A recent amendment to the Rome Statute of the International Criminal Court has drawn unprecedented attention to the war crime of starvation of civilians as a method of warfare. It comes at a time when mass starvation in war is resurgent, devastating populations in Yemen, Ethiopia, Syria, South Sudan, Nigeria, and elsewhere. The practice has also drawn the scrutiny of the United Nations Security Council. And yet, despite this heightened profile and sharpened urgency, what precisely is criminally wrongful about starvation methods remains underspecified.
A common way of thinking about the criminal wrong is as a form of killing or harming civilians. Although its differentiating particularities matter, the basic wrongfulness of the crime inheres, on this view, in it being an attack on those who ought not be attacked. For some, this supports a broad interpretation of the starvation ban. However, for others, the graduality of starvation preserves the continuous possibility of the avoidance or minimization of civilian death or harm in a way that direct kinetic attacks do not. In combination with the method’s purported military utility, this distinctive incrementalism has underpinned arguments for the permissibility of certain forms of siege and other deprivation and a narrow interpretation of the starvation crime.
Drawing on the moral philosophy of torture, this Article offers a different normative theory of the crime. Starvation, like torture, is peculiarly wrongful in its distortion of victims’ biological imperatives against their capacities to formulate and act on higher-order desires, political commitments, and even love. This process does not merely raise the cost of fulfilling those commitments. Instead, starvation tears gradually at the very capacity of those affected to prioritize their most fundamental commitments, regardless of whether they would choose to do so under the conditions necessary to evaluate matters with a “contemplative attitude.” Rather than palliating, the slowness of starvation methods is at the crux of this torturous wrong. Recognizing this redefines the meaning and place of the crime in the framework of international criminal law.
Table of Contents
In 2019, the 123 States Parties to the International Criminal Court (ICC) approved a statutory amendment incorporating the war crime of starvation of civilians as a method of warfare in non-international armed conflicts (NIACs). The new provision marks the first step towards closing a consequential gap in the ICC system. Prior to the amendment, the Rome Statute had criminalized the practice only in international armed conflicts (IACs)—the decidedly less common of the two categories of contemporary war.
The change comes at a moment when the infliction of mass starvation in war is resurgent. Millions of civilians across multiple conflicts have suffered severe deprivation, often in ways that are attributable directly to the strategies of the belligerent parties. Isolated by a Saudi- and Emirati-led blockade on one side and subject to the confiscation of food and medicine by the Houthis on the other, the people of Yemen have endured years of what remains one of the world’s gravest humanitarian crises. In Myanmar, the destruction, pillage, and denial of food and other essentials have been key components of the military’s counterinsurgency strategy, contributing to the ethnic cleansing and alleged genocide of the Rohingya population. In 2020, almost half of the population of South Sudan was in “crisis” or worse due to food deprivation arising in significant part from the actions of the warring parties. The Independent International Commission of Inquiry on Syria has described “modern day sieges in which perpetrators deliberately starved the population along medieval scripts,” imposing “indefensible and shameful restrictions on humanitarian aid.” Most recently, Ethiopian and Eritrean belligerents have used a range of starvation tactics in Tigray in what has quickly become one of the world’s most severe humanitarian catastrophes.
It is in this context that the starvation crime will take shape. The agreement on the new ICC provision was preceded by a landmark resolution in which the U.N. Security Council “strongly condemn[ed]” starvation methods in armed conflict and warned that they “may constitute a war crime.” There is now an unprecedented level of attention among academics and practitioners on a criminal prohibition that had been dormant and largely ignored since its first treaty codification (in IAC form) in 1998.
And yet, much remains uncertain. In light of the IAC crime’s dormancy and the NIAC crime’s recency, there is not the case law or broader jurisprudence that might help to settle incipient, but consequential interpretive disputes. Particularly thorny are questions around intent and purpose. Views differ as to whether the crime would attach to the act of blocking the delivery of essentials, such as food and water, to an encircled and starving population when that obstruction is performed with the goal of starving out ensconced enemy forces. On one view, for the crime to attach, it would need to be established that the besieging party acted with a view to weaponizing the attendant civilian suffering. On an alternative view, engaging in deprivation with the knowledge that it would cause civilians to starve would suffice. These views diverge significantly in terms of both the range of operations to which the crime might attach and the subset for which prosecution would be viable from an evidentiary perspective.
But confusion is not confined to questions of interpretation. Underpinning those disputes lies a deeper lack of normative clarity as to what the starvation crime is really about. Those who assert a narrow definition of the legal prohibition tend to invoke military necessity as an underlying principle of international humanitarian law (IHL) and the guiding norm in this instance. Siege, they argue, is an essential tool of warcraft and cannot be performed effectively without the complete sequestration of the entire encircled population. Whatever one makes of the empirical assumptions embedded in that claim, it raises further normative questions. The invocation of necessity is compelling only if the pro tanto wrong associated with inflicting starvation conditions on civilians is of a kind that is amenable to a necessity justification in the first place. Whether it is depends in part on what makes starving civilians as a method of warfare criminally wrongful.
A potent normative account of the starvation crime would explain not just why the conduct is condemnable, but why it warrants a criminal prohibition separate and distinct from other potentially applicable war crimes or crimes against humanity, such as attacking civilians, murder, extermination, forced displacement, or willfully causing great suffering or serious injury to body or health. Distilling that normative core is important both for honing international criminal law’s message in starvation cases and for clarifying what is at stake in interpretive disputes regarding the scope of the crime.
With that in mind, this Article analyzes criminality in the context of siege starvation—the encircling and cutting off of a defended and populated locality in order to elicit the capitulation of the besieged party and thereby avoid the need to take the area by assault. To focus on this specific form of the method is neither to ignore other modalities of starvation warfare,  nor to dismiss the use of mass starvation to punish, subjugate, or exterminate. It is instead to respond to the status of encirclement deprivation as both the quintessential form of starvation as a method of warfare—prevalent historically and today—and the form most likely to be defended, legally and morally. With sieges arguably becoming “a defining feature of modern warfare,” encirclement deprivation is the form most capable of shedding light on the normative underpinnings of the starvation crime.
The most straightforward way of thinking about starving civilians as a method of warfare is as a specific instantiation of the broader criminal category of attacks on civilians. Indeed, this appears to be the standard view of the crime. Although recognizing the distinctive features of starvation, this conceptualization identifies the method’s wrongfulness as inhering fundamentally in it being an attack on those who ought not be attacked.
This way of understanding the crime can be taken in divergent interpretive directions. On the one hand, it might be thought to underpin a categorical prohibition of encirclement starvation on the analogical basis that it would be prohibited to subject an identical besieged population to kinetic attack without discriminating between its civilian and combatant constituencies. On the other hand, some might single starvation out among attacks on civilians on the grounds that its slow and incremental nature allows for civilian harm to be avoided or minimized in a way that kinetic attacks preclude. That, in turn, might be thought to open the door to the justificatory invocation of necessity, particularly in contexts in which civilians would not be the specific targets of the operation.
Whatever one makes of those divergent lines of reasoning, the central claim of this Article is that their shared normative premise is incomplete. Plainly, the fact that criminal starvation methods can be characterized as harmful and lethal attacks on civilians is an important aspect of their criminality. However, that framing risks obscuring an additional dimension of the wrong—one that is particularly potent in explaining the independent criminalization of starvation methods and in clarifying the normative implications of the method’s graduality. This dimension is best illuminated by way of a societal analogy to torture. 
Both torture and encirclement starvation involve the infliction of suffering (and potentially death). However, the feature that sets these crimes apart from the other legal categories under which they could be prosecuted is the way that these methods turn victims’ biological imperatives against their fundamental capacities to formulate and act on higher-order desires, political commitments, and even love. The process is not simply coercive; the distortion of biological imperatives does not merely raise the cost of fulfilling higher-order commitments. Instead, it slowly crowds out the capacity of victims to decide whether to do so.
Seen in this way, the graduality of starvation, as compared to other forms of attack, is not a mitigating factor that opens the possibility of harm minimization and thus permissibility. Rather, it is central to the distinctively torturous wrong that defines the method. Starvation tears gradually at the capacity of those affected to prioritize their most fundamental commitments, regardless of whether they would choose to do so under the conditions necessary to evaluate matters with a “contemplative attitude.” This process works only through the incremental and sustained accumulation of suffering and the imperative to break free from it.
In advancing this theory of the crime, the Article proceeds as follows. Section II provides a brief outline of the legal framework and some of the key doctrinal debates regarding the scope of the crime. Section III identifies the normative role of military necessity in underpinning arguments for a narrow prohibition, emphasizes the need to consider necessity in light of the moral stakes of the conduct in question, and situates that endeavor within a theory of international criminal law as an expressive regime. Section IV elaborates what appears to be both the standard account of the wrong underpinning the starvation crime and the account assumed by those who argue for its narrow scope on necessity grounds. Section V considers the relevance of the moral philosophy of torture in illuminating its criminal wrongfulness and responding to the use of necessity by torture advocates. Section VI draws on that philosophical insight to offer a normative account of starvation of civilians as a method of warfare, framing it as a war crime of societal torture. Section VII situates that insight in the framework of international criminal law as a whole.
Starvation of civilians as a method of warfare is a relative latecomer to international criminal law. The innovative charge of “[d]eliberate starvation of civilians” was considered fleetingly when the Allies were planning to prosecute their defeated foes after World War I, but international prosecution efforts collapsed, and the crime was never established. Two and a half decades later, senior Nazi defendants were convicted at Nuremberg of starving prisoners of war, occupied populations, the enslaved, and others under their control. However, starvation in the conduct of hostilities was treated quite differently. Members of the German High Command were acquitted of any legal violation arising from the siege of Leningrad, in which over one million Russians died. The American‑run tribunal hearing the case affirmed the legal “propriety of attempting to reduce [a place controlled by the enemy to] starvation,” holding that “the cutting off of every source of sustenance from without” in such a context was fully compatible with existing international law. It was a tactic the Allies had used themselves.
Soon thereafter, international humanitarian law (IHL) underwent a significant revision. However, except in the narrow context of belligerent occupation, the Geneva Conventions of 1949 retained a permissive posture towards starvation tactics. They require that parties allow “essential foodstuffs” through to adversary territory only when those consignments are “intended for children under fifteen, expectant mothers and maternity cases” and the besieging party has no “serious reasons for fearing” that they may be diverted, controlled ineffectively, or provide a “definite advantage” to the adversary by substituting for goods that it would have provided. In the context of siege warfare, parties are required only to “endeavour to conclude local agreements” for the removal of “wounded, sick, infirm, and aged persons, children and maternity cases”; there is no requirement to succeed in those endeavors, no requirement to allow persons in the protected categories out in the absence of an agreement, and no requirement to even try to conclude agreements allowing other civilians to exit. The notion that starvation methods could qualify as a war crime remained a long way off.
It was not until the Additional Protocols were agreed in 1977 that this began to change. Both Protocol I (for IACs) and Protocol II (for NIACs) proscribe “starvation of civilians as a method of warfare [or combat].” The more detailed Protocol I rule elaborates that it is prohibited to “destroy, remove or render useless objects indispensable to the survival of the civilian population. . . . for the specific purpose of denying them for their sustenance value to the civilian population or to the adverse Party.” It clarifies that such deprivation is permissible for the denial of sustenance if the objects are used solely by the armed forces of the adversary and is permissible if done for reasons other than sustenance and the action would not lead to civilian starvation or forced movement. A separate provision specifies that if the civilian population in non‑occupied territory is inadequately supplied, “relief actions which are humanitarian and impartial in character and conducted without any adverse distinction shall be undertaken, subject to the agreement of the Parties concerned in such relief actions.”
The introduction of these rules shifted the law’s posture on starvation in war. At the same time, the ban was excluded from the grave breaches provision of Protocol I. As such, it remained a step removed from war crime status. It took two more decades for the criminality of the method to be codified in an international instrument. Article 8(2)(b)(xxv) of the Rome Statute proscribes “[i]ntentionally using starvation of civilians as a method of warfare by depriving them of objects indispensable to their survival, including wilfully impeding relief supplies as provided for under the Geneva Conventions.” An equivalent provision for NIACs was deleted late in the drafting process in what, remarkably, appears to have been an administrative error. Given the prevalence of the non‑international form of armed conflict, this was a significant omission. A further twenty years passed before it was remedied by amendment.
The ICC, of course, is limited in its reach. Absent Security Council referral or ad hoc state acceptance, the Court’s jurisdiction is restricted to crimes on the territory, or perpetrated by a national, of one of the 123 States Parties. Under current political conditions, this alone precludes ICC action in many of the current or recent conflicts involving mass starvation. Moreover, pursuant to the Statute’s onerous amendment process, the new NIAC provision will enter into force only for States Parties that choose to ratify it. As such, allegations of the use of starvation methods in the NIACs in Nigeria and Mali (both of which are ICC States Parties) remain beyond the Court’s jurisdiction even after the amendment’s incorporation into the Statute.
However, the implications of starvation’s criminalization in the Rome Statute are not defined by the jurisdictional and political limits of the ICC. Despite the Court’s difficulties, the Statute remains the normative focal point of international criminal law. The 1998 codification alone had a catalytic legislative effect. Many states quickly replicated the Rome Statute’s list of international crimes in their domestic penal codes, which often provide for universal or extended jurisdiction. Hybrid, special, and regional tribunals have also drawn on the Rome Statute for their lists of crimes. As a result, the starvation crime is now available in a range of jurisdictions not bound by the specific constraints of the ICC.
Some of these states and authorities followed the initial Rome Statute focus on starvation in the IAC context. However, others, including the Malabo Protocol for the African Court of Justice and Human Rights, have incorporated the war crime for both conflict classifications. In those that restricted jurisdiction to the IAC crime, the recent amendment to the Rome Statute can be expected to have a catalytic effect on legislative expansion to cover NIAC starvation crimes. Similarly, hybrid, special, and regional tribunals created post-2019 are now more likely to include the starvation crime for both IACs and NIACs.
The fact of the Rome Statute amendment can also help to amplify political and civil society scrutiny of starvation in war. As just one element of the unprecedented treaty agreed in Rome in 1998, the initial IAC provision was quickly forgotten and has rarely been invoked. The 2019 amendment, on the other hand, has been incorporated at a time when the practice of starvation in war is resurgent and has drawn the attention of a wide range of actors, from civil society groups to the Security Council. The new amendment has already been invoked in key expert reports regarding ongoing conflicts. As the legal and political implications of the starvation war crime develop further, much will turn on how the prohibition is understood and used.
Perhaps the most consequential question in interpreting the crime is how to think about intent, purpose, and method in the deprivation of objects that sustain both combatants and civilians. Divergent views on that question have profound implications for whether and under what conditions siege warfare can be waged without implicating international criminal law.
At one end of the spectrum, the crime is understood to attach only to acts that seek to weaponize the civilian suffering associated with starvation. Advocates of this view emphasize the Rome Statute language specifying that to perpetrate the crime is to “[i]ntentionally us[e]” the starvation of civilians as a “method of warfare.” The term “method” alone suggests deliberate, purposive action; that the method in question must be used “intentionally” only bolsters that implication. As such, for civilian starvation to be used intentionally as the method by which to advance the war effort, it must be inflicted purposefully with a view to its weaponization. Or so the thinking goes
On such an interpretation, much of the starvation of civilians in war might appear to be beyond the scope of the criminal prohibition. Most obviously, acts pursued with a view to weakening or weaponizing the suffering of combatants could be argued to fall outside the scope of the crime, even when they also cause widespread civilian deprivation, as is likely in a comprehensive starvation siege. As long as the “surrender or starve” message of the siege is directed at combatants, the fact that civilians will also starve would not itself be sufficient for the war crime to attach. Offering civilians safe egress might be thought to emphasize this distinction by indicating that the starvation of those who remain is regretted, rather than weaponized. Understood in that way, the ensuing civilian suffering is analogous to the civilian loss occurring as the collateral damage of an attack on a legitimate military objective. Although potentially subject to the IHL rules of proportionality and precaution, this civilian suffering would not be prohibited by the starvation rule, and it would certainly not qualify as the criminal form of starvation of civilians.
This argument tends to be made most forcefully in contexts of encirclement deprivation. There is a doctrinal reason for this: Article 54 of Additional Protocol I is more clearly expansive in its prohibition of the destruction, removal, or rendering useless of indispensable objects than it is with respect to the impeding of such objects’ delivery to an encircled area. However, the focus on encirclement deprivation in arguments seeking to limit the scope of the starvation ban also reflects a normative determination that sieges involve a particularly acute military imperative to engage in comprehensive and indiscriminate starvation. This necessity claim is central to the question of how to think about the normative underpinnings of the crime. It is addressed in the next Section.
Focusing on the text of the criminal prohibition, others are not persuaded by the purposive characterization of what is proscribed. For them, “method of warfare” has no settled definition in IHL, and is understood most plausibly to do “no more than describe conduct that is part of hostilities.” Moreover, the specification that starvation of civilians be used “intentionally” ought to be understood in accordance with the default mens rea standards in Article 30 of the Rome Statute, which specify that a person should be understood to have “intent” with respect to a criminal consequence when that “person means to cause that consequence or is aware that it will occur in the ordinary course of events.” The threshold for the latter (oblique) form of intent has been defined in ICC jurisprudence as being satisfied when the perpetrator acted with a “virtual certainty” that the result would occur. Using starvation methods on an encircled population with a view to starving out combatants could meet this threshold, because such a technique would entail knowingly starving civilians, even assuming no desire to weaponize their suffering. Although Article 30 does not apply technically to the use of “intentionally” in the starvation provisions, the latter can be read in its light.
A further line of argument in favor of a broad prohibition focuses on the meaning of “starvation of civilians as a method of warfare.” The central objective element of the ICC crime is that civilians are deprived of objects indispensable to their survival, not that they die or suffer any particular harm as a result. Moreover, paragraphs 2–3 of Article 54 of Protocol I (the IHL rule underpinning the war crime) proscribe both engaging in such deprivation actions with a view to denying sustenance to the adverse party (as long as civilians will also be affected) and engaging in such deprivation actions for any other reasons when doing so would cause civilian starvation or force civilians to move.
Understanding the war crime in that context, “intentionally using starvation of civilians as a method of warfare” can be read transitively to refer to the practice of denying civilians sustenance, rather than that of seeking to weaponize the harm they suffer as a result. On this reading, the crime would include the actions of belligerents who engage deliberately in the deprivation of objects indispensable to civilian survival even if they do so without the goal of harming civilians.
In fact, recognizing the starvation method to inhere in the act of deprivation, rather than the weaponization of suffering, has implications even assuming that civilians must be targeted with deprivation for the crime to attach. A civilian population does not lose its civilian character due to the presence of combatants within it. When belligerents deprive an encircled population of essentials in order to starve out the combatants ensconced within, they target the civilian population with deprivation as a necessary predicate to starving those enemy fighters. Similarly, operations targeted at civilians and combatants without discrimination have been understood to be operations targeted at civilians at the same time as they are operations targeted at combatants. 
There is much more to be said doctrinally on both sides of this interpretive debate. However, the elements above identify some of the key points of disagreement. The implications are significant. On the narrow view of the prohibition, it would be rare for a siege to meet the threshold for starvation of civilians as a method of warfare. On the broader interpretation, a siege that involved depriving the encircled population of essential goods would almost certainly satisfy the threshold, unless the population were composed predominantly of combatants or persons directly participating in hostilities. It is not the project of this Article to resolve that doctrinal dispute. However, the debate spotlights a more fundamental tension regarding how to understand the moral stakes of the prohibition.
The central normative claim underpinning objections to a broad and categorical prohibition of encirclement deprivation is that such a rule is insufficiently attentive to the military imperatives that motivate siege warfare. The thesis arising from this argument can take one of two forms. One is internal to the law. In that line of thought, military necessity is invoked as an underlying principle of IHL that demands a narrow interpretation of the rules codified in the Protocols, of their customary analogues, and therefore of the derivative war crime provisions in the Rome Statute and elsewhere. On this view, existing law should be interpreted to minimize the constraints on encirclement starvation. A second approach critiques the law from the external point of view. From that perspective, the current legal rule is correctly interpreted as broadly prohibitive of encirclement deprivation, but states would do well to refrain from ratifying treaties that include the rule and to resist its customary crystallization on the grounds that the rule ignores relevant military imperatives.
The important point at this stage is that these divergent perspectives on the law as it stands share a common normative posture. After detailing that shared underpinning, this Section contends that neither of these positions can be evaluated without reference to the countervailing moral interest at stake in the starvation ban. It is only through identifying the pro tanto wrong associated with starvation methods that we can begin to think through whether and when necessity might justify them.
The empirical claim underpinning each of the necessity arguments is simple. Facing a populated and well-defended locality, the imposition of comprehensive restrictions on what goes in is often deemed the most promising route by which a belligerent might take that locality and defeat the forces that hold it. Most obviously, that method allows the besieging force to coerce capitulation, rather than engaging in the significantly more militarily demanding alternative of trying to take the area by assault. Even when some form of assault is required, the siege can be expected to weaken the enemy sufficiently to reduce the difficulty of such an operation. Thus, despite decrying the “surrender or starve” tactics of siege warfare as “disastrous for civilians,” the U.N. Independent International Commission of Inquiry on Syria recognized the method as having been “successful for overtaking opposition-held territory” in the conflict. In the naval context, Wolff Heintschel von Heinegg asserts, “blockade remains a most efficient method for subduing the enemy.”
Crucially, on at least some accounts, the efficacy of this method is contingent on the total isolation of the besieged force. From that perspective, if a siege is to work, those within the encircled locality must be cut off completely. Anything short of comprehensive sequestration, including from humanitarian aid and other essentials, would give the adversary the lifeline necessary to sustain its defensive posture, potentially extending the siege indefinitely and precluding a decisive victory. And it is in that respect that the starvation of civilians comes into the picture. Ordinarily, the very fact of encirclement indicates that the besieging party lacks granular control over what happens within the encircled area. As such, it is impossible for that party to determine in advance whether food or other essential consignments allowed through would sustain civilians, enemy combatants, or both. Therefore, if the enemy forces must be isolated for the method to work, so too must be the civilian population within which they are ensconced.
From the method’s efficacy, it is claimed, follows a military imperative. “Very often,” Sean Watts argues, “you do not have a choice, you do not have the forces you need to undertake an assault of a city, or the weapon system that you would need to overcome the defenders.” Cutting them off and weakening them to the point of capitulation or maximal vulnerability would seem to offer the most (and possibly the only) viable path forward.
The thesis that much encirclement deprivation either is or ought to be permissible is built on this empirical premise. Responding to those who might question why civilian deaths can be inflicted by encirclement in ways that would be prohibited if they were achieved by kinetic attack, Michael Walzer reasons, “[t]he obvious answer is simply that the capture of cities is often an important military objective . . . and frontal assault failing, the siege is the only remaining means to success.”
As indicated above, some take the necessity of encirclement deprivation in such contexts to expose a serious flaw in the law’s categorical prohibition on starvation of civilians as a method of warfare. Along these lines, Yoram Dinstein has condemned Article 54 of Protocol I, describing its posture on siege warfare as “untenable in practice” because it blocks a viable path to military success, when “no other method of warfare has been devised to bring about the capture of a defended town.” That untenability might be invoked as a reason for declining to ratify the relevant treaties, ratifying with a reservation, or resisting claims that the rule has gained customary status.
For others, the existing rules are more open to interpretation than Dinstein allows. From that perspective, military necessity demands a narrow interpretation of the legal prohibitions, particularly in the context of encirclement warfare. Thus, Watts decries “operationally troubling” efforts to require besieging parties to permit the passage of humanitarian relief to avert civilian starvation, describing that understanding of existing law as driven by a myopic “resort to humanitarian objects or purposes, without equal attention to military necessity.” For him, such interpretations fail to take seriously the possibility that “permissive rules for withholding consent to relief actions reflect not inadequacies but rather the presently-operative balance between humanity and military necessity.”
Similarly, Matthew Waxman centers military necessity as the foundation of what he believes to be the law’s permissive posture on the issue, arguing:
siege methods have long been given leniency in customary law because they were seen as the only viable means of securing certain military objectives. . . . [Therefore,] [t]he international community expressed a reluctance, even among the strongest condemners of Serb practices [in the sieges of Sarajevo and Srebrenica in the 1990s], to accept the wholesale rejection of siege as a legitimate instrument. 
