IRRC No. 919

“Reason to know” in the international law of command responsibility

Download PDF

A recent report by the Australian Defence Force arrived at a conclusion that further investigation was not warranted of commanders regarding their responsibility for failing to investigate suspicious behaviour of subordinates in Afghanistan, who were accused of violations of international humanitarian law. This troubling conclusion calls for a better analysis and understanding of command responsibility in international law and gaps in the law of command responsibility. This article identifies the conflicting precedents and scholarship regarding the law of command responsibility, which create uncertainty, and proposes a clarification of that law, with a special focus on the “reason to know” standard that triggers responsibility for failing to prevent or punish war crimes. It refutes the popular claim that commanders must act wilfully, and it rejects the common dichotomy between a commander who orders or otherwise directly participates in the war crimes of subordinates and one who unwittingly fails to prevent or punish such crimes. Using the empirical psychological literature, the article further explains how commanders can insidiously signal toleration of war crimes without giving direct orders. Finally, the article argues that international law, by absolving commanders who fail to properly train their subordinates to respect the law of armed conflict, misses a rare opportunity to deter war crimes, and offers some suggestions to fill this gap in the law.

* The authors thank Haley Bjorn for her research assistance.


Commanders are responsible for everything their command does or fails to do. … Commanders who assign responsibility and authority to their subordinates still retain the overall responsibility for the actions of their commands.

US Army Regulation 600-201



In November 2020, the Australian Defence Force (ADF) was confronted with the findings of a four-year-long investigation, undertaken by the inspector-general of the ADF and a justice of the New South Wales Supreme Court, Major General Paul Brereton, into allegations of war crimes by ADF special forces. The Afghanistan Inquiry Report, also known as the Brereton Report,2 found that there was credible evidence to support claims that, from 2005 until 2013, some members of the Australian Special Air Services (SAS) had engaged in a pattern of war crimes, including the murders of dozens of detainees and civilians and a subsequent cover-up. One of the practices uncovered in the investigation was the carrying of “throwdowns” – foreign weapons or equipment to be placed with the bodies of an ostensible “enemy killed in action” for the purposes of site exploitation photography, in order to portray that the person killed had been carrying the weapon or other military equipment when engaged and was therefore a legitimate target.3 The Brereton Report found evidence that Command was aware of these practices and had been told of claims by Afghans that SAS soldiers and junior officers were committing war crimes.4 SAS commanders chose neither to investigate the claims nor to alert high command, in part because SAS officers were biased toward disbelieving complaints.5

In applying the law of armed conflict (LOAC) to these events, the Brereton Report arrives at a set of conclusions that, while critical of SAS officers’ handling of the evidence of potential violations, do not favour further investigation of the commanders who themselves chose not to investigate these incidents. The Report characterizes the actions of the SAS as “disgraceful and a profound betrayal” of the ADF's “professional standards and expectations”6 by a number of lower-ranked soldiers, not going so far as to conclude that these actions together indicate an accepted culture of criminality throughout the SAS. The suspicious behaviour of subordinates, the Report concludes, could have been properly interpreted as being indicative not of premeditation for the commission of war crimes, but rather of acts done to avoid unnecessary scrutiny for what could theoretically have been lawful activities.7

The Brereton Report paints the picture of an insular and secretive unit operating without the oversight of Command,8 but concludes that the behaviour was justifiably considered necessary for unit cohesion, stating:

The close-holding of information – frequently referred to as “compartmentalisation” – is a necessary feature of military units generally, and it is accentuated in the sphere of special operations. The security of the nation and the lives of individuals can depend on it.9

Ultimately, the Report explains commanders’ failure to discover, prevent or investigate potential violations, in part, by noting that “few would have imagined some of our elite soldiers would” commit these violations.10


The conclusions of the Brereton Report, on both unit cohesion and the law of command responsibility, make for unsettling reading. They imply that the obligation of military officers to prevent and punish war crimes is secondary to considerations of morale and unit effectiveness. More generally, they reflect a long-standing inconsistency in command responsibility doctrine that removes a critical disincentive to dangerously irresponsible command decisions and thereby undermines command responsibility's deterrent value.

This article examines the law of command responsibility and its relationship to unit cohesion and other extralegal values. In general, the law of command responsibility makes a commander criminally responsible for crimes committed by forces under his or her effective authority and control if the commander knew or, owing to the circumstances at the time, had reason to know that the forces were committing or were about to commit such crimes, yet failed to take all necessary and reasonable measures to prevent or repress the commission of the acts.11 Similarly, a commander who knew or had reason to know of past war crimes of subordinates becomes responsible for failing to take the necessary measures to punish those subordinates.12 Within this seemingly straightforward doctrine lies a confusing vagueness about the mental element of the offence. In particular, the circumstances under which a commander will be deemed to have constructive knowledge of the war crimes of subordinates has occasioned recurrent debates among international criminal tribunals and scholars alike, with inconsistent and sometimes contradictory results.

Nearly all jurists addressing the issue propose a clear demarcation between a commander's intentional participation in a subordinate's war crime, which triggers either direct or command responsibility, and a commander's mere negligent supervision of subordinates who commit war crimes, which generally results in the exoneration of the commander as lacking the necessary scienter. This article challenges the widely assumed dichotomy between participation and neglect as both an oversimplification of human methods of communication and a misapprehension of the dynamics of military organizations. The consequences of this false dichotomy, illustrated in the Brereton Report but by no means unique to it, have proven disturbing from both moral and legal perspectives. The article proposes a reformed concept of a commander's “reason to know” of war crimes by subordinates, in which evidence of intentionality assumes a less prominent role. It concludes by suggesting an alternative to the doctrine's absolution of commanders from any duty to train or supervise subordinates under their command.

The law of indirect command responsibility

If a military commander plays an active role in promoting war crimes by subordinates, the appropriate charge is ordering, soliciting or inducing a war crime,13 aiding, abetting or assisting in a war crime,14 or contribution to a war crime,15 depending on the form the commander's promotion takes. In contrast, command responsibility relates to a superior's failure to take appropriate action to prevent or punish war crimes committed by subordinates. Indirect responsibility is a much more complex doctrine and raises difficult questions of interpretation and application, particularly regarding the mental element of the offence. The essential point of indirect command responsibility is to deter commanders from tolerating war crimes by subordinates or exonerating them after their crimes come to light.

In the Brereton Report, because there were no allegations that ADF officers ordered subordinates to commit war crimes in Afghanistan, only the indirect aspect of command responsibility is implicated. This is unsurprising, not merely because direct commands to commit war crimes are relatively rare. The doctrinal difficulties of the law of indirect command responsibility have so perplexed both scholars and international criminal tribunals that untangling the jurisprudence and the underlying policy rationale for specific approaches to indirect command responsibility demands careful analysis. In this part of the article, we will briefly summarize the conflicting jurisprudence and sources of authority on command responsibility under the LOAC as it has developed since the Second World War.

Early indirect command responsibility in the LOAC

Command responsibility for ordering war crimes by subordinates, or for refusing to punish subordinate war criminals, found its way into early municipal articles of war and military codes,16 but no such law before 1945 held commanders criminally responsible for violations of international law caused by mere toleration of war crimes by subordinates. Suggestions for the incorporation of more robust command responsibility into the international LOAC were floated as early as the Hague Peace Conferences of 1907 and the Versailles Peace Conferences of 1919,17 but even after the Second World War, the Allies were divided about whether a military commander could be criminally responsible for the war crimes of subordinates caused by toleration or neglect, as opposed to ordering the crimes.18

The Yamashita trial

The resolution of that debate, and the origin of the modern law of indirect command responsibility, came with the trial of the commander of the Japanese Army in the Philippines, Tomoyuki Yamashita. General Yamashita became the first officer to be charged based on responsibility for an omission, specifically in permitting officers and troops under his command to plan and commit thousands of war crimes, including mass murder, torture, mutilation and gang rape of civilians as Japanese forces were driven out of Manila.

There was no strong, direct evidence that Yamashita knew or expressed approval of the war crimes,19 and he claimed at trial that he had operational but not disciplinary control over naval land forces. Nonetheless, the US Military Commission found that Yamashita, as the highest military commander in the Philippine Islands, had a responsibility to prevent and investigate war crimes committed by subordinate officers under his command:

The Prosecution presented evidence to show that the crimes were so extensive and widespread, both as to time and area, that they must either have been wilfully permitted by the accused, or secretly ordered by the accused. … [W]here murder and rape and vicious, revengeful actions are widespread offences, and there is no effective attempt by a commander to discover and control the criminal acts, such a commander may be held responsible, even criminally liable, for the lawless acts of his troops, depending upon their nature and the circumstances surrounding them.20

Thus, two new standards for a commander's criminal responsibility were established: first, the commander must not passively tolerate war crimes of which he or she is aware, and second, the commander must supervise and discipline troops under his or her command with regard to detecting and preventing war crimes. As Ilias Bantekas has interpreted the case:

This standard … creates an objective negligence test that takes into account the circumstances at the time. Absence of knowledge is no defence if the superior did not take reasonable steps to acquire such knowledge, which in itself constitutes criminal negligence. Superiors have reason to know if they exercise due diligence. … This inevitably raises a duty to know, rebuttable only through evidence of due diligence, because it is a commander's duty to be apprised of events within his or her command.21

These principles were reaffirmed by the Nuremberg Tribunal and the International Military Tribunal for the Far East (IMTFE), also known as the Tokyo War Crimes Tribunal.22


Additional Protocol I

The Yamashita rule was incorporated with some alterations into Additional Protocol I to the 1949 Geneva Conventions (AP I), in its Article 86:

The fact that a breach of the Conventions or of this Protocol was committed by a subordinate does not absolve his superiors from penal or disciplinary responsibility, as the case may be, if they knew, or had information which should have enabled them to conclude in the circumstances at the time, that he was committing or was going to commit such a breach and if they did not take all feasible measures within their power to prevent or repress the breach.23

The phrasing of AP I – “had information which should have enabled them to conclude” – suggests a high standard of diligence in both supervision of subordinates and prevention of war crimes. The use of the past conditional “should have enabled” indicates that the failure to investigate incomplete evidence or make logical inferences that war crimes were occurring engages the commander's responsibility as surely as if the commander actually knew these facts. Wilful blindness and reckless disregard of incriminating facts are therefore as culpable as actual knowledge,24 consistent with the ancient maxim of non scire quod scire debemus et possumus culpa est.25


The International Committee of the Red Cross (ICRC) Commentary on AP I asserts that a commander's “negligence must be so serious that it is tantamount to malicious intent, apart from any link between the conduct in question and the damage that took place”.26 This interpretation is unfortunate. It contradicts the plain language of Article 86, which requires only that the commander possess inculpatory information and fail to take action as required. Malicious intent is simply not an element of the crime. On the contrary, AP I thus uses the strongest possible language to describe the commander's responsibility once facts have come to his or her attention that warrant further investigation.27 Apathy meets this standard as well as malicious intent, although it could perhaps be argued that malicious intent can be inferred from a commander's self-blinding to the facts available to him or her.

The language of Article 86 seems to require evidence to be in the commander's actual possession (“had information”), not merely available to the commander. It thus may be interpreted to exonerate a commander who, lacking incriminating evidence, takes no measures to supervise direct subordinates in order to ensure that they are not committing war crimes. If this interpretation is correct, it departs from at least one of the credible interpretations of the Nuremberg and Tokyo war crimes precedents. The ICRC Commentary takes the opposite approach; it interprets actual possession of incriminating information as not being required to engage the commander's responsibility. For example, it views a commander's failure to diligently review reports of war crimes during absence from the theatre of combat as not excusing a failure to inform himself or herself and to take appropriate preventive or punitive measures.28

Statutes of the international criminal tribunals

In modern practice, Article 7 of the Statute of the International Criminal Tribunal for the former Yugoslavia (ICTY) used less exacting language with regard to negative duties, providing that a superior would be responsible for the illegal acts of his subordinate

if he knew or had reason to know that the subordinate was about to commit such acts or had done so and the superior failed to take the necessary and reasonable measures to prevent such acts or to punish the perpetrators thereof.29

The Statute of the International Criminal Tribunal for Rwanda (ICTR) included similar language in its Article 6(3). In both cases, the tribunals have interpreted their respective statutes as requiring only general knowledge about possible war crimes30 – a point that will be elaborated later in this essay.


Although “reason to know” is conceptually analogous to (and more elegant than) the AP I standard of scienter, the ICTY Statute dilutes “all feasible measures” into the less exacting “necessary and reasonable measures”. The distinction may be more apparent than real, however. The seriousness of war crimes such as the wilful killing of civilians or the torture of detainees implies that very extreme measures should be considered both necessary and reasonable to prevent such crimes. It is therefore logical to interpret any measure taken by a commander that is weaker than necessary to prevent or punish such crimes as unreasonable.31

The Rome Statute of the International Criminal Court (ICC) echoes, but expands considerably upon, both AP I and the ICTY and ICTR Statutes. Article 28 provides in relevant part:

A military commander or person effectively acting as a military commander shall be criminally responsible for crimes within the jurisdiction of the Court committed by forces under his or her effective command and control, or effective authority and control as the case may be, as a result of his or her failure to exercise control properly over such forces, where:

(i) That military commander or person either knew or, owing to the circumstances at the time, should have known that the forces were committing or about to commit such crimes; and

(ii) That military commander or person failed to take all necessary and reasonable measures within his or her power to prevent or repress their commission or to submit the matter to the competent authorities for investigation and prosecution.32


The “should have known” standard of responsibility, which originates in the Toyoda trial,33 reflects an early (later amended) draft of AP I's Article 86.34 It also parallels the ICTY Statute's “reason to know” and the “all necessary and reasonable measures within his or her power” language, merging the terms of AP I and the ICTY Statute on the obligation of prevention.35 But the Rome Statute also adds a consequential innovation: an explicit reference to the responsibility of military commanders to “exercise control properly” over forces under their command.


Plainly, neither the statutes of international tribunals nor their jurisprudence attribute strict liability to military commanders whenever their subordinates commit war crimes. But precisely what they require of commanders is not entirely apparent from these sources and is made still less certain by the different formulations found in them. Consider the divergent treaty language:


  AP I ICTY/ICTR Statutes36 Rome Statute of the ICC
Scienter “had information which should have enabled them to conclude” “had reason to know” “owing to the circumstances … should have known”
Action required “all feasible measures within their power” “the necessary and reasonable measures” “all necessary and reasonable measures within his or her power”


Interestingly, the ICRC's study on customary international humanitarian law endorses the “knew or had reason to know” language of the ad hoc tribunals.37 At the same time, the ICRC Commentary on AP I takes the surprising position that the “should have enabled” language is essentially inoperative due to a divergence between the English and French versions of AP I, with the French version referring only to information that did in fact enable the commander to conclude (“des informations leur permettant de conclure”).38 The reasoning behind this conclusion, as Jenny Martinez has noted with some charity, is “not entirely clear”.39 If evidence suggests that subordinates are committing war crimes, such evidence “should” enable a commander to conclude that a crime is occurring, and therefore it does “permit” the commander “to conclude” the same.

