The Conduct of Hostilities, Attack Effects, and Criminal Accountability

Abstract

War crimes related to the decision to carry out attacks during the conduct of hostilities are almost always defined in terms of conduct and not result (Article 8(2)(b) of the Rome Statute of the International Criminal Court). Yet it is common for critiques of such decisions to focus on attack results as proof of their alleged illegality. While such results are probative of compliance or non-compliance with international humanitarian law rules regulating the conduct of hostilities, they should rarely be indisputable. This article addresses the challenge of attaching probative value to attack results when assessing responsibility for alleged war crimes based on allegedly illicit attack decisions.

1. Introduction

For more than a century, the international community has aspired to ensure individual accountability for serious violations of the laws and customs of war, today known as international humanitarian law, or IHL. Several considerations have motivated this aspiration, including the retributive and general deterrence interests normally associated with criminal sanction. While there have been periods of rapid evolution in the processes associated with such accountability, preventing impunity for what appear to be relatively obvious war crimes continues to be a significant international legal and diplomatic challenge.¹

Tolerance for such impunity, however, has become increasingly perceived as unacceptable. This itself is a consequence of numerous factors, not the least of which is the speed with which global audiences are exposed to the consequences of such apparent violations, and the density and clarity of that exposure. Images and on-scene reports of the death and destruction inflicted during armed conflicts – especially when civilians are the victims – are transmitted so rapidly that the forces engaged in such hostilities struggle to influence the perceptions and narratives created by such instant access to battlefield information.

Intolerance of perceived impunity for those suspected of committing war crimes has been elevated to a new level in relation to the ongoing conflict between Russia and Ukraine. Calls for criminal accountability have been robust and consistent since Russia launched its most recent campaign of aggression.² These calls are undoubtedly influenced substantially by the visceral reaction to visual evidence of the widespread infliction of suffering on the Ukrainian civilian population, especially as this suffering appears to result from indiscriminate and illegal attacks by Russian armed forces.³

It is difficult to deny the validity of the perception that Russian forces have committed, and continue to commit, serious war crimes in the conduct of these hostilities. However, the near total reliance on attack effects to conclude that war crimes have been committed raises an important question in the mosaic of the criminal accountability process: what is – or perhaps what should be – the probative value of attack effects in the criminal accountability process for allegations of conduct of hostilities crimes? Arguably, while attack effects are relevant to raising suspicion and the assessment of guilt of those subsequently accused of conduct of hostilities crimes, they rarely (if ever) provide conclusive evidence of such guilt. This is because the focus of both the IHL regulation of the conduct of hostilities and the international criminal law proscription of this conduct is on the decision to attack, rather than the result of the attack. Accordingly, it is the judgement of the individual in deciding to launch the attack, within the context of the totality of the circumstances that existed at the time of the attack decision, that defines compliance or non-compliance with both bodies of law.

This article will consider this question of how attack effects should play into the accountability process for war crimes prosecutions. It will begin by outlining why accountability must ultimately turn on the legality of attack decisions and not on their effects. It will then consider what is often the evidentiary dilemma created by a lack of access to direct evidence of the pre-attack state of mind of a criminal defendant when that state of mind is an essential element of criminality. This will lead to consideration of the importance of circumstantial evidence as proof of a subjective criminal mental state. The article will then consider whether attack effects alone can ever legitimately provide sufficient proof of illegal attack decisions. It will then briefly suggest other types of circumstantial evidence that should play an important role in the totality of the evidence mustered in a case where attack effects are the primary evidence of culpability. Finally, the article will raise the question of whether resistance to prosecutorial efforts to gain access to evidence associated with attack decision making should ever justify an adverse inference against a defendant to war crimes.

2. Conduct of hostilities and attack decisions

The effects of an attack are emotive.⁴ They are also nearly always the primary publicly available information related to the process of employing force during the conduct of hostilities.⁵ Furthermore, because hostilities seem increasingly to gravitate towards areas of substantial civilian populations, these effects routinely include widespread destruction of civilian property, death and injury to civilians, and the infliction of substantial second- and third-order negative impacts on the civilian population.⁶ These brutal realities of war are also increasingly broadcast to international audiences in near real time, generating widespread perceptions that the conduct of hostilities is both unlawful and immoral.⁷

Yet such imagery rarely tells the complete story of such military actions.⁸ Why was the attack launched? Why was a particular weapon system used? What was the commander's understanding of the enemy's situation at the time of the attack? What alternate feasible options did the commander have at his or her disposal? These are just a few considerations related to the legality of an attack that are almost never revealed by the imagery of attack effects. Nonetheless, because of the emotive nature of such imagery and the lack of access to all these other considerations, it is logical that so many observers and critics of military operations tend to gravitate towards what I have previously labelled ‘effects-based condemnations’ – the process of drawing ipso facto conclusions that war crimes were committed based exclusively on such effects.⁹ However, assessing compliance or non-compliance with the attack legality framework of IHL based solely on such effects is inconsistent with the focus of legal regulation of hostilities that lies at the core of such compliance.¹⁰

