Introduction
It would be unrealistic and naive to imagine a “clean” warFootnote 1 with no civilian casualties. Just the opposite is commonly observed in the types of conflict prevalent today, often non-international in nature, involving organized armed groups disinclined to respect the customs of war and taking place, in some cases, in the territory of a failed State. It is in these types of situation that civilians suffer the most as a result of the fighting and are sometimes even deliberately targeted, particularly in urban environments, as is the case now in Syria, Gaza, Yemen and eastern Ukraine.
In this context, it is tempting to point to the inadequacy of the rules of international humanitarian law (IHL), which do not a priori take into account the new types of “combatants” and their asymmetric forms of warfare or consider the consequences of the use of new technologies. In my view, however, two preliminary observations need to be made before making such a judgement. First, IHL is a rich body of law, open to interpretation, and we would do well to explore all its possibilities before disparaging it. Second, it should not be forgotten that the failure of a party to a conflict to respect IHL does not release the other from its obligations in this respect, and this tends to reinforce a virtuous circle of practical IHL implementation.
As a legal practitioner for the French armed forces, I would like to share some reflections on the practical difficulties of implementing the principles of IHL (especially distinction, proportionality and precaution) in today's conflicts, particularly in urban areas, and the solutions that would allow the parties to act according to these principles and in keeping with the spirit of IHL.
Difficulty of applying the principle of distinction in urban areas
The very nature of cities makes it complicated for armed forces to apply the principle of distinction in respect of objects.Footnote 2 Cities are, by definition, made up of countless civilian objects (dwellings, shops, schools, hospitals, etc.). Objects that are, by nature, military objectives (barracks, air bases, headquarters, etc.) are sometimes interwoven in the fabric of the city, whether as a result of urban extension or by design. In non-international armed conflicts (NIACs) in which regular State armed forces fight organized armed groups (OAGs), the military infrastructures of the latter tend to be very difficult to identify, because they are often pragmatically located in originally civilian buildings. The first difficulty for armed forces deployed in a conflict in which the adversary is operating in a city is therefore to pinpoint, based on various forms of intelligence, the exact location of enemy military facilities and the objects that, by their nature, could be considered essentially military objectives.
Armed forces are immediately confronted with the problem of dual-use facilities,Footnote 3 particularly so in urban warfare. These are facilities that are civilian by nature but are used for both military and civilian purposes and can therefore be regarded as military objectives. There are many and varied examples, including bridges, roads, power stations and electricity distribution and transmission networks. The challenge for the armed forces, in terms of the intelligence function, is to gather as much information as possible on what use is being made of the facilities and try to determine, as accurately as possible, how this use is shared between the enemy forces and the civilian population. In the case of an oil refinery, for example, it would be necessary to establish the amount of fuel being directly supplied to enemy troops and the precise impact the destruction of the plant would have on the conduct of the enemy's operations. The share of the output destined for civilian use, to provide fuel for vehicles, heating, etc., must then also be calculated. The analysis reveals whether the facility is indispensable to the survival of the population; if so, it is entitled to special protection under IHL.Footnote 4 The analysis also determines the direct and indirect risks that the loss of these resources would have on civilians. Furthermore, the findings of the analysis would be taken into account in determining proportionality, as will be seen below.
Some questions arise in respect of the definition of objects indispensable to the survival of the civilian population provided in Article 54 of Additional Protocol I to the Geneva Conventions (AP I). Article 54 is primarily concerned with food supplies, and its main purpose is to prohibit the use of starvation as a method of warfare. The scope of application of this rule of customary lawFootnote 5 is, however, broader than might first be supposed. The Preparatory Committee for the establishment of a permanent international criminal court recognized that the term “starvation” did not just mean killing by depriving of nourishment and water, but also covered malnutrition, illness and disease caused by a lack of food, medicines and other essential commodities.Footnote 6 Moreover, the list of protected objects given in Article 54(2) of AP I is not exhaustive, as it is preceded by the term “such as”. In the case of conflicts taking place in cities, Article 54(2) must therefore be considered to have a broad meaning and to cover food storage and distribution areas and the water supply system. However, the survival of the population can also depend on other systems, such as the electricity network, which is required to heat homes and power hospitals, for example, but can also make a direct contribution to the adversary's capabilities (communications in particular). The challenge for armed forces is to assess the extent to which the civilian population depends on these dual-use facilities. In order to do this, systemic analyses carried out for intelligence purposes must be precise and fact-based, so that the direct and indirect impact of the planned destruction can be accurately measured.
