Was the Kunduz Hospital Attack a War Crime?

by Jens David Ohlin

The Pentagon has released its report on the U.S. air assault against a hospital in Kunduz, Afghanistan, in October. The picture painted by the Pentagon report is pretty damning. The attack killed 42 people and turned out to be a giant mistake. The U.S. attacked the wrong building.

Initially, some Afghanistan officials suggested that insurgents had taken up positions in the hospital—an allegation that spurred an intense legal debate about whether, and when, the presence of such fighters would render the hospital a legitimate military target under LOAC. The Pentagon report makes clear that these allegations were unfounded. The insurgents were located in a different building, and the U.S. hit the wrong target.

The Pentagon report details a litany of mistakes—not just a single mistake but indeed a “cascade” of errors. The mistakes were clearly evidence of unprofessional behavior and deserving of reprimands. A total of 16 Americans involved in the attacked were officially disciplined administratively.

But was the attack criminal? 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 (more on that later), but the civilian killings were not performed with purpose.

The New York Times had this to say about reckless attacks on civilians:

The failure to bring any criminal charges was “simply put, inexplicable,” said John Sifton, the Asia policy director of Human Rights Watch.

There are legal precedents for war crimes prosecutions based on acts that were committed with recklessness, he added, and recklessness or negligence does not necessarily absolve someone of criminal responsibility under the United States military code.

Is Sifton right about this?

The answer to this question is complicated. I’ve written an entire article about this, Targeting and the Concept of Intent, and I can’t go into that level of detail in a blog post. And even my full-length article did not fully address all angles of the question. The issue is exceedingly complex. But let’s make some preliminary observations.

The Rome Statute includes two particular war crimes of interest to the discussion.

The first provision defines as a war crime:

Intentionally directing attacks against the civilian population as such or against individual civilians not taking direct part in hostilities;

The second provision defines as a war crime:

Intentionally launching an attack in the knowledge that such attack will cause incidental loss of life or injury to civilians or damage to civilian objects or widespread, long-term and severe damage to the natural environment which would be clearly excessive in relation to the concrete and direct overall military advantage anticipated;

The first war crime is a violation of the principle of distinction: intentionally killing civilians. The second war crime is a violation of the principle of proportionality: causing disproportionate collateral damage.

The problem with applying the first war crime provision from the Rome Statute is that the attack against the civilians in the Hospital building in Kunduz did not obviously involve “intentionally directing attacks against the civilian population.”

Now here is where things get complicated. The word “intentionally” does not have a stable meaning across all legal cultures. As I note in my article, the word intentionally is generally understood in common law countries as equivalent to purpose or knowledge, depending on the circumstances. But some criminal lawyers trained in civil law jurisdictions are more likely than their common law counterparts to give the phrase “intentionally” a much wider definition, one that includes not just purpose and knowledge but also recklessness or what civilian lawyers sometimes call dolus eventualis. Now for present purposes I am going to avoid the difficult controversy of whether dolus eventualis is equivalent to recklessness or a higher mental state (residing somewhere above recklessness but well below knowledge), and for the purposes of this discussion simply assume that dolus eventualis and recklessness are similar mental states dealing with risk-taking behavior.

Now here is the key point. In the civilian tradition, the concept of intent is a wider category that in some circumstances might include recklessness. This equation sounds odd to a common-law trained criminal lawyer, because to an American student of criminal law, intent and recklessness are fundamentally different concepts. But just for the sake of argument, what would happen if intent were given this wider meaning? Could the U.S. service members be prosecuted for intentionally directing an attack against the civilian population because “intentionally” includes lower mental states such as dolus eventualis or recklessness?

I worry about this argument. And here’s why. If intent = recklessness, then all cases of legitimate collateral damage would count as violations of the principle of distinction, because in collateral damage cases the attacker kills the civilians with knowledge that the civilians will die. And the rule against disproportionate attacks sanctions this behavior as long as the collateral damage is not disproportionate and the attack is aimed at a legitimate military target. But if intent = recklessness, then I see no reason why the attacking force in that situation couldn’t be prosecuted for the war crime of intentionally directing attacks against civilians, without the court ever addressing or analyzing the question of collateral damage. Because clearly a soldier in that hypothetical situation would “know” that the attack will kill civilians, and knowledge is certainly a higher mental state than recklessness. That result would effectively transform all cases of disproportionate collateral damage into violations of the principle of distinction and relieve the prosecutor of the burden of establishing that the damage was indeed disproportionate, which seems absurd to me.

