The 21st-Century Belligerent’s Trilemma

Abstract

This article introduces three ways in which a state at war can attempt to accommodate the often contradictory demands of military necessity and humanitarianism – three ‘logics’ of waging war. The logics of sufficiency, efficiency and moral liability differently distribute the harm and destruction that waging war inevitably causes. International law demands belligerents follow the logic of sufficiency. Contemporary strategic imperatives, to the contrary, put a premium on waging war efficiently. Cross-culturally shared expectations of proper state conduct, however, mean killing in war ought to fit the logic of moral liability. The latter proves entirely impractic able. Hence, a belligerent faces a choice: (i) renounce the right and capacity to use large-scale collective force in order to meet public expectations of morally appropriate state conduct (logic of liability); (ii) defy those expectations as well as international law and follow strategic im peratives (logic of efficiency) and (iii) follow international law (logic of sufficiency), which is inefficient and will be perceived as illegitimate. This is the 21st-century belligerent’s trilemma.

1 Introduction

A state in war faces at every turn the overwhelming demands of military necessity. Not following them might give the adversary the decisive edge in the struggle for military victory. At the same time, very few states that end up waging war against another state will altogether fail to also perceive an imperative to protect human life.¹ After all, they are members of an international society in which the violation of human rights provides grounds for criticism and often entails reputational costs. Of course, there is no reason to assume that acting on military imperatives always involves sacrificing humanitarian goals. Yet, it is a non-contingent reality of war that it regularly does. How can a belligerent square the circle between the desire to win a war, on the one hand, and the interest in meeting widely shared normative expectations of legitimate state conduct, on the other hand?

This article introduces three ways in which a belligerent can attempt to accommodate these opposing imperatives – three logics of waging war. What shall be called the logics of sufficiency, efficiency and moral liability differently distribute the harm and destruction all wars inevitably inflict on a belligerent society. The article demonstrates that a contextual interpretation of international humanitarian law (IHL) demands belligerents follow the logic of sufficiency. The 21st-century battlefield, to the contrary, appears to put a premium on waging war in accordance with the logic of efficiency. Yet, neither combat operations that follow efficiency considerations nor hostilities conducted in accordance with the strictures of sufficiency meet with public expectations of legitimate state conduct. Cross-culturally shared beliefs about the value of human life require destruction and killing in war to fit the logic of moral liability. However, it proves impossible to wage war while accounting for the moral liability of individuals on the opposing side. Rather than accommodating both humanitarian and military imperatives, the logic of moral liability amounts to a de facto prohibition on the use of force. The article discusses the implications of the observation that law (logic of sufficiency), strategy (logic of efficiency) and legitimacy (logic of moral liability) make diverging demands on states in war. The article focuses in particular on what it means that obedience to international law neither allows a belligerent to pursue military victory in an efficient way nor does it ensure the legitimacy of conduct in war.

2 International Law and Sufficiency

‘War is about killing people and breaking things.’² Distinguishing people that belligerents are allowed to kill from those who are immune from attack and things that belligerents may break from those that are to be left intact is at the heart of the subjection of warfare to legal regulation.³ This section will establish the logic behind the distribution of harm in war that international law envisages. The First Additional Protocol to the Geneva Conventions contains the most recent promulgation of what and who is a legitimate target of attack.⁴ Article 48 enjoins belligerents to ‘direct … operations only against military objectives’. Persons that are combatants in the meaning of Article 43 are military objectives. According to Article 52(2), as far as objects are concerned, military objectives are those ‘which by their nature, location, purpose or use make an effective contribution to the military action and whose partial or total destruction, capture or neutralisation in the circumstances ruling at the time offers a definite military advantage’. Two criteria – an ‘effective contribution to military action’ and a ‘definite military advantage’ – hence determine whether an object can be reckoned a military objective.

In other words, it is the connection of an object to the conduct of combat operations – those of the enemy belligerent (effective contribution) and one’s own (military advantage) – that puts an object into the category of military objectives.⁵ But how close must this connection be? What is the minimum degree of nexus between an object and the enemy’s hostile actions for the object to count as a military objective? We could interpret Article 52(2) to allow the engagement of only those objects that contribute to the enemy’s military effort, meaning the direct, mostly kinetic engagement of enemy forces. Alternatively, a connection to the war effort more broadly could be deemed sufficient.⁶ The latter interpretation creates a wider pool of ‘things to break’ on the opposing side.

Not surprisingly, the degree of nexus between an object and the adversary’s military action often determines the degree of nexus between the attack and the military advantage arising from the engagement of this object.⁷ Can we solve the interpretive question raised earlier by establishing the required degree of nexus between an attack and the military advantage? The text says that the advantage has to be ‘definite’. The ordinary meaning of definite could be tangible, visible or palpable, terms that allude to the existence of an advantage, possibly the likelihood of its emergence. Alternatively, definite could denote precise, determinate, distinct or unequivocal. These words refer to, as it were, the sharpness of the contours of the advantage. None of the synonyms of ‘definite’ describe the advantage’s connection to the attack as designations such as direct, immediate, prompt or instant would.

Claude Pilloud and his colleagues in their commentary on the Protocol require the advantage to be ‘substantial and relatively close’.⁸ This requirement suggests there is a limit to the permissible distance between the advantage and the attack, but ‘relatively close’ leaves room for interpretation. Michael Bothe, Karl Partsch and Waldemar Solf consider ‘definite’ to mean that the military advantage must be ‘concrete and perceptible … rather than … hypothetical’.⁹ Perceptible is another way of describing the quality of the required advantage. ‘Concrete’ and ‘not hypothetical’ could refer either to the latter or to the connection between the attack and the advantage. If concrete and not hypothetical are attributes of the connection between the attack and the advantage, they likewise rule out that any remote connection can bring an object under the definition of military objectives. However, like Pilloud and his colleagues, Bothe, Partsch and Solf do not positively specify a required degree of nexus. Scholars disagree on whether an indirect advantage arising from an attack renders the object in question fair game.¹⁰ The interpretation of both criteria that define a military objective is hence beset by the same interpretive controversy.

