Targeting Dual-Use Structures: An Alternative Interpretation

by | Jun 28, 2021

F-15C Strike Eagles Over San Francisco


​In a recent Articles of War piece, I discussed protection of the media under international humanitarian law, a topic that had come to the fore when the Israel Defense Force (IDF) attacked the Al-Jalaa Tower, a 12-story building in the Gaza Strip on May 15. The building, which was destroyed, housed the Associated Press and Al Jazeera offices, as well as civilian apartments. According to the IDF, the building “contained military assets of the Hamas terror organization,” which used the building for unlawful human shielding. Israel’s Ambassador to the United States later told Associated Press executives that Hamas was using the building to develop a means of disrupting Israel’s Iron Dome system. Israel gave occupants a one-hour warning by various means and the building was evacuated before being struck.

The incident provoked a great deal of debate in the IHL community (see, e.g., here, here, and here). Although my post focused on media protection, all attention centered on a single paragraph of dictum:

There is some disagreement on whether a building that contains both apartments or offices used for civilian purposes and others that have been converted to military use should be considered a military objective in its entirety or as consisting of separate and distinct entities. The better view, but one that does not appear to have achieved universal consensus, is that if an attacker can surgically strike that aspect of the building used for military ends, harm to the remaining sections must be factored into the proportionality analysis.

At the request of numerous scholars and practitioners in the field, allow me to explain my rationale for that assessment.

Points of General Consensus

In a definition of military objective that accurately restates customary law, Article 52(2) of the 1977 Additional Protocol (I) to the 1949 Geneva Conventions provides, “In so far as objects are concerned, military objectives are limited to those objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage.” Whether a civilian structure qualifies as a military objective is, therefore, a two-part inquiry. With regard to the view I expressed, only the first prong is at issue, that is, how to understand the requirement that the object in question effectively contributes to military action. This necessitates identifying the “object” that will be assessed against the military objective standard.

There is universal consensus that if a civilian building is being used to effectively promote or support military operations, at least part of it, if not the entire building, becomes a military objective—so long as the second prong of the military objective test is also met because targeting the building will result in a definite military advantage for the attacker. Moreover, if enemy forces will use a civilian building in the future for military ends, an attacker need not wait until it is so used before striking it, as the building qualifies as a military objective by the purpose criterion. Whether qualification as a military objective is based on the use or purpose criteria has no bearing on the extent to which the building qualifies as a military objective.

Before turning to that matter, it is necessary to dispense with three points that sometimes cause confusion. First, and most importantly, the mere fact that individuals who may be lawfully targeted (combatants, members of an organized armed group, or direct participants in hostilities) are present incidentally in a particular structure, such as a store or non-military-related government building, does not render the structure a military objective. Only if the individuals are somehow using, or intend to use, the structure itself—for instance as a command-and-control or storage facility, an observation post, or a location from which to snipe—does the issue of the building’s status as a military objective arise. In some cases, their presence may be evidence of the object’s use. But if not, damage to the building resulting from an attack on such individuals is collateral damage factored into the proportionality analysis and the requirement to take precautions in attack.

Second, plainly distinct structures must be assessed independently against the military objective standard. For instance, if a tunnel or a covered walkway connects two adjacent buildings, they nevertheless are generally considered separate for the purpose of military objective status even though they are physically connected.

Third, once a structure is no longer used for military purposes, and there is no evidence indicating future use, it reverts to civilian object status; it enjoys immunity from attack and damage to it during an attack on a lawful military objective or lawfully targetable individuals is considered in proportionality and precautions-in-attack assessments.

The Debate

At issue in this article is a structure in which all the components thereof comprise an integral whole, the paradigmatic example being a single apartment building comprised of several apartments or units. However, in the scenario, some aspects of the structure are not, and will not be, used for military purposes. Two general approaches to targeting such “dual-use” structures dominate discussion.