IHL provides a comprehensive legal framework applicable to the conduct of hostilities.¹¹ This framework is best reflected in what is routinely characterised as the ‘targeting’ rules codified in the 1977 Additional Protocol I to the Four Geneva Conventions of 1949 (AP I).¹² These treaty rules reflect near universally recognised principles of customary international law: distinction, proportionality, and, to a somewhat lesser extent, attack precautions.¹³

The principle of distinction, rightly designated as the ‘basic’ rule in AP I, mandates that the only subjects that may be made the deliberate targets of attack are lawful military objectives, members of enemy armed forces and other organised armed groups, and civilians directly participating in hostilities.¹⁴ In so doing, the rule requires participants in hostilities to constantly ‘distinguish’ between these lawful subjects of attack and all other persons.¹⁵ This obligation applies equally to a private deciding whether to shoot someone in battle as it does to the highest-level commander deciding whether to launch a massive missile attack.¹⁶

The principle of distinction also prohibits launching an attack on any object or place that is not a military objective. This limitation is functionally implemented through compliance with the rule of military objective – a customary international law rule codified in AP I.¹⁷ In its most basic conception, this rule establishes the criteria to decide the places and things that qualify prima facie as lawful objects of a deliberate attack.¹⁸ While the military objective test does not textually apply to which people may be attacked, it is generally recognised that the same concept applies to justify attacking combatants and civilians directly participating in hostilities.¹⁹ For combatants, the authority to attack is based on a determination of that ‘status’ and justifies deliberate targeting unless the individual is rendered hors de combat; for civilians directly participating in hostilities, attack authority continues ‘for such time’ as the participation continues.²⁰ There is a divergence of views on how individuals who do not qualify as combatants but are members of organised armed groups are properly categorised, with an ‘increasing tendency to treat all non-state actors as merely a conglomeration of civilians who take a direct part in hostilities’.²¹ Some states, like the United States, utilise a membership approach, characterising civilians who are part of an organised armed group as unprivileged belligerents subject to deliberate attack no differently from combatants;²² other states follow the approach advanced by the International Committee of the Red Cross and consider these individuals as civilians who lose protection from attack only if their role in the organised armed group amounts to a ‘continuous combatant function’²³ (CCF).²⁴ However, under either approach, the civilian becomes subject to deliberate attack once the relevant assessment is made.²⁵

Determining whether a person, place or thing was prima facie subject to deliberate attack is ultimately, however, a matter of judgement.²⁶ Indeed, it is self-evident why the rules establishing who or what is a lawful object of deliberate attack are prospective in nature: they require a reasonable assessment of all available information before the trigger is pulled.²⁷ No commander or soldier can be expected to have perfect situational awareness, and the variables of combat operations will inevitably influence this important decision.²⁸ Accordingly, when making attack decisions, participants in hostilities are obligated to assess all ‘reasonably available’ information to inform the judgement of whether a person, place or thing qualifies as a lawful object of attack.²⁹ If the decision turns out to have been wrong, that alone is not an indication of a legal violation.³⁰ Only if the error was the result of unreasonable decision making will it indicate such a violation.³¹

If a person, place or thing is assessed as not qualifying as a lawful object of attack, or target, an attack is unquestionably prohibited,³² but even when the person or object does so qualify, the legality assessment will normally continue to consider other obligations. These additional steps in the legality assessment are required whenever the attack is reasonably assessed as creating risk to civilians or civilian property.³³ Only when the attack is assessed as creating no such risk may these additional steps of the process be bypassed. In other words, if the target is assessed as a lawful military objective (or individual subject to deliberate attack) and there is no reasonably assessed risk that launching the attack will create such civilian risk, the attack is legally permitted.³⁴ However, because this is rarely the case in contemporary armed conflicts, additional rules intended to protect civilians from the incidental or collateral risk of attack must be implemented, notably precautions and proportionality.³⁵

The principle of proportionality is part of a broader prohibition against launching any attack that qualifies as indiscriminate.³⁶ Again, however, the prohibition is not defined by the result, but by the decision to launch the attack.³⁷ For example, if a commander treats distinct military objectives located in a civilian population centre as one general target – what is often called ‘area targeting’ – launching the attack would be considered indiscriminate.³⁸ The same rule prohibits employing weapons that cannot be effectively directed at a specific military objective (like a missile or rocket that cannot be designated to strike a specific target) or that produce effects that cannot be controlled after release (such as an incendiary weapon).³⁹ However, the proportionality component of this prohibition against launching an indiscriminate attack is most commonly associated with the law's objective of mitigating risk to civilians and civilian property.⁴⁰

Where an attack directed against a military objective is anticipated to result in civilian casualties and/or destruction of civilian property that is assessed as excessive in comparison with the anticipated concrete and direct military advantage, the attack falls within the definition of indiscriminate and is prohibited.⁴¹ Importantly, this means that such casualties and/or destruction are not categorically prohibited, even when they are an anticipated consequence of conducting an attack on a military objective.⁴² So long as these effects are not the intended objects of attack and are not assessed as excessive, the law tolerates them.⁴³