For the application of the law, does it matter that different interpretations prevail regarding the minimum connection between objects and military operations? It does. As indicated, the lower the minimum required degree of nexus, the broader the category of military objective. An example of an object, whose contribution to the military effort is vital, yet indirect, is the food supplier of an enemy belligerent.¹¹ As soldiers need to eat, food suppliers quite literally sustain the adversary’s war effort. By the same token, their engagement ultimately generates a military advantage because hungry forces are less militarily effective. However, this military advantage is not a direct result of the attack. It is two, rather than one, causal steps away from the destruction of the object in question. The result of the attack is that the business is in ruins and food availability decreases: first causal step – soldiers get hungry; second causal step – military effectiveness declines. In turn, the food supplying industry is doubtlessly part of a society’s war effort. Yet, it is two causal steps removed from the enemy’s military effort, meaning the engagement of the enemy belligerent in hostilities. Compare this to an attack on an object more directly related to combat operations, for instance, a power plant producing the energy supply for, inter alia, the opposing armed forces. Contrary to the food industry, power plants generate an output that provides something soldiers need to fight rather than ‘merely’ to live.¹² As a result, the decrease in military effect iveness directly follows from their destruction. It is widely accepted that power plants used by the armed forces are military objectives. Whether food suppliers that service the military can be reckoned prima facie legitimate targets is controversial.¹³

Modern industrialized societies heavily rely on objects the engagement of which potentially yields a significant military advantage, but only in more than one causal step. Two causal steps separate a decrease in military effectiveness from attacks on non-military industry, businesses or other objects used for a taxable economic activity. In Step 1, such an attack decreases the financial resources of the state and, in Step 2, its capacity to, for instance, procure weapons. Three or more causal steps separate a decrease in military effectiveness from attacks on symbolically important sites, the political apparatus of a state at war, infrastructure only used by civilians, and communication links between the government and the civilian population. In Step 1, attacks on such ‘message targets’¹⁴ may ‘inconvenience’¹⁵ civilians. In Step 2, civilians might rethink their contribution to the war effort, communicate disaffection or worry to their relatives ‘at the front’, or withdraw their support from their warmongering political leaders. Only in Step 3 or even further down the line might a decrease in military effectiveness ensue.

Closely related to the interpretive controversy about the minimally required connection between an object and combat operations is the question of how belligerents should define progress during hostilities. The point of reference used to determine a military advantage could be the destruction of one object – a larger, but discrete, step in the process of overcoming the adversary’s military forces – or victory as such. Examples of a point of reference that is more than one attack, but not overall victory, are the capture of a strategically important area of enemy territory or the destruction of the adversary’s air defence system. The interpretive statement for Article 52(2) introduced by the United Kingdom on the occasion of negotiations to the First Additional Protocol avers that the definite military advantage ‘is intended to refer to the advantage anticipated from the attack considered as a whole and not only from isolated or particular parts of the attack’.¹⁶ The British position has been often reproduced and is widely endorsed.¹⁷ Most commentators thus agree that it does not have to be a single air strike or artillery barrage that provides the advantage.

Yet, whether it has to be a discrete step in the pursuit of victory or overall victory that is the most general allowable point of reference for the determination of progress in war is contested. The interpretation that ‘“[m]ilitary advantage” is not restricted to tactical gains, but is linked to the full context of one’s war strategy’¹⁸ conflicts with one commentary’s understanding that ‘an attack as a whole is a finite event, not to be confused with an entire war’.¹⁹ What if the point of reference for the definition of progress in combat was the ‘full context of one’s war strategy’ – that is, overall victory? It is commonplace that states wage war for political reasons rather than as an end in itself. The question would hence arise whether Article 52(2) should be interpreted in light of the desired political end-state a belligerent seeks in war. Advantage is a relational concept, and in order to attribute meaning to the notion of a relative gain over another actor, we need to know what it is we ultimately seek to gain: victory informed by the specific political goals of a war or victory on the battlefield that then needs to be translated into political outcomes?²⁰

That the textual interpretation of positive IHL remains inconclusive on this point has important implications.²¹ If political goals of a war serve as the point of reference for defining progress, different categories of objects count as military objectives in different wars. For instance, in 2003, during Operation Iraqi Freedom, 1,799 desired mean points of impact (targets) served the strategy-to-task mission referred to as ‘the suppression of the regime’s ability to command Iraqi forces and govern the State’.²² The interpretation of military advantage with a view to the goal of undermining the regime’s ability, not only to command its forces but also to govern the state, may have brought Baath party headquarters and information links between the government and the public, such as media facilities, into the remit of legitimate targets.²³ Roughly 9 per cent of all air strikes, the second largest category, were thus chosen in light of the political goal of regime change. Alternatively, during Operation Allied Force, in 1999, NATO practised so called ‘boutique bombing’.²⁴ Attacking civilian industrial and infrastructural targets to interfere with the lives of regime cronies and the larger population was meant to provide an advantage in light of the political goal of pressuring Slobodan Milošević to return to the negotiation table. The connection of these air strikes to the military goal of overcoming Serbian military forces was remote. As an instance of message targeting, boutique bombing also provides examples of attacks that only hypothetically and in three or more causal steps impact military effectiveness.²⁵

Could an interpretation of the text animated by the purpose of the First Additional Protocol not shed light on the required point of reference for the definition of pro gress in war? The Hague Conventions and the Geneva Conventions have attempted to further military pragmatism and humanitarian goals in combat operations respect ively.²⁶ The object and purpose of the First Additional Protocol, in contrast, is split between these two imperatives. It is to prescribe rules for the conduct of hostilities that render warfare as humane as possible given military pragmatism and as militarily expeditious as possible given humanitarian goals.²⁷ Military pragmatism means law cannot make warfare impossible.²⁸ Humanitarianism means law cannot allow more death and destruction than necessary for war to be possible. It follows that IHL must allow no more and no less violence than is sufficient; but sufficient for what exactly? While a purposive interpretation of the First Additional Protocol thus means returning to the question of the right point of reference for defining progress in war, it is the split object and purpose of the Protocol that provides an important insight into the logic that the Protocol envisages combat operations follow. In order to do justice to both humanitarianism and military pragmatism, contemporary IHL permits no more and no less violence than is sufficient. The only question is, sufficient for what?

