Fighting over Kunduz

This is the third in a new series of posts on military violence against hospitals and medical personnel in conflict zones. It examines some of the key issues arising from the US attack on the Trauma Centre run by Médecins Sans Frontières (MSF) in Kunduz on 3 October 2015; it follows directly from my detailed analysis of the attack here and prepares the ground for a still more detailed analysis of attacks on hospitals, doctors and casualties in Syria to follow.

There are at least four main issues arising from the US attack on the MSF Trauma Centre in Kunduz that spiral out into a wider argument about what I will later call ‘The Death of the Clinic’.  I’m treating ‘the clinic’ here as a topological figure that extends from the body of the wounded through the evacuation chain to the hospital itself.  The clinic has been accorded a privileged status within the space of exception that is the modern conflict zone – a complicated, fractured space in which killing is made permissible subject to the protocols of international humanitarian law –  so that the clinic becomes an exception to the exception and its inhabitants granted a conditional immunity from attack.

msf-trauma-centre-burning

It’s important to understand that this legal armature is not immutable, and that changes (and challenges) to it arise through both (geo)political and military actions; international humanitarian law is not a deus ex machina, somehow above the fray, but is thoroughly entangled with the prosecution of military violence.  More on this to come, but for now it will be enough to list some of the major protections accorded to the clinic in war-time.

The first Geneva Convention (1864) (‘the Red Cross Convention’):

Ambulances and military hospitals shall be acknowledged to be neuter, and, as such, shall be protected and respected by belligerents so long as any sick or wounded may be therein.  Such neutrality shall cease if the ambulances or hospitals should be held by a military force … A distinctive and uniform flag shall be adopted for hospitals, ambulances and evacuations.

Under the Hague Regulations (1899/1907) that were in force during the hospital raids in France at the end of the First World War:

… all necessary steps must be taken to spare, as far as possible, … hospitals, and places where the sick and wounded are collected, provided they are not being used at the time for military purposes. It is the duty of the besieged to indicate the presence of such buildings or places by distinctive and visible signs, which shall be notified to the enemy beforehand.

Geneva Conventions 1949 care of woundedUnder the Geneva Conventions (1949) – whose provisions applied to the attack on the MSF Trauma Centre a hundred years later – there is a similar immunity granted to the military-medical machine:

The protection to which fixed establishments and mobile medical units of the Medical Service are entitled shall not cease unless they are used to commit, outside their humanitarian duties, acts harmful to the enemy. Protection may, however, cease only after a due warning has been given, naming, in all appropriate cases, a reasonable time limit and after such warning has remained unheeded.

And this is explicitly extended beyond the military-medical machine to institutions like the MSF Trauma Centre:

Civilian hospitals organized to give care to the wounded and sick, the infirm and maternity cases, may in no circumstances be the object of attack but shall at all times be respected and protected by the Parties to the conflict.

The protection to which civilian hospitals are entitled shall not cease unless they are used to commit, outside their humanitarian duties, acts harmful to the enemy. Protection may, however, cease only after due warning has been given, naming, in all appropriate cases, a reasonable time limit and after such warning has remained unheeded.

In so doing the treatment of hostile combatants is also explicitly provided for and protected:

The fact that sick or wounded members of the armed forces are nursed in these hospitals, or the presence of small arms and ammunition taken from such combatants and not yet been handed to the proper service, shall not be considered to be acts harmful to the enemy.

The language and specifications change, but there is nevertheless a consistent thread running through these provisions.  It has been stretched – and perhaps broken – by the attack on the MSF Trauma Centre, and here I’ll focus on four issues that have proved contentious.  First, the visual identification of the Trauma Centre; second, the alleged breach of its conditional immunity; third, the construal of the attack as a war crime; and fourth, the putative rejection of medical neutrality altogether.

(1) Visual identification

International humanitarian law (IHL) requires those responsible for hospitals ‘to indicate their presence’ – the language varies – in order to ensure their protection, and here the US military investigation made this finding (all page numbers in brackets refer to the redacted report):

The center roof of the MSF Trauma Center was marked with two rectangular MSF flags… The front and sides of the MSF hospital were marked from the street view and a MSF flag flew in the courtyard.  The MSF Trauma Center was not marked with any internationally recognized symbols such as a red cross, red crescent or a red “H”.  If it had been marked with these symbols, it is possible the Trauma Center would not have been engaged. (082)

This counterfactual does not loom large in the report or its recommendations, but Charles Dunlap (at Lawfire) has seized upon it to berate MSF:

Ask yourself: wasn’t it a mistake for [MSF] – and a serious one – not to have marked its facility in accordance with Protocol III to the Geneva Conventions which designates “the only emblems recognized by nations signifying the protected status of individuals or objects bearing them during armed conflict”?  Had, for example, the hospital been marked with large Red Crosses/Red Crescents or one of the other internationally-recognized symbols (as the U.S. does) or something that would make its protected use clear from the air, isn’t it entirely plausible that the aircrew (or someone) might have recognized the error and stopped the attack before it began?

Put another way, isn’t it foreseeable that in an exceptionally chaotic combat situation (where a belligerent is making use of civilian buildings to conduct combat operations) that mistakes could occur in identifying a protected structure absent Protocol III markings or at least something to make it identifiable at a distance, especially when it’s known that attacking aircraft are being used?  Wouldn’t reasonably prudent persons have marked their medical facility with an internationally-recognized symbol or something of similar clarity to the warring parties?  Wouldn’t due care demand it in that situation?

In accusing MSF of ‘imprudence’ and even recklessness Dunlap applies a double standard.  He repeatedly insists that the US and the Afghan militaries confronted ‘an extraordinarily intense situation’ in Kunduz, that they faced ‘terrible urgency’ and ‘enormous pressure’ as they operated ‘in the turmoil of a war zone’ – all of which is undoubtedly true – but he uses this to excuse their mistakes while refusing to extend the same privilege to MSF.

_85907418_85907417

Let me remind you of Dr Kathleen Thomas‘s account of working in the ER (above) once the city had fallen to the Taliban:

The first day was chaos – more than 130 patients poured through our doors in only a few hours. Despite the heroic efforts of all the staff, we were completely overwhelmed. Most patients were civilians, but some were wounded combatants from both sides of the conflict. When I reflect on that day now, what I remember is the smell of blood that permeated through the emergency room, the touch of desperate people pulling at my clothes to get my attention begging me to help their injured loved ones, the wailing, despair and anguish of parents of yet another child lethally injured by a stray bullet whom we could not save, my own sense of panic as another and another and another patient was carried in and laid on the floor of the already packed emergency department, and all the while in the background the tut-tut-tut-tut of machine guns and the occasional large boom from explosions that sounded way too close for comfort.

In any case, MSF had clearly ‘indicated their presence’ to both the US and Afghan authorities by providing them with the GPS co-ordinates of the Trauma Centre (see my previous discussion here). Dunlap finds this ‘commendable’ but ‘legally problematic’.

Instead, he is fixated on the absence of a Red Cross flag from the roof, in which case he might reflect on another passage from the report.  On 2 October, the day before the air strike, MSF phoned the Special Operations Task Force in Bagram to develop a contingency plan: while the Taliban were respecting the neutrality of the Trauma Centre and ‘treating the government casualties well’, they wanted to know the feasibility of extracting their patients should conditions deteriorate.  During that conversation they were advised to ‘take the signs normally affixed to the sides of the trucks and to install them on the top of the vehicles for easy identification by aircraft during this or any future MSF resupply operations‘ (503; my emphasis).  This surely makes it clear that the US military anticipated no difficulty in recognising MSF’s flag and logo as symbols of medical neutrality.

msf153147

(2)  Conditional immunity

IHL makes it clear that treating wounded combatants does not compromise the protections afforded to a medical facility; that occurs only if it is used as a base from which ‘to commit, outside their humanitarian duties, acts harmful to [one of the belligerents]’.  I’ll address the intervening clause – ‘outside their humanitarian duties’ – under (4) and confine my discussion here to the alleged militarisation of the clinic.

msf153705-1

MSF’s internal review found that its unambiguous ‘no weapons‘ policy was adhered to:

All of the MSF staff reported that the no weapons policy was respected in the Trauma Centre. [Since the KTC opened, there were some rare exceptions when a patient was brought to the hospital in a critical condition and the gate was opened to allow the patient to be delivered to the emergency room without those transporting the patient being first searched. In each of these instances, the breach of the no weapon policy was rapidly rectified.]  In the week prior to the airstrikes, the ban of weapons inside the MSF hospital in Kunduz was strictly implemented and controlled at all times and all MSF staff positively reported in their debriefing on the Taliban and Afghan army compliance with the no-weapon policy.