Along the same lines, Françoise Hampson contends, “[i]f well-meaning humanitarians argue that the applicable rules mean that you cannot conduct sieges lawfully, that does not mean that sieges will not happen. Rather, it means that sieges will happen unlawfully or outside the framework of law.” The upshot, in her view, is simple: “sieges are occasionally necessary, and being necessary, it is up to the law to accommodate them.” Thus, in evaluating the siege of Sarajevo, a U.N. Commission of Experts found itself “faced with the unpalatable fact that, unless there is a neutral arbiter, the only way to starve-out a besieged military force, a legitimate act of war, is to starve the civilian population.” Given that “the intermingling of military forces and the civilian population” precluded discriminating in the deprivation methods, the Commission found the criminality of the siege of Sarajevo to be “debatable.”
One line of response to such arguments is to challenge the claim that starvation methods are necessary to overcome well-defended and populated areas. Along these lines, one might question the binary framing according to which the only paths forward are all-out military assault or comprehensive encirclement starvation. A siege or blockade that interdicts contraband but preserves an effective channel for the supply of goods essential to human survival can be an effective way to contain enemy forces within the encircled area, to undermine their capacity to launch an offensive from that location, and to preclude their deployment elsewhere. Under the right circumstances, this can be combined with targeted attacks on specific military objectives to degrade the contained force. On this view, one can affirm the utility of a form of siege warfare, and yet insist that “[s]ieges must still allow for vital foodstuffs and other essential supplies to be delivered to the civilian population.”
Challenging the premise from a different angle, it has been argued that subjecting a heavily populated area to encirclement starvation often fails to precipitate capitulation, serving instead simply to strengthen the hand of the besieged force vis-à-vis the severely impacted civilian population. Additionally, it might be argued that when sieges do succeed in dislodging the adversary, this is because the process of encirclement deprivation is supplemented with the use of “overwhelming” or “indiscriminate” force. 
Objections along these lines go to the empirical keystone of the necessity argument. If alternative (and less destructive) methods are available, it would be misleading to characterize comprehensive siege starvation as militarily essential. Conversely, if encirclement deprivation is largely ineffective or counterproductive, the necessity claim is simply untenable. And if the method works only in contexts in which it is combined with actions that either render it superfluous or implicate the besieging party in clear war crimes, its potential efficacy is not justificatory.
However, resolving the question of the efficacy of siege warfare is not the project of this Article. For even assuming encirclement starvation can be effective, it does not follow that it must be permissible as a matter of legal interpretation or that it should be permissible as a matter of lex ferenda. The normative relevance of military necessity depends on a theory of the pro tanto wrong associated with the starvation of civilians. It depends, in other words, on a normative account of the crime.
Contemporary IHL and the associated war crimes regime have at their core a series of categorical prohibitions that hold irrespective of the potential military advantage associated with their violation. In contrast to the authorizing role of the necessity concept in earlier iterations of the regime,  “the idea today that military necessity may justify deviation from the LOAC [law of armed conflict] cannot be taken as a serious argument.” 
This is a legal reality with significant operational implications. Under the right conditions, one party to a conflict may stand to gain militarily by attacking the enemy’s civilian population to coerce their political leaders into withdrawal or capitulation. In some cases, doing so may be the only plausible route to achieving their ultimate objectives in the conflict. And yet, from the legal point of view, this imperative is irrelevant. The targeting of civilians who are not participating directly in hostilities is categorically prohibited and criminal, irrespective of military necessity. The arguments offered in defense of Allied “terror bombing” in World War II are simply unavailable within the normative framework enshrined in law today. Similarly, belligerents may not employ weapons that cannot be directed at a specific military objective, even if those are the most militarily effective weapons available to them in a context of urban warfare. Those who use such weapons in that context engage in war crimes.
The insufficiency of military necessity as the basis for legal permission is equally clear in the context of capture and detention. It is straightforwardly unlawful and criminal to kill enemy belligerents who have been captured and therefore rendered hors de combat, even if their capture occurs behind enemy lines and taking the prisoners in tow would thwart the mission of the capturing force. In such contexts, the capturing party must either release or detain, regardless of whether such options would be compatible with the goals of its operation. Similarly, if a detaining authority is unable to provide prisoners of war with adequate nutrition and is unable to ensure that standard via other means (such as external assistance), it must release and repatriate the affected prisoners. 
More prominently, and as discussed in greater detail below, efforts to justify torture on necessity grounds have, with rare and highly controversial exceptions, failed to gain legal acceptance. The rejection of pro-torture arguments has not relied exclusively on the empirical counterpoint that the practice is ineffective as a method of eliciting information. On the contrary, it has been maintained even in contexts in which efficacy is assumed. In that vein, Judge Aharon Barak wrote on behalf of the Israeli Supreme Court that a law-abiding state (or, in his framing, a democracy) “must sometimes fight with one hand tied behind its back.” Although the decision in which that claim is embedded has been critiqued persuasively as having the practical effect of legitimating, rather than constraining, detainee mistreatment, Barak was correct that compliance with the law of armed conflict does not guarantee the availability of all militarily advantageous methods.
Just as the military utility of a particular method is insufficient to justify deviation from a clear legal prohibition, that utility cannot be invoked in isolation to resolve ambiguities within the law of armed conflict. The status of military necessity as an underlying normative principle of the regime is itself contested. Moreover, even those who assume that military necessity does indeed perform such a function accept that it operates in productive tension with the often‑countervailing principle of humanity. Identifying that tension as the normative underpinning of the regime precludes the unchecked interpretive dominance of the former no less than it precludes the unchecked interpretive dominance of the latter.  Thus, even for those who assume that the law of armed conflict is best understood as seeking to accommodate “states’ interest in pursuing the object of war,” it is crucial “to avoid the misunderstanding that the LOAC generally, and military necessity in particular, justifies the unencumbered pursuit of war aims.”
Waxman acknowledges that arguments for the permissibility of siege starvation are rooted fundamentally in “similar reasons” to those underpinning purported justifications of indiscriminate bombing in World War II and the Vietnam War. This is telling. Indiscriminate attacks and attacks on civilian populations have been prohibited despite their potential military utility. Analogously, it is entirely possible that “[t]o the extent the rules [on starvation and humanitarian access] complicate the conduct of military operations, such complication is amply warranted by the crucial concerns of humanity served by provision of relief to civilian populations in need.” Whether that impediment to military efficacy is indeed warranted on those grounds depends in significant part on what is morally at stake in the starvation of civilians.
Ultimately, whatever one’s view on the general theory that IHL represents a balance between the imperatives of necessity and humanity, the categorical nature of at least some of the IHL prohibitions discussed above reflects a determination that the associated moral imperatives hold even in the face of significant military or other utility. International human rights law accepts the limitation of rights in pursuit of certain “legitimate aims” only as long as that limitation does not put the right itself in jeopardy. Along similar lines, IHL’s categorical proscriptions might be understood as efforts to protect core rights that cannot be sacrificed or jeopardized, even in the service of a war effort. Certainly, the absolute prohibitions on torturing or executing detainees may appropriately be understood in these terms. Determining whether something similar might be true of starvation as a method of warfare requires getting to grips with the normative underpinnings of the prohibition. It requires asking what, precisely, is criminally wrongful about engaging in the starvation of civilians as a method of warfare.
To ask what is criminally wrongful about a particular violation is to take seriously the notion that designating that violation a war crime entails a moral claim about the wrongfulness of the underlying conduct.  Arguably, this idea underpins criminal law generally. As Anthony Duff puts it:
What is distinctive about criminal law is that it inflicts not just penalties, but punishments—impositions that convey a message of censure or condemnation; the convictions that precede punishment are not mere neutral findings of fact, that this defendant breached this legal rule, but normative judgments that this defendant committed a culpable wrong.
It is now widely asserted that one of international criminal law’s core functions is to give voice to certain moral values and condemn their violation. That function is manifest most clearly in the punishment of those who commit war crimes, crimes against humanity, or genocide while acting in their home territory, pursuant to clear domestic legal authorization (and even obligation), against their co-citizens. Overriding the primary coordinative legal framework (domestic law) in a context in which there is no need to coordinate between that framework and another framework with equal sovereign authority cannot be understood with reference to the coordinating function of law. International criminal law’s requirement that any perpetrator who has a “moral choice” is dutybound to disobey such domestic laws manifests a substantive view about right and wrong.
The institutional agents of international criminal law have long been conscious of its normative posture. In his opening statement before the first major international criminal tribunal, chief American prosecutor at Nuremberg Robert Jackson argued, “When I say that we do not ask for convictions unless we prove crime, I do not mean mere technical or incidental transgression of international conventions. We charge guilt on planned and intended conduct that involves moral as well as legal wrong.” Decades later, amidst the revival of international criminal law, the International Criminal Tribunal for the former Yugoslavia emphasized its role in issuing “public reprobation” through punishment. The ICC has described its sentencing of convicted perpetrators as “an expression of the international community’s condemnation” and refers frequently to perpetrators’ “moral blameworthiness”  and “cruelty” in sentencing.
To recognize the moral expression inherent in international criminalization and punishment is not to deny the ways in which political factors shape the articulation of the expressivist mission and the way it is pursued. Notwithstanding Jackson’s lofty rhetoric, the tribunal before which he delivered that speech was normatively defective because it lacked the jurisdiction to examine crimes committed by the Allies. Although the tribunals that have been created since have not been quite as explicit in their selectivity or partiality, power remains a crucial factor in the distribution of criminal justice, and victors’ justice critiques (or close equivalents) endure.
That political context raises real challenges for the legitimacy of the application of international criminal law. However, it does not contradict the fact of a moral assertion at the heart of the regime. On the contrary, that condemnation is central to what stakeholders seek to control and harness to their political ends. Moreover, critiques that point to the specific political conditions of any given court’s activity as precluding the moral standing necessary for defensible condemnation are meaningful precisely because condemning is what such institutions do.
If punishment amounts to a particular form of moral expression, it is important to all stakeholders that the message inherent in any given criminal prohibition can be identified. Accused persons face the prospect of being followed by the “symbolic and condemnatory” implications of the specific crimes of which they are convicted. Conversely, the law’s capacity to express solidarity with victims is contingent on the relevant criminal categories capturing the wrong they suffered in a morally intelligible way.  Meaning matters also to the broader communities affected or implicated and to legal and institutional actors seeking to operate consistently within the system.
It is in this context that the “fair labeling” principle—the requirement that “widely felt distinctions between kinds of offences and degrees of wrongdoing are respected and signalled by the law”—is gaining purchase as a human rights imperative. It is connected through the principle of legality to the idea of “fair warning,” pursuant to which “citizens should be advised [in advance] as to how their conduct will be censured.” It must be possible for those to whom, and on behalf of whom, the law speaks to “discern the criminal law’s designation of [the] wrong [associated with each crime] as distinct and deserving of its own condemnation.”
If anything, the stakes associated with this principle are magnified in international criminal law, where verdicts are of profound political significance to large constituencies, and the crimes are thought to carry an elevated stigma.  Mirjan Damaška emphasizes the importance of ensuring that distinctions in this domain rest not on legal technicality, but on “moral distinctions shared by ordinary people.” 
The significance of moral meaning in international criminal law is exemplified by the intensity of feeling regarding whether acts that clearly qualify as crimes against humanity also constitute genocide. This distinction was central to the public discourse around several International Criminal Tribunal for the Former Yugoslavia (ICTY) judgments on the status of persecutory violence at Srebrenica (where genocide was affirmed) and in other Bosnian municipalities (where it was not).  The issue provoked a diplomatic fracas at the Security Council over a symbolic resolution commemorating the Srebrenica genocide. The genocidal status of clear crimes against humanity has also been a point of debate with respect to the Khmer Rouge atrocities prosecuted before the Extraordinary Chambers in the Courts of Cambodia.
In none of those instances was criminal liability for the impugned acts ultimately at stake—the crimes against humanity convictions were straightforward and often overdetermined. Equally, the different possible verdict combinations portended no formal penalty differential, and the sentences, in at least some of the cases, were anyway fairly certain to be longer than the defendants’ remaining life expectancy. At stake was simply whether the court would identify the attacks as distinctively genocidal in their wrongfulness. Even when a de facto life sentence is guaranteed, the question of expressive classification remains of enormous importance to stakeholders across the spectrum.
Genocide’s unique status in the popular imagination is such that debates as to its applicability tend to take on a heightened public profile. But other examples pervade the practice of international criminal law, including on issues such as modes of liability. In the area of gender-based and sexual violence, the codification of crimes has developed over time to encompass a broad range of specific offenses, including not only outrages upon personal dignity and rape, but also sexual slavery, enforced prostitution, forced pregnancy, and enforced sterilization. In the first international conviction for the crime against humanity of forced pregnancy, the ICC acknowledged the position of some states in drafting negotiations that “the crime was unnecessary because its elements were already covered by the crimes of rape and unlawful detention in the Statute.” However, the Chamber reasoned, “As with any crime, forced pregnancy must be interpreted in a manner which gives this crime independent meaning from the other sexual and gender based violence crimes in the Statute.”  This, the judges held, “implicates the principle of fair labelling, and how the proper characterisation of the evil committed, that is to say, calling the crime by its true name, is part of the justice sought by the victims. It is not enough to punish it merely as a combination of other crimes.” 
For similar reasons, the Special Court for Sierra Leone (SCSL), the Extraordinary Chambers in the Courts of Cambodia (ECCC), and the ICC have reached beyond the offenses codified explicitly within their statutes to articulate and distinguish under the category of “other inhumane acts” the crime against humanity of forced marriage. Here, too, an individual guilty of that crime is almost certain to be guilty of some combination of other crimes against humanity (such as forced labor, enslavement, sexual slavery, or forced pregnancy) for the same impugned acts. However, forced marriage was elaborated on the grounds that, even in aggregate, those other offenses do not fully capture the wrongfulness of forcibly rendering someone a spouse. In the landmark SCSL judgment, forced marriage was deemed to include “a forced conjugal association” that is distinctive in its exclusivity, its entailment of mutual obligations, and in inflicting a particular form of “long-term social stigmatization” arising from the marital association, with all of the challenges that poses for community reintegration.
Further exemplifying the importance of moral meaning in the expression of international criminal law is the Rome Statute’s approach to the crime against humanity of persecution. Article 7(1)(h) provides that persecution must be committed “in connection with” a “crime within the jurisdiction of the Court.” As such, acts qualifying for this crime are guaranteed also to qualify for at least one other ICC crime. To charge such acts as persecution, rather than the alternative ICC crime, is meaningful only in identifying discriminatory intent. There is no formal penalty differential between the criminal categories, and discriminatory intent can anyway be an exacerbating factor in sentencing for any crime. As such, the value of the distinct category inheres not in its implications for retribution, deterrence, or coordination, but rather in enabling the specific condemnation of discriminatory atrocity.
Ultimately, for a practice to qualify as an international crime, it must be assumed from the legal point of view that it entails a particular kind of wrong. A normative account of the crime responds to the imperative to clarify that moral meaning and to distinguish the meanings of overlapping prohibitions, thus respecting the rights of victims and the accused and ensuring the integrity of criminal law’s expressive function. A normative account can also shed light on interpretive questions, inform charging and other prosecutorial decisions in contexts of scarce resources, and help to sharpen the work of civil society actors seeking to use systems of international criminal justice as part of a broader strategic response to malign policies. 
In contemplating the appropriate account of starvation, it is worth reflecting on the range of kinds of wrongdoing covered by war crimes law. One way of classifying those crimes would map them onto roughly the following issue areas:
- attacking legally protected persons in the conduct of hostilities (such as surrendering or wounded combatants, medical personnel, humanitarian actors, peacekeepers, or civilians [including via disproportionate or indiscriminate attack]);
- mistreating individuals under one’s immediate custody or control (including by inflicting outrages upon their dignity, subjecting them to cruel treatment, mutilation, medical experimentation, murder, or torture, using them as hostages, or prosecuting them without fair trial protections);
- sexual or gender-based violence (including inflicting outrages upon dignity, engaging in rape, sexual slavery, enforced prostitution, forced pregnancy, enforced sterilization, and other forms of sexual violence);
- pillaging property or attacking or causing excessive damage to protected objects (such as civilian property, medical or humanitarian material, units, or transport, cultural heritage, and the environment);
- drawing on or compelling the combatancy of persons protected from performing that function (such as children, enemy nationals, or enemy prisoners of war);
- deporting, forcibly transferring, or inflicting collective punishment on the civilian population;
- abusing the control afforded by belligerent occupation (such as by transferring one’s own population into the occupied territory);
- endangering persons or undermining the law of armed conflict by engaging in certain rule-breaking conduct that is prohibited even when no malign consequence arises (such as using prohibited weapons or declaring no quarter); and
- exploiting the laws of war in ways that put innocent persons at heightened risk (such as by using human shields, engaging in perfidy, or making improper use of flags of truce or medical emblems).
To be clear, this is not a formal classification; one could draw the lines differently. Moreover, there are overlapping normative features across these clusters, as well as important elements that distinguish the individual crimes within each cluster, as the unique expressive value of each crime demands. For example, one might distinguish child soldiering from other forms of compelled participation in armed conflict due to the distinctive concerns associated with the protection of children, as compared to adults. Nonetheless, the categorization above indicates the expressive diversity within the war crimes category.
In contemplating the place of the starvation war crime in this framework, it is worth noting that most manifestations of starvation of civilians as a method of warfare could in principle qualify for multiple war crimes, including attacks on civilians, collective punishment, forced transfer or deportation, indiscriminate (or disproportionate) attacks, and attacks on civilian or other protected objects. Cases of widespread or systematic starvation (which might be thought to include all instances of starvation as a “method” of warfare) could also qualify as one or more crimes against humanity, including, most obviously, murder, extermination, forcible transfer, or other inhumane acts. A persuasive normative account of the starvation crime ought to make sense of it as a war crime, while also identifying the features that warrant its distinct articulation and application.
The simplest account of the core wrong of illegal starvation tactics is that such methods inflict a range of harms, and ultimately death, on a category of persons not liable to be attacked in that way, even in war—namely, civilians who are not taking a direct part in hostilities. At the very least, starvation tactics endanger civilians in those respects. Thus understood, the crime of using starvation of civilians as a method of warfare might be thought to fit primarily and straightforwardly within the first cluster of war crimes in the list above, as a form of attack on a protected category of persons (civilians).
This is a common theory of the crime. One of the leading commentaries on the Rome Statute identifies article 8(2)(b)(xxv) as fundamentally “an application of the prohibition to attack civilians.” Contributors to a key IHL text argue similarly that the starvation ban is “[d]erived from the principle of distinction.”  The International Committee of the Red Cross’s commentary on the underlying rules in the Additional Protocols supports this position, describing the prohibition on starvation methods in Article 14 of Protocol II as “really only a specific application of common Article 3, which imposes on parties to the conflict the obligation to guarantee humane treatment for all persons not participating in hostilities, and in particular prohibits violence to life.” Announcing an indictment related to alleged war crimes perpetrated in Sudan between 1998 and 2003, the Åklagarmyndigheten (Swedish Prosecution Authority) offered the practice of burning crops as an example of an indiscriminate attack or an attack on civilians. The International Commission of Inquiry on Darfur approvingly cited the U.K. Manual of the Law of Armed Conflict for the view that the starvation ban “follows” from the legal fact that “[v]iolence to the life and person of civilians is prohibited, whatever method is adopted to achieve it.” It is, on that interpretation, a prohibition rooted ultimately in the right to life. Notwithstanding the complexity associated with the meaning of “attack” or “violence” in IHL, the U.K. Manual extends this reasoning to sieges, blockades, and the “blocking of relief supplies with the intent of causing starvation.” Along similar lines, the fact that starvation tactics lead ultimately to the infliction of death is thought to implicate extermination (mass‑murder) as the primary manifestation of this wrong in the crimes against humanity category.
On the one hand, this common view of the prohibition’s normative core gains credibility from its fit with the general criminality of attacking civilians and civilian populations. On the other hand, precisely that connection raises the question of whether the starvation ban has a distinct function or meaning that cannot be subsumed by the general prohibition and criminalization of attacks on civilians. Several such functions might be invoked.
First, the separate articulation of the starvation ban eliminates any ambiguity as to whether the principle underpinning the ban on attacking civilians also applies legally to methods that inflict the same fundamental wrong via a different route. Until relatively recently, the application of that underlying principle to encirclement starvation was not obvious. Walzer observed of the normative posture of what he terms the “war convention” as it existed shortly before the agreement of the Additional Protocols, “If there is a general rule that civilian deaths must not be aimed at, the siege is a great exception.”
In light of both the historical permissibility of siege warfare and ongoing debates regarding the meaning of “attack” in IHL, it could prove complicated to rely on the war crime of attacking civilians as the basis for prosecuting encirclement deprivation. Notably, in his defense of the permissibility of such tactics, Watts (like Walzer) is explicit about the status of siege starvation as a normative “outlier” compared to the rules protecting civilians from attack in IHL. Given that context, the specific criminalization of starvation methods, including with explicit reference to deprivation by impeding the delivery of relief supplies, can be understood to serve the function of dispelling doubt about whether such methods are covered by the more general war crimes related to the protection of civilians.
Second, the separate articulation of the starvation crime might be thought to obviate some of the practical obstacles associated with attaching other war crimes and crimes against humanity to starvation methods. At the crux of these difficulties is the fact that the outcome to which the criminal proscription is putatively attentive (namely, death) is several steps removed from the impugned conduct. The ultimate cause of death in famines is typically not starvation itself, but infection with one or more of the communicable diseases that proliferate in such situations.  Depriving populations of essentials enables such proliferation, but outbreaks inevitably involve multiple factors, including complications arising from the policies of the besieged authorities. 
That reality affects the viability of any criminal prohibition that requires establishing that the impugned conduct caused a proscribed consequence. At the ICC, that category of prohibitions includes the war crimes of killing, murder, willfully causing great suffering, inflicting excessive collateral injury or death on civilians, and the crimes against humanity of murder and extermination.  At the ad hoc tribunals, the crime of attacking civilians was considered established only if the attacks could be shown to have caused death or serious injury. The latter requirement has not been incorporated into the corollary ICC crime, but the aforementioned ambiguities regarding the meaning of “attack” remain pertinent in that context.
These obstacles are not hypothetical. The Supreme Court Chamber of the ECCC reversed a finding of extermination associated with the starvation of persons during the transfer of hundreds of thousands under the Khmer Rouge on the grounds that the link between mass deprivation and large-scale death was not established. In the context of the siege of Sarajevo, the U.N. Commission of Experts found the supply of essentials to have been “extremely limited” but determined that there was no indication that this had caused deaths. The implication was clear: “As no one appears to have died of starvation, cold, or dehydration in Sarajevo, it is unlikely anyone could be held liable” for siege starvation. The policies of the besieged Sarajevan authorities further complicated the question of causation. Seen in this light, the ICC starvation crime might be thought to provide distinctive value by focusing on endangerment and thus eliminating the peculiar difficulty of establishing causation in this context.
Third, the distinct articulation of starvation within the normative cluster of attacks on civilians and other protected persons might also be thought to enable the precise identification of a particular form of that wrong. Starvation inflicts specific kinds of physical and psychological suffering, has life-changing impacts on children, heightens the affected population’s vulnerability to the spread of infectious disease, contributes to mass displacement and economic deprivation, portends transgenerational reverberating effects, and ultimately kills via these and other mechanisms. Although kinetic attacks on civilians also cause a spectrum of immediate and reverberating harms, both the longer temporal dimension and the nature of the harms associated with starvation are distinctive. Thus, starvation tactics have been described as inflicting “death in slow motion.”  Characterized in that way, the starvation crime might be understood to serve a function within its war crimes cluster parallel to that served by the act of inflicting conditions of life calculated to bring about physical destruction within the umbrella category of genocide. 
In sum, on this common view, the basic wrong at the heart of the starvation crime is the targeted or indiscriminate endangerment of persons who ought not be attacked in that way even in armed conflict. Codifying a distinct starvation crime within that broader category serves three functions. First, it removes any ambiguity regarding the criminality of starvation by encirclement or by the removal of essentials. Second, it obviates the unique challenge of establishing causation in this context. Third, it allows for a more precise condemnation of the particular forms of suffering and death specific to starvation methods.