It is a systematic or contextual interpretation of Article 52(2) that establishes with regard to what sufficiency has to obtain. The First Additional Protocol, as the first international treaty to regulate the conduct of hostilities in light of the prohibition on the use of force, does not allow appeal to the notion of a causa justa. IHL (jus in bello) is independent from the reasons for resort to force and their legality (jus ad bellum). The Protocol’s preamble unequivocally spells out that its provisions must be applied ‘without any adverse distinction based on the nature or origin of the armed conflict or on the causes espoused by or attributed to the Parties to the conflict’. Law that does not allow any appeal to the causes for which a war is fought in guiding the conduct of hostilities must work from the assumption that there is a stable (if very abstract) concept of military victory that is valid across most wars, notwithstanding their different moral and political contexts. Belligerents then have to achieve their political goals via the advantages that appeal only to such a ‘generic’ military victory, rather than directly to the political or moral reasons for which they are ultimately fighting – for instance, regime change. It follows that it is generic military victory that represents the most general permissible point of reference for the determination of a military advantage according to a contextual interpretation.

In other words, the First Additional Protocol commands belligerents to sharply distinguish between ultimate goals (political or other) and intermediate goals (military). While hostilities are ongoing, belligerents have to bracket their larger political, moral or other aspirations when devising how to act, namely what to attack. I call this action the ‘command of sequencing’ – sequencing the use of force and the pursuit of politics. Underlying this command is the assertion that military victory is sufficient to allow states to subsequently achieve their legitimate political or other goals. But is this true? The legal prescription to use force only with a view to attaining generic military victory rules out war as an effective instrument for the achievement of all those political goals for which a generic military victory is not actually a sufficient condition. It is crucial here to note that the current international legal order rests on a presumption against the use of force as a continuation of politics by other means. For this interpretation to be coherent, if states had a right to use force as a regular expression of sovereign statecraft, we would have to test whether generic military victory is in fact sufficient for the achievement of what are considered just causes. There are no so-called just causes.²⁹ The assertion that generic military victory is sufficient is thus not an empirical observation or assumption. It is an expression of the fact that jus in bello, its self-contained nature notwithstanding, shares in the mission of general international law to limit the usefulness of force in international relations.

The injunction that generic military victory is all that belligerents are allowed to strive for elucidates what kind of violence can be considered sufficient and thus legal. Only one side can win every war. So law does not simply allow all violence that is sufficient for the achievement of even just military victory. It permits violence that is necessary and sufficient for a competition between enemy militaries to proceed. Crucially, this competition has to be geared towards generic military victory only. As a result, it is sufficient to attack objects directly connected to such a competition. After all, a competition between enemy militaries in which one side will ultimately prevail militarily does not have to involve more than the objects and persons directly involved in it. In turn, being allowed to attack those objects and persons whose engagement immediately provides a genuinely military advantage is necessary for the competition between two militaries to proceed and a chance of generic military victory for one side to exist. The First Additional Protocol hence commands belligerents to sharply distinguish objects and persons closely connected to the competition between enemy militaries from everyone and everything else.³⁰ A systematic interpretation of Article 52(2) then requires that no more than one causal step separates an object from military action and an attack from the resulting advantage in the quest for generic military victory. I call this action the ‘command of containment’ – the containment of hostilities.

What does a war fought in accordance with the logic of sufficiency look like? The commands of sequencing and containment mean hostilities will have to focus on objects instrumental in the actual competition between enemy militaries, such as weapons, barracks, military transport and military industry. These are all military objectives by nature and have no, or only marginal, civilian functions. Combat operations may be directed against objects that have both a civilian and a military function in virtue of their purpose, location or use. However, according to the logic of sufficiency, the intent of attacks on what is generally referred to as dual-use objects must be to neutralize only their military function. The foreseeable disruption of objects’ civilian function must count as expected collateral damage, not as military advantage. Finally, combat operations following the logic of sufficiency will not involve objects that do not have a direct (one causal step) connection to the fight. Examples include political infrastructure, symbolic sights, non-military industrial production and communication infrastructure not used for command and control but merely for propaganda. Regardless of the specific political goals of a war, the logic of sufficiency translates into attrition warfare – the attempt to win a war by overcoming an enemy military.³¹

Waging war in this way is highly inefficient. First, the logic of sufficiency precludes the efficient (direct, quick and cheap) pursuit of the belligerents’ ultimate political goals. The goal over which other factors can be minimized is military progress narrowly defined. To be specific, in a war with limited goals, such as Operation Allied Force, sequencing the conduct of war and the pursuit of those goals by focusing on all-out attrition warfare might seem wasteful in terms of time, blood and treasure. Moreover, even generic military victory may only be achieved with the engagement of objects that in one causal step contribute to the competition among opposing militaries. This conclusion remains true even if the attack on other objects promises to contribute to ending the war sooner by generating a political or psychological advantage but only an indirect military advantage (ending the war). Message targeting and attacks on the political fabric of a state and on non-military industry are prohibited by the command of containment. The First Additional Protocol regulates warfare through sequencing and containment, both of which defy the efficient achievement of political goals with force.

3 Strategic Imperatives and Efficiency

Implicit in the above outline of the logic of sufficiency is an alternative way in which to accommodate both military and humanitarian imperatives in the conduct of war as well as an alternative way in which to distribute harm. A logic of efficiency would eschew sequencing the pursuit of military and political victory and demand belligerents choose targets with a view to gaining an advantage in the pursuit of their ultimate overall – often political – goals. In its most radical form, a logic of efficiency would also reject the containment of hostilities and recommend belligerents target exactly those objects and persons – be they civilian or military – whose attack promises the quickest political victory. A moderate version of the logic of efficiency might not abandon the legal obligation to distinguish altogether. Rather, it could broaden the category of military objectives, and, thus, of prima facie legitimate targets, by allowing for a lower degree of nexus to connect objects to military operations.