The US military investigation accepted this was indeed the case:

Evidence provided to the investigation team supports the MSF internal initial report’s characterization that their no-weapons policy was adhered to with rare exceptions (038, note 15).

msf112743

Mathieu Aikins‘s interviewees also confirmed the absence of weapons from the Trauma Centre:

Though the MSF hospital was crowded with fighters, whether patients or caretakers (each patient was allowed one), staff members and civilians who were present said the insurgents respected the rules. They left their weapons outside or handed them over at the gun lockers at the entrance. One employee recalled seeing a fighter give up his weapon but forget his ammunition vest; when the employee nervously approached the fighter about it, the man apologized profusely and handed it over. “We had respect for the hospital, as they were serving the people,” said Shahid, the Taliban commander. “I myself went there once when one of our men was wounded, and before entering we submitted our weapons outside.”

Aikins goes on to report that patients were allowed to retain their cellphones, and some of their caretakers retained hand-held radios whose transmissions were intercepted by Afghan special forces.  They in turn concluded that not only were the Taliban inside the hospital but were using it as a base: ‘They had raised their flag and established their headquarters there.’  On 1 October, presumably in response to these reports, the Pentagon contacted MSF in New York to ask whether ‘they had a large number of Taliban “holed up”’ in the Trauma Centre, and were assured that the only Taliban inside the hospital were wounded patients.

But the suspicions clearly remained, and festered to such a degree that some of those on the ground were convinced that the hospital had been overrun by Taliban fighters.  Associated Press reported that the radio intercepts prompted US analysts to request ‘specific intelligence-gathering flights over the hospital’ – their outcome has never been disclosed – and on 1 October a senior Special Forces commander (whether in Kabul or in Kunduz is unclear) wrote in his daily log that the Trauma Centre was under Taliban control and that he planned to clear it in the coming days.   At least some of the Green Berets in Kunduz agreed with his assessment: ‘They were using it as a C2 node … They had already removed and ransomed the foreign doctors, and they had fired on partnered personnel from there.’  Indeed, after the attack a senior US officer in Kabul was told – by whom has been redacted – that ‘there were three dead Military-Aged Males near the hospital, identified as Taliban by the local population.  They were using the hospital as a command post (using its protected status)’ (275).

observations-from-ac-130-001

But all of this was fantasy, and the investigation discounted it.  Although US intelligence reported that insurgents were present at the hospital at the time of the strike, the investigation accepted that this was for medical treatment and they could trace ‘no specific intelligence reports that confirm[ed] insurgents were using the MSF Trauma Center as an operational  C2 [command and control] node, weapons cache or base of operations’ (085).  In addition, they determined that observations made from the AC-130 revealed neither substantive hostile acts nor demonstrations of hostile intent –  only ‘unarmed individuals walking around [or] sitting in chairs’ (085).  The report describes these as ‘ordinary and innocuous acts’ (055), but to at least one member of the aircrew that was in itself grounds for suspicion: ‘In his experience, when AC-130 aircraft fly over insurgents, they act normally or try to stay normal… [whereas] civilians will not try to be nonchalant when the aircraft is overhead’ (093, note 304).  Damned if you do, and damned if you don’t: when everything is construed as hostile, even the most innocent acts are transformed into somcething sinister.

The claims made by Afghan forces were even wilder.  Here is May Jeong in The Intercept:

On the night of the hospital strike, a unit commander with the Ministry of Defense special forces was at the police headquarters taking fire from the direction of the hospital. “Vehicles were coming out of there, engaging, then retreating,” he told me. When I pointed out that he couldn’t have seen the gate of the hospital from where he was, several hundred meters away, he said that he was sure because he had personally interrogated a cleaner who told him that the hospital was full of “armed men using it as a cover.” The cleaner told the commander that there were Pakistani generals using the hospital as a recollection point and that they had set up a war room there. When I challenged his line of vision again, he responded, “Anyone can claim anything. The truth is different.”

afghanistan_kunduz_3459341b

[Amrullah] Saleh, [former head of the National Security Directorate and] the author of the 200-page Afghan commission report on the fall of Kunduz … believed that the “hospital sanctity had been violated” and held out as evidence 130 hours of recorded conversations with more than 600 interlocutors. “I spoke with the MSF country director,” Saleh told me recently. “They don’t deny that the hospital was infiltrated by the Taliban.”

But of course they did deny it: repeatedly, emphatically and convincingly.

(3) War crimes?

The US military investigation was unequivocal: it found multiple violations of the military’s own Rules of Engagement and of international humanitarian law.

The first rule of customary international humanitarian law, now codified in the Additional Protocols to the Geneva Conventions, is distinction:

The parties to the conflict must at all times distinguish between civilians and combatants. Attacks may only be directed against combatants. Attacks must not be directed against civilians.

The investigation found that both the Ground Force Commander (GFC) and the aircraft commander failed to exercise this core principle:

Neither commander distinguished between combatants and civilians nor a military objective and protected property. Each commander had a duty to know, and available resources to know that the targeted compound was protected property’ (075-6).

A second core principle is proportionality:

Launching an attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated, is prohibited.

The investigation found this to have been disregarded too:

The GFC and the aircraft commander failed to exercise the principle of proportionality in relation to the direct military advantage (076).

Both principles are deceptively simple, and in ‘The Passions of Protection: Sovereign Authority and Humanitarian War’ Anne Orford reminds us that IHL ‘immerses its addressees in a world of military calculations.’  In practical terms the distinction between civilians and combatants in today’s conflicts is rarely straightforward, but in this case the No-Strike List plainly recognised the protected status of the Trauma Centre and there is no convincing evidence that its immunity had been compromised.  In addition, the balance between loss of civilian life and military advantage is weighed on the military’s own scales – ‘expected’; ‘excessive’; ‘anticipated’: these are not self-evident calculations – but even if the GFC or the aircraft commander had grounds to believe the Taliban were firing from the hospital the Pentagon’s own Law of War Manual (which is not without its own controversies: see here and, specifically on proportionality, here and here) advises under §7.10.3.2 that

The obligation to refrain from use of force against a medical unit acting in violation of its mission and protected status without due warning does not prohibit the exercise of the right of self-defense. There may be cases in which, in the exercise of the right of self-defense, a warning is not “due” or a reasonable time limit is not appropriate. For example, forces receiving heavy fire from a hospital may exercise their right of self-defense and return fire.  Such use of force in self-defense against medical units or facilities must be proportionate.

Not only was there was no evidence of hostile let alone ‘heavy fire’ from the Trauma Centre but the AC-130 was also monitoring the progress of the Afghan Special Forces convoy that it was tasked with protecting and knew perfectly well that it was still within the perimeter of the airfield.  This was not a time-sensitive target (the report makes that crystal clear) and neither the GFC nor the aircraft commander had reason to believe that any putative threat to Afghan or US forces was so grave and so sustained that it called for an air strike involving multiple passes by the AC-130 – over 30 minutes according to the US military, an hour according to MSF – delivering such intense fires that the building was virtually destroyed.

For these reasons many commentators – and MSF (‘Under the clear presumption that a war crime has been committed, MSF demands that a full and transparent investigation into the event be conducted by an independent international body’) – have insisted that the attack was a war crime.

not-a-war-crime-3-jpeg

not-a-war-crime-2-jpeg

But others (including the US military) have concluded that it was not.  US Central Command’s initial summary – produced before the redacted report was released – accepted that there had been breaches of both the Rules of Engagement and of IHL (‘the law of armed conflict’) but noted that

the investigation did not conclude that these failures amounted to a war crime.  The label “war crimes” is typically reserved for intentional acts – intentionally targeting of civilians or intentionally targeting protected objects.  The investigation found that the tragic incident resulted from a combination of unintentional human errors, process errors and equipment failures, and that none of the personnel knew that they were striking a medical facility.

The report has been so heavily redacted so that this legal discussion is unavailable (see also the commentary by Sarah Knuckey and two of her students here).  We do know that the investigation team included an unnamed legal advisor from US Central Command (CENTCOM) and that its report was subject to legal review by the Staff Judge Advocate, who accepted its findings as ‘legally sufficient’ with several, redacted exceptions  – though there is no way of knowing what they were (007-009).  We know too that General John Campbell, who ordered the investigation as commander of US Forces in Afghanistan, subsequently disapproved a number of findings and recommendations ‘not related to the proximate cause of the strike’ (002) but, again, the details have been excised.

General Votel at Pentagon press briefing on MSF attack

General Joseph Votel, commander of CENTCOM, repeated the summary statement’s disavowal of war crimes at a Pentagon Press Briefing on 29 April 2016, and in responding to a storm of questions from plainly incredulous reporters (above) he elaborated:

… an unintentional action takes it out of the realm of actually being a deliberate war crime against persons or protected locations…. They were absolutely trying to do the right thing; they were trying to support our Afghan partners; there was no intention on any of their parts to take a short cut, or to violate any rules that were laid out for them. And they were attempting to do the right thing.  Unfortunately, they made a wrong judgment in this particular case…

Jens David Ohlin explains the disputation (which Faye Donnelly helpfully re-casts as one between two contending narratives whose speech-acts struggle to realize their performative force):

The problem is that the killing of the innocent civilians was not intentional, it was accidental. As a matter of criminal law, it was either reckless or negligent … but the civilian killings were not performed with purpose.

