Urban Siege Warfare: A Workshop Report
[Editor’s note: The views expressed in this piece are those of the authors and not necessarily those of the ICRC, PILAC, or other participants.]
Siege is a well-established practice in warfare. It seeks to compel surrender, reduce adversary resistance, or lay the ground for an assault by encircling or otherwise isolating an area and cutting off essential supplies to it. From the razing of Jericho to the sack of Magdeburg during the Thirty Years’ War, the 900-Day Siege of Leningrad during the Second World War, and more recently the Siege of Marawi in the Philippines’ fight against militants affiliated with so-called “Islamic State” (IS), many battles have been waged through extensive deprivation and suffering, including starvation. In the current conflict in Ukraine, it has been widely reported that Russian forces besieged the city of Mariupol and Chernihiv.
Siege itself is not prohibited under the law of armed conflict and it is even permissible to seek to starve enemy forces into submission (U.S. Department of Defense Law of War Manual, §5.19.1). However, sieges of urban areas necessarily entail extraordinary harm to civilians and fighters placed hors de combat. While it is clear that the law of armed conflict applies to urban siege operations, the precise content and contours of the applicable law remain unclear and subject to interpretive disputes. Further, there are practical challenges to align an urban siege operation to applicable rules as various strategic, operational, and tactical levels of decision-making are interlinked in the process of capitulation.
The Lieber Institute, together with Harvard Law School’s Program on International Law and Armed Conflict and the International Committee of the Red Cross, co-hosted a workshop in Spring 2022, to address these legal challenges by examining urban sieges through historical examples and contemporary experiences from a range of military, civilian, and humanitarian perspectives. This post highlights key contentious issues the workshop identified and discussed.
Multi-Level Linkages in Siege Decision-Making
Unlike establishing a naval blockade, siege operations do not require that any legal preconditions be met. Indeed, siege has been variously described as a military tactic, a method of warfare, or an operational strategy. There are different kinds of siege, ranging from complete encirclement aimed at starvation to “porous siege,” where people and goods could move in and out (i.e. incomplete encirclement), and tactical options as “situationally dependent implements of war.” As one workshop participant observed, it is essentially “a game of dominance and resource consumption.”
As a response to an adversary’s action, urban siege is often activated when forces seek refuge in an urban environment. Unlike siege warfare conducted in the past, strategic siege has become a rare event in modern warfighting. Rather, sieges today are operational and tactical tools employed to take advantage of the conditions the urban environment present for military exploitation. As Anthony King argues, the size and posture of military forces equipped with more sophisticated precision munitions shape contemporary urban warfare. It is estimated that siege operations small in scale relative to overall urban areas are likely to make localized sieges a central feature of current and future urban warfare.
Still, the workshop did not conclude that current urban siege is devoid of strategic significance. Besieging forces may exact strategic gains by undermining popular support for the enemy’s war efforts and by avoiding or delaying the need for costly and destructive urban assaults. Paradoxically, however, localized urban siege could also strengthen the enemy force’s control over the civilian population due to increased dependency for sustenance and livelihood. The enemy force could even deliberately expose the civilian population to danger to strategically discredit the encircling force by causing the latter to inflict more civilian casualties.
These strategic considerations weigh heavily in the context of counter-insurgency operations, where winning the support of a local population has greater policy significance. Indeed, in the Siege of Marawi City conducted from May to October 2017, the Philippines armed forces fought IS-affiliated militants and their systematic propaganda campaign simultaneously. This consideration led Philippines armed forces to prioritize the evacuation and protection of civilians, while restricting the use of explosive weapons where civilians were present and ensuring transparent communication regarding the realities and dilemmas of an urban siege operation to limit loss of legitimacy.
Siege and Starvation as a Method of War
When siege is designed to isolate an urban area by cutting off supply, it is likely to cause starvation. Article 54(1) of Additional Protocol I to the Geneva Conventions prohibits starvation of civilians as a method of warfare. Article 54(1) is considered to have crystallized into customary international law (U.S. DoD Law of War Manual, §17.9.2; Customary International Humanitarian Law Study, r. 53). However, neither incarnation of the rule prohibits siege operations per se, as long as they are intended to achieve a military objective (including starvation of enemy forces) and not to starve a civilian population.