B. Necessity Revisited: Graduality and Civilian Harm
Accepting this account, one might think that the straightforward interpretive implication would be that civilian starvation should be understood to be broadly and categorically prohibited. As noted above, attacks targeting civilians are prohibited and criminal. Similarly, it is prohibited for belligerents to engage in attacks that fail to distinguish between civilians and combatants (or other military objectives) or that use means or methods of warfare that cannot be directed at a specific military objective. Indiscriminate attacks are prohibited and criminal under customary international law. Although the Rome Statute does not address such attacks explicitly, an ICC Trial Chamber has determined that attacks inflicted “indiscriminately” on “civilians and fighters alike” amount to conduct undertaken with a dual purpose of targeting both combatants and civilians. Moreover, populations comprised predominantly of civilians retain their protected status under that rule, even when there are combatants ensconced within. The customary and Rome Statute war crime of attacking such a population attaches when the population as a whole is made the target of attack. 
These prohibitions are categorical. It is criminal to attack civilians or the civilian population regardless of the military advantage that such an attack might return, regardless of whether the ultimate goal is to inflict suffering on civilians or to neutralize the combatants ensconced within, and regardless of whether there are other available routes to military success. Although there are ongoing disputes about precisely when explosive weapons with wide-area effects may be used in populated areas,  it is clear that the ban on attacks that fail to distinguish between civilians and combatants or that use means or methods incapable of such distinction applies even if their indiscriminate fire may have been motivated by apparently benign goals, such as force protection, and even if the attacking party lacked alternative means. Military advantage can be invoked to justify attacks inflicting civilian loss only when that loss is part of the collateral damage arising from discriminate operations targeted at specific military objectives. Even then, the civilian loss expected cannot legally be excessive in relation to the concrete and direct military advantage anticipated.
Through these rules, contemporary IHL protects the right of civilians not to be instrumentalized or treated as expendable in war, unless they have acted to forfeit that right through participating directly in hostilities. Assuming the starvation crime to be rooted in that same principle, one might think that the method would be subject to a similarly categorical proscription. On that view, a starvation siege that deprives the encircled (civilian) population of essentials to starve out the combatants within would be straightforwardly prohibited, as would deprivation operations that inflict starvation on combatants and civilians without discrimination. As with any other attack on civilians or the civilian population, military advantage would offer no basis for interpreting the prohibition more permissively.
And yet, as discussed above, military necessity is understood by some to demand a less stringent approach in the starvation context. On that view, military imperatives demand that the targeted starvation of encircled (civilian) populations and the indiscriminate starvation of combatants and civilians are (or at least should be) permissible in siege situations as long as they are performed with a view to starving out the combatants ensconced within, rather than weaponizing civilian suffering. To be persuasive, such arguments must explain why the military utility of siege deprivation is sufficient to override the pro tanto claims of the affected civilians, when those same civilians have normatively and legally dispositive claims against being subjected to analogous kinetic attacks.
The relevant normative distinction between encirclement deprivation and kinetic attacks cannot inhere either in the fact of military utility (which is both contingent and applicable in principle to either method) or in the ultimate impact on the civilian population (which can be similarly severe in either case). Most plausibly, it would inhere instead in the divergence in the methods’ respective temporal scopes and the implications of that divergence for harm, intent, and responsibility.
Kinetic attacks on civilian populations are characterized primarily by their immediate consequences. Even their reverberating effects are not easily reversed after the fact. When a bomb is dropped, the fullness of the harm entailed is inflicted or set irreversibly in motion. Those killed and maimed are killed and maimed in that moment. The suffering of survivors endures and can be mitigated or exacerbated over time, but its fundamental source is set. The immediacy of these consequences is such that engaging deliberately in the attack entails adopting an intentional posture towards the infliction of civilian harm and death.
The infliction of civilian harm in a siege occurs on a different trajectory. Rather than a moment of immediacy, it is a process of steady escalation, the pace of which can be described comparatively as “glacial.” One of the consequences of that graduality is that the apocalyptic conclusion of famine and mass fatality is relatively uncertain and distant in the early stages. Therefore, whereas the commander who orders kinetic attacks on a civilian population cannot avoid an intentional posture with respect to the resultant civilian harm and death, those involved in siege deprivation may seek to characterize their intention in terms of stimulating capitulation before there is any significant civilian loss.
In that vein, Dinstein insists that Protocol I “completely fails to take into account the inherent nature of siege warfare,” namely that “a siege does not generate starvation for the purpose of killing civilians with hunger, but only in order to cause the encircled town to surrender.”  Various authorities, including some military manuals, adopt a similar way of framing the practice, describing the goal as “forcing [ ] surrender,”  eliciting “submission,”  or generating “capitulation,” as explicitly distinct from purposefully “killing [the inhabitants] with hunger.”  Hampson argues that often “the besieging force” structures the deprivation so as “to let [the besieged civilian population] go significantly hungry” but not to starve to death. Although this may entail subjecting civilians to “years with not enough to maintain healthy bodily functions,” she is skeptical that it should be understood as prohibited under the current rules: “starvation is starvation, it is not hunger.”
The point is partly comparative. Encirclement deprivation is framed on this view as “limit[ing] the heavy civilian casualties often associated with urban fighting.”  It is a way to “subdue an enemy while limiting direct hostilities” and “avoiding a large-scale atrocity.” The alternative routes to taking the defended locality may be expected to “increase the intensity of the fighting and the associated risks of incidental harm for civilians,” or, worse, may lead to the temptation to “flatten cities and inflict massive suffering.”
Of course, encirclement deprivation also threatens massive suffering. And although threatening a wrong is generally not as bad as consummating the threat, threats can themselves have psychological consequences sufficiently grave to warrant war crime status. Moreover, starvation methods rarely if ever work as pure threats. Capitulation is a response to the escalation of deprivation, not its prospect. Any effort to distinguish starvation methods must grapple with that reality.
Those who defend the permissibility of encirclement deprivation might argue that there is a normatively meaningful difference between the gradual escalation of suffering in that process and the execution of the threat of a kinetic attack on civilians. The latter entails a binary shift from hypothetical to real killing and maiming. Although the execution of such threats can be incremental at the aggregate level, as in the case of a series of small-scale terror attacks on random civilians, this aggregate incrementalism operates alongside an individual binarism—there is nothing gradual about what is done to the direct victims of such attacks. In contrast, the continuation of a siege entails (at each stage) an escalation of suffering that is not only incremental in aggregate, but incremental for the affected individuals. Parts of the population are affected sooner and more acutely than others, but even for the hardest hit, the consequences develop over time. Irreversible impacts, such as death or permanent health effects, occur not in an explosive instant, but through a cumulative and extended process, arising from the combination of the continuous series of decisions to persist with the siege and various intervening factors that may exacerbate or ameliorate its effects.
With encirclement deprivation understood in these terms, it might be argued that intent vis-à-vis its effects can be updated and reframed continually according to the ever-present possibility of capitulation. If one were to accept the controversial normative premise that suffering inflicted up to that point can be treated as a moral sunk cost, even the continuation of a deprivation siege in the late stages might be framed in terms of the besieging party’s ongoing hope that capitulation is imminent. Although civilians would have a significant pro tanto claim against that continuation, the combination of incrementalism and the permanent offer of a way out might be invoked to suggest the possibility of a different and more accommodating path to the justification of encirclement deprivation than is available for indiscriminate bombardment or other violent attacks on civilian populations.
In sum, whereas attacking the civilian population entails the intentional killing or maiming of civilians who have not lost their protection from targeting, siege deprivation entails the scalar escalation of pressure combined (at least in principle) with the constant offer of a path to the avoidance of the escalating harm. As such, the discrimination between civilians and combatants absent from the initial deprivation is always possible, including through civilians’ own willingness to accept the available egress or through their leadership’s capitulation. Therefore, although the coercive imposition of deprivation entails a pro tanto wrong, it might be argued to be less severe and direct an abrogation of the rights of those affected than would be inflicted by a kinetic attack. Indeed, some argue that civilians who eschew egress when it is offered thereby forfeit any right not to be subject to the starvation regime.
Overall, there appears to be a two-step normative structure to the claim that siege starvation is appropriately subject to a less restrictive framework under IHL than are analogous kinetic operations. First, just as with a bombing campaign or an urban assault, the pro tanto wrongfulness of starvation methods inheres in the targeted or indiscriminate harm and death inflicted on persons not liable to such attacks. As such, starvation methods and direct kinetic attacks can be evaluated on the same dimension. Second, the fact that encirclement starvation inflicts those harms via gradual and incremental escalation, rather than through an act of immediate violence has normatively significant implications for whether and how that wrong is realized. In particular, encirclement starvation reduces the harm intended by the besieging party at any point, allows in principle for the continuous possibility of escape from that harm, and mitigates the besieging party’s responsibility for harm that could have been avoided by capitulation or civilian egress. Or so the argument might go.
Even on its own terms, this line of reasoning has several significant and interrelated infirmities. First, it relies on the assumption that siege starvation can be effective prior to the ultimate infliction of widespread civilian harm or death. Without that assumption, the notion that intent can be framed around imminent capitulation lacks credibility. And yet, the assumed efficiency of the method is belied by experience. Even “successful” encirclements are often extended and devastating in their effects. It is true, of course, that the efficacy and cost of a given method are always variable. However, given the stakes, the invocation of military necessity in this kind of argument depends for its normative credibility on a particularly robust set of empirical presumptions.
Second, it is distorting to evaluate each incremental decision to initiate or continue a siege or other starvation method in isolation. Encirclement deprivation can have a broad and calamitous aggregate impact over time. The length of the timeline on which that cumulative harm manifests may dull our attentiveness to it, but that reality only emphasizes the importance of elevating its urgency in law. Certainly, the hope of a quick and relatively harmless resolution cannot justify evaluating the initiation of a starvation campaign without reference to its likely or expected implications over the full course of its implementation. Similarly, assessing the continuation of a long-standing siege with exclusive reference to the immediate future relies on the controversial notion that the prior accumulation of wrongful harm and suffering can be dismissed as normatively irrelevant to the decision to persist. Particularly when that prior suffering and the projected future impact are intimately connected as the results of the continuous execution of a single overarching threat, such a disaggregated evaluation is not persuasive.
Third, that civilian egress or capitulation could have avoided or limited the harm of a starvation siege does not mitigate the perpetrator’s responsibility for the harm inflicted. Civilians are under no obligation to leave the encircled area. When they decline to do so, it is legally unambiguous that they do not thereby participate in hostilities or become liable to be targeted. When they are forced not to leave by the besieged party, this does not exempt the besieging party from its responsibilities towards them; rather, it indicates the shared responsibility of the two warring parties for those civilians’ fates. That the besieged and besieging authorities share responsibility in such a situation entails neither that responsibility is divided, nor that the responsibility of the besieging party is in any other way diminished; it simply means that multiple actors bear responsibility for the wrong. Even those civilians who do flee to escape starvation may well be victims of the crime against humanity of forced movement.
However, these problems with the two-step argument for a permissive posture with respect to siege starvation are not the most fundamental. The deeper problem concerns the conceptualization of the normative underpinnings of the crime. Siege starvation is not merely an anomalously slow mechanism by which harm or death is inflicted in war. As elaborated in the next two Sections, it is better understood as a process by which biological imperatives are turned against fundamental human capabilities in a manner more normatively reminiscent of torture than it is of a kinetic attack. This torturous nature of mass deprivation shows the graduality of the method to be not a mitigating element, but a key aspect of its distinctive wrongfulness. Understanding the crime in this way provides the basis for an argument for its categorical prohibition.
As reconstructed in the previous Section, the necessity argument for the permissibility of encirclement starvation bears striking resemblance to necessity arguments sometimes invoked to defend the use of torture in interrogation. As elaborated below, the responses to the latter arguments can shed light on how to think about necessity and wrongfulness in siege starvation.
The Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (CAT) states the unequivocal position of international law: “No exceptional circumstances whatsoever, whether a state of war or a threat of war, internal political instability or any other public emergency, may be invoked as a justification of torture.” The European Court of Human Rights takes this to reflect a deeper moral truth: “The philosophical basis underpinning the absolute nature of the right under Article 3 does not allow for any exceptions or justifying factors or balancing of interests.” The ban on torture is among the clearest red lines in international law’s protection of the human being.
This absolutism notwithstanding, skeptics of the categorical ban have long argued for a necessity exception. The “ticking bomb” scenario has so pervaded discussion on the issue as to require no introduction here. It informed the highly criticized and ultimately withdrawn executive branch legal memoranda during the George W. Bush administration; it was invoked by multiple candidates, including the eventual winner, in the 2016 U.S. presidential campaign; and it has informed the controversially permissive posture of the Israeli Supreme Court. As in the context of siege starvation, necessity has been invoked in both claims about proper legal interpretation and claims about what the law ought to be.
Thus far, efforts to claim a necessity exception in law have been rejected by the overwhelming majority of legal authorities. Many have also pushed back on the normative claim. As in the case of siege starvation, some question the empirical premises or utility of the ticking bomb scenario. One common line of argument along these lines challenges the premise that torture can produce accurate information. Another contends that the ticking bomb scenario does not reflect real-world conditions. A third holds that structuring a legal permission around it would lead inevitably to a significantly broader and morally unjustifiable opening for legally sanctioned torture. These are all plausible and persuasive lines of rebuttal.
Another response, however, goes deeper, articulating the wrongfulness of torture in a way that distinguishes it from other forms of violence or harm that are thought to be more readily justifiable with reference to defensive necessity. From this perspective, to get drawn into the empirical debate about whether torture works is to concede too much. The independent and more fundamental moral thesis is that such techniques are categorically wrongful, even if at least some forms could be effective in some circumstances. Two arguments along these lines are worth contemplating here.
B. The Wrongfulness of Torture
In a seminal 1978 paper, Henry Shue responds to a common form of the necessity argument for the permissibility of torture. Shue frames the line of thought to which he is responding as follows: “[S]ince killing is worse than torture [in that the former involves the total destruction of the person, whereas the latter involves only partial destruction], and killing is sometimes permitted, especially in war, we ought sometimes to permit torture.” Rejecting the premise that the two acts can be compared simply along the dimension of harmfulness, Shue sets out to identify qualitatively distinctive aspects of torture that warrant treating it differently from ordinary belligerent killing, even when the detainee is an adversary with information about an imminent threat, or at least appears to be so with as high a probability as is necessary to render someone a legitimate target in war.
Almost three decades later, in another influential contribution on the same topic, David Sussman asked, “What is it about torture that sets it apart even from killing, maiming, or imprisoning someone, such that the circumstances that might justify inflicting such harms would not even begin to justify torture?” Shue’s and Sussman’s answers are distinct in important ways, but they also share elements that illuminate the wrongfulness of mass starvation.
For Shue, the first step in distinguishing torture morally is to recognize that it is an attack on the defenseless. Torture begins, he argues, “only after the fight is—for the victim—finished.”  As such, it is unlike “the killing in battle of a healthy and well-armed foe.” It fails even a “weak constraint of being a ‘fair fight.’” The victim is entirely at the torturer’s mercy, so the assault on her person “sinks below even the well-regulated mutual slaughter of a justly fought war.” Sussman similarly emphasizes the “absolute” “asymmetry of power, knowledge, and prerogative” between interrogator and detainee in a torture situation.
Shue’s and Sussman’s shared premise is reflected in various respects in the law. For torture to be established as a crime against humanity at the ICC, the Rome Statute requires that the victims have been “in the custody or under the control” of the perpetrator at the time of the abuse. Although there is no formal element along those lines in the war crime definition, the crime applies only to the torture of fighters rendered hors de combat, civilians, medical personnel, or religious personnel taking no active part in the hostilities, and it is therefore “difficult to conceive of the crime being committed” without control or custody. Similarly, a leading commentary on the CAT argues that torture presupposes the exercise of “total control” by one person over another.
Notwithstanding this profound asymmetry, Shue recognizes that the torture victim’s defenselessness is, as a philosophical matter, open to objection. Specifically, the defender of permissible torture might insist that a detainee is not “utterly helpless in the face of unrestrainable assault as long as he or she holds in reserve an act of compliance which would satisfy the torturer and bring the torture to an end.” Along these lines, Sussman suggests that in the ticking bomb scenario, one can frame the detainee’s “continued silence” as “part of his attack.”
Shue and Sussman are ultimately unpersuaded by such objections. For Shue, the way out that is putatively available to the detainee with the information necessary to stop the torture is, morally speaking, no way out at all. This is because torture functions here as a way of overriding the victim’s fidelity to her own higher-order commitments. For the detainee, compliance entails a “betrayal of one’s ideals and one’s comrades.” This, Shue argues, is pivotal: “The possibility of betrayal cannot be counted as an escape,” at least not when those commitments are “profoundly held” (as the necessity of torture would seem to suggest they are). In short, the distinctive wrongfulness of torture—that trait which distinguishes it from other forms of bodily harm or death—inheres, on this view, in the fact that it is an assault on the defenseless, whose only escape is a “violation of integrity,” and is thus no escape at all.
Sussman questions Shue’s framing insofar as it suggests that the wrong of torture would appear to be greater the stronger the detainee’s commitment not to comply, which, particularly in the case of someone presumed to have information about a terrorist threat, may be thought a perverse result. Nonetheless, the account he offers in place of Shue’s also defines torture’s singular wrongfulness with reference to the infliction of the conditions of self-betrayal.
In that vein, Sussman draws the following distinction:
Torture should be distinguished from both coercion and brainwashing, even though all three may often overlap in particular cases. What is distinctive about torture is that it aims to manipulate its victims through their own responses, as agents, to the felt experience of their affects and emotions in a context of dependence, vulnerability, and disorientation. Coercion, in contrast, need only exploit the agent’s rational responses to the cognitive content of these feelings. The coercer tries to influence his victims through their own appreciation of their reasons for action. . . . In principle, we could coerce a being that has no emotional life at all (e.g., a corporate agent such as a state or a university), so long as this being had determinate interests that it rationally pursued in part by anticipating the intentions and actions of other rational agents. But we could not in principle torture this sort of artificial person.
The key difference for the victim of torture, as opposed to the victim of coercion, is that the former must confront the dilemma of whether to give in (and thus violate her or his higher commitments) “not merely with respect to the disvalue of pain and fear, but while caught up in the experience of these very feelings themselves.” Being “caught up” in these feelings precludes deciding reflectively in response to one’s higher-order commitments and the threats posed by the coercive agent. As Sussman puts it, “[w]hen sufficiently intense, pain becomes a person’s entire universe and his entire self, crowding out every other aspect of his mental life. Unlike other harms, pain takes its victim’s agency apart ‘from the inside.’” This is because pain triggers a biological imperative—“something like a bodily demand to change something about one’s condition.” As such, “[n]ormally, one cannot adopt a purely contemplative attitude toward one’s own pain.”
Shifting to using the first person to describe the victim’s perspective, Sussman argues that the fact that pain is experienced as a demand from the victim’s own body means that, although inflicted by the torturer and thus “not unproblematically an exercise of my own agency (the way my reflectively adopted commitments might be),” it is also not “something fully distinct from such agency.” As such, “My suffering [as torture victim] is experienced as not just something the torturer inflicts on me, but as something I do to myself, as a kind of self-betrayal worked through my body and its feelings.” To be clear, it is not necessary that the victim actually comply with the torturer’s demand and thus perform the final act of self-betrayal. The wrong does not inhere in that consequence. Rather, it inheres in twisting the body against the person: “Even if the victim does not break, he will still characteristically discover within himself a host of traitorous temptations.”
Together, Shue and Sussman capture crucial features of what makes torture distinctively wrongful. Torture is not simply the infliction of severe pain or suffering on a defenseless victim; it is the weaponization of the biological imperative to avoid pain and suffering to subjugate the victim’s higher-order desires to her body’s demands. It uses the body to turn the victim against herself and undermines one of the traits we take to be most fundamental to our humanity—our capacity for higher-order commitments.
In addition to being philosophically illuminating on their own terms, these arguments offer the foundation for a compelling account of the international legal prohibition and criminalization of torture. The mere infliction of severe pain or suffering is ordinarily not sufficient to establish torture. Rather, in most legal definitions, the infliction of harm must be structured to turn the victim against herself in one way or another.
The most prominent definition of torture is enshrined in Article 1 of the CAT, which requires that the severe pain or suffering be:
intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind.
In each of these purposes is entailed a form of weaponization of the body against the self—as a mechanism by which to cause self-betrayal through confession or other undesired action, as a mechanism by which to bring the body into conflict with past actions through punishment, or as a mechanism by which to bring the body into conflict with identity through discrimination.
Torture’s purpose element is not marginal. As a leading commentary on the CAT explains:
the severity of pain or suffering, although constituting an essential element of the definition of torture, is not a criterion distinguishing torture from cruel and inhuman treatment. . . . Whether or not cruel or inhuman treatment can also be qualified as torture depends . . . above all [on] whether inhuman treatment was used for any of the purposes spelt out.
Although central to the anti-torture regime, the CAT is not the only source for the meaning of torture in international law. Nonetheless, something like the purpose element it articulates has been recognized by a range of human rights authorities, including those interpreting treaties that lack a detailed internal definition of the concept.
There are notable exceptions to this pattern. Indeed, the Rome Statute is internally split on the issue, with the crime against humanity of torture including no purpose element, but the war crime of torture requiring that the prosecutor establish one of the wrongful purposes. However, as things stand, the purpose element remains dominant; both the international criminal tribunals and the ICC (in the war crimes context) have relied on it to distinguish torture from other forms of mistreatment.  Even in the Inter-American system, where the definition of torture is explicitly not purpose-specific as a matter of treaty law, the Inter‑American Court of Human Rights has nonetheless had occasion to emphasize the function of torture in breaking the victim’s will.
By explaining why torture is not simply the most painful form of cruel treatment, but a qualitatively distinctive kind of wrongful treatment—a kind that is not specified by the broader category that already guarantees its criminality—Shue’s and Sussman’s accounts clarify why it is important to condemn that particular wrong with precision. The weaponization of the victim’s biological imperatives against her person is central to that explanation.
In the context of detention, the link between starvation and torture is straightforward—starving detainees is one way of inflicting upon them the severe pain or suffering central to torture. However, as discussed in the first Subsection below, a belligerent using starvation methods in the conduct of hostilities rarely controls individual victims in the manner generally thought to be necessary for the legal applicability of torture. Equally, torture as a legal category does not capture certain morally constitutive features of mass starvation as a method of warfare. Specifically, as discussed in the third Subsection below, the use of starvation in the conduct of hostilities involves a social dimension that has no clear equivalent in the context of detainee mistreatment. Together, these distinctions mean that the starvation of civilians as a method of warfare is likely not, and certainly not simply, a form of torture in the legal sense. And yet, the parallels are illuminating. The torturous quality of starvation as a method of warfare distinguishes it from indiscriminate attacks or attacks on civilian populations.
Consider first the question of control. It is possible that innovative legal arguments to stretch torture (or at least inhuman treatment)  to cover the use of mass starvation in warfare will gain traction. The two most obvious interpretive routes to that outcome would be to eschew a control element when one is not explicitly provided or to interpret “control” loosely to include the capacity to determine a population’s access to a sufficient supply of indispensable objects. However, whether or not such efforts are successful, there are normatively meaningful distinctions between the kind of control traditionally associated with torture and that associated with conduct of hostilities crimes, such as encirclement deprivation. Those differences suggest that there is some value to having different legal categories for the condemnation and punishment of the two wrongs.
Whether isolated or widespread, random or systematic, the infliction of torture is fundamentally an individual process. No matter the number of their victims, those engaged in torture select whether, how, and when to torture each specific detainee. Indeed, they control their detainees’ fates in a deeply intimate way, determining when each individual wakes, when he washes, whether and how much he eats, whether he sits, whether he has the means to read, write, or paint, whether he goes outside, and with whom, if anyone, he interacts.
In contrast, starvation of civilians as a method of warfare is ordinarily inflicted on populations. The besieging commander does not exercise over the besieged population anything close to the comprehensive and forensic level of control exercised by the torturer over the detainee. On the contrary, many aspects of both civilian life in general and the specific lives of individual civilians remain beyond her reach. Even on the limited dimension of controlling the supply of essentials to the encircled community, the commander does not control distributive allocation, sharing, sacrifice, coping strategies, or the arrangement of family and social life around food and other necessities. These are not contingent facts; the putative military necessity of encirclement deprivation is premised on the impossibility of limiting its impact to specific persons (or categories of person) within the besieged population.
This distinction regarding the extent of control and the degree to which it is individualized bears on normatively relevant features of the relationship between perpetrator and victim in the two scenarios. For Sussman, the torturer’s control is essential not just in establishing the victim’s defenselessness, but also in enabling the torturer to create a “protracted process in which pain is both inflicted and withheld (perhaps even assuaged) in a capricious and unpredictable way.” This creates the dynamic whereby the “victim is led to hope (however falsely or unreasonably) that there might be something he could do to appease or mollify his tormentor.” This, in turn, plays an important role in giving the victim “enough freedom and rationality to think of himself as accountable, while he nevertheless finds himself, despite all he can do, to be expressing the will of . . . [an] enemy.”
