The degree to which militaries are capable of minimizing civilian harm hinges largely on what ‘precautionary measures’ commanders take to reduce the calamities of war. ‘Precautionary measures’ is a legal term of art under international humanitarian law (IHL) that requires warring sides to limit their means and methods of warfare in a manner that avoids or at least minimizes harm to civilians and damage to civilian objects.
But the precautionary measures commanders take will largely hinge on what information they have about the cities they fight in and the civilian populations that live in those cities. And the information they have at their disposal about these places and people will largely hinge on the amount of time and the quality of resources they have to collect and analyse that information. The problem, though, is that baseline information about cities, such as population concentrations and locations of civilian structures, can often be hard to find and, even when such information is available and accurate, it can rapidly change in urban settings.
Fortunately, the rules of precautionary measures can be interpreted as having embedded within them a commander’s obligation to obtain information well in advance of mounting attacks. This does not mean that advanced planning is a panacea for all civilian casualties. The United States has often taken considerable precautionary measures only to have those efforts undercut with its broad definitions of who or what constitutes a lawful target. That said, this post unpacks the contours of this obligation and expands on why in situations of urban conflict it is so important for warring parties to implement.
The commander’s obligation to obtain information
Article 57(1) of Additional Protocol I sets out the general rule of precautionary measures, which requires parties to an armed conflict to take constant care to spare the civilian population, civilians, and civilian objects. More specific rules are found, in part, in Article 57(2)(a). This Article requires those who plan or decide an attack to do everything feasible to verify that an object being targeted meets the legal criteria for what can be lawfully targeted; take all feasible measures to make choices about the means (i.e., weapons used) and methods (i.e., tactics used) of warfare with a view to avoiding or at least minimizing such civilian harm; and refrain from launching attacks that are expected to cause civilian harm that may be excessive to the anticipated military advantage, which implicitly requires a commander to assess the attack’s anticipated military advantage and civilian harm.
Articles 57(1) and 57(2)(a) are applicable to international armed conflicts and reflect customary international law in IACs and NIACs. These obligations are therefore also applicable to States that are party to non-international armed conflicts.
The information that a commander has at their disposal has a decisive influence over what Article 57(1) and 57(2)(a) measures they will take when planning an attack. Knowing how and when an enemy navigates through the city; where their headquarters are located; where civilians reside and what times they frequent certain locations over others; and, where protected places—such as hospitals and places of worship—are located, are just a few of the things that can help a commander verify who, or what, can (or cannot) be targeted and how an attack can be designed to spare civilians and civilian objects.
At the same time, the battle tempo of urban warfare is often high, the locations of civilians and fighters can rapidly change, and enemies that refuse to wear uniforms easily blend into the civilian population. Situations may arise when, for example, a commander orders defensive air or ground operations in response to troops that unexpectedly come under fire in urban areas. This can leave commanders with limited time to collect and analyse information about, say, the presence of civilians and identifying civilian buildings in relation to the enemy threat.
What IHL requires a commander to do in this situation is not always an easy question to answer. It some cases it may require the commander to show restraint, cease, or slow down the tempo of combat until more information is obtained. But one of the best ways to address these types of situations is to find ways to reduce the number of times they occur. This is where Articles 57(1) and 57(2)(a) come into play.
The Article 57(1) obligation to take constant care to spare civilians and civilian objects (‘constant care’ for short) does not have a temporal scope. The same is true for the Article 57(2)(a) obligation on commanders to do ‘everything feasible’ to verify that an objective is a lawful target and to take ‘all feasible precautions’ to avoid or minimize civilian harm. This means that in addition to placing obligations on a commander to take constant case and feasible precautionary measures at the time of an attack, these two articles place the additional obligation on commanders to avoid—or at least minimize—civilian harm by engaging in target verification and civilian harm avoidance activities well in advance of deciding, planning or launching an attack. In this way, Articles 57(1) and 57(2)(a) serve to prevent commanders having to operate blindly in situations where they do not have the time to collect and analyse information about, for example, whether sniper fire directed at their soldiers is coming from atop a makeshift medical facility.
State practice and authoritative legal interpretation support this approach. How else, for example, could a French commander implement the instructions in the 1992 Law of Armed Conflict Summary Note to obtain, prior to taking any military action, ‘a maximum of information concerning the nature and the location of protected objects (medical units, cultural objects, installations containing dangerous forces) and concerning any concentration of civilians’. To avoid civilian harm, Australia’s 2006 law of armed conflict manual similarly states that it is ‘important to obtain accurate intelligence before mounting an attack’ and that ‘the best possible intelligence’ is required concerning:
- concentrations of civilians;
- civilians who may be in the vicinity of military objectives;
- the nature of built-up areas such as towns, communities, shelters, etc;
- the existence and nature of important civilian objects and specifically protected objects; and
- the environment.
The type of information that these, and a diverse range of other military instructions, require commanders to obtain is not the type of information that can be quickly gathered and analysed in the same moment as a commander decides how to respond to an unexpected attack on forward operating forces conducting ground patrols in a densely populated area. Such a situation highlights the need for a commander to collect certain types of information on an iterative basis in anticipation that an attack will inevitably have to be launched in the not too distance future.
Most importantly, this information should be up-to-date and should not rely on outdated data from old census figures or city maps, especially when information that is more accurate can be collected. This is why the Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia explained that ‘A military commander must set up an effective intelligence gathering system to collect and evaluate information concerning potential targets’.
