1. Introduction
International humanitarian law (IHL) distinguishes between international armed conflicts (IACs), including situations of military occupation, and non-international armed conflicts (NIACs).Footnote 1 Internal disturbances, tensions, and isolated or sporadic acts of violence do not come within the purview of IHL.Footnote 2 Armed conflicts are regulated by the 1949 Geneva Conventions, the 1977 Additional Protocols I (AP I)Footnote 3 and II (AP II),Footnote 4 and customary IHL.Footnote 5 IHL treats all parties to such armed conflicts equally and demands its equal application by the parties.Footnote 6 However, IHL does not provide clear classification guidelines in the case where a state covertlyFootnote 7 intervenes in a NIAC taking place in another state by supporting an armed group operating in the territory of that other state. IHL instruments contain no specific articles that address and determine the nature of such armed conflicts,Footnote 8 except perhaps Article 29 of Geneva Convention (GC) IV, which foresees a situation of occupation by a state's de facto agents (see below).Footnote 9 The question thus becomes which law applies: the IHL regulating IAC or that pertaining to NIAC? Or is this a situation of two armed conflicts taking place in parallel, or rather a new type of conflict of a mixed nature?Footnote 10
Prima facie internal conflicts with foreign state intervention are subject to much controversy. The parties involved seldom agree on whether there is foreign state intervention.Footnote 11 For some, foreign intervention in a prima facie NIAC leads to an internationalised internal armed conflict.Footnote 12 Verhoeven regards this term to mean ‘armed conflicts waged in a State between government forces and rebel factions or between armed groups with the intervention of one or more foreign states supporting the government, the rebel factions or armed groups’.Footnote 13 Mačák explains it as ‘the process of transformation of the legal nature of a prima facie NIAC, which renders the law of IAC applicable to such a conflict’.Footnote 14 Yet the term ‘internationalisation’ can create confusion in determining the applicable IHL. It suggests a blanket application of the law of IAC.Footnote 15
In order to avoid confusion, the preferred term throughout this article is the notion of a NIAC involving foreign state intervention in support of an armed group fighting in it.Footnote 16 Obviously, what then needs to be clarified is the type (military, logistical, financial, political, and so on) and degree of involvement by the foreign state required to transform a NIAC into an IAC.Footnote 17 The differentiation between the type and degree of such involvement is key: financial or political support provided to an armed group would not make the intervening state a party to the armed conflict.Footnote 18 However, there may come a point when that state goes further and in fact provides the armed group with weapons, intelligence, instructions or even its own troops, thus in turn actually becoming engaged in the armed conflict.Footnote 19
For this reason, some perceive the effective control test, developed by the International Court of Justice (ICJ),Footnote 20 to be a suitable benchmark.Footnote 21 For others, by contrast, it is the overall control test advanced by the International Criminal Tribunal for the former Yugoslavia (ICTY) that is appropriate.Footnote 22 Others again suggest a test that is ‘general (in its scope) and strict (in terms of intensity)’,Footnote 23 which would be met:Footnote 24
if State A had a crucial and leading role in organizing, coordinating and planning military actions of armed group C. It is not sufficient for State A to loosely accommodate the activities of armed group C, nor to merely help in the general planning of its military operations, as required by the overall control test. The room for manoeuvre of C should not be too large, and it must be possible to establish a chain of command between the armed group and the controlling State.
