I read with great interest my friend Rogier Bartels’ long post arguing that it is perfidious to use a bomb planted in a civilian car to kill an enemy soldier. As Rogier notes, the post is his more formal contribution to a recent debate — here and in Opinio Juris — concerning the nature of perfidy involving him, me, Ian Henderson, Ryan Goodman, and Sarah Knuckey.

Rogier’s post is excellent, and I agree with much of it. But I still disagree with his central point. Here is the crux of his argument (emphasis mine):

During the drafting of AP I, the ICRC clarified that perfidy “was based on the objective notion of good faith and on the subjective notion of intention:” the good faith on the side of the adversary, and intent to betray that good faith in order to “create surprise” on the side of the attacker. This goes to the core of my disagreement with the view of Ryan, Sarah, and the others, best captured by Kevin, who submitted that a “positive act to invite confidence” needed for a killing to be perfidy, was hard to discern in the killing of M (from our case study). That is not an unreasonable position. In my view, however, the opposite is true. Camouflaging a bomb as a civilian car does constitute a positive act to invite confidence.

The bomb was not just placed on the street or in a parking lot. It was “disguised” as a civilian object (because that is what effectively is done, when a bomb is planted in or mounted on a civilian car). The car thereby became a military object, but (purposely) no measures were taken to distinguish it as being military. Using a civilian vehicle “invites the belief” that this was indeed a civilian SUV. Seeing a regular civilian car, a combatant should be able to trust (be confident) that he is not going to be attacked from that car (or that it is not “going to attack” him by way of a shaped or targeted detonation). And, of course, a civilian vehicle was used on purpose, rather than a green vehicle with IDF or US army symbols on it, as surely M would not obliviously have walked up such an army vehicle belonging to the enemy.

The key is the bolded text: Rogier believes that by failing to mark the car as military, the attackers in his case study, dubbed A, took advantage of a rule of international humanitarian law (IHL) — that parties to a conflict must distinguish their soldiers and military objects from civilians and civilian objects — to gain an unfair advantage over M, their adversary. M saw the car, had no reason to suspect that it was anything but the civilian object it appeared to be, and thus approached the car believing it would not be used to attack him. Boom. Perfidy.

But is it really true that whenever a soldier sees a civilian object, the rules of IHL entitle him to assume he will not be attacked from it? I don’t think so. To see why, we need only consider three types of attacks — ambushes, camouflage, and landmines — that have never been considered perfidious but involve precisely the same kind of “disguising” of civilian objects that Rogier condemns in the context of a car bomb. 


The first type of attack is an ambush. As long as the soldiers are in uniform (as Rogier notes in his follow-up post), ambushes are not perfidious; they are permissible “ruses of war.” Indeed, the ICRC cites at least 16 states in its discussion of such ruses in the Customary IHL Study — Rule 57 — that consider ambushes permissible ruses of war. Moreover, Bothe, Partsch, and Solf specifically deem ambushes permissible ruses of war in their seminal Commentary to the First Additional Protocol.

The permissibility of ambushes undermines Rogier’s theory of perfidy. If the mere act of using a civilian object to attack an enemy soldier is perfidious, ambushes can only be lawfully launched from civilian objects that are clearly marked as military — which would, of course, defeat the entire point of an ambush, given that soldiers always expect to be targeted from military objects. More specifically, if Rogier is correct, uniformed soldiers engaged in urban warfare cannot launch an ambush from any residential house, cannot launch an ambush from behind any fence attached to a private business, cannot launch an ambush from behind a burned out school bus, and so on. In fact, according to Rogier’s theory of perfidy, no individual soldier could ever use any civilian object as cover, because doing so would conceal the fact that he intended to use the object to attack the enemy and would thereby “invite the confidence” of enemy soldiers that they could safely approach it. I know of no state or IHL scholar who believes that tossing grenades from behind a Toyota in the middle of an intense urban firefight is perfidious.

When I mentioned these types of ambushes to Rogier in the comments to my Opinio Juris post on the US-Israeli operation, he responded by distinguishing between ambushes in urban areas and ambushes that are “laid somewhere in the field, making use of the natural surroundings to hide behind and set up a ‘kill zone’ where the enemy is lured towards.” In his view, urban ambushes are perfidious while “classic” ambushes in the field are not.

But that doesn’t work, for one simple reason: the natural environment is no less a civilian object than a house, the fence of a business, or a school bus. The ICRC makes this clear in Rule 9 of the Customary IHL Study (emphasis added):

State practice considers civilian areas, towns, cities, villages, residential areas, dwellings, buildings and houses and schools, civilian means of transportation, hospitals, medical establishments and medical units, historic monuments, places of worship and cultural property, and the natural environment as prima facie civilian objects, provided, in the final analysis, they have not become military objectives.

Article 8(2)(b)(iv) of the Rome Statute also explicitly criminalizes attacks that cause disproportionate damage to the natural environment, thereby clearly deeming the environment a civilian object.

Because the natural environment is no less a civilian object than a house, the fence of a business, or a school bus, Rogier’s theory cannot distinguish between urban ambushes and ambushes in the field, considering only the former to be perfidious. If enemy soldiers are entitled to believe they will not be attacked from civilian objects, they are just as entitled to assume they will not be attacked from hedgerows or dense clusters of trees in a jungle as they are (according to Rogier) entitled to assume they will not be attacked from a privately-owned car in the middle of a city. In both types of ambush, the attacker is equally (according to Rogier) gaining an unfair advantage over his enemy by concealing the fact that a civilian object is being used for military purposes. But even Rogier admits that ambushes in the field are permissible.


