Kate Clark at Foreign Policy treats us to the confusion that is intelligence in Afghanistan.

On September 2, 2010, ten men in northern Afghanistan were killed in an air attack that was a targeted killing, part of the U.S. Special Forces ‘kill or capture’ strategy. The U.S. military said it had killed the Taliban deputy shadow governor of Takhar, who was also a ‘senior member’ of the Islamic Movement of Uzbekistan (IMU): one Muhammad Amin, as well as “eight or nine other insurgents.”

Many Afghans, including senior government officials, were incredulous. Many knew the man who had actually been targeted — who was not Muhammad Amin, but Zabet Amanullah. He had not fought for the Taliban since 2001 and had been out campaigning for his nephew in Afghanistan’s parliamentary elections with more than a dozen other men, mainly extended family members. That very morning, as per usual, he had called in to the district police chief to check on security before the election campaign convoy set off. The strike was an “obvious mistake,” said the provincial governor, Abdul-Jabar Taqwa. “He was an ordinary person and lived among normal people,” said the Takhar Chief of Police, Shah Jahaan Nuri. “I could have captured him with one phone call.”

U.S. Special Forces got the wrong man, but despite overwhelming evidence, they have remained adamant that they were correct. Senior Special Forces officers’ gave me lengthy accounts of the attack, including the intelligence behind it. That has allowed a piecing together of what went wrong.

Intelligence analysts were monitoring the calls of Muhammad Amin in early 2010 — and confirmed that he really was the Taliban deputy governor of Takhar. They came to believe that one number he had called in Kabul was passed on to him. They believed he began to use this phone and to ‘self-identify’ as Zabet Amanullah. In other words, they believed Muhammad Amin was using the name ‘Zabet Amanullah’ as an alias.

Friends and family have confirmed that Zabet Amanullah, who was living in Kabul, was in occasional telephone contact with active members of the Taliban, but this is not unusual in a country where fortunes change and it is prudent to stay in touch with all sides. Zabet Amanullah also kept in touch with senior members in the government. What may have looked like a suspicious cluster of calls and contacts, in reality, proved nothing about the actual conduct of the caller. Zabet Amanullah’s life, lived quietly and openly at home in Kabul, has been documented in detail. I met him in 2008 when he had just fled Pakistan where he had been working on a human rights project. He had been detained and severely tortured by the Pakistani intelligence agency, he believed, because he was a former Taliban commander who was not fighting.

The Special Forces unit denied that the identities of two different men, Muhammad Amin and Zabet Amanullah, could have been conflated. They insisted the technical evidence that they were one person was irrefutable. When pressed about the existence — and death — of an actual Zabet Amanullah, one officer said, “We were not tracking the names, we were targeting the telephones.”

Yet in the complex political landscape of Afghanistan, it is not enough to track phones. It is certainly not enough to base a targeted killing on. The analysts had not built up a biography of their target, Muhammad Amin —where he was from, what his jihadi background was, and so on. They had not been aware of the existence of a well-known person by the name of Zabet Amanullah. They had not had access to the sort of common, everyday information available to Afghans watching election coverage on television. They had not made even the most basic background checks about a target they had been tracking for months. Instead, they relied on signals intelligence and network analysis (which attempts to map insurgent networks by monitoring phone calls), without cross-checking with any human intelligence.

There are a whole host of problems here, including (a) Ms. Clark concluding with certainty that the target was a case of mistaken identity, (b) it being commonplace to stay in contact with insurgents (that they haven’t been marginalized by now is testimony to the state of the campaign), (c) the supposed need to have absolute certainty in intel before reacting, and (d) the lack of local atmospherics.  But let’s assume the accuracy of Ms. Clark’s report for the sake of argument.  It is this last item I want to discuss for a moment.

Recall in Good Counterinsurgency, Bad Counterinsurgency and Tribes that we discussed the value of local intelligence and direct action kinetics being conducted by the same Soldiers and Marines who patrol villages every day.

