Guest Post: Israel’s Use of Law and Warnings in Gaza
[Janina Dill is a Lecturer at the Department of Politics and International Relations at the University of Oxford.]
In the ongoing military campaign ‘Protective Edge’ the Israeli Defence Forces (IDF) consistently issues warnings before air strikes against targets in Gaza. The population is warned of impending attacks with phone calls, text messages or so called ‘knocks on the roof’ (dropping of non- or low-impact explosives on the intended target). The warnings play a central role in Israel’s claim that, contrary to Palestinian armed groups, namely Hamas; it obeys the strictures of international law. ‘While the IDF goes to extraordinary lengths to avoid civilian casualties, Hamas deliberately puts civilians in the line of fire, the IDF maintains on its official blog. The First Additional Protocol to the Geneva Conventions, the relevant sections of which have the status of customary international law, in Article 57(2) c indeed prescribes that ‘effective advance warning shall be given of attacks which may affect the civilian population, unless circumstances do not permit’.
Warnings, their frequency and form, are at the centre of a narrative that Israel does not simply comply with, but goes beyond the call of international law in its care for the civilians of Gaza. ‘Israel’s use in the Gaza Strip of non-lethal warning shots to the roofs of buildings which constitute military targets … is not legally obligated’, the Military Advocate General’s Corps holds. Media commentary commends Israel for giving civilians a way out without even being obliged to do so. This is a misunderstanding of Article 57. The provision establishes an unequivocal obligation to warn before attacks that implicate the civilian population – as air strikes against a territory as densely populated as Gaza will regularly do. Granted, it is not an absolute obligation. The law recognizes that sometimes it may not be possible to warn. Crucially the provision does not say ‘warn if possible’, but ‘warn unless impossible’. It is open to interpretation when that is the case and reasonable people may disagree, but the default is to issue a warning and it is a failure to do so that requires explanation. Warnings are not acts of charity.
But are the kinds of warnings issued as part of Operation Protective Edge manifestations of the IDF’s commitment to the laws of war? The practice raises two distinct concerns. The first is that the air strikes the IDF carries out after issuing warnings are indicative of a misunderstanding of the legal implications warning before an attack. It has none! The second is the concern that the practice itself violates international law. I discuss them in turn.
The final paragraph of Article 57 provides a first strong indicator that warnings do not influence the legal obligations an attacker otherwise faces. It unequivocally states that ‘[n]o provision in this article may be construed as authorizing any attacks against the civilian population, civilians or civilian objects’. Only if warnings could somehow affect the status of persons or objects as civilian could they hence have legal implications. Can warnings change the status of the objects attacked? There is one and only one way in which an object can become a military objective under the principle of distinction: ‘by its nature, location, purpose or use [it] make[s] an effective contribution to military action’ and its engagement ‘in the circumstances ruling at the time offers a definite military advantage’. This is a test, enshrined in Article 52(2), the necessity or stringency of which is entirely unaffected by advance warnings. Civilian residences do not become military objectives only in virtue of their inhabitants having been warned.
By the same token, warnings do not take care of the attacker’s obligation to weigh expected collateral damage against the anticipated military advantage and make sure the former is not disproportionate, ‘excessive’ in the language of Article 51(5) b. While the IDF has never stated that warnings diminish other legal requirements, during Operation Protective Edge the sheer number of civilian residences attacked and the high ratio of civilian casualties (Human Rights Watch estimates 77%, the United Nations 73%) suggest that civilians who remain in their homes after a warning are disregarded or at least, as it were, attributed less ‘weight’ in the proportionality calculus than civilians who are taken by surprise. The argument for why civilians who remain in their homes, warnings notwithstanding, do not have an unimpaired claim to protection is that they must have decided to serve as human shields to the military objective that is the target of the impending attack. ‘Hamas exploits many of the IDF’s attempts to provide advance warning by using civilians as human shields in order to deter attacks’, the IDF charges.