While a cursory review of these principles and accordant implementing treaty rules, one consistent thread runs through all of them: they are focused on the judgement to launch the attack, not the outcome of the attack.⁴⁴ This focal point of legality obligations related to the conduct of hostilities provides the very foundation for a legal regime that seeks to accommodate both the necessity of employing decisive combat power against an enemy and the humanitarian interest of mitigating the suffering resulting from such combat action.⁴⁵ By focusing on the legality of decisions versus the legality of effects, the law accounts for the realities of war by rejecting an impossible standard of complete accuracy in the exercise of attack decision making (not to mention incentivising the exploitation of the presence of civilians in an effort to shield targets from attack).⁴⁶ Instead, what is demanded is not that decisions to attack are always right but that they are always reasonable under the circumstances.⁴⁷

This focus on regulating attack decisions is also integrated into the international criminal law accountability equation.⁴⁸ These provisions generally align with IHL targeting rules and reflect the effort to define offences that subject attack decision makers to liability for illegal attack decisions.⁴⁹ Because of the importance of defining the constituent elements of such offences, the focus on decision making, and not result, is even more explicit.⁵⁰

Like almost all crimes, war crimes fall into two general categories: conduct crimes, and result crimes.⁵¹ Conduct crimes are defined by the concurrence between a defined criminal state of mind and prohibited conduct, with no requirement to prove a criminal result.⁵² In contrast, result crimes are defined by the concurrence between a defined criminal state of mind and an act or omission that produces a criminal result.⁵³ This is reflected in offences falling within the jurisdiction of the International Criminal Court (ICC), which include both conduct and result crimes.⁵⁴ Consider the war crime of wilful killing, as defined in Article 8(2)(a)(i):⁵⁵

1. (1) The perpetrator killed one or more persons.

2. (2) Such person or persons were protected under one or more of the Geneva Conventions of 1949.

3. (3) The perpetrator was aware of the factual circumstances that established that protected status.

4. (4) The conduct took place in the context of and was associated with an international armed conflict.

5. (5) The perpetrator was aware of factual circumstances that established the existence of an armed conflict.

Note the evidentiary requirement to prove that the alleged criminal act or omission caused the alleged result: the actual death of a protected person.⁵⁶ Numerous other crimes within the Court's jurisdiction are equally result-focused.⁵⁷

In contrast, almost all war crimes related to the conduct of hostilities are conduct-focused.⁵⁸ These crimes generally align with the IHL principles outlined above.⁵⁹ The ‘basic rule,’ or principle of distinction, is reflected in several war crimes, including the war crimes of attacking civilians, attacking civilian objects, and attacking personnel or objects involved in humanitarian assistance or peacekeeping missions.⁶⁰ Each of these war crimes defines the criminal act as directing an attack on the defined individual or object, with no requirement to prove an actual result of achieving the illegitimate objective.⁶¹ For example, Article 8(2)(b)(i) defines the war crime of attacking civilians and lists the material elements of this offence as:⁶²

1. (1) The perpetrator directed an attack.

2. (2) The object of the attack was a civilian population as such or individual civilians not taking direct part in hostilities.

3. (3) The perpetrator intended the civilian population as such or individual civilians not taking direct part in hostilities to be the object of the attack.

4. (4) The conduct took place in the context of and was associated with an international armed conflict.

5. (5) The perpetrator was aware of factual circumstances that established the existence of an armed conflict.

Article 8(2)(e) defines an analogous war crime among the offences applicable to conflicts not of an international character.⁶³ Like Article 8(2)(b)(i), the prohibited act is defined as directing an attack against ‘a civilian population as such or individual civilians not taking direct part in hostilities’, with no requirement to prove a criminal result.⁶⁴

Criminal sanction for launching an indiscriminate attack in violation of the proportionality rule is also codified as a conduct crime, which is belied by the title of the offence.⁶⁵ Specifically, Article 8(2)(b)(iv) is titled the ‘[w]ar crime of excessive incidental death, injury, or damage’.⁶⁶ This suggests that the crime is focused on the result of inflicting this excessive harm.⁶⁷ However, the elements of the offence contradict this suggestion, as they require proof of the following:⁶⁸

1. (1) The perpetrator launched an attack.

2. (2) The attack was such that it would cause incidental death or injury to civilians or damage to civilian objects or widespread, long-term and severe damage to the natural environment, and that such death, injury or damage would be of such an extent as to be clearly excessive in relation to the concrete and direct overall military advantage anticipated.

3. (3) The perpetrator knew that the attack would cause incidental death or injury to civilians or damage to civilian objects or widespread, long-term and severe damage to the natural environment and that such death, injury or damage would be of such an extent as to be clearly excessive in relation to the concrete and direct overall military advantage anticipated.

4. (4) The conduct took place in the context of and was associated with an international armed conflict.

5. (5) The perpetrator was aware of factual circumstances that established the existence of an armed conflict.

