March 1916. Pancho Villa — a folk hero turned bandit again after losing the revolution to Carranza — raids Columbus, New Mexico, and kills seventeen American soldiers and civilians.1
Woodrow Wilson's advisers make the case for a response, and the reasoning is worth keeping because it is the reasoning that always gets made:
For a power as large as the United States, they argued, not to strike back at an army that had invaded its territory would send the worst kind of signal. Furthermore… many Americans saw Wilson as a pacifist, a principle the public doubted as a response to violence; he needed to prove his mettle and manliness by ordering the use of force.2
He sends ten thousand men under General Pershing. By that summer the expedition has swelled to 123,000.3
They never catch him. In January 1917 Wilson orders the withdrawal, and rebel forces pursue the retreating army so persistently that "the U.S. Army [used] airplanes to protect its rear flanks."4
The single most instructive detail is easy to skim past.
The Americans offered a $50,000 reward for information leading to Villa's capture. But the Mexican people, who had been disillusioned with Villa when he had returned to banditry, now idolized him for facing this mighty American army. They began to give Pershing false leads.5
Villa's own people had already turned on him. The banditry had cost him the popularity the revolution gave him — which means the problem was solving itself, and Greene's Keys name this as one of the available options: wait for the Mexicans themselves to do away with Villa.
The expedition reversed a decline that was already underway. Not by failing — by existing. The moment the contest was Villa against the United States rather than Villa against his own reputation, the question changed and so did the answer.
⚠ And the false leads are the mechanism made concrete. The reward converted an entire population into an active counter-intelligence asset, and it did so by offering money to people who had just been given a reason to prefer him.
the imbalance between the size of the pursuer and the size of the pursued—who still managed to stay free—made the affair a joke.6
This is the finding, and it generalises well beyond the case.
A large force pursuing a small one is a claim about the small one's importance, made continuously and publicly, and the claim gets louder the longer the pursuit lasts. Every reinforcement is a new statement. The escalation from 10,000 to 123,000 is a twelvefold public upgrade of Pancho Villa's significance, paid for by the people upgrading him.
And the outcome is asymmetric in a way that cannot be fixed by competence: the small party wins by continuing to exist. Villa does not have to defeat Pershing. He has to remain uncaught, which requires only geography and a friendly population — and the pursuit supplied the second.
⚠ Note that Pershing was not incompetent. He had "defeated guerrillas in the Philippines and Native Americans in the American Southwest," the troops had radios and aerial reconnaissance, and none of it addressed the actual constraint, which was that nobody would tell them where Villa was.
The Punitive Expedition became a sensational story, and carloads of U.S. reporters followed Pershing into action. The campaign, they wrote, would be a test of American power.7
The reporters are not incidental colour. They convert the operation into something that can be lost publicly, and they do it before it starts.
Two consequences the chapter does not separate.
The framing was set in advance and by other people. A test of American power is a standard nobody in the administration chose and everybody was now measured against. Catching Villa becomes the minimum acceptable outcome rather than a good one.
And the daily coverage supplies the tempo. A pursuit with no reporters has no interim results; a pursuit with a press corps generates a public non-event every day, and the accumulation is what turned it into "a kind of slapstick farce."
⚠ The administration invited them. This is the amplification decision at its clearest: the response to a raid that had reached a small audience was an operation designed from the outset to reach a very large one.
⚠ Seventeen Americans were killed at Columbus. Greene's framing treats this as the "minor annoyance" that should have been ignored, and the chapter's own parenthesis concedes the difficulty — "Pancho Villa had in fact killed American citizens."
🚩 The case for the law rests on the raid being small enough to disdain, and the chapter never argues that; it asserts it and then supplies four alternatives that are mostly covert operations rather than inaction. The honest version of this case is not Wilson should have ignored it but Wilson should have responded in a way that could not be watched.
🚩 SECONDARY WITHOUT PRIMARY · 🚩 SINGLE SOURCE — no historian named for any figure. The $130 million and 123,000 figures come from Greene and from a Villa quotation printed as a marginal, which is the adversary's own account of his adversary's costs.
🚩 The Mexican Revolution, Carranza's government and its consent to the expedition are compressed into clauses. The corpus should hold the mechanism and not the history, which is a substantial and contested field represented here by six paragraphs.
Just think—it cost your government $130 million to try to get me. I took them over rough, hilly country. Sometimes for fifty miles at a stretch they had no water. They had nothing but the sun and mosquitoes… And nothing was gained.8
⚠ Worth flagging as a structural point: Greene prints the target's own summary as the law's supporting evidence. And nothing was gained is the conclusion the chapter wants, delivered by the person with the strongest possible interest in that conclusion being believed.
It is also the sharpest statement of the finding, which is the awkward part. Villa's accounting is by cost and outcome — dollars spent, ground covered, nothing achieved — and it is a better argument than anything in the Interpretation. 🚩 MOTIVATED REASONING, and correct.
