Thousands of people worked under George C. Marshall in the War Department and in the field, and most of them never met him. They felt him anyway — in the terseness of the reports that reached them, in how fast their questions got answered, in the way officers like Eisenhower handled them.
Few of them could have explained how he did it. That was rather the point.
This is Robert Greene's positive case for Strategy 5, the Command-and-Control Strategy: what it looks like when someone treats the chain of command itself as the thing to be built, rather than as a given to be shouted through.1
In 1939, Roosevelt had to choose an army chief of staff, and it was a critical appointment — war had begun in Europe and he believed American involvement was inevitable. The army was small, badly trained, and running doctrine unsuited to modern technology. Marshall, then a brigadier general, had been arguing loudly for exactly the reforms Roosevelt thought were needed, so Roosevelt bypassed more senior men and gave him the job.1
It was a curse disguised as a promotion. The War Department was hopelessly dysfunctional: generals with monstrous egos and the power to enforce their preferences, senior officers who took department jobs instead of retiring and built protected fiefdoms, feuds, waste, overlapping responsibilities, communication breakdowns.1
Marshall's problem was not strategy. It was that he could not control the institution he was supposed to use.
Greene's reading pivots on a restraint most people wouldn't manage. The temptation was to fight — to take on the recalcitrant generals, the political feuds, the layers of waste, problem by problem.
Marshall was too smart for it, for three reasons.1 There were too many battles and they would exhaust him. Micromanaging the department would embroil him in petty entanglements and cost him the larger picture. And doing both would make him look like a bully.
The only way to kill a many-headed monster, he concluded, was to step back and rule indirectly — with a touch light enough that nobody quite noticed how completely he dominated.
The mechanism was in place before he needed it, which is the part most retellings skip.
Ten years earlier, as assistant commander of the Infantry School at Fort Benning, Marshall had trained a great many officers — and kept a notebook recording the names of the promising ones.1
Once he became chief of staff, he began retiring the older officers in the War Department and replacing them with the young men from that notebook. They were ambitious, they already shared his reform instincts, and he actively encouraged them to speak their minds and show initiative. Omar Bradley and Mark Clark came out of that group. So did Dwight Eisenhower.
The relationship began days after Pearl Harbor, when Marshall asked Eisenhower — then a colonel — to write a report on what should be done in the Far East. The report showed Marshall that this man already thought the way he did about running a war.1
For months afterward Marshall kept Eisenhower in the War Plans Division and watched him closely. The two met daily, and in that time Eisenhower absorbed Marshall's leadership style and his way of getting things done. Marshall also tested him — indicating he'd be kept in Washington rather than given the field command he desperately wanted. Eisenhower passed. Like Marshall, he got along with other officers while remaining quietly forceful.1
In July 1942, Marshall surprised everyone by naming the still relatively unknown Eisenhower commander of the European Theater. When the Americans fared badly in North Africa and the British demanded a replacement, Marshall stood by him — and offered one crucial piece of advice: develop your own protégé, a roving deputy who thinks the way you do and acts as your go-between with subordinates.
Marshall suggested Bradley. Eisenhower took the suggestion, essentially duplicating the staff structure Marshall had built in Washington. With Bradley in place, Marshall left Eisenhower alone.1
Marshall placed protégés throughout the department, where they quietly spread his methods. To make that easier, he attacked the waste with what Greene calls utter ruthlessness — reducing the number of deputies reporting to him from sixty to six.1
He hated excess. His reports to Roosevelt were famous for compressing complex situations into a few pages, and the six men who reported to him learned fast that any report a page too long simply went unread.
His method of correction was almost entirely non-verbal, and brutally efficient.
He would listen to oral presentations with complete attention — but the moment a deputy wandered off topic or said something not thought through, Marshall would look away, bored and uninterested. It was an expression they came to dread. Without a word he had made it clear they'd displeased him and it was time to leave.1
The six began to think like him, and to demand from their own subordinates the same efficiency and compression he demanded of them. Greene's estimate of the result: the speed of information flow up and down the line quadrupled.
The most-quoted moment in the account is a small one. Brigadier General Leslie Groves — military director of the atom bomb project — came to get Marshall's signature on a $100 million expenditure.
Finding Marshall absorbed in paperwork, Groves waited while the chief of staff carefully compared documents and made notes. Marshall finally put down his pen, examined the $100 million request, signed it, and handed it back without a word.