Modern jurisprudence developing the law of command responsibility

Between 1949 and the formation of the ICTY, no international tribunal further developed the law of command responsibility appreciably. However, the issue came up regularly before both the ICTY and ICTR, beginning in the mid-1990s. Soon, these tribunals began treating command responsibility as customary international law, despite the nearly fifty-year gap in international criminal jurisprudence. In the Čelebići case, the ICTY noted the incorporation of command responsibility into AP I and declared that the criminal responsibility of “military commanders and other persons occupying positions of superior authority … for the unlawful conduct of their subordinates is a well-established norm of customary and conventional international law.”40

In evaluating whether war crimes by subordinates engaged a commander's indirect responsibility, the ICTY adopted a three-step approach. First, there needed to be a superior–subordinate command relationship. Second, the commander must have had the requisite mental state, as discussed below. Third, the superior must have failed to take reasonable and appropriate measures to prevent or punish international crimes committed by a subordinate or subordinates.41 A fourth step, showing that the commander caused or contributed to the subordinate's war crime by failing to prevent or punish the crime, has sometimes been discussed in the ICTY and ICTR jurisprudence, but it was not clearly required until the Rome Statute embraced a causation requirement.42

The development of command responsibility jurisprudence beyond these basic points has proved nettlesome. In some cases, the tribunals have treated command responsibility as a dereliction of duty by commanders, who assume responsibility for their own acts only.43 In others, the tribunals have followed the Nuremberg and IMTFE precedents in treating command responsibility as a form of vicarious liability that made the commander complicit in the crimes of subordinates.44 This jurisprudence has left international criminal law plagued with mixed messages, but the trend in the ICTY and ICTR Trial Chambers was toward treating a dereliction of duty as sufficient to trigger command responsibility,45 while the Appeals Chambers of the ICTY favoured an approach that required the commander to possess some information which would suggest that war crimes were planned, were being committed, or had been committed by subordinates.46 This section will summarize the elements of command responsibility as they have developed in modern international criminal law jurisprudence.

Superior–subordinate relationship

An officer or other commander may be held liable for the acts of subordinates only if the subordinates are under his or her formal command47 or, in the case of civilian commanders, effective authority and control.48 This is a factual test. An officer holding higher rank in the same service as a subordinate and in the same unit of military organization may have effective control over the subordinate, but formal military authority is not dispositive.49 An officer or civilian superior may have effective control even without formal superiority in rank, and therefore military titles have limited relevance. Moreover, superior officers may be tasked with supervising only specific junior officers and troops. The jurisprudence of the ICTY accordingly focuses on “de facto command”, meaning actual ability to control the behaviour of subordinates, at the time of the commission of the relevant acts.50 Command responsibility cannot, therefore, be presumed from de iure command.

“Effective control” must be assessed in the broader context of a situation of command or authority, with command being defined as authority over forces, and authority being defined as “the power or right to give orders and enforce obedience”.51 The ICC has stated that effective control can be ascertained through examination of objective factors, such as the capacity to issue orders, whether orders are in fact followed, the authority to issue disciplinary measures, and the power to terminate the employment of subordinates.52 It is obviously easier to prove the existence of a superior–subordinate relationship in the context of a military chain of command, where rank is clearly delineated within a hierarchical structure. However, because effective control is a factual test, there are certain universal relevant factors that can be examined to find the requisite relationship, even in the absence of a formal military hierarchy. The ICC in the Bemba case outlined the relevant factors as including:

(i) the official position of the commander within the military structure and the actual tasks that he carried out; (ii) his power to issue orders, including his capacity to order forces or units under his command, whether under his immediate command or at lower levels, to engage in hostilities; (iii) his capacity to ensure compliance with orders including consideration of whether the orders were actually followed; (iv) his capacity to re-subordinate units or make changes to command structure; (v) his power to promote, replace, remove, or discipline any member of the forces, and to initiate investigations; (vi) his authority to send forces to locations where hostilities take place and withdraw them at any given moment; (vii) his independent access to, and control over, the means to wage war, such as communication equipment and weapons; (viii) his control over finances; (ix) the capacity to represent the forces in negotiations or interact with external bodies or individuals on behalf of the group; and (x) whether he represents the ideology of the movement to which the subordinates adhere and has a certain level of profile, manifested through public appearances and statements.53

It should also be noted that, in at least some US cases involving command responsibility, courts have held, consistent with equitable principles, that a military commander “cannot escape liability where his own action or inaction causes or significantly contributes to a lack of effective control over his subordinates”.54 Determining whether a commander has made a “significant contribution” to undermining his or her own control inevitably requires a fact-dependent inquiry, but such behaviour as the delegation of substantial authority to subordinates without adequate supervision, a chronic failure to punish infractions or insubordination, or displaying a high degree of passivity when leadership is needed would all seem to qualify as relevant factors.


Mental state and scienter

As noted, a commander who orders or otherwise directly contributes to a subordinate's war crime becomes a principal and active participant in the crime under various doctrines establishing responsibility for ordering, facilitating or contributing to a war crime.55 Similarly, a commander who knew of war crimes and did not take adequate measures to prevent or punish them incurs indirect command responsibility. These standards are doctrinally straightforward, and most difficult questions turn on the availability and persuasiveness of evidence of the relevant facts.

The second situation, and the one arising most commonly in practice, occurs under the scenario of incomplete information about possible war crimes by a subordinate. When a commander becomes aware of ambiguous facts which raise a suspicion that subordinates might have committed war crimes or might commit them in the future, the concept of “should have known” or “reason to know”56 comes into play and precludes the plea that a commander can only assume responsibility when observing a war crime flagrante delicto or with incontrovertible evidence, such as contemporaneous video footage of the crime. Confusion and discord in the doctrine of command responsibility rest principally on disagreement about the nature and extent of the commander's responsibility to act when information available to the commander may suggest, without proving, that his or her subordinates are planning, committing, or have committed a war crime.

Part of the confusion arises from the fact that “reason to know” is not a unitary concept; it is a spectrum, ranging from highly probative information confirmed by multiple independent and reliable sources at one end, to the merest unsubstantiated innuendo from a single unknown or biased source at the other. At the root of this discord is moral doubt about imposing a criminal punishment for a person's incompetence or passivity, or worse, for an exercise of questionable judgment in assessing uncertain evidence of war crimes, rather than intentional wrongdoing. The early international criminal tribunals overcame this doubt to a degree by rejecting the need for proof of the commander's positive knowledge about a subordinate's war crime. In the trial of Wilhelm List, often known as the Hostages case, the Nuremberg Tribunal affirmed that a commander need not be aware of war crimes committed by his subordinates to incur liability; the commander's failure to review reports of war crimes and to order investigation alone could make the commander criminally responsible.57 If a commander

fails to require and obtain complete information, the dereliction of duty rests upon him and he is in no position to plead his own dereliction as a defense. … Want of knowledge of the contents of reports made to him is not a defense. Reports to commanding generals are made for their special benefits. Any failure to acquaint themselves with the contents of such reports, or a failure to require additional reports where inadequacy appears on their face, constitutes a dereliction of duty which he cannot use in his own behalf.58

A British military court took a similar approach in the trial of Major Karl Rauer, whose subordinates had been convicted of executing British prisoners of war while traveling to a prison camp in several instances. Although it was not claimed that Rauer ordered or even knew of the executions, the subordinates repeatedly reported shootings during escape attempts and Rauer failed to investigate. Although Rauer was acquitted of the charge for the first such murder, apparently on the basis that he had no reason to disbelieve his subordinates, the court found Rauer guilty of the subsequent charges, because he had set a tone favouring war crimes by expressing hostility toward captured enemies, and after the first report by subordinates, he should have investigated the shootings to prevent their continuation. Rauer was sentenced to death by hanging, which was ultimately commuted to life imprisonment.59


In the Araki judgment, the IMTFE similarly held that a commander is responsible for having “failed to acquire” knowledge, through “negligence or supineness”, that war crimes were being committed by subordinates.60 It was insufficient that a commander “accepted assurances from others more directly associated with [the facts on the ground] if having regard to the position of those others … he should have been put upon further enquiry as to whether those assurances were true or untrue”.61 In the Toyoda case, the IMTFE reaffirmed this standard, elaborating that if the commander

knew, or should have known, by use of reasonable diligence, of the commission by his troops of atrocities and if he did not do everything within his power and capacity under the existing circumstances to prevent their occurrence and punish the offenders, he was derelict in his duties.62

Command responsibility can thus arise from constructive or imputed knowledge that subordinates were committing or were about to commit war crimes, and the imputation is not defeated by mere assurances from subordinate officers or reports unless these views are vindicated by the commander's reasonably diligent investigation.


By contrast, in the High Command case, the Nuremberg Tribunal insisted that command responsibility must result from

a personal dereliction. That can occur only where the act is directly traceable to [the commander] or where his failure to properly supervise his subordinates constitutes criminal negligence on his part. In the latter case it must be a personal neglect amounting to a wanton, immoral disregard of the action of his subordinates amounting to acquiescence.63

This is the language repeated in the ICRC Commentary on AP I. It is unclear how negligence can amount to acquiescence; the two concepts are mutually exclusive. It appears likely that the Nuremberg Tribunal was attempting to raise the standard of command responsibility to apply only in the most severe cases, but perhaps did not wish to impose upon prosecutors an obligation to provide evidence that the commander actually approved or tolerated known war crimes.


It is technically possible to reconcile these opinions by imputing a “wanton, immoral disregard” of a subordinate's war crime to any disregard of credible information implicating such crimes. However, such an interpretation seems contrary to the emphases of the respective opinions. It may instead be that the High Command case is an outlier, and that all that can be concluded from the early jurisprudence of the war crimes tribunals is that it supports a spectrum of opinions about the commander's requisite scienter in the face of evidence of war crimes by subordinates, ranging from “negligence” and “supineness” on one end to “wanton disregard” and “acquiescence” on the other. This is perhaps dependent to some extent on the degree of removal of the commander from the subordinate, with officers at the top of the chain of command being held to a less exacting standard than lower-ranking officers with more direct supervisory responsibilities over the guilty subordinates.

As noted, the ICTY and ICTR have also been at odds with themselves on the mental element of the doctrine. The Appeals Chambers of both tribunals agree that the commander's actual knowledge of past or future war crimes by subordinates need not be proved. It suffices that the accused “had ‘some general information in his possession, which would put him on notice of possible unlawful acts by his subordinates’”.64 The adjectives used suggest that the evidence need not be exceptionally strong to engage the commander's duty. Because the information need only be “general”, it would seem unnecessary for a commander to know the identity of the specific subordinates involved, the time or date of the crime, the identity of the target or victim, or other details. As noted in the Čelebići case:

This information does not need to provide specific information about unlawful acts committed or about to be committed. For instance, a military commander who has received information that some of the soldiers under his command have a violent or unstable character, or have been drinking prior to being sent on a mission, may be considered as having the required knowledge.65

Because the commander need be alerted to no more than “possible unlawful acts”, the reliability of the information apparently does not need to be high.


Yet, the Appeals Chambers have also shied away from full criminal responsibility in cases in which a commander egregiously failed to supervise troops who committed war crimes. The jurisprudence of the early Trial Chambers proposed multifactor tests to determine whether a commander could be held responsible for negligence. In the Blaskić case, for example, the ICTY Trial Chamber held that “ignorance cannot be a defence where the absence of knowledge is the result of negligence in the discharge of [the commander's] duties: this commander had reason to know within the meaning of the [ICTY] Statute”.66 The Chamber elaborated that knowledge may be proved by circumstantial evidence such as the

number, type and scope of the illegal acts; the time during which the illegal acts occurred; the type and number of troops involved; the logistics involved, if any; the geographic location of the acts; the widespread occurrence of the acts; the speed of the operations; the modus operandi of similar illegal acts; the officers and staff involved; and the location of the commander at the time.67

The ICTY Appeals Chamber has drawn the line at the “possession” of information, however, and rejected a commander's criminal responsibility for failure to supervise criminal subordinates, even if such dereliction enabled or contributed to the commission of their war crimes. Most prominently, in the Čelebići case, the Appeals Chamber rejected the reliance on IMTFE precedents and held that a commander does not assume responsibility for war crimes committed by subordinates without actual possession of some incriminating knowledge.68 In the ICTY's view, a military commander has no legal obligation to supervise the compliance of his or her direct subordinates with the LOAC;69 a commander's responsibility can arise from “deliberately refraining” from investigating information in his or her possession about subordinates’ war crimes, but not for “negligently failing” to gather such information in the first place through inadequate or non-existent training, supervision, or both.70


The municipal military laws of States are no more consistent than the international decisions. The US Department of Defense Law of War Manual adopts the standard in the High Command case, favouring the language imposing the least stringent duty on commanders to prevent war crimes: “The commander's personal dereliction must have contributed to or failed to prevent the offense; there must be a personal neglect amounting to a wanton, immoral disregard of the action of his or her subordinates amounting to acquiescence in the crimes.”71 US military practice thus ignores both contemporaneous and subsequent jurisprudence applying a higher standard to commanders.72

The Australian Criminal Code Act 1995 uses a much more forgiving standard: the commander who fails to “exercise control properly over” forces under his or her command must either know or be “reckless as to whether the forces were committing or about to commit” war crimes.73 The United Kingdom goes still further. The UK Manual of the Law of Armed Conflict uses the Rome Statute's “reason to know” language on the commander's mental state,74 but UK criminal law provides that a military commander must have either known “or consciously disregarded information which clearly indicated that the subordinates were committing or about to commit … offences”.75 Conscious disregard and “clearly indicated” are a far cry from negligent supervision or even wilful blindness to information suggesting war crimes. It appears that the UK manual adopts the scienter standard for civilian commanders under the Rome Statute. The Statute provides that military commanders will be liable if “owing to the circumstances at the time, [they] should have known”76 that forces under their control were committing or planning a war crime, whereas civilian commanders become responsible if they “knew, or consciously disregarded information which clearly indicated, that the subordinates were committing or about to commit such crimes; [and] … [t]he crimes concerned activities that were within the effective responsibility and control of the superior”.77

Not all State military and penal codes adopt such forgiving standards, however. For example, the French Penal Code reproduces the language of the Rome Statute,78 and German criminal law holds a commander liable if he or she “omits to prevent” a subordinate from committing a war crime or “negligently omits properly to supervise a subordinate under his or her command or effective control.”79

In short, with regard to “reason to know,” international criminal law has established no definitive statement of where, on the spectrum of reliability, information must fall for purposes of indirect command responsibility, and how much (if anything) a commander must do to investigate or to supervise subordinates in the absence of any inculpatory information. Such inconsistent standards as “wanton disregard”, “reckless disregard”, “criminal negligence”, “supineness” or “dereliction of duty” are used to characterize the commander's reaction to the information available in various cases. At most, it can be said that customary international law appears to have converged on an interpretation of “reason to know” which encompasses general information, including from outside sources such as media reports, that does not need to be complete or to include highly dependable sources of evidence, but which is nonetheless “sufficiently alarming”80 that it puts the commander on notice of possible war crimes by subordinates.81 If a military commander chooses not to investigate such facts, that failure must engage the commander's criminal responsibility.82 At a minimum, then, a commander who blinds himself or herself to the specific facts relevant to possible war crimes by his or her subordinates is not safe from responsibility, because the mere awareness of alarming information suffices to trigger the duty to act.

Whether the relevant information includes general information relating to the characteristics of subordinates, such as their ages, training, experiences, past criminal convictions, service records and attitudes, is unclear.83 The existing jurisprudence focuses more on evidence of facts indicating that war crimes are actually being planned or executed, or have actually been committed. However, the jurisprudence of the international criminal tribunals has not unambiguously ruled out the relevance of general information.