Note that the prohibited criminal act is the launching of the attack, again with no requirement to prove a criminal result, only the anticipation that the attack will result in the prohibited result.⁶⁹ Thus, a commander who launches an attack with the requisite knowledge that it is anticipated to inflict clearly excessive civilian death, injury or destruction of civilian property, has violated this provision, even if that outcome does not come to fruition.⁷⁰ In contrast, the commander who launches an attack based on a reasonable assessment that it will not result in clearly excessive civilian death, injury or destruction of civilian property has not violated this proscription, even if it turns out that the actual result does give rise to that level of harm.⁷¹

It is unfortunate that the Rome Statute does not enumerate a war crime based on a violation of the obligation to implement feasible civilian risk mitigation precautions. Such an offence, however, would need to focus on an alleged criminal omission rather than on an alleged criminal result.⁷² Depending on the requisite criminal mental state, such an offence would require proof either that the defendant knew that implementing such precautions was feasible and omitted to do so, or omitted to do so as the result of a reckless or grossly negligent omission.⁷³

Nonetheless, while not a stand-alone crime, evidence of such an omission would almost certainly be relevant in proving the crime of launching an indiscriminate attack. Specifically, where an attack decision maker omitted to implement feasible precautions, it contributes to the finding that the decision maker knew that launching an attack would result in clearly excessive civilian harm.⁷⁴ Importantly, however, even if the IHL precautions obligation is considered functionally embedded within Article 8(2)(b)(iv),⁷⁵ the focal point of accountability remains on conduct, and not result. The relevance of this consideration is not whether precautionary measures were in fact feasible to implement, but whether the decision maker perceived they were and then omitted to do so. Thus, whether considering the basis for a stand-alone IHL violation or an aspect of proving the war crime of launching an indiscriminate attack, the focal point of criminal responsibility remains conduct, and not result.

Accordingly, while both IHL and the ICC make attack conduct the focal point of legal compliance and not attack results, the mental state related to compliance or violation is not identical. When assessing the legality of attack decisions within the IHL framework, it is generally understood that attack decision makers are held to a standard of reasonableness.⁷⁶ Pursuant to this standard, the ultimate focus of the inquiry is whether their decisions fell within a margin of objective reasonableness based on the circumstances ruling at the time of the decisions.⁷⁷ Accordingly, it is logical that war crimes based on violation of these fundamental targeting rules focus on attack decisions and not attack outcomes.⁷⁸ In war, outcomes are often beyond the control of commanders and others who decide to launch attacks.⁷⁹ In contrast, the judgement of the decision to do so is obviously within their control, and therefore it is the state of mind related to that judgement that is the logical focal point of criminal liability.⁸⁰

An unreasonable attack decision, however, is a necessary but insufficient basis for proving guilt for an attack-related war crime. This is because the ICC imposes a culpability standard that is more demanding than that of the reasonable commander standard.⁸¹ This heightened standard is reflected in the mental elements required for proving violation of targeting-based war crimes.⁸² The requirement to prove that an attack decision violated distinction or proportionality requires, as noted above, that the defendant launched the attack with intent to target civilians or civilian property, or knowledge that the attack on a military objective would result in clearly excessive incidental injury or collateral damage.⁸³ These subjective criminal mental states exceed the reasonable commander standard because both require proof of more than simply an unreasonable decision:⁸⁴ proof that the decision to attack was reckless or even grossly negligent would indicate that it was unreasonable, but neither of these findings prove the intent or knowledge required for the ICC offences.⁸⁵

Of course, investigating and punishing individuals who make unreasonable attack decisions is not the exclusive or even the primary responsibility of the ICC. States bear an independent responsibility to hold attack decision makers accountable for unlawful conduct.⁸⁶ In that context, proof of a different criminal state of mind, such as recklessness or gross negligence – imposing a less demanding burden on the prosecution for alleged violations of targeting rules – may satisfy the requirements for domestic criminal accountability.⁸⁷ This may also, however, necessitate proof of a prohibited result depending on how the domestic crime is defined.⁸⁸ Thus, for example, a grossly negligent attack decision would satisfy the requisite mental element for the crime of manslaughter in violation of the Uniform Code of Military Justice applicable to US armed forces, but because manslaughter requires proof of the unlawful killing of a human being, the decision alone would be insufficient to prove this offence⁸⁹ (or perhaps reckless or grossly negligent attack decisions could provide the basis for other conduct crimes, such as reckless endangerment or dereliction of duty).⁹⁰

Comparing the crime of reckless manslaughter with the crime of launching an attack with the intent to target civilians or individual civilians reinforces the dichotomy between conduct and result targeting violations.⁹¹ As noted, the former offence is defined in terms of result: the unlawful killing of a human being.⁹² In contrast, the latter offence is defined by conduct: the directing of an attack intending to target civilians.⁹³ Accordingly, proving a harmful result is not essential for proving the conduct crime.⁹⁴ However, this does not mean such a result is irrelevant in seeking to prove a conduct violation, but just how should attack results be factored into the culpability equation?

