There were too many bodies to bury by hand alone.
The news had spread across south-eastern Afghanistan in a series of phone calls: there had been an airstrike near the village of Bati Tana.
As mourners made their way to the hardscrabble borderlands where the missiles struck, the charred corpses lay bundled in burial cloth. The dry soil of the cemetery had to be broken up by a farming machine before so many individual graves could be dug.
The US aircraft attack on June 5 2015 hit two pickup trucks as they drove on a rock-strewn track near the border with Pakistan. Photos taken shortly afterwards show the vehicles burnt down to their mangled frames. Scorched scraps of rubber and metal had been flung meters from the main wreckage. No survivors were reported. Fourteen people were killed.
There is no consensus on who these 14 really were. Both the UN and Nato say they know – but their versions are startlingly different.
Khost province, where Bati Tana lies, has seen many US airstrikes over the years. It sits next to Pakistan’s Waziristan region and militants of all stripes rub shoulders with pastoral nomads in its mountainous border areas. It was in Khost that the Taliban handed over captured US soldier Bowe Bergdahl in 2014. The most recent strike there was in November, when at least 10 militants were reported killed.
But there was something different about the June 5 strike. Within hours of the attack, the identity of the dead was being disputed in the local media. An alarming claim was being put forward by some of the nomads: those killed were civilian mourners attending a local man’s funeral.
Both the UN and Nato launched inquiries to find out the truth. Neither have published detailed findings, but Nato says all 14 were Taliban insurgents.
The UN has come to another conclusion, the Bureau discovered during a four month investigation into the strike. A UN source told the Bureau it has classified all of them civilians – a view echoed by the Afghan government’s lead investigator.
Which of these organisations is correct matters, and not just to the relatives of the dead.
If the UN’s findings on Khost are accurate, the 14 deaths were, on the face of it, part of a dramatic escalation in the rate of civilian casualties caused by international airstrikes. Both the Khost incident and the more high profile killing of 42 civilians in a Medecins Sans Frontieres hospital in October contributed to an overall civilian death toll that reached 103 in 2015, according to the UN’s latest data.
These 103 civilian deaths came from 411 US airstrikes, an average rate of one civilian killed every four strikes. The year before, when most US troops were still in the country, the rate was roughly one civilian death every 11 strikes.
This rise comes amid growing pressure on the US to start offering air support to beleaguered Afghan security forces – a move that could increase the number of civilian deaths further.
The UN and Nato will only say what they believe the combatant status of those killed to be, but the Bureau has dug further and traced the events of the day. We conducted telephone interviews in Pashto with nomad elders and Afghan officials, and gathered photos of the wreckage, bodies and graves, as well as what were said to be the ID cards of the victims.
We discovered three distinct narratives about what happened on June 5.
One version has the strike attacking a Taliban commander’s funeral. Another has it hitting innocent villagers aged between 18 and 73 as they headed back for lunch after digging a local man’s grave. And in the third version, the strike hit insurgents who had attacked a border police checkpoint.
Much remains murky about that day, but one thing is clear – at least two of these narratives must be wrong.
The strike affected one of Afghanistan’s sub-groups in particular – the Kochis. The Kochis are Pashto-speaking nomads with good political connections. They are spread out across the country, and sometimes but don’t always have the surname Kochai.
The June 5 strike hit one of the places where the Kochi graze their animals during the winter months. In the summer they move to cooler provinces such as Logar and Paktia, but they also leave a few people behind in places such as the village of Bati Tana to guard the houses.
Gul Marjan Farooqzoi is one of the Kochis’ community leaders. He was in Logar on June 5. He told us he was performing Friday prayers when the strike happened. When he came out of the mosque, he saw he had a number of missed calls on his phone. It started ringing again. On the other end of the line was a Kochi villager from the Bati Tana area.
Farooqzoi recalls the villager being furious, shouting “shame on you!” and complaining that he had taken so long to answer his phone when there were so many dead to deal with. The villager asked him to arrange coffins and to bring other people to help.
Farooqzoi then spoke to Meer Zaman, a Kochi from a town near Bati Tana, and told him to head to the site immediately. Farooqzoi himself would not get there for another two or three hours.
Zaman obeyed. He told the Bureau the bodies were still in the cars when he arrived. None of the villagers were said to have a camera-phone, and Zaman was reluctant to take pictures himself because it seemed disrespectful while people were crying around him.
Later, he spoke to Farooqzoi again, who stressed it was vital to document the attack. Zaman then sent Farooqzoi photos via Viber and Whatsapp showing the bodies bound up in cloth – photos we have obtained.
When Farooqzoi arrived at Bati Tana, he went to inspect the bodies himself. “Two or three people could not be recognised at all,” he said. “Most bodies were badly burned.”