Failure to take measures to prevent or punish

A superior is liable both for a failure to prevent a foreseeable crime by subordinates and a failure to punish one that has occurred. These are separate obligations, and a commander will be responsible for forsaking the duty to prevent a foreseeable crime even if he or she punished the crime afterward.

In the Čelebići case, the ICTY Trial Chamber expressed scepticism that a satisfactory general standard of preventive action could be formulated. According to the Chamber, “any evaluation of the action taken by a superior to determine whether this duty has been met is so inextricably linked to the facts of each particular situation that any attempt to formulate a general standard in abstracto would not be meaningful”.84 Other judgments of the ICTY have taken a less pessimistic view and relied upon four factors:

  1. the degree of effective control a superior has over the conduct of subordinates – different superiors will have different degrees of power and control, and this will affect what measures they are expected to take;

  2. the extent to which a measure is necessary and reasonable under the circumstances;

  3. the severity and imminence of the war crime – more grievous or imminent potential war crimes require the commander to react more expeditiously and decisively; and

  4. the actual authority and ability of the commander to prevent the crime – impossibilium nulla obligatio est.85

The ICC Appeals Chamber made clear in the Bemba case that it is not enough to suggest in the abstract that a commander could have done more. Instead, a tribunal “must specifically identify what a commander should have done in concreto” to prevent the subordinate's war crimes.86


As for past war crimes, the ICTY Appeals Chamber in the Hadžihasanović case has held that a superior officer's knowledge of and failure to punish the past offences of subordinates cannot as a matter of law justify imputing future war crimes by the subordinates to the superior officer.87 However, although the Chamber was correct in distinguishing the duty to punish from the duty to prevent, the two are not factually unrelated. A superior's conscious tolerance of crime by subordinates will predictably promote more crime. In holding that the Trial Chamber had committed an error of law in inferring command responsibility based on failure to punish past war crimes, the Appeals Chamber removed a potent disincentive for military commanders to punish war crimes committed by subordinates. The Appeals Chamber thereby made it easier for military commanders to escape liability for fostering a culture of tolerance for war crimes, and to compound the abuse by implicitly endorsing it.88 As will be explained below, this line of reasoning is especially problematical because not all communication within the context of a military command occurs through explicit orders or guidance.

One point at which international criminal tribunals and domestic laws have been repeatedly at odds is whether the commander's failure to prevent or punish makes the commander responsible for the subordinate's war crime itself or for an independent crime of neglect of duty. The ICTY Appeals Chamber wrote in the Krnojelac case that “where superior responsibility is concerned, an accused is not charged with the crimes of his subordinates but with his failure to carry out his duty as a superior to exercise control”.89 Although a few scholars and jurists have endorsed this viewpoint,90 it contradicts much of the jurisprudence of the post-Second World War military tribunals, and it has not been followed by the relevant treaties or consistently in State practice.91


The final element – only explicit under the Rome Statute92 – is that of causation: a superior is criminally responsible for crimes committed by subordinates under his or her command and control, or effective authority and control, as a result of his or her failure to exercise proper control over those subordinates. The commander's acts or omissions need not be the entire cause of the subordinate's war crime; it suffices that the commander's behaviour was a significant contributing factor.

In the case of prevention, a causation may seem anomalous. In both Blaškić93 and Orić,94 the ICTY noted the impossibility of a commander's failure to punish a subordinate's antecedent crime retroactively “causing” that crime. The obvious solution would be to treat causation as prospective, in the sense that failure to punish may contribute to future war crimes by subordinates. Given the crucial function of punishment as a general deterrent, it would be difficult to imagine a situation in which the toleration of past war crimes by subordinates would fail to signal an equal toleration of any future war crimes they might be contemplating.95 As will be discussed below, however, even the failure to punish a subordinate's isolated past crime, with no possibility of future repetition, plays a role in causing injury relating to that crime.

Command responsibility as a distinctively international doctrine

“Reason to know” of war crimes by subordinates

The inconsistencies and contradictions in the command responsibility doctrines of treaties and statutes, international criminal jurisprudence and custom open a wide door to confining responsibility for war crimes to the direct actors while exonerating officers who may have subtly pressured subordinates to commit war crimes, intentionally or negligently communicated tolerance of war crimes, or simply shown no interest in preventing or punishing war crimes. They have also frustrated military and international lawyers seeking to clarify the commander's concrete responsibilities under the LOAC.96

The doctrinal points on which consistency is most elusive are the extent of the commander's obligations (1) to investigate any partial information actually presented to the commander which suggests that subordinates are planning or committing, or have committed, war crimes, and (2) to pre-empt war crimes by subordinates through training and the implementation of systems and procedures to detect, prevent and punish war crimes. The first point calls for a more nuanced interpretation of the “reason to know” prong of the commander's scienter. The second asks the fundamental question of whether a military commander satisfies their legal obligations by relying on colleagues, subordinates, the media or other sources to bring war crimes to their attention, or, more broadly, whether the commander has any affirmative duty to train and supervise their subordinates in order to ensure that they are not committing war crimes and have not committed war crimes with impunity. We will address the first point in this section and the second further below.

Command responsibility in context

To seek a reconciliation between these doctrines, the temptation to turn to municipal criminal laws for parallels has proved great. Such laws suggest that responsibility for a subordinate's war crimes must depend on the commander's own conscious intentions to further the crime through encouragement, or at least conscious inaction. A commander's mere passive failure to supervise the subordinates or negligence in investigating war crimes, in this view, does not involve sufficient moral culpability to characterize the commander as a war criminal himself or herself. Instead, such failures should be referred to the municipal disciplinary code of the officer's own country for whatever action that country's military authorities might wish to take, if any.97

Variations on this argument have been advanced by numerous scholars98 and approved by some international criminal tribunals,99 based on the claim that such liability is unknown in municipal criminal law. Municipal law, they believe, does not recognize criminal responsibility without a specific intent to further the criminal act, because an actor's criminal responsibility is tied strictly to his or her mens rea.100 Some have gone further and claimed that command responsibility requires a military commander to actually or constructively approve or at least acquiesce to the crimes of subordinates.101 Others have argued as well that a commander's approval of a subordinate's war crime ex post facto and failure to prosecute it should not engage the commander's criminal responsibility for the war crime itself due to the lack of contemporaneous mens rea.102

The claim that municipal criminal laws never recognize criminal responsibility without specific intent is factually inaccurate,103 but the flaw in these arguments runs much deeper. The analogy between command responsibility in international law and the requirement of mens rea in municipal criminal law is necessarily a false one, because nearly every aspect of the context in which war crimes occur is radically different from any context in a municipal criminal law setting. The idea that the mens rea of command responsibility must conform to municipal criminal law concepts is indeed based on a basic misconception about international law itself.104 Command responsibility is not a creature of municipal law; it arose in response to a perceived need to create disincentives for military commanders to order, encourage or tolerate the war crimes of subordinates in the specific context of armed conflict, which is to say, a breakdown of civilized social behaviour.105

The commander plays a unique role in that context. A commander's responsibilities under international criminal law are embedded in an exceptionally comprehensive hierarchical order that exists nowhere in municipal law. To the extent that the superior officer acts within the framework of the command structure and the military organization's legal order (and sometimes when acting outside of that framework), the commander's authority over his or her subordinates is absolute in a sense unknown to municipal law.106 The obedience of military subordinates to the lawful orders of their commanders is a basic tenet of military service, inculcated throughout a soldier or officer's training, and always enforced by criminal penalties.107 In most State military organizations, insubordination or disobedience may be punished severely.108 To mitigate the risks of insubordination and war crimes, in most countries commanders are obligated by military law to supervise the performance of their orders by subordinates and to adjust their orders to account for changing circumstances.109 Commanders are also responsible for assessing risks to their subordinates during operations for purposes of force preservation.110 Understandably, no comparable relationship exists in a municipal criminal law context.

A propensity for obedience is not the only characteristic engrained in military troops that increases the risk of war crimes; military personnel are also systematically trained to develop aggressive personality traits and desensitized to lethal violence against other human beings.111 Their training may also entail exhortations to subordinates that are intended to dehumanize opposing combatants (and, in some cases, all persons of the same nationality, ethnicity, religion or political ideology). Such communications serve the need of reducing moral doubt in subordinates about the killing of fellow human beings, but they may easily trigger tribalistic instincts of fear and hatred that can result in indiscriminate killing, torture, or other war crimes against opposing combatants or civilians belonging to the same group.112

The circumstances of armed conflict are comparably dissimilar to the circumstances under which municipal crimes occur. In the context of combat, the commander's responsibility is concomitant to his or her legal authority to direct the extrajudicial killing of other human beings. As one jurist put it,

intensified legal obligations are commonly placed upon persons who engage in inherently dangerous activities. The military commander is entrusted with the inherently dangerous activity of supervising persons with training in violence who have access to weapons and other equipment to carry out violence, and who have undergone indoctrination to reduce their inhibitions against violence. The law grants the commander privileges, but it also requires her to be vigilant in remaining informed and taking measures to prevent and repress violations. Thus, the commander entrusted with such an inherently dangerous activity cannot argue that she was “merely” criminally negligent in creating her own ignorance. Her indifference, in the context of her responsible relation to a clear public danger, is, arguably, sufficiently blameworthy in a desert-based account.113

The commander–subordinate relationship precludes any useful analogy between municipal and international criminal law.114 Colonel William Eckhardt observed decades ago that “[t]he wisdom of civilian law never really contemplated the judging of criminal actions in battlefield related circumstances”.115 Colonel Kenneth A. Howard later amplified this notion, stating that “[d]omestic law has not been required to contemplate a military commander's duty in a battlefield situation to control and regulate the actions of his subordinates short of the legal theory of principals”.116 The responsibilities and authority of a supervisor in a civilian context (such as in an employment situation) are not remotely analogous to those of a military commander, and the consequences of inadequately supervising a subordinate in the two cases are far from equivalent.117 Nearly all modern discussions of command responsibility fail to recognize these differences and are therefore inappropriately tethered to domestic analogies.


The differences between the two situations have moral as well as legal consequences. The commander's enhanced legal responsibility for the actions of subordinates, far from representing a radical departure from the ethics underlying municipal criminal law, is integral to military culture and organization. The strict parallelism between municipal criminal law and command responsibility would, in fact, undermine the structure of and justification for the LOAC itself. As Jenny Martinez has observed: “The moral logic of the law of war breaks down if the commander has no duty to acquire knowledge of what the killing machines he has unleashed and whom he ostensibly controls are doing with the power he has conferred on them.”118 Martinez has likewise argued that a reasonably prudent commander is not justified in assuming that subordinates, predominantly young men in dangerous and charged situations, armed with weapons designed for mass killing, will stay strictly within the boundaries of lawful violence without “constant monitoring” and supervision.119 To exonerate a commander who falls short of positively “acquiescing” in the crimes of subordinates120 releases the commander from his or her institutional responsibilities too easily, with what must inevitably prove disastrous consequences for civilians and persons hors de combat.

In consequence, the responsibility to actively investigate any information which may suggest that subordinates are planning or committing a war crime, or have committed one, is the minimum standard to which international law can hold a commander consistent with the moral obligation to protect civilians and persons hors de combat from war crimes. The commander's approval or even tolerance of war crimes by subordinates is, and should be, epiphenomenal, a point that both AP I and the Rome Statute support.

The false dichotomy between participation and innocence

Understanding the institutional context of military organization and operations also leads to a more nuanced appreciation of the role that commanders can play in indirectly enabling or encouraging war crimes by subordinates. The hierarchical military relationship is no different from any human relationship to the extent that it rarely confines itself to explicit communications without subtext or secondary meaning. A sharp line cannot always reliably be drawn between active and passive failures of a commander, or between negligent and reckless encouragement of war crimes.121 Ignoring the unique nature of the relationship between commander and subordinate in military organizations, and the equally unique nature of the situations in which war crimes occur, tends to result in oversimplification of the ways in which a commander can contribute indirectly to war crimes by subordinates, either purposely or unwittingly.

At least on the conceptual level, the distinction between reckless toleration of a war crime committed by subordinates and failure to adequately supervise subordinates is defensible.122 The respective consequences of each offence would seem to follow logically as well; in the case of toleration, the commander's punishment should be on par with the subordinate's due to their equivalent intentions and the commander's ability to prevent the crime by taking appropriate action. In contrast, the commander's mere failure to supervise subordinates is a dereliction of duty that may have tragic but presumably unforeseeable results, and the commander's punishment should be accordingly less severe, if it should be criminal at all.123

Yet, the dichotomy between conscious toleration and neglect of duty is often a false one. It distorts the dynamics of military command, and indeed of human communication and interaction in general, in the service of a simplistic legal doctrine. Human beings use a wide variety of techniques of signalling with each other to communicate beliefs and intentions indirectly and often indistinctly.124 Military officers are frequently well aware of these modes of communication. In approving the criminal conviction of General Jacob Smith for inciting and permitting subordinates to commit war crimes during US counter-insurgency operations in the occupied Philippines in the early 1900s, President Theodore Roosevelt (himself a veteran cavalry officer) emphasized that officers must be:

peculiarly careful in their bearing and conduct so as to keep a moral check over any acts of improper character by their subordinates. … Loose and violent talk by an officer of high rank is always likely to excite to wrongdoing those among his subordinates whose wills are weak or whose passions are strong.125

The basis for these points requires some development by reference to the social psychology research on group dynamics and implicit communication. Implicit meaning is a normal feature of human communication among persons of normal or higher intelligence. There are many reasons why a person might communicate a message, the face value of which differs from the subtext.126 In the context of war crimes, motivations might include the commander's wish for deniability of the order and the felt need to avoid triggering a moral reaction to a barbarous act by explicitly naming the act. How much more civilized it sounds to “teach him a lesson” than to “brutalize him” or “beat him to a pulp”.


