The greatest intellectual challenge in Fourth Generation war—war against opponents that are not states—is how to fight it at the operational level. NATO in Afghanistan, like the Soviets three decades ago, has been unable to solve that riddle. But the Taliban appears to have done so.

The operational level of war lies between strategy and tactics. While great commanders have always thought and fought at the operational level, the concept was not formally recognized until the 19th century. As usual, it was the Prussian army that led the way. Some historians think the operational level may have been formalized by Field Marshal von Moltke himself in the Franco-Prussian war as a way to keep Bismarck out of his business. (“Yes, my dear Bismarck, you are in charge of strategy, but you simply must not interfere in operational matters.”)

The U.S. Army did not officially recognize the operational level of war until 1982, but the tsarist Russian army and later the Soviets picked up on it. By 1944-45, the Red Army was as competent at what they called “operational art” as the Wehrmacht. That was never true of the Western allies.

The Russian term, operational art, is a good one, because unlike tactics or strategy it is not a thing but a link. It is the art, not science, of using tactical events, battles and refusals to give battle, victories and sometimes also defeats (from the North Vietnamese perspective, the Tet offensive was a tactical defeat but a decisive operational victory) to strike as directly as possible at the enemy’s strategic center. Because it resorts to battle only when and where necessary, operational art is a great economizer of fighting strength—even a battle won eats up soldiers, fuel, equipment, and, most importantly, time.

A brilliant example of its application comes from General Heinz Guderian’s XIXth Panzer Corps in the 1940 campaign against France. Guderian led the famous advance through the Ardennes mountains’ weakest point, the junction between the strong forces the French had pushed forward into Belgium and those manning the Maginot fortifications. After Guderian crossed the Meuse river at Sedan, he faced French forces coming up from the south. He could have stayed there and fought them. Instead, thinking operationally, he held the crossing with minimum force and threw everything he had north toward the English Channel. That collapsed the “hinge” between the French and British forces in Belgium and those in France, winning the campaign in one stroke. France, which by everyone’s account had the best army in the world, went down to defeat in six weeks.

Were war to remain in its Third Generation incarnation, a matter of fast-moving campaigns led by tank armies, the U.S. military might eventually get operational art. But war has moved on: tank armies are now as irrelevant as armies of mounted bowmen. So the question must be asked anew—how do you link tactical events to winning strategically?

The Soviet army focused its best talent on operational art. But in Afghanistan, it failed, just as we have failed. Like the Soviets, we can take and hold any piece of Afghan ground. And doing so brings us, like the Soviets, not one step closer to strategic victory. The Taliban, by contrast, have found an elegant way to connect strategy and tactics in decentralized modern warfare.

What passes for NATO’s strategy is to train sufficient Afghan forces to hold off the Taliban once we pull out. The Taliban’s response has been to have men in Afghan uniform— many of whom actually are Afghan government soldiers or police—turn their guns on their NATO advisers. That is a fatal blow against our strategy because it makes the training mission impossible. Behold operational art in Fourth Generation war.

According to a May 16 article by Matthew Rosenberg in the New York Times, 22 NATO soldiers have been killed so far this year by men in Afghan uniforms, compared to 35 in all of last year. The report went on to describe one incident in detail—detail NATO is anxious to suppress. There were three Afghan attackers, two of whom were Afghan army soldiers. Two Americans were killed. The battle—and it was a battle, not just a drive-by shooting—lasted almost an hour.

What is operationally meaningful was less the incident than its aftermath. The trust that existed between American soldiers and the Afghans they were supposed to train was shattered. Immediately after the episode, the Times reported, the Americans instituted new security procedures that alienated their native allies, and while some of these measure were later withdrawn,

Afghan soldiers still complain of being kept at a distance by the Americans, figuratively and literally. The Americans, for instance, have put up towering concrete barriers to separate their small, plywood command center from the outpost’s Afghan encampment.

Also still in place is a rule imposed by the Afghan Army after the attack requiring most of its soldiers to lock up their weapons when on base. The Afghan commanding officer keeps the keys.

One American soldier nonetheless advised a visitor to take an armed escort to the Afghan side of the base, which was about 100 feet away, ‘just in case.’

Multiply the aftermath of this incident 22 times since the beginning of the year and it becomes operationally important. Each incident quickly becomes known to all NATO troops in Afghanistan, which spreads the impact. Just a few hundred more such “green on blue” attacks will effectively end our training mission.

The Taliban know this technique is operational, not just tactical. They can be expected to put all their effort into it. What counter do we have? Just order our troops to pretend it is not happening—to keep trusting their Afghan counterparts. That order, if enforced, will put our soldiers in such an untenable position that morale will collapse.

So powerful is this taste of Taliban operational art that Washington may fear the example it sets. During a recent visit by Secretary of Defense Leon Panetta to Afghanistan, no American soldiers were allowed to get near him with loaded weapons. Might the Pentagon be worried that our own troops could learn from the Taliban? Were I an American soldier who had been told to hand over or unload his weapon before approaching Secretary Panetta, I would certainly have read it that way.

William S. Lind is director of the American Conservative Center for Public Transportation.