This article argues that the requirements regarding the type and level of control needed for an intervening foreign state supporting an armed group to become a party to a prima facie NIAC, and thereby transform it into an IAC, are contained within IHL itself. IHL has to be viewed as a self-contained system;Footnote 25 the level of control enshrined in the provisions of Article 4A(2) of GC III and Article 29 of GC IV corresponds to the type of control envisaged by the overall control test.Footnote 26 This, however, does not mean that IHL is an isolated regime incapable of dialogue with other branches of international law,Footnote 27 particularly with that of state responsibility (discussed below at Section 3.2). Failure to consider other branches risks increasing the fragmentation of international law.Footnote 28 Thus, the ICTY Appeals Chamber in the Tadić case should have gone beyond the ICJ in the Nicaragua case, using GC III Article 4 and GC IV Article 29 as primary sources for its analysis when identifying the role of the intervening state in a prima facie NIAC. This would have formed a bridge towards the regime of state responsibility. The ICJ, in turn, should have identified common grounds between IHL and the law of state responsibility. In the Nicaragua and Genocide Footnote 29 cases, the Court should not have kept separate the matters concerning conflict classification and state responsibility.
The second section of this article examines the effective and overall control tests as understood by the ICJ and the ICTY respectively. It demonstrates that the tension created between the law of state responsibility and that of conflict classification is artificial. The following section analyses GC III Article 4 and GC IV Article 29. From these the required level of control can be drawn in order to determine whether a foreign state's participation in a prima facie NIAC in support of an armed group engaged in it transforms that conflict into an IAC. This article concludes by upholding the overall control test for both conflict classification and state responsibility purposes. It maintains that the overall control test is more realistic to ascertain, thereby making it harder for states to escape their IHL obligations.
2. Complete Dependency and Effective Control versus Overall Control
This section focuses on the type and degree of control needed for attributing to a state the responsibility for wrongful acts committed in the context of a prima facie NIAC by armed groups which are not organs of that state. The tests of complete dependency and effective control as developed by the ICJ in the Nicaragua case are contrasted with the overall control test developed by the ICTY Appeals Chamber in the Tadić case.
2.1. Complete Dependency and Effective Control Tests
Whether a state is responsible for an internationally wrongful act is determined based on the identification of the wrongful act and its attribution to the state as codified in the International Law Commission (ILC) Articles on the Laws of State Responsibility (ARSIWA).Footnote 30 Articles 4 and 8 provide important tests of control through which state responsibility can be examined. These articles require different levels of control and refer to various scenarios.
Pursuant to Article 4, an organ of the state ‘includes any person or entity which has that status in accordance with the internal law of the State’.Footnote 31 The level of control is the highest and most stringent as that provision identifies state organs stricto sensu. The type of relationship that subsists between a state and its organs/agents thus amounts to complete dependency.Footnote 32 Based on Article 4, the conduct of an organ of the state (as per its internal law) can trigger the responsibility of the state.Footnote 33
Persons or entities that do not legally form part of a state under its internal law may be regarded as ‘de facto’ organs, such that all of their conduct is attributable to it,Footnote 34 on the basis ‘of a factual link with that organisation’.Footnote 35 ARSIWA Article 4 also provides for the possibility of state responsibility for the conduct of a group or individual based on ‘the role that they in fact perform within the structure of the state’.Footnote 36
Pursuant to ARSIWA, the conduct of private persons or entities is not generally attributable to a state.Footnote 37 However, ARSIWA Article 8 states:
The conduct of a person or group of persons shall be considered an act of a State under international law if the person or group of persons is in fact acting on the instructions of, or under the direction or control of, that State in carrying out the conduct.
The commentary to Article 8 does not explain what ‘acting on the instructions’ and ‘under the direction or control’ mean. Nonetheless, it considers the terms ‘instructions’, ‘direction’ and ‘control’ to be disjunctive, as ‘it is sufficient to establish any one of them’.Footnote 38 All three constitute instances of control exerted by the state over private persons or entities. However, the type and degree of control that should be exercised by the state over such entities remains unspecified.