Over the last two decades, military doctrine, specifically among NATO countries, has evolved to ever more closely reflect such a moderate version of the logic of efficiency, which rejects sequencing and relaxes containment.³² The introduction and rise to prominence of effects-based operations (EBO) in Western militaries is a clear indicator of this trend. The doctrine of EBO explicitly demands efficiency in war. The prescription is that mission accomplishment should be ‘sought while minimizing cost in lives, treasure, time, and/or opportunities’,³³ seeking ‘to achieve objectives most effectively, then most efficiently’.³⁴ The following paragraphs show that EBO challenges both the sequencing and the containment command of the logic of sufficiency and the First Additional Protocol.

In defiance of sequencing, EBO is most successfully executed if belligerents achieve their ultimate political goal while not having to destroy the enemy’s military forces.³⁵ The doctrine recommends that those targets are to be selected that ‘contribute directly to the achievement of strategic objectives’.³⁶ ‘Strategic’ is defined as ‘the highest level of an enemy system that, if affected, will contribute most directly to the achievement of our national security objectives’.³⁷ Hence, the doctrine advocates choosing objects as targets that are linked not to generic military victory but, rather, to the specific strategic [read political] goals of a war. Accordingly, ‘offensive action [is allowed and welcomed] against a target – whether [it is] military, political, economic, or other’.³⁸ Manuals contrast effects-based targeting with attrition warfare, which is shunned for its lack of effectiveness.³⁹ Rather than limiting combat operations as much as possible to the competition between opposing militaries, as the logic of sufficiency does, effects-based targeting avoids this competition as much as possible.

In defiance of containment, the doctrine urges commanders to ‘consider all possible types of effects’ when selecting targets.⁴⁰ It advocates producing ‘political effects beyond the mere destruction of those targets’⁴¹ because those indirect effects are considered to be often more important than the immediate kinetic results of an attack.⁴² In other words, an effects-based approach does not follow the sharp distinction between military objectives narrowly defined and the rest of a belligerent society as prescribed by the logic of sufficiency. Military manuals explicitly credit effects-based thinking for inspiring the engagement of objects other than ‘“traditional” wartime targets’.⁴³ The prescriptions to avoid the engagement of enemy military forces and to strive for other than kinetic effects is taken even further by the doctrine of ‘achieving rapid dominance’, colloquially known as ‘shock and awe’. The latter identifies the civilian population as the most promising object of psychological warfare.⁴⁴

The rise to prominence of military doctrine that challenges the logic that warfare ought to follow as envisaged by the First Additional Protocol has not left the interpretation of international law unaffected. In many countries, IHL exegesis and the conception of military doctrine are in different hands. In the USA, however, so-called operational law and military doctrine are devised by the same bureaucracy, the Department of Defense.⁴⁵ In addition, shock and awe and EBO both originated in the USA, and it is the US military, specifically the air force, that has most explicitly turned its back on what I refer to as the logic of sufficiency. Not surprisingly then, in the development of US operational law, we find a neat imprint of the shift from sufficiency to efficiency.

Although it did not ratify the First Additional Protocol, the USA adopted the language of Article 52(2) into operational law manuals.⁴⁶ This move signalled the USA’s acceptance of the Protocol’s addition of a definition of military objectives to the customary obligation to distinguish binding general international law. Yet, two decades later, military manuals started to feature a change in the wording of the definition of military objectives. In the 1997 field manual on the joint targeting process, the attribute ‘war-sustaining’⁴⁷ emerged as a criterion for mission assessment.⁴⁸ The Joint Doctrine for Targeting of 2002 used the term to explain the definition of a military objective.⁴⁹ The term entered this definition in Military Commission Instruction No. 2 of 2003. The document defines military objectives as those objects that ‘effectively contribute to the opposing force’s war-fighting or war-sustaining capability’, as opposed to the original criterion of ‘an effective contribution to military action’.⁵⁰ According to the new formulation, a link to military action properly so-called is no longer the only way an object can become a military objective. Another way is to contribute to an enemy’s ‘war-sustaining capability’. The military advantage that may ultimately arise from attacks on objects that can conceivably be construed as sustaining a belligerent’s capabilities to wage war can have a very low degree of nexus to the attack itself.⁵¹ The discussion in the previous section showed that anything from the civilian political apparatus to business activities and morale sustains the war in some way.

The shift towards a lower degree of nexus is also visible in the yearly Operational Law Handbooks issued by the Judge Advocate General School. Since 2004, the handbooks have started the definition of military objectives with a verbatim repetition of Article 52(2) and have then elaborated: ‘The connection of some objects to an enemy’s war fighting or war-sustaining effort may be direct, indirect, or even discrete. A decision as to classification of an object as a military objective ... is dependent upon its value to an enemy nation’s war fighting or war-sustaining effort (including its ability to be converted to a more direct connection), and not solely to its overt or present connection or use.’⁵² The criterion of having value to an enemy’s war-sustaining effort is thereby framed as a faithful interpretation of the uncontroversial formulation defining a military objective according to Article 52(2).

Whether or not this is the case is controversial. Prominent military expert, Hays Parks, in a public lecture at Chatham House, referred to the issue as ‘more of an intellectual argument between various semantic alternatives which does not make a real practical difference’.⁵³ Michael Schmitt likewise maintains that the use of the attribute war-sustaining as a feature defining military objectives has not led to problematic targeting choices.⁵⁴ The International Humanitarian Law Research Initiative abstained from using the term in its 2009 Manual on Air and Missile Warfare. The commentary describes it as ‘a matter of dispute’ whether ‘the definition includes objects which indirectly yet effectively support military operations’. The Initiative goes on to reject the argument that objects that merely sustain the war should fall under the definition contained in Article 52(2).⁵⁵ Yoram Dinstein considers ‘war-fighting’ capability largely synonymous with military action, but, like Frits Kalshoven, he rejects ‘war-sustaining’ as being too broad.⁵⁶

Admiral Horace Robertson is a lone voice in arguing that the change in language presents a deliberate challenge to the customary status quo: ‘[T]he inference that one may draw from this change in wording is that the United States ... has rejected the presumptively narrower definition contained in Article 52 of Protocol Additional I in favour of one that, at least arguably, encompasses a broader range of objects and products.’⁵⁷ Corroborating the view that the criterion ‘sustaining a war effort’ creates a more inclusive category of military objectives than the Protocol’s condition that the object ‘makes an effective contribution to military action’, the Operational Law Handbooks of 2006–2008 list, without qualification, ‘(1) Power (2), Industry (war supporting manufacturing/export/import), (3) Transportation’⁵⁸ as military object ives. None of the objects that fall into these three categories are military objectives by nature. The First Additional Protocol therefore requires that a link to military action via their location, purpose or use be established before they can legally be attacked.