The correct result, it seems to me, is to explicitly codify a new war crime of recklessly attacking civilians, and the codification of such a crime should use the word “recklessly” rather than use the word “intentionally.” The idea would be to create a duty on the part of attacking forces and then penalize them for failing to live up to it. (Of course, the scope and content of that duty would then have to be elucidated through case law adjudication.) And the existence of a separate war crime would help signal the moral difference between intentionally killing civilians and recklessly killing them.

If such a hypothetical prosecution were to take place, is there sufficient evidence that the attacking force was reckless in the Kunduz hospital case? Unfortunately yes. Among the factual issues are:

1. The targeting system of the AC-130 gunship was not operating correctly because the gunship had to take evasive maneuvers due to ground fire.

2. The targeting system therefore identified the target as an empty field, which forced the gunship’s crew to locate the correct target visually.

3. Using visual confirmation, the crew located the wrong building—the hospital—instead of the actual building where the insurgents were located.

4. Apparently the crew of the gunship either did not have a list of no-fire targets on board or failed to check the hospital coordinates against the list.

5. Commanders at HQ failed to check the coordinates of the hospital target with the coordinates on their no-fire target list. Had they done so, they would have realized that the gunship was about to engage a no-fire target.

6. After the attack began, hospital workers and MSF officials began frantically calling and texting the U.S. military to stop the attack, but there was a substantial delay before the attack was finally halted.

7. According to the Times report, at least one commander was hesitant to stop the attack when they did not have “situational awareness” (SA) on the ground. Apparently he was concerned that friendly ground forces might remain in danger even as they called off the attack. Of course, the opposite turned out to be true: because they lacked SA, they continued to attack the wrong target without a firm understanding of who or what they were really attacking. Obviously it was a mistake for them to have attacked the target in the first place given that they had no SA.

Does all of this add up to a crime of recklessness? I don’t know. That would be for the fact-finder to decide, but a prosecutor could certainly make out a prima facie case that targeting “best practices” were not followed in this case, leading to the identification of the wrong target, and the loss of 42 innocent lives. But I don’t think this is a Rome Statute case. There may be sound moral reasons to create a new war crime provision for accidents of this type, but I don’t think this conduct falls under the existing law as it stands now.


36 Responses

  1. Thanks for the post , It seems that the respectable author of the post , despite the terrible complication of the issue presented , has ignored two crucial points :

    First : The doctrine of : justification !! under universal principles of criminal law , collateral loss or damage , mounts to : intentional killing , right !! But , typically , there is no justification in criminal law , for collateral loss or damage . Yet , in war , there is such !! And why ?? very simple :

    The inherent right, of self defense, would become then senseless!! A state, wouldn’t be able to exercise that right, due to the simple fact, that an appropriate conduct of war, even with minimum efficiency, wouldn’t be possible ! Always , there are collateral damages expected , in every war almost .

    Second (and in accordance) that is how, negligence should be observed . For that purpose, one must distinguish between, preliminary preparations , where negligence should be tested by more severe standards (due to lack of battlefield stress) and, negligence in the battlefield itself , under stress and fire, where, the judgment of commander actions , shall and must be, much less severe, due to probable necessities and improvisations on the ground .

    And so , If : negligence, is revealed prior to battlefield stress, and : could be avoided by carefully or reasonably sticking to written protocols, and : huge loss could be anticipated in advance due to it , then:

    The prosecutor in Hague , could find great interest in it .For less than that , we have an issue , your are right !!