As Farooqzoi surveyed the scene, an account of why those people were killed was taking shape in the local media.
This narrative first cropped up in an interview given to Voice of America by Khost’s deputy governor, Abdul Wahid Patan, on the day of the attack.
In that interview Patan said the pickup trucks were part of a funeral procession for Taliban commander Ameen Kochai, who was supposed to have been killed the previous day.
Separately, Pajhwok, an Afghan news agency, reported the next day that another Taliban commander, Bahram Kochai, had been in the funeral procession when it was struck.
But when we spoke to deputy governor Patan several months later, he seemed less sure about what had happened.
He said: “They (the villagers) said someone had died in Pakistan and they brought him here for burial. On the other hand the government sources said it was a Taliban’s funeral, but we could not prove either of these claims.”
He explained that when he spoke to the media on June 5, he had been briefed by the local branch of the National Directorate of Security (NDS). He said the NDS coordinates with the American military base in Bagram.
There is one significant problem with this version of events: Ameen Kochai, the Taliban commander whose funeral it was supposed to be, is still alive.
The Bureau found video of a man identifying himself as Ameen Kochai saying he had been arrested by the NDS. The NDS confirmed they had him in custody.
The idea that a funeral was supposed to take place that day was repeated by a number of sources. However, they were talking about a different funeral entirely.
According to residents of the area, the funeral was for an ordinary man from the village, who lived on the other side of the Pakistan border. His real name was Meer, but everyone knew him as Jaakha.
This narrative was put forward in the Afghan media by Haidar Jan Naeemzoy, a Kochi MP in the Afghan parliament.
The Bureau heard the same story – that the funeral was for Jaakha, not a Taliban commander – from three people from the Bati Tana area, all of them members of the tight-knit Kochi community. None were there at the time the strike hit.
Ramazan Kochai, a 47-year-old community leader who acts as a mediator in some of the Kochis’ land disputes, arrived the day after the strike. He told us the strike hit people gathered for the funeral ceremony – all of them men.
He explained that the dead came mainly from two families – a middle-aged man called Gulab Shah and his descendants, and two branches of the Khan family.
“In the house of Haji Gulab Shah there are only six people left – most of them children under seven years old,” he added. He said the Khan families had been left with only one breadwinner. “In these families there is not even a person left who would go to the market and buy tea.”
And with rising passion, he added: “I swear what I tell you is the truth, and if you could prove it wrong slit my throat.”
Gul Marjan Farooqzoi, the community leader who had arrived on the scene a few hours after the strike, said he knew Jaakha personally and provided a photo of him.
We spoke to Farooqzoi on a number of occasions to clarify and substantiate his account of the day.
He told us the day’s events began when Jaakha’s relatives called people in the Bati Tana area to say they would be bringing his body over for burial.
The villagers started digging the grave in a cemetery a short drive away, Farooqzoi said, and by noon they were finished. Because the body had still not arrived by then, they went home for lunch and Friday prayers. The Shah and Khan families drove back to their village, while others walked. The strike happened shortly after the two vehicles set off, near a checkpoint.
The Bureau spoke to a relative of Jaakha’s, who was in neighbouring Paktia province at the time of the strike. He confirmed that Jaakha’s brothers were driving the body over to Bati Tana when it occurred.
Though Farooqzoi could not make out the faces of some of the victims because they were so badly burned, he says he went to the families afterwards and photographed the victims’ IDs, which he sent to the Bureau.
Another villager gave us the names of all 14 victims. He said one of them was a Pakistani citizen named Fazil whose body was sent back across the border. The 13 Afghan names he provided matched the ID cards we were given.
There is nothing about the ID cards themselves that either suggests or rules out the idea the men were insurgents.
There is an inconsistency in the villagers’ version of events: one said that Jaakha was already buried when the strike happened, while the others insisted his body had not arrived at the graveyard by then.
Khost’s deputy governor, Abdul Wahid Patan, said that when he investigated the strike, he too was puzzled about where Jaakha’s body had been.
Farooqzoi told the Bureau that Jaakha was buried in a different part of the graveyard after the strike.
He provided photos of 13 freshly dug plots, and then later sent over a photo of what he said was Jaakha’s coffin by the side of his grave.
A third version of events has also been suggested. A well-connected and independent source in Khost province, who asked to remain anonymous because of the sensitivity of the subject matter, said the strike followed an attack on the border police checkpoint.
Taliban commander Bahram Kochai was killed in the strike, the source said, and three of the 14 killed were from Waziristan.
Mohammad Yaqub Mandozai, the provincial police security director, told the Turkish news agency Anadolu on the day of the strike that it hit as Taliban fighters were running away “after fighting between frontier police forces and the insurgents who had crossed the border”.
