A group of Columbia Pictures executives comes to their president's office with bad news. One of their employees, the screenwriter John Howard Lawson, has been named as a communist. He has to go immediately or the studio takes the committee's wrath.1
The president is Harry Cohn. Greene establishes exactly who is being asked:
Harry Cohn was no bleeding-heart liberal; in fact, he had always been a die-hard Republican. His favorite politician was Benito Mussolini, whom he had once visited, and whose framed photo hung on his wall. If there was someone he hated Cohn would call him a "Communist bastard."2
He visited Mussolini. He kept the photograph on the wall. Communist bastard was his standing insult.
He refused to fire Lawson.
Greene closes off both comfortable explanations in one sentence.
Not principle — the man's politics are established above, at length, precisely so this reading is unavailable.
And not merit: "He did not keep the screenwriter on because he was a good writer—there were many good writers in Hollywood."1
So Lawson was, by every ordinary measure, maximally disposable. Politically radioactive at the worst possible moment. Personally loathsome to his employer. And professionally substitutable in a town full of writers.
He kept his job.
He kept him because of a chain of dependence: Lawson was Humphrey Bogart's writer and Bogart was Columbia's star. If Cohn messed with Lawson he would ruin an immensely profitable relationship. That was worth more than the terrible publicity brought to him by his defiance of the committee.1
Three links. Cohn needs Bogart. Bogart needs Lawson. Therefore Cohn cannot touch Lawson.
Nothing in that chain runs on anyone's opinion of anyone. Cohn's contempt for Lawson is real and entirely irrelevant, because the pressure reaches Lawson through a party who has no interest in the politics at all.
Notice the arithmetic Greene attributes to Cohn: the cost of defying a congressional committee, weighed against the cost of losing a star, and the star wins. This is not a man taking a stand. It is a man pricing two disasters and choosing the cheaper one.
This is the mechanism worth extracting, because it inverts what the chapter around it recommends.
Law 11's instruction is to make the person with power depend on you. Every other case in the chapter follows it: Bismarck makes kings need him, Michelangelo makes a pope need him, the astrologer makes Louis XI need him.
Lawson does not appear in Harry Cohn's calculation at all except as a cost. He supplies Cohn nothing that another writer could not supply.
What he has is a link to someone who is themselves unremovable — and the immunity conducts down the chain without the protected party doing anything.
That is a genuinely different architecture, and it has one enormous practical advantage: it requires no exceptional capability. Michelangelo's route needs you to be Michelangelo. Kissinger's needs institutional seniority. Lawson's needs only that one irreplaceable person actually depend on you, specifically, in a way that is visible and hard to sever.
Greene states the principle a few pages earlier in a single clause — "you do not necessarily have to entwine yourself around the master; another person will do, as long as he or she too is indispensable in the chain"3 — and then illustrates it with the best case in the chapter without marking it as the democratic version.
The phrase does a lot of work and deserves unpacking.
Lawson is not simply a writer whom Bogart likes. He is Bogart's writer — an established, publicly known working relationship, of a kind that is expensive to reconstruct.
That specificity is what makes the link hold. A vague good relationship with a star transmits no protection, because it can be replaced without friction. A working dependency — this person writes the films this star makes — cannot be swapped without the star bearing a real cost, and the star's discomfort is the currency.
So the protection is proportional to how much the irreplaceable person would lose by your removal, not to how much they like you.
Which gives the honest version of the strategy: the goal is not proximity to power, and not being liked by someone important. It is becoming a genuine input to work that a structurally protected person needs done.
Greene does not say this, and it is the obvious limit.
Lawson's security rests entirely on Bogart's continued irreplaceability at Columbia. It is borrowed, and the lender can call it in without doing anything deliberate — by leaving, by declining, by a change in what audiences want.
The protected party also has no ability to maintain the anchor. Michelangelo could keep his hands working. Bismarck could keep handling enemies. Lawson could do nothing whatsoever to keep Bogart a star.
Borrowed immunity is the cheapest to acquire and the only kind whose collapse you cannot see coming or prevent, which is a real trade rather than a free lunch, and the chapter presents it as a straightforward win.
You are replaceable, and you know it.
First, stop trying to be valuable to the most senior person available. That is a crowded market and you will lose it to someone more talented. Look instead for whoever in your organisation genuinely cannot be removed — and rank is a poor guide to that. Bogart was not Cohn's superior.
Second, become a real input to their work, not a pleasant presence in their life. Bogart's writer is a job description. Liking transmits nothing; a working dependency transmits everything.
Third, make the link legible. The protection operates through other people's calculations, so it only works if those people can see the connection. A dependency nobody has noticed provides no cover.
Fourth, be honest that you have borrowed something. You cannot maintain your own anchor. Check periodically whether the person you are attached to is still structurally protected, because your position moves when theirs does and you will get no warning.