The Rome Statute of the International Criminal Court provides for war crimes prosecutions for ‘intentionally directing’ or ‘intentionally launching’ attacks that contravene international humanitarian law (in effect, criminalizing the rules of IHL).  Jens discusses this in relation to attacks on civilians, but the Statute also proscribes ‘intentionally directing attacks against buildings, material, medical units and personnel’ or against ‘personnel, installations, material, units or vehicles involved in a humanitarian assistance or peacekeeping mission’.

In every case the emphasis is on intentionality, and yet intentionality – as philosophers have demonstrated time and time again – is not the simple, settled matter some legal scholars assume it to be.  Jens’s central point is that common-law cultures identify intentionality with purpose or knowledge whereas civil-law cultures widen its sphere to include a conscious disregard of risk or ‘recklessness’.  The full argument is here – including an intricate disection of the (geo)politics involved in drafting the Geneva Conventions and the Additional Protocols – but the sharp conclusion is that (for Jens, at least) the strike on the Trauma Centre would not constitute a war crime under the first count (he accepts that neither the GFC nor the aircraft commander possessed the knowledge or the purpose) but could under the second (their actions, and those of others, were reckless).  I should add that he recommends the recognition of a new war crime to explicitly address the second count and thereby signal ‘the moral difference between intentionally killing civilians and recklessly killing them.’

kunduz-msf-a-tragedy-of-errors-001

 

The investigation report provides endless, explicit examples of a thoroughly compromised ‘risk management process’ by multiple actors at multiple sites, and this dispersal of responsibility in Kunduz (see map above) and Bagram further complicates the legal situation.  Peter Margulies – who does not accept that ‘the lack of intent among US personnel is determinative’ – concedes that ‘the cascading systemic errors in the hospital attack impede the attribution of culpable awareness to one or more specific individuals.’ In his view,

CENTCOM would have been better served by acknowledging that intent was not required [for the commitment of a war crime], but that awareness of risk was distributed among many organizational components, without full awareness concentrated in one or more individuals who could be charged criminally.

Adil Ahmad Haque notes that Additional Protocol I to the Geneva Conventions requires attackers to do ‘everything feasible’ to verify that their target is a military objective and instructs them in cases of doubt to presume that it is civilian – the Law of War Manual doesn’t follow this standard, but the investigation report does –  and here there is such clear evidence of recklessness on the part of many of the US forces involved (whose evidence is shot through with technical failures and radical uncertainty) that, in his view, their decision to press on with the attack ‘was unlawful, irrespective of their good faith.’

(4) Medical neutrality at risk

I noted above that hospitals only lose their protected status if they are used ‘to commit, outside their humanitarian duties, acts harmful to [one of the belligerents]’.  It’s a telling provision because its intermediate clause can be read as a tacit acknowledgement that those humanitarian duties – treating the sick and wounded – could otherwise be construed as acts harmful to their enemies.

And there is evidence that this is exactly how both the Afghan government and its military viewed MSF’s activities.  When Mathieu Aikins visited Kunduz after the air strike he reported:

Some members of the Afghan government and security forces there had little respect for MSF’s neutrality and resented its treatment of wounded Taliban. When I visited Kunduz in November, their anger was still surprisingly raw, despite the recent destruction of the hospital. “They give them medicine; they transport and treat their injured,” [Colonel Abdullah] Gard, the commander of the [Ministry of Interior’s] quick-reaction force, told me. “Their existence is a big problem for us…. The people that work there are traitors, all of them.”

abdullah-gard

Gard (seen above) and one of his colleagues told May Jeong exactly the same:

Gard spoke of MSF with the personal hatred reserved for the truly perfidious. He accused the group of “patching up fighters and sending them back out,” a line I heard repeatedly. Cmdr. Abdul Wahab, head of the unit that guarded the provincial chief of police compound, told me he could not understand why in battle an insurgent could be killed, but the minute he was injured, he would be taken to a hospital and given protective status. Wouldn’t it be easier, he asked, wouldn’t the war be less protracted or bloody if they were allowed to march in and take men when they were most compromised? He had visited the MSF hospital three times to complain. Each time a foreign doctor explained the hospital’s neutral status and its no-weapons policy, which mystified him.

In short, it seems that some (perhaps many) in the Afghan security forces – particularly after the humiliation of being forced out of Kunduz – believed that the Taliban were legitimate targets wherever they were and that the fight against them was being hamstrung by what one officer described to Jeong as a ‘silly rule’.

kunduz-a-silly-rule-001

This becomes material because, as I showed previously, the intended target for the air strike was a National Directorate of Security compound whose co-ordinates had been passed to Major Michael Hutchinson, the Ground Force Commander, by Afghan Special Security Forces (who, like him, were unfamiliar with the city): they had identified the NDS compound as a Taliban command and control node, and planned to clear it on their return from the airfield following a casualty evacuation.  The AC-130 was supposed to provide Close Air Support, but a series of technical difficulties compromised the accuracy of its sensors and several hours after the Afghan Special Security Forces had left in an armoured convoy with their three casualties the aircrew were still unsure of the location of the target and so requested a verbal description of the NDS compound.  This was provided by Afghan forces still inside the Provincial Chief of Police compound: their description matched the MSF Trauma Centre much more closely than the NDS compound, and the aircrew fixed on this as their target.

doctors-with-enemies-nyt

It’s impossible to know whether this was a mistake or misdirection, and the report fails to identify who provided the description.  Hutchinson was shown various photographs but explained that ‘it was dark when everything happened’; he couldn’t remember the name of the Afghan liaison officer and – the redactions make his responses difficult to follow at this point – he wasn’t always sure who he was talking to since he had to rely on interpreters (387-8).  But the description obviously had to come from someone who knew the city (which would include Gard and Wahab), and both Aikins and Jeong clearly believe that misdirection is not only possible but also extremely likely.  ‘That hospital is in the service of the Taliban,’ Gard told Aikins. ‘I swear to God, if they make it a hundred times, we’ll destroy it a hundred times.’  Hence the headline for Aikins’s searching New York Times report (above): ‘Doctors with enemies: did Afghan forces target the MSF hospital?‘ I should note that David Glazier dismisses all this as ‘highly speculative’ and insists that ‘it simply defies logic’.  While the claim is speculative it surely doesn’t defy logic, and Aikins and Jeong make at least a plausible if not definitive case).  They are clearly not alone in their suspicions: MSF’s very first question in response to the investigation report was this:

‘What was the physical description of the intended target provided by the Afghan forces and how did it match the description of the MSF hospital?’

gard-that-hospital-is-in-the-service-of-the-taliban-001

The studied refusal to recognise medical neutrality – if that is what this was – emerges from a long history of friction between MSF and the government in Kabul, and it threads its way out into a wider history and geography of deliberate attacks against medical facilities elsewhere in Afghanistan, and – among other recent places – in Gaza (by the Israeli military), in Syria (primarily by the Russian and Syrian Arab Air Forces) and in Yemen (primarily by the Saudi-led coalition which is advised by the United Kingdom on targeting).  In February 2014 Thanassis Cambanis was already writing that ‘medical care is now a tool of war’, and in a report issued in May 2016 the World Health Organisation counted almost 600 attacks against medical facilities, doctors and nurses in 19 countries since then.  These shocking statistics, which are inevitably imperfect, include attacks by non-state actors, among them the Taliban and Islamic State, but I have emphasised the complicity of state actors – including leading members of the UN Security Council – because it is their actions that determine the course of international humanitarian law and because they are ultimately responsible for what MSF’s president Joanne Liu saw in the wake of the attack on its Trauma Centre in Kunduz as  ‘not just an attack on our hospital’ but ‘an attack on the Geneva Conventions.’

msf-kunduz-battlefields-without-doctors

In a letter to the UN Security Council issued on 22 September 2016, after a direct and deliberate attack on a humanitarian convoy delivering aid to eastern Aleppo, more than 100 humanitarian organisations noted that:

‘Each [head of state] that accepts a lack of accountability for perpetrators and facilitators of war crimes colludes in the ongoing dissolution of international humanitarian law.’

That dissolution can be seen as a defiant reassertion of the absolutism of sovereign power, because ‘medical neutrality’ is not a neutral claim.  Adia Benton and Sa’ed Ashtan persuasively argue that:

The health worker’s claim to impartiality may itself be a stance against the state’s insistence that it is the sole arbiter of who can live and who can die. The local health worker’s claim to an international norm … may be understood as a direct challenge to the state’s claim to sovereignty…. Can ‘‘medical neutrality’’ accurately describe a situation where there is no neutral ground upon which to stand? (‘‘‘Even War has Rules’’: On Medical Neutrality and Legitimate Non-violence’ , Cult. Med. Psychiatry 40 (2016) 151-158).

