Breakthrough Tactics

The war ended in November 1918. By January the delegations had gathered at Paris, and every general staff in Europe was asking the same question, though none of them knew it was the same question, because the answers they arrived at would be so different that it would take another war to determine which one was right.

The problem was this: four years of industrialized slaughter on the Western Front had produced a condition nobody’s doctrine could explain. The armies of 1914 had marched into Belgium and northern France expecting a war of maneuver, expecting Sedan again, expecting the decisive battle that all the prewar theory promised. What they got instead was a line. Seven hundred kilometers of trench from the Channel to Switzerland, and for four years that line barely moved. Millions of men were fed into it. The line absorbed them and held. Offensives were launched with the concentrated fury of entire national economies and gained, in the best cases, a few kilometers of mud that would be lost again within the month. The Somme. Verdun. Passchendaele. Names that became shorthand not for battles but for the discovery that battle, as the nineteenth century understood it, had stopped working.

Every country that fought in that war drew a lesson from it. Every lesson was different. Every lesson made sense from inside the experience that produced it. Most of them were wrong, but not in the same way, and not for the same reasons, and the specific shape of each wrongness tells you more about the country than the war.

The British invented the tank and then could not decide what they had invented.

On September 15, 1916, forty-nine Mark I tanks rolled toward the German lines at Flers-Courcelette, on the Somme. Most broke down. Several ditched in shell craters. A handful reached the German trenches and the infantry following them advanced further that morning than any British attack had managed in two years.¹ The machine was slow, unreliable, half-blind, and loud enough to announce itself from a mile away. It was also the first weapon since 1914 that had crossed no man’s land without being stopped by the machine gun. The question was what that meant.

J.F.C. Fuller thought it meant everything. Fuller was the chief of staff of the Tank Corps, a man of vast intelligence and increasingly strange politics who would later attend Hitler’s birthday party and develop an enthusiasm for Aleister Crowley. But in 1917 he was simply the person who understood what the tank could become before anyone else in the British Army was ready to hear it.² His Plan 1919, drawn up for an offensive that the armistice rendered unnecessary, called for massed armored formations striking deep into the German rear, bypassing the front line entirely, paralyzing command and communication at the source. Not attrition. Not the infantry slog. A different kind of war, built around the machine rather than the man.

B.H. Liddell Hart arrived at a similar conclusion by a different route. He had been gassed on the Somme in 1916 and spent the rest of his career developing what he called the indirect approach: the argument that decisive results in war come not from frontal confrontation but from dislocation, from striking where the enemy isn’t, from using speed and surprise to collapse the structure before the structure can respond. His histories traced the principle from Scipio Africanus through Sherman’s march through Georgia. The tank was, for Liddell Hart, the instrument that finally made the indirect approach mechanically feasible.³

The British Army listened to both men, studied both arguments, conducted exercises and wrote papers and debated the question across two decades of interwar calm. And then it built an army that couldn’t do any of it.

The problem wasn’t intellectual. The British understood armored warfare as well as anyone, probably better. The problem was imperial. Britain in the interwar period was a global empire managing colonial commitments across three continents, and the army that managed those commitments needed to be configured for policing, not for breakthrough. Small garrisons, light equipment, expeditionary flexibility. The kind of army that could suppress unrest on the Northwest Frontier or maintain order in Palestine, not the kind that could concentrate seventy-ton tanks at a Schwerpunkt on the plains of northern Europe. The intellectual tradition pointed one way. The institutional incentives pointed another. The institution won, because institutions always win, because the garrison needs to be staffed this quarter and the hypothetical armored corps for a hypothetical European war can be built later, and later never quite arrives until it arrives all at once and the army you have is not the army you need.⁴

When the war came in 1939, the British Expeditionary Force that went to France had tanks, but organized piecemeal, scattered across infantry divisions as support weapons rather than concentrated into armored formations capable of independent action. The British had theorized the breakthrough. They sent an army designed for something else.

The Japanese drew perhaps the strangest lesson of all, and they drew it from the farthest distance.

Japan had entered the Great War on the Allied side and seized German colonial possessions in the Pacific and at Tsingtao with minimal effort. The Western Front was not their war. But they sent observers, and the observers watched, and what they reported back entered a military culture that was already organized around a proposition the Western Front should have destroyed: that spiritual force could overcome material disadvantage.