In the case of conflicts in urban areas, there are two examples that clearly illustrate the practical difficulties currently facing armed forces. The issue of targeting communication systems is particularly complex. In the conduct of combat operations, the transmission of information through chains of command and control (abbreviated as C2) is vital. The destruction or neutralization of C2 systems therefore provides a real and precise military advantage. However, such networks, whether electrical, electromagnetic or digital, are most commonly dual-use facilities. It must therefore first be determined whether they can be considered indispensable to the population, taking into account that they are required, among other things, for the operation of the emergency response chain. Destroying or disabling phone towers used by the adversary, for example, is often a complicated matter. The issue of cyber-targeting in order to neutralize C2 systems is even more complex, as the effects are difficult to measure and control.Footnote 7
The second example concerns schools, which are, by nature, civilian objects and of crucial importance to the civilian population in the short and long term, and for the country's development. The use of school premises for military purposes by regular armed forces or organized armed groups, which is becoming increasingly frequent, can turn them into military objectives, because they do not enjoy the special protection granted to certain sites, such as places of worship and cultural property.Footnote 8 Parties to a conflict often use schools deliberately, seeking to shield themselves in a civilian facility and also as a means of discrediting the adversary in the eyes of public opinion, when it is driven to attack a school. Sometimes, however, armed forces are compelled to use schools to temporarily store weapons and equipment or to provide shelter for their soldiers. This solution is clearly only conceivable when the premises are not being used for educational purposes. Many humanitarian organizations are justifiably concerned about the use of schools and other places of education for military purposes. One of them is the Global Coalition to Protect Education from Attack (GCPEA),Footnote 9 which in 2013 drew up the Draft Lucens Guidelines.Footnote 10 The purpose of the guidelines was to have States incorporate, first in their domestic legislation and eventually in the Geneva Conventions and their Additional Protocols, a specific ban on the use of educational establishments for military purposes. This initiative, however laudable, focuses on requirements that are not easy to reconcile: on the one hand, the protection of objects that are especially exposed and fragile, particularly in urban warfare, and whose destruction would damage prospects for future development; and, on the other, the leeway required by armed forces conducting modern operations against adversaries who have no qualms about taking advantage of the fact that the other party complies with the rules of IHL. Adding new rules to the existing body of IHL would undoubtedly be overly restrictive, as it would often be difficult to comply with them for reasons of operational effectiveness.
The specific characteristics of urban areas could also lead to a rekindling of the debate about cities as targets in themselves and the lawfulness of siege warfare.Footnote 11 Cities have always been an important factor in the power game, and in an armed conflict the taking or destruction of them can become a symbol or an end in itself. This can result in the parties ignoring basic principles of IHL, in particular the obligation to distinguish between military objectives and civilian objects. In the ongoing conflict in the Levant,Footnote 12 it is clear that efforts to take and maintain control over cities (Mosul, Raqqa, Ramadi, Palmyra, etc.) are driven by a combination of military, strategic, political and symbolic aims. This is particularly true in the case of Daesh, with its ideology of State-like territorial domination, but also for the States involved, which are seeking to restore their sovereignty. In this drive to take cities, sometimes involving prolonged sieges, the parties often overlook their obligation to take into account the provisions of the Geneva Conventions,Footnote 13 which provide for agreements to be made to arrange for the evacuation and exchange of the sick and wounded and to facilitate the passage of humanitarian aid.
Distinguishing between individuals is more complicated in urban areas
While the nature of cities complicates the practical application of the principle of distinction in respect of objects, the difficulties are no fewer when it comes to distinguishing between individuals and identifying the adversary, particularly in NIAC situations. As the notion of “combatant” does not exist in the rules applicable to NIACs, the solution lies solely in determining the direct participation of individuals in hostilities.Footnote 14 The notion of direct participation in hostilities is not properly defined in the Additional Protocols, but the consequences are clear: civilians lose the protection accorded to them for the duration of their direct participation in hostilities. The International Committee of the Red Cross (ICRC) provides some insight into this notion of direct participation in hostilities in the Interpretive Guidance published in 2009,Footnote 15 which explains that it refers to specific acts meeting three cumulative criteria: a threshold of harm, a direct causal link between the act and the harm likely to result from it,Footnote 16 and the belligerent nexus (the act must be carried out in support of a party to the conflict and to the detriment of another).
In spite of these clarifications, however, the definition of direct participation in hostilities remains unsatisfactory. It is, for example, a complicated matter to draw a clear line between direct and indirect participation, particularly in conflicts in urban areas, where the precise characterization of activities can be complex. According to the Interpretive Guidance, supplying weapons to organized armed groups is not an act amounting to direct participation in hostilities, unless they are directly delivered to the zone where armed operations are taking place.Footnote 17 In urban warfare, it can be difficult to establish the facts and determine what can be considered the indirect provision of weapons and what amounts to delivery for immediate combat purposes, given the blurred and changing boundaries of combat zones and their location in the midst of civilian objects and facilities. There is therefore a danger that all transportation of weapons and munitions could be considered direct participation in hostilities, making direct attack permissible.
It is important for armed forces not to target civilians who are taking an indirect part in hostilities, by supporting the “war effort”, for example. Economic and financial aid, political support, demonstrations, etc. do not amount to acts of direct participation in hostilities, and those who engage in such activities must not be targeted.Footnote 18
It should also be noted that the general protection granted to civilians in NIACs does not cover people fighting as part of an organized armed group – that is, a group with a certain level of organization that uses armed warfare, amongst other methods, to fight against one or more parties to the conflict. Armed forces must therefore make every effort to gather intelligence on membership of such groups, obtaining up-to-date information on participation in armed operations or meetings with influential members of the armed group, roles in the chain of command and control, the recruitment and training of fighters, etc.