  2. “If intent = recklessness, then all cases of legitimate collateral damage would count as violations of the principle of distinction, because in collateral damage cases the attacker kills the civilians with knowledge that the civilians will die. And the rule against disproportionate attacks sanctions this behavior as long as the collateral damage is not disproportionate and the attack is aimed at a legitimate military target. But if intent = recklessness, then I see no reason why the attacking force in that situation couldn’t be prosecuted for the war crime of intentionally directing attacks against civilians, without the court ever addressing or analyzing the question of collateral damage.”

    The reason I do not believe this would be the outcome is because I think the mens rea is being applied to the wrong act. A “reckless attack” can be reckless under the principal of distinction, the principal of proportionality, or under both principals.

    If it is excepted that Article 8(2)(b)(i) criminalizes violations of the principal of distinction, then the mens rea should only apply to the act of discriminating between military targets and civilians.

    If recklessness were the mens rea, there would not be a violation of Article 8(2)(b)(i) if a military target is the object of the attack and collateral damage ensues. Instead, a violation of Article 8(2)(b)(i) with a recklessness standard would be an attack in which the actor cannot identify a target as either military or civilian, but proceeds to attack regardless. The crucial question under Article 8(2)(b)(i), and a material element of the crime, is whether civilians were the object of the attack, not whether they died as a result of an attack.

  3. Jens: I’ve only just glanced through the report, but noticed that it does appear to conclude that there were violations of the requirements of precaution, distinction and proportionality, wholly apart from any criminal culpability. Is that your understanding?

  4. I agree with Zach

  5. This doesn’t solve the problem because knowledge is a higher mental state than recklessness. In a situation where the attacking force is aware with virtual certainty that the attack will harm civilians, they will have made civilians the “object of the attack,” (if intent equates recklessness), regardless of whether the attack is disproportionate or not.

    As for this statement: “Instead, a violation of Article 8(2)(b)(i) with a recklessness standard would be an attack in which the actor cannot identify a target as either military or civilian, but proceeds to attack regardless.” In the Kunduz case, the attacking force identified the target as military, but did so erroneously (and possibly recklessly). The class of cases we are interested in are ones where the status of the target is misidentified in this manner.

  6. Hi Jens,

    You write that “where the attacking force is aware with virtual certainty that the attack will harm civilians, they will have made civilians the “object of the attack”.” I just don’t see it. If you attack A, knowing that B will be incidentally or collaterally harmed, then you do not thereby make B the object of your attack. You directed your attack at A, not at B.

    As I understand it, roughly, the war crime of recklessly attacking civilians involves (i) directing an attack (ii) at persons who are in fact civilians (iii) disregarding a grave risk that they are civilians. In contrast, the war crime of recklessly causing excessive civilian losses involves (i) launching an attack [at one target] (ii) disregarding a grave risk of causing incidental civilian losses [in or near the target] (iii) that would be excessive in relation to the reasonably anticipated military advantage. What’s the problem with that?

  7. Jens (and interested readers ) may read on the same subject , at ” just security ” blog ( by Alex whiting ) :



  8. Because the conclusion that “You directed your attack at A, not at B” is only possible with reference to the conceptual distinction between purpose and knowledge. But if the word “intentionally” in the relevant provision is interpreted to mean recklessness (as we are discussing), then the purpose-based understanding of “intentionally directing” is off the table.

    I have no problem with codifying a future war crime provision for recklessly attacking civilians, but I don’t think one can find such a crime in the current provision “Intentionally directing attacks against the civilian population as
    such or against individual civilians not taking direct part in hostilities” … This is especially true in light of Article 30 of the Rome Statute.

  9. Sometimes excessiveness is needed. US bombed Hiroshima and Nagasaki killed many but saved American lives by bringing Japan to its knees. USA had that right and duty given Japan attacked America.

  10. Hi Jens,

    As I understand it, your argument is

    1. If it is a war crime to recklessly direct attacks against civilians then “recklessly” modifies “direct attacks”, not “against civilians”.
    2. However, you can’t direct an attack against A without intending to attack A.
    Therefore, there can be no conceptually sound war crime of recklessly directing attacks against civilians.