In some circumstances and contexts, a commander's innuendo, joke, facial expression or body language could implicitly convey toleration or encouragement of war crimes as effectively as a direct order. Seemingly general affirmations of unit cohesion and military solidarity or affirmations to subordinates of trust or grants of independence can be intended and interpreted as encouragement or toleration of war crimes. In practice, orders to commit war crimes are sometimes phrased ambiguously with no diminution in the clarity. For example, during the Second World War, Adolf Hitler instructed his high command that armies of the Soviet Union were to be “annihilated”, which was interpreted to apply not only to the weakening of Soviet military power but to the extermination of the Russian people.127 During Operation Barbarossa, commanders of the Einsatzgruppen used similar language to subordinates, stating during evacuations that “nothing could be done” with burdensome old and sick persons, who subordinates then murdered in order to avoid the inconvenience of transporting them away from the zone of operations.128 Similarly, irresponsible statements by US military commanders at Haditha, Iraq, in 2005 “created a climate that minimized the importance of Iraqi lives”, which likely contributed to the war crimes committed by US troops there.129

By similar means, the US war crimes in Afghanistan and Iraq resulted from commanders giving orders focused on the desired results, such as obtaining information, along with hints and innuendos which suggested that the means for obtaining those results mattered little (“You have carte blanche”; “Soldiers are dying, get the information”; “Do whatever is necessary”).130 Statements of commanders to subordinates portraying the military mission or objectives as all-important may be designed to order a disregard of the LOAC and may be heard as such. For example, statements by President George W. Bush, Vice-President Dick Cheney and Secretary of Defense Donald Rumsfeld insisting that US interrogators of prisoners of war and detainees must obtain “results” from the interrogations were interpreted by subordinates as commands to engage in torture for that purpose, and they may have been intended to communicate precisely that.131

Euphemism has been a particularly effective means of conveying a commander's wishes for subordinates to commit war crimes and is especially useful for dehumanizing potential war crime victims, creating the psychological conditions for war crimes without the need for direct orders. Such techniques were used by the Nazis during the Second World War, Hutu genocidaires in Rwanda, and US commanders in Abu Ghraib.132 For example, rather than explicitly inciting genocide, Hutu leaders and influencers referred to the need to “go to work” or “sweep the dirt outside”.133 US commanders referred to “enhanced interrogation” of “unlawful combatants” to communicate their desire that detainees be treated inhumanely.134

It could be argued that the fault for interpreting such instructions as orders to commit war crimes lies with subordinates, and indeed in some cases commanders may use ambiguous or dehumanizing language with no intention of encouraging war crimes. However, as the quoted language from President Roosevelt suggests, irresponsible statements by officers carry a known danger. Even when no intention to order war crimes can be shown, military commanders must realize that subordinates will tend to interpret vague or ambiguous orders in light of the specific characteristics inculcated in military personnel through training and in light of the perilous combat situation they are experiencing.

Psychological pressures to obey authority exist even without the additional pressures of military training and combat stress;135 most human beings are primed by nature to obey the commands of perceived authorities.136 But obedience obviously exerts a far more powerful pull in the military context, where, as noted, that trait is systemically drilled into combatants until it becomes nearly instinctive.137 Human beings are also strongly inclined to conform to group opinions and behaviour in order to avoid unpopularity or ostracism as a minority,138 another trait that assumes exaggerated importance in the military context.

This dynamic may be particularly pronounced in elite regiments, where unit cohesion, solidarity and the notion of the unit being separate and superior to “regular” soldiers is particularly pronounced. In the psychological literature, this dynamic is known as “deviant cohesion”;139 it occurs when sub-unit solidarity leads to a breakdown of command pathways as superior organizational goals are undermined by the sub-unit, whose members feel a greater loyalty to one another than to the hierarchy and its mission goals.

Deviant cohesion is not an uncommon phenomenon in military contexts,140 and cannot be dismissed as individual misconduct by “lone wolf” soldiers with psychological disorders that could not be known by the blameless commander. Instead, it is a common form of military misconduct, in which the “actors involved believe that their misconduct was serving some military purpose, which is notably perceptible in the ways in which they frame what they had done at the time and retrospectively”.141 Indeed, the very nature of elite units seems designed to foster deviant cohesion – the elite soldier, intentionally separated from his or her fellow “regular” soldiers, is consistently reminded of their special status, and frequently given more situational autonomy than their comrades. Elite units in the field “rely less on formal authority and more on personal rapport, fostering a more informal approach to leadership. This relative autonomy can also become a double-edge sword, however, as it can create a permissive environment favouring misconduct.”142 Exonerating commanders who fail to counteract such pressures, or indeed who foster them, misses an important opportunity to deter war crimes by subordinates.

The hallmarks of deviant cohesion can be clearly seen in the Brereton Report. There is an acceptance, both within the unit and without, of the SAS as being somehow above or separate to the rules that other soldiers must follow. Throughout the Report, repeated reference is made to patrol commanders being considered infallible, and to the notion that the duty owed by the soldier was to the commander and not to the mission or the law. For example, the Report observes that

to a junior Special Air Service Regiment trooper, the patrol commander is a “demigod”, and one who can make or break the career of a trooper, who is trained to obey and to implement their superior commander's intent. … [T]o such a trooper, who has invested a great deal in gaining entry into Special Air Service Regiment, the prospect of being characterised as a “lemon” and not doing what was expected of them was a terrible one, which could jeopardise everything for which they had worked. … … … Some domestic commanders of Special Air Service Regiment … embraced or fostered the “warrior culture” and empowered, or did not restrain, the clique of non-commissioned officers who propagated it. That responsibility is to some extent shared by those who, in misconceived loyalty to their Regiment, or their mates, have not been prepared to “call out” criminal conduct or, even to this day, decline to accept that it occurred in the face of incontrovertible evidence, or seek to offer obscure and unconvincing justifications and mitigations for it.143

This seeming acceptance of a culture of deviant cohesion was evidenced at higher command levels also:

[C]ommanders trusted their subordinates: including to make responsible and difficult good faith decisions under rules of engagement; and to report accurately. Such trust is an important and inherent feature of command. However, an aura was attached to the operators who went “outside-the-wire”, and whose lives were in jeopardy. There was a perception – encouraged by them and accepted by others – that it was not for those “inside-the-wire” to question the accounts and explanations provided by those operators. This was reinforced by a culture of secrecy and compartmentalisation in which information was kept and controlled within patrols, and outsiders did not pry into the affairs of other patrols. These matters combined to create a profound reticence to question, let alone challenge, any account given by an operator who was “on the ground.” As a result, accounts provided by operators were taken at face value, and what might at least in retrospect be considered suspicious circumstances were not scrutinised. Even if suspicions were aroused in some, they were not only in no position to dispute reported facts, but there was a reticence to do so, as it was seen as disloyal to doubt the front line operators who were risking their lives.144

Group solidarity represents a powerful force in military culture that can lead to greater unit cohesion and effectiveness but also to mutual support in committing, tolerating or covering up war crimes.145


Similarly, military culture is particularly adapted to “groupthink” – the situation in which high group cohesiveness and a perceived need for unanimity override an individual's independent judgment and motivation to think realistically and rationally.146 The result is overestimation of the group's judgment, closed-mindedness, and enhanced pressure toward cognitive conformity.147 In the military context of extreme threats to the group, anyone who questions an order or a group consensus about appropriate measures may be seen as disloyal and pressured into conformity through fear of ostracization or worse.148

There is, finally, notable pressure from politics and personal ambition in military organizations. As Amy Sepinwall has noted,

political expediency may … lead a commander to pass over his troops’ crime; where, for example, support for the military effort is waning, a commander may seek to avoid the negative publicity that investigation into an atrocity will undoubtedly invite. Then again, a commander may be motivated to forego punishment not for the sake of some larger national goal, but instead for the sake of personal ambition and, in particular, a fear that his subordinates’ offense will taint his future professional prospects.149

Career ambition of this kind apparently joined with instincts for unit cohesion to prevent reporting of the war crimes by the Australian SAS. According to the Brereton Report:

It is evident that fear of the consequences of reporting misconduct to the chain of command has deterred some from doing so. In most cases, this is fear for career prospects, although in some there has been fear of physical reprisals. In any event, experience shows that where a complaint or report is adverse to a member's chain of command, there are powerful practical constraints on making it.150

The result was that junior officers aware of suspicious practices by subordinates refrained from alerting superiors or investigating the circumstances vigorously themselves.


Combined, the psychological forces described here can exert intense pressure on service personnel who interpret vague or ambiguous communications from commanders as authorizing or ordering war crimes, and who feel compelled to execute the perceived will of the commander. Service personnel can be expected to face difficulty resisting such pressure, and this may have some explanatory power for the disturbing frequency of war crimes committed without explicit orders from a commander.

The doctrine of command responsibility, as interpreted in much scholarship and the appellate jurisprudence of international criminal tribunals, thus overlooks that a commander who “does nothing” can clearly signal a message to subordinates, just as Sherlock Holmes concluded that the dog's silence carried more meaning than if it had barked.151 Omission can contribute to causing a subordinate's war crime as effectively as committing an act of complicity, as Colonel Kenneth A. Howard observed long ago.152 The ICTY alluded to this possibility in Blaškić, where it noted that the failure to prevent or punish a subordinate's war crime “conveys some tolerance or even approval on the part of the commander towards the commission of crimes by his subordinates and thus contributes to encouraging the commission of new crimes”.153 Similarly, in Aleksovski, the Trial Chamber observed that, although a commander's presence at the scene of a subordinate's war crime does not automatically indicate encouragement of the crime,

the presence of an individual with uncontested authority over the perpetrators of the unlawful act may, in some circumstances, be interpreted as approval of that conduct … [taking] into account the accused's prior or concomitant behaviour or statements …. Moreover, it can hardly be doubted that the presence of an individual with authority will frequently be perceived by the perpetrators of the criminal act as a sign of encouragement likely to have a significant or even decisive effect on promoting its commission.154

As noted, such methods of communication are not necessarily malevolent. The concept of wilful blindness is too narrow to capture the full range of circumstances in which supervisory negligence could encourage subordinate war crimes. A commander might avoid asking questions or reviewing reports in which subordinate war crimes might be revealed, not from a conscious desire to approve of such war crimes but from other motivations, such as anxiety that knowledge would result in personal responsibility or guilt, fear of unpopularity with subordinates, or a desire not to draw negative attention from superiors. Tales of this form of command failure and “misguided loyalty” to the unit, fuelled by anxiety at rocking the boat, suffuse the Brereton Report.155 As Peter Rowe has stated:

At the command level misconduct in the form of failing to deal with allegations of misconduct by those lower in the chain of command may be due to personal reasons, to misplaced loyalty to superior commanders or to an old-fashioned attempt to cover up alleged wrongdoing. Each is likely to encompass fear for one's reputation, career or promotion prospects. In any event, it is likely to be a very uncomfortable process for commanders.156

Yet, motivations for not investigating suspicious facts are doctrinally irrelevant once the commander has reason to know that war crimes might be committed. It is reasonable, then, to ask why blindness must be “wilful” to trigger command responsibility.


Although it may seem that the foregoing discussion does not apply to a failure to punish a subordinate's past war crimes, because the failure cannot necessarily be viewed as implicitly communicating approval of future behaviour, there are compelling reasons to view the commander's behaviour as potentially criminal in such cases as well. The most obvious situation, and the one most readily admitted to satisfy the exigencies of municipal criminal law, is that the commander's failure to punish subordinates guilty of war crimes may embolden them to repeat their crimes by signalling implicit approval, thus contributing causally to future crimes.157

Some believe this logic cannot extend to a subordinate's isolated war crime with no possibility of repetition.158 Whether this position is justifiable depends on how the injury from a war crime is conceived. Although the immediate and direct effect of the war crime cannot be enhanced or facilitated retroactively, it is reasonable to view the broader injury caused by a war crime, both to the victim and to the rule of law, as continuing until justice is visited upon the war criminal. Amy Sepinwall has argued that “expressive injuries” resulting from unpunished war crimes cause harm that cannot be ignored without undermining the humanitarian function of the LOAC.159 As long as the commander fails in his or her duty to punish the subordinate, a war criminal escapes punishment, the victim's interest in justice goes unsatisfied, and the LOAC's force and authority are degraded; even a failure to prosecute an isolated war crime produces significant material and moral harm that justifies punishing the commander as partly responsible for the war crime's effects, if not the war crime itself. Holding a commander responsible for failing to punish a war crime thus serves the interests of the international community by providing a general deterrent for future war crimes, upholding the rule of law, and vindicating the interests of the victims in ensuring that their abusers do not escape justice.

Does a commander's failure to train and supervise subordinates make the commander complicit in their war crimes?

The discussion of the lex lata of command responsibility above noted that no international treaty or criminal tribunal has unambiguously endorsed the criminal responsibility of military commanders under international law for prospective or past war crimes of subordinates when the commander possessed no information about those crimes, but when the commander's failure to adequately train subordinates in the LOAC and to supervise their compliance with the LOAC resulted in, or contributed to, the commission of the war crimes by subordinates. The large preponderance of jurisprudence and most of the academic commentary treat an officer's neglect of such duties as a matter for internal military disciplinary proceedings at most, not a subject for international criminal law.160 The idea of a commander having a legal duty under the LOAC to supervise subordinates was explicitly rejected by the ICTY Appeals Chamber in the Čelebići case:

Neglect of a duty to acquire [knowledge of war crimes], however, does not feature in the provision [ICTY Statute Article 7(3)] as a separate offence, and a superior is not therefore liable under the provision for such failures but only for failing to take necessary and reasonable measures to prevent or to punish. … The point here should not be that knowledge may be presumed if a person fails in his duty to obtain the relevant information of a crime, but that it may be presumed if he had the means to obtain the knowledge but deliberately refrained from doing so. … [A]lthough a commander's failure to remain apprised of his subordinates’ action, or to set up a monitoring system may constitute a neglect of duty which results in liability within the military disciplinary framework, it will not necessarily result in criminal liability.161

Similarly, the ICTR Appeals Chamber has insisted that only an officer who deliberately fails to prevent or punish war crimes by subordinates, or who “culpably or willfully” disregards his or her duty to prevent or punish, may be held liable.162 As a result, “criminal negligence is not a basis of liability in the context of command responsibility”.163


Indeed, one jurist has gone so far as to assert that because failure to supervise subordinates is an “omission”, it “can never, strictly speaking, be causal to an effect. Ex nihilo nihil fit.164 To hold otherwise, some believe, would inevitably subject commanders to a repugnant standard of strict liability for the war crimes of their subordinates.165 This claim is of course an overstatement – neglect of duty and strict liability are mutually exclusive concepts – but it arises from a healthy concern with ensuring that military commanders are not maligned and punished as war criminals for a failing that is not a proximate cause of the war crime but is at most a contributing factor and, from an evidentiary standpoint, a speculative one.166 In many cases it may be impossible to know whether adequate training and supervision would have prevented subordinates from committing a war crime, because this requires considering a counterfactual situation. The danger of subjecting a commander to unfair hindsight bias is considerable.

Certainly, it would be unrealistic and counterproductive to expect a commander's omniscience with regard to the activities of his or her subordinates. As the Nuremberg Tribunal observed, “[a] high commander cannot keep completely informed of the details of military operations of subordinates and most assuredly not of every administrative measure”.167 Moreover, “the distinction between excusable and culpable lack of information may be a fine line in practice”,168 and fine distinctions supply a precarious foundation for criminal liability.

Yet, much worse than holding commanders to an exacting standard of supervision is releasing commanders from the obligation to protect defenceless persons against whom those same commanders unleash lethal violence. The commander's duty to prevent or punish war crimes begins not with discrete acts of ordering, tolerating or consciously ignoring war crimes, but with the very tone and attitude the commander takes toward the LOAC and military professionalism. As General Douglas MacArthur once observed: “Soldiers of an army invariably reflect the attitude of their general. The leader is the essence.”169 The same may be said mutatis mutandis of lower-ranking officers relative to subordinates under their command.

The treaties and statutes articulating the law of command responsibility are at best vague on the relevance of tone, training and supervision, and the jurisprudence of the international criminal tribunals has not treated the question of whether a commander created a culture of compliance with the LOAC as a crucial factor in command responsibility analysis. Indeed, the Čelebići case approach to protecting civilians naively treats commanders as somehow hermetically sealed away from subordinate war crimes short of an intrepid informant alerting them to the facts about a future, contemporaneous or past war crime. The neglect of such factors leads to a doctrine that privileges military commanders with the right to command their troops to kill and maim without assuming any responsibility to ensure that they do so legally, then treats them as blameless when their fecklessness results in horrific acts by subordinates. This was the reasoning that led the Trial Chamber in Blaskić to insist that the role of commanders “obliges them to be constantly informed of the way in which their subordinates carry out the tasks entrusted them, and to take necessary measures for this purpose”, when there is a danger of subordinates committing war crimes170 – but that decision was overturned on appeal.