The first considers the entire building a military objective and therefore concludes that an attacker need not consider damage to the structure in the proportionality and precautions-in-attack analyses. This has always been the U.S. approach, as I understand it. The updated 2016 Department of Defense’s Law of War Manual provides “If an object is a military objective, it is not a civilian object and may be made the object of attack” (sec Similarly, the 2019 Army/Marine Corps Commander’s Handbook on the Law of Land Warfare [Army Field Manual (FM) 6-27/Marine Corps Tactical Publication (MCTP) 11-10C] clarifies that “[t]he principle of proportionality does not impose an obligation to reduce the risk of harm to military objectives” (para 2.72). Thus, if the entire building is a military objective, damage due to an attack is not collateral damage in proportionality calculations or precautions-in-attack analysis.

This is likewise the position of the IDF. Note that the IDF is of the view that harm to civilian objects within the building, such as furniture, qualifies as collateral damage, a reasonable view. Moreover, IDF Military Advocate General officers with whom I have spoken emphasize that when such situations occur, the IDF tries to avoid damage to the components of the building that the enemy is not using for military purposes, for instance by employing precision-guided munitions. In some cases, the IDF decides not to attack at all. These are policy, not legal, decisions.

From decades of working with armed forces worldwide, I can confirm that many militaries take the same approach or a slight variant thereof.  For instance, the 2020 Danish Manual provides,

As far as dual-use objects are concerned, the entire object constitutes a military objective. Under international law, this means that damage to the dual-use object is not regarded as collateral either in whole or in part if the object is effectively indivisible. As a rule, the non-military ‘share’ of the object should not be taken into consideration in the proportionality assessment (sec 3.1).

Like Israel, however, Denmark imposes greater restrictions as a matter of practice and policy.  It directs its forces to treat the non-military share of the objective as collateral damage when it is of “particular and direct importance to protected persons.”

The ICRC has articulated a second approach that has found favor among many academics. Like the first approach, it characterizes the entire structure as a military object, for “from a legal perspective, an object is either a military objective or a civilian object, and there is no intermediate category of dual-use objects.” But the ICRC then points to a troubling aspect of the first approach: “[I]f a fairly minor military use has turned a civilian object into a military objective (assuming that it fulfills the definition of Article 52), the damage caused to the remaining civilian part—however important it is—would have no bearing on the decision to launch an attack.”

To resolve this dilemma, the ICRC suggests that “while the dual use object is a military objective, the impact of the attack on the civilian part or component of the object (such as apartments in a building whose basement is used as a munitions depot) or on the simultaneous civilian use or function of the object (such as in the case of a bridge or electricity station used for both military and civilian purposes) must also be taken into consideration in the assessment of proportionality.”

In support of its approach, the ICRC cites the International Criminal Tribunal for the former Yugoslavia’s Trial Chamber analysis of the 1993 destruction of the Old Bridge in Mostar in the case of Prlić et al. Although the tribunal found the bridge to be a military objective, its destruction made it “impossible for [the residents of the Muslim enclave on the right bank of the Neretva] to get food and medical supplies resulting in a serious deterioration of the humanitarian situation for the population living there.” This made the bridge’s destruction, in the estimation of the ICRC, disproportionate, although, as will be explained, the reference to the case was misdirected because the case did not involve the parsing of an object.

Assessing the Prevailing Approaches

Neither approach comports neatly with IHL’s object and purpose, which is to achieve an appropriate balancing of military and humanitarian considerations. As a body of law crafted by States, IHL is intended to allow States effectively to fight while maximizing the protection of civilian, civilian objects, and other protected persons and objects.

The first approach risks sometimes skewing that balance in precisely the manner suggested by the ICRC. There is no denying the possibility that in fact it could lead to situations in which the real impact on civilians of an attack far outweighs the military need to attack the target, but the attack would nevertheless be lawful. The existence of the proportionality rule, by which important military attacks are sometimes forbidden due to the likelihood of causing excessive collateral damage, reflects IHL’s recognition, and accordingly that of States, that such a result is unacceptable.