2.2. The Nicaragua Case
The basis for any discussion of the complete dependency and effective control tests is the judgment of the ICJ in Nicaragua v United States of America. Nicaragua argued that the United States was recruiting, training, arming, equipping, financing, supplying, encouraging, supporting, aiding, and directing military and paramilitary actions in and against Nicaragua. According to Nicaragua, the US government was effectively in control of the Contras paramilitary group, formulating its strategy and guiding its tactics, with the initial purpose of overthrowing the government of Nicaragua. Nicaragua also accused the US of having mined Nicaragua's territorial waters and attacked its ports and oil installations.Footnote 39 Nicaragua claimed that these acts were committed by US military personnel or persons with the nationality of unidentified Latin American countries, who were paid by and acted on the direct instructions of US military or intelligence personnel: ‘These persons were apparently referred to in the vocabulary of the CIA as “UCLAs” – “Unilaterally Controlled Latino Assets”’.Footnote 40
The ICJ established that in late 1983 and early 1984, acts against Nicaragua were indeed carried out by UCLAs and civilian US pilots contracted by the CIA.Footnote 41 Even though it was not proven that US military personnel took a direct part in the operations, US agents had participated in the planning, direction, support and execution of operations, while the execution was the task of the UCLAs. Hence, responsibility on the part of the US for these attacks was established.Footnote 42 The Court then proceeded to examine the genesis, development and activities of the Contras, which Nicaragua maintained the US had ‘conceived, created and organized [as] a mercenary army’.Footnote 43
The ICJ thus queried – based on ARSIWA Article 4 – whether the relationship of the Contras to the US government was ‘so much one of dependence on the one side and control on the other that it would be right to equate the Contras, for legal purposes, with an organ of the United States Government, or as acting on behalf of that Government’.Footnote 44 The Court decided that the evidence available to it indicated that the various forms of assistance provided to the Contras by the US were crucial for their activities, yet this was insufficient to show and establish their complete dependence on it.Footnote 45 The ICJ considered that despite the fact that the US heavily subsidised and provided support, there was no clear evidence that it had actually exercised such a degree of control in all fields so as to substantiate that the Contras were to be treated as acting on its behalf.Footnote 46 The ICJ therefore concluded that the Contras could not be equated, for legal purposes, with US armed forces proper.Footnote 47
Next, the ICJ considered whether the US – based on its financing, organising, training, equipping, selection of military and paramilitary objects, and planning of the Contras’ operations – was responsible for specific violations of IHL committed by the Contras. It decided that in order for the acts of the Contras to give rise to responsibility on the part of the US, ‘[i]t would in principle have to be proved that that State had effective control of the military or paramilitary operations in the course of which the alleged violations were committed’.Footnote 48 The ICJ took the view that the Contras alone (and not the US) were responsible for their acts, and that the US was responsible for its own actions ‘vis-à-vis Nicaragua, including the conduct related to the acts of Contras’.Footnote 49 According to the Court, while the UCLAs had been under complete US control, the Contras were not subject to the same oversight and control by the US, despite the latter providing a great amount of material and financial support. The ICJ thus concluded that the effective control test was not met either, as the Contras’ actions against Nicaragua were not carried out under US operational control.Footnote 50 Thus, the US did not control each and every operation by the Contras.
In sum, the ICJ, finding the US not responsible for the acts committed by the Contras in the course of their military or paramilitary operations in Nicaragua,Footnote 51 classified the conflict between the Contras and the government of Nicaragua as a NIAC. As for US actions in and against Nicaragua, according to the Court they fell under the rules of IAC.Footnote 52 Had the Court found that the US was responsible also for the acts of the Contras, it would by implication have considered the Contras to have been de facto agents of the US.Footnote 53 This would have prompted the Court not to split the law between the engaged parties and ultimately classify the prima facie NIAC between Nicaragua and the US as an IAC. This, in turn, would have paved the way towards more effectively forming a deterrent for states to covertly intervene in a NIAC on the territory of other states.