    As I see it, the first premise is false. Consider that the Rome Statute defines the war crime of attacking civilians as *intentionally* directing attacks against civilians. Yet, according to the Elements of Crimes, the elements of that war crime are

    1. The perpetrator *directed* an attack.
    2. The object of the attack was a civilian population as such or individual civilians not taking direct part in hostilities.
    3. The perpetrator *intended* the civilian population as such or individual civilians not taking direct part in hostilities to be the object of the attack.

    According to the EoC, the mental element “intentionally” doesn’t modify “directing” (which needs no modification since it is inherently intentional). Instead, “intentionally” modifies “against civilians”. Accordingly, the elements of a war crime of recklessly directing attacks against civilians would be

    1. The perpetrator *directed* an attack.
    2. The object of the attack was a civilian population as such or individual civilians not taking direct part in hostilities.
    3. The perpetrator [*consciously disregarded a grave risk that*] the civilian population as such or individual civilians not taking direct part in hostilities [were] the object of the attack.

    On this approach, “recklessly” doesn’t modify “directed”, it modifies “against civilians”. Thoughts?

  11. Very helpful diagram of the issue.

    For me it does not change the outcome of the analysis, and the EOC only strengthens my conclusion.

    Consider the following situation: The attacking force knows that 10 HVT combatants are located in an apartment. The attacking force has only one available tactical option for defeating the combatants, and they realize that this tactical option will, with practical certainty, result in the death of 1 civilian who is also located in the apartment. Let’s stipulate that the collateral damage here is not disproportionate. If intent is redefined, this attack would fall under prong 3.

    Part of the issue here is that I think the word “object” is intimately tied to the mental state of acting with intent. I think that in prong 3 of your helpful diagram, intentionally modifies everything that comes after it, including “to be the object…”

    By the way, I should add that on Twitter, KJH has taken a different approach, and asked whether under the circumstances I have described, perhaps it is correct to conclude that there is a violation of the principle of distinction. (He s asking a question, not taking a firm stance on the question.) I take it that this is not your view.

  12. I think both Jens and Alex Whiting have done a great job at illuminating a complicated issue, and I really don’t see their conclusions as at odds. FWIW, I also don’t see the facts as supporting “recklessness” (but still digging through the report). BTW, take a look at Kevin Jon Heller’s discussion of mistake of fact in a different “war crime” case, but perhaps worth a scan re our current discussion: http://opiniojuris.org/2014/08/11/mh-17-framed-murder-war-crime/.

    Marty makes a good point, but I think those references are more about ROE (which typically includes restraints that are dictated by policy, not law). Regardless, I don’t think that any US jury – to include a civilian one (compare: http://articles.latimes.com/2008/aug/29/local/me-marine29 ) – would convict anyone here given the intense combat situation, and the total absence of any evidence that anyone involved knew it was a protected facility. IMHO, we need to be careful about judging combat from hindsight, and from our (relatively!) comfortable offices far from the enormous stress of intense warfighting. As the WSJ points, out “decisions on the battlefield must be made in real time with lives in the balance, and failing to act can also result in casualties.”

    This “failing to act” aspect is especially relevant as a couple of days before that attack, Amnesty International reported that the Taliban was conducting a “reign of terror” in Kunduz, and said it had “credible reports of killings, rapes and other horrors.” Understandably, they called for “Afghan authorities to do more now to protect civilians, in particular in areas where more fighting appears imminent.” I think we all might be able to appreciate that there is a lot of stress in that kind of situation to try to evict the Taliban in order to protect the civilians.

    Just FYI, I talk more about all this on my own blog, to include my disappointment with the human rights community

  13. Adil, Jens: I think it might help here to specify what sources of law are relevant, and what, exactly, it was that was recklessly or intentionally done.

    1. As I read it, the DoD Report itself concludes that some of the parties involved violated the laws of war — not merely ROE, Charlie — and did so “intentionally”; in particular:

    — the ground force commander and aircraft commander violated the principle of distinction by striking without distinguishing *whether* the target was military or civilian, i.e., by targeting the MSF Trauma Center without any reason to think it was a lawful military objective and when “it should have been presumed to be a civilian compound.” They “arbitrarily close the building they engaged.”