The relationship between the besieging commander and besieged civilians lacks this intimacy. Operating on the population as a whole and accumulating slowly, the aggregate effects of encirclement deprivation cannot be turned on and off at will. There may be times during a prolonged siege when the besieging party allows humanitarian supplies through, whether to divert scrutiny or otherwise, but this mechanism is too blunt to carefully calibrate and control the swings between suffering and relief. Once supplies go in, there is little external influence over how they are allocated, used, or preserved. Amidst the combination of initiating and intervening causes and the complexity of the relationship between individual and collective suffering within the besieged population, civilians in that community are unlikely to experience an identifiable “will lurking behind [their] suffering” to which they may develop an allegiance in the form of their bodies’ demands for relief.
And yet, despite these meaningful distinctions, there are also deep connections between the use of starvation as a method of warfare and torture. Although not comprehensively defenseless against that attack in the manner of the torture victim, the civilian population is defenseless in the thinner sense that its members cannot fight back on their own behalf without losing their protection against becoming lawful targets. Indeed, on some accounts, the defenselessness of civilians is normatively central to why they are not legitimate targets in war. It is reflected in the legal protection accorded them until such time as they take on the means to defend themselves and others against an enemy force by joining the armed forces or participating directly in hostilities.
For those who decline to participate directly in hostilities, neither the fact that they are protected in principle by the combatants charged with defending their society, nor the fact that they could (unlike detainees) choose to pick up arms and fight for themselves is accepted as a reason for their own legal protection to be diminished. Shue’s focus on detainee defenselessness as a normatively significant feature of torture reflects this—the contrast he draws is not between a detainee and an undesignated person, but between a detainee and an enemy fighter. There is a meaningful sense in which attacks on civilians, like attacks on detainees, are attacks on the defenseless.
Moreover, although lacking the intricate and total control of the torturer, the premise of a viable siege is that the besieging commander exercises control over the encircled society’s overall access to essentials. It is in the exercise of that blunter control that the besieging commander in a starvation siege inflicts a wrong that is both analogous to and distinct from the wrong of torture. Like torture, starvation methods distort biological imperatives against the self by crowding out meaningful access to higher-order desires and commitments. Unlike torture, that distortion occurs at the macro level, functions in at least two quite different ways, and is not always purposeful. In these respects, mass deprivation might be characterized as a societal analogue of torture.
The first dimension of the distortion of mass starvation is fundamentally political and more closely resembles the purposeful weaponization of the body in torture. It is on this dimension that a besieging force deprives the population of essentials with the purpose of breaking its will to resist. The second dimension is social. It manifests in the foreseeable effect of starvation conditions in undermining our capacities for the kind of commitments and bonds that underpin community, friendship, and love. It is largely independent from the goals of the besieging party.
Starvation crimes are pursued for diverse purposes, including control, collective punishment, ethnic cleansing, theft, and extermination. However, paralleling the legal and theoretical focus on interrogational (rather than sadistic) torture, the analysis here focuses specifically on the use of starvation to compel the adversary to change its behavior. Like interrogational torture, this mode of starvation is both the form most likely to be defended, in all of the ways already discussed, and the quintessential form of the crime. Moreover, the wrongfulness of other forms of starvation in war is captured straightforwardly by other criminal categories, such as murder, forcible transfer, extermination, or genocide.
Siege deprivation implicates the decision‑making of the civilian population, the military forces ensconced within, and their leadership. In the most straightforward case, these actors are unified in their posture. As Walzer writes, “[p]olitical integration and civic discipline make for cities whose inhabitants expect to be defended and are prepared, morally if not always materially, to endure the burdens of a siege.” Faced with a population unified in this way, the besieging party seeks to compel a shift from unified resistance to capitulation.
Our commitments not to abandon our homes, our community, and/or our system of political authority run deep. Like the detainee in Shue’s torture case, some significant proportion of besieged civilians may hold higher-order desires and commitments to resist the siege irrespective of the costs. Indeed, something like this is the presumption underpinning Walzer’s ideal type. The inhabitants in the scenario he imagines are sufficiently invested in one or the other of these higher‑order commitments to want to resist despite the cost threatened by the besieging party. That is what it means to be “prepared[ ] morally” to endure the promised deprivation.
The goal of the besieging party in such a scenario is to shift that posture through depriving the encircled population of indispensable objects and thus escalating the burdens of the siege. It may be that the besieging commander regrets the degree to which that entails inflicting starvation conditions on civilians, rather than on military forces or the political apex of the command structure. However, in a context in which the population is integrated and the deprivation is inflicted on the population as a whole, the commander’s regret does not preclude purpose with respect to civilian deprivation. As such, even when the ultimate goal is to starve out the military forces, a question arises with respect to how to understand the prior deprivation of the population, particularly given that it is well understood that military forces will not be the first to starve.
Short of literally starving the besieged population in order to get at the military forces ensconced within (itself a form of extermination as a crime against humanity), the alternative intentions (more in line with efforts to justify encirclement starvation) would involve compelling a change in the civilian population’s posture in order to change the posture of the military adversary. In a population that is broadly unified behind its political and military leaders, that would entail shifting the population’s posture away from affirmation of resistance to the siege and towards capitulation. That shift could manifest in the limited form of civilians’ abandonment of their homes and political community, or in the more comprehensive form of their political agitation in favor of surrender.
Whichever of these ends it serves, encirclement starvation is not simply coercive. The method does not function solely by raising the cost of resistance to the point at which the besieged population calculates that the benefits of resistance are no longer worth it. Rather, siege warfare aims to break the higher-order political commitment to resist through weaponizing the biological imperatives of the population against their deeply held commitments and thus to compel submission. It works, to the extent it does, by replicating what Sussman describes of torture: the suffering of hunger and the need to satisfy it is used to “crowd[ ] out every other aspect” of the mental lives of those affected, precluding a “purely contemplative attitude toward” the question of whether that suffering is a cost they are willing to bear in order to realize their higher-order political commitments.
This weaponization of biology against will is not unique to the Walzerian ideal of a politically unified besieged community. Even when the besieged population and the besieged force are at odds on whether to resist or surrender, the decision of individuals within that community not to attempt to overthrow or flee the besieged force still reflects a strategic decision on the part of those civilians. It is, in other words, a decision made in response to reasons. Those reasons may be pragmatic, rather than rooted in deep political ideals, and may themselves be responsive to a significant wrong (in the form of subjugation within the besieged community). However, as commitments made in response to reasons, they reflect the wills of those who make them. To the extent a starvation siege would “crowd out” the capacity to make even such pragmatic decisions in a way that is responsive to persons’ ordering of desires and commitments, it too would turn the victims’ bodies against their wills.
To recognize this reality is not to mitigate the wrongfulness of a besieged party subjugating the civilians under its control or denying them egress. As discussed above, the wrongdoing in such scenarios is not exhausted by one party. The fact that civilians were coerced into remaining does not eliminate the distorting wrong of using mass deprivation to compel their desperate exit or their revolt against that coercive authority.
As in situations of detainee torture, many of those subjected to starvation conditions in a siege may find it within themselves to resist until death or until the purpose of the perpetrator is frustrated. Beehner, Berti, and Jackson describe besieged civilians “weather[ing] severe hardships almost indefinitely.” However, even when those targeted resist successfully, they are still likely to experience the war between body and will. As does Sussman’s torture victim, they may “discover within [themselves] a host of traitorous temptations.” The perpetrator’s wrong inheres in weaponizing the victims’ bodies in this way, regardless of whether their will to resist is ultimately broken. The wrong is not contingent on the method’s success.
Understood in this way, starvation is distinctive from even a credible threat of instantaneous death through kinetic attack. Until delivered, the latter exists primarily at the conceptual level—where it remains subject to evaluation with a “contemplative attitude,” and, ultimately, to control by the will. In contrast, starvation tactics work slowly on the body and the mind, eroding the capacity to prioritize higher-order desires, no matter how much the individual wants to at the conceptual level. In other words, it is precisely because of its slowness relative to other methods of killing and harming in war that starvation entails a distinct and torturous wrong.
The second dimension of the twisting of body against self in mass starvation is both closely related to the first and yet distinct in certain key respects. In crowding out persons’ responsiveness to higher-order reasons, starvation tears not just at the capability of victims to uphold political commitments, but also at the human capacity for friendship and love. It “constrict[s] the reach of our social concerns and solidarities,”  limiting the “ambit of [victims’ social] reciprocity” within their communities. Although the erosion of social bonds through the effects of mass deprivation is likely to weaken political unity within the besieged population and thus contribute to its capitulation, this is unlikely to be experienced as part of a capitulation to the adversary in the manner of departure or surrender. Similarly, the besieging party is less likely to have sought to weaponize this form of distortion. Nonetheless, it is the inevitable upshot of prolonged starvation methods. Moreover, although there may be no specific experience of alliance with the enemy, the war between body and self on the social dimension may be even more wrenching due to the central place of social attachment in human life.
At the crux of the rupture of both social and political capabilities is the demand made by one’s body in a state of severe hunger to take “drastic efforts to preserve one’s own being.” Recounting the impact of that demand, Cormac Ó Gráda writes that situations of mass starvation:
invariably entail much antisocial behavior, as the bonds of family and neighborhood break down. Famine victims become desperate and self‑absorbed, and lack shame, their baser instincts prompting actions that would be unthinkable in normal times. Famines erode hospitality, solidarity, and community, and examples abound of appalling inhumanity and heartlessness among victims.
Observing such collapse, Pitrim Sorokin concludes, “to hunger, nothing is sacred.” These effects are systemic—families in contexts of siege deprivation are “forced to make impossible choices with the little food and water available.” 
This can often take on a specifically gendered dimension. Alex de Waal recounts that the “central women’s experience” of the famine in what is now Malawi (and was then Nyasaland) in the 1940s “was abandonment by their menfolk, leaving them to care for children and the uncertain mercies of government feeding schemes.” Reports from recent situations of mass starvation include stories of similar breakdowns. 
De Waal cautions against the temptation “to suppose a singularity here: that human beings are reduced to animals, and that extreme hunger displaces all other emotion or thought.” Even in conditions of the most extreme deprivation people make sacrifices, express love, and cooperate, often in heroic ways. Despite detailing examples of the kind of social and familial disintegration described above, a 2014 report by the Commission of Inquiry on Human Rights in the Democratic People’s Republic of Korea also notes that mothers “experienced severe deterioration in their health, largely because they either skipped or reduced portions of their meals for the benefit of other family members.” There are likely multiple factors at play in such results, including the displacing psychological pain of witnessing the suffering of one’s own children. Nonetheless, it is clear that even in the most severe famine conditions, there is “never a point at which humanity disappears.”
This observation, however, serves only to emphasize the torturous parallel. The wrongful distortion at the heart of torture depends on the detainee retaining enough of her agency to maintain a sense of the commitments that her body demands that she violate. Similarly, in the context of mass starvation, it is because fundamental human commitments have an enduring reality that the countervailing pain and suffering inflicted on the mind and body of the victims puts them at war with themselves. That distortion occurs even for those able, at some level, to resist the demands of their bodies and act on their higher-order commitments.
This distortion within persons combines into an aggregate distortion of the affected society against itself. Notwithstanding his caution regarding overstating the dehumanizing effects of starvation, de Waal emphasizes that famine “gnaws away at [communal, familial, and intrapersonal] bonds of affection, respect, and trust,” and argues that as a measure of the suffering inflicted in famine, “the severing of community relations may prove to be even more significant in the longer term” than is the risk of death. He affirms what social anthropologists have termed the “accordion effect” of famine, whereby each individual’s “web of obligations” contracts and moral restraints degenerate even as “recognizably social” functioning endures in “much reduced” form. 
The analysis above suggests a situation of significant moral tension and complexity. Writing of the distinctive deprivations of Nazi death camps, Primo Levi argues that kapos (prisoner functionaries within the camps) may be “the rightful owners of a quota of guilt (which grows apace with their freedom of choice),” but that:
in the vast majority of cases, their behavior was rigidly preordained. In the space of a few weeks or months the deprivations to which they were subjected led them to a condition of pure survival, a daily struggle against hunger, cold, fatigue, and blows in which the room for choices (especially moral choices) was reduced to zero.
In famine, Breandán Mac Suibhne reasons, this context of moral ambivalence is such that judgment is, “at best, not easy” and “sometimes, but not always . . . impossible.” Similar effects may be expected in a besieged community.
The moral complexity of human relations within a situation of mass deprivation arises from the fact that those who violate their own higher‑order commitments do so as victims of the besieging party’s wrongful infliction of the conditions in which they act. More specifically, they are victims of the torturous wrong of having their biological imperatives turned against their fundamental commitments to their neighbors, friends, and loved ones. In other words, they suffer the wrong of being driven to act against their own values through the systematic crowding out of the space for responding to moral reasons.
The normative structure of this dynamic is illuminated in part by considering the situation not simply in terms of wrongfulness and culpability, but in terms of the first- and second-personal moral dimensions—burden and blame. When persons within the besieged society are driven to violate key commitments by the crowding effects of extreme biological imperative, those who suffer the immediate effects of that violation may appropriately feel the kind of moral rupture that underpins blame. This follows from having suffered a violation of the shared commitments that underpin their bond with those who engaged in the violation, notwithstanding the extraordinary conditions within which the breach occurred.
However, it does not follow that it would be appropriate for external third parties to adopt a similar posture of blame regarding that relational deterioration. On the contrary, removed from the distorting effects of starvation, external third parties lack the moral standing necessary to appropriately blame or condemn those whose will is broken in such circumstances. To blame without oneself suffering either the violation in question or the crowding effect of the deprivation of essentials would express an unreflective self-righteousness.
Those involved in perpetrating the siege are also differently situated on the second-personal dimension of blame. Not only do they lack standing to blame those who may have violated deeply held commitments under starvation conditions; they (the perpetrators of the siege) are themselves the appropriate recipients of condemnation from all affected by the siege, including those persons who were unable to uphold their higher-order commitments in that context. Indeed, one of the key wrongs for which such condemnation may be issued is the wrong of distorting the biological needs of the besieged against their capacities to uphold higher-order commitments.
And yet, having suffered that torturous wrong, those who acted in those conditions of crowding may nonetheless be burdened by their actions in that context. Here, the question is not who may condemn, but whether the individual who acted “can live comfortably with herself,” recognizing the acts in question to be hers. The notion of burden reflects individuals’ ownership of their conduct and the endurance of their higher‑order commitments, even as they violate those commitments in the cognitively crowded situation of starvation. Individuals in that situation may feel both a sense of moral outrage with respect to those who inflicted the starvation conditions and a sense of self‑betrayal or self‑estrangement for having been unable to resist the force of those conditions. The cruel unfairness of that burden is itself a central aspect of the wrongfulness of the deprivation.
In addition to the struggle associated with having breached one’s own commitments, the conditions of starvation may also elicit the unwarranted feeling of failure (or the converse feeling of having been failed) that can come from being unable to provide for one’s dependents. Here, the fact of deprivation attacks the viability of social bonds and turns individuals and families against themselves. This occurs not necessarily through crowding the individual’s responsiveness to higher-order commitments or reasons (although the “traitorous temptations” identified by Sussman in the torture context may exacerbate the tendency towards self-blame, or even other-blame), but through overwhelming the capacity to relate to oneself fairly.
In short, mass starvation in war inflicts at least two distinctive forms of torturous wrong on the affected population. First, it places the bodies and biological imperatives of the targeted population at war with their political commitments, seeking to break the will of those with the capacity to make political choices—whether those choices involve staying or fleeing, supporting the besieged regime or resisting it, or advocating or opposing capitulation. Second, it places the bodies and biological imperatives of the targeted population at war with their social and familial commitments, leading predictably to a narrowing and breaking down of at least some of those commitments. Even when commitments are upheld, they are upheld over the biological appeal of self-betrayal. In addition, the suffering of dependents can itself provoke feelings of failure that overwhelm the capacity to relate fairly to oneself and to others.
The political dimension is more central to the reason for imposing a starvation siege. However, the second form of torturous wrong is important in part because the social dimension is not contingent on starvation being imposed as a mechanism by which to break the will of the besieged population. Social distortion arises from the conflict between biological imperatives and social and family commitments that have little to do with the besieging party.
The social dimension of distortion is important also because it is more distinctive to starvation, as compared to detainee torture. Torture may be used to override familial and social commitments (in addition to, or including, political commitments). However, the unique rupture in a situation of mass starvation is that the betrayal (or even potential betrayal) of one’s higher-order desire to cooperate with, prioritize, and love others is not abstracted from quotidian participation in those relationships. In contrast to the torture victim, who is ordinarily isolated from those who may depend on her resistance—abstracted by the distance of detention—the victim of mass starvation continues to participate actively in the relationships against which his body tears at the very moment of that distortion. Moreover, that distortion is likely to be experienced simultaneously by all parties to the relationships affected.
The capacity to form and act upon higher-order desires and the specific manifestation of that capacity in love, social connection, and political cooperation are essential aspects of our humanity. Just as torture is a mechanism by which the victim is turned against herself, starvation is a mechanism by which society is turned against itself, with individuals suffering a direct assault on their fundamental human capacities. That assault is a wrong that cannot be captured solely by the harm, suffering, or even death in which it may culminate, even as those aspects are, of course, key elements of what makes the practice condemnable.
Recognizing the torturous aspect of starvation of civilians as a method of warfare has several implications for how to understand the war crime within international criminal law. It suggests the possibility of a category of war crimes focused on the kind of distortion that links torture and mass starvation. It raises questions as to how to think about the other wrongful facets of starvation crimes. It is of potential interpretive significance. It could motivate the development of a parallel crime against humanity. And it hints at how prosecutors ought to think about investigations and charges in contexts of mass starvation.
Identifying the normative connections between torture and starvation as a method of warfare suggests a genus of war crimes defined loosely by the distortion of bodies or biological imperatives against selves. Given the way in which distinctions within the category map onto other ways of classifying war crimes, the category may be best understood as cutting across several of the types of war crimes discussed above, rather than as offering an alternative to them. Thus, torture might be conceptualized most effectively as both a crime of abuse against persons under the total control of the perpetrator and a crime that distorts the victim’s biological imperatives against herself. Similarly, starvation might be best understood as both an attack on civilians and a crime that distorts the biological imperatives of the starved against their capacities for higher-order commitments. Although a full exploration of the range of crimes that might join torture and starvation in this cross-cutting category is beyond the scope of this discussion, there are several that are notable immediately for their analogous features.
In the realm of customary international law, the war crime of intentionally inflicting terror on a civilian population shares notable features. Part of the function of inflicting terror is to create the kind of accumulating and overwhelming fear or panic that can crowd out the capacity of affected persons to consider their higher-order commitments with a contemplative attitude. The infliction of terror might, in other words, be understood to serve not only to coerce, but to “break the will”  of the adversary population through the steady drumbeat of attacks over time.  That, at least, is plausibly central to what distinguishes such methods from other forms of indiscriminate or targeted attacks on civilians. In contrast to starvation, where the mechanism of distortion is physical, the weaponization of victims against themselves in the war crime of inflicting terror operates primarily on the psychological dimension. Additionally, the latter crime is unlikely to cause the social distortion that is central to the second dimension of the starvation wrong. Nonetheless, at least in scenarios (such as the siege of Sarajevo) in which a system of terror is maintained over time to manipulate the psyche of those affected, there are significant parallels to the torturous aspects of the starvation crime. 
Other crimes cause distortion in a different way, forcing or coercing the body into roles that contradict core commitments or principles of autonomy, rather than using the body to crowd out responsiveness to those commitments. The consequence can be an enduring internal dissonance and trauma. Rape, sexual slavery, enforced prostitution, forced pregnancy, enforced sterilization, other forms of sexual violence, and the crime against humanity of forced marriage all include a form of weaponization of body against self, alongside other wrongs associated with the gendered nature of such crimes and the discrimination or subjugation attached to them.
Notably in this respect, rape has been equated explicitly with starvation in genocide jurisprudence.  It has also been recognized as a form of torture. At the philosophical level, Sussman frames his account of torture as “more akin to rape than other kinds of violence characteristic of warfare or police action.” This is not to say that rape or other crimes of sexual violence are reducible to torture. As noted above, the differentiation of offenses within the category of sexual and gender-based violence crimes is itself explicable in significant part with reference to the value of precision in the moral expression of criminal law. However, the overlap spotlights the importance of the torturous wrong across several war crimes.
A plausible case could also be made for thinking about several additional crimes in this category of war crimes. For example, the war crimes of enlisting, conscripting, or using child soldiers, compelling a protected person to serve in the forces of a hostile power, and possibly taking hostages might all be thought to manifest aspects of the distortion of bodies against selves. Here, too, there is no manipulation of biological imperatives to crowd out the victim’s responsiveness to higher-order commitments. Rather, the victims in such scenarios are forced or coerced into roles that place their bodies into functions (whether a combat function, or that of a bargaining chip) that run contrary to their presumed political commitments or to the moral boundaries of childhood.
B. The Multifaceted Nature of the Starvation Wrong and the Interpretive Significance of the Torturous Element
To identify this common thread is not to assert its normative exclusiveness. Starvation entails the infliction of severe physical suffering, disease, stunting, transgenerational harm, mass displacement, and death on those not liable to be targeted (or indiscriminately afflicted) with such harms, often affecting the most marginalized members of society first and most severely.  These are important reasons for condemning the method and holding perpetrators accountable. The same can be said of the severe pain and suffering inflicted during torture. The way in which starvation methods put victims in a posture of internal conflict may be distinctively torturous among attacks on civilian populations, but the wrongful infliction of severe pain and suffering is itself sufficient for criminality and the condemnation attached thereto.
Nonetheless, there are good reasons not to lose sight of the distorting aspects of starvation in making sense of the crime. Although the normative argument is not always transparent, those who defend certain forms of encirclement starvation appear to assume civilian death and suffering to be the key elements of the pro tanto wrong and then to discount the degree to which those wrongful harms inhere in starvation tactics on two grounds. First, those harms are not the goal of starvation campaigns that are targeted ultimately at the combatants ensconced within the affected population and that aim for a surrender that would obviate the escalation towards extreme suffering or death. Second, due to their incremental escalation, the full scope of those harms is avoidable by the targeted society through capitulation of one form or another.
On the account offered here, the slowness of starvation, as opposed to other forms of military attack, is not a virtue, or a reason to appraise it as less harmful than those alternatives; it is instead what underpins the infliction of a torturous wrong that operates in intimate connection with the suffering described above, but that is not captured by it. It is through the cumulative denial of biological imperatives that those imperatives are weaponized to crowd out higher‑order commitments and reasons. The system of breaking the will of the besieged people and tearing at their social fabric is possible only because the deprivation endures over time.
Torture is prohibited categorically in the context of detainee treatment. Recognizing an analogous wrong to be a central and distinctive feature of the war crime of starvation can help to account for why that method of warfare may appropriately be the subject of an equally categorical ban—a ban unaffected by the possible military advantage of its use in a particular context and perhaps more (rather than less) robust than that applicable to kinetic attacks. It offers an explanation for the “growing perception that, of all the ways in which death can result from an armed conflict, starvation is one which humanely we cannot accept.”
Emphasizing the torturous wrong of the starvation war crime also clarifies its role vis-à-vis the expressive function of international criminal law. That function is dependent on a criminal code that captures the distinctiveness of the wrongs condemned and that classifies wrongful conduct into different criminal categories according to morally meaningful distinctions. Precision in that respect is important if the system of international criminal justice is to articulate the object of the global community’s non-acquiescence and to stand in solidarity with victims and survivors in a way that affirms the scope of the wrong suffered.
In many instances, those responsible for starvation campaigns could, in principle, be prosecuted for a range of international crimes. However, in omitting the torturous feature of starvation, alternative categories would be expressively incomplete, even when evidentiarily straightforward. To prosecute a starvation campaign only as extermination, for example, would miss an important dimension of the wrong.