The Imperial Japanese Army in the interwar period developed a doctrine that looked, from a distance, like a deliberate refusal to learn from Europe. Where the Europeans had concluded that firepower dominated the modern battlefield, the Japanese concluded that firepower could be negated by superior morale, superior training, and an offensive spirit so total that the individual soldier’s willingness to die became itself a tactical instrument.⁵ The bayonet charge. Night attacks. Close combat in jungle and mountain terrain where the machine gun’s range advantage was compressed. An entire theory of war built on the assumption that the will to fight, pushed to its absolute limit, could break the equations that governed the Western Front.

This sounds like mysticism, and in part it was. The Imperial Army’s field manuals spoke of seishin, fighting spirit, as though it were a material resource that could be manufactured through training and ideology. And there was something in this that resembled Mishima more than Clausewitz, the aestheticization of sacrifice, the body offered up as proof of a metaphysical claim about national destiny. But it wasn’t only mysticism. The Japanese had a strategic problem that no amount of rational analysis could solve: they were an island nation with roughly a tenth of America’s industrial capacity, contemplating a war against opponents who could outproduce them by any material metric. If the war was about production, they lost before it started. If the war was about something else, about willingness to absorb casualties, about the defender’s advantage in jungle terrain, about breaking the enemy’s will before the enemy’s factories could reach full output, then maybe. Not a good bet. But the only bet available.⁶

It worked in China, where the opponent was less organized and the terrain permitted the offensive doctrine to function. It worked in Malaya, where a force of thirty-six thousand defeated a garrison of eighty-five thousand by moving faster through terrain the defenders had assumed was impassable, bicycles on jungle paths, an advance so rapid it outran its own supply lines because the troops were trained to forage and the doctrine assumed they would. Singapore fell in a week.

Then came Guadalcanal, and the Kokoda Track, and Iwo Jima, and the doctrine met an opponent with unlimited material, functional logistics, and no intention of breaking. The night charges into American automatic weapons fire on Henderson Field. The banzai attacks on Saipan, entire battalions running into prepared positions because the doctrine said spirit would carry the position and the doctrine could not be wrong because to question it was itself a failure of spirit, a recursive trap that made the theory unfalsifiable from inside. Three thousand men charged the American lines on Saipan on July 7, 1944. Most of them died within an hour. The theory that willpower transcended firepower was tested against the reality of interlocking fields of fire and the test was not ambiguous.

The lesson the Japanese drew from the Western Front was not exactly wrong. Morale does matter. Offensive spirit does matter. Willingness to close with the enemy at close range in difficult terrain is a real tactical advantage. What was wrong was the extrapolation, the conversion of a partial truth into a total doctrine, the decision that because morale mattered it could substitute for everything else. A true observation, generalized past its domain of validity, until it became a machine for producing courage and corpses in equal measure.

The Italians fought their own private Western Front, and it broke them in a way that is still legible.

Eleven battles of the Isonzo. The name itself has the quality of a diagnosis. Between June 1915 and September 1917, the Italian army launched eleven offensives across the Isonzo River against Austrian positions in the Julian Alps, and each time the result was the same: a preparatory bombardment, an infantry advance into limestone karst where the defenders had dug into rock rather than mud, casualties that mounted in the specific way casualties mount when men are asked to climb uphill into machine gun fire, and gains measured in hundreds of meters.⁷ Luigi Cadorna commanded these offensives with the particular rigidity of a man who had one idea and eleven opportunities to discover it wasn’t working. He did not discover this. He ran the same attack eleven times and got the same result and concluded each time that the problem was insufficient commitment rather than insufficient method, which is also a kind of overfitting, though the training data in this case was not the last war but his own prior failure, replayed at higher cost.