    — Because there was *no* established military value, the decision to`1 strike also, by definition, violated the requirement of proportionality.

    — The aircrew violated the requirement of feasible precautions by failing “to reduce the risk of harm to individuals they could not positively identify as combatants,” and by “consistently engag[ing] individuals that it did not positively identify as a threat.”

    These LOAC violations were all deliberate and intentional, whether or not the officers *knew* (or suspected) the building was a civilian object.

    Moreover, the Report repeatedly concludes that they acted unreasonably and in violation of numerous ROEs.

    2. The Central Command summary states that Gen. Campbell “did not conclude that these failures amounted to a war crime,” because “the label ‘war crimes’ is typically reserved for intentional acts — intentionally targeting civilians or intentionally targeting protected objects.” That’a an unfortunate, imprecise passage. Typically reserved by whom? And I suppose by “intentionally targeting” it means “deliberately targeting.” But even if that is what “typically” warrants the “label,” so what? Forget the label — the pertinent question is whether the Report describes conduct that can and/dor should be prosecuted.

    3. I’d assume the most important *legal* question is culpability under the UCMJ. Does that require deliberate targeting of civilians, or “merely” intentional, reckless and unreasonable violations of the LOAC? (I presume the latter, but I ahven’t studied the question.)

    4. Only then *might* one want to ask about culpability under the Rome statute (although it’s probably a purely academic question). And if “intention” is required under the Rome statute, then one needs to ask: intention *to do what*? To target individuals or objects *known* not to be a legitimate military objective? To intentionally target a building indiscriminately? To target it intentionally without having satisfied the principles of distinction and precaution?

  14. Marty, If you are saying that the aircrew simply attacked any building in Kunduz, I think that is a misreading. I don’t think there is any serious question that they thought it was an enemy facility – obviously mistaken – but their honest belief. Sure, there are processes, etc., that they should have used, but in the midst of combat people can make hard decisions when they think they are doing the right thing to save/protect friendlies. Imagine if, in fact, it was the correct facility, but they circled overhead without firing because some piece of equipment wasn’t working while people on the ground were being killed. Yes, the rule may be you can’t attack without this or that piece of equipment working, but combat can create different – and honestly held – imperatives when lives are at stake. To be clear, mistakes were made and a terrible tragedy occurred, and corrective/disciplinary action is mandated here, but I’d really like to see an experienced trial lawyer who would say that a criminal conviction would be the likely result in a war crimes trial. But I accept that I am in a minority of one in my view of this one!

  15. Charlie: As I read the Report, of course the choice of target was not *completely* random; but it was so indiscriminate as to be “arbitrary” (Report’s term), because it was apparently was based entirely on the fact that the hospital (like other buildings in the area) was t-shaped, which was not remotely enough to overcome the presumption that it was a civilian object.

  16. Hi Jens,

    As I understand your scenario, there are two possibilities:

    1. The attackers directed an attack at the apartment, which is a military objective.
    2. The object of the attack was *not* the individual civilian but the apartment. Presumably, if she ran out of the apartment then attackers would attack the apartment and not her.
    3. The attackers did *not* disregard a risk of making the civilian the object of the attack, though they expect her to suffer incidental harm since she is inside the object of attack.


    1′. The attackers directed an attack at the combatants (suppose everyone is standing in the courtyard and can be seen from above)
    2′. The object of the attack was *not* the individual civilian but the combatants. Presumably, if she ran far away from the combatants then the attackers would the combatants and not her.
    3. The attackers did *not* intend the civilian to be the object of the attack, though they expect her to suffer incidental harm since she is near the objects of attack.

    Of course, both “directed at” and “object of attack” are inherently intentional. The question is whether you can be reckless about whether the person at whom you are directing your attack, the person who is the object of your attack, is civilian.

  17. Just to re-emphasize my point, I don’t believe that the aircrew at the time of the attack, thought they were striking anything but a legitimate target (given that in combat there are very few things that are a certainty). Should they have done more? Not attacked until all their equipment was working? Sure, we can say all that now, and they have been disciplined for it – but the disciplinary authority – like prosecutors everywhere – took into account the extreme circumstances and stress under which they were operating in determining that administrative sanctions (which can be career ending) were the right disposition. (From a practical point of view, I’ve tried a lot of cases, and I just don’t think that a conviction could/would be forthcoming in this instance. Among other things, look at the convoluted discussion the experts here have been having as to what the law is – can we or should we expect nonlawyers to have a better grasp?)