Of course, just as the criminalization of starvation captures dimensions of wrongfulness omitted by other categories, the converse is also true. Particularly since it can attach without any killing at all, the starvation war crime is not itself expressive of the wrong of mass killing. As such, relying on starvation charges alone might obscure an important feature of the wrong associated with a massively lethal starvation campaign. A full condemnation of the wrong in such a scenario may require cumulative starvation and extermination convictions for the same acts of deprivation. When such a campaign is inflicted with genocidal intent, a conviction for genocide may also be appropriate. In cases in which crimes of sexual or gender-based violence are part of the mechanism of starvation (used, for example, to prevent women and girls from venturing to sources of food or water) those, too, warrant separate and distinct recognition.
Doctrinally, the basis for cumulative convictions in such situations is that the crimes have mutually distinct elements. The fact of that mutual distinction is not necessary for, and need not underpin, heightened sentencing. International criminal courts and tribunals issue sentences based on an overarching assessment of conduct and culpability, including aggravating and mitigating factors not captured by the elements of the criminal convictions.  In exercising their broad discretion in that determination, they have eschewed issuing consecutive sentences for cumulative convictions attached to the same underlying conduct,  seemingly in part due to concerns about double-counting. Seen in that light, the value of cumulative convictions inheres not in the deterrent or retributive functions of criminal law, but instead in its expressive function.  The objective is a verdict that captures each facet of the wrongs done.
Expressive precision can also help to make normative sense of the absence of a human consequence element in the starvation crime. As noted above, that lack of a consequence element can be explained in part with reference to the evidentiary difficulty of establishing causation in a context defined by stretched temporality and a complex multiplicity of overlapping causal factors. However, the denial of objects indispensable to civilian survival is also the first step in the distortion of bodies against selves. As in the case of rape as a form of torture, the wrongful harm arising from that distortion can be presumed given the action and context, without needing a further showing of specific indicia of suffering.
Recognizing the torturous features of starvation that warrant distinct expression among war crimes spotlights the lack of an equivalent crime against humanity. That is of note both in addressing the infliction of starvation outside situations of armed conflict and in scenarios in which starvation occurs in a conflict zone but without the nexus necessary for it to qualify as a war crime. Crises relating to the supply of essentials in Venezuela and North Korea exemplify the phenomenon of contemporary mass starvation outside of armed conflict. Meanwhile, a recent report of the Independent Fact-Finding Mission on Myanmar, found that “in the majority of cases” government restrictions on movement and “deprivation of food and denials of humanitarian relief” were “not directly connected” with the armed conflict taking place in Myanmar at the time.
The most viable path to recognizing the unique harms associated with starvation in crimes against humanity law would be via the residual category of other inhumane acts. To qualify for inclusion in this category, an act must be of a “similar character” to other crimes against humanity and involve “intentionally causing great suffering, or serious injury to body or to mental or physical health.” This consequence threshold, although more legally demanding than the starvation war crime, does not pose the same evidentiary challenges with respect to establishing causation as is posed by extermination. Indeed, it may be that serious injury to mental or physical health could be presumed in cases of the severe deprivation of essentials.
There is already some developing jurisprudence of relevance in this respect. The U.N. Commission of Inquiry on Human Rights in North Korea has reasoned that that “knowingly causing prolonged starvation” is a crime against humanity under the “other inhumane acts” category.  Additionally, the Iraqi High Tribunal recognized the 1982 razing of the orchards around Dujail as qualifying under the “other inhumane acts” category of its crimes against humanity provision. Affirming and developing the notion of starvation as an inhumane act within the residual category would allow for the development of a specific set of elements and a direct recognition of that wrong. As discussed above, precisely this kind of development has occurred on the issue of forced marriage.
The return of the use of mass starvation in war has reignited discussions regarding the ways in which the law might respond to such deprivation. Particularly since the amendment of the Rome Statute to include the starvation war crime for NIACs, the criminality of starvation has taken on an unprecedented political, legal, and academic prominence.
And yet, confusion endures regarding both the scope of the prohibition and what the crime is fundamentally about. A doctrinal case can be made for a broad and categorical understanding of the starvation war crime. However, this Article has focused instead on the underlying normative foundations of the crime.
For many, the pro tanto wrongfulness of starvation of civilians as a method of warfare inheres in its status as a form of attack on civilians. That itself could plausibly provide the basis for a comprehensive and categorical prohibition. However, advocates of a narrow prohibition argue that the harm and death inflicted in starvation contexts is mitigated by its graduality. On this view, it is easier to justify subjecting an encircled area to starvation conditions than to indiscriminate bombardment, because the slowness of the former harm allows for its avoidance, most obviously through civilian exit or the capitulation of the ensconced adversary.
That standard account fails to understand the role of graduality in the starvation crime (and particularly in its instantiation in situations of siege deprivation). The harm and death inflicted on civilians in a starvation siege are clearly an important part of the wrong. However, starvation is more complicated in its moral content. Reducing it simply to pain, suffering, and death obscures a sense in which it inflicts a qualitatively different kind of wrong on affected persons—a wrong much closer in kind to that which lies at the heart of torture. Clarifying that distinctive feature of starvation can illuminate some of the uncertainty surrounding the scope of the crime and clarify the function of the crime within the framework of war crimes law more broadly.
Beyond its intrinsic value and its interpretive implications, clarity on this point also has implications for how international criminal law is practiced. A number of commentators have called for the ICC Prosecutor to focus not on seeking those most responsible for the gravest wrongs globally, but instead on proceeding selectively with prosecutions with a view to drawing attention to under-recognized criminal categories as a way of affirming their status, bringing visibility to their criminality, and prompting judicial elaboration of their contours. In a related vein, Barrie Sander observes what he terms “strategic expressivism” in various stakeholders’ approaches to the institutions and processes of international criminal law. On this account, international criminal justice is a domain “in which different actors compete for the legitimation of their preferred messages and narratives.” In that field of political contestation, civil society actors seek to rely “on the language and institutions of international criminal justice to achieve specific expressive benefits for their particular communities of interest.” Whether it informs prosecutors or other actors engaged in the politics of international criminal law, clarifying the fundamental normative meaning of the starvation crime can shape how it is invoked and applied.
 Resolution on Amendments to Article 8 of the Rome Statute of the International Criminal Court, ICC-ASP/18/Res.5 (Dec. 6, 2019); Rome Statute of the International Criminal Court art. 8(2)(e)(xix), July 17, 1998, 2187 U.N.T.S. 90 (last amended Dec. 6, 2019, depository notification C.N.394.2020.TREATIES-XVIII.10.g) [hereinafter Rome Statute].
 It is just the first step because the amendment will only enter into force on a state-by-state basis, as States Parties ratify the amendment individually. Rome Statute, supra note 1, art. 121(5).
 See Rome Statute, supra note 1, art. 8(2)(b)(xxv). The amendment had been proposed by Switzerland. See Annex IV, in Int’l Criminal Ct. Assembly of States Parties, Report of the Working Group on Amendments, ICC-ASP/17/35 (December 5–12, 2018); Int’l Criminal Ct. Assembly of States Parties, Report of the Working Group on Amendments, ICC-ASP/18/32 (Dec. 3, 2019). On the pre-history to this codification, see Nicholas Mulder & Boyd van Dijk, Why Did Starvation Not Become the Paradigmatic War Crime?, in Contingency in International Law 370 (Kevin Jon Heller & Ingo Venzke eds., 2021). On the peculiar lack of clarity regarding the reasons for the initial restriction of the crime to IACs, see infra note 50.
 On the resurgence, see Alex de Waal, Mass Starvation: The History and Future of Famine ch. 10 (2018). On the history of starvation tactics in war, see, for example, id. ch. 4; Esbjörn Rosenblad, Starvation as a Method of Warfare: Conditions for Regulation by Convention, 7 Int'l Lawyer 252, 255–56 (1973); Charles A. Allen, Civilian Starvation and Relief during Armed Conflict: The Modern Humanitarian Law, 19 Ga J. Int’l & Comp. L. 1, 4–8, 31–32 (1989); James Kraska, Siege, in Max Planck Encyclopedia of Public International Law ¶¶ 3, 8 (2009); Wolff Heintschel von Heinegg, Blockade, in Max Planck Encyclopedia of Public International Law ¶¶ 6–22 (2015); Bridget Conley and Alex de Waal, The Purposes of Starvation, 17 J. Int'l Crim. Just. 699, 709–11 (2019).
 In 2017, the Security Council was briefed that more than 20 million people across conflicts in Somalia, Yemen, South Sudan, and Nigeria “face[d] starvation and famine.” Associated Press, World Faces Worst Humanitarian Crisis since 1945, says UN Official, Guardian (Mar. 10, 2017), https://perma.cc/LD78-3Q5D. Alex de Waal described O’Brien’s framing as “hyperbolic” but agreed that “2017 marks a critical turning point, a moment at which famine could return.” De Waal, supra note 4, at 154. Briefings in 2020 were similarly dire. Mark Lowcock, Under‑Secretary‑General for Humanitarian Affairs and Emergency Relief Coordinator, Remarks to the Security Council on “Protection of civilians in armed conflict: indispensable civilian objects” (Apr. 27, 2021), https://perma.cc/Y2XW-GQJD; Peter Maurer, President of the International Committee of the Red Cross, Remarks to the U.N. Security Council Open Debate on the Protection of Objects Indispensable to the survival of the civilian population (April 27, 2021), https://perma.cc/P2M5-HSRA.
 See De Waal, supra note 4, at 6–7. For earlier iterations of the claim that famine is caused by political choices, see Alex de Waal, Famine Crimes: Politics and the Disaster Relief Industry in Africa 7 (1997); Amartya Sen, Development as Freedom 175 (1999); David Marcus, Famine Crimes in International Law, 97 Am. J. Int'l L. 245, 245 (2003).
 See Hannah Summers, Yemen on Brink of 'World's Worst Famine In 100 Years' If War Continues, Guardian (Oct. 15, 2018), https://perma.cc/Z6NL-3UUR; Hum. Rts. Council, Rep. of the Detailed Findings of the Group of Eminent International and Regional Experts on Yemen, ¶¶ 768–71, U.N. Doc. A/HRC/42/CRP .1* (Sept. 3, 2019) [hereinafter, Group of Experts on Yemen, Detailed Findings (2019)]; Jane Ferguson, Is Intentional Starvation the Future of War?, New Yorker (July 11, 2018). In 2019, World Food Programme spokesperson Herve Verhoosel warned that approximately 10 million Yemenis were “one step away from famine.” 10 Million Yemenis ‘One Step Away from Famine’, UN Food Agency Calls for ‘Unhindered Access” to Frontline Regions, U.N. News (Mar. 26, 2019), https://perma.cc/9Q2Y-P7TH. Cuts to essential aid and the COVID-19 pandemic exacerbated the crisis, see generally Vivian Yee, Yemen Aid Falls Short, Threatening Food and Health Programs, N.Y. Times (June 2, 2020), https://perma.cc/HQX4-8CAM; Mark Lowcock, Under‑Secretary-General for Humanitarian Affairs and Emergency Relief Coordinator, Briefing to the Security Council on the humanitarian situation in Yemen (Apr. 15, 2021), https://perma.cc/F6KK-VN7S (describing the situation in Yemen as “the world’s largest humanitarian crisis” and one that was “getting worse”).
 See, e.g., Hum. Rts. Council, Detailed Findings of the Independent International Fact-Finding Mission on Myanmar, ¶¶ 10, 49–50, 411, U.N. Doc. A/HRC/42/CRP.5 (Sept. 16, 2019) [hereinafter, IIFFMM, Detailed Findings (2019)]; Hum. Rts. Council, Rep. of the Detailed Findings of the Independent International Fact-Finding Mission on Myanmar, ¶¶ 1367–69, U.N. Doc. A/HRC/39/CRP.2 (Sept 17, 2018) [hereinafter, IIFFMM, Detailed Findings (2018)]; Andray Abrahamian, The Tamadaw Returns to the ‘Four Cuts’ Doctrine, Interpreter (Sept. 4, 2017), https://perma.cc/9X2W-6Q7S.
 Comm’n. on Hum. Rts. South Sudan, “There Is Nothing Left For Us”: Starvation As a Method of Warfare in South Sudan, ¶ 4, U.N. Doc. A/HRC/45/CRP.3 (Oct. 5, 2020).
 Hum. Rts. Council, Rep. of the Independent International Commission of Inquiry on the Syrian Arab Republic ¶ 96, U.N. Doc. A/HRC/46/54 (Jan. 21, 2021) [hereinafter IICISyria Report 2021].
 See World Peace Found., Starving Tigray: How Armed Conflict and Mass Atrocities Have Destroyed an Ethiopian Region’s Economy and Food System and Are Threatening Famine 23–52 (2021), https://perma.cc/ENL6-HK5A; Tom Dannenbaum, Famine in Tigray, Humanitarian Access, and the War Crime of Starvation, Just Sec. (July 26, 2021), https://perma.cc/LZ76-XLAB.
 S.C. Res. 2417 (May 24, 2018).
 The IAC codification at the ICC, supra note 3, was the first codification of the starvation crime in an international instrument. Exemplifying the renewed attention, see generally Special Issue Starvation in International Law, 17 J. Int’l Crim. Just. 673–929 (2019); A Pandemic of Hunger: Implementing UN Security Council Resolution 2417, Glob. Rts. Compliance (May 2021), https://perma.cc/54CZ-7HEB.
 For a limited domestic exception to the lack of case law, see Public Prosecutor v. M.P. et al (K. 74/96), verdict, District Court in Zadar, Croatia (Apr. 24, 1997), unofficial translation of the verdict available at https://perma.cc/U4M9-HC54. The Swedish Prosecution Authority’s recently announced intention to prosecute Lundin Oil AB executives for complicity in alleged war crimes in Sudan between 1999 and 2003 includes mention of the “burning [of] . . . crops so that people did not have anything to live by” as a “grave war crime.” Press Release, Åklagarmyndigheten (Swedish Prosecution Authority), Prosecution for Complicity in Grave War Crimes in Sudan (Nov. 11, 2021), https://perma.cc/8ZQC-SM6T. However, it does not appear that this will be prosecuted as the specific crime of starvation of civilians as a method of warfare. For a comprehensive overview of jurisprudence relevant to starvation, see Starvation Jurisprudence Digest, Glob. Rts. Compliance, https://perma.cc/27RQ-JAWJ.
 See infra Section II.B.
 See infra Section III.A.
 See infra Section III. On the notion of a normative account of an international crime, see, for example, Tom Dannenbaum, Why Have We Criminalized Aggressive War?, 126 Yale L.J. 1242, 1249–54 (2017).
 On identifying various paths to criminal accountability for the infliction of mass starvation, see, for example, Federica D’Alessandra & Matthew Gillett, The War Crime of Starvation in Non‑International Armed Conflict, 17 J. Int’l Crim. Just. 815, 826 (2019); Rogier Bartels, Denying Humanitarian Access as an International Crime in Times of Non-International Armed Conflict, 48 Israel L. Rev. 281, 299–301 (2015).
 See infra Sections II.B, III.D, III.F.
 There is no legal definition of “siege” or “encirclement” (unlike blockade). On the legal elements of blockade, see San Remo Manual on International Law Applicable to Armed Conflicts at Sea ¶¶ 93–104 (Louise Doswald-Beck ed., 1995); Declaration Concerning the Laws of Naval War, Feb. 26, 1909, 208 Consol. T.S. 338, reprinted in The Declaration of London 112, arts. 1–21. (James B. Scott ed., 1919).
 Other modalities may be part of encirclement warfare but can also occur independently. They include attacks on humanitarian workers; the destruction, looting, or rendering useless of livestock, crops, farmland, fishing systems, water and irrigation infrastructure, markets, and humanitarian aid; the impeding of pastoralists’ rights of free movement; the disruption of coping strategies; and attacks on civilians seeking to access essentials. See Tom Dannenbaum, Starvation in an Age of Mass Deprivation in War, 54 Vand. J. Transnat’l L. (forthcoming 2022).
 Bridget Conley and Alex de Waal identify nine reasons that those with the power to do so inflict starvation on others: (i) extermination or genocide; (ii) control through weakening a population; (iii) gaining territorial control; (iv) flushing out a population; (v) punishment; (vi) material extraction or theft; (vii) extreme exploitation; (viii) war provisioning; and (ix) comprehensive societal transformation. Conley & de Waal, supra note 4.
 See supra note 4.
 See generally Comm’n. on Hum. Rts. South Sudan, supra note 9, ¶¶ 101, 105–06, 109, 124–27; IIFFMM, Detailed Findings (2019), supra note 8, ¶¶ 156–60, 168–69, 584–85 (impediments to transferring essentials or delivering humanitarian aid), 404–06 (import bans); IIFFMM, Detailed Findings (2018), supra note 8, ¶¶ 575–77 (denial of humanitarian access), 1165–68 (encirclement); Hum. Rts. Council, Rep. of the Int’l Comm’n of Inquiry on Libya, ¶¶ 21, 88, 90, 128, 160, 542, 563, U.N. Doc. A/HRC/19/68 (Mar. 2, 2012); IICISyria Report 2021, supra note 10, ¶¶ 39, 46; Hum. Rts. Council, Findings of the Group of Eminent International and Regional Experts on Yemen, U.N. Doc. A/HRC/39/43 (Aug. 17, 2018); Rep. of the Panel of Experts on the Sudan (2020), transmitted by Letter dated 14 January 2020 from the Panel of Experts. Established Pursuant to Resolution 1591 (2005) Concerning the Sudan Addressed to the President of the Security Council, ¶ 132, U.N. Doc. S/2020/36 (Jan. 14, 2020); Rep. of the Panel of Experts on Somalia (2020), transmitted by Letter dated 28 September 2020 from the Chairman of the Security Council Comm. Established Pursuant to Resolution 751 (1992) Concerning Somalia Addressed to the President of the Security Council, ¶ 124, U.N. Doc. S/2020/949 (Sept. 28, 2020); Martha Mundy, Report Strategies of the Coalition in the Yemen War 2–3, 7 (Oct. 9, 2018), https://perma.cc/V9DT-SUJL; Amnesty Int’l, We Leave or We Die: Forced Displacement under Syria’s Reconciliation Agreements 18 (2017) https://perma.cc/FK8L-GBDR; Prosecutor v. Dragomir Milošević, Case No. IT‑98-29/1-T, Judgment, ¶¶ 208, 937 (Dec. 12, 2007); Prosecutor v. Krstić, Case No. IT-98-33-T, Judgment, ¶¶ 26, 28, 566 (Aug. 2, 2001); Prosecutor v. Karadžić, Case No. MICT-13-55-A, Judgment, ¶¶ 565–73 (Mechanism for Int'l Crim. Tribs. Mar. 20, 2019).
 See infra Sections II.B–III.A.
 Amos C. Fox, The Reemergence of the Siege, 18 Inst. Land Warfare 2 (2018). Others suggest that siege warfare is poised to become “even more relevant” over time. Lionel M. Beehner, Benedetta Berti, & Michael T. Jackson, The Strategic Logic of Sieges in Counterinsurgencies, 47 Parameters 77, 78 (2017).
 See infra Section IV.
 See infra text accompanying notes 216–22.
 See infra Section IV.B.
 See infra Section VI.
 See infra Sections V–VI.
 See infra note 293 and accompanying text.
 Report of the Commission on Responsibility of the Authors of the War and on Enforcement of Penalties, Report Presented to the Preliminary Peace Conference (March 29, 1919), reprinted in 14 Am J. Int’l L. 95, 114 (1920).
 On the efforts to pursue criminal justice for alleged German perpetrators after World War I, see Matthias Neuner, When Justice Is Left to the Losers: The Leipzig War Crimes Trials, in 1 Historical Origins of International Criminal Law 333 (Morten Bergsmo, Cheah Wui Ling & Yi Ping, eds., 2014); Gary Jonathan Bass, Stay the Hand of Vengeance: The Politics of War Crimes Tribunals ch. 3 (2002).
 See, e.g., Judgment (Sept. 30, 1946), in 22 Trial of the Major War Criminals Before the International Military Tribunal, 411, 456, 472, 474, 478, 480, 482, 484, 495, 542–44, 541 (1948).
 United States v. Wilhelm von Leeb et al. (Oct. 27, 1948), in 11 Trials of War Criminals Before the Nuernberg Military Tribunals Under Control Council Law No. 10 462, 555, 563 (1949) (citing Charles Cheney Hyde, 3 International Law 1802–03 (2d ed. 1945)).
 See Mulder & van Dijk, supra note 3, at 377–80.
 Any power engaged in belligerent occupation of foreign territory has “the duty of ensuring the food and medical supplies of the population; it should, in particular, bring in the necessary foodstuffs, medical stores and other articles if the resources of the occupied territory are inadequate.” Convention (IV) Relative to the Protection of Civilian Persons in Time of War art. 55, Aug. 12, 1949, 6 U.S.T. 3516, 75 U.N.T.S. 287 [hereinafter GC IV]. It has a clear duty to consent to and facilitate relief schemes (particularly of food, medical supplies, and clothing) for that population if it is “inadequately supplied.” Id. art. 59.
 See Marcus, supra note 6, at 266 (critiquing the Fourth Convention framework for non-occupied territory); René Provost, Starvation as a Weapon, 30 Colum. J. Transnat’l L. 577, 592 (1992).
 GC IV, supra note 39, art. 23.
 Id. art. 17.
 None of even the limited requirements related to starvation and humanitarian access were included in the Fourth Convention’s grave breaches provision. Id. art. 147.
 Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts art. 54, June 8, 1977, 1125 U.N.T.S 3 [hereinafter AP I]; Protocol Additional to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of Non-International Armed Conflicts art. 14, June 8, 1977, 1125 U.N.T.S. 609 [hereinafter AP II].
 AP I, supra note 44, art. 54(2).
 See id. art. 54(3).
 Id. art. 70(1). See also AP II, supra note 44, art. 18.
 See AP I, supra note 44, art. 85.
 Rome Statute, supra note 1, art. 8(2)(b)(xxv).
 In the most compelling explanation, Rogier Bartels argues that the omission was likely “an oversight, perhaps caused by the unfortunate placing of the proposed crime together with various versions of disproportionate use of force,” which, unlike the starvation provision, lacked underlying codification in Protocol II. Bartels, supra note 18, at 298. Ultimately, however, the reasons remain hazy even for those who participated in the final Rome Conference. Panel on Starvation in Armed Conflicts (panelists: Federica D’Alessandra, Brian Lander, Matthias Lanz & Charles Garraway) Geneva Academy (May 2, 2019), https://perma.cc/RW6G-2T4Y (May 6, 2019) (Charles Garraway, who was at the Rome Conference, at 28:50: “my own belief is it fell just below the threshold, bearing in mind there were still states in Rome who were not prepared to accept any, any offenses [in NIACs] except perhaps common article 3”; Matthias Lanz, having explored the travaux in preparation for the 2018 Swiss proposal, at 36:18: “you find . . . no real information in the travaux préparatoires on why this happened, so I think . . . it was chaotic and there was a lot of politics involved is probably what comes closest to the truth. . . . We believe that [the NIAC draft provision] was [deleted from what became the final Statute as] part of a bigger package where tradeoffs were made.”); Michael Cottier & Emilia Richard, Article 8(2)(b)(xxv), in The Rome Statute of the International Criminal Court: A Commentary 508, 510 (Otto Triffterer & Kai Ambos eds., 3rd ed. 2016); William A. Schabas, An Introduction to the International Criminal Court 285 (5th ed. 2017).
 Rome Statute, supra note 1, art. 8(2)(e)(xix).
 Id. arts. 12(3), 13(b). Sudan’s recently announced cooperation with the ICC Prosecutor in the arrest and transfer of Omar al-Bashir emphasizes the degree to which such contingencies can change swiftly and unexpectedly. @IntlCrimCourt, Twitter (Aug. 16, 2021, 11:16 PM) https://perma.cc/8RMP-FGNH. As final editing on this Article was concluding, a coup in Sudan has introduced further flux and uncertainty into that situation. See, e.g., Rebecca Hamilton, After the Coup in Sudan: Key (Short-Term) Indicators for Democratic Survival, Just Sec. (Oct. 25, 2021), https://perma.cc/NJ6M-BLKX.
 See Rome Statute, supra note 1, arts. 12–13 (providing for the Court’s jurisdiction in situations not referred by the U.N. Security Council only when the crimes occur on the territory of a state party or are perpetrated by the national of a State Party or any state that accepts the Court’s jurisdiction on an ad hoc basis).
 See supra notes 4–11. None of Eritrea, Ethiopia, Myanmar, South Sudan, Syria, or Yemen is party to the Statute. Nor are the vast majority of states that have participated in the war in Yemen, including Saudi Arabia and the United Arab Emirates.