2.3. The Tadić Case
The approach of the ICJ in examining US state responsibility in the Nicaragua case was scrutinised by the ICTY Appeals Chamber in the Tadić case in the course of classifying the armed conflict in Bosnia-Herzegovina.Footnote 54
In order to determine whether the grave breaches regime applied (the latter being applicable only in times of IAC),Footnote 55 the Appeals Chamber needed to re-examine the classification of the conflict, and indeed found it to be an IAC.Footnote 56 The question before the Tribunal was whether Serbia (or, from 1992 to 2003, the ‘Federal Republic of Yugoslavia’ (FRY)) could be regarded as participating in the armed conflict after 19 May 1992. To this end, the ICTY analysed whether the acts of the Bosnian-Serb army, the VRS, could be attributed to the FRY. Because conventional IHL is silent on this link, the ICTY looked at the attribution mechanism in the law of state responsibility for the acts of an armed group which is not the de jure organ of that state. It examined the relationship through which such attributability arises.
The ICTY examined the notion of control and its threshold as propounded by the ICJ in the Nicaragua case. There, effective control in terms of attribution required a high level of control over the course of the operation, including decision making and total operational control, specifically controlling each and every step of the operation and its participants. However, while the ICTY Trial Chamber had adhered to the ICJ text, the Appeals Chamber came up with a completely novel approach – namely overall control – thus lowering the bar set by the ICJ.
The ICTY Appeals Chamber upheld the Trial Chamber in a finding of fact that the armed conflict prior to 19 May 1992 was an IAC; but whether the armed conflict after 19 May 1992 continued to be an IAC or had descended into a NIAC depended on whether Bosnian Serb forces (in whose hands the Bosnian victims found themselves) could be considered de jure or de facto organs of a foreign state, namely the FRY.Footnote 57 The reasoning engaged here by the ICTY reflected ARSIWA Article 8. In other words, for the conflict to have become international, it must be proved that the VRS and the Republika Srpska – the entity set up by ethnic Serbs on the territory of Bosnia in 1992 – were acting not on their own but on behalf of the FRY, meaning that both were under its control. The question then becomes what kind of control.Footnote 58
The ICTY Appeals Chamber stated that IHL had no criteria for determining when a group of individuals could be said to be under the control of a state (that is, acting as de facto state officials/agents). Hence, it examined the notion of control by a state over individuals as laid down in general international law (such as in the law of state responsibility), trying to ascertain whether certain individuals can be regarded as acting as de facto state officials.Footnote 59 The ICTY analysed the notion of control as propounded by the ICJ in Nicaragua. According to the ICTY Appeals Chamber:Footnote 60
In both cases, what is at issue is not the distinction between State responsibility and individual criminal responsibility. Rather, the question is that of establishing the criteria for the legal imputability to a State of acts performed by individuals not having the status of State officials. In the one case these acts, if they prove to be attributable to a State, will give rise to the international responsibility of that State; in the other case, they will ensure that the armed conflict must be classified as international.
Thus, the ICTY considered and contemplated the same test for state responsibility and conflict classification. It held that the degree of control that rendered an individual or a group ‘agents’ of a state could vary depending on the circumstances.Footnote 61 In the case of individuals who are members of an organised armed group – such as a military unit or, in the case of a NIAC, armed bands of irregulars or rebels – overall control by the state is required.Footnote 62 Such control ‘to some extent equates the group with state organs proper’.Footnote 63 For the acts of a military or paramilitary group to be attributed to a state, it must be shown that:Footnote 64
the State wields overall control over the group, not only by equipping and financing the group, but also by coordinating or helping in the general planning of its military activity … it is not necessary that, in addition, the State should also issue, either to the head or to members of the group, instructions for the commission of specific acts contrary to international law.