As Groves turned to leave, Marshall spoke: "It may interest you to know what I was doing: I was writing the check for $3.52 for grass seed for my lawn."1
Greene's reading of this is exact. An officer might go home chuckling about the boss fussing over a gardening bill — and then slowly realize that if he wasted a penny, Marshall would know.
You've inherited a team or department that doesn't work, staffed with people you didn't hire, several of whom actively prefer the current arrangement. The instinct is to start fixing problems: address the worst performer, untangle the worst process, win the most visible fight first.
Marshall's move is to resist that entirely and ask a different question — not what's broken, but who, in this structure, will still be here in a year, and do they think the way I need them to think?
The specific sequence matters. He replaced people before he tried to change behavior, because behavior change in an unwilling person costs enormous energy and reverts the moment you look away. He cut the volume of things demanding his attention — sixty deputies to six — before trying to raise the quality of any of them. And he communicated standards through consequence rather than instruction: no memo about report length, just reports past a page going unread until people adjusted.
The uncomfortable part is the timeline. Marshall's notebook was ten years old when he needed it. The version of this you can start today isn't the replacement — it's beginning to notice, now, which of the people around you would be in the notebook.
Marshall's reorganization of the War Department and his role in developing the American high command are well documented, and his reputation as the architect behind Eisenhower is standard military history rather than Greene's invention. [POPULAR SOURCE] for the framing and the anecdotes.
Open tension: Greene presents "rule indirectly through cloned protégés" as straightforwardly superior leadership, and within its own frame it worked spectacularly. What the chapter doesn't examine is the obvious failure mode — a leader who fills every key position with people trained to think exactly as he does has also built an organization structurally incapable of telling him he's wrong.
Marshall got away with it because his judgment was, by most historical accounts, unusually good. The method itself carries no protection against a leader whose judgment isn't. Greene's own Groupthink warning earlier in the same chapter is about groups converging on bad decisions through social pressure — and an organization of deliberate clones is that failure mode installed as architecture, not avoided. The chapter never notices the collision.
Set this beside Gallipoli and the Broken Chain, the same chapter's failure case. Hamilton and Marshall had structurally identical problems — a large organization, subordinates chosen by a seniority system rather than by them, and a need to move faster than the enemy.
The difference is what each man treated as fixed. Hamilton took his chain as a given and tried to send good orders through it. Marshall treated the chain as the primary object of work and rebuilt it before sending any orders that mattered.
Read together, they suggest the chapter's real claim isn't about communication skill at all. Hamilton's orders failed not because he phrased them badly in isolation, but because he was speaking through people whose defining traits he had never adapted to and could not change. Marshall solved the same problem upstream, by changing who was in the chain — which is slower, less heroic, and the only one of the two approaches that worked.
Business → Founder Mode vs. Manager Mode. The recurring argument about whether leaders should stay deep in the details or delegate broadly usually gets framed as a personality question. Marshall's case reframes it as a sequencing question: he was ruthlessly hands-on about who was in the room and about cutting reporting volume, then deliberately hands-off about how those people did their jobs.
The insight the pairing produces: "delegate more" and "stay in the details" are both incomplete instructions, because they don't specify which details. Marshall micromanaged personnel selection and information architecture — the two things that determine everything downstream — and then genuinely let go of the work itself. The $3.52 grass-seed line is not a story about a man who couldn't delegate; it's a man deliberately advertising which category of thing he still watched.
Psychology → Modeling and Vicarious Learning. Social learning research holds that people acquire complex behavioral repertoires far more effectively by observing a model over time than by receiving explicit instruction — which is precisely the mechanism of the daily Marshall-Eisenhower meetings, and of Eisenhower then reproducing the same structure with Bradley.
The insight: Marshall wasn't transmitting rules, which degrade in transmission the way Hamilton's order did. He was transmitting a style through sustained proximity, and style survives the handoff intact in a way instructions don't. That's why the structure replicated one level further down without Marshall having to design that layer himself.
Marshall's real leverage was created a decade before he had any authority to use it, in a notebook nobody asked him to keep. By the time he needed a hundred loyal, capable officers who already thought the way he did, the only question left was which names to call — and the people he outmaneuvered in 1939 had no way to compete with a decision he'd made in 1929.