Then came Caporetto. October 1917. An Austro-German force using the same infiltration tactics that would later be called Hutier’s method, the same short bombardment and deep penetration that the Germans were developing for the Spring Offensives, struck at the junction of two Italian armies in the upper Isonzo valley. The Italian line didn’t bend. It evaporated. Three hundred thousand prisoners in three weeks. A retreat of over a hundred kilometers. Entire divisions dissolved. The army that had spent two years grinding against the same positions was unable to hold any position at all once the nature of the attack changed, because the army had been built for one kind of war, the attritional slog, and had no institutional reflex for another.⁸

Caporetto produced two contradictory impulses in Italian military thinking, and neither won cleanly. The first was the lesson of trauma: never again. Defensive depth, caution, the Maginot instinct. The second was a strange inversion: the belief that what had broken the Italian army was not the method but the morale, that the real failure was spiritual weakness, that what Italy needed was not better doctrine but harder men. Mussolini, who was injured on the Isonzo in 1917, built an entire political movement on the second interpretation. The fascist reading of Caporetto was that Italy had been stabbed in the back by defeatism, that the army’s collapse was a moral failure that could be cured by national regeneration, by the forging of the new Italian man who would not break.

The army Mussolini built reflected both impulses and neither coherently. It talked about guerra di rapido corso, rapid decisive war, speed and shock and the offensive spirit, a doctrine that sounded modern and aggressive. But the industrial base couldn’t support it. The tanks were too light. The logistics were underbuilt. The officer corps was politicized. The doctrine was a costume draped over an army that could not execute it, which produced, in North Africa and Greece and the Balkans, the specific kind of failure that occurs when an institution’s self-image has diverged from its actual capabilities and no one inside has the political room to say so.⁹

Italy learned from Caporetto that morale was the decisive factor. Japan learned from the Western Front that spirit could overcome material. These are not the same conclusion, but they share a family resemblance, and the family resemblance is this: both took a partial truth, that human factors matter, and inflated it into a total explanation that conveniently displaced the material questions they couldn’t answer. If the problem is spiritual, the solution is ideological. If the solution is ideological, you don’t need to count the tanks.

The Americans arrived late and drew a lesson from arriving late.

The AEF reached France in 1917 and fought in earnest for roughly six months, from the spring to the autumn of 1918. This is not enough time to develop an institutional trauma. The British and French had been bleeding for three years by the time the Americans showed up, and the American experience, while violent, was compressed and conclusive in a way that the European experience was not. The Meuse-Argonne was brutal. It was also the last act. The war ended while the lessons were still forming.

John Pershing insisted on two things that shaped how the American army thought about the war afterward. First, that American forces would fight as a unified command under American officers, not as replacements fed into British and French units. Second, that the American soldier’s advantage was his aggressiveness, his willingness to attack, his rifle marksmanship, his frontier directness, and that the trench warfare the Europeans had settled into was a product of exhaustion and timidity rather than tactical necessity.¹⁰

This was, in many ways, the same error the Japanese would make from a greater distance, the belief that offensive spirit could override the defensive advantage of modern weapons. But Pershing had something the Japanese didn’t: unlimited industrial capacity behind him. If you combined aggressive infantry doctrine with overwhelming material superiority, the offensive spirit didn’t need to do all the work. It just needed to do enough. The factories would do the rest.

The interwar American army was small, underfunded, and largely committed to continental defense and the Philippines. But the doctrine it carried into the next war was essentially Pershing’s, updated: the assault supported by overwhelming firepower and logistics so deep that the enemy couldn’t outlast them. Not elegant. Not subtle. Not the indirect approach. The direct approach, applied with such material preponderance that finesse became optional.

This worked in Europe, more or less, not because it was a good doctrine but because the American industrial base was so overwhelming that a mediocre doctrine backed by unlimited supply could achieve what a brilliant doctrine backed by limited supply could not. Normandy to the Rhine was not a campaign of operational art. It was a campaign of production, of logistics, of the brute capacity to replace losses faster than the enemy could inflict them. Patton was the exception. Eisenhower was the rule. The broad-front strategy that spread Allied forces across the entire line rather than concentrating at a Schwerpunkt was the opposite of what the Germans would have done. It worked because the Allies could afford to be everywhere at once, and the Germans couldn’t afford to be anywhere enough.¹¹

The Pacific war corrected some of this. Island-hopping required exactly the kind of combined-arms coordination and concentrated assault that the European campaign didn’t demand, and the Marines in particular developed an operational culture of initiative and improvisation that Auftragstaktik would have recognized. But the dominant American lesson from the Second World War was the one the industrial base taught: mass wins. Production wins. Logistics wins. Apply enough force across a broad enough front and the details sort themselves out.

