Last week, I left off my exploration of the International Committee of the Red Cross’s (ICRC) 2016 Updated Commentaries on the First Geneva Convention of 1949 by asking about the significance of consent where governments have lost control over portions of their territory to a non-state actor or actors who might use that ungoverned territory to threaten the interests of other States? The situation in Syria immediately comes to mind. If a threatened State, such as Turkey, sends forces into ungoverned territory to deal with the non-state actors without the consent of the State which has lost control of that land, such as Syria, should the two States automatically be considered in an international armed conflict with each other, as the ICRC’s “formula” suggests?

Consent seems more relevant to the issue of whether access to the territory must otherwise be based on the lawful exercise of State self-defense, rather than form the basis for conflict categorization.  In other words, as a lawful recourse to force issue. This is evident when looking at the theory upon which the ICRC approach is based. That theory, set out in Dapo Akande’s “Classification of Armed Conflicts: Relevant Legal Concepts” found in the 2012 book International Law and the Classification of Conflicts, raises two points particularly worth noting.  First, in that book (at p. 74) the lack of consent leading to an international armed conflict classification is based on the use of force being in breach of the intervening State’s obligations under “the principles of the non-use of force and non-intervention” as is reflected in the Armed Activities case. This reliance on a breach of Article 2(4) of the UN Charter, part of the law governing the recourse to war, appears at odds with the clear statement in the Updated Commentaries that the jus ad bellum (State self-defense) must be separated from the jus in bello (humanitarian law). Further, the Commentaries note the determination of the existence of an armed conflict should depend on the prevailing situation on the ground, and not whether the use of force is permitted under the Charter (para. 215).  Ultimately this reliance on State consent, as the basis for conflict categorization, makes it difficult, if not impossible, to separate it from the law governing the recourse to war.

Secondly, Dapo’s theory (p. 74) views responses against non-State actors under Article 51 of the Charter in this context as creating an international armed conflict, since “the use of force would be against the territorial integrity of another State.” In focusing on the territorial integrity this theory leaves out State responses against NSAs on the high seas, the use of force against an NSA inside another State with that State’s consent, or the use of force against an NSA within one’s own territory. It concentrates on “sovereignty” rather than the nature of the protagonists. Not all breaches of sovereignty necessarily mean there is “automatically” an armed conflict between the States involved. This can be seen in the conflict categorization morass that is the Syrian conflict. In response to the Turkish incursion against Islamic State and Kurdish armed groups in northern Syria that began on 24 August 2016 the Syrian government asserted a “flagrant violation of Syrian sovereignty”.  Setting aside the issue of how it is a violation of sovereignty if Turkey is lawful exercising its right of State self-defense, the Syrian Foreign Ministry also stated “[a]ny party conducting a battle against terrorism on Syrian soil must do so in coordination with the Syrian government and the Syrian army, who have been fighting this war for five years”.  Nothing in this response suggests Syria viewed itself as being in an international armed conflict with Turkey as a result of this non-consensual intervention. This undermines an assertion that Turkey, and potentially the U.S. which provided combat air support against Islamic State targets in Syria, are engaged in an inter-State armed conflict with Syria.  That makes sense given the States involved are taking action against NSAs who are in also conflict with the Syrian government.  It was Kurdish authorities who viewed the Turkish action as a declaration of war

There is little doubt the ICRC adopted theory will find favour with those international lawyers who continue to be reluctant to accept the premise that defensive action can be taken against NSAs independent of any connection to a State.  However, the interpretation favoured by those lawyers is no longer the majority view of the international law community post 9/11.  The reality of transnational jihadist threat has resulted in a much broader acceptance that State self-defense can be directed uniquely at NSAs. In a recent article on classifying the Syrian conflict Terry Gill also addresses the consent issue.  He highlights that entry into another State may not violate its sovereignty, and the law governing State self-defense should be kept separate from how conflict is categorized under international humanitarian law.  Importantly, the article also notes the lack of any reference in the Geneva Conventions (Common Articles 2 and 3) to a violation of sovereignty being a criterion for conflict categorization, the absence of State practice supporting the consent based theory, and that conflict classification primarily focuses on the nature of parties involved. Professor Gill does indicate because the ICRC has adopted this theory it cannot be conclusively rejected, however, on balance the automatic creation of an international armed conflict based on non-consensual intervention seems considerably less persuasive than an approach based on the factual situation.