The broadest conception of military objectives in US doctrine, however, is implied by the permission to intentionally attack ‘objects that contribute to an opposing state’s ability to wage war’, which has been featured in the Operational Law Handbooks since 2010.⁵⁹ Ultimately, the entire civilian infrastructure can be considered to contribute to a state’s ability to wage war. The statement includes no specification of the required nexus between an object and the actual military effort. The same operational law manuals acknowledge that the USA defines ‘“definite military advantage” very broadly’.⁶⁰ A preventive argument against critics is implicit in the statement that ‘[s]tates may come to different conclusions regarding whether certain objects are military objectives in accordance’ with Article 52(2).⁶¹ It is certainly safe to conclude that US operational law allows for a lower degree of nexus between an object and the enemy’s military action – and, hence, between an attack and the resulting military advantage – than the First Additional Protocol, which demands containment of hostilities within one causal step.

This brief overview of a recent trend in military doctrine and of the development specif ically of US operational law suggests that what seems like an outright rejection of distinction in military strategy corresponds instead to a lowering of the required degree of nexus between objects and military operations in operational law. The USA has not turned its back on distinction, but it does no longer recognize the requirement of a direct causal connection between an object and the competition among enemy militaries.⁶² It follows that the ultimate point of reference for the definition of the first order effects of an attack as progress in war is not necessarily generic military victory. Military strategy openly defies the command of sequencing when it prescribes that belligerents endeavour to attain their desired political end state as directly as possible. Operational law has abandoned it by implication.

In order to shed light on why this alternative logic of efficiency has emerged and risen to prominence in military and operational legal doctrine over the last two decades, an enquiry into the factors that may have altered warfare at the turn of the 20th to the 21st century is necessary.⁶³ The most obvious impulses for changes to the character of war come from progress in technology. A large body of scholarship explores the effects of increasingly precise weapons and improved capabilities for reconnaissance, surveillance and intelligence gathering (RSI) on combat operations.⁶⁴ Often overlooked is the question of how these alterations to the material means of waging war interact with moral and legal imperatives. It is the progress in technology that enhances RSI that has also improved telecommunication and that affords the international public an ever better view onto the battlefield. It is not immediately obvious why an international public opinion⁶⁵ should prefer that wars be waged according to the logic of efficiency though. What it means for the conduct of war that the public is watching requires an exploration of shared ideas about violence in international relations.

In order to explain the shift in logics, the considered judgments on war of the international society would have to have changed. The first tangible result of such a change in the shared beliefs about violence in international relations in the 20th century was arguably the formal legal prohibition on the use of force in 1945. While outlawing war had been a project in pacifist and feminist circles as well as among liberal politicians and scholars for a while, it was only after the two World Wars that circumstances aligned to make it a reality. Spurred by the human suffering just witnessed, a majority of actors in the international community, including the major powers of the moment, were able to reach consensus on the establishment of a legal obstacle to the resort to force in international relations.⁶⁶ Over the following decades, this ideational change entailed the creation, elaboration and increased justiciability of a global human rights regime under international law. The end of the Cold War, moreover, inspired a flurry of United Nations (UN) peacekeeping missions to ease the consequences of conflict and rebuild war-torn societies. The post-Cold War order also saw the revival of international criminal law. Martha Finnemore and Kathryn Sikkink have diagnosed international relations with ‘a long-term trend toward humanizing the “other,” or “moral progress”’.⁶⁷ What is alternatively called the ‘individualisation’⁶⁸ of international relations or the ‘humanisation’⁶⁹ of general international law arguably reached a peak with the tentative qualification of state sovereignty for the sake of an emerging right to humanitarian intervention towards the end of the 20th century.⁷⁰ At the heart of this ideational change lies what Theodor Meron calls the ‘advent of a general distaste for the waste of human life.’⁷¹ International relations at the turn of the century were characterized by the increasingly cross-cultural scope of physical integrity right.⁷²

This ideational change, namely the rise of an individual rights-based moral standard in international relations, explains the rising and now notorious casualty aversion specifically among Western societies. There seems almost universal agreement among scholars of international relations that ‘Western societies can now only fight wars which minimise human suffering.’⁷³ The ‘zeitgeist that requires the reduction of human risk’⁷⁴ has been most thoroughly documented to prevail in US society.⁷⁵ A project analysing various opinion polls discovered that the American public, as we might expect, displays a double standard in that losses among American troops are felt more keenly than fatalities among civilians in the country under attack or enemy combatants.⁷⁶ Crucially, however, the plight of civilians in the theatres of US military operations abroad receives some media attention in the USA these days and by no means is met with complete indifference.⁷⁷

That deaths of enemy combatants can even entail reputational costs for the belligerent causing them became evident when on 2 March 1991, shortly before the end of Operation Desert Storm, the USA killed scores of Iraqi troops on Highway 80 between Kuwait and Basra. The legality of the attacks depended on whether the Iraqi troops were withdrawing (illegal) or merely retreating (legal) and on whether they had violated a prior cease fire agreement as the commanding general had alleged.⁷⁸ These subtleties were of little importance to the media, which broadcast gruelling images of smoking tanks and charred bodies. While the resulting indignation about the waste of human life was an international phenomenon, critical voices were heard also in the USA. If we took the USA to be representative of a Western belligerent with the strategic offensive, we could unequivocally state that in such a society at war loss of life in none of the three groups – friendly combatants, civilians and opposing combatants – is simply considered unproblematic nowadays.

In third countries not involved in a war, civilian casualties tend to generate the most abhorrence, more so than combatant deaths on either side. After all, troop fatalities are rarely individually reported outside the dead combatants’ countries of origin. The so-called ‘Highway of Death’ incident in Iraq was an exception in this respect. It evidenced, however, that even combatants’ lives are no longer considered dispensable and their large-scale killing does not go unnoticed or uncriticized.