The American patrol set out from a base in Yahya Khel district center at 6 p.m. Tuesday, planning to provoke a fight with a team of Taliban sharpshooters suspected to be operating around the village of Palau. The troops, from Angel Company, 3rd Battalion, 187th Infantry Regiment, dropped off a team at a small Afghan army outpost and then moved by foot toward the village.

Just before dusk, the patrol was ambushed, not by the expected long-range marksmen, but by a team of gunmen who attacked with rifles and grenades from as close as 50 feet away. Two American soldiers were wounded. Half an hour later, at the outpost, Angel Company’s commander, Capt. Joshua Powers, received permission over the radio from Col. David Fivecoat, the battalion commander, to call in fire from attack helicopters. The pilots had watched a group of fighters move from the area of the gun battle to a courtyard in a small village north of Palau. They told Captain Powers that they could make out a machine gun and several rifles. At 8:38 p.m., one of the helicopters fired a Hellfire missile into the cluster, then shot another man who was on the roof of the building abutting the courtyard. Over the next half hour the helicopters attacked two more groups of suspected fighters in the area with cannon fire.

In the dark, Angel Company walked north from the outpost to assess the damage. In the courtyard, the corpses of two men were illuminated by burning weapons and motorcycles. While his medic tended to a third man, severely wounded and clad in camouflage, Captain Powers radioed his battalion with bad news: The building by the courtyard was a mosque. The pilots had not known, since no loudspeakers were visible and identifying writing was visible only from the ground. There was shrapnel damage to the walls, and the roof had a hole in it from cannon rounds.

The patrol, along with a group of Afghan soldiers and their commander, Lt. Col. Mir Wais, stayed the night outside the mosque. The Taliban would undoubtedly claim that civilians had been killed, Captain Powers explained, and he wanted to be there when the villagers woke up to show them the weapons and combat gear. “If we hold this ground, we can show them the evidence right away,” he said. “The first story is usually the one that sticks.”

The pilots thought they had killed half a dozen fighters at a second site the helicopters had attacked, but the bodies were already gone when the patrol arrived. Captain Powers acknowledged that this meant there was no way to know for sure whether civilians had been killed, but thought it unlikely: the site was secluded, and among charred motorcycles there were rocket-propelled grenades and camouflage vests with rifle magazines. At the first site, all four bodies — the two in the courtyard, the one on the roof, and the wounded man, who later died — wore camouflage fatigues and similar vests, containing grenades, ammunition, makeshift handcuffs and a manual on making homemade explosives.

Around 5 a.m., the men of the village started to congregate by the mosque. Captain Powers and Colonel Mir Wais addressed them, telling their story of what had happened. The men complained that the strike had frightened their wives and children and damaged the mosque, and that they were trapped between the pressures of the Americans and the Taliban. But they did not suggest that any residents of the village had been wounded or killed, and did not claim the bodies. Later in the morning, the district subgovernor, Ali Muhammad, described the night’s events to citizens gathered in the Yahya Khel bazaar. He also signed, along with Captain Powers, a letter about the attack  that would be distributed in the area after dark: a counterpoint to the Taliban’s infamous “night letters.”

And as I previously observed, the same people who ordered the strike were there to explain it in the morning, just as I suggested should happen.  The same people who fight by night are there for the locals to look at in the morning.  And look into their eyes.  If they see cut and run, they will side with the insurgents, or someone else, whomever that may be.  If they see victory and determination, they will side with the stronger horse.  We need to be the stronger horse.

Once again this explains why a small, SOF-driven, raid-based hunt for high value targets can never work as a strategy in Afghanistan, in spite of how much we might like to withdraw and let dark operations take over.  In addition to the lack of logistical support (where do you think the SOF troopers get their supplies?), the most significant problem with a small footprint in Afghanistan is that the intel will be worthless.  Unless we have boots on the ground, patrolling every day, living amongst the population, looking into their eyes, we will never know who should be courted and who should be killed.