What is the exact meaning of the ‘overall control’ threshold carved out by the ICTY Appeals Chamber? What amounts to this control of an outside state over armed groups? Clearly such control is lower than both complete dependency and effective control as envisaged by the ICJ, as it does not include specific orders or directing each and every operation; but it is also higher than the mere provision of financial assistance, military equipment or the training of soldiers. It seems that the state has to have ‘a role’ and generally direct the actions of such groups. It is therefore not necessary that the controlling authorities plan all the operations of the dependent units, choose every target, or issue specific instructions on the conduct of military operations.Footnote 65 Pursuant to the overall control test, the ICTY Appeals Chamber found the actions of the Bosnian Serbs in fact to be attributable to the FRY and the conflict in Bosnia-Herzegovina after 19 May 1992 to be an IAC.Footnote 66
To summarise, the ICTY Appeals Chamber in Tadić diverged from the ICJ Nicaragua standard of complete dependency and effective control for the purposes of establishing state responsibility, thus crafting a single test for both state responsibility and conflict classification. However, the ICTY's actions should be examined pursuant to Article 92 of the UN Charter, which regards the ICJ as ‘the principal judicial organ of the United Nations’.Footnote 67 The ICJ deals with general matters of public international law, whereas the jurisdiction of the ICTY concerns only matters of individual criminal responsibility. The approach of the ICTY was based on the absence of a clear test in IHL that could deal with conflict classification given outside state involvement in a prima facie NIAC. Whether the ICTY was too audacious or gave a fresh look at the matterFootnote 68 and paved the way to harmonise the issues of state responsibility and conflict classification that are separate but nevertheless very close to one another will be discussed next through the ICJ Genocide case.
2.4. The Genocide Case
In 2006, when the ICJ adjudicated Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro),Footnote 69 it also opined on the stance of the ICTY Appeals Chamber as expressed in Tadić. The ICJ examined the question of responsibility for events in Srebrenica, notably whether the responsibility of the FRY was implicated through the actions of persons or entities that were not its de jure organs.Footnote 70 The ICJ employed ARSIWA Articles 4 and 8.Footnote 71 Unlike the Nicaragua case, the Court on this occasion dealt with the matter more clearly, and in two steps.Footnote 72
The Court first found that the criteria of ARSIWA Article 4 were not met. It had not been shown that the FRY army itself had taken part in the massacres, nor that its political leaders had had a hand in preparing, planning, or in any way carrying them out. While there was much evidence on the direct or indirect participation of the FRY official army, together with the VRS, in military operations in Bosnia and Herzegovina in the years prior to the events in Srebrenica, it was not shown that there had been such participation in the actual massacres. Also, ‘neither the Republika Srpska, nor the VRS were de jure organs of the FRY, since none of them had the status of organ of that State under its internal law’.Footnote 73 The ICJ further stated that although the salaries of Republika Srpska officials and some VRS officers were paid by the FRY, the former were not the de jure organs of the latter. The ICJ also considered whether the FRY was responsible for the actions of the Scorpions paramilitary unit. Based on the materials presented to it, the ICJ was unable to find that the Scorpions were de jure organs of the FRY in mid-1995.Footnote 74
The Court then turned to consider whether it was possible, still under Article 4, to attribute to the FRY the conduct of (groups of) persons who did not have the legal status of state organ, but which in fact acted under such strict control by the state that they had to be treated as such. After repeating the test enunciated in the Nicaragua case on equating an armed group with a state organ,Footnote 75 the Court held that pursuant to its jurisprudence persons, groups of persons or entities, for purposes of international responsibility, may be equated with state organs even if that status was not granted by the internal law of the state, ‘provided that in fact the persons, groups or entities act in “complete dependence” on the State, of which they are ultimately merely the instrument’.Footnote 76 Such an equation, according to the Court, was exceptional and required ‘proof of a particularly great degree of State control over them’.Footnote 77 The Court did not find the existence of such dependence on the FRY. According to the Court, differences in military strategy between the Yugoslav authorities and the Bosnian Serb leaders evidenced that the latter had some real ‘margin of independence’.Footnote 78
Finally, the ICJ considered attributing the Srebrenica genocide to the FRY via the ‘instructions’, ‘direction’ or ‘control’ of Article 8 ARSIWA.Footnote 79 In the Court's view, whether the perpetrators of the genocide were acting on the instructions of the FRY, or under its direction or control, was a completely separate issue. An affirmative answer to this question would not imply that the perpetrators should be regarded as organs of the FRY. It would, however, mean that the FRY was internationally responsible based on the conduct of those of its organs that issued the instructions or exercised the control that resulted in the breaches of its international obligations. Accordingly, the Court had to determine whether FRY organs having that status under the state's own internal law had ordered the genocide via instructions given to the perpetrators or via exercising direction or control.Footnote 80