But on the other hand, the second view is flawed in three ways. First, as an operational matter, it could incentivize the use of civilian facilities (especially large, complex, multi-faceted structures with significant civilian use) by fighting forces because, in some cases, the interpretation could bar enemy attack through operation of the proportionality rule. This would place civilians and civilian property at greater risk and exacerbate the already contentious issue of interpreting the prohibition of human shielding (especially voluntary versus involuntary shielding).

Second, it is simply wrong by the plain text of the law. The rule of proportionality provides, “[a]n attack which 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” is prohibited.” Even if an attack complies with the proportionality rule, the attacker must, by the precautions in attack requirement, “[t]ake 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.”

To begin with a minor point, note that the loss of use of a structure does not qualify as collateral damage. As the text clarifies, only damage qualifies, although the “value” of that damage may be assessed by reference to its use. But much more importantly, the plain text of both rules is crystal clear. Only harm to civilian objects is factored into the assessments. And according to Article 52(1) of the Additional Protocol I, which is well-accepted as reflecting customary law,  “[c]ivilian objects are all objects which are not military objectives as defined in paragraph 2” [the paragraph extracted above].

If the whole structure is a military objective, as both views assert, then neither the rule of proportionality nor the requirement to take precautions in attack comes into play. The sole remaining protection would be found in the Article 57(1) of Additional Protocol I requirement that “[i]n the conduct of military operations, constant care shall be taken to spare the civilian population, civilians and civilian objects.”  As I have explained elsewhere, I believe the obligation is customary and interpret the term “spare” to extend to harm beyond the death, injury, or damage criteria found in the other two rules, which are limited to military operations qualifying as “attacks.”

Lastly, the ICRC citation of the Prlić et. al. judgment does not support the assertion that it is necessary to consider the “damage caused to the remaining civilian part” of a structure in the proportionality analysis. Unlike a structure  only part of which serves a military purpose, the entire Old Bridge did so. Thus, the trial chamber actually addressed reverberating effects in the proportionality analysis, and in my view, correctly so. The blockage of food and medical supplies to an isolated civilian population would foreseeably cause consequences that would place the affected civilian population at risk of starvation and medical issues. These results amount to “injury,” as that term appears in the proportionality rule, and are accordingly appropriate to factor into the proportionality analysis.

Although in decades past there had been debates as to whether indirect effects factor into the proportionality and precautions in attack analyses, today the prevailing view is that they do so long as they are foreseeable and not too remote (see, e.g., DoD Law of War Manual sec If the destruction of a building qualifying as a military objective causes the requisite harm to civilians or civilian objects other than the building itself, it is appropriate to consider it as collateral damage when applying the proportionality and precautions rules. Indeed, it is this very point that the military manuals cited by the ICRC in support of its position, including those of the United States, make. Many academics have made the same mistake in interpreting those manuals and other statements by States.

An Alternative?

Considering the shortcomings of these two prevailing approaches, I believe States should consider a third interpretation of the law. In advancing the approach, I want to be clear that no State has espoused it to my knowledge. But I equally want to be clear that I am not proposing new law. Instead, my proposal is about how to interpret the notion of “military objective.”

Rules of international law should be construed in a manner that best achieves their object and purpose in light of the context in which they apply. In IHL, rules accordingly must remain responsive to the operational and humanitarian environment if they are to be respected by parties to the conflict. In this case, I do not suggest that the text of the rule defining military objectives should be ignored, but only that its good faith interpretation and application must reflect the reality of the conduct that the rule is meant to govern.

On its face, the proportionality rule only applies to civilians and civilian objects. While there may be interpretive play in the meaning of those two terms, if civilian objects are defined negatively as those objects that are not military objectives, one cannot characterize a structure as a military objective and at the same time apply the rule.