    Though it is a bit beyond our discussion I might also add a couple of things knowing that they won’t sit well with some readers, but might be helpful in some way.

    While things here went terribly awry in this case, the U.S. military prides itself with being aggressive and pushing initiative down to the lowest levels. Decisions made with “ground truth” are often the best. Obviously, that wasn’t the case here, but too draconian a punishment may send the wrong message to the force, since it could create such a fear a mistakes that combat effectiveness will be compromised over the long run. Too much fixation on playing it safe is not the mindset you want, because combat necessarily and inevitably involves great uncertainty and risk.

    Minimizing mistakes and developing prudent risk-taking is obviously the goal, but the fact remains that things like Kunduz do happen and, I’m afraid, will continue to happen, in the chaos that is war. When something like that happens, you certainly need to learn everything you can to try to prevent a repetition, but you also have to press on fighting the enemy with real determination and resourcefulness.

    I’ve bored everyone enough, so I’ll sign off!

  18. Sorry, the second reading of your scenario should end with

    3′. The attackers did *not* disregard a risk of making the civilian the object of the attack, though they expect her to suffer incidental harm since she is near the objects of attack.

  19. Adil,

    I see where you are going with this, and I think I understand the account you are outlining. If I understand you correctly, you want to limit the issue of recklessness to the identification of the status of the target, but everything else would remain intentional in the sense of purpose. If we do that, then I think we avoid the problem that I am worried about. However, doing so requires understanding the conceptual of intentionality in one sense and then in a different sense in the very same provision, which as a matter of textual interpretation is difficult to square with the existing text. In other words, I don’t see a lot of evidence that the crime you are describing is the crime codified in the Rome Statute.

    However, I am pretty sure that your scheme is conceptually coherent. But again, if that’s what we want to codify, I would have written the provision in a different way, which is why I suggest creating a new crime.

    I also worry that courts won’t be scrupulous about limiting the wider notion of intent to the limiting circumstance you describe, but will apply it to other elements beyond the question of the status of the targets.

  20. Oh I didn’t mean to suggest that the Rome Statute recognizes a war crime of recklessly directing attacks against civilians. As you know, one of my objections to the Rome Statute is its failure to do so. I was just using the EoC to illustrate my account of the CIL/ICTY crime. Apologies for any confusion.

  21. If one wants to analyze non-Rome norms, there is the grave breaches provision of AP I, which outlaws “(a) making the civilian population or individual civilians the object of attack;” (art 85).

    While reasonable people can disagree, my reading of this provision is that it requires the purpose to attack civilians, since that is what is meant by the notion of making civilians “the object of attack.”

    Not sure how Adil, Marty, Kevin, or Charlie will interpret it, though.

  22. Oh, I think you can guess how I interpret article 85. Under article 85, it’s a grave breach to *wilfully* make civilians the object of attack, which the ICTY, following the ICRC, interprets to mean *recklessly* make civilians the object of attack. On this view, it’s a grave breach if (i) the object of your attack is in fact a civilian and (ii) you are reckless with respect to whether they are civilian or military.

    API has 174 state parties. Have *any* of them publicly rejected the ICRC/ICTY interpretation?

  23. I can’t speak to how many nations have accepted or rejected this interpretation. But I would simply say that the phrase “willfully making civilians the object of attack” sounds to me like dolus directus. I understand the ICRC interpretation, but to me it does not accord with the plain meaning of the sentence. I would also add that the original ICRC Commentary cites no legal support for its proposition that willfully in this context means recklessly. Kevin, what say you?

    I would also add that the Rome Statute has a long list of signatories, and as you have pointed out, the Rome Statute codifies a more much limited mens rea requirement…

  24. I should also add that the ICRC Commentary does not clarify what “recklessly” modifies, i.e. does not do the careful work that you did in your prior responses. It’s therefore not clear to me whether the ICRC would accept or would not accept the limit you have placed on where recklessness is relevant and where it is not relevant (only relevant as to status).