3. Attack effects as circumstantial evidence of criminal intent

3.1. Proof beyond reasonable doubt and the criminal state of mind

Proof beyond reasonable doubt requires evidence to satisfy both a burden of production and a burden of persuasion.⁹⁵ The burden of production is satisfied when the evidence, viewed in the light most favourable to the prosecution, supports a rational finding that all elements of the alleged crime have been established.⁹⁶ Attack effects resulting in the death of civilians, injury to civilians, and/or destruction of civilian property logically serve as important circumstantial evidence to satisfy this prima facie burden.⁹⁷ Of course, the more widespread the injury and/or damage, the more rational is the inference that intent to inflict such results or knowledge that the attack would produce indiscriminate results.⁹⁸

Yet while attack effects may often justify criminal investigation and charge, and satisfy the prima facie burden of production, can they, standing alone, also satisfy the burden of persuasion?⁹⁹ In theory, as long as the inference of criminal intent is the only rational inference derived from such effects, such a finding would be justified.¹⁰⁰ However, when the circumstantial evidence associated with the attack effects points to the alternate reasonable inference, it would be far less likely that those effects alone would satisfy this burden because those effects must be considered in the context of the attack situation.¹⁰¹

This necessitates two qualifications to reliance on attack effects as the principal or exclusive circumstantial evidence of guilt. First, the risk of potential over-reliance on such effects should constantly be emphasised to the jury or judges in deciding guilt. Second, the effects must be assessed in the context of the situation as perceived by the defendant at the time of the attack decision, meaning consideration of available information related to the nature of the target and the circumstances related to the attack. This is because the complexity of combat operations and the myriad of influences on attack decisions will often indicate that effects alone are insufficient to rule out the alternative reasonable conclusion: that the attack – and the resulting effects – were not the result of the requisite criminal intent or knowledge.¹⁰²

3.2. Translating the legal standard to trial process

Effective war crimes prosecutions involve more than just an understanding of the requisite elements of a charged offence; they require proving these elements to a degree sufficient to overcome the legal presumption of innocence.¹⁰³ This is the point at which the evidentiary value of attack results intersects with the challenge of proving conduct crimes.¹⁰⁴ Proof that satisfies this burden of persuasion will normally fall into two broad categories: direct evidence and circumstantial (to include opinion) evidence.¹⁰⁵ Contrary to lay misconceptions, these categories of evidence are not, as a matter of law, valued differently.¹⁰⁶ Instead, it is the plenary prerogative of the fact finder to allocate probative value to all evidence.¹⁰⁷ The following pattern instruction used in US military trials is an example of this evidentiary principle:¹⁰⁸

Evidence may be direct or circumstantial. ‘Direct evidence’ is evidence that tends directly to prove or disprove a fact in issue. If a fact in issue was whether it rained during the evening, testimony by a witness that he/she saw it rain would be direct evidence that it rained.

On the other hand, ‘circumstantial evidence’ is evidence that tends to prove some other fact from which, either alone or together with some other facts or circumstances, you may reasonably infer the existence or nonexistence of a fact in issue. If there was evidence the street was wet in the morning, that would be circumstantial evidence from which you might reasonably infer it rained during the night.

There is no general rule for determining or comparing the weight to be given to direct or circumstantial evidence. You should give all the evidence the weight and value you believe it deserves.

Because the war crimes related to illegally launching attacks require proof of a defendant's subjective criminal mental state, direct evidence of that element will often be elusive.¹⁰⁹ This is simply a truism of any criminal accountability process: proving a defendant's subjective criminal mental state is never easy short of an admission or confession.¹¹⁰ When such direct evidence of the defendant's intent or knowledge is admissible and reliable, the likelihood of conviction is at its strongest.¹¹¹ However, when such evidence is available to a prosecutor, the reality is that it creates a powerful incentive for a defendant to offer to plead guilty to secure either a favourable negotiated plea bargain or the mitigating value of contrition.¹¹²

In practice, this means that in most cases that involve an allegation of a crime requiring proof of intent or knowledge, a plea of not guilty and contested trial will often mean that direct evidence of the criminal mental state is either not available to the prosecution or, if available, is of dubious credibility,¹¹³ but, when available, direct evidence of criminal intent when launching an attack makes prosecution most compelling. A commander or other attack decision maker whose illicit use of force is so conclusively established by admission – either contemporaneous with the decision or in the form of a post-attack statement – is certainly deserving of punishment. The reality, though, is that this will rarely be the case,¹¹⁴ as it is not common for prosecutors to gain access to admissions of the intent to attack civilians or actual knowledge that launching an attack will result in an indiscriminate result.¹¹⁵

Furthermore, even when available, such direct evidence may be insufficient to satisfy the burden of persuasion when considered within the broader context of the ‘totality of the circumstances’.¹¹⁶ This is because the complexities of combat operations may dilute the probative value of such a statement.¹¹⁷ A good example of this was the attack order issued by General Ante Gotovina and introduced against him in his trial before the International Criminal Tribunal for the Former Yugoslavia (ICTY).¹¹⁸ That order directed his artillery commander to attack the city of Knin.¹¹⁹ For the prosecution, this order seemed to provide near conclusive proof of an intent to attack civilians and civilian objects, as the entire city of Knin was obviously not itself a lawful military objective;¹²⁰ but other evidence placed that order into a broader and less incriminating context.¹²¹ The artillery commander recounted that, prior to the attack, military objectives in the city had been identified and designated for artillery attack.¹²² Accordingly, when he received the order, he understood it to direct that he place those designated targets under attack as previously planned.¹²³

