unreliable. The two together imply a high-variance strategy, and the chapter never prices the error.
Tension: the case is one sentence supporting a general theory of organisations. Richelieu and the queen mother carry the entire weight, with no second instance.
Tension: the corpus contains its own confounder. Law 21 recommends appearing not to hold power; Law 23 recommends detecting who really holds it. A reader following both would be systematically misled by other readers following the first.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no date, no source, no detail on how the attachment was made or what Richelieu supplied; the queen mother is not named. [POPULAR SOURCE]
Open questions. How do you distinguish a genuine figurehead from someone performing harmlessness — the chapter offers no test, and Law 21 guarantees the second category exists. And: does undocumented authority survive being identified? Marie de' Medici's power ran on informal access; a person whose influence becomes widely recognised is usually formalised or removed.
Against the single patron, this passage supplies the missing qualifier. That page's law rests on attaching yourself to a "single, appropriate source of power" and never defines appropriate. This is the definition, arriving separately: appropriate means the one who actually decides. The two passages are three paragraphs apart and Greene does not link them, so the law's central undefined term is defined elsewhere in its own chapter.
Set beside Law 21's advice to play the fool the corpus produces a genuine conflict. One law tells you to look powerless while holding ambition; the next tells you to identify who is really in charge. Both are addressed to the same reader, and the first manufactures exactly the false signal the second must penetrate — with no acknowledgement in either chapter.
And against the arrow image, this passage is the chapter's only treatment of aim, and it undercuts the image's confidence. The arrow assumes the target is known. Here the target is the hard part, hidden by design, and possibly performing.
History — The Fixer-Broker Network
That page maps the shadow architecture around Rasputin between 1914 and 1916 — bankers and socialites operating through his proximity to the Empress to move war contracts, ministerial appointments and markets, with the charismatic figure at the centre and the operators around him.
It is Richelieu's insight documented at full resolution, and it corrects the chapter in a specific way.
What the pair produces: the real decision architecture is usually a network rather than a hidden individual, and Greene's framing sends you looking for the wrong shape. "Who is the real director behind the scenes" implies a single substitute target — swap the king for his mother and you are done.
The Rasputin network shows something messier: a proximity node, a set of professional intermediaries monetising access to it, and an outcome produced by the arrangement rather than decided by any member of it. Nobody in that structure was the decider, and attaching yourself to any one of them would have bought a share of an unstable mechanism.
Which sharpens the diagnostic considerably. The question is not who decides but what path a decision travels, and how many hands it passes through that can alter it. A single hidden decider is the easy case and the rare one — and it is the only one the chapter describes.
Business — BANT Qualification — Budget, Authority, Need, Timing
That page carries IBM's framework for whether a prospect can actually buy, with Authority as one of the four gates: does the person you are talking to have the power to say yes.
Sales solved Greene's problem formally, decades ago, because it had to — the cost of the error is measurable there.
What the pair produces: the discipline is not detection but refusal to proceed without detection, and that is what Greene's version lacks. BANT's contribution is procedural — you do not advance the opportunity until authority is established, because everything spent before that point is spent on someone who cannot complete the transaction.
Greene tells you to find the real decider and then says nothing about what to do while you are unsure. The chapter's whole pressure is toward committing your forces.
So the transferable move is a gate rather than an insight: name the moment before which you will not concentrate. And BANT's placement of authority alongside timing and budget adds the part Richelieu's story omits — the queen mother had to be able to act, not merely willing. A correctly identified decider without the resources or the moment to decide is the same dead end as a figurehead, arrived at by a more sophisticated route.
Sharpest implication. The higher your concentration, the more expensive a targeting error becomes — and the chapter that most strongly recommends concentration is the one that reveals targets are hidden by design.
Which inverts the correct order of operations: diffuse observation first, concentrated effort second. Greene's law implies the opposite sequence and never states either.
Generative questions.