The ICJ thus viewed the ‘effective control’ test for the purposes of state responsibility to be different from complete dependency in two respects. First, it was not necessary to show that the persons who performed the alleged acts were in a general relationship of ‘complete dependence’ on the state in question. Instead, it had to be ascertained that they acted in accordance with the instructions of the organs of that state or under their ‘effective control’. Second, according to the Court, it had to be demonstrated that such ‘effective control’ was indeed exercised – that is, that specific instructions were issued with regard to each operation and that they were not just generally given with regard to the overall actions taken by the persons or groups of person that committed the violations.Footnote 81 As the Court suggests, the difference between the tests of total and effective control lies in both the type and degree of control they embody. The former envisages these groups to be subject completely to the state's will, leaving no room for independence in their actions; the group is therefore a mere instrument of that state. The latter test, in turn, is different in kind and also possesses a lower threshold: state organs merely instruct or control a group's specific actions.
Because Bosnia-Herzegovina questioned the validity of the application of the Nicaragua ruling to its case and instead drew attention to the Tadić case, the ICJ, without further elaboration, separated the issue of state responsibility from other matters, notably conflict classification. The ICJ observed that in Tadić the ICTY had not been called upon, nor was it in general called upon, to rule on the issue of state responsibility, as its jurisdiction was limited to criminal matters that extended to persons only.Footnote 82 The ICJ held that the ICTY test of overall control had broadened the scope of state responsibility too much and that the ‘overall control’ test was therefore unsuitable as it stretched ‘too far, almost to breaking point, the connection which must exist between the conduct of a State's organs and its international responsibility’.Footnote 83 As regards determining whether the armed conflict was of an international or non-international character, however, the overall control test may still be ‘applicable and suitable’.Footnote 84 Hence, the ICJ maintained two different tests for two different undertakings.
2.5. Democratic Republic of the Congo v Uganda
Another case (which preceded the Genocide case) in which the ICJ dealt with concerns regarding a foreign state supporting a dissident armed group involved in an internal conflict, was the Case Concerning Armed Activities on the Territory of the Congo (Democratic Republic of the Congo (DRC) v Uganda). The DRC claimed that Uganda had engaged in military and paramilitary activities against the DRC, occupied its territory and militarily, logistically and financially supported irregular forces operating there.Footnote 85 Here, too, the ICJ found no credible evidence that Uganda had formed the MLC armed group. According to the Court, the conduct of the MLC was not that of ‘an organ’ of Uganda (ARSIWA Article 4), it did not fall under an entity that was exercising elements of governmental authority on its behalf (ARSIWA Article 5), nor was it considered to be acting ‘on the instructions of, or under the direction or control of’ Uganda (ARSIWA Article 8). None of the tests for holding the actions of paramilitaries attributable to the intervening state had been met.Footnote 86
The Court also remarked that the DRC claimed ‘indirect administration’ by Uganda through various Congolese rebel factions and supervision by Ugandan officers over local elections in the territories under Ugandan army control. However, according to the Court, the DRC had provided no concrete evidence to demonstrate that authority was exercised by Ugandan armed forces in any areas other than Ituri. There was no evidence to allow the ICJ to characterise the presence of Ugandan troops stationed at Kisangani Airport as occupation pursuant to Article 42 of the 1907 Hague Regulations;Footnote 87 nor did the ICJ uphold the DRC claim that Uganda was an occupant outside Ituri district, in areas controlled and administered by Congolese rebel movements, as no supportive evidence was found that these groups were ‘under the control’ of Uganda.Footnote 88
This case demonstrates that an armed group could occupy a territory militarily if acting as a de facto organ of an outside state, exercising elements of governmental authority on its behalf, or acting ‘on the instructions of, or under the direction or control’ of that outside state. The ICJ could indeed have characterised the situation as occupation had the evidence demonstrated that Congolese rebels were under the effective control of Uganda. Had this been proved, the group would have had to exercise effective control over the territory pursuant to Article 42 of the Hague Regulations. It is through this armed group that the state in question would have exercised control over the territory, thus making the armed conflict international. Similarly, had the overall control test been met, this would have been enough to classify the armed conflict as an IAC – although, pursuant to the ICJ, meeting the overall control test would not be enough to establish state responsibility.