The prohibition of starvation of civilians imposes significant restrictions on the way in which siege operations may be conducted. However, the degree of its impact depends on how it is interpreted in the context of siege.
First, there is a view that the prohibition of starvation under Article 54(1) of Additional Protocol I must be interpreted in conjunction with the rest of the article. It provides specific prohibitions on the attack, destruction, removal or otherwise rendering useless objects indispensable to the survival of the civilian population including foodstuffs, agricultural areas, and water facilities. In other words, the besieging force is prohibited from attacking these objects, but is not otherwise held liable for the infliction of starvation as an outcome. This focus on transitive measures with the effect of causing starvation has the benefit of circumventing the difficulties with tracing the specific cause of starvation in a particular context as it is often a consequence of a gradual process where a combination of pre-existing local conditions and intervening factors is at play.
However, as Tom Dannenbaum discussed elsewhere, this interpretation stands apart from more traditional views that siege is prohibited only if it is specifically intended to starve civilians, or alternatively, when it is reasonably expected to cause starvation indiscriminately such that the magnitude of civilian starvation is clearly excessive in relation to the anticipated military advantage gained from enemy submission (U.S. DoD Law of War Manual, §5.20.2). Even these narrow interpretations create significant barriers to isolating an urban area from life-sustaining supplies, where enemy forces are entrapped together with many civilians, without in parallel pursuing compensatory measures, such as establishing humanitarian evacuation corridors and allowing for humanitarian relief operations.
Second, there is a question regarding whether the besieging force is required to provide humanitarian corridors to evacuate civilians from besieged areas. Under Article 17 of 1949 Geneva Convention IV, States have a limited obligation to “endeavour to conclude local agreements for the removal from besieged or encircled areas,” but this obligation only applies to certain classes of people, such as those wounded, sick, infirm, aged persons, children, and maternity cases. And belligerent parties are required only to make attempts to conclude an agreement. They do not have to reach an agreement if other considerations overweigh under the attendant circumstances (p. 21).
Under the modern law of armed conflict, it can be considered that attacking or forcibly blocking civilians from leaving a besieged area is incompatible with the duty to take feasible precautions for the protection of civilians (U.S. DoD Law of War Manual, §184.108.40.206). Nevertheless, there is a risk that besieged forces manipulate an evacuation arrangement, which could invite a forcible response and endanger civilian lives. Repeated disruptions of humanitarian corridors agreed between Ukraine and Russia illustrated this challenge to meeting humanitarian demands without compromising the effectiveness of the siege operation.
Third, the same dilemma arises regarding the authorization for humanitarian relief. The relevant rule is found in Article 70 of Additional Protocol I to the Geneva Conventions, which reads in part:
If the civilian population of any territory under the control of a Party to the conflict, other than occupied territory, is not adequately provided with the supplies mentioned in Article 69 [which refers to clothing, bedding, means of shelter, other supplies essential to the survival of the civilian population], relief actions which are humanitarian and impartial in character and conducted without any adverse distinction shall be undertaken, subject to the agreement of the Parties concerned in such relief actions…
The provision of humanitarian relief in besieged urban areas is thus conditional upon the besieging force’s agreement to the delivery of supplies.
There is a view that this agreement must not be arbitrarily withheld because the word “shall” in the provision suggests that “the acceptance of humanitarian relief is not entirely discretionary” (p. 21). However, this reading of the provision has been contested due to a host of interpretive issues (pp. 27-35). It remains to be seen whether there is an unqualified duty to permit humanitarian relief operations beyond the situations where besieging forces themselves have approved the provision of humanitarian supplies.