 Rome Statute, supra note 1, art. 121(5). For the current list of ratifying states, see Amendment to article 8 of the Rome Statute of the International Criminal Court (Intentionally using starvation of civilians), U. N. Treaty Collection, https://perma.cc/P7X5-5FQR. Criticizing this aspect of the amendment system, see Kevin Jon Heller, The Rome Statute’s Flawed Amendment Regime: Starvation in NIAC Edition, Opinio Juris (Dec. 7, 2019), https://perma.cc/PA5X-L53E.
 See, e.g., Displaced in Northeast Nigeria ‘Knocking on Door of Starvation’: WFP, U.N. News (Oct. 15, 2021), https://perma.cc/KY65-2MP3; Rep. of the Panel of Experts on Mali (2020), transmitted by Letter dated 7 August 2020 from the Panel of Experts. Established Pursuant to Resolution 2374 (2017) and Extended by Resolution 2484 (2019) Concerning Mali Addressed to the President of the Security Council, ¶¶ 129–35, U.N. Doc. S/2020/785/Rev.1 (Aug. 7, 2020).
 On the Court’s difficulties, see, for example, Independent Expert Review of the International Criminal Court and the Rome Statute System: Final Report, ICC-ASP/19/16 (Sept. 30, 2020); Douglas Guilfoyle, This is Not Fine: The International Criminal Court in Trouble, Parts I-III, EJIL:Talk! (Mar. 2019), https://perma.cc/7YZV-GNBN. On the enduring significance of the Statute, see, for example, Alex Whiting, Crime of Aggression Activated at the ICC: Does it Matter?, Just Sec. (Dec. 19, 2017), https://perma.cc/2GHC-U9BA.
 See generally National Implementing Legislation Database, Int’l Crim. Ct. Legal Tools Database, https://perma.cc/2Z5D-UAPS; ICRC, Practice Relating to Rule 53. Starvation as a Method of Warfare, https://perma.cc/SD6U-Y5C6.
 See, e.g., U.N. Transitional Administration in East Timor, Regulation No. 2000/15 on the Establishment of Panels with Exclusive Jurisdiction over Serious Criminal Offenses, art. 6(1)(b)(xxv), U.N. Doc. UNTAET/REG/2000/15 (June 6, 2000); The Statute of the Iraqi Special Tribunal of 2003, art. 13(b)(xxv); Law on Specialist Chambers and Specialist Prosecutor’s Office No. 05/L-053 of 2015, art. 14(1)(b)(xxv) (Kos.).
 Protocol on Amendments to the Protocol on the Statute of the African Court of Justice and Human Rights, Annex: Statute of the African Court of Justice and Human and Peoples’ Rights, arts. 28D(b)(xxvi), 28D(e)(xvi), June 27, 2014. The Malabo Protocol has yet to enter into force. Based on a search of the ICC legal tools database (at legal-tools.org), and the ICRC’s customary IHL database (ICRC, supra note 58) states codifying starvation as a war crime, regardless of conflict classification include: Belgium, Bosnia and Herzegovina, Croatia, Ethiopia, Norway, the Philippines, Rwanda, Spain. States that provide explicitly that starvation can be perpetrated in a NIAC include: Azerbaijan, Cambodia, Germany, Kosovo, the Netherlands, Peru, Portugal, Philippines, South Korea, Switzerland, Uruguay. See also D’Alessandra & Gillett, supra note 18, at 820 n.20 (also including Austria, North Macedonia, Romania, Serbia, Sweden).
 See Comm’n. on Hum. Rts. South Sudan, supra note 9, ¶ 148(e), recommending the inclusion of the war crime of starvation as a method of warfare in the Statute of the promised (but not yet created) Hybrid Court for South Sudan. On the status of the court, see Nyagoa Tut Pur, A Glimmer of Hope for South Sudan’s Victims, Hum. Rts. Watch (Jan. 31, 2021), https://perma.cc/P7XA-VDA7; Robbie Gramer, U.S. Quietly Gives Up on South Sudan War Crimes Court, Foreign Pol’y (July 20, 2021), https://perma.cc/2TS9-WRC9.
 See supra notes 1–11 and accompanying text.
 See, e.g., Comm’n. Hum. Rts. South Sudan, supra note 9, ¶¶ 35, 89; Hum. Rts. Council, Rep. of the Comm’n. on Human Rights in South Sudan, ¶ 71, U.N. Doc. A/HRC/43/56 (Jan. 31, 2020); id. Annex II, ¶ 44; World Peace Found., supra note 11, at 10 n.23.
 I discuss the criminal law debates in greater detail in Tom Dannenbaum, Criminalizing Starvation in an Age of Mass Deprivation in War, 54 Vand. J. Transnat’l L. (forthcoming 2022).
 Rome Statute, supra note 1, arts. 8(2)(b)(xxv), 8(2)(e)(xix).
 On reading purpose into “method” in the underlying IHL prohibition (replicated in this respect in the war crime), see Phillip J. Drew, Can We Starve the Civilians?, 95 Int’l L. Studies 302, 314 (2019); Sean Watts, Humanitarian Logic and the Law of Siege, 94 Int’l L. Studies 1, 18–19 (2019). See also U.S. Dep’t of Def., Law of War Manual §§ 5.20.1, 220.127.116.11 (rev. ed. Dec. 2016), https://perma.cc/AC36-LQCK [hereinafter U.S. DoD Law of War Manual]; Jean-Marie Henckaerts & Louise Doswald-Beck, 1 Customary International Humanitarian Law 186 (2005) (describing rule fifty-three). On reading purpose into the use of “intentionally” (or “intentionally” combined with “method”) in the war crime, see Randle C. DeFalco, Conceptualizing Famine as a Subject of International Criminal Justice, 38 U. Pa. J. Int’l L. 1113, 1145 (2017); Gerhard Werle & Florian Jeßberger, Principles of International Criminal Law 506 (3d ed. 2014); Daniel Frank, The Elements of War Crimes – Article 8(2)(b)(xxv), in The International Criminal Court, Elements of Crimes and Rules of Procedure and Evidence 203, 205 (Roy S. Lee ed., 2001).
 See IIFFMM, Detailed Findings (2019), supra note 8 at ¶ 420; Rep. of the Comm’n of Experts Established Pursuant to Resolution 780 (1992), transmitted by Letter dated 24 May 1994 from the Secretary General to the President of the Security Council, Annex VI.B: The Battle of Sarajevo and the Law of Armed Conflict, ¶ 76, U.N. Doc S/1994/674/Add.2 (vol. III) (Dec. 28, 1994) [hereinafter UN Commission Report Sarajevo]; Watts, supra note 66, at 19.
 Such tactics are sometimes referred to using the terms “kneel or starve,” “surrender or starve,” or “surrender-or-die.” Glob. Rts. Compliance & World Peace Found., The Crime of Starvation and Methods of Prosecution and Accountability ¶ 105 (June 13, 2019), https://perma.cc/V353-98MH; Amnesty Int’l, supra note 24, at 15; Brian Lander & Rebecca Vetharaniam Richards, Addressing Hunger and Starvation in Situations of Armed Conflict, 17 J. Int’l Crim Just. 675, 677 (2019).
 Relatedly, see U. K. Ministry of Defence, The Manual of the Law of Armed Conflict ¶¶ 5.27.1, 5.34.3 (2004), https://perma.cc/QS59-2DED [hereinafter U.K. LOAC Manual]; Kraska, supra note 4, ¶ 21; Yoram Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict 255–56 (3d ed. 2016); Provost, supra note 40, at 618; Matthew C. Waxman, Siegecraft and Surrender, 39 Va. J. Int’l. L. 353, 363 (1999).
 Even assuming proportionality does serve as a backstop for a narrow reading of the starvation ban in IHL, the war crime attaches only to “clearly” disproportionate attacks, and there is no NIAC war crime of disproportionate attacks in the Rome Statute. Rome Statute, supra note 1, art. 8(2)(b)(iv). Contesting the application of proportionality to siege warfare on the grounds that it applies only to “attacks,” which are defined in Article 49 of Protocol I as “acts of violence against the adversary, whether in offence or in defence,” (AP I, supra note 43, art. 49(1)), see Watts, supra note 66, at 19; Drew, supra note 66, at 319–20.
 See, e.g., Watts, supra note 66; Waxman, supra note 69; Drew, supra note 66.
 See Jacob Turkel et al., The Public Commission to Examine the Maritime Incident of 31 May 2010 ¶ 78 (2011), https://www.gov.il/BlobFolder/generalpage/downloads_eng1/en/ENG_turkel_eng_a.pdf (last visited Jan. 23, 2022) [hereinafter the Turkel Commission Report]; Drew, supra note 66, at 314.
 The Commentary to Protocol I provides only, “[t]he term ‘means of combat’ or ‘means of warfare’ . . . generally refers to the weapons being used, while the expression ‘methods of combat’ generally refers to the way in which such weapons are used.” Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949 ¶ 1957 (Yves Sandoz, Christophe Swinarski & Bruno Zimmermann eds., 1987). Although comprehensible in framing what differentiates means from methods, this is plainly not a viable definition of the term. The Protocol itself identifies methods of warfare that involve no direct use of weapons. Gloria Gaggioli & Nils Melzer, Methods of Warfare, in Oxford Guide to International Humanitarian Law 235 (Dapo Akande & Ben Saul eds., 2020).
 Wayne Jordash, Catriona Murdoch & Joe Holmes, Strategies for Prosecuting Mass Starvation, 17 J. Int’l Crim. Just. 849, 862 (2019); Glob. Rts. Compliance & World Peace Found., supra note 68, ¶ 78.
 D’Alessandra & Gillett, supra note 18, at 841; Jordash, Murdoch, & Holmes, supra note 74, at 854, 858–60. More equivocally, Cottier and Richard argue that the relevant language is designed to exclude “starvation as a result of unintended mismanagement,” Cottier & Richard, supra note 50, at 518, and suggest, in line with oblique intent, that “if the outcome of impeding humanitarian assistance is obvious according to the ordinary course of events, the intention can be inferred.” Id. at 519. See also the equivocation in Manuel J. Ventura, Prosecuting Starvation under International Criminal Law: Exploring the Legal Possibilities, 17 J. Int’l Crim. Just. 781, 785–88 (2019).
 Rome Statute, supra note 1, article 30(2)(b) [emphasis added].
 Prosecutor v. Lubanga, Case No. ICC-01/04-01/06-3121, Appeal Judgment, ¶¶ 447–50 (Dec. 1, 2014), https://perma.cc/22B2-XCZP.
 Prosecutor v. Bemba, Case No. ICC-01/05-01/13-2275-Red, Appeal Judgment, ¶ 677 (Mar. 8, 2018), https://perma.cc/8846-552S; Prosecutor v. Bemba, Case No. ICC-01/05-01/08-424, Decision on the Confirmation of Charges at ¶¶ 194–95 (June 15, 2009), https://perma.cc/8S27-9UER; Prosecutor v. Ongwen, Case No. ICC-02/04-01/15, Trial Judgment, ¶¶ 2706, 2865–74 (Feb. 4, 2021), https://perma.cc/58EV-2RW3. Werle & Jeßberger, supra note 66, at 352, 365, 387.
 Dannenbaum, supra note 64.
 See Int’l Crim. Ct., Elements of Crimes, Doc. No. ICC-PIDS-LT-03-002/11_Eng, at 31 (2011) [hereinafter ICC Elements of Crimes].
 See AP I, supra note 43, art. 54(2–3).
 Dannenbaum, supra note 64.
 See supra notes 220–21 and accompanying text.
 See supra note 219 and accompanying text.
 See Dannenbaum, supra note 64.
 See, e.g., Françoise Hampson, A Terminological Issue: What is a Siege?, 46 Collegium 91, 93 (2016) (“It is generally admitted that the ratio is between three and five attacking forces to one defending force in order to take a town by assault, whereas in the case of a siege, the ratio is said to be one to one.”). See also Beehner, Berti, & Jackson, supra note 26, at 81; Fox, supra note 26, at 4.
 Hum. Rts. Council, Rep. of the Independent Int’l Comm’n of Inquiry on the Syrian Arab Republic, U.N. Doc. A/HRC/34/64 ¶ 95 (Feb. 2, 2017) [hereinafter Independent Int’l Comm’n of Inquiry on Syria Rep.].
 Heintschel von Heinegg, supra note 4, ¶ 27.
 See Watts, supra note 66, at 48.
 See id.
 See id.; Sean Watts, Siege in Contemporary Doctrine: How Does it Work?, 46 Collegium 95, 96 (2016).
 There are rare containment encirclements where the blockading state has the capacity to enter and control the territory but chooses instead to control the perimeter. The Israeli blockade of Gaza arguably falls into this category. Yoram Dinstein, The International Law of Belligerent Occupation 276–85 (2009).
 See U.N. Comm’n. Rep. Sarajevo, supra note 67, ¶¶ 76, 87; Provost, supra note 40, at 617; Françoise Hampson, The Prohibition of Starvation: Additional Elements? 46 Collegium 99, 99 (2016); Watts, supra note 66, at 19.
 Watts, supra note 91, at 95.
 Michael Walzer, Just and Unjust Wars 162 (1977).
 Yoram Dinstein, Siege Warfare and the Starvation of Civilians, in Humanitarian Law of Armed Conflict: Challenges Ahead 145, 151 (A.M. Delissen et al. eds., 1991). See also Dinstein, supra note 69, at 253–57; Kraska, supra note 4, ¶¶ 10, 21; William J. Fenrick, Attacking the Enemy Civilian as a Punishable Offense, 7 Duke J. Comp. & Int’l L. 539, 557 (1997). Although Watts later builds on this to argue for a narrow interpretation of the law, some of his earlier work adopts a posture not unlike Dinstein’s. See Watts, supra note 91, at 98; Sean Watts, Some Comments About the United Nations Security Council Resolution 2139 (2014), 46 Collegium 104, 105 (2016).
 Memorandum from the Joint Chiefs of Staff, Review of the 1977 First Additional Protocol to the Geneva Conventions of 1949, to Secretary of Defense, at 72–73 (May 3, 1985), https://perma.cc/T2HL-PTS3 (Article 70 of Protocol I, which supplements Article 54 by emphasizing the need for humanitarian access in cases of inadequate supply, was deemed acceptable only subject to the caveat that humanitarian access could be denied on the grounds of “imperative considerations of military necessity,” to preclude a “radical, if perhaps unintended, change in the customary law of siege and blockade warfare, which has always allowed the besieging and blockading power to cut off all supplies going to areas of enemy control.”).
 Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts (Protocol I), Reservation / Declaration (U.K., ¶ (p)), https://perma.cc/67W4-QMVP; Id. (Fr., ¶ 17), https://perma.cc/QUD9-MH2C.
 Questioning whether “all of [the] particulars” of Article 54 are customary, even as it accepts the customary status of the core ban. U.S. DoD Law of War Manual, supra note 66, § 5.20.4.
 Watts, supra note 66, at 6.
 Id. at 46.
 Id. at 48.
 Waxman, supra note 69, at 421.
 Hampson, supra note 86, at 92.
 U.N. Comm’n Rep. Sarajevo, supra note 67, ¶ 76.
 Letter dated 24 May 1994 from the Secretary General to the President of the Security Council, ¶ 204 U.N. Doc S/1994/674 (May 27, 1994).
 KJ Riordan, Shelling, Sniping and Starvation, 41 Victoria Univ. Wellington L. Rev. 149, 178 (2010).
 Consider the Turkel Commission’s framing of an Israeli policy change in 2010. Turkel Commission Report, supra note 72, ¶ 97 (“[I]t should be noted that in June 2010, the Israeli Government changed the land border crossings policy from a policy in which only the transfer of limited humanitarian supplies was allowed to a policy where only the entry of goods that have military purposes is prohibited.”). On the legality of interdicting contraband in the absence of a viable comprehensive blockade, see Heintschel von Heinegg, supra note 4, ¶ 2; Daniel P. O’Connell, 2 The International Law of the Sea 1150 (1984).
 This was how the Israeli High Court of Justice evaluated the situation in Gaza in 2008. HCJ 9132/07 Al-Bassiouni Ahmed v. Prime Minister of Israel, ¶ 22 (2008) (Isr.), unofficial translation available at https://perma.cc/CBC6-7TJ9. See also Turkel Commission Report, supra note 72, ¶¶ 73, 76.
 See Emanuela-Chiara Gillard, Sieges, the Law, and Protecting Civilians 2 (2019); Beehner, Berti & Jackson, supra note 26, at 80–81.
 Sec. Council, The International Commission of Inquiry on the Central African Republic, Final Report, U.N. Doc. S/2014/928 (Dec. 22, 2014), ¶ 233. See also supra note 110.
 See Mark Lattimer, Can Incidental Starvation of Civilians be Lawful under IHL?, EJIL:Talk! (Mar. 26, 2019), https://perma.cc/25HL-JHRL; Provost, supra note 40, at 620; Allen, supra note 4, at 82.
 See Robert Pape, Dying to Win 27–37, 79–101 6 (2005). Efficacy here is highly contextual. See e.g., Tom Dannenbaum, Bombs, Ballots, and Coercion, 20 Sec. Stud. 303 (2011).
 See AP I, supra note 44, art. 51(2); Rome Statute, supra note 1, arts. 8(2)(b)(i), 8(2)(c)(i), 8(2)(e)(i); Prosecutor v. Galić, Case No. IT-98-29-A, Appeal Judgment, ¶¶ 129–30 (Nov. 30, 2006); Prosecutor v. Blaškić, Case No. IT-95-14-A, Judgment at ¶ 109 (Int’l Crim. Trib. for the Former Yugoslavia July 29, 2004).
 Cf. Walzer, supra note 95, ch. 16 (evaluating justifications offered for the “terror bombing” of German and Japanese cities, rejecting many, but providing narrow support for the use of the method in extreme emergency, such as the imminent risk of comprehensive and devastating Nazi victory).
 See 2 Independent International Fact-Finding Mission on the Conflict in Georgia 338–43 (2009), https://perma.cc/VTQ9-5MRL; Kraska, supra note 4, ¶ 13; Hum. Rts. Council, Rep. of the Independent Commission of Inquiry on the 2014 Gaza Conflict, ¶¶ 31, 33, U.N. Doc. A/HRC/29/52 (2015).
 See Prosecutor v. Martić, Case No. IT-95-11-T, Judgment, ¶¶ 461–63, 472 (June 12, 2007); Prosecutor v. Martić, Case No. IT-95-11-A, Appeal Judgment, ¶¶ 247–52 (Oct. 8, 2008).
 See Rome Statute, supra note 1, art. 8(2)(b)(vi); AP I, supra note 44, arts. 41, 85(3)(e); Hague Convention (IV) respecting the Laws and Customs of War on Land, Oct. 18, 1907, annex, art. 23(c), 36 Stat. 2277; Michael N. Schmitt, Military Necessity and Humanity in International Humanitarian Law, 50 Va. J. Int’l L. 795, 805 (2010).
 See Schmitt, supra note 122, at 805.
 See Jean-Marie Henckaerts et al., Commentary on the Third Geneva Convention ¶ 2119 (2021).
 See infra Section V.A.
 For evidence of the inefficacy of torture, see generally Shane O’Mara, Why Torture Doesn’t Work: The Neuroscience of Interrogation (2015).
 HCJ 5100/94 Pub. Comm. Against Torture in Israel v. State of Israel, 53(4) PD 3, 36–37 (1999).
 See, e.g., Itamar Mann & Omer Shatz, The Necessity Procedure: Laws of Torture in Israel and Beyond, 1987–2009, 6 Unbound: Harv. J. Legal Left (2010). That critique has now been confirmed, as the Supreme Court seems to have untied the government’s hand completely. See generally HCJ 9018/17 Tbeish & Pub. Comm. Against Torture in Israel v. Attorney General (2018) (Isr.), unofficial translation of the verdict available at https://perma.cc/38RM-U747.
 See Adil Haque, Law and Morality at War 32–34 (2017).
 See generally Schmitt, supra note 122.
 See David Luban, Military Necessity and the Cultures of Military Law, 26 Leiden J. Int’l L. 315, 339 (2013) (on the best understanding of IHL today, “civilian interests matter as much as military interests.”). But see generally Watts, supra note 66 (condemning the invocation of humanity at the expense of necessity).
 Lawless, supra note 116, at 312.
 Waxman, supra note 69, at 407. See also Watts, supra note 66, at 16 (integrating an analysis of the efficacy of encirclement deprivation and of indiscriminate bombardment).
 See AP I, supra note 44, at arts. 51(2), 51(5)(a); Rome Statute, supra note 1, arts. 8(2)(b)(i), 8(2)(e)(i). See also supra notes 118–22 and accompanying text.
 Allen, supra note 4, at 83.
 See, e.g., Hum. Rts. Comm., General Comment 34, ¶ 21, U.N. Doc CCPR/C/GC/34 (Sept. 12, 2011); Hum. Rts. Comm., General Comment 10: Freedom of Expression (Art. 19), ¶ 4 (June 29, 1983).
 See Gäfgen v. Germany, Eur. Ct. H.R. [GC], App. No. 22978/05, ¶ 107 (June 1, 2010). See also infra Section V.
 Emphasizing the importance of engaging in such inquiry, Naz Modirzadeh argues, “The hard part in our field is not the technical law math. The hard part is the morality, ethics, and the stakes of decision making. How to talk about those in a legal language, and how to actually say thoughtful or critical or constructive things about those questions. That is the hard part.” Naz Modirzadeh, Cut These Words: Passion and International Law of War Scholarship, 61 Harv. Int’l L.J. 1, 58–59 (2020). For Modirzadeh, contemporary scholars could learn from those who wrote on IHL during the Vietnam War, many of whom were driven to connect their legal analysis to their understanding of the moral purpose of the rules and the regime as a whole. Id.
 On the relationship between war crimes law and just war theory, see Tom Dannenbaum, War Crimes and Just War Theory, in The Palgrave Handbook of Applied Ethics and the Criminal Law 775 (Kimberly Kessler Ferzan & Larry Alexander eds., Palgrave Macmillan 2019). On the relationship between just war theory and the underlying law of armed conflict, see Haque, supra note 129.
 R.A. Duff, Towards a Modest Legal Moralism, 8 Crim. L. & Phil. 217, 219–20 (2014). See also R.A. Duff, Trials and Punishments, 81–82 (1986); Andrew Simester & Andreas von Hirsch, Crimes, Harms, and Wrongs: On the Principles of Criminalisation 19 (2011); John Kleinig, The Hardness of Hard Treatment, in Fundamentals of Sentencing Theory 273 (Andrew Ashworth ed., 1998); Dan M. Kahan, What Do Alternative Sanctions Mean?, 63 U. Chi. L. Rev. 591, 599 (1996); Victor Tadros, The Ends of Harm: The Moral Foundations of the Criminal Law 91 (2011); On expressivism at the domestic level, see Elizabeth S. Anderson & Richard H. Pildes, Expressive Theories of Law, 148 U. Pa. L. Rev. 1503 (2000); Joel Feinberg, The Expressive Function of Punishment, 49 Monist 397 (1965).
 See, e.g., Mark Drumbl, Atrocity, Punishment and International Law 3, 12, 61, 173–79 (2007); David Luban, Fairness to Rightness: Jurisdiction, Legality, and the Legitimacy of International Criminal Law, in The Philosophy of International Law 569, 575–76 (Samantha Besson & John Tasioulas eds., 2010); Robert D. Sloane, The Expressive Capacity of International Punishment, 43 Stan. J. Int’l L. 39 (2007); Diane Marie Amann, Group Mentality, Expressivism, and Genocide, 2 Int’l Crim. L. Rev. 93, 117 (2002); Margaret M. deGuzman, Choosing to Prosecute, 33 Mich. J. Int’l L. 265 (2012); Ralph Henham, Some Issues for Sentencing in the International Criminal Court, 52 Int’l & Comp. L. Q. 81, 111 (2003); Bill Wringe, Why Punish War Crimes?, 25 No. 2 L. & Phil. 151 (2006). For a comprehensive overview of expressivism in international criminal law, see Carsten Stahn, Justice as a Message: Expressivist Foundations of Criminal Civil Justice (2020).