As the spectral presence of Giorgio Agamben in my slide below implies, this has the liveliest implications for how we are to understand the space of exception and the refusal of its victims to be reduced to the passivity of ‘bare life’.

medical-neutrality-sovereign-power-and-biopolitics-001

As I will show in the next post in this series, Syria is the most egregious contemporary case.  If some members of the Afghan security forces wilfully misled the Americans into targeting the MSF Trauma Centre in Kunduz, then it seems clear that they objected to the protection extended by IHL to wounded Taliban combatants and those who treat them (though many of those killed in the attack were civilians, despite Alan Dershowitz‘s tawdry attempt to suggest that MSF ‘favoured Taliban fighters over civilian patients’: see Kevin Jon Heller’s magisterial response here).  But in Syria not only has the provision of medical aid to those in rebel-held areas been explicitly criminalised by the state’s new Counterterrorism Law which came into effect in July 2012 – the space of exception is far from being a ‘legal “black hole” – but the ban extends to those providing medical aid to sick or wounded civilians.  And make no mistake: there have been no mistakes.  The murder of doctors and nurses and the bombing of hospitals and clinics in Syria has been deliberate and systematic.  The exception to the exception contracts to its vanishing point.

To be continued.

 

Killing over Kunduz

This is the second in a new series of posts on military violence against hospitals and medical personnel in conflict zones.  It examines the US attack on the Trauma Centre run by Médecins Sans Frontières (MSF) in Kunduz on 3 October 2015.  I provided preliminary discussions here (on the conduct of US military investigations into civilian casualty incidents), here (on MSF’s own investigation into the attack), here (on the Executive Summary of the US military investigation), here (on two first-hand accounts from MSF personnel), here and here on the final report, and here (on the likelihood that the attack constituted a war crime).   This post draws on those discussions but also on a close reading of the redacted report of the US military investigation [all page references refer to that report], on work by investigative journalists, and on ancillary materials and commentaries. 

spooky-mainOne year ago today, in the early hours of the morning of 3 October 2015, a US AC-130U gunship (‘Spooky’) launched a concentrated attack on the Trauma Centre in Kunduz run by Médecins Sans Frontières.  In an otherwise probing report on what happened,  the Washington Post claimed that the gunship has sensors ‘that give it a “God’s eye” of the battlefield’.  Here  I explore some of the multiple ways in which such a view was – and remains – impossible.  For militarized vision, like any other optical modality, is never a purely technical affair.  A series of cascading technical errors bedevilled the US attempt to re-take Kunduz from the Taliban, who had swept into the city a few days earlier, but these were compounded by a series of profoundly human decisions and interactions and it was the intimate entanglement of the technical and the human that determined the hideous outcome.

At least 42 people were killed, including 24 patients, 14 medical staff and 4 caretakers.  Many others were wounded and traumatized.  Here is Dr Evangeline Cua, a Philippina surgeon who was on duty when the attack started:

msf164464-profileWe were like two headless chickens running in total darkness — me and the surgeon who assisted me in an operation. The nurses who were with us a moment ago had run outside the building, braving the volley of gunshots coming from above. I was coughing, half-choked by dust swirling around the area. Behind my surgical mask, my mouth was gritty, as if somebody forced me to eat sand. I could hear my breath rasping in and out. Layers of smoke coming from a nearby room made it hard to see where we were. Blinking around, I caught sight of a glow, from a man’s hand holding a phone. He seemed mortally wounded but was still trying to send a message…perhaps to a loved one?

I stood transfixed, not knowing where to turn or what to do. All around us, bombing continued in regular intervals, shaking the ground, sending debris sweeping and flying. One. Two. Three. I tried to count but there seems to be no abatement to the explosions. I stopped counting at eight and silently prayed that we could get out of there alive.

Fire licked at the roof at one end of the building, dancing and sparkling in the dark, reaching towards the branches of the trees nearby. The ICU was burning.  Outside, only the constant humming from above pointed to the presence of something. An aircraft? Airstrike? Why the hospital? Why us? Then, without warning, another tremendous, ear splitting blast shook the building. The ceiling came crashing down on us and the last remaining lights were turned off, sending us to total darkness. I screamed in terror as wires pinned me to the ground. That was the last thing I could remember.

What follows is an attempt to answer those questions.  It is fraught with uncertainty: the most detailed investigation to date has been carried out by the US military, but the redacted version of the final report that has been released to the public is (by the standards of other US military investigations) profoundly unsatisfactory – redacted with a brutishly heavy hand.  Time and time again, ironically, references to the time of events have been removed; transcripts of radio communications and interviews by the investigating panel that have been released in other cases have been suppressed; and some redactions seem to have been made not for reasons of privacy or security but to avoid embarrassment (more here; you can download the report from US Central Command’s Freedom of Information Act (FOIA) library here).

All of this reinforces MSF’s original call for an independent investigation.  I understand May Jeong‘s pessimism:

A former Afghan special forces commander who was at the command and control center in Kunduz during the fight assured me I would never get to the bottom of the attack. The reason why I couldn’t figure out exactly what had happened, he said, was the fog of war. “Ground truth is impossible to know. Even those who were there wouldn’t be able to tell you what they saw.”

But when the ‘fog of war’ – so often a convenient cover for all manner of horrors – is deliberately thickened – when visibility is ruthlessly reduced by redaction – then perfectly proper public interest is trumped by political and military expediency.

***

When the NATO-led combat mission to Afghanistan conducted by the International Security Assistance Force (ISAF) finished at the end of 2014 it was replaced by a much smaller advisory mission, Resolute Support, which was ‘to provide further training, advice and assistance for the Afghan security forces and institutions’.  Resolute Support was authorized by a Status of Forces agreement between NATO and the Afghan government in Kabul.  Its central hub was Kabul/Bagram, with four ‘spokes’ formed by four other ‘Train Assist Advise’ Commands to support four Afghan National Army Corps outside the capital (more here and here):

resolute-support

US troops were the major contributor to Resolute Support, but they were also assigned to the United States’s continuing (‘concurrent and complementary’) counter-terrorism mission now designated as ‘Operation Freedom’s Sentinel’.  Until March 2016 both missions were under the overall command of General John Campbell.

By September 2015 the focus of US concern in Afghanistan was Helmand in the south – where the Taliban were on the ascendant, forcing Afghan government forces to retreat as they seized control of key districts and gained control of the Kajaki dam – and US Special Forces were rushed to Camp Bastion after the fall of Musa Qala gave the insurgents a strategic advantage.

By contrast, Kunduz in the north was regarded as ‘secure’ [135] after a series of combat operations at the start of the fighting season earlier in the year.  As late as 13 August Brigadier-General Wilson Shoffner, Deputy Chief of Staff for Communications with Resolute Support, declared that although there had been ‘an attempt by the Taliban to try to stretch the Afghan security forces in the north’ the city of Kunduz ‘is not now and has not been in danger of being overrun by the Taliban’ (he also described the situation at Kajaki as merely a ‘local security challenge’).  But those previous operations in Kunduz had targeted Taliban operations areas and did not extend to support zones outside the city.

isw-taliban-attack-and-supply-zones-in-kunduz

Obeid Ali reports that during the summer the Taliban continued to make inroads until they controlled areas to the south west, north west and south east of the city.

On 28 September 2015, the Taliban stormed various ANSF locations in Kunduz city from the three different directions they had spent so long preparing… The simultaneous attacks on the city and the collapse of check posts at the city ‘gateways’ destroyed the confidence of the ANSF inside the city in their ability to stand against this unexpected offensive. In the face of the well-organised and coordinated insurgent operation, most held out for only a few hours. A chaotic environment quickly spread and government officials, ALP [Afghan Local Police] commanders and some of the ANA [Afghan National Army] officers, fled to the military base at the airport [Camp Pamir], leaving Kunduz effectively leaderless.

taliban-has-captured-the-city-of-kunduz-late-sep-28-monday

Kunduz was a spectacular, strategic prize: the first city to fall to the mujaheddin in 1998 and the first time the Taliban had seized a major city since 2001, its capture signalled both a resilient Taliban and a faltering government footprint in the region.

•••

On 28 September there was a detachment of US Special Forces (‘Green Berets’) based at Kunduz airfield as part of the Train, Assist, Advise mission.  Like every Operational Detachment – Alpha (OD-A) it consisted of just 12 soldiers, all cross-trained and capable of operating for extended periods of time with little or no support.  On 29 September their superior command – the US Special Operations Task Force in Afghanistan – ordered two other OD-As to Kunduz.  While they were in the air the OD-A on the ground sketched out a contingency plan (‘Kunduz City Foothold Establishment’) to assist the Afghan forces to return to the city and secure the Kunduz City Hospital and the Prison.  There were repeated US airstrikes against Taliban positions in and around the city throughout the day, but by the time the OD-A reinforcements, together with other Afghan troops including Afghan Special Security Forces (below), arrived in the evening it was clear that the original plan was unworkable and their immediate priority had to be the defence of the airfield [032, 382].

afghan-special-forces-arrive-at-kunduz-airfield-29-september-2015

The US reinforcements included Major Michael Hutchinson, who assumed overall command of the combined OD-As (he was identified as the Ground Force Commander by the New York Times).  He had misgivings about the mission but accepted that ‘we can’t lose the provincial capital’ [377].   The next day a revised plan (‘Kunduz Clearing Patrol’) was submitted to the Special Operations Task Force for approval, which was granted that night, and the OD-As requested that Afghan Special Security Forces be accorded ‘designated special status’ that would permit the Green Berets to extend their own envelope of self-defence and assume a direct combat role (including calling in air strikes) to defend their partner forces if they came under attack [046-7].