In the case of the conflict against Daesh in the Levant, for example, the structure of the group means that the principle of distinction must be strictly respected, as Daesh is seeking to turn itself into a State and therefore endeavouring to acquire all the attributes of one, in particular by administrating large swathes of Iraqi territory.Footnote 19 It is therefore necessary to ensure the protection of civilians, even if they work, maybe under duress and against their will, in the administrations established by Daesh (banks, courts, law enforcement, etc.). In this context, direct participation in hostilities is not as clear-cut as in other conflicts, and attention should be paid primarily to distinguishing between the political wing and the armed wing of the non-State party to the conflict.
The situation is further complicated when, in today's so-called “asymmetric” conflicts, organized armed groups and even regular State forces seek tactical advantages in complex urban environments, using, for example, human shields, voluntary or otherwise. Human shields can be used to protect high-value targets (command centres), convoys and even whole cities when they become the last refuge of armed forcesFootnote 20 or are placed under siege in the course of a conflict. Evidently, in such cases, the challenge is to respect the traditional principles governing the conduct of hostilities, which allow a military objective to be attacked, provided that the attacker fulfils its obligation to take all feasible precautions and respect the principle of proportionality in order to avoid, or at least minimize, incidental losses. In the case of voluntary human shielding,Footnote 21 where civilians deliberately and voluntarily position themselves to protect a military objective by their presence, it could be argued that these civilians are abusing their status as protected persons. If it turned out that they had voluntarily chosen to act in this way, they could be considered, according to some sources, to be taking a direct part in hostilities for such time as their action lasted, and could therefore be deemed a legitimate target.Footnote 22 However, this view is not upheld in part of the doctrine,Footnote 23 particularly when the physical obstacle created by these individuals fails to reach the threshold of harm required for the act to qualify as direct participation in hostilities.Footnote 24
Practical difficulties common to both facets of the principle of distinction
There are considerably greater difficulties when the attacking State does not have troops on the ground to distinguish civilians from fighters directly participating in hostilities or to confirm or refute intelligence collected for that purpose. For reasons generally based on political considerations, States are reluctant to deploy troops on the ground.Footnote 25
In this context, one of the possible solutions, as I see it, would be to move conventional warfare out of cities, for example by limiting air strikes in urban areas and putting a very high-precision targeting system in place for cases in which it is decided that a strike must be carried out in a city.Footnote 26 This would involve planning air strikes on support “upstream” of enemy military operations, respecting the principles governing the conduct of hostilities, and wiping out the leadership of armed groups, using very precise intelligence and, in many cases, special forces to ensure highly targeted operations. In cities, the use of non-lethal means could also be promoted (psychological operations and perception-changing and cyber-activities), ensuring that the principle of distinction is fulfilled at all times.
Challenges of minimizing collateral damage in urban areas
The principle of proportionality calls for collateral damage to be minimized
The principle of proportionality is defined in paragraph 2(a)(iii) of Article 57 of AP I.Footnote 27 For the protection of civilians and civilian objects to be effective, it is necessary to ensure that all precautions are taken to reduce any incidental damage that could be caused by combat actions. Collateral damage is not prohibited by IHL, but it must be minimized. This is not only a legal requirement but also a strategic one, because a military force that causes civilian casualties could discredit the entire operation and undermine the legitimacy of the action taken.
In order to define what is meant by collateral damage, a semantic distinction must first be made, based on the degree of intentionality. Three categories can be identified, which are unforeseen, incidental and deliberate civilian harm.Footnote 28 The latter category does not fall within the definition of collateral damage, as it is always caused knowingly. In addition, as will be seen below, the implementation of a targeting oversight process, aimed specifically at reducing the risk of collateral damage, is in itself a guarantee that deliberate harm will not be done.Footnote 29 I will not therefore address the issue of deliberate harm, which results from the failure to take into account the principle of proportionality in the conduct of operations and to respect IHL in a broader sense.
Unforeseen harm can be caused, for example, by human errors or technical failures.Footnote 30 In this respect, the analysis of the US air strike carried out on 3 October 2015 on the Médecins Sans Frontières (MSF) hospital in Kunduz, Afghanistan, is particularly telling,Footnote 31 revealing an accumulation of human errors and technical failures: an AC-130 aircraft, diverted to provide air support to US and Afghan troops on the ground that had come under enemy fire, failed to implement standard “no-strike designation” procedures; the aircraft's communication system malfunctioned, preventing information exchanges with headquarters; the onboard electronic systems were degraded as the result of an alert, which forced the AC-130 to effect an emergency change of course; there was no visual target acquisition prior to the strike; the decision to carry out the strike was approved by a non-competent authority at headquarters; and the real coordinates of the site were not verified prior to the strike. In view of all these factors, the analysis concluded that the strike had caused unforeseen civilian harm.Footnote 32 In order to reduce this type of harm, armed forces must focus on improving their internal procedures and training.