  25. Surely if the ICRC and ICTY’s interpretation was correct, there would at least ONE record of a military court martial convicting some soldier for the grave breach of recklessly killing civilians? After all, so many countries have signed AP I and it was in force for nearly two decades before the ICTY’s creative interpretation. Have there been any court martial’s based on a recklessness theory since then?

    If you can’t tell from the question, I am skeptical, but quite willing to be corrected.

  26. Hi Jens,

    Those are all fair points. I would note that all of the ICTY rulings that recklessly attacking civilians is a war crime post-date the Rome Statute, so the judges clearly didn’t think that the Rome Statute reflects CIL. As far as I know, no state party to the Rome Statute has complained that the ICTY misrepresented customary law. In any event, someone should really do a comprehensive review of state practice on this point. Not me, though. I have a book to finish.



  27. Hi random,

    Good question. One complicating factor is that states may elect to punish acts that would qualify as war crimes under their domestic military law. Accordingly, the US and UK both punish recklessly killing civilians under the headings of murder or manslaughter.

    Hope that helps,


  28. Had the Kunduz Hospital been taken by members of the Afghan armed groups and emptied of civilians a few hours before the attack was launched, and had the US launched the attack taking the same precautions (or lack thereof, as alleged in the report, i.e. using visual confirmation of the target, without checking the no-fire target list previously and without having SA on the ground), would this attack have been a war crime?

    That is, can a “reckless” attack that doesn’t harm civilians in any way be still criminal simply because it was “reckless”?

  29. I agree with Charlie Dunlap. In war, it is very difficult to analyse or scrutinise minutely whether the combatants intended to target a civilian or military facility. While engaged in counter-insurgency or counter-guerilla operations it is likely that combatants may commit a mistake because the response time is very short. It is easy to be wise after the event for lawyers, but it is extremely stressful in the field of battle.

  30. random,

    Here’s an example. More of an op-ed, but you can gather the facts.



  31. Anon,

    Great question. Grave breaches of API require causation of death or injury. The Rome Statute doesn’t, and I believe the same is true under customary law. So if you intentionally direct an attack at civilians then you’ve committed a war crime even if civilians miraculously emerge unharmed. Same with launching attacks expected to cause clearly excessive harm to civilians. So presumably a war crime of recklessly attacking civilians would not require a harmful result.

    Now, as a normative matter, I would like to restructure ICL so that we start with result crimes and then derive inchoate crimes (attempt, reckless endangerment) from those result crimes. But that’s not the lex lata.

    You can read about my view here: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1439803



  32. Response…
    Thanks to Jens and all the commenters- the analysis here has been really insightful.

    I only have a dilettante’s understanding of the law, so I’ve got a few questions for whoever is still interested in keeping this thread alive. I’m really trying clarify my own understanding of this issue, so these questions aren’t rhetorical.

    1. The discussion so far has focused on recklessly “making the civilian population or individual civilians the object of attack” rather than whether one was reckless in doing “everything feasible to verify that the objectives to be attacked are neither civilians nor civilian objects and are not subject to special protection but are military objectives within the meaning of paragraph 2 of Article 52…”. Is that because the former is a grave breach of API (and therefore part of ICL) while the latter is not?

    I ask because it seems to me that a hypothetical crime of “recklessly failing to do everything feasible to verify…” would proscribe precisely the sort of behaviour described in the Kunduz report. Importantly, it would do so without having to stretch the mens rea required to convict a commander of making civilians the object of attack.

    2. Is it significant that the state representatives at the Diplomatic Conference chose not to include ‘precautions in attack’ into the grave breaches regime in this way?

    Perhaps as merely a LOAC violation, rather than an international crime, the architects of API intended for the states to punish violations of ‘precautions in attack’ arising from reckless or negligent behaviour?