By this time Médecins Sans Frontières had been in contact with both US and Afghan forces to ensure that they were aware of the location and status of its Trauma Centre in Kunduz.  It was in the eye of the storm.  Dr Kathleen Thomas, an Australian doctor in charge of the Emergency Room and the Intensive Care Unit, explained:

We all knew that at times, our hospital was in the middle of the rapidly changing front line – we could feel it. When the fighting was close – the shooting and explosions vibrated the walls. I was scared – we were all scared. When a loud “BOOM” would sound a bit closer to the hospital, we would all drop to the floor away from the large windows that lined the ICU walls. We also tried to move the patients and large (flammable) oxygen bottles away all from the windows, but the layout of the ICU prohibited doing this effectively. I worried constantly about the exposure from those windows – yet never thought to worry about the exposure from the roof.

kunduz_2

Most of the patients were civilians.  Of the combatants, MSF reported that most of them were from the Afghan army and police, as had been the case since the Trauma Centre opened, but once the city fell on 28 September ‘this shifted to primarily wounded Taliban combatants.’  The Afghan government speedily arranged the transfer of all its patients (apart from the most severely wounded cases) to another hospital.

By that night the Taliban announced that it was in control of the district.  Kathleen Thomas described the scene:

The first day was chaos – more than 130 patients poured through our doors in only a few hours. Despite the heroic efforts of all the staff, we were completely overwhelmed. Most patients were civilians, but some were wounded combatants from both sides of the conflict. When I reflect on that day now, what I remember is the smell of blood that permeated through the emergency room, the touch of desperate people pulling at my clothes to get my attention begging me to help their injured loved ones, the wailing, despair and anguish of parents of yet another child lethally injured by a stray bullet whom we could not save, my own sense of panic as another and another and another patient was carried in and laid on the floor of the already packed emergency department, and all the while in the background the tut-tut-tut-tut of machine guns and the occasional large boom from explosions that sounded way too close for comfort.

Although the Trauma Centre had been on US Central Command’s ‘No-Strike List’ since October 2014 MSF now re-supplied its GPS coordinates and reminded the Ministry of Defence in Kabul that ‘MSF and its personnel observes strict neutrality and impartiality in the name of universal medical ethics and rights of populations affected by conflicts to humanitarian assistance’ and claimed ‘full respect of these principles and rules in order to be able to continue responding to the humanitarian and medical needs of all Afghans’ [144].  On 29 September MSF issued what would prove to be a remarkably optimistic statement:

We are in contact with all parties to the conflict and have received assurances that our medical personnel, patients, hospital and ambulances will be respected.  With the government provincial hospital not currently functioning, MSF’s hospital is now the only place in Kunduz where people in need of urgent trauma care can receive it.

msf153705-1

MSF had withdrawn from Afghanistan in August 2004 – after the targeted killing of five of its aid workers in June, the government’s failure to arrest those responsible, and Taliban threats to target organizations like MSF that they falsely claimed ‘work for US interests’ – and returned five years later with agreements from the US-led coalition, the Afghan government and the Taliban to respect the de-militarization of its hospitals (including a strict ‘no-weapons’ policy inside them). Initially MSF assumed responsibility for two public hospitals in Kabul and Helmand; two years later it opened its Trauma Centre in Kunduz inside the old Spinzer cotton factory.  It soon became immensely important:

operative-volume-at-msf-trauma-centre-kunduz-2011-2015-trelles

[Source: Miguel Trelles, Barclay T Stewart and others, ‘Averted health burden over 4 years at Médecins Sans Frontières (MSF) Trauma Centre in Kunduz, Afghanistan, prior to its closure in 2015’, Surgery (2016) in press]

Between August 2011 and August 2015 the Trauma Centre cared for 6,685 patients; roughly one-third were suffering from ‘violence-related trauma’, which included land mines and bomb blasts, gunshots, stabbings, assaults, rape and torture; one quarter of those were children.  Procedures for complex wounds were the most common – debridement (excision), removed of shrapnel, care of burns – followed by orthopaedic procedures (including amputation).  Those injuries increased dramatically in the months before the city fell to the Taliban.  Miguel Trelles and his collaborators estimate that during this period the Trauma Centre averted 154, 254 ‘Disability Adjusted Life Years’; more prosaically:

The MSF Kunduz Trauma Centre provided surgical care for a large number of wounded and injured patients in the region. The surgical epidemiology is consistent with reports from other areas of prolonged insecurity in that unintentional, traumatic, non–war-related injuries generally outnumber those from violence. Nevertheless … the Trauma Centre provided surgical care for many adults and children injured directly by conflict (eg, injuries due to gunshots, land mines, bomb blasts). The health burden averted by surgical care at the Trauma Centre was large…

And yet, despite the importance of the Trauma Centre and its inclusion on a centralized No-Strike List that database was not consulted during the operational vetting and legal approval of the two plans drawn up by the OD-As [032, 045] (which, to be fair, had never been in the city and had no direct knowledge of the terrain; their Joint Terminal Attack Controllers had tried to print hard copy of ISR imagery before they set out from Camp Pamir but the base’s only printer was so old all it could produce were ‘giant magenta blobs’  that were completely useless [383] – so initially they relied on a single 1:50,00 map to plan and execute their operations [048]).

In fact – the irony is extraordinary – one member of the Special Operations Task Force testified that even they had no access to the No-Strike List and only discovered the existence of the Trauma Centre by accident, when ‘somebody was looking for additional medical facilities for use as emergency means to treat our own casualties’ if they could not make it back to Camp Pamir and the Forward Surgical Team based there [217, 219].  It was only then, late in the night of 29 September, that the Trauma Centre was added to the database maintained by the ISR [Intelligence, Surveillance and Reconnaissance] Tactical Controllers at the Special Operations Task Force at Bagram; early the next morning e-mails with this information were sent to ‘all ISR assets supporting operations in Kunduz’ [213].

***

us-army-personnel-leave-a-truck-inside-afghan-military-base-kunduz-1-oct-2015

On 30 September there was a secure videoconference between General Campbell, his Afghan counterpart and Major Hutchinson.  It was clear that Campbell was exasperated at the conduct of the Afghan forces and attached great importance to re-taking the city.  Fired up, Hutchinson briefed his men on the planned Kunduz Clearing Patrol, relaying the spirit of Campbell’s comments and telling them this was ‘a no fail mission’, that ‘all of the civilians have fled and only the Taliban are in the city’, and that ‘everything is a threat’ [256].  That night, once the mission had been approved, the Green Berets fought their way into the city alongside the Afghan Special Security Forces, with Close Air Support from US aircraft including an AC-130 gunship that ‘continuously called out and engaged [Taliban] ambush sites’ [325].  This seems to have been the same aircraft and crew that returned on 2/3 October; the sensor operator described that fateful mission as their third flight over Kunduz, following two others on 2 September and 30 September, the last when they provided armed overwatch for a US convoy into the city centre and engaged the Taliban at multiple locations.  Indeed, he claimed that those previous missions had provided them with ‘good situational awareness’ of Kunduz and the ‘patterns of life’ of both civilians and insurgents [362].

Before dawn on 1 October the US and Afghan troops had cleared several key buildings and established a defensive strongpoint in the Provincial Chief of Police Compound [PCOP].  They hunkered down and came under repeated mortar, rocket-propelled grenade and automatic weapons fire, and throughout that day and the next their Joint Terminal Attack Controller called in multiple strikes from F-16 aircraft, many of them ‘danger close’, in immediate proximity to the PCOP [332].

nyt-map-kunduz-ops

By the end of the afternoon on 2 October several Afghan troops had been wounded.  Their commander was all for taking the casualties back to Camp Pamir immediately, but Hutchinson persuaded them that this was madness: they were stable so the medical evacuation should wait for the cover of darkness.  The Afghan Special Security Forces agreed; while they were at Pamir they would re-supply and then return to attack a command and control centre they said had been established by the Taliban in the National Directorate of Security compound (NDS) to the south west of the PCOP which the Afghan SSF also referred to as ‘the NDS prison’ [386-8].  The investigation report includes this map showing the relationship of the PCOP to the NDS Compound and the MSF Trauma Centre:

map-pcop-msf-and-nds-in-kunduz

[A similar map included in a detailed analysis by The Intercept mis-locates the PCOP – almost certainly confusing it with the NDS Prison that the Operations Center in Bagram wrongly assumed was the intended target of the air/ground operation: see below]

The Afghan Special Security Forces were assured that Close Air Support would be extended to the convoy once they had returned to the ‘self-defence perimeter’ beyond the PCOP – a ‘bubble’, Hutchinson called it, roughly defined by the range of the heavy machine guns and anti-aircraft guns used by the Taliban [387].