Voluntary human shields may, under certain circumstances, be considered to directly participate in hostilities and hence, for such time as they do, lose their protection from attack under Article 51(3). Yet, staying put after a warning is not sufficient evidence for the warned civilians’ qualification as voluntary human shields. Staying in your home is not an affirmative step to block an attack either physically or legally. International law does not know of an ‘obligation to leave’ after being warned. That is true even if a warning is effective in the sense that it is unmistakable and civilians have the time and means to get to safety, all of which has been questioned in the case at hand. That children can never be considered voluntary human shields is self-evident.
All international laws asks of civilians is not to directly participate in hostilities, but that obligation cannot be considered violated by ‘not fleeing’. No hostile intent can be inferred from the decision not to leave one’s home in a war zone. Direct participation in hostilities is an exception to the most fundamental rule on which the entire idea of a legal regulation of warfare depends: civilians are immune from direct attack. It must be interpreted restrictively. Much as a person, in case of doubt, is a civilian, a civilian, in case of doubt, is not directly participating in hostilities.
By the same token, a belligerent may not simply assume that warned civilians leave so that the expected collateral damage of an attack that is preceded by a warning is zero or even at all diminished. Article 57(2) b demands that belligerents ‘cancel or suspend’ an attack if it becomes evident that ‘the attack may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects or a combination thereof which would be excessive in relation to the concrete and direct military advantage anticipated’. There is no reason why that may not become evident after a warning when it turns out that the warned civilians have not in fact left.
At the end of the day, any interpretation of Article 57 that imposes an additional obligation on civilians, but weakens the obligations the attacker faces is an interpretation contra legem. While by no means all rules of international law for the conduct of hostilities have an overriding humanitarian purpose, Article 57 does. Its regulative goal is to spell out what it means to do justice to distinction and proportionality by protecting civilians as much as militarily possible. It does not spell out how the attacker’s obligations fall by the wayside if the presence of the civilian population interferes with a belligerent’s ability to conduct effective military operations.
A second concern is not that warnings are used to weaken other legal obligations, but that the practice itself violates international law. In circumstances in which the civilian population has in fact no means to leave do warnings violate Article 51(2)? The second sentence prohibits ‘acts or threats of violence the primary purpose of which is to spread terror among the civilian population’. It is deeply plausible that a warning that one’s house will be bombed, in the absence of a real possibility to get to safety, does just that: it induces terror. Of course, from the practice alone we cannot infer that terrorising the civilian population is its primary purpose. Is their consistent ineffectiveness even counter-productiveness, as alleged by Amnesty International, Human Rights Watch and the United Nations, evidence enough that protecting civilians is not the primary purpose of warnings?
International law here asks us to make assumptions about the motivations of a belligerent, which is difficult enough in a court room. On a battlefield what we are left to do is look to a belligerent’s general concern for human life and commitment to international law. We tend to grant those to Israel and deny them to Hamas. When Hamas advises civilians to ignore warnings the IDF asks us to believe that the motivation is to use civilians as human shields, not to avoid panic and mass displacement as Amnesty International suggests might be the case. When Israel warns civilians we are meant to believe that the motivation is to protect civilians, not the old idea of terror-bombing civilians into withdrawing their support from a war-mongering government.
While Hamas has done little to suggest concern for human life and nothing to demonstrate commitment to international law, recent statements by Israeli politicians would make it naïve to accept warnings as simply the practice of a belligerent that selflessly upholds international law against an enemy that only uses it to score points. Lawfare, ‘the use of law as a weapon of war’, is usually understood to mean imposing reputational costs on the belligerent trying to comply with law by alleging violations or making compliance difficult, a charge routinely levelled against Hamas. Using law to circumvent legal obligations in order to be able to continue to claim compliance is a new form of the ‘negative manipulation of international … laws to accomplish purposes … contrary to … those for which they were originally enacted’. Using warnings as a means to legitimise the disproportionate victimization of civilians is lawfare 2.0.