Coming back to the two logics, the normative imperative of saving lives provides an explanation for the perceived strategic imperative of efficiency. In both logics, civilians are protected from direct attack but open to being incidentally harmed. The logic of sufficiency makes no claim at all to protecting combatants’ lives. In fact, it endows killing combatants with the legitimacy of a legally privileged course of action, an acceptable way to attain a military advantage. The logic of efficiency does, of course, likewise involve deliberate attacks on combatants. However, it lays a claim to the protection of human life that the logic of sufficiency does not make: minimizing combatant and possibly even civilian losses by getting a war over with as quickly as possible. At first sight, waging war according to the logic of efficiency then enjoys an advantage over the logic of sufficiency in the eyes of a belligerent concerned with being perceived as legitimate.⁷⁹ The fewer air strikes it takes to end a war, the less likely it is to receive criticism from a global public that is scandalized by the loss of life (if it is reported) and from the public at home, which is anxious to protect the troops.

It is plausible that casualty aversion means contemporary military operations are kept as brief as possible and other than ‘traditional’ military objectives are targeted to that end (defiance of containment). It is less clear why efficiency should be sought with regard to the achievement of the overall political goals of a war (defiance of sequencing). Human suffering, at least the kind directly inflicted by combat operations, presumably ends with one side achieving generic military victory.⁸⁰ Why then focus on the desired political end state, as EBO and shock and awe prescribe? For an explanation, we have to return to the prohibition on the use of force. This legal watershed rendered war a state of exception in need of a legal and political apology. If something is prima facie prohibited, unless it pursues one of a very small number of specific legitimizing goals, those goals are naturally crucial – the more so, the keener a belligerent is on upholding an aura of legitimacy.

The only legal apology available to a state unilaterally resorting to force is self-defence. However, over the last seventy-odd years, the exercise of self-defence very rarely has meant the straightforward defeat of an invading army. It is often a specific political end state that affords the neutralization of the threat that has triggered the perceived need to resort to force from the point of view of the ‘defender’.⁸¹ Achieving this political end state, which grounds a state’s inherent right to self-defence, exceptionally removes the stigma of aggression from the use of force.⁸² It is only natural that this source of legitimacy is kept at the forefront of combat operations. Every air strike that inflicts civilian casualties and/or puts troops at risk ought to also contribute to the achievement of the exceptional goals for which a state claims to rightfully deviate from the Grundnorm of peaceful international relations.

In this reading, the removal of war from the spectrum of regular international politics has paradoxically encouraged appeal in the conduct of combat operations to political aims. This explains why a belligerent who has an interest in avoiding the reputational costs of being branded an aggressor is tempted to lend expression to the importance of its overall goals with every attack. To sum up, it is an ideational change – the emergence of an individual rights-based morality and therefore an evolution of shared normative beliefs about violence in international relations – that drives those perceived strategic imperatives to which developments in military doctrine attempt to do justice and that a change in the interpretation of IHL accommodates.

4 Morality and Moral Liability

If ideational changes (casualty aversion and the internalization of the prohibition on the use of force) drive perceived strategic imperatives (quick, direct achievement of overall political goals), the question arises whether the logic of efficiency actually addresses the normative concerns that, I argue, have inspired its emergence in military doctrine and operational law.⁸³ In order to find out whether the logic of efficiency lessens the tension between 21st-century liberal moral values concerned with the individual and large-scale, collective state on state violence in war, we need to investigate what it looks like to conduct combat operations following the logic of efficiency.⁸⁴

The logic of sufficiency and efficiency can be respectively translated into the credos ‘contained wars are the least destructive’ and ‘sharp wars are brief’. Whereas the logic of sufficiency manifests on the battlefield as attrition warfare, belligerents following the logic of efficiency will avoid it. Instead, they will focus on dual-use targets, message targets, the adversary’s political infrastructure and other objects connected to the specific political end that the state has sought. It is analytically plausible that a war fought according to the logic of efficiency might be shorter, but that the air strikes that efficiency considerations demand – less attrition and more targeting of dual-use and civilian infrastructure – create a higher risk of collateral damage than ‘a similar’ war fought according to sufficiency considerations.⁸⁵ Yet, we do not know whether either of the logics pans out – that is, whether sharp wars are in fact brief enough to warrant their increased sharpness or whether wars fought under the strictures of sufficiency are contained enough to warrant their increased length. In efficient wars, does enough of a society remain intact once one side has achieved their political goals to make up for the fact that fewer objects are immune and that civilians are less protected from the start? In a contained war, does enough of a society remain immune from attack once sufficient objects and persons have been declared fair game to make up for the fact that this competition is likely to take longer?

It is not possible to establish whether a shift from sufficiency to efficiency saves lives or increases the casualty count. As a result, in order to gauge whether the logic of efficiency meets the expectation of legitimate state conduct, we have to focus on public reactions to instances of combat operations following efficiency considerations. Even more so than air strikes in accordance with EBO or shock and awe, targeted killings, often with unmanned aerial vehicles, reflect the strategic pressures of the 21st-century battlefield induced by changed shared ideas about violence in international relations. Targeted killings afford friendly forces absolute protection⁸⁶ and circumvent killing large numbers of enemy troops. In addition, that they directly channel violence towards an enemy framed as culpable, probably even deserving of lethal attack, makes what we colloquially refer to as ‘drone strikes’ the epitome of warfare, according to the logic of efficiency.⁸⁷ There is no more obvious way to keep the goal of self-defence at the forefront of violent action than to direct that violence against the individual who is the source of the supposed threat. Drone strikes heed the liberal demand that ‘criminals, not an impoverished nation, should be on the receiving end of punishment’.⁸⁸ At first sight, the logic of efficiency has therefore considerable appeal.