This is a true lesson in the specific domain where it was learned. It is also a catastrophically overfit lesson, as Korea and Vietnam and Iraq would demonstrate in sequence, each time the American military applying the overwhelming-force model to a conflict whose decisive features were not about force at all. But that is a different essay, and this one is already long enough.

Serbia is the case that doesn’t fit, which is why it matters.

Serbia entered the Great War in 1914 as a nation of roughly 4.5 million people and came out of it missing a quarter of them. Not a quarter of its army. A quarter of its population. The war for Serbia was not a question of doctrine or analysis or institutional learning. It was a question of survival, and the answer to that question was not theoretical.

In the autumn of 1915, a combined Austro-Hungarian, German, and Bulgarian force invaded from three sides. The Serbian army, already depleted by a year of fighting and a typhus epidemic that had killed more soldiers than combat, could not hold. King Peter I, seventy-one years old and too ill to ride a horse, was carried in an oxcart. The army and a substantial portion of the civilian population retreated through the mountains of Albania in winter, through terrain that was not passable for an army and that they passed through anyway, losing men to cold and starvation and Albanian guerrillas and the simple mechanical fact that a body moving through snow at altitude eventually stops. Somewhere between a hundred thousand and two hundred thousand people died on the retreat. The exact number is not known. The Adriatic was the destination, and the survivors who reached it were evacuated to Corfu by Allied ships, reconstituted, and sent to fight again at Salonika, where they held a front for two more years and participated in the final breakthrough in September 1918 that knocked Bulgaria out of the war.¹²

What does a nation learn from this? Not what the British learned, or the French, or the Japanese. Serbia did not have the luxury of drawing doctrinal conclusions from the Western Front because Serbia’s war was not the Western Front. Serbia’s war was about whether Serbia would continue to exist, and the answer had nothing to do with infiltration tactics or defensive fortification or fighting spirit as a substitute for industrial capacity. The answer was that a nation can be reduced to a column of people walking through mountains in winter, losing everything except the decision to keep walking, and that this is sufficient if you survive it and insufficient if you don’t, and there is no doctrine that covers the difference.

The Kingdom of Serbs, Croats, and Slovenes that emerged from the war drew a lesson that was not about tactics at all. It was about geography, about the vulnerability of a small state surrounded by larger ones, about the fact that no amount of military excellence can substitute for strategic depth when you don’t have any. The interwar Yugoslav state built its military around the assumption that it would be invaded and could not stop the invasion and would need to fight a guerrilla war from the mountains. This is not a lesson about breakthrough. It is a lesson about what happens to countries for whom breakthrough is irrelevant because the problem is not how to break the enemy’s line but how to survive the enemy breaking yours.¹³

In April 1941, the Wehrmacht invaded Yugoslavia and defeated the conventional army in eleven days. The guerrilla war that followed lasted four years. The lesson, such as it was, had been learned.

The Nivelle Offensive broke something in France. Not the line, which held. The army behind it.

In the spring of 1917, Robert Nivelle promised a breakthrough along the Chemin des Dames within forty-eight hours. He had the artillery, the planning, the confidence of a man who had studied his own success at Verdun and concluded it was a method rather than a circumstance. The attack gained almost nothing. Casualties ran to 187,000 in the first ten days. By June, units across the French army were refusing orders. Not deserting, that would have required somewhere to go. They simply stopped. Held the trenches, manned the positions, and declined to walk forward into the thing again. Somewhere between mutiny and strike, a distinction the high command preferred not to examine too closely.

Nivelle was removed. Pétain replaced him, restored order through a combination of improved conditions and selective executions, and the army recovered enough to finish the war. But the mutinies of 1917 had deposited something in the institutional memory of the French military that no subsequent victory could dislodge. France had spent 1.4 million men. Ten percent of its prewar male population. The officer class that survived drew a conclusion with the clarity of people who have watched the bill arrive and know exactly what it cost: the offensive was finished. Modernity, the machine gun and the artillery barrage and the wire, had made attacking more expensive than defending by a ratio that no amount of courage or planning could overcome. The next war, if it came, would be fought from concrete.