Incidental civilian harm best illustrates the meaning of the military jargon term “collateral damage”, and it is with the aim of reducing this type of harm that armed forces put in place oversight processes. The risk of incidental harm is particularly high in urban areas, owing to the very nature of cities and the distinction difficulties described above. Incidental harm is, however, a foreseeable risk, and efforts to ensure the continuous improvement of targeting procedures, including the choice of weapons employed, should focus on strengthening studies that combine the gathering and use of intelligence, systemic analyses and technical studies on essential functions and infrastructure, particularly in urban areas (organization and weaknesses in water and electricity networks, organization of emergency services to deal with indirect incidental damage, etc.).
Indirect collateral damage is another issue to be considered. It is not always easy to determine the cause of such damage with scientific accuracy, particularly if the strike has a delayed or cascading effect. Here too, the question of the foreseeability of damage arises.
Processes adopted to minimize collateral damage
The practical implementation of the principle of proportionality by the armed forces is therefore based mainly on the adoption of strict standard targeting procedures, particularly for planned strikes. According to French doctrine, targeting is a rational, integrated process that involves identifying and then choosing the targets to be attacked, using a range of capabilities to achieve the desired outcome. The legal constraints established by IHL are fully integrated into every stage of the process. In assessing the legitimacy of the target, the principle of distinction is applied.Footnote 33 The next step, the collateral damage estimate (CDE), ensures that the principle of proportionality is respected.Footnote 34 This method, resembling a scientific risk analysis, takes into account the effects that weapons can reasonably be expected to have, and a level of responsibility is assigned to each level of risk.Footnote 35
This analysis is based on knowledge of the sites to be attacked. I will not return here to the question of establishing the certainty that the object is a legitimate military target, but once this has been determined, the structure of the site must be studied (type of construction, potential weaknesses, etc.) with a view to assessing the effects of the weapons to be used, as must its surroundings (protection of nearby objects, type of buildings and infrastructure in the area around the target, state of communication, water and electricity networks, actual use of all these objects and resources, etc.).
In the case of unplanned attacks (dynamic targeting or on-the-spot combat actions), armed forces take into account the principle of proportionality in two ways. First, the rules of engagement (ROE) require combatants to implement the principle of proportionality, indicating what means are authorized in different cases.Footnote 36 Second, providing military personnel with IHL training, however general, prior to or during deployment, is a means of ensuring that IHL principles and their implementation are properly understood and fully incorporated into the conduct of operations.
Practical limitations of implementation in urban areas
The limitations inherent in the targeting process have been outlined above in relation to indirect effects. It can be argued that in the “fog of war”, military planners and decision-makers cannot reasonably be expected to assess cascading effects, as the information needed to do so is never fully available in the required time.Footnote 37 This is particularly true in an inevitably complex, dense and changing urban setting. The complexity of the situation is further compounded by the deliberate use of civilian objects and human shields, often not readily visible at first sight, with a view to discrediting the attacker and “pushing” it into causing collateral damage.
Another issue to be addressed is whether proportionality should be assessed for a single strike or for a series of strikes,Footnote 38 or even the entire military campaign. Military decision-makers are more inclined to consider the effects of attacks as part of an overall plan, designed to achieve a final outcome, which is to defeat the adversary. There is therefore a natural tendency to put the focus on the principle of necessity and to weigh up potential collateral damage against an overall military advantage. The role of the legal adviser is crucial here in reconciling the different operational and legal requirements.
Modern types of conflict (with the use of means of asymmetric warfare in urban areas, including improvised explosive devices and human shields, as mentioned above) and the frequently adopted “no boots on the ground” approach make it difficult to apply the principle of proportionality in targeting processes in urban settings. Even when armed forces have deployed soldiers in the territory where the conflict is taking place, the extent to which military commanders are obligated to expose their own forces to danger in order to limit collateral damage is a question open to debate. Arguably, limiting the risks to which combatants are exposed can in itself provide a concrete military advantage.Footnote 39 This assertion is, however, only valid if the technological resources available to these same armed forces (tracking radars, drones, etc.) do not provide them with sufficient protection from the risks posed.Footnote 40
Repeated tragedies involving civilians in urban conflict areas have prompted calls for changes to the law and practice, with a view to perhaps prohibiting collateral damage even when it is not unlawful or at least urging States to go “beyond what is required by IHL”.Footnote 41 This understandable humanitarian concern is amplified by the media, which often provides greater coverage of conflicts in urban settings, partly for practical reasons of access, but also because of the impact of such events on public opinion. This extensive media coverage sometimes masks the efforts of armed forces to reduce collateral damage and to implement protective measures.