Something small and public has happened at your expense, and everyone is telling you that not responding will look weak.
Notice that "we must respond or we look weak" is an argument about your reputation and not about the problem. Wilson's advisers made it explicitly — he needed to prove his mettle. That is a real consideration and it is a separate one, and merging them is how the response gets sized wrong.
Check whether the problem is already resolving. Villa's own people had turned on him. This is the most commonly missed fact in these situations, because the moment of maximum provocation is the moment you are least likely to notice the other party's position deteriorating on its own. Ask who currently dislikes them and what that was trending toward before you got involved.
Price the response by audience size, not by proportionality. The instinct is to make the reply fit the offence. The correct calculation is how many people will know about this because of my response — and a proportionate response with a large audience is worse than a disproportionate one with none.
And do not invite the press. Whatever the equivalent is in your situation — announcing the review, copying the wider list, raising it in the meeting rather than after it. Once there is an audience the framing is set by them, the standard becomes catching Villa, and every day without a result is a public result.
⚠ The version that actually works is usually the one Greene puts in parentheses: a quiet response, through channels with no spectators, executed by people who were already inclined against them.
Strongest evidence. A dated case with a clear escalation record (10,000 → 123,000), a documented reversal of a pre-existing decline in the target's popularity, and a specific mechanism for that reversal — the reward that produced false leads.
Tension — the chapter's prescription is not what its own counterfactuals recommend. Three of the four are covert operations.
Tension — the raid killed seventeen people and the framing requires it to be a minor annoyance.
Tension — the supporting evidence includes the target's own press statement.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no historian, and the financial and troop figures trace to Greene and to Villa. 🚩 MOTIVATED REASONING in the marginal.
Open questions. Would a small covert operation have worked? Greene asserts it as the better path and the case for it is untested — the Carranza government had already consented to the expedition and had not caught Villa itself in the preceding period, which is at least weak evidence that the quiet options were harder than the chapter implies.
This is the Transgression for Law 36 and it argues the law's repaired version rather than its stated one.
Against Law 35's power sticks to those who conclude a revolution there is a direct connection the chapter misses. Villa is a furtherer who lost — Carranza concluded the Mexican Revolution and Villa returned to banditry, which is precisely the fate that sentence predicts. ⚠ The expedition then rescued him from it, which is the more interesting finding: an external intervention can restore a role the domestic sequence had already ended.
And against Law 34's Selassie the pair is exact and inverted. Selassie is the small party who wins the comparison by standing still. Villa is the small party who wins it by running. ⚠ Same mechanism — the audience prices the difference between the two parties — and the small party's correct behaviour is opposite in the two cases. The variable appears to be whether the contest is moral or physical, and two cases is not enough to state it.
Insurgency and asymmetric conflict — winning by not losing. Boer War Commandos and Asymmetric Tactics documents small mobile forces imposing disproportionate cost on a far larger conventional army, with the terrain and the sympathies of the local population as the decisive assets rather than any tactical advantage.
Villa's year in Chihuahua is that pattern with an American army in it.
The insight neither produces alone: the asymmetric-warfare frame explains the military outcome and Law 36 explains why the political outcome was worse than the military one. Pershing's failure to catch Villa is unremarkable — a mobile irregular in friendly terrain is not catchable by a conventional force, and this was known. What the pairing exposes is that the operation's real product was not a military result at all: the expedition manufactured the friendly population it then failed against. So the standard counter-insurgency lesson (the population is the terrain) sharpens into something specific about punitive operations — a force sent to punish rather than to hold is a force whose only mechanism for affecting the population is antagonising it, and it therefore degrades its own operating conditions at a rate proportional to its size. 123,000 men is not twelve times more effective than 10,000; it is twelve times more antagonising.
Public relations — the standard set by other people. Asymmetric Reputation Warfare treats reputational conflicts as governed by who establishes the criteria for success, and holds that the party who frames the contest usually determines the outcome regardless of the facts.
The reporters set "a test of American power" before a shot was fired.
What the pairing produces: Wilson's error is normally read as strategic overreach — too many men, too little chance. The framing lens says the loss was structural and preceded the strategy: the press established a binary standard (catch him or fail) that no available operation could satisfy, and once that standard existed, every outcome except capture was a defeat. Which means the decisive moment was the decision to make the operation visible, taken before the decision about its size. And that inverts the usual sequencing of these decisions: you choose the audience first and the response second, because the audience determines what counts as success — and most organisations do it the other way round, sizing the response and then discovering who is watching.
Sharpest implication. Villa's own people had already turned on him — the expedition reversed a decline that was underway, and the $50,000 reward converted the population into a counter-intelligence asset by paying people who had just been given a reason to prefer him. And the escalation is a public revaluation: 10,000 to 123,000 is a twelvefold upgrade of Villa's importance, paid for by the party upgrading him. A punitive force's only mechanism for affecting a population is antagonising it, so it degrades its own conditions at a rate proportional to its size.
Generative questions.