The Gaza Report’s Treatment of Warnings: A Response to Blank

Laurie Blank published a post yesterday at Lawfare entitled “The UN Gaza Report: Heads I Win, Tails You Lose.” The post accuses the Independent Commission of Inquiry’s report on Operation Protective Edge (“Gaza Report”) of “completely undermin[ing] the foundational notion of equal application of the law” with regard to three areas of IHL: warnings, civilian vs military objects, and compliance. None of Blank’s criticisms are convincing, but in this post I want to focus solely on her first topic, warnings. Here is what she says about the Commission’s discussion of whether Israel complied with its obligation under IHL to provide civilians in Gaza with “effective advance warning” prior to attack:

First, consider the report’s treatment of warnings, one of the precautions set out in Article 57 of Additional Protocol I.  Article 57 mandates that when launching attacks, “effective advance warning shall be given of attacks which may affect the civilian population, unless circumstances do not permit.”  The Commission examines Israel’s warnings in great detail, including leaflets, telephone calls, texts and roof-knocks, noting that the warnings often did lead to successful evacuation and save many lives.  However, the Commission found in many cases that specific phone warnings were not effective as required by LOAC, because the individuals in the targeted building would not know “in what direction to escape.” (¶ 237).

However, LOAC contains no requirement that the civilian population be able to act on the warnings in order to find them effective.  Instead, the legally correct approach is to examine whether the warnings generally informed civilians that they were at risk and should seek shelter. In other words, the legal issue is whether they were effective in transmitting a warning, not whether the civilians actually heeded them. The Israel Defense Forces routinely made individualized, specific phone calls to warn the residents of buildings to seek safety in advance of an attack on a particular building, far exceeding the requirements of LOAC. Yet the Commission bases its conclusions on the post-hoc question of whether civilians actually found shelter, which ultimately depends on a host of considerations outside the control of the attacking party.

Unfortunately both paragraphs misrepresent the Gaza Report. Let’s consider Blank’s claims one-by-one.

[T]he Commission found in many cases that specific phone warnings were not effective as required by LOAC, because the individuals in the targeted building would not know “in what direction to escape.”

The Israel Defense Forces routinely made individualized, specific phone calls to warn the residents of buildings to seek safety in advance of an attack on a particular building, far exceeding the requirements of LOAC.

These statements are misleading. The subsection of the Gaza Report that Blank criticises focuses on Israel’s controversial use of “roof knocking,” not on its use of phone calls to civilians located in or near buildings about to be attacked. (The subsection is entitled “Roof Knock Warnings.”) Indeed, the entire point of the subsection is to explain why roof-knocking does not provide civilians with effective advance notice unless it is combined with a phone call or “other specific warnings” (¶ 239). Blank does not challenge the Commission’s conclusion in that regard. She does not even acknowledge it.

Yet the Commission bases its conclusions on the post-hoc question of whether civilians actually found shelter, which ultimately depends on a host of considerations outside the control of the attacking party.

This claim, which is central to Blank’s argument, is simply false — which is probably why this paragraph in her post is neither quotes nor cites the Gaza Report. At no point in the “Roof Knock Warnings” subsection does the Commission infer the ineffectiveness of a warning from the bare fact that civilians were killed in an attack. On the contrary, the Commission emphasises again and again that, in the context of Gaza, roof knocks simply do not provide civilians with adequate notice that their home or a nearby building is about to be attacked. Consider, for example, the following statement (¶ 236):

The confusion caused by “roof-knocks” is understandable in such a densely populated area. Often, there are several buildings of 4, 5 or 6 floors adjoined.   When a roof is hit by a strike it thus becomes very difficult for people located on the lower floors to understand whether the warning was targeting their own building, or the one to the left, to the right, behind or in front, or across the narrow street; thus the “roof-knock” warning raises the question whether fleeing might put one more at risk than staying put.

The Commission illustrates this problem by referencing Israel’s notorious attack on the Dheir home, which killed 19 civilians, including nine children. According to the Commission, the individuals died because they “did not understand that their house had been the subject of a ‘roof-knock’ and only fled once “they were told by a neighbour that they had to flee” (¶ 236).

This is the kind of nuanced analysis that Blank dismisses as the Commission “bas[ing] its conclusions on the post-hoc question of whether civilians actually found shelter.” Nothing could be further from the truth, as any fair reading of the paragraphs in question demonstrates.

LOAC contains no requirement that the civilian population be able to act on the warnings in order to find them effective. Instead, the legally correct approach is to examine whether the warnings generally informed civilians that they were at risk and should seek shelter. In other words, the legal issue is whether they were effective in transmitting a warning, not whether the civilians actually heeded them.

There are numerous problems with this statement. To begin with, the Commission never claims — nor even suggests — that warnings cannot be effective unless civilians actually heed them. The Commission’s point is that civilians were not capable of heeding Israel’s warnings, because those warnings were not specific enough to allow them to find safety. The quote above makes that inordinately clear, as does the Commission’s discussion of whether Israel gave civilians enough time to heed warnings to leave their homes (¶ 237):

The short time lapse between the “roof-knock” strikes and the actual strike that resulted in the destruction of the building is also cause for concern. In cases documented by the commission and by NGOs only a few minutes (between 3 and 5) elapsed between the two strikes. If the “roof-knock” is the first warning, a few minutes are clearly not sufficient to allow a multi-storey building inhabited by families with children and elderly and sometimes disabled persons to be evacuated, taking into account the time required to realize that the strike was meant as a warning. In one case documented by an IDF video presenting an audio recording of a conversation between an IDF soldier and a resident of Gaza, while showing the image of a building surrounded by many others, the IDF officer only provides five minutes for the person to leave the house before a neighbouring building is destroyed.