 See, e.g., United States v. Göring (Oct. 1, 1946), Judgment of the International Military Tribunal, in 22 Trial of the Major War Criminals Before the International Military Tribunal 411, 466, (1948) [hereinafter Göring]; United States v. Alstötter, 3 Trials of War Criminals before the Nuernberg Military Tribunals under Control Council Law No. 10 954, 1011 (1949); Rome Statute, supra note 1, art. 33(1); Statute of the International Criminal Tribunal for the Former Yugoslavia, art. 7(4), U.N. Doc. S/RES/827 (1993). On the duty to disobey internationally criminal orders, see Tom Dannenbaum, The Crime of Aggression, Humanity, and the Soldier 20–25 (2018).
 Göring supra note 142, at 466. See also United States v. Otto Ohlendorf (Apr. 8–9, 1948), in 4 Trials of War Criminals Before The Nuernberg Military Tribunal Under Control Council Law No. 10 411, 470–88 (1949).
 Opening Statement at Nuremberg by Justice Robert H. Jackson, Chief Prosecutor for the United States, United States v. Göring (Nov. 21, 1945), in 2 Trial of the Major War Criminals Before the International Military Tribunal 98, 102 (1947).
 Prosecutor v. Blaškić, Case No. IT-95-14-A, Judgment, ¶ 678 (Int’l Crim. Trib. for the Former Yugoslavia July 29, 2004).
 Prosecutor v. Al Mahdi, ICC-01/12-01/15, Trial Judgment, ¶ 67 (Sept. 27, 2016), https://perma.cc/GCC9-YK9A.
 See, e.g., Prosecutor v. Bemba, Case No. ICC-01/05–01/08, Decision on Sentence pursuant to Article 76 of the Statute, ¶ 17 (June 21, 2016), https://perma.cc/W7RM-YAAA.
 See, e.g., Prosecutor v. Katanga, ICC-01/04–01/07, Decision on Sentence pursuant to Article 76 of the Statute, ¶¶ 49, 69–71, 143 (May 23, 2014), https://perma.cc/EYF4-2F9D.
 See Barrie Sander, The Expressive Turn of International Criminal Justice: A Field in Search of Meaning, 32 Leiden J. Int’l L. 851, 862–66 (2019).
 Charter of the International Military Tribunal, 82 U.N.T.S 279, arts 1, 6 (Aug. 8, 1945). On the critique, see, for example, William A. Schabas, Victor’s Justice: Selecting Situations at the International Criminal Court, 43 John Marshall L. Rev. 535, 536 (2010). This point was made most pointedly in the dissenting opinion of Radhabinod Pal in the similarly restricted tribunal established in Tokyo for the trial of defeated Japanese war criminals. Judgment of the Honorable Mr. Justice Pal, Member from India, Documents on The Tokyo International Military Tribunal 809, 826–27 (Neil Boister & Robert Cryer eds., 2008); Charter of the International Tribunal for the Far East (Jan. 19, 1946) 1589 T.I.A.S. 20, arts 1, 5.
 See, e.g., Letter from Philip Roth, Human Rights Watch to ICTR Chief Prosecutor Hassan Jallow in Response to His Letter on the Prosecution of RPF Crimes, Hum. Rts. Watch (Aug. 14, 2009), https://perma.cc/8QS8-LF5V; Withdrawal Strategy Document, African Union Doc. EX.CL/1006(XXX) (Jan. 2017); Asad G. Kiyani, Third World Approaches to International Criminal Law, 109 Am. J. Int’l L. Unbound 255 (2016).
 See Sander, supra note 149, at 867.
 See Tom Dannenbaum, Legitimacy in War and Punishment, in The Oxford Handbook of International Criminal Law 129, 134–36 (Kevin Jon Heller et al. eds., 2020).
 Matthew Gibson, Deceptive Sexual Relations, 40 Oxford J. Legal Studs. 82, 106 (2020); see also James Chalmers & Fiona Leverick, Fair Labelling in Criminal Law, 71 Modern L. Rev. 217, 231–34, 246 (2008).
 See Jeremy Horder, Rethinking Non-Fatal Offences Against the Person, 14 Oxford J. Legal Stud. 335, 351 (1994).
 See Chalmers & Leverick, supra note 154, at 228, 231–34, 246; Gibson, supra note 154, at 105; Victor Tadros, The Homicide Ladder, 69 Modern L. Rev. 601, 615, 618 (2006).
 Andrew Ashworth, Principles of Criminal Law 78 (6th ed. 2009). See also Simester & von Hirsch, supra note 140, at 19; Victor Tadros, Responses, 32 L. & Phil. 241, 248 (2013); Glanville Williams, Convictions and Fair Labelling, 42 Cambridge L.J. 85 (1983).
 Prosecutor v. Ruto & Sang, Case No. ICC-01/09-01/11-2027-Red, Trial Chamber V(A), Reasons of Judge Eboe-Osuji, Decision on Defence Applications for Judgments of Acquittal, ¶ 328 (Apr. 5, 2016), https://perma.cc/WT9E-B9RV; Talita de Souza Dias, The Retroactive Application of the Rome Statute in Cases of Security Council Referrals and Ad hoc Declarations, 16 J. Int’l Crim. Justice 65, 81 (2018); Ricardo Canese v. Paraguay, Inter‑Am. Ct. H.R. (ser. C) No. 111, ¶ 175 (Aug. 31, 2004).
 Gibson, supra note 154, at 108 (emphasis omitted).
 Id. at 107 (emphasis omitted). See also Tadros, supra note 157, at 248.
 See Leena Grover, Interpreting Crimes in the Rome Statute of the International Court 164 (2014).
 Mirjan Damaška, What is the Point of International Criminal Justice?, 83 Chi.-Kent L. Rev. 329, 353 (2008). Similarly, Dias argues that the principle would demand that the conduct underpinning international crimes “is labelled and described in a manner that is proportionate to the accused’s level of blameworthiness.” Dias, supra note 158, at 82.
 See David Luban, Calling Genocide by its Rightful Name: Lemkin’s Word, Darfur, and the UN Report, 7 Chi. J. Int’l L. 303 (2006); Amann, supra note 141, at 142–43.
 See, e.g., Prosecutor v. Karadžić, Case No. IT-95-5/18-T, Trial Judgment at 537–38, 551–55, 2446–64, 2614–26, 5655–73 (Int’l Crim. Trib. for the Former Yugoslavia, Mar. 24, 2016); Denic Džidić, Radovan Karadžić Convicted of Genocide, Balkan Transitional Just. (Mar. 24, 2016), https://perma.cc/F8PT-7FTA.
 See Marko Milanovic, Establishing the Facts About Mass Atrocities, 47 Geo. J. Int’l L. 1321, 1357–58 (2016).
 See, e.g., Co-Prosecutors v. Nuon Chea & Khieu Samphan, Case File No. 002/19-09-2007/ECCC/TC, Judgment (Extraordinary Chambers in the Courts of Cambodia, Nov. 16, 2018) (Ottara J., sep. op.); Stéphanie Giry, The Genocide That Wasn’t, N.Y. Rev. Books (Aug. 25, 2014), https://perma.cc/GMD4-6EPP.
 See, e.g., Statute of the International Criminal Tribunal for the Former Yugoslavia, supra note 142, at art. 24; Prosecutor v. Kayishema & Ruzindana, Case No. ICTR-95-1-A, Appeal Judgment, ¶ 367 (Dec. 4, 2001).
 Such courts and tribunals may, of course, choose to sentence more harshly for certain crimes, but there is no requirement that they do so, and they are anyway capable of doing that for discriminatory intent even in the absence of a separate criminal category. See Sloane, supra note 141, at 67–69; Jens David Ohlin, Towards a Unique Theory of International Criminal Sentencing, in International Criminal Procedure: Towards a Coherent Body of Law (Goran Sluiter & Sergey Vasiliev, eds., 2009). See also Rome Statute, supra note 1, arts. 77–78.
 See Radovan Karadžić, for example, was 73 when his sentence was extended from forty years to life on the basis that others responsible for the same crimes (both those convicted of genocide and those not convicted of genocide) were sentenced to life. Prosecutor v. Karadžić, Case No. MICT-13-55-A, Appeal Judgment, ¶¶ 761–76 (Int’l. Crim. Trib. for the Former Yugoslavia, Mar. 20, 2019).
 See Luban, supra note 163.
 Prosecutor v. Kvočka, Case No. IT-98-30/1-A, Appeal Judgment, ¶ 92 (Int’l. Crim. Trib. for the Former Yugoslavia, Feb. 28, 2005).
 Rome Statute, supra note 1, arts. 7(1)(g), 8(2)(b)(xxi), 8(2)(b)(xxii), 8(2)(c)(ii), 8(2)(e)(vi). On the shift over time towards legal categories that more better reflect the nuances across of the distinct wrongs entailed in these violations, see, for example, Christine Chinkin, Gender and Armed Conflict, in The Oxford Handbook of International Law in Armed Conflict 675, 681–91 (Andrew Clapham & Paola Gaeta eds., 2014).
 Prosecutor v. Ongwen, ICC-02/04-01/15-1762-Red, Trial Judgment, ¶ 2719 (Feb. 4, 2021), https://perma.cc/58EV-2RW3 (citation omitted).
 Id. ¶ 2722.
 See Prosecutor v. Brima, Kamara & Kanu, SCSL-2004-16-A, Appeal Judgment, ¶¶ 190–99 (Special Court for Sierra Leone, Feb. 22, 2008); Prosecutor v. Sesay, Kallon & Gbao, Case No. SCSL-04-15-A, Appeal Judgment, ¶¶ 735–36 (Special Court for Sierra Leone, Oct. 26, 2009); Co-Prosecutors v. Nuon Chea & Khieu Samphan, Case File No. 002/19-09-2007/ECCC/TC, Judgment ¶¶ 3686–95 (Extraordinary Chambers in the Courts of Cambodia, Nov. 16, 2018); Ongwen, ICC-02/04-01/15-1762-Red, Trial Judgment at ¶¶ 2741–53 (Feb. 4, 2021); Leila Nadya Sadat, Crimes against Humanity in the Modern Age, 107 Am. J. Int’l L. 334, 350 (2013).
 See Ongwen, ICC-02/04-01/15-1762-Red, Trial Judgment at ¶¶ 2747–50 (Feb. 4, 2021).
 Brima, SCSL-2004-16-A, Appeal Judgment at ¶¶ 190–99.
 Rome Statute, supra note 1, art. 7(1)(h).
 See International Criminal Court, Rules of Procedure and Evidence, ICC-PIDS-LT-02-002/13, r. 145(2)(b)(v) (2d ed. 2013).
 See Sadat, supra note 176, at 344.
 See Sander, supra note 149. On the interpretive implications of a normative account, see Dannenbaum, supra note 17, at 1301–17; Tom Dannenbaum, The Criminalization of Aggression and Soldiers’ Rights, 29 Eur. J. Int’l L. 859 (2018).
 See supra Section III.E.
 See, e.g., D’Alessandra & Gillett, supra note 18, at 826; DeFalco, supra note 66, at 1159, 1142–49.
 On systematicity as an element of crimes against humanity, see, for example, Prosecutor v. Kunarac, Case No. IT-96-23-T & IT-96-23/1-A, Appeal Judgment, ¶ 94 (Int’l Crim. Trib. For the Former Yugoslavia, June 12, 2002).
 See DeFalco, supra note 66, at 1159–75. But note the challenges discussed at infra note 211 and accompanying text.
 Cottier & Richard, supra note 50, at 509.
 Gloria Gaggioli & Nils Melzer, Methods of Warfare, in The Oxford Guide to International Humanitarian Law 235, 248 (Ben Saul & Dapo Akande eds., 2020).
 Sandoz, Swinarski & Zimmermann, supra note 73, ¶ 4794.
 See Åklagarmyndigheten, supra note 14.
 U.K. LOAC Manual, supra note 69, ¶ 15.19.1; Rep. of the Int’l Comm’n of Inquiry on Darfur to the U.N. Secretary-General, Pursuant to Security Council Resolution 1564 of 18 September 2004, ¶ 166(x) n.90 (Jan. 25, 2005) [hereinafter ICID Report].
 See U.K. LOAC Manual, supra note 69, ¶ 15.19.1; ICID Report, supra note 191, ¶ 166(x), n.90.
 On arguments for a narrow reading of “attack” that would exclude encirclements, see supra note 70. On the other hand, some have argued for a broader understanding of the term and the ICC jurisprudence on the issue is currently murky. Ori Pomson, Ntaganda Appeals Chamber Divided on Meaning of ‘Attack,’ Articles of War (May 12, 2021), https://perma.cc/4GY6-EDAU. On encirclement deprivation as an “attack,” see Gloria Gaggioli, Joint Blog Series on International Law and Armed Conflict: Are Sieges Prohibited under Contemporary IHL?, EJIL:Talk! (Jan. 30, 2019), https://perma.cc/SX2R-Q67D; Gillard, supra note 111, at 8.
 U.K. LOAC Manual, supra note 69, ¶ 15.19.1.
 See Marcus, supra note 6, at 273.
 See supra Section III.E.
 See supra Section II.A.
 By war convention, Walzer refers to “the set of articulated norms, customs, professional codes, legal precepts, religious and philosophical principles, and reciprocal arrangements that shape our judgements of military conduct.” Walzer, supra note 95, at 44.
 Id. at 162.
 See supra note 193.
 Watts, supra note 66, at 47.
 See Conley & de Waal, supra note 4, at 701.
 See ‘It is Necessary that Those Who are Responsible for These Famines Fear that They Could be Prosecuted for Their Crimes’: An Interview with Jane Ferguson, 17 J. Int’l Crim. Justice 907, 909 (2019).
 See De Waal, supra note 4, at 22–23.
 See ICC Elements of Crimes, supra note 80, at 5–6, 13, 15, 19, 31.
 See Prosecutor v. Kordić & Čerkez, Case No. IT-95-14/2-A, Appeal Judgment, ¶¶ 66–67 (Int’l Crim. Trib. for the Former Yugoslavia, Dec. 17, 2004).
 See ICC Elements of Crimes, supra note 80, at 18, 34. See supra note 193.
 See Co-Prosecutors v. Nuon Chea & Khieu Samphan, Case File No. 002/19-09-2007/ECCC/SC F36, Appeal Judgment, ¶ 536–57 (Extraordinary Chambers in the Courts of Cambodia, Nov. 23, 2016).
 U.N. Comm. Rep. Sarajevo, supra note 67, ¶ 75.
 Id. ¶ 77.
 See id. ¶ 87. The Commission was considering the siege through the lens of starvation of civilians as a method of warfare. However, it did so prior to the crime’s codification. When codification occurred through the Rome Statute, it was made clear that civilian harm or death is not a consequence element of the crime. ICC Elements of Crimes, supra note 80, at 31.
 See, e.g., Conley & de Waal, supra note 4, at 701–03; Jordash, Murdoch & Holmes, supra note 74, at 850; Antonio Coco, Jérôme de Hemptinne & Brian Lander, International Law Against Starvation in Armed Conflict: Epilogue to a Multi-faceted Study, 17 J. Int'l Crim. Just. 913, 913 (2019); IIFFMM, Detailed Findings (2019), supra note 8; Rep. of the Detailed Findings of the Comm’n of Inquiry on Human Rights in the Democratic People’s Republic of Korea, U.N. Doc. A/HRC/25/CRP.1, ¶ 542–553 (Feb. 7, 2014) [hereinafter DPRK Report Detailed Findings]; Cesar G. Victora et al., Maternal and Child Undernutrition: Consequences for Adult Health and Human Capital, 371 Lancet 340 (2008).
 On the reverberating effects of kinetic attacks, see, for example, Isabel Robinson & Ellen Nohle, Proportionality and Precautions in Attack, 98 Int’l Rev. Red Cross 107 (2017).
 Ventura, supra note 75, at 792. See id. at 798 (starvation and death are “two sides of the same coin”).
 Describing this underlying act as genocide by “measures of slow death,” see Independent International Fact-Finding Mission on Myanmar, U.N. Doc. A/HRC/39/CRP.2, ¶ 1400 (Sept. 17, 2018). Identifying starvation methods as potentially genocidal via this underlying act, see ICC Elements of Crimes, supra note 80, at 3 n.4; Prosecutor v. Tolimir, Case No. IT-05-88/2-A, Appeal Judgment, ¶ 226 (Int’l Crim. Trib. for the Former Yugoslavia, Apr. 8, 2015); Prosecutor v. Akayesu, Case No. ICTR-96-4-T, Trial Judgment, ¶ 506 (Sept. 2, 1998); Ventura, supra note 75, at 810 (noting that this category “matches [starvation] perfectly”).
 See supra note 118 and accompanying text.
 See supra notes 120–21 and accompanying text.
 See AP I, supra note 44, art 51(4–5); Hum. Rts. Council, Situation of Hum. Rts. in Yemen, Including Violations and Abuses Since September 2014: Rep. of the Group of Eminent Int’l and Reg’l Experts as submitted to the U.N. High Comm’r for Hum. Rts., ¶¶ 32–34, U.N. Doc. A/HRC/42/17 (Aug. 9, 2019); Group of Experts on Yemen, Detailed Findings (2019), supra note 7, ¶ 316. See also supra note 121.
 Prosecutor v. Ntaganda, Case No. ICC-01/04-02/06-2359, Judgment, ¶ 923 (July 8, 2019), https://perma.cc/LF76-5ZDK. See also id. at ¶¶ 926, 1057; Prosecutor v. Katanga, Case No. ICC-01/04-01/07-3436, Judgment, ¶ 802 (Mar. 7, 2014), https://perma.cc/J4BE-ENA3.
 See AP I, supra note 44, art. 50(3). The ICTY held repeatedly that “a population may qualify as civilian as long as it is predominantly civilian,” finding on that basis that “the population of the urban areas inside the confrontation lines of Sarajevo between 1992 and 1995 had civilian status as a whole,” notwithstanding the presence of combatants in those areas. Prosecutor v. Karadžić, Case No. IT-95-5/18-T, Judgment ¶ 4610 n.15510 (Int’l Crim. Trib. for the Former Yugoslavia Mar. 24, 2016). See also id. ¶ 474; Prosecutor v. Galić, Case No. IT-98-29-A, Appeal Judgment, ¶¶ 135–38 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 30, 2006); Prosecutor v. Kordić & Čerkez, Case No. IT-95-14/2-A, Judgment, ¶ 50 (Int’l Crim. Trib. for the Former Yugoslavia Dec.17, 2004); Prosecutor v. Blaškić, Case No. IT-95-14-A, Judgment, ¶ 115 (Int’l Crim. Trib. for the Former Yugoslavia July 29, 2004). The ICC has affirmed this principle in its own jurisprudence. Katanga, ICC-01/04-01/07-3436, Judgment at ¶ 1105.
 On the IHL prohibition of attacks on civilian populations, as distinct from specific civilians, see AP I, supra note 44, art. 51(2); AP II, supra note 44, art. 13(2). For the criminality of attacking civilian populations as distinct from specific civilians, see AP I, supra note 44, art. 85(3)(a); Rome Statute, supra note 1, arts. 8(2)(b)(i), (e)(i). The Ntaganda Appeals Chamber at the ICC clarified recently that the targeting of a civilian population without distinction between civilians and combatants can qualify as an attack directed against a civilian population. Ntaganda, ICC‑01/04-02/06-2666-Red, Appeal Judgment at ¶¶ 418, 424, 491 (Mar. 30, 2021), https://perma.cc/6P2E-EJ8G. See also Galić, Case No. IT-98-29-A, Appeal Judgment at ¶¶ 131–38.
 Cf. Int’l Comm. of the Red Cross, International Humanitarian Law and the Challenges of Contemporary Armed Conflicts 19–22 (2019), https://perma.cc/23TP-ZYSH [hereinafter ICRC Challenges 2019]; Int’l Comm. of the Red Cross, International Humanitarian Law and the Challenges of Contemporary Armed Conflicts 47–53 (2015), https://perma.cc/RT7W-34L9; Int’l Comm. of the Red Cross, International Humanitarian Law and the Challenges of Contemporary Armed Conflicts 40–42 (2011), https://perma.cc/Z8MF-AVKP; Laurent Gisel, The Use of Explosive Weapons in Densely Populated Areas and the Prohibition of Indiscriminate Attacks, in Conduct of Hostilities: the Practice, the Law and the Future 100, 103 (Edoardo Greppi ed., 2014); Remarks by Department of Defense General Counsel Paul C. Ney Jr. on the Law of War, Just Sec. (May 28, 2019), https://perma.cc/N4DA-SBAP.
 See ICRC Challenges 2019, supra note 222, at 21.
 See supra note 120.
 See AP I, supra note 44, art. 51(5)(b); Rome Statute, supra note 1, art. 8(2)(b)(iv).
 Seeking to reconcile this principle with the permissibility of the infliction of proportionate collateral harm arising from discriminate operations, see Haque, supra note 129, ch. 8.
 See supra Sections II.B & III.A.
 On the distinction between those who think they are and those who think they ought to be, see supra notes 96–107 and accompanying text.
 Occasionally, those advocating the permissibility of encirclement deprivation also advocate the permissibility of indiscriminate kinetic attack. See supra note 133.
 Reverberating effects could include the harm arising from the dispersal of unexploded ordnance or the destruction of units providing essential services. Robinson & Nohle, supra note 213; IIFFMM, Detailed Findings (2019), supra note 8, ¶¶ 407, 571; ICRC Challenges 2019, supra note 222, at 20.
 Beehner, Berti & Jackson, supra note 26, at 82. The approach of the besieging party is to “achieve a decision . . . or to slowly destroy the besieged.” Fox, supra note 26, at 2.
 Dinstein, supra note 96, at 151. Invoking and building on Dinstein’s argument here, see Waxman, supra note 69, at 418–19.
 U.K. LOAC Manual, supra note 69, ¶ 5.34.1 (on the objective of “forcing . . . surrender”).
 U.S. DoD Law of War Manual, supra note 66, § 5.19.1 (on aiming at “submission”).
 Watts, supra note 66, at 15 (on speeding “capitulation”).
 George Alfred Mudge, Starvation as a Means of Warfare, 4 Int’l L. 228, 246 (1970) (noting that “[t]he definition does not read ‘for the purpose of killing them with hunger’ but only ‘for the purpose of starving them into surrender.’ . . . The implicit definition of starvation assumes that there will be a military effect, i.e., capitulation of the besieged locality”).
 Hampson, supra note 93, at 99 (2016). Dov Weisglass, senior advisor to former Israeli Prime Minister Ehud Olmert, made a similar point in particularly callous terms, defending the Israeli blockade of Gaza by describing the goal as “to put the Palestinians on a diet, but not to make them die of hunger.” Conal Urquhart, Gaza on Brink of Implosion as Aid Cut-Off Starts to Bite, Guardian (Apr. 15, 2006), https://perma.cc/HPV9-USXU. See also Cedric Ryngaert, Humanitarian Assistance and the Conundrum of Consent: A Legal Perspective, 5 Amsterdam L. Forum 5, 8 (2013) (“One can argue that denying access to outside humanitarian actors could lead to starvation and thus amount not only to a violation of the humanitarian law prohibition of starvation, but also of the (human) right to life . . . From this perspective states could only deny access if such denial will not result in the deaths of individuals deprived of assistance.”).
 Hampson, supra note 93, at 99. Notably, Hampson does also worry that too much focus (legally) is placed on starvation at the expense of this lesser harm. Id.
 ICRC Challenges 2019, supra note 222, at 23. On the dangers of urban combat for civilians, see Id. at 16–18; Nathalie Durhin, Protecting Civilians in Urban Areas, 98 Int’l Rev. Red Cross 177, 178–79 (2017).
 Beehner, Berti & Jackson, supra note 26, at 78.
 Id. at 80.
 ICRC Challenges 2019, supra note 222, at 23.
 Fox, supra note 26, at 4.
 See Prosecutor v. Dragomir Milošević, Case No. IT-98-29/1-A, Judgment, ¶¶ 32–36 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 12, 2009); Prosecutor v. Galić, Case No. IT-98-29-A, Appeal Judgment, ¶¶ 91, 100–02 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 30, 2006).
 See, e.g., supra notes 4–11 and accompanying text.
 See Prosecutor v. Galić, Case No. IT-98-29-T, Judgment, ¶¶ 582–94, 597 (Int’l Crim. Trib. for the Former Yugoslavia Dec. 5, 2003).
 On the complexity of causation in this context, see supra notes 202–11 and accompanying text.
 This is the “prospective view” pursuant to which past harms are moral sunk costs and the moral viability of the action going forward depends exclusively on harms expected in the future. Seth Lazar, Moral Sunk Costs, 68 Phil. Q. 841, 842, 844 (2018).
 See Dinstein, supra note 69, at 256–57. On the offer of egress, see supra note 69.