    My naïve sense is that career-affecting non-judicial punishment was the ‘right’ answer here and the fact that the US has pursued that course of action seems like a positive (albeit tiny) step towards establishing some sort of state custom of repressing violations of ‘precautions in attack’ this way. It is particularly laudable given the significant political pressure for states to sweep such incidents under the rug or to otherwise minimise culpability to the greatest extent possible.

    In contrast, I worry that no case of ‘recklessness in precautions’ could ever achieve the sort of gravity required to get a hearing at an international tribunal/court. Therefore:

    3. What would be the value of criminalising ‘recklessness in precautions’ under ICL?

    Is it that states might try to camouflage attacks against the civilian population as recklessness in precautions?

    It might be that these are completely the wrong questions to be asking and they might be entirely too focused on lex ferenda, but I’m really eager to know an expert’s take on them if anyone is game.

  33. Hi Noel,

    All excellent questions. Happy to engage. This is all useful for something that I’m writing.

    1. Yes, in a way. Both the ICTY and the Rome Statute draw on the grave breach provision, as well as on API 51(2), in framing the war crime of attacking civilians. I agree that a war crime of recklessly failing to take feasible precautions would cover much of the same ground, although it sounds rather inchoate. What if precautions are not taken, but the attack is called off anyway? What if those potentially harmed are not in fact civilians? Would the elements be satisfied?

    2. Yes, but not very. For example, it is not a grave breach to use civilians as human shields, but that is clearly a war crime under both customary law and the Rome Statute.

    FYI, the ICRC Commentary on Protocol I states that reckless conduct may be criminal but that merely negligent conduct should only be disciplined.

    I’m sympathetic to your sense that “non-judicial punishment was the ‘right’ answer here.” One explanation might be that war crimes should require not merely conscious disregard of a substantial risk (which the crew clearly had) but also an attitude of extreme indifference to the value of civilian life or some such thing (which I presume the crew lacked).

    3. Well, failure to take precautions, like the use of indiscriminate weapons, can be indirect evidence of intentionally attacking civilians. But I think the value of separately criminalizing reckless conduct is simply that far more civilians are killed recklessly than intentionally (Daesh notwithstanding). So if we want ICL to play a role in protecting civilians then that’s where most of the action is.

    Hope that helps,


  34. Adil,

    If it would be criminal to recklessly attack even if no civilians are hurt, then maybe one could say that the crime we’re talking about is not “reckless attacks against civilians” but rather “reckless attacks”.

    If so, what would the threshold for an attack, any attack, to be considered to be “reckless”? Maybe, it could be said that an attack is ” reckless” if the attacker lacks information to assess whether the attack would be proportional, i.e. if an attack is launched without having information to know if the anticipated damage to civilians would be greater than the direct military advantage brought by it?

    But if the above is correct, could it be possible that armed forces would be obliged to have the infrastructure necessary to make such an assessment before even considering whether to fight a war? Do most militaries have the means to carry out a proper evaluation of proportionality in every situation? Or the requirement would be different depending on the circumstances surrounding the armed force at hand?

  35. Hi Anon,

    The war crime that I’m discussing is recklessly making civilians the object of attack, that is, (i) directing an attack (ii) against persons who are in fact civilians (iii) with reckless disregard for the risk that they might be civilians. So I’m still dealing with the principle of distinction, not proportionality. The risk is that the persons attacked are civilians, not that civilians near the persons attacked might be collaterally harmed.

    With respect to proportionality, IHL requires attackers to do everything feasible to asses whether an attack may be expected to cause excessive harm to civilians, where feasibility depends on what is practically possible for the attacker at the time given the resources and information available. So while a state is blameworthy for not equipping its forces with adequate resources, combatants are responsible only for im/properly using the resources they’ve been given.

    That said, I follow the ICRC’s view that, in case of doubt regarding the proportionality of an attack, the attack should not be carried out. So if an ill-equipped force uses its limited resources but remains in grave doubt regarding the proportionality of an attack then they must refrain. In that way, the failure of a state to equip its forces can limit what its forces may lawfully do.

    Hope that helps,


Trackbacks and Pingbacks

  1. There are no trackbacks or pingbacks associated with this post at this time.