Ultimately this indicates the evidentiary challenge of proving an intent to attack civilians or civilian objects and/or knowledge that an attack will result in an indiscriminate effect – the requisite subjective state of mind required to prove war crimes based on attack decisions. Absent direct evidence of those criminal mental states, prosecutors must rely almost exclusively on circumstantial evidence and the persuasiveness of the state of mind inferences supported by that evidence.¹²⁴ While there may be nothing remarkable about this in the broad context of criminal law, this does frame the critical question related to attack-based war crimes: where do the effects of such an attack fit into this equation?

When considering the totality of available evidence to establish the state of mind of someone accused of an unlawful attack decision, it would be illogical to exclude attack effects. Indeed, both the effects of the alleged illegal attack and the effects of other attacks ordered by the defendant are relevant in drawing inferences about the defendant's state of mind. However, while often highly persuasive, over-reliance on such effects creates the risk of shifting the culpability focus from the attack decision to the attack result.¹²⁵ This also fails to account for the fact that attack decisions involve a wide range of other relevant factors in assessing the defendant's state of mind at the time the attack was launched.¹²⁶ The Gotovina example above illustrates this point.¹²⁷ In isolation, the terms of the attack order, coupled with the fact that there was significant damage to civilian property, suggested an illegal attack decision.¹²⁸ When considered in the broader context, however, a reasonable alternative explanation emerged: that the attack had been intentionally directed against buildings and areas assessed as military objectives.¹²⁹

Accordingly, deliberation guidance that emphasises the importance of considering other aspects of circumstantial evidence in relation to attack effects would be useful in guiding the accountability judgement. Specifically, because the circumstantial evidence would have to prove either an intent to attack civilians or civilian property, or knowledge that the attack would result in clearly excessive indiscriminate effects, any such guidance must caution against treating attack effects as conclusive or near-conclusive evidence of guilt.¹³⁰ Instead, the inquiry should focus on how other circumstantial evidence related to the attack decision aligns with or contradicts the inference of criminal intent or knowledge; for example, evidence indicating that nothing in the nature of the target subjected to attack could plausibly have been assessed as a lawful military objective.¹³¹

Such deliberative guidance should direct the inference inquiry towards common situational aspects that frame attack decisions. There may be cases where an attack inflicts civilian casualties and/or destruction of civilian property in an area with absolutely no objective indicia of military value. Those effects in that context, absent any other evidence to suggest a genuine mistake on the part of the defendant, could justify a singular reasonable inference: the attack was intended to inflict that harm.¹³² Thus, while there are situations where circumstantial evidence derived from attack effects may provide near conclusive proof of an intent to launch an unlawful attack, this will be the case only where nothing supports an alternate rational inference that the attack may have been the result of an erroneous or mistaken judgement of legality, one resulting from the complexity of the combat situation.¹³³

Proving that an attack was launched with knowledge that it would result in a clearly excessive incidental effect on civilians and/or civilian property, based on an inference derived from the effects themselves, is even more complex.¹³⁴ This is because, unlike intentionally attacking civilians and/or civilian property, such an allegation presupposes the attack was launched at an actual military objective.¹³⁵ Accordingly, circumstantial evidence in the form of civilian casualties and/or destruction of civilian property should rarely, standing alone, be sufficient to satisfy the burden of persuasion.¹³⁶ This is because this offence requires proof beyond a reasonable doubt of a knowingly invalid judgement as to the permissible balance between anticipated military advantage and anticipated civilian risk.¹³⁷

More importantly, any finding that the attack decision maker launched the attack with such criminal knowledge demands some consideration of the military advantage component of the legality equation.¹³⁸ Accordingly, attack effects may certainly justify suspicion of such requisite criminal knowledge, but it is difficult to imagine they can be conclusive absent consideration of the totality of the value/risk equation.¹³⁹ Guilt would necessitate evidence that supports only one rational inference: the commander knew the military advantage derived from launching the attack would in no possible scenario ‘pay its way’ and justify the anticipated excessive incidental civilian harm.¹⁴⁰

Circumstantial evidence would play a critical role in any such assessment.¹⁴¹ The objective disparity between the nature of the intended target and the manner of the attack compared with the expected harmful effects on civilians and civilian property should be the primary source of such evidence.¹⁴² For example, a field kitchen serving an enemy unit would certainly qualify as a military objective.¹⁴³ If, however, the commander knew the field kitchen was located in an elementary school full of children and knew that the means and method of his planned attack would be likely to result in killing or wounding dozens of children, the imbalance between the concrete and direct anticipated military advantage and the anticipated civilian harm would justify a finding of knowledge of excessive effect.¹⁴⁴