But by then the F-16s providing close air support were running low on munitions and in the early evening, with the situation in Kunduz remaining precarious, the Special Operations Task Force scrambled the AC-130 gunship from Bagram to take over.

***

The AC-130 (call-sign ‘Hammer’) was a mission in a hurry and the aircraft took off without a proper briefing or any geospatial intelligence products.  All the aircrew had was the grid location of the PCOP and the call sign and contact frequency for the OD-As [052].  By then, superior commands had received the e-mail detailing the location of the Trauma Centre, and at 1847 the Fires Officer from Combined Joint Special Operations e-mailed a package of ‘mission products’ to the Electronic Warfare Officer onboard the AC-130 which included that information.  But en route to Kunduz one of the aircraft’s communications systems failed and the message never arrived; when the aircrew did not acknowledge receipt, the Fires Officer at Bagram made no attempt to pass the information over the radio (which was working) [052].

At 0130 on 3 October the Afghan convoy left on its evacuation and re-supply mission, and Hutchinson contacted the AC-130 through his Joint Terminal Attack Controller (JTAC) to ask them to carry out a ‘defensive [infrared] scan’ of the area of operations.  Specifically, he wanted to prepare the ground ahead of the convoy’s return: if they were ambushed and ‘got fixed in place what I wanted to do was to reduce heavy weapons and strongpoints so that they would be able to effectively maneuver on to the objective’ [390].  To that end he supplied the aircraft with a grid location for the NDS compound.

It is unclear – from the redacted report, at least – how the co-ordinates of the target were obtained. Hutchinson said that when the Afghan Special Security Forces showed him their plan for securing the NDS compound it included ‘a grid [which] said, I think, NDS prison’, but when he plotted the location he realised it was not the Prison to the south that was one of the objectives included in the original plan to establish a foothold in the city. Hutchinson riffed on the multiple NDS facilities throughout Kunduz, but this begs a crucial question: how did he plot the grid to confirm the location?  He claimed to have been working from the 1:50,00 map spread out on the hood of his armoured vehicle, which could hardly have provided the co-ordinates required for a precision strike.  The Joint Terminal Attack Controllers would have had access to digital imagery stored on their laptops, but by this stage they were running low on batteries and cannibalising the radios of other Green Berets to keep communications with the AC-130 open [334, 383].  ‘The worst part of it,’ Hutchinson said, was that the day after the strike they found a detailed 1:10,000 map produced by a Provincial Reconstruction Team in 2013 ‘with nice crisp imagery, and it had everything labelled with 10-digit grids’ [397].  The commanders of the other two OD-As remembered it differently, both testifying that the map was found in the provincial governor’s office on 1 October.

ac-130u_sensor_operator

All this matters because when the TV sensor operator on the AC-130 (above) inputted the grids that were passed by Hutchinson via his JTAC he found ‘it put me in a field with residential buildings’.  The AC-130 has a sophisticated sensor suite, including high resolution sensors (an All Light Level Television system, infrared detection set and strike radar to permit all weather/night target acquisition).  But reading between the redactions in the investigation report there is some suggestion that there are also known technical issues with the system (perhaps distortion introduced by the aircraft’s height and/or orbit, because the AC-130 had been forced out of its overhead orbit at 2220 by taking evasive action against a surface-to-air threat): ‘Nothing in the immediate location matched the target but from training I was aware that at significant [redacted]…’ [363].

So the sensor operator widened the search and found a large compound 300 metres to the south that appeared to match the description of the target.  It was not difficult to find: the Trauma Centre had its own generator and was the only building in the city that was still brightly illuminated.  At first sight the sensor operator said ‘there was nothing else near the original location that could match the description of a prison.’

msf-kunduz-attack

‘As we got closer,’ s/he continued,

I observed multiple [redacted: this is surely MAMs or ‘military-aged males’, a term the US military was supposed to have discontinued, which would explain the otherwise puzzling deletion] walking in between buildings [redacted] entrances with [redacted: guards?] posted. After passing back the information to the JTAC he said the compound was under enemy control and that those [redacted: MAMs?] were declared hostile [363].

The navigator had informed the JTAC that the grids had originally plotted to the middle of a field but they now had a large compound in their sights, a T-shaped structure with an arch gate and nine people ‘roaming outside’.  The Green Berets conferred with the Afghan Special Security Forces in the PCOP who confirmed that this was the NDS compound, and the Fire Control Officer on the AC-130 adjusted the target location in the fire control system accordingly [054, 242].

unless-the-grids-are-off

But the sensor operator, more mindful of the Tactical Guidance issued by General Campbell (below), testified that he wanted ‘to make sure we were not inadvertently declaring civilians hostile’.

resolute-support-tactical-guidance-jpeg

So he re-entered the original co-ordinates (‘to determine any system [redacted: error?]’) – by then the AC-130 had moved to a more accurate, overhead orbit [057] – and this time the sensor homed in on a second compound:

a much smaller compound with two large buildings, what appeared to be a third smaller shack, two overhangs, a wall surrounding, what appeared to be guard towers at the four corners with a single entrance on the south side of the compound and was unable to observe any movement in that compound [363].

This underscored his concerns.  ‘Now that we are closer,’ he told the rest of the crew,

even though that compound [is] the only one that’s limited and has activity, if you look in the TV’s screen you can see this hardened structure [the second compound] that looks very large and could also be more like a prison with cells. So I just want to verify that before we start declaring people hostile, that we are 100 per cent sure that this is the correct compound [057].

He asked the navigator to request a more detailed target description from the Joint Terminal Attack Controller (though the JTAC was not told that the aircraft’s sensors had now identified two different compounds from the same grids).

The JTAC came back with a target description of multiple buildings with a wall surrounding, and a main gate with an arch shape. I asked for further clarification on which side of the compound that gate was on, to which he replied the North side of the compound. The gate I was able to make out at the first compound was on the north side and matched the target description [363].

The first compound was the MSF Trauma Centre; the second was the NDS compound.

***

The redacted version of the investigation report includes a satellite image of the MSF Trauma Centre but conspicuously failed to include a corresponding image of the NDS compound.  Yet from TerraServer’s satellite imagery (below) it is clear that the two are radically different, and in fact the gate on the NDS compound faced south not north.

kunduz-imagery-001

kunduz-imagery-002

kunduz-imagery-003

Neither Hutchinson nor his JTAC had access to real-time imagery from the AC-130 because the same antenna that prevented the aircraft receiving the e-mail with the No-Strike List also prevented it from transmitting a video feed to the JTAC’s laptop, and so both the aircrew and the US forces on the ground had to rely on verbal descriptions.  The investigation report calls the characterisation of the target building ‘a vague description’ [034] but, as Mathieu Aikins pointed out in a superb analysis of the strike, ‘it’s actually a rather specific description that corresponds to MSF’s distinctive layout.’  Indeed, when the aircrew compared the two compounds they were persuaded by the description of a ‘T-shaped structure’ that they had identified the correct target.

Once the AC-130 aircrew’s description of the Trauma Centre had been confirmed by the Afghan Special Security Forces as the NDS compound, the circle was closed.  As Hutchinson testified, he had a report from the AC-130 ‘that describes a target, the disposition of the target and the pattern of life on it that’s completely consistent with what I’ve heard from the Afghans…’  Whether the Afghans deliberately substituted a description of the Trauma Centre for the NDS compound remains an open question.  From their own (separate) interviews in Kunduz, both May Jeong and Mathieu Aikins repeatedly raise this as a distinct possibility.  Some informants insisted that the Trauma Centre had been overrun by the Taliban, confirmed (so they said) by raw intelligence and communications intercepts, even that it was being used as a firing position – a claim that was repeated by the government in Kabul in the immediate aftermath of the strike – while others complained that MSF treated Taliban casualties who then returned to the fray: ‘patching up fighters and sending them back out.’  Much of this is ex post facto rationalisation; clearly many Afghans regarded the attack on the Trauma Centre as perfectly justified.  But Aikins asks a more pointed question: Did Afghan forces, out of longstanding mistrust of MSF, draw the United States into a terrible tragedy?’

If they did, then it had to have been a spur-of-the-moment decision to take advantage of a developing situation, since the Afghan Special Security Forces had originally provided the correct grids for the NDS compound.

More telling, I suspect, is that from 0100 until well into the attack on the Trauma Centre the only people who had the co-ordinates for the target now in the sights of the AC-130 were the aircrew, who did not pass the grid location for what they had incorrectly identified as the target back to Hutchinson.  And yet the ground force commander had already told the navigator he had ‘great confidence in the grids passed [057]’, and it is astonishing that this did not prompt a more extensive discussion among the aircrew since the original grids had plotted first to an open field and second to the NDS compound but never to the Trauma Centre that had now been designated as the target.