Precautions in attack
The general rule on precautions in attack is found in Article 57 of AP I, which provides in general terms, in paragraph 1, that “[i]n the conduct of military operations, constant care shall be taken to spare the civilian population, civilians and civilian objects”.Footnote 42 In addition to the principle of proportionality established in paragraph 2(a)(iii), as explained above, Article 57 contains rules concerning verification that the target is a military objective, the choice of means and methods of attack, the possibility of interrupting an attack, choosing between objectives, and advance warning.
Importance of intelligence gathering
Paragraph 2(a)(i) of Article 57 indicates that it is necessary to “do everything feasible” to verify that the target is a military objective,Footnote 43 which entails gathering complete and accurate intelligence. This can take various forms: imagery (aerial, satellite), radio, electromagnetic, human, and so on. As far as possible, information should be gathered from several different sources (for example, images corroborated by human intelligence) and should be recent (satellite images no more than several days old, communications recently intercepted, etc.). Generally, military intelligence services use standard codes to “track” the accuracy of intelligence. There are, however, some practical difficulties in terms of intelligence, which is the cornerstone of a precise targeting process that respects the applicable rules.
The gathering of relevant intelligence requires well-developed, reliable and even redundant technical resources. Not all States, for example, are able to acquire high-quality satellite images, and military satellite capabilities vary significantly from one State to another (in terms of image resolution, in particular). The quality of intelligence is enhanced by the physical presence of human gatherers (conventional or special forces) in the theatre of operations. In the case of “no boots on the ground” operations, this human intelligence is lacking, and it is unrealistic to think that technology alone can resolve this deficiency. When the intelligence comes from “local” sources (allied forces, rebel fighters, etc.), the party involved must be able to verify the accuracy and objectivity of the information before taking responsibility for a strike based on the information received.
In a coalition military operation, it must be possible to share intelligence and for the intelligence provided by the coalition or States belonging to it to be “questioned” and controlled by national means of verification. Difficulties exist on two levels: on a technical level (means of verification at least equivalent to those of allied forces) and in terms of classification (information exchange agreements in order to access all target documentation).
The importance of today's open sources for obtaining intelligence should not be underestimated: the Internet (Google Maps), NGOs operating on the ground, think-tank reports, journalists, etc. While such sources should be fully taken into account, closer monitoring of their accuracy is nonetheless required, and this is a limitation that today's armed forces must be aware of.
Issues relating to choice of weapon in urban areas
Paragraph 2(a)(ii) of Article 57 of AP I is concerned with the choice of means and methods of warfare,Footnote 44 which must be made with a view to limiting collateral damage. Even before the targeting process, armed forces are limited in their choice of weapons by the provisions of international law. For States party to the conventions concerned, the use of certain weaponsFootnote 45 is regulated or, in some cases, banned. For example, a State such as France could not include the use of cluster bombs in a deliberate targeting process, because it has ratified the Oslo Convention.Footnote 46
Two precautions are generally taken when it comes to choosing permitted or regulated weapons. First, the most precise weapon should be chosen.Footnote 47 This not only ensures the application of the principle of proportionality, but is also a guarantee of technical, operational and even financial effectiveness. Second, with a view to minimizing potential collateral damage, all the effects of the weapon must be taken into account: the effects of the actual impact and also the blast and fragment effects caused by the impact. These effects are the subject of very precise scientific studies, using input from the evaluation of damage caused by past strikes (known as battle damage assessment), which ensure that predictive calculations are as accurate as possible.
Lastly, I would like to comment on the current trend, undeniably commendable in its intention to increase protection for civilians, towards seeking to impose more constraints on armed forces which they would find difficult to implement for operational reasons and which would be applied indistinctly to all armed forces, without taking sufficient account of efforts made to respect the relevant provisions of IHL. A telling example is the current mobilization against the use of explosive weapons in urban areas. While there is no question about the destructive effects of explosive weapons in densely populated areas, such effects are largely due to a lack of proper oversight in the use of these weapons or to too broad an interpretation of the principles of distinction and proportionality. In truth, strict application of the principle of precaution will, in the vast majority of cases, rule out the use of this type of weapon in populated areas.
Establishment of specific procedures and precautions
Article 57 also requires procedures to be put in place for suspending or cancelling an attackFootnote 48 and for choosing between several targets.Footnote 49 The fact that there are highly standardized targeting procedures and that target dossiers are studied before choices are made by decision-making authorities is an implicit acknowledgement of the possibility of choosing between targets and also allows for the implementation of decision-making processes taking into account the level of risk involved. In addition, the use of video footage from drones, for example, makes it possible to follow a strike live, which means that it is now easier to suspend a strike when a change in the environment is detected (for example, the presence of civilians).
In the process of approving a target pack, other precautions can also be taken, whether technical (aircraft angle of attack, use of weapons with a delay fuse) or of a more general nature (night strikes, for example). The timing of the attack is an essential factor in the application of the principle of precaution. With a view to choosing the right time, accurate data must be collected about the situation on the ground, in particular information about the way of life and customs of the civilians at risk.