 See, e.g., U.K. LOAC Manual, supra note 69, ¶ 5.34.3; J. M. Spaight, War Rights on Land 164 (1911). Channeling arguments to this effect, see Mudge, supra note 236, at 236, 241.
 See Conley & de Waal, supra note 4, at 709–11; Beehner, Berti & Jackson, supra note 26, at 81; Provost, supra note 40, at 619–22. Recall the aforementioned report on Syria, which recognized the method as having been notably “successful for overtaking opposition-held territory” in the conflict but at the cost of proving “disastrous for civilians.” Independent Int’l Comm’n of Inquiry on Syria Rep., supra note 87, ¶ 95; Hum Rts. Council, The Siege and Recapture of Eastern Ghouta: Conf. Room Paper of the Indep. Int’ Comm’n of Inquiry on the Syrian Arab Republic, ¶¶ 31–37, U.N. Doc. A/HTC/38/CRP.3 (June 20, 2018). See also supra notes 4–11 and accompanying text.
 See supra note 251. Indeed, many who argue that encirclement deprivation remains lawful are clear‑eyed about this reality. Watts, supra note 66, at 4; Drew, supra note 66, at 309 n.34.
 See Brian Lander & Rebecca Vetharaniam Richards, Addressing Hunger and Starvation in Situations of Armed Conflict - Laying the Foundations for Peace, 17 J. Int’l Crim. Just. 675, 683–84 (2019) (noting the slowness with which officials react to the slow devastation of mass starvation.).
 Critiquing this perspective in other contexts, such as jus ad bellum proportionality, see generally Lazar, supra note 248.
 See Gillard, supra note 111, at 12. The direct participation in hostilities threshold is provided in AP I, supra note 44, art. 51(3). What surpasses that threshold is subject to debate, with voluntary human shields falling into the zone of disagreement. Compare Int’l Comm. Red Cross, Interpretive Guidance on the Notion of Direct Participation in Hostilities under International Humanitarian Law 56–57 (2009), with HCJ 769/02, Pub. Comm. Against Torture in Israel v. Gov’t of Israel, (2) IsrLR 459, 498 (2006). However, whatever one makes of that debate, it is implausible to hold that declining to leave one’s home could qualify a person as a voluntary human shield and thereby a lawful target. See, e.g., Hum. Rts. Council, Rep. of the Special Rapporteur on Extrajudicial or Arbitrary Executions, Philip Alston; Special Rapporteur on the Right of Everyone to the Enjoyment of the Highest Attainable Standard of Physical and Mental Health, Paul Hunt; Representative of the Secretary-General on Hum. Rts. of Internally Displaced Persons, Walter Kälin; Special Rapporteur on Adequate Housing as a Component of the Right to an Adequate Standard of Living, Miloon Kothari: Mission to Lebanon and Israel, ¶ 41, U.N. Doc. A/HRC/2/7 (Oct. 2, 2006); ICRC Challenges 2019, supra note 222, at 17, 25. Also on the importance of preserving respect for civilian status in such circumstances, see Provost, supra note 40, at 619.
 On the besieged party’s duties, see AP I, supra note 44, arts. 51(7), 58. On irrelevance of their violation for duties of besieging party, see id. art. 51(8); Gillard, supra note 111, at 8.
 Rome Statute, supra note 1, arts. 7(1)(d), (2)(d); AP I, supra note 44, art. 54(3)(b).
 Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment art. 2(2), 1456 U.N.T.S. 113 (Dec. 10, 1984) [hereinafter CAT].
 Gäfgen, Eur. Ct. H.R. [GC], App. No. 22978/05, ¶ 107.
 For a wide range of views on the issue, see generally Torture: A Collection (Sanford Levinson ed. 2004).
 See Jay S. Bybee, Office Legal Counsel, Department of Justice, Memorandum for Alberto R. Gonzales, Counsel to the President Re: Standards of Conduct for Interrogation Under 18 U.S.C §§ 2340–2340A, 31–33, 39–45 (Aug. 1, 2002); CIA Office General Counsel, Draft: Legal Appendix, Hostile Interrogations: Legal Considerations for CIA Officers (Nov. 26, 2001); Senate Select Comm. on Intel., Committee Study of the Central Intelligence Agency’s Detention and Interrogation Program: Executive Summary 181, 197 n.1155 (2014), https://perma.cc/WDZ6-GYYF.
 See Jonathan Swan, Trump Calls for ‘Hell of a lot Worse than Waterboarding’, Hill (Feb. 6, 2016), https://perma.cc/R5XT-HBYP.
 See HCJ 5100/94 Pub. Comm. Against Torture in Israel v. State of Israel, 53(4) PD ¶¶ 3638 (1999) (allowing for the possibility that an official who engages in otherwise criminal interrogation techniques may “find refuge under [the ‘necessity’ defence’s] wings” on an ad hoc basis in light of “a given set of facts,” but emphasizing that necessity “cannot serve as [the] basis for [an ex ante administrative] authority” to employ such techniques); HCJ 9018/17 Tbeish v. Attorney General, ¶¶ 61–66 (2018) (Isr.) (holding that officials who engage in such techniques on the basis of necessity “are entitled to an appropriate measure of legal certainty,” such that the Attorney General may “detail[ ] [ ] circumstances that will be considered in reaching a decision whether the interrogators’ action falls within the scope of the necessity defense” and interrogators may act pursuant to those guidelines and in consultation with their superiors without this system thereby falling afoul of the principle that necessity cannot underpin an ex ante administrative authority).
 See O’Mara, supra note 126, at 234.
 See Kim Lane Scheppele, Hypothetical Torture in the ‘War on Terrorism,’ 1 J. Nat’l Sec. L. & Pol’y 285, 293–95, 305–07 (2005). See generally Henry Shue, Torture in Dreamland, 37 Case W. Res. J. Int’l L. 231 (2006).
 See Scheppele, supra note 265, at 307–18; Jeremy Bentham, Of Torture, reprinted in W.L. Twining & P.E. Twining, Bentham on Torture, 24 N. Ir. L. Q. 305, 308 (1973); Id. at 311 (“The great objection against Torture is, that it is so liable to abuse.”); Oren Gross, The Prohibition on Torture and the Limits of the Law, in Torture: A Collection 229 (Sanford Levinson ed. 2004).
 Henry Shue, Torture, 7 Phil. & Pub. Aff. 124, 125 (1978).
 The probability point is important because a common variant on the practical rebuttals to the necessity claim noted above is that the interrogator does not know with absolute certainty that the person interrogated in fact has the relevant information, and so risks inflicting suffering to no end. This, of course, is also true in the context of lethal targeting in war. On the normative challenges related to killing under uncertainty in war, see Haque, supra note 129, ch. 5; Dannenbaum, supra note 142, at 147–53.
 David Sussman, What’s Wrong with Torture?, 33 Phil. & Pub. Aff. 1, 3 (2005).
 See Shue, supra note 267, at 125.
 Id. at 130.
 Sussman, supra note 270, at 7.
 Rome Statute, supra note 1, art. 7(2)(e).
 See ICC Elements of Crimes, supra note 80, at 32–33.
 William A. Schabas, The International Criminal Court: A Commentary on the Rome Statute 244 (2d ed. 2016).
 Manfred Nowak & Elizabeth McArthur, The United Nations Convention against Torture: A Commentary 22, 76 (2008).
 Shue, supra note 267, at 130.
 Sussman, supra note 270, at 16.
 Shue, supra note 267, at 135.
 Id. at 141.
 See Sussman, supra note 270, at 18.
 Sussman writes,
[T]orture forces its victim into the position of colluding against himself through his own affects and emotions, so that he experiences himself as simultaneously powerless and yet actively complicit in his own violation. So construed, torture turns out to be not just an extreme form of cruelty, but the pre-eminent instance of a kind of forced self-betrayal, more akin to rape than other kinds of violence characteristic of warfare or police action.
Id. at 4.
 Id. at 8–9.
 Id. at 10.
 Id. at 14.
 Id. at 20.
 Id. at 21.
 Id. at 29.
 CAT, supra note 258, art 1(1).
 Nowak & McArthur, supra note 280, at 69. See also id. at 558; IIFFMM, Detailed Findings (2018), supra note 8, ¶ 162.
 See Hum. Rts. Comm., CCPR General Comment No. 20: Article 7 (Prohibition of Torture, or Other Cruel, Inhuman or Degrading Treatment or Punishment), ¶ 4, UN Doc. HRI/GEN/1/Rev.7 (1994).
 See Inter-American Convention to Prevent and Punish Torture, art. 2, Dec. 9, 1985, O.A.S.T.S. No. 67 (includes the CAT list but adds: “or for any other purpose”); Huri-Laws v. Nigeria, Communication No. 225/98, African Comm. Hum. & Peoples’ Rts. [Afr. Comm’n H.P.R.], ¶ 41 (Nov. 6, 2000). For a long time, the European Court of Human Rights distinguished torture from other cruel treatment primarily with reference to the severity of the pain inflicted, and not the purpose (overriding in that respect a prior decision by the retired European Commission of Human Rights). See, e.g., Ireland v. U.K., App. No. 5310/71, ¶ 167 (Jan. 18, 1978), https://perma.cc/8KNA-8RJE. However, the Court has since re-emphasized the purposive element as an important distinguishing feature. See, e.g., Gäfgen, App. No. 22978/05, ¶¶ 88, 90, 105.
 See ICC Elements of Crimes, supra note 80, at 7, 14, 32.
 Prosecutor v. Krnojelac, Case No. IT-97-25-T, Judgment, ¶ 180 (Int’l Crim. Trib. for the Former Yugoslavia Mar. 15, 2002); Prosecutor v. Delalić, Case No. IT-96-21-T, Judgment, ¶ 542 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 16, 1998); Prosecutor v. Bemba, ICC-01/05-01/08‑424, Decision on the Confirmation of Charges, ¶¶ 291, 293–94 (June 15, 2009), https://perma.cc/8S27-9UER.
 Tibi v. Ecuador, Preliminary Objections, Merits, Reparations and Costs, Inter-Am. Ct. H.R. (ser. C) No. 114, ¶¶ 146, 148 (Sept. 7, 2004).
 Notably, since various other forms of mistreatment are both illegal and criminal under international law, the purpose of this distinction may primarily be expressive. At the same time, it is worth noting that this line can sometimes be blurred. Although focused on the line between categorical prohibitions and those allowing exception (rather than the line between torture and cruel treatment), the Israeli Supreme Court’s examination of sleep deprivation draws on a similar principle, distinguishing sleep deprivation that is a “side effect” of certain forms of interrogation from that inflicted “for the purpose of tiring [the detainee] out or ‘breaking’ him.” HCJ 5100/94 Pub. Comm. Against Torture in Israel v. State of Israel, 53(4) PD 3, 28–29 (1999). See also Gäfgen, App. No. 22978/05, ¶ 89 (on treatment qualifying as degrading when it has the effect of “possibly breaking [the victim’s] physical or moral resistance.”).
 See, e.g., Prosecutor v. Kvočka, Case No. IT-98-30/1-T, Judgment, ¶ 144 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 2, 2001). Controversially, feeding detainees a “reduced diet” during their detention and “pending interrogations” was found by the European Court of Human Rights not to constitute torture in 1978. Ireland v. U.K., App. No. 5310/71, ¶¶ 96, 246 (Jan. 18, 1978), (holding that subjecting detainees to a “reduced diet during their stay at the centre and pending interrogations” qualified as inhuman and degrading treatment, but not torture). However, this continues to be litigated, and the Court of Appeal in Northern Ireland ruled in 2019 that precisely the techniques at stake in the 1978 case were in fact torture. McGuigan’s & McKenna’s Application,  NICA 46, ¶ 116. Moreover, the European Court of Human Rights has itself hinted that the threshold for torture may now include such techniques. Selmouni v. France, App. No. 25803/94, ¶ 101 (July 28, 1999), https://perma.cc/XMC9-BG8N. In refusing to revisit the 1978 judgment, the Court acknowledged that its “case-law on the notion of torture has evolved” and emphasized that it was relying on the fact that “a request for revision is not meant to allow a party to seek a review in the light of the Court’s subsequent case‑law.” Ireland v. U.K., App. No. 5310/71, ¶¶ 124–25 (Mar. 20, 2018), https://perma.cc/57XE-LBQR.
 See supra note 280 and accompanying text.
 Suggesting inhuman treatment may apply, see Group of Experts on Yemen, Detailed Findings (2019), supra note 7, ¶ 746; Jelena Pejic, The Right to Food in Situations of Armed Conflict, 83 Int’l Rev. Red Cross 1097, 1106 (2001).
 See supra notes 279–81.
 See Ventura, supra note 75, at 794.
 See supra notes 92–93 and accompanying text. See also Hampson, supra note 86, at 93.
 Sussman, supra note 270, at 24.
 Id. at 23.
 Id. at 29.
 On the use of isolated relief to deflect public attention, see Dannenbaum, supra note 11.
 Sussman, supra note 270, at 29.
 See AP I, supra note 44, art. 51(3).
 See Seth Lazar, Sparing Civilians 101–23 (2015). It is worth noting that both the defenselessness of civilians and their non-target status has been questioned in just war theory. Richard J. Arneson, Just Warfare Theory and Noncombatant Immunity, 39 Cornell Int’l L.J. 663 (2006); Helen Frowe, Defensive Killing chs. 5–6 (2014); Jeff McMahan, Killing in War chs. 4–5 (2009).
 See AP I, supra note 44, arts. 43, 50(1), 51(3).
 See supra note 255. As noted above, some just war theorists question the moral non-liability of civilians to attack. See supra note 317. On the relationship of those debates to existing law, see Tom Dannenbaum, War Crimes and Just War Theory, in The Palgrave Handbook of Applied Ethics 775 (Larry Alexander & Kimberly Kessler Ferzan eds., 2019).
 See Shue, supra note 267, at 128–30.
 Chris Dolan argues for the idea of “social torture” in a case study of atrocities committed in Northern Uganda during the period 1986–2006. See generally Chris Dolan, Social Torture: The Case of Northern Uganda, 1986–2006 (2009). It is a powerful argument, albeit one that is quite different from that elaborated here in that Dolan is interested in the application of torture to the scenario he examines, not just in its utility as a normative analogue, and the situation he examines involves the detailed control of victims by perpetrators in the torturous moments, insofar as the latter are able to return to and abuse specific individuals over time. See, e.g., id. at 1–2, 6–8.
 See supra note 22.
 See Shue, supra note 267, at 133.
 See supra notes 18, 186 and accompanying text.
 Walzer, supra note 95, at 163.
 See Tom Dannenbaum, Encirclement, Deprivation, and Humanity, 97 Int’l L. Stud. 307, 360–64 (2021).
 Sussman, supra note 270, at 14.
 Id. at 20.
 See AP I, supra note 44, arts 51(7), 58; Walzer suggests that resistance to a siege is altogether precluded when it entails subjecting the civilian population to starvation conditions without their consent. Walzer, supra note 95, at 163 (“Consent clears the defenders, and only consent can do so.”).
 See supra note 256–258 and accompanying text.
 Beehner, Berti & Jackson, supra note 26, at 82.
 Sussman, supra note 270, at 29.
 Equating the two, see Ventura, supra note 75, at 782 (“[T]here is little to distinguish between a person facing death through starvation and another threatened with arbitrary execution because of her political beliefs” (quoting with approval U.N. Off. of the High Comm’r for Hum. Rts., Fact Sheet No. 20: Human Rights and Refugees (July 1993) https://www.ohchr.org/Documents/Publications/FactSheet20en.pdf)). It may be that the psychological impact of iterated terror bombing or other forms of terror attack have a psychological crowding effect similar to that of starvation. See supra note 244 and infra notes 369–71 and accompanying text.
 See supra note 293 and accompanying text.
 De Waal, supra note 4, at 18.
 Id. at 20.
 Cormac Ó Gráda, Famine: A Short History 48 (2009).
 Id. See also generally, id. ch. 2.
 Pitrim Sorokin, Hunger as a Factor in Human Affairs 148 (1975).
 ICRC Challenges 2019, supra note 222, at 23.
 De Waal, supra note 4, at 41.
 See DPRK Report Detailed Findings, supra note 212, ¶¶ 518, 528, 565, 567.
 De Waal, supra note 4, at 18.
 See id. at 17–21.
 DPRK Report Detailed Findings, supra note 212, ¶ 560.
 De Waal, supra note 4, at 21.
 See supra Section V.B.
 De Waal, supra note 4, at 41.
 Id. at 20–21. See also Conley & de Waal, supra note 4, at 700–01.
 Primo Levi, The Drowned and the Saved 38 (Raymond Rosenthal, trans., Simon & Schuster 1988, reissue ed. 2017) (1986).
 Breandán Mac Suibhne, The End of Outrage: Post-Famine Adjustment in Rural Ireland 17 (2017).
 On the wrong of being forced to do wrong by coercion, rather than by torturous crowding, see Dannenbaum, supra note 142, at 25–30, 37–49, 126–31, 253–56.
 On standing to blame, see G. A. Cohen, Casting the First Stone: Who Can and Who Can’t, Condemn the Terrorists, in Finding Oneself in the Other ch. 6 (Michael Otsuka ed. 2013); Marilyn Friedman, How to Blame People Responsibly, 47 J. Value Inquiry 271 (2013); T. M. Scanlon, Moral Dimensions: Permissibility, Meaning, Blame 175–79 (2008). On burden, see Dannenbaum, supra note 142, at 126–31.
 See Scanlon, supra note 355, at 138, 146, 150.
 See generally supra note 355.
 See Dannenbaum, supra note 142, at 127.
 As discussed above, that wrong attaches both when the distortion leads to the breaking of will and when it does not. See supra notes 296, 332–33.
 See Susan Wolf, The Moral of Moral Luck, 31 Phil. Exch. 1, 4, 7, 9 (2001); Margaret Urban Walker, Moral Luck and the Virtues of Impure Agency, 22 Metaphilosophy 14, 18, 21 (1991); Dannenbaum, supra note 142, at 127–31.
 Dannenbaum, supra note 142, at 127.
 On the moral importance of one’s connection to specific acts and consequences, see, for example, Martha C. Nussbaum, Luck and Ethics, in Moral Luck 73, 74 (Daniel Statman ed. 1993); Bernard Williams, Moral Luck, in Moral Luck: Philosophical Papers 1973–1980, at 20, 29–30 (1981); Wolf, supra note 360, at 12–18, 20–23; Walker, supra note 360, at 25 (1991); Scanlon, supra note 355, at 126–28.
 In a different context, Tim O’Brien wrote that he “detested” the coercion imposed by the society that pushed him into Vietnam and “held them personally and individually responsible” for the decision, Tim O’Brien, The Things They Carried 49 (1990), and yet that he felt the profoundest “self-hatred and self-betrayal” for succumbing to that coercion. Tim O’Brien, The Vietnam in Me, N.Y. Times Mag. (Oct. 2, 1994), https://perma.cc/VE76-43HV.
 Cf. Dannenbaum, supra note 142, at 131, 254–55.
 See De Waal, supra note 349.
 See supra notes 296, 333 and accompanying text.
 Cf. supra Section III.F.
 Riordan, supra note 108, at 165.
 See Prosecutor v. Dragomir Milošević, Case No. IT-98-29/1-T, Judgment, ¶ 910 (Int’l Crim. Trib. for the Former Yugoslavia Dec. 12, 2007).
 In Sarajevo, the two methods operated in conjunction, even though it was the terror sniping that underpinned criminal accountability. Dragomir Milošević, Case No. IT-98-29/1-T, Judgment at ¶ 910 (Int’l Crim. Trib. For the Former Yugoslavia Dec. 12, 2007).
 Rape in genocide has been recognized as a condition of life designed to bring about its destruction, and in that form has been equated with “the starving of a group of people, reducing required medical services below a minimum, and withholding sufficient living accommodation for a reasonable period.” Prosecutor v. Kayishema, Case No. ICTR-95-1-T, Judgment, ¶ 116 (May 21, 1999). See also IIFFMM, Detailed Findings (2018), supra note 8, ¶ 1406.
 See Prosecutor v. Delalić, Case No. IT-96-21-T, Judgment, ¶¶ 494–96 (Int’l Crim. Trib. for the Former Yugoslavia Nov. 16, 1998).
 Sussman, supra note 270, at 4.
 See supra notes 172–78 and accompanying text.
 Rome Statute, supra note 1, arts. 8(2)(a)(v, viii), (b)(xv, xxvi), (c)(iii), (e)(vii).
 See supra note 212 and accompanying text.
 See supra Section IV.B.
 See Dannenbaum, supra note 327, at 368–69.
 Coco, Hemptinne & Lander, supra note 212, at 913.
 See supra note 186.
 Cf. supra notes 208–11.
 See Conley & de Waal, supra note 4, at 700–01.
 See Prosecutor v. Delalić, Case No. IT-96-21-A, Judgment, ¶¶ 412–13 (Int’l Crim. Trib. for the Former Yugoslavia Feb. 20, 2001); Prosecutor v. Musema, Case No. ICTR-96-13-A, Judgment, ¶¶ 361–63 (Nov. 16, 2001); Prosecutor v. Ntaganda, Case No. ICC‑01/04-02/06-2666-Red, Appeal Judgment, ¶ 1202 (Mar. 30, 2021).
 International criminal trial chambers have significant discretion to determine the appropriate sentence in any given case in light of the “totality of the criminal conduct and overall culpability of the offender.” Delalić, Case No. IT-96-21-A, Judgment at ¶ 430 (Int’l Crim. Trib. for the Former Yugoslavia Feb. 20, 2001). At the ICC, this includes “[b]alanc[ing] all the relevant factors, including any mitigating and aggravating factors, and consider[ing] the circumstances both of the convicted person and of the crime.” International Criminal Court, Rules of Procedure and Evidence, supra note 180, r. 145(b).
 See Carl-Friedrich Stuckenberg, Cumulative Charges and Cumulative Convictions, in The Law and Practice of the International Criminal Court 840, 851 (Carsten Stahn ed., 2015). See, e.g., Ntaganda, Case No. ICC-01/04-02/06-2667-Red, Public redacted version of Judgment on the appeal of Mr Bosco Ntaganda against the decision of Trial Chamber VI of 7 November 2019 entitled ‘Sentencing judgment’ at ¶¶ 133–40 (Mar. 30, 2021).
 Ntaganda, Case No. ICC-01/04-02/06, Trial Judgment at ¶ 1202; Ntaganda, Case No. ICC-01/04-02/06, Sentencing Decision at ¶¶ 31, 249; Ntaganda, Case No. ICC-01/04-02/06-2667-Red, Public redacted version of Judgment on the appeal of Mr. Bosco Ntaganda against the decision of Trial Chamber VI of 7 November 2019 entitled ‘Sentencing judgment’ at ¶¶ 134–36, 139.
 See Stuckenberg, supra note 385, 844 (“if all provisions are applicable, then cumulative convictions are necessary in order to adequately express what the accused did. It is a separate question how to determine the sentences based on these multiple convictions, especially how to calculate the total amount of punishment.”).
 See supra notes 208–11 and accompanying text.
 See De Waal, supra note 4, at 30–31, 33.
 See Prosecutor v. Kunarac, Case Nos. IT-96-23 & IT- 96-23/1-A, Judgment, ¶¶ 150–51 (Int’l Crim. Trib. for the Former Yugoslavia June 12, 2002).
 IIFFMM Detailed Findings (2019), supra note 8, ¶ 174.
 For a view advocating for the use of other inhumane acts category in combination with extermination, see DeFalco, supra note 66, at 1176.
 Rome Statute, supra note 1, art. 7(1)(k).
 See supra notes 202–11 and accompanying text.
 On the likelihood this could be established, see Ventura, supra note 75, at 792. On the validity of similarly presuming severe pain or suffering to arise from rape (thus satisfying the consequence element of torture), see supra note 390.
 Hum. Rts. Council, Rep. of the Comm. of Inquiry on Human Rights in the Democratic People’s Republic of Korea, ¶ 76, U.N. Doc A/HRC/25/63 (Feb. 7, 2014).
 The Public Prosecutor in the High Iraqi Court v. Hussein, Judgment, Case No. 1/C1/2005, 45–46 (Nov. 4, 2005).
 See supra notes 176–78 and accompanying text.
 See Dannenbaum, supra note 64.
 See generally Margaret M. deGuzman, Choosing to Prosecute, 33 Mich. J. Int’l L. 265 (2012).
 Sander, supra note 149, at 866.
 Id. at 867.