But in the case at hand, I believe there is interpretive play in the term “military objective.” Nothing in the accepted definition of that term in Article 52 necessitates treating an entire structure as a single entity. Rather, the term “object” in the provision can be understood to refer to a distinct, tangible entity. The mere fact that distinct entities are physically connected does not preclude treating them as separate. Thus, in the same way that a single apartment complex consists of individual buildings (objects) even if connected structurally by walkways or tunnels, a building can be conceived of as a collection of individual apartments (objects), in which there are separate rooms, together with other discrete features such as a lobby, basement, roof, and stairwells. The question is whether these separate entities may be disaggregated such that the use of one or more of them does not necessitate treating all of them as parts of a single military objective (the building) in IHL terms.

My view is that this should depend on the capabilities of an attacking force. If the attacker either cannot identify that part of the structure the enemy is using for military purposes or individually strike that part it can locate, then treating the entire building as the military objective, as both approaches mentioned above do, is appropriate. The IDF asserted that this was the case with the Al-Jalaa Tower: “There was no way of taking down only the Hamas facilities that were in the building. They occupied several floors in the building and it was impossible only to take down those floors. It was deemed necessary to take down the whole building.”

In such cases, damage to the aspects of the building that are de facto civilian, but de jure part of a military objective, would play no role in the proportionality or precautions in attack analysis (absent foreseeable indirect effects that qualify as incidental injury or collateral damage). Of course, the attacker may consider that damage as a matter of policy when deciding whether and how to attack, as Israeli and Danish forces do. Still, the lex scripta would appear clear: qualification as a civilian or a civilian object is a condition precedent to applying those two IHL rules.

Sometimes, though, it is possible to surgically strike entities that the enemy is using within a structure. To take a simple example, if the upper floor of a building is the only one being used for military purposes, it might be possible to use a low blast precision munition to collapse the ceiling without causing extensive harm to other parts of the building. Similarly, a helicopter might employ a rocket or gun against an apartment on the exterior of a building in a manner unlikely to harm the surrounding apartments. Of course, the feasibility of conducting such an attack depends on an array of factors. They range from reliability of intelligence and weapons systems availability to location of the entity within the building and the structure’s composition.

In these situations, I suggest that the aspect of the structure the enemy is using qualifies as a military objective, but its separate and distinct components that are not being used for military purposes retain their civilian character. The issue is one of interpretation. Nothing in the definition of “military objective” contemplates a feasibility assessment, but nor is there any aspect of the accepted definition that precludes its role in identifying a military objective. Moreover, this interpretation reflects the general principle of military necessity that informs many IHL rules—that military operations should not be more destructive than they need to be to fight effectively.

Indeed, the approach accords with operational reality. When those who plan and execute an attack are able to attack surgically, they tend to think of the “target” as that particular aspect of the building being used, not the building in its entirety. In their mind, they are taking out, for instance, the command and control capability located in the apartment, floor, etc., used for that purpose. By contrast, if the mission planners cannot conduct such an operation, for instance because they do not know which part of a building the enemy is using, they see the entire building as the target; this is because they need to drop the building to achieve the desired effect.

If clearly distinct components of a structure are treated as separate entities as a matter of law, the prohibition on indiscriminate attack would apply and be breached if the decision was taken to attack the building in its entirety when striking only the part constituting a military objective is feasible. This is because attacks must be “directed at a specific military objective.” But from an operational perspective, an attack against the entire building in such circumstances usually would be tactically unsound.

My proposed interpretation could also affect the proportionality analysis. For example, should damage to aspects of a structure other than the military objective (the floor, apartment, etc., being used by the enemy) be factored into the assessment, the operation might be precluded altogether because expected collateral damage would be excessive relative to the anticipated military advantage likely to result from the attack. That said, if the attacker cannot strike surgically, the interpretation preserves the option of striking the structure, so long as other IHL rules are satisfied.

The approach would have little effect on the requirement to take precautions in attack, for if civilians or civilian objects are within the building, expected harm to them would already necessitate employing the weapons or attack tactics likely to minimize that harm so long as no military advantage is lost (e.g., likelihood of mission success).

I see the proposed interpretation as consistent with other IHL normative dynamics. For instance, it is congruent with operation of the rule of proportionality, for an attack is sometimes acceptable in the face of civilian harm given the anticipated military advantage, and sometimes it is not.  Therefore, States adopting this approach would not be buying into a more operationally restrictive dynamic than that which already applies to most of their attacks.