Once again, we see here the Commission emphasising that Israel’s warnings did not give civilians a chance to get to safety, even when they wanted to. Blank’s claim that the Commission infers the legal inadequacy of warnings from the fact that civilians failed to heed them is thus simply false — a caricature of what the Report actually says.

Finally, it is worth noting that Blank’s understanding of the “legally correct” approach to Art. 57(2)(c) is far from incontestable. Recall her argument:

LOAC contains no requirement that the civilian population be able to act on the warnings in order to find them effective. Instead, the legally correct approach is to examine whether the warnings generally informed civilians that they were at risk and should seek shelter.

This is difficult to accept, at least on its face. Consider the following situation: Attacker X wants to bomb building Y, a legitimate military objective. 100 civilians are trapped against their will in building Y. Attacker X is aware of that fact, but bombs building Y anyway after warning the trapped civilians they are about to be attacked. According to Blank, attacker X has fully complied with Art. 57(2)(c)’s “effective advance warning” requirement.

That is an exceedingly counterintuitive understanding of “effective advance warning,” one that essentially reads “effective” out of the requirement. Moreover, Blank’s understanding is difficult to reconcile with Art. 57(2)(a)(ii) of AP I, which provides that an attacker should “take all feasible precautions in the choice of means and methods of attack with a view to avoiding, and in any event to minimizing, incidental loss of civilian life, injury to civilians and damage to civilian objects.” Feasible precautions do not necessarily require effective advance warnings: if the attacker can ensure that civilians will not be incidentally harmed in an attack even without warning them, no warnings are necessary. But it is difficult to see how an attacker can be said to have taken taken “all feasible precautions in the choice of means and methods of attack” if it bombs a building despite knowing that it contains civilians and that the civilians will not be able to heed its warnings to leave. In such a situation, where the attacker knows that it is virtually certain the civilians will be killed despite advance warning, Art. 57(2)(a)(ii) would seem to obligate the attacker to consider a different method of attack on the building.

That is obviously an extreme situation, one that will rarely occur in practice. In many cases, perhaps most, an attacker will be uncertain whether a warning will be effective because the civilians may be able to escape the attack. But the Commission took that uncertainty into account — once again reflecting the sophistication of its analysis. Here is the final paragraph in the “Roof Knock Warnings” subsection, also ignored by Blank, in which the Commission notes that Israel had an obligation under Art. 57(2)(a)(ii) to modify its methods of attack once it became clear its warnings were not being effective:

242. The limited effectiveness of these precautionary measures must have become abundantly clear during the early days of the operation, given that many buildings were completely destroyed together with their inhabitants. The apparent lack of steps taken to re-examine these measures in light of the mounting civilian toll suggests that Israel did not fully comply with its obligation to take all feasible precautions in attack.

To be sure, IHL rules concerning precautions in attack are notoriously murky. And Blank is simply following the US approach to precautions, which does not require an attacker to specifically warn civilians of an impending attack, much less provide them with information about how to avoid the attack. Here is what the new Law of War Manual says:

5.11.1.1 Effective Advance Warning. There is no set form for warnings. Warnings may be general, communicated to the national leadership of the enemy State, or delivered to the civilian population through military information support operations (such as broadcast or leaflets) advising the civilian population of risk of injury if they remain near military objectives. Giving the specific time and place of an attack is not required.

The US position, however, is far from universal. Indeed, the UK — which is not exactly known for adopting a touchy-feely approach to IHL — takes a very different position. Here is its Manual of the Law of Armed Conflict (emphasis mine):

5.32.8 There is a duty to give advance warning of an attack that “may” affect the civilian population, unless circumstances do not permit. Obviously, the point does not arise as a matter of law if military operations are being conducted in an area where there is no civilian population nor if the attack is not going to affect the civilian population at all. In other cases, the warning must be given in advance and it must be effective. The object of warnings is to enable civilians to take shelter or leave the area and to enable the civil defence authorities to take appropriate measures. To be effective the warning must be in time, sufficiently specific and comprehensible to enable them to do this.

The UK thus rejects the idea that “LOAC contains no requirement that the civilian population be able to act on the warnings in order to find them effective.”

Moreover, even some Israeli scholars reject Blank’s (and the US’s) insistence that a general warning to civilians is enough to satisfy Art. 57(2)(c). Consider, for example, the position of Pnina Sharvit Baruch, a retired IDF Colonel, and Noam Neuman, a Lt. Colonel in the IDF. Although they do not accept the idea that IHL always requires an attacker to explain to civilians “what they should do to avoid harm” and “[a]s far as possible… state the location to be affected and where the civilians should seek safety,” they acknowledge that there will be situations in which a warning cannot be considered legally effective unless it provides such information (pp. 386-87):

Having said that, we recognize that in order to fulfill the aim of enabling civilians to protect themselves from attack, there might be situations in which warnings need to include some guidance and instruction as to how civilians should act following receipt of the warnings. Such circumstances may exist when, without such information, the warnings would not give the civilians sufficient understanding of what they need to do in order to protect themselves and the attacking side has the ability of clarifying the situation without compromising its military concerns.

There is no question that IHL’s rules regarding warnings and precautions are uncertain and difficult to apply, so it is not surprising that scholars have taken issue with how the Commission has applied them in the context of Operation Protective Edge. It is also more than possible that some of the Commission’s conclusions are wrong. Blank, however, consistently misrepresents the Gaza Report and criticises it on the basis of an understanding of warnings that is anything but settled. Keep both problems in mind when you read her post.