Current doctrinal developments focus on what is referred to as “full-spectrum targeting”, including both kinetic and non-kinetic (cyber, influence, etc.) operations, whether carried out separately, sequentially or concurrently. Different means are ultimately chosen and employed, taking into account the risk of collateral damage and fulfilment of the principle of precaution, with a view to achieving the same desired end-state while minimizing the risks.
The complex question of giving advance warning
In accordance with the principle of precaution, the attacker can also urge civilians to leave the combat or strike area by issuing an advance warning, as provided for in paragraph 2(c) of Article 57 of AP I.Footnote 50 However, for these warnings to be effective, they must meet numerous criteria:Footnote 51 a range of means of communication must be used (television and radio broadcasts, telephone calls, leaflets, etc.) to ensure that those at risk from attacks are reached; the warnings must indicate the time and place of the attack as clearly as possible; they must give sufficient time to react; they must contain precise instructions on how to avoid the attack (where to go, evacuation measures, etc.); and the messages must be credible. Given these constraints, in my view, the level of effectiveness of such warnings is very low, particularly in certain circumstances when the territory is small and there is intense fighting everywhere.Footnote 52 It should not be forgotten either that a party to a conflict has an obligation to give advance warning “unless circumstances do not permit” – that is, provided that the warning does not put the party's own forces in danger and/or spoil the element of surprise.
In modern asymmetric conflicts, however, the element of surprise is not always such a decisive factor. It would be easier for armed forces with advanced technological resources to fulfil the obligation to give advance warning, even when they do not exercise effective control over a territory. This was a conclusion that emerged from a United Nations (UN) document published in 2006 on the situation in the occupied territories, in which the Secretary-General suggested that the greater the resources the attacker has at its disposal (communication and surveillance systems, in particular), the greater its obligation to ensure that the principle of precaution is fulfilled by giving advance warning, particularly in conflicts taking place in urban areas.Footnote 53
Precautions against the effects of attacks
The general rule concerning protection against the effects of attacks is provided in Article 58 of AP I.Footnote 54 This article is not about regulating the behaviour of combatants in attacks, but rather encompasses all the measures that all powers should take in their own territory or territory that they control in order to protect the people there. They are therefore preventive measures aimed at ensuring effective protection for civilians. The requirement to take precautions against the effects of attacks is a rule to be taken into account in peacetime, but also takes on particular significance in the planning of attacks and in the conduct of operations. It is not, however, an absolute rule, as Article 58 starts by providing that the parties to the conflict should implement it “to the maximum extent feasible” and specifies that it refers to territories and populations “under their control”. As indicated in the ICRC's study on customary international humanitarian law, the rule concerning protection from the effects of attacks can be divided into two sub-rules, specifically the need to locate military objectives away from civiliansFootnote 55 and the need to remove civilians from the vicinity of military objectives.Footnote 56
Locating military objectives away from civilians
The principle stated in paragraph (b) of Article 58 of AP I involves taking preventive or protective measures in peacetime. It is therefore necessary for the party that can be referred to as the “defender” to ensure that certain types of building intended for military use (barracks, ammunition depots, etc.) are not erected in certain places (specifically in towns or cities), to avoid locating military objectives in certain areas and to keep dangerous points away from civilians. This obligation is, however, limited to what is feasible. Many States interpret this notion of “to the maximum extent feasible” as meaning that they have to take the precautions that are materially or practically possible, taking into account the prevailing circumstances, including military considerations (no questioning of the overall defence strategy, for example) and humanitarian considerations (the measures should not create too much hardship for civilians). It should be noted that demographic changes and the extension of urban areas can prove an obstacle to locating military facilities outside towns and cities in the long term. The principle does not apply to civilian objects that can also be used for military purposes (stations, airports, etc.).
It is worth noting that this obligation to take precautions is harder to fulfil in NIAC situations, particularly by organized armed groups, which may be of recent or spontaneous creation and have no pre-existing military facilities. Such groups are therefore limited in the choice of where to locate their military facilities and may have limited resources, making them unable to implement preventive measures.
It is easier to implement the principle of separation in the case of fixed objectives than in the case of mobile ones (troops, vehicles, etc.) because, for the latter, precautions must be taken while operations are in progress. The challenge for parties to a conflict is clearly identifying their military forces without compromising the safety of those forces. Precautions could therefore include decisions to limit the passage of convoys through cities (bypassing them, where possible), setting up camps and positions outside cities, ensuring zones are clearly marked, restricting access, training military personnel so that they can inform civilians, etc. Evidently, the main risk of such identification measures is that the party is giving the adversary a clear indication of the location of its military objectives. The solution is to carry out a case-by-case analysis in order to weigh up the military expediency of locating troops in a specific, clearly identified place and the threat that this poses to those forces if the location is too visible.