And capabilities have long mattered in the application of IHL. The paradigmatic example is the prohibition on attacks that are indiscriminate because they “employ a method or means of combat which cannot be directed at a specific military objective.” As weapons have become more accurate and weapons systems have become more precise, our sense of when to characterize them as indiscriminate on this basis has evolved; weapons that may have been considered sufficiently precise in the 1950’s such as high-altitude gravity bombs, would likely be regarded as indiscriminate if used today.

Admittedly, it has been claimed that the approach (see here at 335) infuses IHL with relativity because some armed forces can conduct surgical attacks on structures while others cannot. It risks constraining the former’s operations more than the latter’s. This understandably troubles technologically advanced militaries.

Yet, IHL is already replete with relativity based on an attacker’s capabilities. It has long been the case, for example, that “[a]ttacks by bombardment by any method or means which treats as a single military objective a number of clearly separated and distinct military objectives located in a city, town, village or other area containing a similar concentration of civilians or civilian objects are prohibited.” Whether an area must be treated for attack purposes as encompassing separate and distinct military objectives depends on weapons, intelligence, and other capabilities. After all, if an armed force cannot strike the military objective individually, the attack is not unlawful on this basis, but if it wields pinpoint attack capability, it cannot mount an area attack. The ICRC’s AP I commentary on the prohibition supports just such an interpretation: “[W]hen the distance separating two military objectives is sufficient for them to be attacked separately, taking into account the means available, the rule should be fully applied.” The legal issue differs from that under consideration here, but the rule demonstrates that capability and proximity matter in the law governing attacks.

Finally, application of the requirement to take precautions in attack is expressly dependent on feasibility, which includes consideration of available means for verifying the target, methods and means of warfare options for attacking it, the existence of any viable alternate targets the attack on which can achieve the desired effect, and the attacker’s ability to issue an effective warning in the circumstances. Thus, the requirement for an attacker to adjust a planned attack to minimize incidental injury and collateral damage depends on, among other factors, capabilities. Since different militaries field different capabilities, an attack that would be lawful for one party may not be for another that has a greater capability to avoid some of the resulting harm to civilians and civilian objects.  Again, capabilities may determine whether an IHL violation has occurred. Clearly, the proposed approach is not the introduction of relativity into this body of law for the first time.


I realize there is something for both sides of the military necessity–humanitarian considerations balancing to like, and to dislike, in my suggested approach. But IHL necessitates compromise in the balancing of these oft-competing values. In my view, the proposed interpretation affords greater protection to factually distinct civilian entities while avoiding unduly tying the hands of States on the battlefield.

It want to reemphasize that I am not proposing new law. Instead, the proposal derives from the fact that the interpretation of existing IHL rules must evolve in lockstep with changes in warfare and the values that States seek to protect through IHL. I believe this interpretation does just that. Even if States do not adopt it as a matter of legal policy, they should consider doing so from a policy perspective, both for operational reasons and to limit the enemy’s ability to engage in lawfare.

Finally, there is a tendency in IHL to focus on the law governing the operations of attackers. This focus often skews the reality of the conflict. In the situation under consideration here, it must be remembered that the defender elected to locate its operations in a structure that was already protected as a civilian object; that party’s actions alone caused the structure to lose its IHL protection. In many cases, doing so is a violation of the defender’s passive precautions obligation to “protect the civilian population and civilian objects under their control against the effects of attacks” or the prohibition on human shielding, or both. Balance is necessary in the evaluation of any conflict; sadly, it is often missing.


Michael N. Schmitt is the G. Norman Lieber Distinguished Scholar at the United States Military Academy. He is also Professor of Public International Law at the University of Reading, Strauss Center Distinguished Scholar and Visiting Professor of Law at the University of Texas, and Charles H. Stockton Distinguished Scholar-in-Residence at the United States Naval War College.