In most cases, however, evidence of such obvious disparity is likely to be elusive.¹⁴⁵ This points to one ultimate conclusion: the burden of persuasion will require the prosecution to muster all available evidence related to the attack decision to meet the demanding burden of proof related to the attack decision.¹⁴⁶

4. A totality approach to assessing attack decisions

It seems relatively clear that proving that a defendant launched an attack intending to target civilians and/or civilian property, or an attack she knew would produce civilian harm that was clearly excessive in relation to the anticipated military advantage, is a daunting challenge. It seems equally clear that absent the unusual situation of a prosecutor being able to offer an admission from the defendant, circumstantial evidence and the inferences drawn therefrom will, by necessity, be the primary, if not exclusive, evidence of guilt.¹⁴⁷ Like all questions of evidentiary value, it is impossible to assign arbitrary weight to any single category of circumstantial evidence; that judgement is entrusted to the court (in the common law tradition, this is the function of the fact-finder responsible for determining the true facts, whether that fact-finder is a trial jury or a judge or panel of judges responsible for weighing evidence and determining guilt).¹⁴⁸ It is therefore useful to consider broad categories of circumstantial evidence that should be considered in assessing the requisite criminal state of mind related to an alleged unlawful attack decision. While, as noted above, attack effects are clearly included within this category, what follows is a proposed template for the presentation of other potentially probative categories related to this key issue.¹⁴⁹

4.1. The enemy situation

One of the most important sources of evidence related to assessing the legality of an attack decision is what the decision maker knew of the enemy situation, as it is the state of mind at the time the attack is conducted that is the focal point of the legality assessment.¹⁵⁰ It may be difficult to completely replicate that picture, but objective evidence of that situation must play a significant role in the fact-finding process.¹⁵¹ Still, the highly contextual nature of any such information makes it essential that the fact finder exercises caution to avoid considering the actual instead of the reasonably perceived situation at the time of the attack.¹⁵²

Assessing such evidence is essential for a credible determination of guilt.¹⁵³ This is because there is really no object that is immune from being transformed into a military objective, and because whether a decision maker knew an attack would result in excessive civilian harm depends on the anticipated military advantage.¹⁵⁴

4.2. Precautionary measures

As noted above, there is no existing crime within the ICC's jurisdiction for failing to implement feasible precautions. Nonetheless, the implementation of such measures – or the omission to do so – provides significant insight into a commander's overall approach to civilian risk mitigation.¹⁵⁵ Implementation of such civilian risk mitigation measures is evidence of good faith commitment to IHL obligations.¹⁵⁶ While certainly not indisputable as to the assessment of guilt for launching an unlawful attack, the more extensively a commander implemented precautionary measures in relation to the alleged unlawful attack and in relation to overall operations, the less plausible it is to conclude that the commander acted with intent or knowledge to attack civilians or inflict clearly excessive civilian harm.¹⁵⁷

This evidence extends beyond the measures taken in relation to the alleged criminal attack. Instead, it includes all civilian risk mitigation measures implemented in relation to the preparation for and conduct of operations.¹⁵⁸ This would include evidence of training, command policies and procedures related to civilian risk mitigation, investigations into possible failures of mitigation efforts and incorporation of lessons learned into future operations, and the leveraging of available technologies to maximise the risk mitigation objective.¹⁵⁹

4.3. Evidence of patterns of conduct

Illegal attacks are rarely a ‘one-off’ in combat; a commander acting with the intent to kill or injure civilians, destroy civilian property, or the knowledge of inflicting a clearly excessive harmful effect on civilians is one who can be assumed to have virtually no regard for humanitarian obligations. It is therefore logical to expect that the conduct of operations under his or her command will include other incidents of suspected illegality or indifference to violations committed by subordinates.¹⁶⁰ Such evidence, even if insufficient to justify stand-alone charges, is probative in assessing whether an alleged illegal attack was the result of criminal intent or knowledge.¹⁶¹

The circumstantial value of a pattern of prior criminal or wrongful acts, or modus operandi evidence, aids in the determination of whether the alleged criminal conduct was or was not set in motion by the requisite criminal intent or knowledge.¹⁶² Where the defendant has engaged in such prior bad acts – especially when those bad acts share similar characteristics with the alleged crime – they suggest that the harmful effects were not the result of a mistake or accident but were indeed calculated to inflict the civilian harm.¹⁶³

4.4. Record of command interest in investigating alleged violations

How a commander responds to indications that subordinates violated IHL obligations and/or related command policies and directives is a reflection of the commander's overall commitment to compliance.¹⁶⁴ The very notion of ‘responsible command’ includes the obligation to set and enforce command standards through the integration of internal disciplinary processes.¹⁶⁵ Indeed, the very qualification as a ‘privileged’ belligerent – an individual entitled by international law to engage in hostilities – is contingent in part on being part of a unit subject to responsible command which respects the laws and customs of war.¹⁶⁶