Looking at this issue from the perspective of the party that can be referred to as the “attacker”, it is important to remember, first of all, that it is not unlawful to target military objectives if the defender fails to take the necessary precautions or if it deliberately uses civilians to shield its military operations. In such cases, the attacker must take precautions in attack and respect the principle of proportionality, even if the defender does not comply with the rules of IHL.Footnote 57
One solution that can be adopted by the attacker with a view to implementing the principle of precaution against the effects of attacks is to drive the adversary towards isolated areas. This was, in some respects, what France sought to do in Operation Serval, conducted in Mali from January 2013 onwards at the request of the Malian State.Footnote 58 Faced with an adversary intent on taking control of the country's cities, including the capital Bamako, the French armed forces sought to repel the enemy by engaging in minimal urban combat, limiting the scope and number of operations through the deployment of special forces.Footnote 59 Local forces were given the responsibility of retaking the cities, while the special forces focused their efforts on pursuing the armed groups, monitoring and controlling supply points to prevent the enemy from replenishing supplies in the cities and maintaining a show of force in urban areas while at the same time continuing to fight the enemy in isolated areas. This strategy does, however, have its limitations, particularly when conflicts extend over a long period, as it can result in all of the fighting taking place outside urban areas, turning the cities into a safe haven for the enemy.
As it is so difficult, if not impossible, for defenders and attackers to separate military objectives from civilians, they might find it preferable to move civilians away from military objectives.
Moving civilians away from military objectives
Compliance with paragraph (a) of Article 58 also requires, in the case of the defender, the implementation of preparatory measures such as the construction of dedicated infrastructure for civilians (shelters, buildings, camps, food depots, etc.), the involvement of civil society (regular training for civil defence services) and the regular dissemination of information and warnings. Once again, these measures are easier for States to implement than armed groups, particularly when the latter's structures are weak.
The attacker can also contribute indirectly to moving civilians away from the fighting by establishing air and maritime exclusion zones, which involves delimiting areas that the enemy is not permitted to enter and where civilians are relatively safe, although violations of such zones can lead to fighting that will have repercussions for civilians. In any event, the establishment of exclusion zones raises the question of access to humanitarian aid, which the attacker must also deal with, in addition to complying with overflight/passage prohibitions. The difficulties involved in managing such zones should not be underestimated, and it should be taken into account that they require significant resources for distinction and information activities. This difficulty is compounded when humanitarian organizations or States delivering aid do so without first notifying the authorities responsible for managing the zone, a problem NATO forces had occasion to observe during operations in Libya in 2011.
Another precaution is evacuation operations that can be carried out directly by the attacker, particularly if it controls the conflict area. Evacuation of the nationals of a particular StateFootnote 60 not only raises issues with regard to jus ad bellum,Footnote 61 but also poses problems of discrimination. Is it legitimate, or even lawful, for a party to evacuate only its own nationalsFootnote 62 and do nothing for foreign nationals suffering abuse, for example? There is, however, no positive obligation in law to evacuate civilians from a country affected by conflict. If civilians had to be evacuated by an attacker in a conflict area, the operation would have to meet the same criteria as those outlined above for defenders, although to a lesser extent, and its implementation would therefore present the same practical difficulties.
The evacuation of civilians by the defender raises numerous questions. First, does the obligation extend to all civilians or just some categories of people entitled to special protection (women, children, the sick, the disabled, etc.)? Apart from the fact that it is very difficult for a party to a conflict to distinguish between people in this way, it should also be noted that the legitimate desire to protect the weaker sectors of the population may conflict with the principle of respecting family unity, which is to be found in numerous provisions of IHL and human rights law.
If it is assumed that civilians in the immediate vicinity of the fighting should be evacuated, the question also arises of how to precisely delimit the boundaries of the combat zone and what to do when the territory is very small. A number of questions arise in relation to the place that civilians are evacuated to, normally a place that they are familiar with and that poses no danger to them. Should civilians be informed in advance of the place they are being evacuated to, so that they can make an informed choice about whether or not to go? How can safety zones be maintained without putting too great a strain on already stretched military resources? And, once again, is this feasible in a small, conflict-ridden territory?
Lastly, with regard to the timing of the evacuation, here again caution is paramount. Evacuating civilians during the fighting can be dangerous, particularly in urban areas, but waiting until the fighting is over can prove disastrous. In any event, parties to a conflict are only required to do what is “feasible”, and the precautions to be taken in evacuating civilians should not go beyond the point where their lives are made difficult or even impossible.
It is important to note that there are provisions prohibiting the forced movement of civilians, which are applicable both in international armed conflicts (IACs) and NIACs.Footnote 63 In both types of conflict, forced displacement is prohibited unless the security of the civilians involved or imperative military reasons require them to be moved.Footnote 64 Even when permitted, the movement of civilians is limited, in the case of IACs, to displacement within the occupied territory, unless materially impossible, and in the case of NIACs, within the national territory. Clearly, the imperative military reasons that permit an exception to this rule do not include the movement of civilians as a means of persecution. The most sensitive issue here, however, is evaluating the safety of civilians to determine whether it is necessary to evacuate them. Who should make this assessment, and what criteria should be used?