The law of command responsibility cannot achieve effective deterrence while conceiving of officers as mere passive participants in the organized violence of their subordinates. No well-regulated army functions in such a manner. Enforcing military discipline and compliance with the LOAC is not a last resort taken only after great hesitation and deliberation, but a normal and continuing obligation of military commanders. An officer in any military organization is required to ensure that soldiers under his or her command are trained in the LOAC; that they are issued rules of engagement (ROE) cards and reminded of their duties; that treatment of detainees and prisoners is properly supervised; and that systems are adopted for supervising compliance with the ROE and LOAC generally, for reporting breaches of conduct, and for the detection and punishment of war crimes.171 Failure to do so not only increases the risk of the commission of war crimes by subordinates through ignorance of the LOAC or through deception by soldiers intent on committing war crimes; it also fails to acculturate subordinates to condemnation of war crimes, as the ICTY Trial Chamber suggested in Blaskić.172 And, as the discussion above indicates, such a failure in training and supervision may in fact reflect a commander's unspoken approval or tolerance of war crimes by subordinates.

The commander, who sets the tone of the military organization under his or her command and who structures the lives of subordinates in a manner that can communicate either hostility or apathy toward the LOAC or conscientious respect for it, thus cannot be treated as immaterial when ill-trained or unsupervised subordinates commit war crimes. It would show slight respect for the LOAC if the mere ignorance of specific war crimes exculpated a commander who created or substantially contributed to the conditions that made such crimes likely.

In assessing a commander's responsibility for the war crimes of subordinates, it is therefore relevant to inquire, at a minimum, what type of training was ordered or given by the commander, what behavioural expectations were communicated directly and indirectly to subordinates, and whether any statements were made which might suggest that opposing forces are unworthy of respect or rights, or that specific results must be obtained regardless of the means.173 If General MacArthur's dictum is accurate, a commander may contribute to creating a culture of tolerance for war crimes by subordinates even before obtaining any relevant information about a subordinate's planned or past war crimes. As Mills has observed, “most war crimes are not only individual acts of atrocity. They are also command failures.”174

Although modern international criminal jurisprudence rejects the idea that commanders become responsible for the war crimes of their subordinates merely by inadequate supervision, it is noteworthy that failure to supervise subordinates was a factor in assessing command responsibility in the early international criminal law jurisprudence. As mentioned previously, the IMTFE treated a high commander's dereliction of duty as sufficient grounds for command responsibility in the Yamashita, Toyoda and Araki trials, and the Allied military tribunals in Europe did the same in the Hostages case and the Rauer trial. In the Roechling case, the Superior Military Government Court of the French Occupation Zone in Germany observed, albeit with too broad a brush, that it is a commander's “duty to know what occurs in his organization, and lack of knowledge, therefore, can only be the result of criminal negligence”.175 More recently, the Rome Statute of the ICC mentions a commander's duty to “exercise control properly over” forces under his or her command as a factor in command responsibility, although this has not been interpreted as a major factor in command responsibility analysis.176 And as a municipal law implementing international criminal law, the Canadian Crimes against Humanity and War Crimes Act provides for commander responsibility for a commander's failure “to exercise control properly” over persons under his or her effective command, and for not only knowing but being “criminally negligent in failing to know” of a planned war crime or a war crime in the process of commission.177

It does not follow that a general rule should be recognized presuming that commanders are aware of war crimes by subordinates.178 However, what this analysis does suggest is that a commander who fosters a culture of disregard for the LOAC or implicitly communicates tolerance of war crimes by a serious failure to train and supervise subordinates under his or her direct command may contribute substantially and concretely to the commission of war crimes by subordinates. The fact that the LOAC imposes no responsibility on a commander who, through various indirect failures, creates conditions propitious to the commission of war crimes leaves a gap in the doctrine that is dangerous to civilians and persons hors de combat. The circumstances and allegations described in the Brereton Report, though still pending further criminal investigation, may prove to illustrate this dynamic and its tragic consequences.

Towards a more nuanced regime of command responsibility

The optimal standard of indirect command responsibility

The Brereton Report, with its disjunction between the suspicious behaviour of subordinates who committed war crimes and the exoneration of commanders reluctant to investigate the evidence, provides a timely reminder of the need for reform of the international law of command responsibility. It illustrates how national military organizations are often quick to excuse commanders who indirectly contribute to war crimes by subordinates, and how a consequential gap in the law of command responsibility can be used to justify that exoneration. The costs of holding military commanders to such a relaxed standard of duty are unacceptably high, not only to the LOAC but to the functioning of military organizations themselves. The backlash from high-ranking military officers after President Donald Trump's pardoning of convicted war criminals and granting of amnesties to accused war criminals in 2019 and 2020179 illustrates how deference to soldiers by those charged with enforcing the LOAC not only puts innocent civilians, detainees and prisoners of war at risk of horrendous crimes, but also undermines justice and the military order. As these generals knew, armies thrive not on sycophancy but on discipline, and command responsibility is an essential component of that discipline.

The difficulty of balancing fairness to military commanders with recognition of the indirect role they may play in promoting war crimes is surmountable with a comprehension of the multifaceted role of military commanders in controlling their subordinates. We reject the fatalism of exonerating military commanders from responsibility because of the difficulty of predicting when subordinates may commit war crimes.180 The difficulty of an important task is no excuse for neglecting it, particularly for those engaged in a profession reliant on lethal force. The risk that soldiers will commit war crimes is present in every armed conflict, and therefore the risk of war crimes is always foreseeable in a general way.181 This does not mean that every individual war crime is foreseeable or preventable by a higher officer. What it does mean is that the law of command responsibility is hampered by an all-or-nothing mentality that is incompatible with the realities of military command. A person threatened by lawless soldiers has no reason to care whether their commander acquiesced in war crimes, felt no concern, refused to believe the evidence, or was too preoccupied to be bothered. The result is the same for the victims.

As discussed, the desire to remould the law of command responsibility in the image of municipal criminal law arises from a misapprehension that ignores the unique characteristics of military training and the dynamics of military organizations, especially those intrinsic to the relationship between commanders and subordinates. The violent nature of armed conflict and the commander's relatively comprehensive control over direct subordinates under circumstances that carry an inherent risk of war crimes justifies holding commanders to a high standard of diligence and care with regard to restraining the violence that they are charged with unleashing on others. Military commanders are usually the only significant restraint on war crimes by soldiers and officers beneath them. Any attempt to relieve them of the obligation to create a culture of compliance with the LOAC, much less to investigate any information suggesting that subordinates may commit or have committed war crimes and to prevent and punish such crimes, leaves civilians, war prisoners and other defenceless individuals with very little protection against the barbarities of which soldiers have historically proved capable.

This article has also explored the dynamics of military command, including the implicit means of communicating approval or tolerance of war crimes available to commanders. The common assumption that commanders who overlook the war crimes of subordinates are merely remiss in their supervisory duties and not active participants in those crimes imposes a false dichotomy on a situation that is frequently much more fluid, ranging from indirect but clearly understood expressions of approval for war crimes to passivity motivated by personal ambition, denial of reality, or cowardice at the possibility of exposing an alarming truth.

At the same time, a pure standard of criminal negligence will be too strict to hold a commander responsible for the war crimes of subordinates in at least some situations. Notwithstanding the justifiability of expecting commanders to meet a very high standard of training and supervising their subordinates, it would be disproportionate to hold a commander responsible for such war crimes as torturing a prisoner or murdering a civilian because the commander did not always thoroughly train and supervise all those under his or her direct command. The problem is not necessarily that no moral theory can justify a criminal penalty against the commander under such circumstances. As observed, a military commander who neglects his or her training and supervisory duties can reasonably foresee that subordinates may try to commit war crimes; relying merely on faith in their moral probity is dangerously naive at best. Yet, negligence or neglect of duty is not a forgiving standard, while war crimes are some of the most reprehensible acts a human being can commit. It would be disproportionate to visit the most extreme penalties of international criminal law on a negligent commander and an active participant in a war crime in equal measure.

The optimal approach, in light of the dynamics of military organization and human psychology discussed here, is to hew to an interpretation of “reason to know” similar to that adopted by the ICTY in Čelebići. Specifically, a commander should be equally responsible for the war crimes of subordinates if he or she disregards or inadequately investigates any evidence suggesting that subordinates are involved in war crimes. Such evidence must be treated with the utmost gravity and should never be dismissed as inconsistent with what the commander believes or wishes about subordinates, as not credible because it is contrary to the commander's bare assumptions, or as inconvenient for the cohesion of the unit or some other objective.

In effect, this is a kind of recklessness standard, in which the concept of recklessness involves failing to observe a strict duty to investigate adequately any evidence of subordinate war crimes, but not a duty to constantly monitor subordinates proactively in order to ensure that they never commit war crimes. In 1973, Major William Parks published an article proposing that the standard of culpability for a commander should be “wanton negligence”, presumably meaning something like recklessness, in order to hold the commander responsible for aiding and abetting a war crime.182 Punishing a commander who does not share the culpable mentality of his or her subordinates to at least some degree would brand as a war criminal an individual who is merely incompetent, poorly trained and supervised as an officer, or overburdened and distracted. Parks’ proposal is thus consistent with our reasoning in a general way. A reckless disregard standard is morally justifiable because of the unique characteristics of military command in a way that neither a negligence standard nor an inferred acquiescence standard captures.

At the same time, nothing in this standard justifies requiring direct evidence of a commander's overt tolerance or approval of war crimes. As discussed above, there are many ways in which a commander can communicate implicit tolerance of war crimes by subordinates. Omission can be as influential as suggestion, and a euphemism can be intended and interpreted as an order. The only practical case that defenders of a higher bar for command responsibility have made is that if commanders are held responsible for training and supervising subordinates against war crimes by punishments comparable to those given to subordinates, it may be difficult to recruit officers.183 Such in terrorem arguments are speculative on the facts, but worse, they attribute more moral weight to the exigencies of military staffing than to the potential commission of war crimes. There is reliable evidence showing that proper training is effective at reducing the risk of war crimes;184 there is no such evidence to suggest that holding commanders responsible for properly training and supervising their troops would cause any military organization to collapse.

The necessity of training and supervision

As also noted, failure to create a culture of compliance with the LOAC, including adequate training and supervision of subordinates, is one potentially telling indicator of the commander's complicity in subordinate war crimes. It should therefore assume a more prominent role in command responsibility analysis. This is not to argue that such neglect should constitute a new, independent basis for a war crime, but the claim that a commander's general duty to supervise subordinates and to assiduously investigate allegations of planned or past war crimes unfairly penalizes commanders who “failed to keep properly informed” skews the moral calculus indefensibly.185 It privileges fairness to one actor, who has assumed exceptional responsibility over the safety and lives of the entire universe of enemy combatants, as well as defenceless civilians and persons hors de combat, over innocent persons whose lives could be seriously degraded or simply ended through the commander's negligence. There is no credible moral argument for striking the balance in the commander's favour in such circumstances.186 It is not enough for a military commander to “assume that” subordinates “would properly perform the function which had been entrusted to them by higher authorities”.187 True, high command is not field supervision – but no military commander has the “right”188 to trust in the unimpeachable integrity and legal scrupulousness of subordinates, particularly in the necessarily perilous context of armed conflict, because that right will be enjoyed at the potential expense of defenceless persons.

Limiting command responsibility for war crimes to cases of reckless disregard and ignoring the military culture created by commanders leaves a gap in international criminal law for commanders who use implicit means to communicate encouragement or toleration of war crimes by subordinates, and for commanders who through carelessness or incompetence put at risk the lives of civilians and persons hors de combat. The consequence of mitigating the punishment of superior officers in cases of serious negligence is to trivialize their moral and legal responsibility to ensure that armed subordinates perpetrating deadly violence, and their superior officers directly responsible for supervising them, comply with minimally civilized standards of behaviour. Indeed, it creates an incentive for unscrupulous generals to informally pressure lower-ranking officers to commit war crimes while preserving plausible deniability, with the greatest risk being an anodyne accusation of “dereliction of duty” under municipal military law that may or may not result in some form of discipline within the commander's own military organization.

To address this gap would require a significant innovation in international law beyond firmly embedding evidence of command neglect as an important factor in command responsibility analysis in international criminal law. One additional option would be to create a lesser charge under international law for a commander who inadequately trains or supervises troops under his or her direct command, when those troops subsequently commit a war crime.189 Colonel Howard observed that a commander who negligently allows a subordinate to commit premeditated murder in violation of the LOAC may be charged with involuntary manslaughter under a military code.190 Although international criminal law does not recognize an involuntary manslaughter charge, the introduction of a lesser offence of this kind, such as “gross neglect of duty”, might be worth considering. To mitigate the risk of hindsight bias, such a charge should require strong evidence of a substantial and sustained neglect of training and supervision, as opposed to training and supervision that are merely considered inadequate in light of the war crimes actually committed by subordinates.

It should be clear that the idea of a lesser offence of neglect of duty would in no way overlap with cases in which the commander participates in a war crime. When the commander had information that should have put him or her on notice of a possible war crime and failed to take preventive or punitive action, command responsibility is the appropriate paradigm. The proposed lesser offence becomes relevant only when the commander had no reasonable notice that a subordinate was planning or committing a war crime.

Alternatively, or in addition, the LOAC could be amended to require States to ensure that commanders properly train and supervise subordinates, to require that States implement auditing or other supervisory systems to assess and correct a commander's neglect in this regard, and to hold States responsible for failing to demote or discharge commanders found guilty of such neglect. Such requirements would have to operate independent of any actual crimes committed by subordinates, because prevention of war crimes will always be more desirable than punishment of those who commit or contribute to war crimes.

It may be asked why this matter should not be left solely to the disciplinary system of the national military organization. The answer is that war crimes are among the gravest offences of which a person is capable, and given the dearth of other safeguards against them, ensuring that commanders give minimally adequate training and supervision of subordinates merits international concern. Although there are counterexamples, history has repeatedly shown that States tend to be reluctant to try, much less to convict, their own military officers for war crimes. The Brereton Report, which largely does not recommend further investigation of SAS Command's failure to investigate evidence of subordinate war crimes, serves as an instructive example. The function of international criminal law is to end impunity for the most serious offences against the LOAC, and this can be accomplished most reliably not only by punishing individual soldiers who commit murders, torture and other war crimes ex post facto, but by creating the conditions that discourage the commission of war crimes in the first place.