Fifth, use the time to build something of your own. The chain buys shelter, not a career. It is the right move when you have no leverage, and a bad place to still be in ten years.
And the plain record: Lawson was protected from a congressional committee by a studio head who despised him, because of an accounting decision about a movie star. The protection was real, it was total, and it had nothing to do with him.
Strongest support. The case is unusually well-controlled: Greene closes off principle (the Mussolini photograph) and merit ("many good writers") explicitly, which isolates the chain as the only operating variable. That is more rigour than most of this book's anecdotes carry.
Tension — it is the chapter's most useful case and gets the least space. Ten lines, no interpretation section, no Image, no place in the law's stated architecture. The mechanism available to ordinary readers is the one Greene passes over fastest.
Tension — the borrowed immunity has no maintenance path and Greene does not flag it. Every other case in the chapter gives the protected party something they can keep doing. This one does not.
Tension — the outcome is asserted without a source. Whether Lawson kept his job at Columbia in these circumstances, and whether this reasoning is why, is stated flatly with nothing behind it.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no citation, no date beyond 1951, no memoir or history named. Cohn's politics and Lawson's blacklisting are historically real; the office scene, Cohn's reasoning, and the causal role of Bogart are given entirely on Greene's word. Note also that Lawson was one of the Hollywood Ten, was convicted of contempt of Congress in 1948 and imprisoned in 1950 — context Greene omits, and which sits awkwardly with an account of him quietly retained at a studio in 1951. [POPULAR SOURCE] · 🚩 The chronology deserves checking before this page is promoted.
Open questions.
Set this against The Condottieri and the Replaceable Servant and the pair produces the chapter's most practical finding, which neither case states alone.
Carmagnola's problem, in Greene's own diagnosis: "there were so many other condottieri as able and valiant as they were. They were replaceable. Nothing was lost by killing them."
Lawson's situation is identical. Many good writers in Hollywood. Nothing lost by firing him — and considerable political benefit in doing so.
Same market conditions, opposite outcomes, and the difference is not talent or strategy. Carmagnola tried to solve replaceability on his own axis, competing against every other condottiere in Italy — a race with a deep field and a rising price, which he could not win.
Lawson did not compete at all. His protection was imported from a different market entirely, one where the field was empty.
So the pair yields a rule the chapter never writes: when you are in a deep market, do not try to win it — attach to someone in a shallow one. Carmagnola's error was not insufficient effort or insufficient entwining; it was accepting the terms of a contest he was structurally going to lose. Every condottiere who tried harder had the same problem.
→ behavioral-mechanics · Intensive Versus Extensive Dependence
Greene's two named architectures: Michelangelo's intensive power (one irreplaceable skill, portable) and Kissinger's extensive power (entangled across many departments, located).
Lawson has neither, and the three together map the field properly.
Intensive power is owned. Extensive power is occupied. Borrowed power is conducted.
The insight the pair produces is about entry cost against maintenance cost, and the three sit at opposite ends. Intensive power is expensive to acquire — years of capability — and cheap to maintain, because you carry it. Borrowed power is nearly free to acquire and impossible to maintain, because the thing sustaining it is another person's market position, which you cannot influence at all.
Greene ranks intensive above extensive on portability and never places the borrowed form on the scale. Placed properly, it is the correct move for exactly one situation: you have no leverage and need shelter now. It buys time. It does not accumulate, and a decade of it leaves you exactly where you started, still holding somebody else's immunity.
→ history · Bismarck and the Weak King
Bismarck survived two kings, one of whom "disliked Bismarck intensely and had no intention of keeping him around."
Cohn "would call him a Communist bastard." Same structure: protection running through active personal hostility.
The pair establishes something the chapter asserts but never demonstrates cleanly — that dependency and affection are fully independent variables, and that a book of this kind can therefore give advice that ignores whether anyone likes you.
But the two cases differ on where the hostile party's need is located, and it decides the durability. William's need was for Bismarck himself — nobody else was handling the enemies. Cohn's need was for Bogart, and Lawson merely sat in the path of it.
The insight neither yields alone: hostility is survivable in both, but only the direct form gives you a hand on the variable. Bismarck could deepen William's need by continuing to be the man who did the job. Lawson could do nothing to deepen Cohn's need for Bogart. Direct dependency converts an enemy into a permanent client; borrowed dependency merely makes you briefly unprofitable to attack — and the moment the arithmetic shifts, the hostility that was always there executes immediately.
Sharpest implication. Lawson was protected by a man who despised him, at the worst possible political moment, without supplying that man anything of value — because the immunity conducted through a third party who had no stake in the argument. That is the only mechanism in Law 11 that works for someone with no exceptional talent and no institutional position, and Greene gives it one clause and a ten-line anecdote. The chapter is written for people who can become Michelangelo; the case buried in it is the one that works if you cannot.
Generative questions.