Indirect contribution to the protection of civilians
With a view to ensuring the protection of civilians, IHL provides for the possibility of establishing protected areas for the purpose of moving the inhabitants of cities away from the fighting.Footnote 65 There are three types of protected area: “hospital zones” for wounded and sick members of the armed forces (Article 23 of the First Geneva Convention), “hospital and safety zones and localities” for civilians entitled to particular protection (Article 14 of the Fourth Geneva Convention) and “neutralized zones” for people hors de combat (Article 15 of the Fourth Geneva Convention). Parties to a conflict also have the possibility of declaring a place a “non-defended locality” (Article 59 of AP I) and establishing agreements to demilitarize certain zones (Article 60 of AP I). Although there is no provision for such zones in the law applicable to NIACs, it is clear that an area which contains only wounded and sick people, medical and religious personnel, humanitarian relief personnel and civilians may not be attacked, because there are specific rules applicable in NIACs protecting these categories of people.Footnote 66
These provisions list the rights and obligations of the parties in great detail, with a clear distribution of responsibilities for the protection of the people gathered in such zones. However, the sine qua non condition for their effectiveness is that they are established with the full knowledge and express consent of all the actors.
The advantages of establishing protected zones are patently evident: they allow a clear distinction to be made between military objectives, on the one hand, and civilians and civilian objects, on the other, and facilitate the arrangement and delivery of humanitarian assistance. However, there are also numerous risks, as they concentrate large numbers of defenceless civilians in one place. In addition, if the agreement between the parties is deficient or flawed in some way or there are insufficient resources to protect the zones, the consequences could be tragic. There is no forgetting the case of the Srebrenica enclave, a “safe zone” established unilaterally by Resolution 819 of April 1993,Footnote 67 because the UN Security Council refused to engage in talks with the Republika Srpska. The fact that there were not enough UN troops (600 Blue Helmets) to protect it and the many other shortcomings revealed since resulted in the infamous massacres of July 1995.Footnote 68
In the context of the Syrian conflict, there have been renewed calls for the establishment of “free zones”Footnote 69 on the Turkish border in the north and/or on the Jordanian border in the south, as sanctuaries for refugees. The intention is commendable, and such an initiative would allow certain principles of IHL to be fulfilled. There are, however, many unresolved issues: how can an agreement be secured between the belligerents, and between which parties would it be established (there are several NIACs taking place simultaneously in SyriaFootnote 70 )? What can be done to prevent such zones from being misused, for example, as safe havens by rebel groups? How can the implementation of these humanitarian principles be achieved in a conflict in which the modus operandi of the different parties involves deliberately targeting civilians?
Protection of civilians by armed forces can also be included in the conduct of operations as a general guideline and not solely in the implementation of the principle of precaution against the effects of attacks. This would involve a comprehensive approach, including support and protection measures, to ensure that civilians do not suffer unduly from the consequences of the fighting. It would include so-called civil-military activities, aimed at rehabilitating infrastructure, improving access to health-care services and directly providing medical care to civilians through military facilities (hospitals and medical personnel). This comprehensive vision, with a long-term focus on rebuilding a State and restoring its institutions, must also be accompanied by measures aimed at protecting civilians from violations and ensuring that perpetrators do not enjoy impunity. The military must be permitted to use force and intervene in order to prevent and end human rights violations and capture those responsible, while at the same time respecting the sovereign prerogatives of the State in whose territory it is intervening.
Conclusion
The conduct of military operations in urban areas today presents a daunting challenge. Efforts to implement the principle of distinction effectively and reduce collateral damage must be radically increased. This is particularly complicated when there are no troops on the ground, and realistically, technology alone cannot be expected to overcome all the difficulties posed. Must this reasoning then be taken to its logical conclusion, which would be that the effective implementation of the principles of IHL requires a physical presence on the ground and therefore the deployment of armed forces in massive numbers to the territories where conflicts are taking place, just the opposite of today's prevailing “no boots on the ground” approach? This would, in fact, also be a sound application of the principle of military necessity.
The difficulties encountered in implementing preventive and protective precautionary measures often arise from the collapse of State structures, whether the cause of the conflict or one of its consequences. A comprehensive vision with a long-term and not exclusively military focus should therefore prevail in this type of situation, with a view to rebuilding institutions and securing the future. This reflection is in keeping with the current line of thinking that has emerged in the updating of urban warfare doctrines.Footnote 71
French doctrine on urban operations,Footnote 72 which has undergone significant developments since 2010, emphasizes the fact that a city is a space replete with restrictions and risks. The “Three Block War” conceptFootnote 73 reflected in this doctrine highlights the fact that, in cities, military units have to engage successively or simultaneously in enforcement, peacekeeping and humanitarian operations, compressed in the same time and space. The implications are threefold. First, integration and cooperation with civilian actors is necessary, particularly with regard to protection. Second, it is vital to maintain essential services and structures indispensable to restoring governance, as part of a comprehensive approach. Lastly, the strategy must include the means to isolate the adversary from civilians, particularly the urban population, by limiting its resources, undermining its freedom of action and conducting information and influence operations.
The implementation of such principles can then contribute to ensuring protection for civilians in urban areas and application of the rules of IHL concerning precautions against the effects of attacks.