Implications of the ‘feasibility’ standard
To be sure, the rules of constant care and precautionary measures are not absolute. For example, many of the obligations in Article 57(2) rest on what is feasible. Many States have interpreted the obligations to take ‘feasible’ precautions as ‘being limited to those precautions which are practicable or practically possible, taking into account all circumstances ruling at the time, including humanitarian and military considerations’. Support for this interpretation is found in treaty law on the use of certain weapons. (See here and here.)
One consequence of this feasibility standard is thatnot all those who plan and decide attacks have the authority to put in place or allocate the resources that are necessary to verify targets and avoid civilian harm. What might be a feasible measure for a very high-level commander to take might be well beyond the authority of a lower-lever commander. For that reason, the obligation to obtain information in advance of any attack, when applied correctly, can run up the chain of command to potentially the highest levels.
Another consequence of the feasibility standard is that it may not be appropriate to judge warring parties with limited resources on par with warring parties that have at their disposal high-end technologies that assist with target verification and civilian harm analysis. Time may also place constraints on what precautionary measures are feasible for a commander who wants to take immediate unplanned defensive or offensive action. The ‘feasibility’ standard is high nonetheless and under-resourced commanders and commanders with time limitations do not have carte blanche to target whatever or however they wish. The importance of such a high standard is especially obvious in densely populated urban settings where attacks have an almost inherently high potential for civilian harm and damage to civilian objects. As noted above, in some cases that law may require a commander not to attack unless additional information is gathered.
Methods and innovations for obtaining information to spare the civilian population
The resources that a commander will have to deploy under the rules of constant care and precautionary measures will be heavily context specific. In densely populated settings where baseline data is unavailable or unreliable, a commander might want to rely on skilled personnel such as urban planners, engineers and demolition experts to assess the integrity of structures to determine their likelihood of incidentally collapsing onto civilians during attacks. These professions might also provide insights into a city’s flow (e.g., rush hour traffic patterns and how street layouts dictate a population’s urban movements) and whether electricity grids that serve enemy forces provide indispensable electricity to, say, hospitals.
Pattern-of-life analysts can help educate militaries in important ways that are also vital to reducing civilian harm. Rarely does urban combat not involve civilians and fighters living amongst each other in ways that make it difficult to distinguish who is who. Pattern-of-life analysis can help discern what behaviours civilians and enemy fighters share so as not to mistake the behaviour of a civilian necessarily as the behaviour of an enemy fighter. This type of analysis can also expose how civilians behave under the duress of enemy forces so as not to assume that civilians have enemy associations that do not exist.
Putting in place civilian harm complaint and investigation mechanisms, as well as ensuring they are properly resourced, have also been proven to improve civilian harm mitigation techniques. The emerging field of forensic architecture may also prove to be a useful precautionary measures tool.
In urban settings with cluttered buildings, populating and updating databases that contain the locations and markings of humanitarian organizations and protected objects and infrastructure can be vital to ensuring civilians and civilian objects do not become the object of direct attack or suffer incidental harm or damage. But even that is no easy task; militaries may have to find new ways to adapt to the fluidity of urban environments. General Stephen Townsend, who was the commander of Combined Joint Task Force in Iraq, explained from his urban combat experience how the targeting systems became ‘outdated quickly because the urban landscape was changing faster than we could update our imagery’.
There is another potential complication that arises when a commander wants to gather and update vast amounts of information from the civilian population. Employing, for example, automated decision algorithms and the collection of bulk communications data as precautionary measures tools can raise perplexing and sensitive privacy and other concerns for a civilian population. To that end, Peru’s 2004 IHL manual reminds us that ‘a distinction must be made between permitted and prohibited methods of gathering information’.
As cities continue to grow faster than ever before and as wars more frequently take place in urban settings, experts have called on militaries to improve how they engage and defeat their enemies. Alongside this shift, an equal emphasis and innovative thinking needs to promote the ability of commanders to gather and analyse large amounts of information about civilian populations and the cities they live in. This is both a military imperative and a vital part of a military’s constant care and precautionary measures obligations under IHL. Some militaries are doing this better than others, and it is impossible to expect a commander to have access to the infinite amount of data that a city contains. There remains considerable room, however, for improving the ways warring parties prepare themselves to spare civilians the harms of war in urban environment where information is constantly in flux and, on top of that, where a high operational tempo means that commanders may often find it difficult to overcome information gaps that emerge during the heat of battle.
This blog post was written in the author’s personal capacity.
Upcoming posts in the Joint Series include:
- Sieges, Evacuations and Urban Warfare: Thoughts from the Transatlantic Workshop on International Law and Armed Conflict, Laurie Blank (Clinical Professor of Law, Director of the Center for International and Comparative Law and Director of the International Humanitarian Law Clinic at Emory University School of Law) – EJIL Talk!
- Better Safe than Sorry: Transferring Detainees Safely to Coalition Partners, Tilman Rodenhauser (Legal Advisor, ICRC) – Lawfare
- Medical Care in Armed Conflict, Marco Sassoli (Director of the Geneva Academy of International Humanitarian Law and Human Rights and Professor of International Law at the Faculty of Law of the University of Geneva) – ICRC Humanitarian Law & Policy Blog
- And more….!
For past Transatlantic Workshop Series
DISCLAIMER: Posts and discussion on the Humanitarian Law & Policy blog may not be interpreted as positioning the ICRC in any way, nor does the blog’s content amount to formal policy or doctrine, unless specifically indicated.