Interestingly, contrary to the ICRC approach, Dapo Akande (at p. 77) indicates that the international armed conflict would extend to the violence used against the NSA, since “the conflict with the non-state group will be so bound up with the international armed conflict between the States that it will be impossible to separate the two conflicts.”  Every attempt to target the NSA, “will at the same time be a use of force against the territorial State.” The ICRC willingness to accept parallel international and non-international armed conflicts under its version of this theory introduces an unfortunate level of complexity, and potential for confusion. It represents the type of theoretical result that causes military commander’s eyes to glaze over as this state of legal affairs is explained to them. What does suggesting there are two apparently intimately intertwined armed conflicts existing at the same time mean in practice?  It is an approach that seems overly wedded to maintaining distinct formal categories of conflict. Warfare is complex enough. The focus should be on reducing complexity rather than increasing it.

The idea that defensive force directed at NSAs almost invariably causes an international armed conflict, for which there is no intensity threshold, also raises questions about what the situation is when States intervene to protect nationals. This increasingly includes rescuing hostages taken by criminal gangs as well as organized armed groups. The taking of hostages, whether by jihadist groups or criminal gangs has become an acknowledged insurgent/terrorist financing tactic. It should therefore come as no surprise that “rescue” operations are an integral part of contemporary military practice. In January 2013 French Special Forces unsuccessfully attempted to rescue an intelligence agent being held by al-Shabaab 70 miles south of Mogadishu in an operation where the agent and two soldiers were killed. Since this operation was subsequently condemned by the Somali government it would appear, under the ICRC theory, to have automatically created an international armed conflict between Somalia and France.  However, it is hard to see this action as anything other than a legitimate act of self-defense in an area where the Somali government was incapable of exercising any control.

International law has always struggled with the concept of intervening to defend nationals, but ultimately has tolerated a certain level of such activity. As early as 1958, D.W. Bowett, in his book Self-Defense in International Law, identified (at p. 88) that a territorial State being “unable, or unwilling” to prevent injury to another State’s persons or property provided the legal basis for such action. This raises the question under the ICRC theory of whether an international armed conflict would be automatically in existence if the United States, the United Kingdom, Canada, or Australia carried out non-consensual non-combatant evacuation missions to protect their nationals threatened in another State by a NSA. For example, would an international armed conflict still be caused by the intervening State acting without consent, even if it does not use any force at all, or only applies the more restricted levels of force contemplated by human rights law solely to rescue their citizens held in a rebel controlled territory?

If the goal is to find there is an international armed conflict because the intervening State may overstay its welcome, thereby setting up a situation of occupation, that possibility is already provided for under international law. This may be a very real possibility should the August 2016 Turkish intervention be extended to long term control over the area of operations, such as the possible creation of a buffer zone within Syria.  A very positive development in the Updated Commentaries is that the ICRC has walked back (paras. 301-22) from the broad assertion of the applicability of the law of occupation during an “invasion” phase found in the 1952 Pictet commentaries. It would seem inconsistent to expand that body of law to limited defensive interventions. Interventions would normally only be for periods of very limited duration.  What is not clear is why the human rights obligations of the intervening State, whether based on treaty or customary international law, or the application of humanitarian law in situations of a non-international armed conflict would be insufficient to protect the civilian population during such operations.

It may be that the ICRC approach reflects a desire to rely on other protections available under the Geneva Conventions such as POW status for State military forces. However, this possibility seems limited given the position a conflict with the NSA would be a non-international one for which such status is not applicable at law. This is an area where what is to gained from an increase in the number of “theoretical” international armed conflicts should be more clearly stated, and as importantly the downside of expanding notional inter-State armed conflicts more fully considered. As Terry Gill has noted (p. 378) in a Syrian context, determining that Coalition States are automatically involved in an armed conflict with Syria, which is in turn supported by Russia and Iran “would have consequences, both legal and non-legal, which go far beyond the scope of the conflicts with Syria.”  Included in that concern is the “geography of armed conflict” and the potential expansion of the war.  Consideration must also be given to the other potential international diplomatic, commercial and legal consequences that can result from an armed conflict between States.

Further, where a State acts in lawful self-defense the delict that could ultimately result in an international armed conflict is not the non-consensual crossing of a border, but rather the detention by the territorial State of a member of the armed forces acting in accordance with international law.  A premature claim that prisoner of war status is available because two States are “at war” may exacerbate existing tensions, and potentially increase the possibility of actual inter-State clashes.  It may also encourage, or harden the views of segments of the populations of both States that seek broader conflict against the other State involved.  As the Updated Commentaries stress a strength of international humanitarian law is its separation from the body of law governing the recourse to war. However, interpretations of humanitarian law cannot be applied in isolation.  A holistic approach must be adopted that takes into account all of the international law applicable to the situation, as well as the practical consequences that can result from favoring a particular interpretation of the law.