  • 1. US Department of the Army, Army Regulation 600-20: Army Command Policy, 24 July 2020, para. 2-(b).
  • 2. Inspector-General of the Australian Defence Force Afghanistan Inquiry Report, 00 (Brereton Report), available at:… (all internet references were accessed in March 0).
  • 3. Ibid., p. 29, paras 18–19.
  • 4. Ibid., p. 31, para. 30.
  • 5. Ibid., p. 464, paras 336, 338–339; p. 490, paras 49–0.
  • 6. Ibid., p. 41, para. 77.
  • 7. Ibid., p. 31, para. 30.
  • 8. Ibid., p. 325.
  • 9. Ibid., p. 332, para. 15.
  • 10. Ibid., p. 489, para. 42.
  • 11. Rome Statute of the International Criminal Court, 2187 UNTS 3, 17 July 1998 (Rome Statute), Art. 28. Section 268. of the Australian Criminal Code Act 1995 (Cth), which is discussed in the Brereton Report, largely but not precisely replicates the command responsibility provision of the Rome Statute.
  • 12. Rome Statute, Art. 28.
  • 13. See ibid., Art. 25(3)(b).
  • 14. See ibid., Art. 25(3)(c); see also e.g. International Criminal Tribunal for Rwanda (ICTR), Prosecutor v. Akayesu, Case No. ICTR-96-4-T, Judgment (Trial Chamber), 2 September 1998, paras 693–694 (finding the mayor of Taba Commune in Rwanda guilty of aiding and abetting widespread and systematic rapes through his “words of encouragement”, which “sent a clear signal of official tolerance for sexual violence, without which these acts would not have taken place”).
  • 15. See Rome Statute, Art. 25(d). Cf. Beatrice I. Bonafé, “Finding a Proper Role for Command Responsibility”, International Journal of Criminal Justice, Vol. 5, No. 3, 2007, p. 6; Chantal Meloni, Command Responsibility in International Criminal Law, Springer, New York, 2010, pp. 216–233 (discussing the International Criminal Tribunal for the former Yugoslavia (ICTY) decisions imposing responsibility for aiding and abetting, conspiracy and joint criminal enterprise).
  • 16. For detailed historical summaries of State practice, see William H. Parks, “Command Responsibility for War Crimes”, Military Law Review, Vol. 62, No. 1, 1973, pp. 2–; Matthew Lippman, “Humanitarian Law: The Uncertain Contours of Command Responsibility”, Tulsa Journal of Comparative and International Law, Vol. 9, No. 1, 2001.
  • 17. Commission on the Responsibility of the Authors of the War and on Enforcement of Penalties, Report Presented to the Preliminary Peace Conference, Versailles, 29 March 1919, reprinted in American Journal of International Law, Vol. 14, No. 1, 1920, p. 121.
  • 18. W. H. Parks, above note 16, p. 20.
  • 19. Several authors have argued that there was credible evidence that Yamashita was aware of the war crimes (see e.g. W. H. Parks, above note 16, pp. 22–38; William V. O'Brien, “The Law of War, Command Responsibility and Vietnam”, Georgetown Law Journal, Vol. 60, No. 3, 72, pp. 625–627; William G. Eckhardt, “Command Criminal Responsibility: A Plea for a Workable Standard”, Military Law Review, Vol. 97, No. 1, 82, p. ), but the tribunal never in fact found direct evidence that Yamashita had knowledge. It held instead that the war crimes were so open, systematic and in propinquity to Yamashita's location that knowledge could reasonably be imputed to him on the facts. The tribunal put great emphasis on those duties of a commander that, if properly exercised, would have led to the discovery of the war crimes, but this emphasis is more consistent with a “reason to know” standard than with an actual knowledge standard.
  • 20. US Military Commission, Manila, Trial of General Tomoyuki Yamashita, Case No. 21, Judgment, 4 February, 1946, in UN War Crimes Commission, Law Reports of Trials of War Criminals, Vol. 4, 1948, pp. 34–35.
  • 21. Ilias Bantekas, Principles of Direct and Superior Responsibility in International Humanitarian Law, Juris, New York, 2002, pp. 113–114 (footnotes omitted). But see Guénaël Mettraux, The Law of Command Responsibility, Oxford University Press, Oxford, 2009, pp. 7–8 (interpreting ICTY precedents, including the Čelebići case decision, to exonerate commanders for whom evidence of the war crimes of subordinates is readily available, but who take no steps whatsoever to acquire it).
  • 22. See e.g. United Nations War Crimes Commission, United States v. von Leeb, in Law Reports of Trials of War Criminals, Vol. 9, 1949, p. 512. That said, some tribunals varied the phrasing of the duty of command responsibility. Thus, for example, in the Toyoda trial, the IMTFE characterized the commander's duty as one of “the exercise of ordinary diligence” or “use of reasonable diligence” to learn of the commission of crimes by subordinates. See IMTFE, United States v. Toyoda, in Records of the Trial of Accused War Criminal Soemu Toyoda, Tried by a Military Tribunal Appointed by the Supreme Commander of the Allied Powers, Tokyo, Japan, 1948–1949, National Archives and Records Administration, M1729, 7 rolls; Transcripts from the Case of the United States of America vs. Soemu Toyoda and Hiroshi Tamura, 1946–1948, M1661, 4 rolls (Toyoda Transcript).
  • 23. Protocol Additional (I) to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts, 1125 UNTS 3, 8 June 1977 (entered into force 7 December 1978) (AP I), Art. 86(2).
  • 24. Anne-Marie Boisvert, Hélène Dumont and Martin Petrov, “Quand les crimes des sous-fifres engage la responsabilité de leur chef: La doctrine de la responsabilité du supérieur hiérarchique en droit pénal international”, Canadian Criminal Law Review, Vol. 9, No. 1, 2004, p. 117.
  • 25. See ICTR, Prosecutor v. Bagilishema, Case No. ICTR-95-1A-A, Judgment (Appeals Chamber), 3 July 2002, para. 28.
  • 26. Claude Pilloud, Yves Sandoz, Christophe Swinarski and Bruno Zimmermann (eds), Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949, ICRC, Geneva, 1987 (ICRC Commentary on APs), p. 1012.
  • 27. It is thus even less justified to argue that a commander must have verified proof of a potential war crime by subordinates to incur liability for failing to prevent those crimes, and he must consciously choose not to act. See G. Mettraux, above note 21, pp. 208, 217, 223. Such an argument excuses both total neglect of supervision on the commander's part and his wilful blindness to indicators of possible war crimes, in contradiction to the plain words of AP I.
  • 28. ICRC Commentary on APs, above note 26, p. 1014.
  • 29. Updated Statute of the International Criminal Tribunal for the former Yugoslavia, September 2009, Art. 7(3).
  • 30. ICTR, Bagilishema, above note 25, para. 28 (quoting ICTY, Prosecutor v. Čelebići [sic: Delalić], Case No. IT-96-21-A, Judgment (Appeals Chamber), 20 February 2001, para. 238).
  • 31. See, generally, Elies van Sliedregt, “Command Responsibility at the ICTY: Three Generations of Case-Law and Still Ambiguity”, in Bert Swart, Alexander Zahar and Göran Sluiter (eds), The Legacy of the International Criminal Tribunal for the Former Yugoslavia, Oxford University Press, Oxford, 2011.
  • 32. Rome Statute, Art. 28(a).
  • 33. See below text accompanying note 62.
  • 34. There was insufficient published debate at the Diplomatic Conference to explain why the original “should have known” language was amended to “had information which should have enabled them to conclude”. The United States had proposed altering the phrase to “should reasonably have known”, but one can only guess as to why the “information” language was added. See Howard S. Levie, “Command Responsibility”, U.S. Air Force Academy Journal of Legal Studies, Vol. 8, 1997–98, pp. 8–9.
  • 35. Some writers have interpreted the “should have known” standard as somehow more relaxed than the “reason to know” standard. It appears that Guénaël Mettraux in particular has conflated “reason to know” with the widely accepted responsibility of a commander to supervise his or her subordinates in order to prevent war crimes, and has concluded that the former is unique to and a product of the latter, resulting in a “legal fiction” of knowledge. G. Mettraux, above note 21, pp. 77–78, 210–212. There is no support for such an interpretation. The confusion can be easily dispelled by pointing out that the commander should have known of a war crime when he or she had reason to know of it, and the commander has reason to know of a war crime when ordinary supervision of his or her subordinates produces information that would alert a reasonable person that subordinates planned to commit or had committed a war crime. No fictional imputation of actual knowledge to the commander is necessary.
  • 36. The Statute of the Special Court for Sierra Leone, 2002, uses identical language in its Article 6.
  • 37. Jean-Marie Henckaerts and Louise Doswald-Beck (eds), Customary International Humanitarian Law, Vol. 1: Rules, Cambridge University Press, Cambridge, 2005 (ICRC Customary Law Study), pp. 558–562, available at:
  • 38. ICRC Commentary on APs, above note 26, pp. 1013–1014. This interpretation of the difference in meaning between the English and French versions misunderstands the significance of the English past conditional tense, but ultimately this misapprehension does not affect the Commentary's conclusion.
  • 39. Jenny S. Martinez, “Understanding Mens Rea in Command Responsibility”, Journal of International Criminal Justice, Vol. 5, No. 3, 2007, pp. 653–654.
  • 40. ICTY, Prosecutor v. Delalić, Case No. IT-96-21-T, Judgment (Trial Chamber), 16 November 1998, para. 333.
  • 41. The case of Prosecutor v Orić expanded the test to include a fourth limb: that “an act or omission incurring criminal responsibility according to Articles 2 to 5 and 7(1) of the [ICTY] Statute has been committed by other(s) than the accused (‘principal crime’)”. ICTY, Prosecutor v Orić, Case No. IT-03-68, Judgment (Trial Chamber), 30 June 2006, para. 294. See also Tilman Blumenstock and Wayde Pittman, “Prosecutor v. Naser Orić: The International Criminal Tribunal for the Former Yugoslavia Judgment of Srebrenica's Muslim Wartime Commander”, Leiden Journal of International Law, Vol. 19, No. 4, 2006.
  • 42. Rome Statute, Art. 28.
  • 43. See ICTY, Prosecutor v. Krnojelac, Case No. IT-97-25-A, Judgment (Appeals Chamber), 17 September 2003, para. 171, in International Legal Materials, Vol. , 2004, p. 330; accord ICTY, Orić, above note 41, para. 293 (“neglect of duty”).
  • 44. See e.g. ICC, Prosecutor v. Bemba, Case No. ICC-01/05-01/08, Decision Pursuant to Article 61(7)(a) and (b) of the Rome Statute on the Charges of the Prosecutor against Jean-Pierre Bemba Gombo, 15 June 2009, para. 436.
  • 45. See also ICTY, Prosecutor v Halilović, Case No. IT-01-48-T, Judgment (Trial Chamber), 16 November 2005, para. 54; ICTY, Prosecutor v Hadžihasanović, Case No. IT-01-47-T, Judgment (Trial Chamber), 15 March 2006, paras 74–75; ICTY, Krnojelac, above note 43, para. 171 (“It cannot be overemphasised that, where responsibility is concerned, an accused is not charged with the crimes of his subordinates but with his failure to carry out his duty as a superior to exercise control”).
  • 46. See e.g. ICTY, Prosecutor v. Delalić, Case No. IT-96-21-A, Judgment (Appeals Chamber), 20 February 2001.
  • 47. AP I, Art. 87.
  • 48. Rome Statute, Art. 28.
  • 49. ICTY, Prosecutor v. Halilović, Case No. IT-01-48-A, Judgement (Appeals Chamber), 16 October 2007, para. 59.
  • 50. ICTY, Delalić, above note 46, paras 193, 197.
  • 51. ICC, Prosecutor v. Bemba, Case No. ICC-01/05-01/08-3343, Judgment (Trial Chamber III), 21 March 2016, para. 180.
  • 52. The ability to terminate employment was considered critical in the Musema case at the ICTR for determining de facto and de jure control. See ICTR, Prosecutor v. Musema, Case No. ICTR-96-13-T, Judgment (Trial Chamber), 27 January 2000, para. 880.
  • 53. ICC, Bemba, above note 51, para. 188.
  • 54. See e.g. US District Court, Southern District of Florida, Romagoza Arce v. Guillermo Garcia, Case No. 99-8364-CIV, Jury Instructions, 19 July 2002, p. 8 (unrep.), available at:….
  • 55. See W. G. Eckhardt, above note 19, p. 4; Rome Statute, Art. 25(3).
  • 56. See Jamie Allan Williamson, “Some Considerations on Command Responsibility and Criminal Liability”, International Review of the Red Cross, Vol. 90, No. 870, 2008, p. 308.
  • 57. U.S. Military Tribunal, Nuremberg, Trial of Wilhelm List and Others, Case No. 47, Judgment, 19 February 1948, p. 71, in Trials of War Criminals before the Nuernberg Military Tribunals under Control Council Law No. 10, Vol. 11, US Government Printing Office, Washington, DC, 1950, pp. 1271–1272.
  • 58. Ibid., p. 1271.
  • 59. British Military Court, Wuppertal, Germany, Trial of Major Karl Rauer and Six Others, Case No. 23, Judgment, 18 February 1946, in United Nations War Crimes Commission, Law Reports of Trials of War Criminals, Vol. 4, Her Majesty's Stationery Office, London, 1948, pp. 113–117.
  • 60. IMTFE, United States of America v. Araki and Others, Judgment, 4 November 1948, p. 48,445.
  • 61. Ibid.
  • 62. Toyoda Transcript, above note 22, Vol. 19, p. 5006.
  • 63. Nuremberg Military Tribunal, United States of America v. von Leeb et al., Case No. 72, Judgment, 28 October 1948 (High Command), in Trials of War Criminals before the Nuernberg Military Tribunals under Control Council Law No. 10, Vol. 11, US Government Printing Office, Washington, DC, 1950, pp. 543–544 (emphasis added).
  • 64. ICTR, Bagilishema, above note 25, para. 28, quoting ICTY, Čelebići, above note 30, para. 238.
  • 65. ICTY, Delalić, above note 46, para. 238. The quoted language refutes those who reject a commander's legal duty to acquire knowledge altogether and claim instead that the commander must have positive knowledge of an incipient or past crime. See G. Mettraux, above note 21, pp. 76–77, 209; B. I. Bonafé, above note 15, pp. 606–607; A.-M. Boisvert, H. Dumont and M. Petrov, above note 24, pp. 126–127.
  • 66. ICTY, Prosecutor v. Blaskić, Case No. IT-95-14-T, Judgment (Trial Chamber), 3 March 2000, para. 332.
  • 67. Ibid., para. 307 (quoting Final Report of the Commission of Experts Established pursuant to Security Council Resolution 780 (1992), UN Doc. S/1994/4, 27 May 1994, Annex, para. 58).
  • 68. ICTY, Delalić, above note 46, paras 388–393.
  • 69. Ibid.; accord ICTY, Prosecutor v. Blaskić, Case No. IT-95-14-A, Judgment (Appeals Chamber), 29 July 2004, para. 62.
  • 70. ICTY, Blaskić, above note 69, para. 406. As noted earlier, civilian commanders are held to a different standard.
  • 71. US Department of Defense, Law of War Manual, December 2016, p. 1141, §