Neither did the aircrew pass the revised grids back to the Special Operations Task Force who were monitoring events from Bagram.  Repeating Hutchinson’s earlier mistaken assumption, the staff in the Operations Center at Bagram believed the target (‘the NDS prison’) was the Prison in the south of the city which had been included in the original Kunduz City Foothold Establishment plan, and they tasked an MQ-1 Predator to provide surveillance over that location [059].

nds-compound-and-prison-in-kunduz

Hutchinson could not view the video feed from the Predator, since the laptops in the PCOP were desperately short of batteries, but the Special Operations Task Force did have access to the drone’s real-time imagery.  Nothing was happening around the Prison, and confident that this was the strike location nobody at Bagram attempted to confirm the coordinates until the attack on the Trauma Center was well under way.  At 0207 they heard a sudden, direct transmission from the AC-130 – ‘unreadable numbers followed by going hot/rounds away’ – and ‘the quickness of the going hot call’ suggested to one experienced JTAC at Bagram that ‘there was possibly a dire situation on the ground.’  But ‘the passing of engagement grids was broken, unreadable’, and s/he immediately ‘made multiple attempts to get a resend of [the] grid of engagement’.  Those requests ‘were either not acknowledged or met with “still engaging/hot”‘, but this was ‘not uncommon due to the task saturation during coordination and employment by ground JTACs and aircraft’.   Meanwhile another JTAC in the Operations Centre, realising that ‘no activity was noted at the facility’ – presumably by the Predator on station over the Prison; the Taliban had reportedly freed all the prisoners when they took the city – tasked the Predator crew to ‘find the engagement area’ [261-4].  At 0220 they were successful, and once the new grids had been checked the Operations Center realised that the AC-130 was attacking the Trauma Center.

***

Hutchinson provided two contradictory rationales for the attack.  One was offensive: his JTAC relayed to the AC-130 that Hutchinson’s intent was to ‘soften the target’ (meaning the NDS compound) for the Afghan convoy returning from Camp Pamir.  When the aircrew asked for clarification they were told they were to ‘destroy targets of opportunity that may impede partner forces’ success’ [059].  When he was questioned by the investigating officers, Hutchinson represented this as pre-emptive and precautionary: ‘If they were going to take contact I did not want to play twenty questions while they were taking fire’ [391].  The other was unambiguously defensive: the immediate trigger for Hutchinson to clear the AC-130 to open fire was the sound of automatic gunfire from the east-west road near the NDS compound.

What did it for me in the end was when I believed the [redacted] convoy to be at that parallel cross street … or the perpendicular cross street … to the facility, I heard sustained automatic weapons fire … and it was coming from that general direction. And so I asked the [redacted] are they in contact yet. He can’t get through [to] them at first, and so I think okay, so that’s a sign they’re probably in contact… Fire continues and I ask him again and he says strike now, assume they are decisively engaged’ [393-4].

It’s not clear from the redactions who Hutchinson was talking to, but it was almost certainly someone from the Afghan Special Security Forces in the PCOP.  What is certain – and known to the aircrew on the AC-130, who were also tracking the convoy, but not to Hutchinson – was that the convoy was nowhere near the NDS compound or even the Trauma Centre at that time but 9 km away, still within the northern perimeter of the airfield.

Hutchinson’s attention was on the sound of gunfire.  He explained that most of the fire directed against his forces in the PCOP compound had been from the west, and it was ‘unthinkable’ that ‘there would have been anything functional over there in terms of essential services’ [394] – like a hospital.

And so, at 0202 Hutchinson had his JTAC instruct the AC-130 to strike the ‘objective building first’ and then to provide ‘suppressing fire’ (which the JTAC later described as a ‘PAX cocktail’ and the aircrew translated as ‘MAMs’ [military-aged males]).   Again the aircrew sought clarification; they wanted to be sure that their target was the ‘large T-shaped building in the centre of the compound’ and that ‘we are [also] cleared on people in this compound.’  It was and they were; at 0208 the first round was fired as the Electronic Warfare Officer announced the grids over the radio: the garbled transmission received at Bagram [064-6].

***

ac-130h-u_007-ts600

The AC-130 made five passes over the Trauma Centre at 15-minute intervals, firing a total of 211 rounds.  But ’rounds’ fails to convey the scale of the ordnance involved.  As May Jeong notes, the AC-130 is ‘built around a gun’; it is, after all, a gunship.  It has a 105 mm M102 Howitzer that fires high explosive shells at 10 rounds a minute (reputedly the largest gun ever operated from a US aircraft); a 40 mm Bofors cannon that fires 120 rounds a minute; and a 25mm 5-barrelled Gatling cannon that fires incendiary rounds at 1,800 rounds a minute.  YouTube has a video of a live-firing exercise carried out by the 4th Special Operations Squadron in 2016 that is truly chilling:

All these weapons are side-firing; the AC-130 performs a slow left-banking pylon turn in a five-mile orbit to keep its weapons on target for much longer than a conventional strike aircraft:

w2300-kunduzhosp1011-plane

The results on the ground were catastrophic.

15hospital-1-master1050

6988426-3x2-940x627

All the patients in the ICU died except one, alongside the caretakers who were with them; one doctor, three nurses and a cleaner who were in the ICU were also killed.  Here is Kathleen Thomas again:

I hope with all my heart that the three sedated patients in ICU, including our ER nurse Lal Mohammad, were deep enough to be unaware of their deaths — but this is unlikely. They were trapped in their beds, engulfed in flames.

The same horror that rocked the ICU rocked the rest of the main building as the plane hit with alarming precision. Our ER nurse Mohibulla died. Our ER cleaner Najibulla died. Dr. Amin suffered major injuries but managed to escape the main building, only to then die an hour later in the arms of his colleagues as we desperately tried to save his life in the makeshift operating theater set up in the kitchen next to the morning meeting room.  The OT nurse, Abdul Salam, died. The strikes continued further down the building, tearing through the outpatients department, which had become a temporary sleeping area for staff. Dr. Satar died. The medical records officer Abdul Maqsood died. Our pharmacist Tahseel was lethally injured. He also made it to safety in the morning meeting room, only to die soon after, having bled to death. Two of the hospital watchmen Zabib and Shafiq also died.

Our colleagues didn’t die peacefully like in the movies. They died painfully, slowly, some of them screaming out for help that never came, alone and terrified, knowing the extent of their own injuries and aware of their impending death. Countless other staff and patients were injured; limbs blown off, shrapnel rocketed through their bodies, burns, pressure wave injuries of the lungs, eyes, and ears. Many of these injures have left permanent disability. It was a scene of nightmarish horror that will be forever etched in my mind.

cvwb2clwuaaw4si-jpg-medium

***

The loss of life and the destruction of the hospital was appalling.  But the effects of the air strike have reverberated far beyond the Trauma Centre and the events of 3 October.  In February this year Sophia Jones told the troubling story of a father of four who lost his right arm and the sight of one eye when he was caught in cross-fire between the Taliban and the Afghan army.  With the destruction of the Trauma Centre in Kunduz there were no local hospitals capable of treating his life-threatening injuries, and it took him two agonising days to travel 200 miles to the Surgical Center for War Victims run by another NGO, Emergency, in Kabul – now ‘the only free, specialized trauma hospital of its kind treating war victims in Afghanistan.’  Like MSF, Emergency is absolutely clear that ‘we cannot be on one side of the war’: ‘a patient is a patient’.  Like MSF, most of Emergency’s patients are civilians.  But, as Luke Mogelson found in the spring of 2012,

At Emergency’s hospital in Kabul, it’s not unusual to find Afghan national security forces recovering in the same ward as Taliban insurgents, and after a while, the ideas that make enemies of the two men lose their relevance; the daily spectacle of their impact on human bodies invalidates them.

That was then.  ‘After Kunduz’, Emergency’s program co-ordinator now concedes, ‘anything is possible.’  It would be truly, desperately awful if one of the casualties of the air strike on the Trauma Centre turned out to be the core principle of medical neutrality.

One year after Kunduz, Christopher Stokes, MSF’s General Director, warned that ‘A war without limits leads to a battlefield without doctors.’  MSF pledged not to allow that to happen.  They must not stand alone.

To be continued

The Federal Administration of Military Violence

On 15 June – one week after the attack on Karachi’s international airport by the Pakistan Taliban (Tehrik-i-Taliban or TTP) and the Islamic Movement of Uzbekistan (UMI) – the Pakistan military announced its ‘comprehensive’ Operation Zarb-i-Azb in the Federally Administered Tribal Areas (FATA).  From their bases in North Waziristan, the statement announced, militants had ‘waged a war against the state of Pakistan’ and the military had been ‘tasked to eliminate these terrorists regardless of hue and color, along with their sanctuaries.’  Although the press release insisted that ‘these enemies of the state will be denied space anywhere across the country’ the epicentre of the operation was and remains North Waziristan.

NWA_PakMil_locations_all_1200

There are reportedly 60,000 troops in the region, and the main Pakistan military installations in North Waziristan are shown on the map above (taken from the AEI spinoff site, Critical Threats), but the prelude to ground operations was a concerted attack by the Pakistan Air Force on eight targets linked to planning the assault on Karachi airport.