The Maginot Line is what that conclusion looks like poured into the ground. Steel and reinforced concrete running the length of the German border, underground barracks connected by tunnels, retractable gun turrets, artillery casemates with interlocking fields of fire, ammunition hoists, ventilation systems, mess halls sixty feet below the surface.¹⁴ It was the Western Front made permanent. The trench system that had held for four years, cleaned up, hardened, given electricity and institutional confidence. Command was centralized. Initiative was not merely discouraged but structurally eliminated. No unit would advance without coordinated artillery support approved through the chain of command, because the last time units had advanced on their own initiative, the Chemin des Dames had happened, and before that the Somme, and before that the frontiers, and the French army had concluded, with the finality of an organism that has touched a hot surface, that the cost of attacking was not a problem to be solved but a fact to be accepted.

Every part of this made sense if the next war was the last war. The problem, which would become visible in six weeks in 1940, was that it wasn’t.

The Germans looked at the same Western Front and saw something else entirely. And because they were the ones who got it right, or closest to right, the story of what they saw and what they built from it is worth telling at length, because the mechanism is more interesting than the outcome.

The Reichswehr that emerged from Versailles was limited to a hundred thousand men, forbidden tanks, forbidden an air force, forbidden a general staff, which it reconstituted immediately under a different name because the prohibition was a formality and everyone involved understood this. Hans von Seeckt, who ran the new army, had a specific idea about what the hundred thousand should be: not an army but a cadre. Every soldier trained to perform the duties of the rank above him. Every officer a potential commander. The constraint of Versailles, which was meant to be a punishment, became a selection pressure that produced the most professionally dense military force in Europe.¹⁵

They studied the war differently than the French. Not from the top, not from the perspective of the general surveying the map and tallying the casualties, but from the bottom. What had actually happened at the point of contact? What had the stormtroop units of 1917 and 1918 done differently? Because something had changed in the last year of the war, something the Allies had felt without fully understanding.

The stormtroops were Oskar von Hutier’s contribution, though the method had precursors and the name is slightly misleading.¹⁶ The idea: instead of attacking along the entire line after days of preparatory bombardment that eliminated surprise and churned the ground into impassable mud, you found the weak points. Short, intense artillery preparation, heavy on gas and smoke rather than high explosive. Small, heavily armed units infiltrating through the gaps rather than advancing in waves. Bypass the strong points. Let them wither. Push through into the operational depth, where the enemy’s command structure and supply lines and reserves lived, and collapse the system from inside rather than trying to batter down the front door.

It had worked, tactically. The Spring Offensives of 1918 gained more ground in weeks than the previous three years of fighting had managed in total. It failed strategically because the German army was exhausted, the logistics couldn’t sustain the advance, and there were two million American soldiers arriving who hadn’t spent four years being ground down. But the method itself. The Reichswehr looked at the method and concluded it was correct. Not a stopgap. Not a desperation measure. The foundation of something.

From this came two ideas that would reshape how wars were fought.

The first was Auftragstaktik, mission-type tactics. The commander gives the objective and the reason for the objective. The subordinate decides how to achieve it. This sounds simple, almost obvious, until you consider that it is the structural opposite of what the French were building. The French system centralized decision-making because initiative had been too costly. The German system decentralized it because rigidity had been too slow. Same war. Same data. Opposite conclusions. The difference was which failure each army found more intolerable: the cost of doing too much, or the cost of doing too little.

The second was Schwerpunkt, the point of main effort. Not just concentration of force, which is as old as warfare, but the idea that the entire organization should orient itself around a single decisive point, that every element from logistics to air support to reserves should be structured to feed the breakthrough rather than distributed evenly along the line. The opposite of the Western Front, where everything had been spread across seven hundred kilometers and nothing had been concentrated enough to break through. The Schwerpunkt doctrine said: find the weak point, mass everything there, punch through, and exploit before the enemy can reconsolidate. Accept risk everywhere else. The line will be thin where it doesn’t matter. It will be overwhelming where it does.

Put these together, Auftragstaktik and Schwerpunkt, initiative at every level and concentration at the decisive point, and what you have is a military philosophy built on the conviction that the Western Front was an aberration, not a new norm. That the stalemate was a product of specific conditions, symmetrical forces with no ability to exploit a breach, and not a permanent feature of industrial warfare. That breakthrough was still possible. You just had to be faster and more concentrated than the defense could respond to.

The question was whether any of this was true, or whether it was the wish of a defeated army dressed in the language of analysis.

On May 10, 1940, the question was answered.