3. Assessment and Possible Solutions in IHL for Transforming a NIAC into an IAC
The diverging stances of the ICJ and the ICTY raise the broader question of whether establishing state responsibility and conflict classification are matters that are sufficiently similar to warrant a single test.Footnote 89 While the ICTY in Tadić offered one test for both conflict classification and establishing state responsibility, the ICJ in the Genocide case kept the two issues apart without further explanation. In my opinion, the ICJ had to state explicitly that IHL consisted of primary rules which provided the answer for conflict classification because this branch of public international law regulates these types of situation. In sum, the route taken by the ICJ in the Genocide case explored a narrow question within the context of an armed conflict. In my understanding the ICJ had first to identify the context in which the contended violations of the convention occurred and only then deal with the issue of state responsibility.Footnote 90
In the Genocide case, the following three questions should therefore have been elaborated upon by the Court. First, when does an armed conflict exist? Second, through whom does a state act in a situation of armed conflict? Third, how may the responsibility of a state be triggered in an armed conflict when it is involved through informal forces? The first and second questions share a close bond: the first opens the door to the second question, which establishes the responsibility of a state acting through its de jure armed forces. The third question then deals with the de facto organs of the state, if present. All three questions derive from IHL as primary rulesFootnote 91 and are based on the inter-application of Common Article 2 of the 1949 Geneva Conventions, Articles 1 and 43 of AP I, customary IHL rule 149,Footnote 92 as well as GC III Article 4 and GC IV Article 29. Each question is addressed in turn.
3.1. When Does an Armed Conflict Exist?
Pursuant to IHL, an armed conflict exists – be it an IAC or a NIAC – when there is use of force resulting in hostilities.Footnote 93 The recent commentaries to Common Article 2 of the 1949 Geneva Conventions hold that ‘the determination of the existence of an armed conflict … must be based solely on the prevailing facts demonstrating the de facto existence of hostilities between the belligerents, even without a declaration of war’.Footnote 94 More specifically, a NIAC exists when parties resort to violence that reaches a certain kind and level of intensity;Footnote 95 and the armed group involved displays a minimum level of organisation.Footnote 96 Whether there is an armed conflict and its type determine the applicable legal framework.
3.2. Through Whom Does a State Act?
States usually act, and are held responsible, through their armed forces.Footnote 97 These are:Footnote 98
persons, groups and units which are under a command responsible to [a party to a conflict] for the conduct of its subordinates … Such armed forces shall be subject to an internal disciplinary system which, inter alia, shall enforce compliance with the rules of international law applicable in armed conflict.
To ascertain whether the armed forces belong to a state in question, it should be established whether they receive instructions and directions from that state to carry out certain military manoeuvres or engage in specific military actions. The instructions and directions are premised on a level of control which demonstrates that the armed forces actually belong to the state in question.Footnote 99 Such affiliation does not exist solely because the armed forces have their status pursuant to domestic law, but because of the existing link of control exerted by a state over the armed forces outside the understanding of its internal organ.Footnote 100 This is in line with the reading of customary IHL rule 149 (which applies in both IAC and NIAC), according to which a state would be responsible for IHL violations attributable to it, which include:Footnote 101
(a) violations committed by its organs, including its armed forces;
(b) violations committed by persons or entities it empowered to exercise elements of governmental authority;
(c) violations committed by persons or groups acting in fact on its instructions, or under its direction or control; and
(d) violations committed by private persons or groups which it acknowledges and adopts as its own conduct.
