The Union army had a discipline problem that wasn't really a discipline problem.
Unlike the well-trained Confederate troops, many Northern soldiers had been conscripted at the last minute. They were pioneers and frontiersmen — fiercely independent men who had spent their lives answering to nobody, now expected to stand in formation and obey.1
Some Union generals tried desperately to instill conventional discipline, and mostly failed. Others gave up on the men entirely and focused on map strategy, while their armies kept performing badly.1
William Tecumseh Sherman did something neither group considered.
Sherman's solution was to stop fighting his soldiers' character and redesign the organization around it.1
He built a more democratic army. He encouraged initiative in his officers rather than obedience. He let his men dress as they saw fit. He deliberately loosened outward discipline — the parade-ground kind — in order to build morale and group spirit instead.1
And he took the trait that made them hardest to command and converted it into the army's primary weapon. Like frontiersmen generally, his soldiers were restless and nomadic. So Sherman exploited that: he kept his army in perpetual motion, always marching faster than his enemies could manage.1
Of all the Union armies, Sherman's were the most feared and performed the best.1
Greene's instruction from the case is direct: do not struggle with your people's idiosyncrasies — turn them into a virtue, a way to increase your potential force.1
The chapter frames this as the final requirement of the Controlled-Chaos Strategy. You must structure your group according to your soldiers' actual strengths and weaknesses and their social circumstances, which means being genuinely attuned to the human side of the people you lead. You have to understand them, and the spirit of the times, inside and out.1
The closing instruction is about the leader rather than the organization: keep your mind as fluid and adaptable as the army you want to lead.
The move requires giving up a specific and satisfying belief — that there is a correct way to run an organization, and that the people who don't fit it are the problem.
Sherman's peers weren't lazy. They were doing the obviously right thing: instilling discipline, which is what armies do and what had worked for centuries. The reason it failed had nothing to do with effort. The men in front of them simply were not going to become the kind of soldiers the method assumed, and no amount of insistence was going to change that.
Sherman's advantage was accepting an unwelcome fact about his raw material and building from there rather than from the manual.
You've inherited a team that doesn't work the way you know teams should work. They resist process. They skip the standup. They go off and build things nobody asked for, and some of those things turn out to be the best work anyone does that quarter.
The default response is to install the discipline: enforce the process, hold the line on the ceremony, make them into the team the methodology describes. That's what most of Sherman's peers did, and it produced armies that performed badly while their generals were technically correct.
The Sherman question is different: what is this group actually good at, and what structure would make that the point rather than the exception?
If they're building unasked-for things that turn out valuable, the trait causing your process problems is initiative — and the answer might be less specification and more explicit space for it, with results as the judgment rather than adherence. If they resist coordination overhead but move fast in small groups, the answer might be fewer, larger blocks of autonomy rather than tighter check-ins.
The discipline required here isn't the team's, it's yours: you have to be willing to run something that doesn't look like what you'd design in the abstract, and to be judged on outcomes rather than on the tidiness of the arrangement.
Sherman's unconventional command style, his troops' reputation, and the mobility of his armies — particularly in the Atlanta campaign and the March to the Sea — are documented military history. [POPULAR SOURCE] for Greene's framing and causal emphasis.
Open tension: Greene attributes Sherman's success primarily to organizational adaptation, which is one real factor among several. Sherman's armies also benefited from Union material superiority, from operating later in the war against a degraded opponent, and from a strategic situation that rewarded exactly the kind of mobile, destructive campaign he preferred. The causal weight placed on "he adapted structure to personality" is heavier than a full military history would support.
A second tension the chapter creates and then buries: this case is the practical content of Strategy 6's Reversal, which concedes that "some people respond better to rigid authority" and that even a loose organization sometimes has to tighten.1 The Sherman case is genuinely about matching structure to people — which cuts both ways and implies that for a different group, the correct adaptation would have been more rigidity, not less.
Greene's chapter spends most of its length arguing that decentralization is structurally superior, then closes with a case whose actual lesson is that no structure is superior in the abstract. Those are different claims, and the second one substantially undercuts the first.
Set this beside The Prussian General Staff, the same chapter's other organizational model, and the contrast is sharp enough to be uncomfortable.
The Prussian approach was to build an institution that would shape the officers who passed through it — decades of common training, a shared doctrine installed deliberately, people made to fit the structure. Sherman's approach was the exact inverse: take people as they arrive and build the structure around what they already are.
Greene presents both admiringly in the same chapter without noting that they're opposite answers to the same question. What the pairing produces: the Prussian method scales and persists — it survived every individual officer and ran for over a century — but requires enormous time and institutional control. Sherman's method works immediately with whoever you actually have, and doesn't transfer at all; the moment Sherman leaves, the arrangement built around his specific men has no independent existence.
That's a real strategic choice with real trade-offs, and the chapter obscures it by treating both as illustrations of a single principle called fluidity.
Business → Hiring for Culture Fit vs. Building Around Talent. Organizations face Sherman's dilemma constantly: adjust people to the system, or adjust the system to the people you can actually attract.
The insight the pairing produces: the choice is usually framed as a values question and is really a timescale question. The Prussian route — shape people to the institution — has a payback period measured in years and produces something that outlasts everyone. The Sherman route pays off immediately and evaporates when the specific people leave. Startups reliably choose Sherman and then are surprised when their culture doesn't survive their first wave of departures; large institutions choose Prussia and are surprised when talented individualists refuse to join.
Psychology → Person-Environment Fit. Organizational psychology's fit literature finds that performance and retention depend less on individual traits or environmental features in isolation than on the match between them — the same person thrives in one structure and fails in another.
The insight: Sherman's peers were running an implicit theory that discipline is universally productive, and the fit literature says there is no such universal. What made Sherman unusual wasn't that he was more permissive; it was that he treated the structure as the adjustable variable, when every incentive in a military hierarchy says the soldier is the thing that adjusts.
Sherman's peers were doing the right thing correctly and losing, because the right thing was calibrated for soldiers they did not have. Being technically correct about how organizations should work is worth nothing if the people in yours were never going to be the people the theory assumes.