Applicability of Targeting Law
The destruction of buildings and other structures is an inevitable part of urban siege operations. Such “attacks” conducted in and against besieged areas must comply with targeting rules under the law of armed conflict. This means that besieging forces are prohibited from deliberately killing or injuring civilians, including those fleeing besieged areas, causing incidental civilian harm when it is expected to be excessive in relation to the anticipated military advantage to be gained, or engaging in acts or threats of violence for the purpose of spreading terror among the civilian population (U.S. DoD Law of War Manual, §5.2.2; Customary International Humanitarian Law Study, rr. 2, 11, 14). The entire area of siege must not be made the object of attack.
Views at the workshop were divided, however, regarding the applicability of targeting law to the siege operation as a whole, beyond each engagement that amounts to an “attack.” A careful choice of wording between “attacks” and more general “military operations” in the text of Additional Protocol I makes it clear that specific targeting rules regarding distinction, proportionality, and the duty to exercise feasible precautions are only meant to apply to the conduct of hostilities that qualifies as an “attack.” Article 49 of Additional Protocol I defines “attacks” as acts of violence against the adversary, which are understood to encompass any operation that is reasonably expected to cause injury or death to persons or damage or destruction to objects (Tallinn Manual 2.0, r. 92; Harvard Manual, r. 1(e) para. 7).
It thus follows that targeting rules have no application as constraints on the modality and scale of an urban siege operation itself insofar as the positioning of forces to isolate an area does not necessarily lead to injury or death to persons or damage or destruction to objects. According to this view, held by Sean Watts, the general conduct of an urban siege operation is subject to specific prohibitions on starvation as a method of warfare and the measure of intimidation to spread terror among the civilian population. For States that are a party to Additional Protocol I, their siege operation must also be guided by the general obligation to take constant care to spare the civilian population, civilians, and civilian objects (Article 57(1)).
This view is countered by those who consider that the end goal of siege is starvation to death, which qualifies as an “attack” and thus triggers the application of targeting rules. The argument goes that an urban siege operation is prohibited in situations where its effect cannot be controlled to discriminate civilians from enemy combatants or the expected magnitude of civilian casualties is excessive in relation to the anticipated military objective to be gained from it. However, this position is criticized as misconstruing the ultimate objective, which is rather to compel enemy forces into submission before they experience starvation. The utility of this argument is also questioned when the military advantages anticipated from siege are aggregated to outweigh humanitarian considerations.
At any rate, there is general agreement that greater cautions need to be exercised in conducting a siege operation in a densely populated urban area, where entrapped civilians can easily record the scene of destruction and capture images of despair for instant dissemination world-wide. Even though decisional analysis is essential to sound enforcement of targeting law, images of destructive effects and devastating humanitarian consequences embroil besieging forces in the “moral game,” as one of the workshop participants described it. This game will play out in the form of information warfare, where both besieging and besieged forces blame each other for the dire humanitarian situation resulting from urban siege.
Siege warfare continues to be a persistent aspect of urban warfare on the modern battlefield. While the decision to engage in this method of warfare is influenced by operational and tactical demands, the legal issues have strategic consequences. Interpretative differences among scholars and practitioners on applying the law of armed conflict to this method of warfare will impact military operations and the perceptions about these operations. This workshop provided an excellent opportunity for a diverse group of academics, practitioners, and combat leaders to gain a broad understanding of siege warfare and the varied perspectives on the application of law to these operations. Awareness and understanding of these perspectives can influence policy, operational, and tactical decisions in a meaningful way as combatants seek to balance military necessity with humanity in the complex crucible of urban siege warfare. The Lieber Institute is grateful to its event partners at Harvard Law and the ICRC and looks forward to conducting other events in support of our mission to foster a deeper understanding of the complex and evolving relationship between law and warfare.
Hitoshi Nasu is a Professor of Law in the Department of Law at the United States Military Academy.
COL Winston Williams is Head of the Department of Law at the United States Military Academy, Co-Director of the Lieber Institute for Law and Land Warfare at West Point, and Co-Editor-in-Chief of Articles of War.
Photo credit: 1st Lt. Derrick Garner, U.S. Army