If the Syrian government were to shoot down a U.S. plane operating over Syria a claim could be made that an armed conflict exists between the two States and any captured aircrew would have prisoner of war status. A similar situation occurred in 1982 when a U.S. plane was shot down over Lebanon by Syrian forces resulting in the death of the pilot and the capture of the navigator. As is explained in a 2016 book, U.S. Military Operations: Law, Policy, and Practice (p. 74), while the U.S. government did not indicate it was at war with Syria the limited engagement was viewed as sufficient to claim prisoner of war status as a matter of law under Common Article 2. Following negotiations, the aviator was released after a month in captivity. However, that case was different than a situation where the military operations are solely directed at a non-State actor.  If a French Special Forces operator had entered a Somali government controlled area following the 2013 hostage rescue operation, or if a Coalition pilot reaches Syrian lines today after being shot down by the Islamic State, the obligation on the territorial governments would be to return the detained personnel to their own forces. Issues related to the status of intervening State personnel are better dealt with on a case-by-case basis after a full assessment of the existing facts, as well as the law, rather than viewing all cross-border interventions as creating an international armed conflict.  In other words, privileging substance over form.

There is no doubt that prisoner of war status provides an exceptionally high standard of treatment, and as the Manuel Noriega case highlights, it is possible to hold a single person as a prisoner of war.  However, prisoner of war status is a consequence of a particular type of conflict being in existence, and not a reason for categorizing an incident a particular way.  There is nothing to stop a State from demanding its detained personnel be treated to the standard of a prisoner of war until the situation is resolved. Further, human rights law also provides a high standard of protection to be applied in situations where a territorial State is holding foreign military personnel until their status can be confirmed.  The challenge is not whether humane standards exit, it is the willingness to apply them.

The ICRC approach is all the more interesting because it is based on just one of a number of possible classification approaches regarding armed conflict with non-State actors.  Professor Akande acknowledges (at p. 75) that his theory has some support from scholars, but is probably not the majority view in the existing literature. As the Commentaries note (para. 214), one of the challenges of international law is that “there is no central authority under international law to identify or classify a situation as an armed conflict”. As a result, classification theories, particularly post-9/11, are numerous and sometimes novel. In this respect, the ICRC selected approach is just one of many. The possibilities include: Additional Protocol I, article 1(4) conflicts with NSAs during “wars of national liberation”; the now dated “recognition of belligerency” theory; NSAs being controlled by a State; the harboring and active support of NSAs by a State; the Israeli Targeted Killing Case view that cross border NSA attacks are “conflicts of an international character”; Yoram Dinstein’s extra-territorial law enforcement theory; post 9/11 “transnational armed conflicts”; the United States, Hamdan vs. Rumsfeld, decision referring to NSA conflicts as non-international armed conflict seemingly unbounded by State territory or borders; non-international armed conflicts being limited by State territory; and spillover non-international armed conflicts.

The Updated Commentaries address, or at least make reference to, some of these approaches. However, what is noteworthy are the theories that are not discussed, or only tangentially addressed. These include post 9/11 theories that have been actively applied during a now 15 year armed conflict with NSAs. Perhaps because it is not popular with some segments of the international law community, the “harboring and active support” theory is not included even though it is widely recognized for providing the rationale for linking the Taliban-led government to al-Qaeda post 9/11. This link made the resulting armed conflict an international one. Further, it can find a theoretical basis in the 1980 International Court of Justice decision in Case Concerning United States Diplomatic and Consular Staff in Tehran (United States v. Iran) to the extent the subsequent approval by Iranian officials of the actions of militants who seized the United States embassy and took hostages made their actions attributable to the State.  The Updated Commentaries Article 3 analysis (para. 482) too readily minimizes the Hamdan v. Rumsfeld approach noting “the practice of States party to the Geneva Conventions in support of a global or transnational non-international armed conflicts remains isolated.” This is unfortunate for a number of reasons.  First, the US Supreme Court decision underpins a significant portion of military action being taken against a substantial and wide-spread transnational jihadist threat.  As I have commented elsewhere there appears to be a transatlantic divide between Europe and North America regarding approaches to legal interpretation based on differing legal systems, as well as geographic and experiential factors.  It is a divide that will not be bridged by effectively relegating the Hamdan case to the footnotes.