When commanders ignore this duty to respond to indications of subordinate misconduct or, even worse, encourage or condone such misconduct, they violate their command obligation and set the conditions for future violations.¹⁶⁷ The doctrine of command responsibility subjects the commander to criminal responsibility for the foreseeable war crimes that result from such a dereliction of duty.¹⁶⁸ In the context of prosecuting a commander for an unlawful attack decision, such indifference or condoning of subordinate misconduct provides relevant insight into the commander's attitude and approach to the law.¹⁶⁹ When coupled with other evidence, this can bolster the inference of the requisite subjective criminal mental state related to an attack.¹⁷⁰

4.5. Motive

Motive is among the most common types of circumstantial evidence used to prove subjective intent or knowledge. Motive is not intent, but is rather the reason why a defendant engaged in certain actions.¹⁷¹ In the context of war crimes, the relevance of an illicit motive – for example, to ‘cleanse’ an area of civilians or to terrorise the civilian population – may provide especially persuasive circumstantial evidence of criminal intent or knowledge.¹⁷² Indeed, when such motive proves intent to target civilians or civilian property, or knowledge that an attack would produce clearly excessive civilian harm, guilt is established by proof of the decision to launch the attack regardless of the effects.¹⁷³ That a commander fails in such a situation to produce the intended result is no shield to criminal responsibility.¹⁷⁴ Thus, such an illicit motive would justify a charge and conviction even absent any death or injury to civilians or destruction of civilian property.¹⁷⁵

Such evidence of motive can come in a number of forms, to include evidence related to the overall strategic objectives of the operation.¹⁷⁶ Again, while not conclusive, motive should be emphasised as an important basis for inferring a defendant's state of mind related to attacks.

5. Conclusion

There is a substantial difference between information that warrants suspicion of war crimes based on violation of ‘targeting’ principles and proving such violations beyond a reasonable doubt. This is especially true for the war crimes within the jurisdiction of the ICC, proscribing unlawfully directing attacks against civilians and/or civilian property and launching indiscriminate attacks. This is because, unlike the standard of objective reasonableness at the foundation of the IHL targeting legality framework, these offences require proof of a subjective criminal mental state that actuates the proscribed conduct: intent to attack civilians and/or civilian property or knowledge that the attack will result in clearly excessive incidental injury to civilians and/or collateral damage to civilian property.

When attacks produce such results, it is tempting to assume that the attack must have been unlawful, but these offences are not defined in terms of result. Instead, it is the conduct of directing or launching the attack that is the proscribed criminal act. As a result, satisfying the burden of proof to convict a commander or other attack decision maker for such a crime requires proof that the decision to direct or launch the attack was actuated by the requisite subjective criminal mental state.

Proving each of these criminal mental states will almost inevitably necessitate reliance on circumstantial evidence and the inferences it supports. While attack effects certainly fall within this category of evidence, because these crimes are defined in terms of conduct and not result, it is dangerous to assume that effects alone establish guilt. Instead, attack effects are better understood as one category of circumstantial evidence within a totality equation that contributes to drawing inferences about the decision maker's state of mind at the time the attack was launched. Recognising the qualified probative value of such effects is important for two reasons. First, it guards against conclusions of illegal attack decisions based on an ‘after the fact’ assessment of results when the proper focus of inquiry is the assessment of the attack decision based on the situation that existed at the time the attack was launched. Second, it serves as a reminder that culpability for directing or launching an illegal attack is not contingent on the result of the harmful attack but on the criminal mental state of the decision maker at the time of the attack.

Ultimately, when assessing the legality of an attack, it is both logical and justified to rely on attack effects as the basis for reasonably suspecting attack-related war crimes. It will also, in some cases, be valid to rely on such effects to satisfy the prima facie burden of production in a criminal prosecution, as the infliction of civilian casualties and destruction of civilian property may justify at least a rational conclusion that the decision to attack was unlawful. However, it should rarely be the case that such effects, standing alone, satisfy the requirement to prove guilt beyond reasonable doubt, a burden that requires the proof to exclude all reasonable conclusions other than guilt. Relying exclusively on attack effects to satisfy this aspect of the burden of persuasion is dubious whenever there is any plausible basis to conclude that the decision was intended to attack a military objective, or that the decision was based on a good faith mistake as to the true nature of the target, the military advantage derived from the attack, or the risk of incidental injury and/or collateral damage.

Accordingly, whether assessing the reasonableness of an attack decision pursuant to the IHL objective standard or whether the attack was set in motion as the result of the criminal intent or knowledge standard of the ICC, its attack effects should be considered as only one factor in a broader category of relevant circumstantial evidence. This evidence includes all the information reasonably available to the commander at the time of the decision to attack, to include the enemy situation, civilian situation, implementation of feasible precautions, patterns of compliance or non-compliance, motive, and command commitment to imposing accountability for subordinate violations. All of these factors contribute to recreating the context of the attack decision, and yet it should be rare that any single category – to include attack effects – will justifiably be accorded indisputable value in determining the decision maker's state of mind at the time of the attack. Focusing the culpability assessment on this totality equation will enhance fundamental fairness in the criminal accountability process and encourage prosecutors to avoid the temptation of endorsing the invalid process of effects-based condemnations.¹⁷⁷