The German army attacked through the Ardennes. The Ardennes was supposed to be impassable. This was not a casual assumption. French military planning had studied the terrain, the narrow roads through dense forest, the river crossings, the bottleneck at Sedan, and concluded that a major armored thrust through the region was logistically impossible. The forest was too thick, the roads too few, the columns would be too vulnerable to air attack, and even if an armored force somehow reached the Meuse it would face fortified positions on the far bank with no room to maneuver. The analysis was competent. It was also the last war’s analysis applied to the next war’s technology, and the difference between the two was about to become visible in the most violent way the twentieth century had yet produced.

Heinz Guderian’s XIX Panzer Corps reached the Meuse at Sedan on May 12. The columns through the Ardennes had stretched back over a hundred kilometers, bumper to bumper along forest roads, the single greatest concentration of armored vehicles in history packed into a space that a competent air force could have turned into a graveyard.¹⁷ The French air force did not respond in time. The reconnaissance was late, the reports were disbelieved, the chain of command that was supposed to prevent reckless action also prevented rapid reaction, because those are the same mechanism seen from different ends.

On May 13, Guderian crossed the Meuse. The textbook said to wait for the infantry to catch up, to consolidate the bridgehead, to bring the artillery forward. Guderian did not wait. Auftragstaktik: the objective was to cross. He decided how. He requested Luftwaffe support in place of artillery and sent his armor across on pontoon bridges while engineers worked under fire and the French positions on the far bank, which had been designed to stop an infantry assault supported by artillery, received instead an assault supported by dive bombers whose contribution was less about physical destruction than about the sound, the Jericho Trumpets on the Stukas, a scream designed to do to the nervous system what the bombs did to the fortifications.¹⁸

The breach was narrow. A few kilometers. In the French doctrine, a narrow breach was containable: you identified it, moved reserves to the flanks, counterattacked, sealed it. This is exactly what the French attempted. The problem was speed. The reserves were not positioned near Sedan because Sedan was not supposed to be a point of attack. The orders to move reserves traveled up the chain of command and back down again, because the system required coordination and coordination required time, and time was the one thing the German doctrine was designed to eliminate. By the time the French reserves received their orders, the breach was no longer narrow. Guderian’s tanks were through and running west, into open country, toward the Channel, and the Schwerpunkt that had been a point was becoming a line and then a flood.

In six weeks France fell. The Maginot Line was never breached. It didn’t need to be. The Germans went around it, through the Ardennes, exactly where the defense was weakest because the defense was weakest where the analysis said an attack was impossible. The most sophisticated fortification system ever built, the physical embodiment of everything the French had learned from the last war, sat intact and irrelevant while the war it was designed to prevent happened somewhere else.

The obvious reading is that the Germans were right and everyone else was wrong. Maneuver beats defense. Initiative beats rigidity. The Reichswehr’s analysis was correct and the Maginot Line was a monument to institutional failure.

This is mostly true and entirely insufficient.

The French were not wrong about the data. They were wrong about the domain. The Western Front really had favored the defense. The machine gun and the artillery barrage really had made frontal assault prohibitively expensive. The Maginot Line really would have held against the kind of attack the last war had produced. The analysis was rigorous. The conclusion was valid. The problem was that the conclusion was valid for a war that was over, and the next war had different parameters, and the features that made the defense dominant in 1914 to 1918, symmetric force structures, no operational exploitation capability, slow communication, horse-drawn logistics, were not permanent features of warfare but contingent features of a specific technological moment that was already passing.

The Germans saw this. Or more precisely, the Germans looked at the same evidence and asked a different question. The French asked: what does the data say about the relationship between offense and defense? The Germans asked: under what conditions did the offense fail, and can those conditions be changed? The first question produces a fixed conclusion. The second produces a research program. The French built a wall. The Germans built a method.