Yet, the identification of targeted killings as the purest implementation of the logic of efficiency should give us reason to pause. Protests against drone strikes in Pakistan match, if not exceed, those against ‘regular’ air strikes in Afghanistan.⁸⁹ In other parts of the world, targeted killings have galvanized political observers, non-governmental organizations and political activists.⁹⁰ The secrecy around who is attacked and the complete lack of international institutional oversight likely has played a role in explaining these reactions. Moreover, public outrage may be spurred on by the remote nature of the infliction of violence and by the challenge that drone strikes cause to the spatial limitation of wars. Yet the sheer intensity of the rejection of what is essentially an attempt to make individuals ‘rather than distinct collectives’ … the “target” in conflict’,⁹¹ warrants further exploration.

At the heart of many critical statements about the practice of targeted killing is concern about collateral damage. Specifically, the current US administration has gone to great lengths to foster the narrative of the surgical precision of drone strikes. Yet, there has been a steady trickle of information revealing that these attacks do cause incidental harm to civilians.⁹² However, do targeted killings cause more collateral damage than all-out attrition warfare against Pakistan, Yemen and Somalia would? If the latter is what we compare drone strikes to, the answer is likely no. In such a comparison, targeted killings save combatants’ as well as civilians’ lives. However, in the counterfactual scenario in which the option of remotely executed targeted killings is unavailable, these wars are extremely unlikely to take place, due to their material but also their reputational costs. Drone strikes lower the threshold for the use of force, precisely because they circumvent the hard-won normative obstacle to war in international relations that is a shared normative belief that combatants’ lives are not simply dispensable.

This threshold effect certainly preoccupies academics and possibly political and military commentators.⁹³ However, it is unlikely to be at the heart of wider public criticism. Marc Herold expresses a more visceral reaction to targeted killings when he writes that ‘[i]t is simply unacceptable for civilians to be slaughtered as a side-effect of an intentional strike against a specified target. ... Slaughter is slaughter. Killing civilians even if unintentional is criminal.’⁹⁴ Behind this abhorrence is not a comparison of the collateral damage that drones cause with the incidental harm associated with all-out war or with other reactions to a perceived threat of terrorism. It is simply a rejection of any and all collateral damage as immoral and as criminal. If we recall that it is ideas centred on the value of the individual and his or her right to life and bodily integrity that grounds the perceived utility of efficient warfare, this rejection should not be a surprise. There is nothing more obviously at odds with the deontological premise of individual rights than killing human beings as a means to a military end.⁹⁵

This elementary rejection of collateral damage should be visible then in regular war as well. It is, but to a slightly lesser extent. Gabriella Blum writes that in recent years criticism of collateral damage ‘has been so fierce as to bring some commentators in the US to advocate the abrogation of the principle [of proportionality] altogether’.⁹⁶ The impact of the reputational costs of collateral damage on the conduct of hostilities is best documented with regard to Operation Enduring Freedom. In July 2009, General Stanley McChrystal, upon taking command in Afghanistan, issued a directive that significantly restricted military action from the air in order to reduce civilian casualties. His predecessor General David McKiernan had already repeatedly curtailed pre-planned air strikes to avoid collateral damage. However, McChrystal also questioned the paradigm that troops in contact ought to receive all available air support, notwithstanding the considerable risk to civilians that called-in air strikes pose. He stipulated that ‘[t]he use of air-to-ground munitions and indirect fires against residential compounds is [sic] only authorized under very limited and prescribed conditions’.⁹⁷ From then on, as a matter of course, pre-planned targets were only approved if the expectation was that no civilians would get hurt. Time sensitive air-to-ground support for troops under fire was subjected to stricter oversight.

How do we explain that air power has not been ruled too costly in modern theatres but that it continues to be ‘the distinctively American form of military intimidation’?⁹⁸ It is the strength of the countervailing imperative to protect friendly forces. The McChrystal directive achieved its objective to limit civilian casualties.⁹⁹ However, it also coincided with a significant increase in fatalities among American troops: from 18 per month before the directive to an average of 33 per month thereafter. ¹⁰⁰ Criticism of McChrystal’s policy evoked not only the imperative of force protection in general but also a right of troops to defend their own lives.¹⁰¹ Combatants do not have a legal right of self-defence against other combatants under IHL, which would be altogether absurd. Yet, against the backdrop of an individual rights-based morality, self-defence is the only legal justification for killing a person.¹⁰² The norm enjoys considerable momentum. The notion that force protection during troops-in-contact situations maps onto this more fundamental norm that a person ought to be able to defend his or her life explains the pressures not to curtail air support.¹⁰³ By the same token, it explains why some collateral damage in all-out war may continue to be acceptable even to an increasingly human rights-conscious international public. The aspiration of zero-casualty warfare has shaped US combat operations across 21st century combat theatres,¹⁰⁴ but it has remained that: an aspiration.

Targeted killings with remotely piloted aerial vehicles do not raise a force-protection concern. Neither do they evoke the right of combatants, whose lives are no longer presumed to be at the disposition of the state, to effectively preserve their own lives. Against the backdrop of the norm that it is immoral to deprive individuals of their right to life as a means to an end, other than in self-defence, the specific moral rejection of targeted killings is plausible. Remotely piloted attacks on individuals respond to the pressures of saving the lives of combatants (on both sides). Yet, as a reaction to shared ideas about violence in international relations that are informed by an individual rights-based morality drone strikes present a fallacy. Perfect force protection takes away the last apology for collateral damage in a liberal order. Drone strikes therefore do not solve the 21st-century belligerent’s legitimacy problems.

From the point of view of an individual rights-based morality, how should a belligerent wage war then? A condition of legitimate state conduct would certainly be a complete avoidance of collateral damage, unless possibly in situations where troops are directly at risk. Those situations themselves need to be minimized according to the same moral standard, which upholds a combatant’s right to life. The avoidance of collateral damage drastically reduces the pool of objects a belligerent would be able to attack. As there are few ‘things to break,’ a belligerent will have to focus on ‘killing people’. Yet because combatants retain their right to life, whom is it that a belligerent may attack? According to an individual rights-based morality, where should belligerents direct the deliberate death and destruction that all wars inevitably involve?