Green Ground Red Drones Blue PAF strikes DAWN 20 June 2014

‘Operation Zarb-i-Azb’ refers to Mohammed’s sword, and its political imagery is artfully dissected by Afiya Shehrbano Zia:

‘It refers to the (‘sharp/cutting’) sword of the prophet of Islam and is a brilliant usurpation of the religious metaphor. It upstages the religious imaginary for which the Tehreek e Taliban Pakistan (TTP) claim to be fighting. After all, who would dare to vanquish the Prophet’s metaphorical sword? The appellation justifies its cause for the defense of the Islamic state, and quells the lesser purpose of the Taliban in one fell swoop. As in all cases in the instrumentalisation of religion as a propaganda tool, it also excites nationalists and seeks to rationalise another round of military operations, killings and displacements that will follow.’

There’s much more of value in her commentary, but – as Zia also acknowledges – the genealogy and geography of the offensive is no less complicated (my map comes from Dawn, 20 June 2014; green circles are Pakistan military land operations; blue are Pakistan Air Force strikes; red are US drone strikes).

First, it’s not clear whether the Pakistan military finally has the Afghan Taliban in its sights too – regarded by Islamabad as the ‘good Taliban’ because, far from threatening the state of Pakistan, it has long been used by both the military and (particularly) the intelligence service as a counter to any Indian influence over Kabul once US and ISAF forces complete their withdrawal.  And it is of course the Afghan Taliban (along with the Haqqani network) which is the principal concern of the United States.

Second, the Pakistan military – and especially the Air Force – has a long history of offensive operations in the FATA, as I’ve discussed in detail before: see here and here.  Now other commentators have noticed this: the Bureau of Investigative Journalism has tracked 15 Pakistan Air Force strikes carried out by helicopter gunships and F-16 fighters between 19 December 2013 and 15 June 2014, which killed 291-540 people (including 16-112 civilians).

The significance of this is not only that it precedes the current offensive but also that it coincides with the so-called ‘pause’ in CIA-directed drone strikes against targets in the FATA.  Chris Woods notes that PAF strikes are ‘generating casualties far in excess of any caused by CIA drones strikes’, and one resident of Mir Ali recited a grim military timetable:

“It’s like doomsday for people in Mir Ali, where death is everywhere since Saturday… They start the day with artillery shelling early in the morning. Gunship helicopters come for shelling during the day and jets strike at around 2:00-2:30 in the night.”

The military denies all reports of civilian casualties but this beggars belief, and the Bureau reports that some residents have even concluded that the drone strikes were preferable:

‘The difference between the drone strikes and the military strikes is that drones target specifically who they want to target… the wanted terrorists… people are saying that drone attacks were good compared to the military strikes.  Personally I agree, because I have seen drones, they are in the air 24 hours and they don’t attack as randomly… the place of the attack was always an area where the Taliban or terrorists were living.’

But whatever one makes of this – a calculation that would imply that the CIA had abandoned its anonymous ‘signature strikes’ – drones have not been absent from the skies over Waziristan. Pakistan has its own reconnaissance drones, and they have repeatedly been used to direct strike aircraft onto their targets (though with what accuracy it is impossible to know) and to support ground operations: the PAF boasted of their use in May, when hundreds of houses and shops were destroyed in Machis Camp and in the bazaar at Mir Ali.  And – the third complication – the US resumed its drone war on 11 and 12 June when two UAVs fired six missiles at compounds near Miram Shah, supposedly killing ten members of UMI and the Haqqani network, and again on 18 June when three compounds near Dargah Mandi were hit.  Whether these strikes were co-ordinated with Pakistan is unclear – the Foreign Office has issued its ritual denial, but it’s difficult to believe they were not connected to Pakistan’s own military operations, and here too there is a long history of what I’ve called ‘dirty dancing‘ between Washington and Islamabad that continued until at least the end of 2011.  It seems highly unlikely that the dance has ended.

_75732447_75732446

Finally, the shock waves from these various operations ripple far beyond their ostensible targets. Hundreds of thousands of people have fled, some in advance of military operations (which had been telegraphed for months), many more when the military temporarily loosened its curfew on the region. As on previous occasions, most of them fled to Bannu, some to government camps (‘Only the poorest of the poor would go to a camp in such hot and humid weather‘) but the majority to stay with family members, while some refugees have even crossed into Khost in Afghanistan to seek sanctuary. The map below is an early trace (18 June), and it shows only those who are officially registered so it excludes those lodging with their extended families; but even this anticipates hundreds of thousands more displaced people to come.

Pakistan Displaced Persons June 2014

There is also the real fear that, as Ismail Khan and Declan Walsh reported earlier this month, Taliban reprisals will focus on the Punjab, the electoral base of Prime Minister Nawaz Sharif.

All of this suggests the importance of unravelling the intimate connections between the political constitution of the Federally Administered Tribal Areas and the administration of military violence there.  This is clearly not limited to CIA-directed drone strikes, and here Zia’s reflections on a question posed by a feminist friend are worth repeating:

‘She wonders, “why this obsession with drones?” Obviously, the interest is due to a host of factors, but her query reflects the difference in modes of analysis. Her position reflects the views of women’s rights/human rights groups who consider specific military operations in one part of Pakistan as just one cog in a broader narrative about the source of the conflict. For them, this has been the cosy nexus and mutually beneficial relationship between the military establishment and the jihadi groups.

Those like Imran Khan, who foreground drones in their analysis of ‘conflict’, consider US intervention and occupation of Afghanistan as the drivers of conflict in Pakistan. But local progressive groups argue that even if militants in FATA are subdued, or US interventions are resisted, unless the policy of patronage and nurturing of jihadi groups in the rest of Pakistan is dismantled and buried, conflict at all levels will never end – drones or no drones.

This doesn’t mean that military technologies are unimportant nor that drone strikes are of marginal concern (inside or outside an ‘area of active hostilities‘): it means that we need to direct our attention to the larger matrix of political and military violence within which they are deployed, transnational and national, and to its genealogies and geographies.

Logistics and dialectics

Shortly before Pakistan re-opened its borders with Afghanistan to NATO’s military convoys, I described the (political and economic) frictions of distance involved in supplying the war in Afghanistan in an essay for Open Democracy.  I described the two main supply lines, the Pakistan Ground Lines of Communication and the Northern Distribution Network, and the intricate system of political concessions and pay-offs each involved.

The border crossings reopened on 5 July, after a break of seven months, but the convoys have been reduced to a trickle by bureaucratic delays and by drivers’ demands for compensation for the long lay-off.

Re-opening the border provoked angry demonstrations in Pakistan.  Standing at the Torkham Gate at the Khyber Pass a local leader of Jamaat-e-Islami, a right-wing Islamist party, declared that ‘NATO supply is haram [forbidden] and against sharia’ and promised to issue a fatwa against it.

But, according to an AP report this morning, it is not only the US military that is relieved at the opening of its supply lines: so too are the Taliban.  Previous reports in The Nation by Aram Roston, together with a scathing Congressional investigation, Warlord Inc., documented the routes through which the Pentagon’s logistics contracts made provision for payments to insurgents not to attack their convoys.  The central mechanism for the privatisation of the supply chain was Host Nation Trucking, which was cancelled in August 2011 (three months before the border closed).

It was replaced by a new National Afghan Trucking contract, but more than half of the 20 contractors involved in the new scheme had been prime or subcontractors under the previous contract, and convoy security was still in the hands of private contractors.  John Tierney, the Democrat chair of the original Congressional investigation, was exasperated: ‘We are right back to the same people that were involved in the problem that instigated the investigation.’  And, as the AP report suggests, this includes the Taliban:

 ‘The insurgents have earned millions of dollars from Afghan security firms that illegally paid them not to attack trucks making the perilous journey from Pakistan to coalition bases throughout Afghanistan… Pakistan’s decision to close its border to NATO supplies in November in retaliation for U.S. airstrikes that killed 24 Pakistani troops significantly reduced the flow of cash to militants operating in southern and eastern Afghanistan, where the convoys travel up from Pakistan, said Taliban commanders…

 “Stopping these supplies caused us real trouble,” a Taliban commander who leads about 60 insurgents in eastern Ghazni province told The Associated Press in an interview. “Earnings dropped down pretty badly. Therefore the rebellion was not as strong as we had planned.” A second Taliban commander who controls several dozen fighters in southern Kandahar province said the money from security companies was a key source of financing for the insurgency, which uses it to pay fighters and buy weapons, ammunition and other supplies.  “We are able to make money in bundles,” the commander told the AP by telephone. “Therefore, the NATO supply is very important for us.”

[The] commanders said they were determined to get their cut as the flow of trucks resumes from Pakistan…  “We charge these trucks as they pass through every area, and they are forced to pay,” said the commander operating in Ghazni. “If they don’t, the supplies never arrive, or they face the consequence of heavy attacks. … We have had to wait these past seven months for the supply lines to reopen and our income to start again… Now work is back to normal.”‘