When we contacted Mandozai several months later to ask what he knew about the checkpoint attack, he said he didn’t want to comment because he had been transferred to a new role.
Neither were we unable to confirm that Bahram Kochai had been killed. One villager said he thought he had been arrested in Pakistan. Khost’s police chief would not give us any information on him, and the local governor’s office didn’t even know who he was.
There may have been one witness to the border checkpoint attack – a US drone. Nato told us they were able to disprove the civilian casualty claim by “pre-strike and strike footage”. But if drone footage did capture a checkpoint attack, Nato would not say so publicly. When asked whether such an attack had occurred, a Nato spokesman said operational security prevented him from giving any details other than it was a counter-terrorism strike.
In their own ways, all three stories are problematic. There is little evidence to support either narrative in which the dead were militants. Meanwhile, the version in which they were civilians rests largely on the testimony of a handful of people from the Bati Tana area, all of them from the Kochi community.
We were prevented from travelling to the scene of the strike by the security situation, a problem shared by Afghan investigators. Also, our interviews with people from the area were carried out over the phone. It is possible that Kochi elders leant on them to support a version of events in which the dead were nothing to do with the Taliban.
There is a noticeable gap in the villagers’ documentation of June 5 – photographs of the immediate aftermath of the strike. The explanation given is that the only person with a camera phone to get to the scene before the bodies were carried off the road was afraid of offending people by taking pictures. But it is also possible that the scene was cleared of potentially incriminating material before it was photographed.
Nonetheless, the Bati Tana residents’ version of events was remarkably detailed and consistent over multiple interviews, and Kochi elders do not claim civilian deaths every time there is a drone strike in eastern Khost.
The mystery of the differing assessments from Nato and the UN could even come down to different definitions of what it means to be a civilian. The US will not disclose how it defines this, but the UN sticks closely to definitions set out in international humanitarian law.
According to the International Committee of the Red Cross’s interpretation of the law, only those “whose continuous function it is to take a direct part in hostilities” constitute members of an armed group during non-international conflict.
This could theoretically mean that villagers who had links to the Taliban or even those who had fought for them on occasion would still be entitled to civilian status at the time of the attack.
Whatever the truth, what no-one disputes is that the US killed 14 people, and that it has failed to persuade either the Afghan government or the UN they were militants.
In the first half of 2015, Nato investigated 21 incidents in which civilian casualties were alleged to have occurred.
Seventeen were disproved or discounted, while three were passed on for national investigation by US authorities.
The Khost strike, meanwhile, involved a joint investigation between Nato, which thought the people killed were insurgents, and the government of Afghanistan, which didn’t, and was simply designated “disputed”.
Resolute Support, as the Nato mission is known, would not give details on how they concluded the Khost dead were insurgents, citing operational security. It would not even say what kind of aircraft was used, though UN data from the time suggests it was likely a drone.
Nato did confirm it was a counter-terrorism strike, and that they considered “all available evidence, including…interviews” when investigating the claim there were civilian casualties.
The United Nations Assistance Mission in Afghanistan (Unama), which requires three different types of source to declare a civilian casualty, would not specify how it had decided that the 14 were civilians.
However, the Afghan government’s lead investigator Dilbar Jan Arman did give the Bureau his reasons for thinking so.
He said Bahram Kochai was not dead, undermining the initial claim that the Taliban commander was in the funeral procession that was struck.
He also explained that the victims came mainly from two families, and it would be unusual for so many members of the same family to be fighters, presumably because it would leave dependents vulnerable in the event of their deaths.
And he pointed out that no weapons were found at the scene.
But he would not comment on the Americans’ divergent assessment of the situation. He said: “We sat together, we spoke about it, but they insisted on their position and we on ours.”
In terms of PR, 2015 was not a good year for the US in Afghanistan. Although the number of strikes plunged following the end of combat operations in 2014, the average number of UN-recorded civilian casualties per attack rose to its highest level since 2008.
The US may end up dramatically expanding airstrikes in Afghanistan in 2016. Afghan security forces have struggled to contain the resurgent Taliban since US troops withdrew, and there has been pressure on the Pentagon to offer air support. Former US General David Petraeus wrote an article in the Washington Post last month calling for America to “unleash” its airpower without expanding the troop presence.
But as the 2015 casualty numbers suggest, “light footprint” air wars can be messy. Even if the US is convinced it killed insurgents at Khost, the lack of persuasive, transparent and publicly available evidence to support this has made the strike perfect propaganda for the Taliban.
Back in Bati Tana, the burnt-out trucks have yet to be removed. They sprawl like carrion on the side of the dirt track, a mute reminder to occasional passers-by of the dangers of sudden violence from the skies.
This article was originally published on The Mystery of Khost: Did a US drone kill 14 innocent Afghans as they prepared a funeral?