But there is something uncomfortable in this, which is that the German analysis was not obviously more rational than the French. It was more correct, which is not the same thing. The French had better reasons for their conclusion than the Germans had for theirs. The evidence for defensive dominance was overwhelming: four years and millions of dead. The evidence for the stormtroop alternative was thin: a few months of tactical success that had failed to produce strategic results. If you had shown both analyses to a neutral evaluator in 1925 and asked which one was better supported by the evidence, the honest answer would have been the French one. The German conclusion required a leap, a belief that the tactical kernel of the stormtroop method could be scaled and mechanized into something the 1918 version had not been. That belief turned out to be correct. It was also, at the time, closer to a hypothesis than a finding.¹⁹

And the German answer, even where it was correct, was correct only within its domain. Sedan worked. Barbarossa worked for five months and then met the Russian winter and the vast strategic depth that no Schwerpunkt could compress. Stalingrad was the Schwerpunkt applied to a city, where concentration at a single point meant the flanks were held by Romanian and Italian armies that couldn’t hold them, and Zhukov’s Operation Uranus struck precisely where the Germans were weakest because the German doctrine required them to be weak somewhere and the Soviets had finally learned, after twenty million dead, where to look.²⁰ The method that had been vindicated in France was tested at continental scale and the test revealed its limits. The breakthrough doctrine could win campaigns. It could not win a war against an opponent with sufficient depth to absorb the breakthrough and sufficient reserves to counterattack the flanks the breakthrough exposed. The Germans had solved the tactical problem of the Western Front. They had not solved the strategic problem of fighting an opponent who could trade space for time, and the confidence produced by the tactical solution made the strategic problem invisible until it wasn’t.

Which means the lesson is not simply “be like the Germans.” The lesson is stranger and less comfortable than that. Sometimes the analysis that is less supported by the evidence is the one that is correct, because the evidence is drawn from conditions that are changing, and the analysis that fits the evidence best is the one most tightly coupled to conditions that no longer hold. And sometimes even the correct analysis is correct only within a range, and the confidence it produces is what carries you past the boundary of that range into the place where it fails.

Seven countries, seven lessons, all drawn from the same war.

The British understood what the next war would require and built an army for a different purpose. The Japanese saw the limits of material warfare and concluded that spirit could transcend those limits and were right in the specific and wrong in the general. The Italians experienced a catastrophic breakthrough against them and concluded the problem was moral rather than structural and built an army around a self-image it could not sustain. The Americans arrived with enough material to make doctrine optional and concluded that material was doctrine. The Serbians learned that small countries die or they don’t and that no analysis of breakthrough tactics is relevant to the experience of walking through Albanian mountains in winter carrying your king in a cart. The French learned the defense wins, poured that lesson into concrete, and discovered that the next war went around it. The Germans learned the offense could still work, proved it at Sedan, and then drove the proof into Russia until the proof broke.

Each lesson was correct about something. None was correct about enough.

What I keep returning to is not the individual lessons but the mechanism. The same dataset, the same war, the same four years of evidence. Each country extracted the lesson its institutional structure and national trauma and strategic position allowed it to extract. The extraction was not random. It was determined by what the country needed to believe, what it could afford to build, what its officer class was trained to see, what its recent dead had seemed to die for. The data was the same. The priors were different. And the priors dominated.

This is, I think, the uncomfortable thing about learning from experience. The experience is real. The lessons are real. They are also artifacts of the position from which the experience was observed, and the position is not the world, it is a point in the world, and from any given point some things are visible and others are not, and the things that are not visible are not announced as absent, they are simply not there, and you cannot miss what you cannot see.

The next war, when it came, was not any of the wars they had prepared for. It was all of them and none of them. The breakthrough happened where the doctrine said it would and also where the doctrine said it couldn’t. The spirit mattered where the spirit theorists said it would and also failed where they said it wouldn’t. The material won where the materialists predicted and also couldn’t reach the places where the war was decided by other means. Every lesson was partially vindicated and totally insufficient.

Maybe that’s the only honest conclusion. That the war you prepare for is never the war you get. That the model trained on the last dataset will fail on the next one in ways the model cannot predict because the model’s blindnesses are structural, built into the point from which the observation was made. That you prepare anyway, because the alternative is not preparing, which is worse, and that the preparation will be wrong in ways you discover only after the preparation meets the thing it was not designed for.

Guderian crossed the Meuse because he believed the textbook was wrong about what was possible. He was right. But “I believe the textbook is wrong” is also what every failed general in history has said, right up until the moment the river didn’t cooperate or the flank collapsed or the reserves arrived. The difference between Guderian at Sedan and Guderian’s doctrine at Stalingrad is not a difference of method. It is a difference of context. From inside, both looked correct.²¹

The river is in front of you. Every army that has ever crossed one thought it knew what was on the other side. None of them were entirely right. Most of them crossed anyway.