  • 72. US military jurisprudence under the Uniform Code of Military Justice similarly frequently ignores the “reason to know” standard and requires “actual knowledge” of the subordinates’ intentions to commit war crimes for the criminal conviction of the commanding officer. See e.g. United States v. Medina, CM 427162, 1971; United States v. Flaherty, 12 CMR 466, 1953, p. 469. For criticism of this approach as incompatible with international law (and, indeed, with US military law at the time and since), see Joseph Goldstein et al., The My Lai Massacre and Its Cover-Up: Beyond the Reach of Law?, Free Press, New York, 1976; W. G. Eckhardt, above note 19, pp. 11–22; Roger S. Clark, “Medina: An Essay on the Principles of Criminal Liability for Homicide”, Rutgers-Camden Law Journal, Vol. 5, No. 1, 1973; Michael L. Smidt, “Yamashita, Medina, and Beyond: Command Responsibility in Contemporary Military Operations”, Military Law Review, Vol. 164, 2000, pp. 211–234; Andrew D. Mitchell, “Failure to Halt, Prevent or Punish: The Doctrine of Command Responsibility for War Crimes”, Sydney Law Review, Vol. 22, No. 3, 2000, pp. 396–397.
  • 73. Criminal Code Act 1995, No. 12, as amended up to 20 April 2019, § 268.115.
  • 74. UK Ministry of Defence, The Manual of the Law of Armed Conflict, 2004, §§ 16.36, 16.36.6. With uncharacteristic optimism, the Manual also asserts that, despite the various formulations of command responsibility, “there is general agreement on the nature of command and the degree of knowledge required” (§ 16.36.2).
  • 75. International Criminal Court Act 2001, § 65(2)(3), Chap. 17, 11 May 2001.
  • 76. Rome Statute, Art. 28(a)(i).
  • 77. Ibid., Art. 28(b)(i–ii). See generally Maria Nybondas, Command Responsibility and Its Applicability to Civilian Superiors, TMC Asser Press, The Hague, 2010.
  • 78. Loi No. 2010-930, 9 August 2010, Art. 7, codified in Code Pénal, Art. 462-7.
  • 79. Gesetz zur Einführung des Völkerstrafgesetzbuches, 26 June 2002, §§ 4, 13, in Bundesgesetzblatt, Part 1, No. 42, 2002, p. 2254 (authors’ translation).
  • 80. ICTY, Delalić, above note 46, para. 155.
  • 81. See Yoram Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict, 2nd ed., Cambridge University Press, Cambridge, 2010, p. 275; Bing Bing Jia, “The Doctrine of Command Responsibility: Current Problems”, Yearbook of International Humanitarian Law, Vol. 3, 2000, pp. 159–160; Diane A. Desierto, “The Contours of Command Responsibility: Philippine Incorporation and Customary Evolution”, Asia-Pacific Yearbook of International Humanitarian Law, Vol. 2, 2006, pp. 227–228.
  • 82. Cf. Daryl A. Mundis, “Crimes of the Commander: Superior Responsibility under Article 7(3) of the ICTY Statute”, in Gideon Boas and William A. Schabas (eds), International Criminal Law Developments in the Case Law of the ICTY, Martinus Nijhoff, Leiden, 2003, pp. 239, 262.
  • 83. See I. Bantekas, above note 21, pp. 116–117. But see G. Mettraux, above note 21, p. 201 (asserting that a commander's awareness of “criminal propensities among some subordinates” triggers no legal duty of supervision to prevent possible war crimes).
  • 84. ICTY, Delalić, above note 40, para. 394. Some scholars similarly despair of a general rule on the standard of conduct that should be required of commanders to address possible war crimes by subordinates: see e.g. J. A. Williamson, above note 56, p. 310.
  • 85. ICTY, Orić, above note 41, para. 329; ICTY, Čelebići, above note 84, para. 395; accord ICC, Prosecutor v. Bemba, Case No. ICC-01/05-01/08 A, Judgment (Appeals Chamber), 8 June 2018, para. 167.
  • 86. ICC, Prosecutor v. Bemba, Case No. ICC-01/05-01/08, Decision (Appeals Chamber), 8 June 2018, para. 170.
  • 87. ICTY, Prosecutor v. Hadžihasanović, Case No. IT-01-47-A, Judgment (Appeals Chamber), 22 April 2008, paras 30–31.
  • 88. See Amy J. Sepinwall, “Failures to Punish: Command Responsibility in Domestic and International Law”, Michigan Journal of International Law, Vol. 30, No. 1, 2009, pp. 294–295.
  • 89. ICTY, Krnojelac, above note 43, para. 171; accord ICTY, Orić, above note 41, para. 293 (“neglect of duty”).
  • 90. See Mirjan Damaška, “The Shadow Side of Command Responsibility”, American Journal of International Law, Vol. 49, No. 3, 2001, p. 467. This was also the position of Judge Shahabudeen in his separate opinion in ICTY, Prosecutor v. Hadžihasanović, Case No. IT-01-47-AR72, Judgment (Appeals Chamber), 16 July 2003, para. 32.
  • 91. See e.g. Rome Statute, Art. 28; ICTY, Halilović, above note 49, para. 53; International Criminal Court Act 2001, § 65; Military Commission Act of 2009, Title XVIII, § 950q, HR 2647; Final Report of the Commission of Experts, UN Doc. S/1994/674, 27 May 1994.
  • 92. As Bing Bing Jia has observed, the language of Rome Statute Article 28 conditions the responsibility of the commander for subordinate war crimes on a causative connection between the failure to properly control the subordinate and the commission of the war crime: B. B. Jia, above note 81, p. 15. A negligent commander who could not have prevented a subordinate's war crime if he or she had (counterfactually) tried to prevent it, therefore, cannot be held criminally liable under the Statute. In contrast, causation was not viewed as obligatory in the jurisprudence of the ICTY and ICTR: see ICTY, Blaškić, above note 69, paras 73 ff. The Trial Chamber in the Čelebići case posited that a causal relationship between the commander's acts and the subordinate's war crime “may be considered to be inherent” in command responsibility, but it also admitted finding no support for this proposition in customary international law. ICTY, Delalić, above note 40, paras 398–400.
  • 93. ICTY, Blaškić, above note 69, para. 83.
  • 94. ICTY, Orić, above note 41, para. 338.
  • 95. Cf. Darryl Robinson, “How Command Responsibility Got So Complicated: A Culpability Contradiction, Its Obfuscation, and a Simple Solution”, Melbourne Journal of International Law, Vol. 13, No. 1, 2000, pp. 16–17.
  • 96. See e.g. W. G. Eckhardt, above note 19; Matthew Lippman, “Humanitarian Law: The Uncertain Contours of Command Responsibility”, Tulsa Journal of Comparative and International Law, Vol. 9, No. 1, 2001; Arthur Thomas O'Reilly, “A Call to Realign Doctrine with Principles”, American University International Law Review, Vol. 20, No. 1, 2004; B. I. Bonafe, above note 15; A.-M. Boisvert, H. Dumont and M. Petrov, above note 24, p. 122; A. D. Mitchell, above note 72, p. 402.
  • 97. See e.g. Antonio Cassese, International Criminal Law, 3rd ed., Oxford University Press, Oxford, 2013; Cassandra Steer, Translating Guilt: Identifying Leadership Liability for Mass Atrocity Crimes, TMC Asser Press, The Hague, 2017, p. 264; Ilias Bantekas, “The Contemporary Law of Superior Responsibility”, American Journal of International Law, Vol. 39, No. 3, 1999, p. 577; B. B. Jia, above note 81, pp. 32–33; Stefan Trechsel, “Command Responsibility as a Separate Offense”, Berkeley Journal of International Law Publicist, Vol. 3, 2009, pp. 29–30, 34–35; cf. G. Mettraux, above note 21, pp. 38–41, 45, 223 (“To be liable, the commander must, therefore, have been aware of the criminal character of his [own] action and, with that awareness, he must have consciously decided not to fulfil his obligations”).
  • 98. See e.g. M. Damaška, above note 90, passim; G. Mettraux, above note 21, pp. 86, 221; C. Meloni, above note 101, pp. 633–636; Arthur Thomas O'Reilly, “Command Responsibility: A Call to Realign the Doctrine with Principles of Individual Accountability and Retributive Justice”, Gonzaga Law Review, Vol. 40, No. 1, 2004–05; see also A. J. Sepinwall, above note 88, pp. 255–256.
  • 99. See e.g. ICTR, Akayesu, above note 14, para. 78; ICTY, Prosecutor v. Naletilić, Case No. IT-98-34-A, Judgment (Appeals Chamber), 3 May 2006, para. 114.
  • 100. See e.g. M. Damaška, above note 90, pp. 463–467; D. Robinson, above note 95, pp. 30–31.
  • 101. See e.g. Nuremberg Military Tribunal, High Command, above note 63, pp. 543–544; G. Mettraux, above note 21, p. 42. See also Chantal Meloni, “Command Responsibility”, Journal of International Criminal Justice, Vol. 5, No. 3, 2007, pp. 621–623 (noting that the ICTY interpreted the post-Second World War legislation on command responsibility as a form of accomplice liability, and correctly noting that the Tokyo judgment took a different view).
  • 102. M. Damaška, above note 90, pp. 468–469 (“Used as a vehicle for vicarious liability, approval of a transgression is alien to the tenets of modern criminal law”).
  • 103. Many jurisdictions recognize general intent crimes characterized by criminal negligence or recklessness, such as reckless arson or driving under the influence of alcohol. Some also recognize strict liability crimes, such as statutory rape. See e.g. Danish Penal Code, No. 871, 2014, § 216, available at:; Penal Code of Japan, 2017, Art. 177, available at:; UK Sexual Offences Act 2003, c. 42, § 5.
  • 104. On the pitfalls of trying to fit international criminal law into the Procrustean bed of municipal criminal law, see generally Mark A. Drumbl, Atrocity, Punishment and International Law, Cambridge University Press, Cambridge, 2007, pp. 5–9, 38–39; Mark Osiel, “The Banality of Good: Aligning Incentives against Mass Atrocity”, Columbia Law Review, Vol. 105, No. 6, 2005, pp. 1753, 1768.
  • 105. Cf. Gerhard Werle and Florian Jessberger, Principles of International Criminal Law, 4th ed., Oxford University Press, Oxford, 2020. Surprisingly, the ICTR Trial Chamber claimed without support that command responsibility did derive from the individual criminal responsibility in dicta in the Akayesu judgment. ICTR, Akayesu, above note 14, para. 78.
  • 106. See Major Trenton W. Powell, “Command Responsibility: How the International Criminal Court's Jean-Pierre Bemba Gombo Conviction Exposes the Uniform Code of Military Justice”, Military Law Review, Vol. 225, 2017, pp. 837–838.
  • 107. See e.g. Defence Force Discipline Act 1982, No. 152, C2019C00, Compilation No. 33, 2 March 2019, § 27, available at: (Australia); Uniform Code of Military Justice, 2021, Art. 90, 10 USC § 890 (United States); Décret No. 2005-796 relatif à la discipline générale militaire, 15 July 2005, Arts 5(1), 7, available at: (France); Código de Justicia Militar, 16 July 1951, Art. 667, Sanción Ley 14029, available at:… (Argentina). See, generally, Mark J. Osiel, Obeying Orders: Atrocity, Military Discipline, and the Law of War, Routledge, London, 2017, p. 241 fn. 21.
  • 108. See e.g. Uniform Code of Military Justice, 2021, Art. 90, 10 USC § 890 (United States); Code de Justice Militaire (Nouveau), 2021, Arts L323-6 to L323-8 (France); Army Act, 1995, §§ 34, 36, 71, 85 (United Kingdom); Military Justice Law 5715, 1955, §§ 122–124, 133 (Israel).
  • 109. See e.g. US Department of the Army, Field Manual 22-100: Military Leadership, 1965, paras 19, 25; US Department of the Army, Field Manual 100-5: Operation of Army Forces in the Field, 1968, paras 3–7.
  • 110. See e.g. US Department of the Army, Field Manual 5-19: Composite Risk Management, 2006, paras 1-0, 1–17, 1–18.
  • 111. See Charlie Barnao, “Military Training: Group Culture, Total Institution, and Torture”, Italian Sociological Review, Vol. 9, No. 2, 2019, p. 295.
  • 112. This dynamic has been studied extensively by social and clinical psychologists. See e.g. Leonard S. Newman and Ralph Erber, Understanding Genocide: The Social Psychology of the Holocaust, Oxford University Press, New York, 2002; Donald G. Dutton, The Psychology of Genocide, Massacres, and Extreme Violence: Why “Normal” People Come to Commit Atrocities, Praeger, Westport, CT, 2007; Paul A. Roth, “Social Psychology and Genocide”, in Donald Bloxham and A. Dirk Moses (eds), The Oxford Handbook of Genocide Studies, Oxford University Press, Oxford, 2010, p. 198.
  • 113. D. Robinson, above note 95, p. 11 (footnotes omitted).
  • 114. This is the case a fortiori in the context of collective war crimes, such as genocide. Cf. A.-M. Boisvert, H. Dumont and M. Petrov, above note 24, p. 122 (observing in the context of collective crimes: “Le droit pénal classique des pays occidentaux, centré sur la répression d'un acte précis en fonction d'une certaine conception philosophique de l’être humain, convient mal en effet à la répression de la criminalité de groupe”).
  • 115. W. G. Eckhardt, above note 19, p. 4.
  • 116. Kenneth A. Howard, “Command Responsibility for War Crimes”, Journal of Public Law, Vol. 21, No. 1, 1972, p. 21.
  • 117. See J. S. Martinez, above note 39, p. 662. Cf. Committee of the Whole: Summary Record of the 1st Meeting, 16 June 1998, UN Doc. A/CONF.183/C.1/SR.1, 20 November 1998, paras 67 ff. (statements of the United States justifying the ICC regime of separate command responsibility).
  • 118. Cf. J. S. Martinez, above note 39, p. 662.
  • 119. Cf. ibid., p. 663.
  • 120. See e.g. G. Mettraux, above note 21, p. 73; ICRC Commentary on APs, above note 26, p. 1012.
  • 121. See e.g. M. Damaška, above note 90, p. 480.
  • 122. See Y. Dinstein, above note 81, p. 271.
  • 123. See e.g. Victor Hansen, “Creating and Improving Legal Incentives for Law of War Compliance”, New England Law Review, Vol. 42, No. 2, 2008, pp. 248–258.
  • 124. See, generally, Myron Wish and Susan J. Kaplan, “Toward an Implicit Theory of Interpersonal Communication”, Sociometry, Vol. 40, No. 3, 1977; Miles L. Patterson, “Strategic Functions of Nonverbal Exchange”, in John A. Daly and Mohn M. Wiemann (eds), Strategic Interpersonal Communication, Erlbaum, Hillsdale, NJ, 1994, p. 273.
  • 125. US S. Doc. 213, 57th Cong., 2nd Sess., p. 5. The background to this conviction can be reviewed in John Bassett Moore, A Digest of International Law, Vol. 7, US Government Printing Office, Washington, DC, 1906, p. 187.
  • 126. See, generally, M. Wish and S. J. Kaplan, above note 124; Francisco Yus, “Misunderstandings and Explicit/Implicit Communication”, Pragmatics, Vol. 9, No. 4, 1999, p. 487; Jan Hoogervorst, Henk van de Flier and Paul L. Koopman, “Implicit Communication in Organisations”, Journal of Managerial Psychology, Vol. 19, No. 3, 2004.
  • 127. See Ian Kershaw, Hitler 1936–1945: Nemesis, Penguin Press, London, 2000, pp. 355–389.
  • 128. See Sophie Richardot, “‘You Know what to Do with Them’: The Formulation of Orders and Engagement in War Crimes”, Aggression and Violent Behavior, Vol. 19, No. 2, 2014, p. 87.
  • 129. See A. J. Sepinwall, above note 88, p. 291; see also Melissa Epstein Mills, “Brass-Collar Crime: A Corporate Model for Command Responsibility”, Willamette Law Review, Vol. 47, No. 1, 2010, pp. 39–50 (description of command failures at Haditha and Abu Ghraib). The command responsibility doctrine was not invoked by the US courts martial in either the Haditha or Abu Ghraib incidents. Ibid., p. 26.
  • 130. See S. Richardot, above note 128, p. 86; Jerrold M. Post and Lara K. Panis, “Crimes of Obedience: ‘Groupthink’ at Abu Ghraib”, International Journal of Group Psychotherapy, Vol. 61, No. 1, 2011, p. 54.