The second type of attack that is inconsistent with Rogier’s theory of perfidy is the use of camouflage, which is specifically deemed a permissible ruse of war by Article 37(2) of the First Additional Protocol. The military camouflages military objects for one reason, and one reason only: to trick the enemy into believing that they are in a non-military area. In other words, the entire point of camouflage is to make the enemy believe it is surrounded only by civilian objects, whether those objects are city streets, residential houses, or clusters of trees in the jungle. So I don’t see how Rogier can plausibly maintain that the mere act of failing to distinguish a military object from a civilian object and then using that military object to launch an attack qualifies as perfidy, unless he believes that camouflage is a permissible ruse of war only in areas that are otherwise marked as military — an idea that is counterintuitive, has no basis in conventional IHL, and is difficult to reconcile with state practice. After all, numerous states involved in WW II camouflaged the tanks they used to fight in villages, forests, jungles, etc. Similarly, Rogier’s theory of perfidy would seem to require individual soldiers dressed in camouflage to avoid fighting from the vicinity of any residential house, business, car, etc. during an urban firefight, lest the enemy be invited to wrongly believe that they cannot be attacked from that civilian object.


The third and final type of attack that is difficult to reconcile with Rogier’s theory of perfidy is the use of landmines. Aside from the former not requiring remote detonation, placing a landmine under (say) an unpaved road in a city is indistinguishable from placing a bomb in a privately-owned car. According to Rogier’s theory, therefore, in both situations an enemy soldier would be entitled to assume he could not be attacked from the civilian object — after all, a city street is no less a civilian object than a privately-owned car, as Rule 9 makes clear.

But states are, in fact, entitled to place unmarked landmines under city streets — at least during or immediately before combat. Here is Article 4 of Geneva Protocol II on Prohibitions or Restrictions on the Use of Mines, Booby-Traps, and Other Devices (emphasis added):

1. This Article applies to: (a) mines other than remotely delivered mines; (b) booby-traps; and (c) other devices.

2. It is prohibited to use weapons to which this Article applies in any city, town, village or other area containing a similar concentration of civilians in which combat between ground forces is not taking place or does not appear to be imminent, unless either: (a) they are placed on or in the close vicinity of a military objective belonging to or under the control of an adverse party; or (b) measures are taken to protect civilians from their effects, for example, the posting of warning signs, the posting of sentries, the issue of warnings or the provision of fences.

In other words, it is permissible to use landmines in a city, town, village, or other area containing significant numbers of civilians as long as ground combat is taking place there or is about to take place. But if that is the case, then it cannot be perfidious to place a landmine under a city street, despite the fact that the entire point of doing so is to gain an advantage over enemy soldiers by inviting them to believe that they can cross the street safely. Indeed, one of the most striking aspects of Geneva Protocol II is that the provision on booby-traps (Article 6) provides that the rules therein are “[w]ithout prejudice to the rules of international law applicable in armed conflict relating to treachery and perfidy,” while the provision on landmines does not. (The same is true of the Chemical Weapon Convention’s Protocol on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices, which Rogier discusses in his follow-up post.) So once again Rogier’s theory of perfidy proves overinclusive.


At the risk of sounding like an armchair psychologist, I’d like to suggest an explanation for why an excellent scholar like Rogier adopts a theory of perfidy that, in my view, cannot be correct. The problem, I think, is the nature of the attack that gave rise to our lively debate: a bomb placed in a privately-owned car in the middle of a generally peaceful city. Such an attack simply doesn’t seem fair; of course a “combatant” — even a high-ranking member of Hezbollah — is entitled to feel safe walking by a car on “a quiet nighttime street in Damascus after dinner at a nearby restaurant,” as the Washington Post put it. Indeed, like Rogier, I am skeptical that IHL even applied to the bombing.

But just as hard cases make bad law, unusual situations generate problematic rules. Once we try to apply Rogier’s theory of perfidy to the “normal” combat situation, its plausibility falls apart. Although the same military/civilian distinctions apply, those distinctions take on a very different sheen during street-by-street, house-by-house fighting in a city virtually destroyed by armed conflict. You expect to be able to walk by a Mercedes in a Damascus suburb without being blown up, even if you are a soldier; but if you are a soldier in downtown Fallujah, the last thing you are going to do is walk casually past that burned out, overturned Mazda sitting in the middle of the city’s main road. Yet that Mazda is no less a civilian object than the Mercedes, and as long as IHL applies there is no legal difference between planting a bomb in the Mazda and planting a bomb in the Mercedes. Either both car bombs are perfidious or neither of them is. And it is very difficult to argue that planting a bomb in a burned-out, overturned Mazda in downtown Fallujah — or placing an ambush behind it, or using it for cover, or blending into it with camouflage, or placing a landmine near it — is an act of perfidy.

I share Rogier’s concern with the Israel/US operation that killed the Hezbollah leader, and I understand his unease — from a civilian protection standpoint — with many of the kinds of attacks I’ve discussed in this post. Any proposal to expand the definition of perfidy, however, must acknowledge the (ugly) reality of combat, particularly in urban areas. The general distinction between perfidy and ruses of war is a sensible one, even if we can — and should — debate precisely where the line between the two is drawn.