Secondly, the jihadist threat is increasingly recognized as a global one, as is reflected in UN Security Council Resolution 2249 (2015).  Thirdly, as Terry Gill notes, there is considerable merit in looking at the threat posed NSAs in terms of status rather than cross-border uses of force.  It is conflict between two States (two “proper” authorities) and not necessarily cross border action that makes a war an international one.  Similarly, it is the status of the State as a proper authority, and that lack of status for an NSA, that most logically governs the categorization of that type of armed conflict rather than focusing on territorial borders. Finally, the Hamdan v. Rumsfeld, and the Israeli Targeted Killing Case represent what international legal jurists have not been asked to do, reach decisions on fighting NSAs who attack across borders.  Those decisions deserve to be accorded a more prominent and detailed analysis.  Given the sheer number of these classification theories, it should come as no surprise that practitioners are apt, in terms of determining what force can be used, to rely more readily on a determination that an “armed conflict” is in existence rather than rush to figure out the nature of the armed conflict.

I agree with Sean Watts’ analysis that the practical consequences of the ICRC approach are, “unlikely to materialize in reality.” In this context, it is doubtful the ICRC notion of a parallel international armed conflict being created when no force is actually being used against territorial State security forces or its institutions will gain much traction, let alone acceptance, amongst States.  Rather the focus is often on the existence of a generic “armed conflict” particularly when the State has to act under the significant time pressures normally associated with the exercise of self-defense. This can mean considering if the group is an organized one engaged in hostilities; and assessing the intensity of the violence, including whether it represents more than, in the wording of Additional Protocol II, “internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature”.

It is in this aspect of conflict categorization where the Updated Commentaries have made significant strides in addressing when an armed conflict between a State and NSA is in existence.  For example, in the Article 3 analysis there is recognition that the duration of hostilities is a factor better suited to after the fact judicial proceedings than on the spot assessments of the intensity of the violence.  Importantly, it is noted (para 439):

From the perspective of the practical application of humanitarian law, an independent requirement of duration could, in contrast, lead to a situation of uncertainty regarding the applicability of humanitarian law during the initial phase of fighting among those expected to respect the law, or to a belated application in situations where its regulatory force was in fact already required at an earlier moment.

Further, “hostilities of only a brief duration may still reach the intensity level of a non-international armed conflict if, in a particular case, there are other indicators of hostilities of a sufficient intensity to require and justify such an assessment.”  This, combined with the recognition that “when the government is obliged to use military force against the insurgents, instead of mere police forces” may represent the requisite degree of intensity is met, more accurately reflects practical factors that are brought into play by States (para. 431).

This is reflective of an approach that relies on a broader, practically focused, “totality of the circumstances” assessment rather than narrow adherence to the more formalist application of the 1995 Tadić criteria of “protracted armed violence” and “organized armed groups”. The “totality of the circumstances” test is reflected in the work of Laurie Blank and Geoffrey Corn, as well as the 2008 ICTY Boškoski case (para. 257).  Added to these factors can be the tactics and weapons used by the organized armed group, and the type of action reasonably required by States to defeat the threat. Of significance, the Updated Commentaries approach also has a closer affinity with the short duration style of conflict recognized in 1997 Juan Carlos Abella v. Argentina Inter-American Commission case.  Ultimately, the emphasis on practicality in this portion of the Commentaries more realistically situates States to deal with the “one off” attacks, like the 2008 Mumbai, 2013 Westgate Mall, 2014 Peshawar School and Garrissa University and 2015 Paris assaults, that are increasingly becoming part of the contemporary security challenges facing States.

In the final analysis, the Updated Commentaries reflect a broader struggle taking place in the international legal arena. There is a requirement move on from the traditional 20th Century near obsession with inter-State conflict to a more pragmatically oriented approach that concentrates on contemporary NSA threats. To be relevant, commentaries such as this one, like the US DoD manual, must endeavor to clearly link theory to practical effect.  Chuck Allen, Karl Chang, and Matt McCormack have done excellent work in completing the US manual over the past five years since taking up the baton after Hays’ retirement.  Importantly, exceptional leadership has also been demonstrated in their willingness of consider amendments to that manual. As further assessments and critiques are written concerning the ICRC initiative, it is hoped the Updated Commentaries will also continue to be just that: updated.  The opportunity to do so will undoubtedly arise as the remaining Commentaries are developed and published “on line”.  Once again the digital format will facilitate such as step.