As mentioned, in liberal individual rights-affirming societies, without prior due process, individuals may only be justifiably killed in self- (or other) defence. It is from this analogy that we can develop a list of the conditions under which an individual is morally liable to potentially lethal attack by another in war. An individual is generally considered liable to be killed in self-defence only if she is (i) responsible for (ii) contributing to (iii) an unjustified threat and (iv) a lethal attack is a proportionate and necessary response to the contribution.¹⁰⁵ I will focus on the first three conditions. Crucially, they concern the conduct of the individual and her resulting moral status, not her membership, in a group – for instance, the armed forces. In the words of Jeff McMahan: ‘To say that a person is morally liable to being harmed in a certain way is to say that his own action has made it the case that to harm him in that way would not wrong him, or contravene his rights.’¹⁰⁶

If it were rigorously implemented, a logic of moral liability would ensure that warfare, as far as deliberate attacks are concerned, does not involve violations of individual rights. Given that the legitimacy of combat operations seems to hinge on the moral imperative to protect individual rights, combat operations in accordance with the logic of moral liability have a much better chance at meeting expectations of legitimate state conduct than hostilities that follow either the logic of sufficiency or efficiency. The next logical step is then to change IHL with the aim of imposing on combat operations the logic of moral liability. Yet, bringing IHL in line with the logic of moral liability proves impossible. The following paragraphs elaborate why.¹⁰⁷

In order to distribute harm in war in accordance with individuals’ moral liability, we would have to determine which side is actually justified in their resort to force. Hence, we would have to look at the morality of resort to determine the morality of conduct. Combatants who fight without a just cause or who resist a just attack contribute to an unjustified threat to the combatants on the other side and are hence liable to defensive harm (including deliberate attack).¹⁰⁸ By implication, if combatants use force in defence of a just cause, they do not forfeit their right to life and should remain immune.¹⁰⁹ The only legal justification for resort to force without a mandate from the UN Security Council is self-defence. On the assumption that only one side, if any, in every war acts in self-defence, IHL that is no longer independent from questions of resort would have to relinquish symmetry.¹¹⁰ It would have to allow one belligerent actions that are prohibited to the other.

Belligerents often enter into wars because they mistakenly believe they are legally permitted to do so. Even if one side was aware that they were in want of a legal justification, the decision to nevertheless go to war suggests that the stakes are high.¹¹¹ Moreover, such a deliberate aggressor would likely lack scruples that could prevent him from using the law applicable to just belligerents. The law that is less permissive for one side would then never be applied. The asymmetry that the logic of moral liability requires – law is consistently differentially permissive for opposing belligerents – undermines compliance with international law.

The second obstacle to changing law so that it would be able to impose the logic of moral liability is that the logic challenges the principle of non-combatant immunity without offering a viable alternative category of persons that are legitimate targets. Contribution can be defined based on the significance of an action for the military effort, the magnitude of a contribution or the proximity of an action to the engagement of enemy forces. In whichever way it is conceived, it is not immediately obvious that civilians on the whole contribute less to a war than combatants. The argument made in the first section of this article about objects is true also for persons. With the possible exception of children, the elderly, ill people and outright dissidents or saboteurs, all members of a society somehow contribute to a war. The significance and magnitude of the contribution of many civilians will outweigh that of many combatants. Compare the nuclear scientist to the ill-motivated 18-year-old conscript. If we focus on proximity to military action in the definition of contribution, we come closer to tracking the distinction between combatants and civilians. Yet some combatants contribute only indirectly to military action – for instance, military chefs. In turn, munitions workers directly contribute to hostilities but are not considered combatants purely in virtue of their profession.

There is arguably a rough correlation between combatant status and a one causal step contribution to military action – a correlation that might justify keeping the principle of non-combatant immunity as the basis of distinction. However, I have so far neglected that the contribution has to be made ‘responsibly’. The fundamental obstacle to meeting the condition of responsibility is epistemic. It is difficult in the heat of battle to determine who exactly contributes to the fighting; it is impossible to determine who does so responsibly. Responsibility refers to a state of mind. We simply do not know the extent to which any combatant, or, for that matter, a civilian who contributes to a threat, has the requisite information to be aware that it is an unjustified or illegal threat. Even if we knew that combatants have the means to judge the legality and/or the moral justification of a war, outright or indirect coercion might excuse individuals and at least cast doubt over their moral liability to lethal attack.

In order to be practicable and complied with, IHL has to remain symmetrical, and it has to retain the principle of non-combatant immunity as the basis of distinction. This means it is incapable of meeting the standard of moral liability according to an individual rights-based morality.¹¹² Targeted killings decollectivize killing in war, but this brief discussion of the complexity of individual liability to lethal attack brings into sharp relief that we would need to know much more about the individuals under attack to be able to affirm their individual moral liability and perceive the attacks as justified. Yet, the fact that targeted killings come so much closer to implementing a logic of moral liability, yet meet with so much criticism, corroborates the earlier point that it is the infliction of collateral damage without the countervailing imperative of force protection and its association with self-defence that accounts for the vehement condemnation of drone strikes.

5 Conclusion

War inevitably jeopardizes human rights on a large scale and often marks the breakdown of order. As the current international system lacks an overarching authority with a monopoly on violence, the use of force by states against states is also sometimes the only available means of maintaining order or protecting human life. Regulating the conduct of war through law seems to offer a way to reconcile non-pacifist foreign policies with a society’s liberal identity. Yet, the paper demonstrates that IHL does not alter the Janus-faced role of war in international relations.

Complying with IHL, as we might expect, not only militates against the efficient pursuit of military victory, but it also fails to ward off moral reproof. This article demonstrates that strategic imperatives in war are chiefly determined by shared normative beliefs. Yet it is impossible to fully accommodate those normative beliefs, and a strategy of war needs to be practicable. Attempting to actually meet shared normative expectations about legitimate state conduct does not merely interfere with a state’s quest for military victory. For a state that does not want to fall foul of 21st-century human rights standards, attaining political goals with military force is an altogether unavailable tool of statecraft. To sum up, a state is left with a threefold choice: (i) acting legally by following the logic of sufficiency, which will be perceived as inefficient and immoral; (ii) following the strategic yet ultimately fallacious imperative to wage war efficiently and break the law in the process or (iii) renouncing the legal right and the capacity of self- (or other) defence by committing to pacifism in order to meet widespread expectations of legitimate state conduct. This is the trilemma of the 21st-century belligerent.