  • 131. See M. Cherif Bassiouni, “The Institutionalization of Torture under the Bush Administration”, Case Western Reserve Journal of International Law, Vol. 37, No. 2, 2006, p. 390; Christopher Kutz, “The Lawyers Know Sin: Complicity in Torture”, in Karen J. Greenberg (ed.), The Torture Debate in America, Cambridge University Press, Cambridge, 2005, pp. 242–243.
  • 132. See Gregory S. Gordon, Atrocity Speech Law: Foundation, Fragmentation, Fruition, Oxford University Press, New York, 2017, p. 289; S. Richardot, above note 128, pp. 87–88; J. M. Post and L. K. Panis, above note 130, pp. 62–64.
  • 133. See G. S. Gordon, above note 132, p. 287.
  • 134. J. M. Post and L. K. Panis, above note 130, pp. 62–64.
  • 135. In Stanley Milgram's experiments on obedience to authority, one subject shocked a person to death without direct instructions from the experimenter, seemingly based on the belief that the experimenter wished the shocks to continue when the “learner” resisted answering the experimenter's questions. See Stephen Gibson, “Obedience without Orders: Expanding Social Psychology's Conception of ‘Obedience’”, British Journal of Social Psychology, Vol. 58, No. 1, 2019, pp. 241, 250.
  • 136. See generally Stanley Milgram, Obedience to Authority: An Experimental View, Harper, New York, 1974 (describing a series of experiments showing that a large majority of persons will obey instructions of apparent authorities to torture and ultimately kill another person).
  • 137. See Emilie A. Caspar et al., “The Effect of Military Training on the Sense of Agency and Outcome Processing”, Nature Communications, 31 August 2020, available at:; Silvia da Costa et al., “Obedience to Authority, Cognitive and Affective Responses and Leadership Style in relation to a Non-Normative Order: The Milgram Experiment”, Revista de Psicología, Vol. 39, No. 2, 2021.
  • 138. See, generally, Solomon E. Asch, “Opinions and Social Pressure”, Scientific American, Vol. 193, No. 5, 1955; Knud S. Larsen, “The Asch Conformity Experiment: Replication and Transhistorical Comparisons”, Journal of Social Behavior and Personality, Vol. 5, No. 4, 1990.
  • 139. Pascal Vennesson, “Cohesion and Misconduct: The French Army and the Mahé Affair”, in Anthony King (ed.), Frontline: Combat and Cohesion in the Twenty-First Century, Oxford University Press, Oxford, 2015, p. 235.
  • 140. There are numerous examples of deliberate criminal activities committed by military personnel, perceived as being part of a broader mission: see Peter Rowe, “Military Misconduct during International Armed Operations: ‘Bad Apples’ or Systemic Failure?”, Journal of Conflict and Security Law, Vol. 13, No. 2, 2008; Jim Frederick, Black Hearts: One Platoon's Descent into Madness in Iraq's Triangle of Death, Harmony Books, New York, 2010; A. T. Williams, A Very British Killing: The Death of Baha Mousa, Jonathan Cape, London, 2012.
  • 141. P. Vennesson, above note 139, p. 242. See also Michael Walzer, Arguing about War, Yale University Press, New Haven, CT, 2006, p. 31, who writes that systemic criminal acts done pursuant to military objectives can be considered as “purposive crimes” rather than “crimes of indiscipline”.
  • 142. Ibid., pp. 242–243. See also P. Rowe, above note 140, pp. 170–182.
  • 143. Brereton Report, above 2, p. 31, para. 27; p. 33, para. 34.
  • 144. Ibid., p. 34, para. 40.
  • 145. See A. J. Sepinwall, above note 88, p. 289.
  • 146. See, generally, Irving L. Janis, Groupthink: Psychological Studies of Policy Decisions and Fiascoes, Houghton Mifflin, Boston, MA, 1982, pp. 5–9.
  • 147. Ibid., p. 60.
  • 148. See e.g. J. M. Post and L. K. Panis, above note 130, p. 61 (context of US torture of detainees at Abu Ghraib military prison); D. G. Dutton, above note 112, pp. 102–103, 111.
  • 149. A. J. Sepinwall, above note 88, p. 290.
  • 150. Brereton Report, above note 2, p. 326.
  • 151. Arthur Conan Doyle, “The Adventure of Silver Blaze”, Strand Magazine, London, 1892.
  • 152. K. A. Howard, above note 116, p. 17. See also A. J. Sepinwall, above note 88, p. 289. In his treatise on international criminal law, Judge Cassese insightfully observes that a commander's wilful failure to prevent a subordinate's war crime need not involve positive action; “it may happen that the commander by his inaction aimed in fact at furthering the crime of the subordinate”. A. Cassese, above note 97, p. 244. Similarly, he notes that it can be argued that failure to exercise the duty of supervision may “in some way” contribute to the war crime. Ibid., p. 245.
  • 153. ICTY, Blaškić, above note 66, p. 789.
  • 154. ICTY, Prosecutor v. Aleksovski, Case No. IT-95-14/1, Judgment (Trial Chamber), 25 June 1999, para. 65.
  • 155. See e.g. Brereton Report, above note 2, p. 325.
  • 156. Peter Rowe, “Military Misconduct during International Armed Operations”, Journal of Conflict and Security Law, Vol. 13, No. 1, 2008, p. 179. This very sentiment was echoed in the aftermath of the My Lai massacre by US forces in Vietnam. During the trial of Lieutenant Calley, his superior Captain Medina gave four reasons as to why he did not report the massacre to his superiors: “The four reasons that I did not report the shooting of any innocent or noncombatants at the village of My Lai four and the reason that I suppressed the information from the brigade commander when I was questioned are as follows: Number one, I realized that instead of going in and doing combat with an armed enemy, the intelligence information was faulty and we found nothing but women and children in the village of My Lai four, and, seeing what had happened, I realized exactly the disgrace that was being brought upon the Army uniform that I am very proud to wear. Number two, I also realized the repercussions that it would have against the United States of America. Three, my family, and number four, lastly, myself, sir.” See “Captain Ernest Medina, Witness of the Court”, Famous Trials, available at:
  • 157. M. Damaška, above note 90, p. 467.
  • 158. See e.g. M. Damaška, above note 90; D. Robinson, above note 95, pp. 18–23.
  • 159. See A. J. Sepinwall, above note 88, pp. 298–302. Darryl Robinson has argued that the municipal criminal law concept of “accessory after the fact” could not justify holding commanders responsible for failing to punish the war crimes of subordinates: D. Robinson, above note 95, p. 48. While technically correct, this argument is irrelevant. As discussed, municipal criminal law analogies have no application in international criminal law due to the very different contexts in which the respective legal systems operate.
  • 160. See e.g. ICTY, Delalić, above note 46, para. 226; G. Mettraux, above note 21, pp. 248 (arguing that a commander's failure to adopt general measures to prevent war crimes is not relevant to command responsibility), 225 (arguing that, if an officer receives “contradictory reports about allegations of crimes” of subordinates, the officer is free to ignore the more disturbing report and rely on the “optimistic and calming report” without investigation, without incurring command responsibility if the reassuring reports turn out to be false). See also Amy Sepinwall's discussion of how the United States repeatedly declined to prosecute officers who intentionally or recklessly ignored war crimes by subordinates in Iraq: A. J. Sepinwall, above note 88, pp. 258–260, 275–279, 284–285.
  • 161. ICTY, Delalić, above note 46, para. 226 (emphasis in original); accord ICTY, Blaskić, above note 69, para. 62; ICTR, Bagilishema, above note 25, para. 42. In Kordić, the Trial Chamber interpreted the Čelebići appellate judgment to excuse commanders from a general duty of supervision of their subordinates. ICTY, Prosecutor v. Kordić, Case No. IT-95-14/2-T, Judgment (Trial Chamber), 26 February 2001, paras 432–437.
  • 162. ICTR, Bagilishema, above note 25, paras 35–37.
  • 163. ICTY, Halilović, above note 45, para. 71.
  • 164. S. Trechsel, above note 97, p. 29.
  • 165. See e.g. ICTY, Blaskić, above note 69, para. 332; G. Mettraux, above note 21, p. 45.
  • 166. See e.g. S. Trechsel, above note 97 p. 32; G. Mettraux, above note 21, p. 225; A.-M. Boisvert, H. Dumont and M. Petrov, above note 24, p. 127.
  • 167. Nuremberg Military Tribunal, High Command, above note 63, pp. 543–544.
  • 168. Sigrid Redse Johansen, The Military Commander's Necessity: The Law of Armed Conflict and Its Limits, Cambridge University Press, Cambridge, 2019, p. 78.
  • 169. Douglas MacArthur, Reminiscences, Da Capo Press, Burlington, VT, 1964, p. 298.
  • 170. ICTY, Prosecutor v. Blaskić, Case No. IT-95-14-T, Decision (Trial Chamber), 3 March 2000, paras 329–332.
  • 171. See I. Bantekas, above note 21, pp. 67–70; cf. W. G. Eckhardt, above note 19, pp. 22–24.
  • 172. An informative case arose following the Canadian intervention in Somalia during the mid-1990s. In Morneault v. Canada (2000 CarswellNat 980), the Appeals Division of the Canadian Federal Court reviewed the decision of a Commission of Inquiry to discipline a lieutenant colonel whose subordinate soldiers beat to death a detainee and shot civilians during a deployment in Somalia. Although another commander had ordered his subordinates to “abuse” any Somali intruders into the base, the Commission did not find that Morneault had specifically ordered or participated in war crimes. It instead concluded that he had failed in his duty as a commander to train and supervise his subordinates, and to ensure that his subordinates knew their obligations under the international law of armed conflict in general and with regard to detainees in particular. See Dishonoured Legacy: Lessons of the Somalia Affair: Report of the Commission of Inquiry into the Deployment of Canadian Forces to Somalia, Vol. 4, 1997, pp. 1029–1032. See also L. C. Green, “Command Responsibility in International Humanitarian Law”, Transnational Law & Contemporary Problems, Vol. 5, 1995, pp. 370–371.
  • 173. It is thus incorrect to argue that because any responsibility a commander may have to train troops is subject to municipal military law and policies, the nature of that training or other preventive measures, or the absence altogether of training, is irrelevant to the commander's responsibility for the war crimes of subordinates: see G. Mettraux, above note 21, pp. 69–70, 248. The ICC arrived at the opposite conclusion, finding a commander responsible for failing to properly train his troops and disseminate a code of conduct prohibiting pillage: ICC, Prosecutor v. Bemba, Case No. ICC-01/05-01/08, Judgment, 21 March 2016, paras 736–737.
  • 174. M. E. Mills, above note 129, pp. 25–26.
  • 175. Superior Military Government Court of the French Occupation Zone in Germany, Case of Hermann Roechling and Others, Judgment, 25 January 1949, reprinted in Trials of the War Criminals Before the Nuremberg Military Tribunals under Control Council Law No. 10, Vol. 14, 1949, pp. 1097, 1106.
  • 176. Rome Statute, Art. 28.
  • 177. Crimes against Humanity and War Crimes Act, SC 2000, c. 24, para. 7. “Criminal negligence”, it should be noted, is a higher standard under Canadian law than simple negligence; it requires wanton or reckless disregard for a legal duty. See Canadian Criminal Code, RSC 1985, c. C-46, § 219.
  • 178. This is how Parks interprets the Muto and Yamashita holdings: see W. H. Parks, above note 16, pp. 89–90. We disagree. We interpret these cases as drawing the inference of knowledge under the specific facts of the cases, involving war crimes so systematic, widespread and repeated that an inference of knowledge was properly drawn in the cases at bar.
  • 179. In May 2019, Trump pardoned several officers and a security contractor who had either been convicted of murdering civilians and unarmed prisoners of war or had been accused of such crimes and were awaiting trial. He also restored the full rank and pay of a Navy SEAL who had committed war crimes. See Leo Shane III, “Trump Grants Clemency to Troops in Three Controversial War Crimes Cases”, Military Times, 22 November 2019; Dan Maurer, “Should There Be a War Crime Pardon Exception?”,, 3 December 2019, available at: In 2020, he also pardoned private military contractors who murdered civilians and children in Iraq. See “Trump Grants Clemency to Former Blackwater Contractors Convicted of War Crimes in Iraq and Associates Prosecuted Following the Mueller Investigation”, American Journal of International Law, Vol. 115, No. 2, 2021; Leo Shane III, “Trump Pardons Former Rep. Duncan Hunter and Four Iraq War Vets Convicted in Blackwater Shooting”, Military Times, 22 December 2020. High-ranking Pentagon officers and veteran soldiers alike expressed dismay about and opposition to the pardons and clemencies for undermining military discipline and the LOAC. See David S. Cloud, “Senior Military Officers Rebel against Trump Plan to Pardon Troops Accused of War Crimes”, Los Angeles Times, 22 May 2019, available at:…; Alex Johnson, “Gallagher Case Reveals Trump's Ignorance of the Military, Fired Navy Secretary Writes”, NBC News, 27 November 2019, available at:…. One jurist has made a strong case for Trump's status as a war criminal based on command responsibility. See Gabor Rona, “Can a Pardon Be a War Crime? When Pardons Themselves Violate the Laws of War”, Just Security, 25 May 2019 and 24 December 2020, available at:….
  • 180. See Jeremy Dunnaback, “Command Responsibility: A Small-Unit Leader's Perspective”, Northwestern University Law Review, Vol. 108, No. 4, 2014, pp. 1411, 1419.
  • 181. See K. A. Howard, above note 116, p. 21; cf. J. Dunnaback, above note 180, p. 1420 (“There is always some level of risk that war crimes are about to be committed”).
  • 182. W. H. Parks, above note 16, p. 103.
  • 183. See e.g. M. Damaška, above note 90, p. 474; J. Dunnaback, above note 180, p. 1414.
  • 184. See e.g. Christopher H. Warner , “Effectiveness of Battlefield-Ethics Training during Combat Deployment: A Programme Assessment”, The Lancet, Vol. 378, No. 9794, 2011.
  • 185. See G. Mettraux, above note 21, p. 211. Mettraux also argues that obligating commanders to fulfil their military duties of supervision “has the practical effect of shifting the burden of proof” to the defendant. However, this claim assumes that only evidence of the commander's actual knowledge of war crimes is relevant. As noted, the commander is factually and morally responsible for a broader range of attitudes and behaviour by subordinates.
  • 186. Cf. M. Osiel, above note 107, p. 193.
  • 187. Nuremberg Military Tribunal, High Command, above note 63, p. 558.
  • 188. Ibid.
  • 189. See C. Meloni, above note 101, p. 636.
  • 190. K. A. Howard, above note 116, p. 20.

Continue reading #IRRC No. 919

More about International Humanitarian Law (IHL), IHL, International Criminal Law, War crimes