This rule reflects ARSIWA. It lists persons who are also presumed in its Articles 4, 5, 8 and 11. Specifically, rule 149 suggests that ARSIWA would govern the question despite the debate on how to interpret the level of control needed for identifying de facto organs.Footnote 102 It is this customary IHL rule that allows a dialogue between the laws of state responsibility and of IHL.
3.3. How May the Responsibility of the State Be Engaged in an Armed Conflict When It Is Not Its Own Formal Forces that Are Involved?
In light of rule 149, clarification is needed in terms of identifying when a state would be engaged in an armed conflict when it does not act through its own armed forces. This question is particularly important in the context of NIACs with foreign state participation. Based on rule 149, from the viewpoint of the armed forces that bear the status of organs under domestic law, the level of control exerted by the state presupposes complete and total control over their actions. As for establishing whether the state is a party to an armed conflict through ‘persons or groups acting in fact on its instructions, or under its direction or control’, it is submitted that IHL is capable of offering its own solution. This solution specifies the required level of control to prove foreign state intervention in a prima facie NIAC in support of a non-state actor.Footnote 103 The following discussion looks at the level of control contained in GC III Article 4 and GC IV Article 29. Both articles apply in times of IAC and unveil potential foreign state involvement.
3.3.1. GC III Article 4
GC III Article 4 determines who is considered a prisoner of war in times of IAC.Footnote 104 Seemingly a very specific article, it has general validity because it establishes a link between a state party to the armed conflict and non-formal armed forces. Article 4A(2) includes persons who are not part of the armed forces of the state stricto sensu:
Members of other militias and members of other volunteer corps, including those of organized resistance movements, belonging to a Party to the conflict and operating in or outside their own territory, even if this territory is occupied, provided that such militias or volunteer corps, including such organized resistance movements, fulfil the following conditions: …
The formulation ‘belonging to a Party to the conflict’Footnote 105 is ‘best understood as a comment on the nature of the authority the militia or volunteers corps in question answers to overall’Footnote 106 – that is, the relationship that exists between the armed group and a sponsoring state.Footnote 107 The Appeals Chamber in Tadić considered the conditions under which armed forces may be assimilated to organs of a state other than that on whose territory they operate.Footnote 108 Under GC III, militias or paramilitary groups and units can be regarded as combatants – and therefore eligible for the status of POW – if they form ‘part of the armed forces’ of a party to the conflict (Article 4A(1)) or ‘belong to a Party to the conflict’ (Article 4A(2)) and satisfy the other four requirements (a) to (d) provided for in that article.Footnote 109
According to the ICTY Appeals Chamber, one of the logical consequences of Article 4 was that if, in an armed conflict, paramilitary units ‘belong’ to a state other than the one against which they are fighting, the conflict is of an international nature.Footnote 110 Consequently, serious violations of the Geneva Conventions may be classified as ‘grave breaches’Footnote 111 – even if committed by armed groups in situations of prima facie NIAC. The rationale for this provision is that after the Second World War it was agreed that states should be legally responsible for the conduct of all irregular forces they sponsored.Footnote 112 So that the irregulars were classed as combatants, it was required that a party to an IAC ‘controlled’ them. This type of control would result in a ‘relationship of dependence and allegiance of these irregulars vis-à-vis that Party to the conflict. These then may be regarded as the ingredients of the term “belonging to a Party to the conflict”’.Footnote 113
The Appeals Chamber, therefore, considered that the requirement of ‘belonging to a Party to the conflict’ in Article 4 was an implicit reference to a control test.Footnote 114 The 1960 Commentary to GC III also states that paragraph 4A(2) is based on experience from the Second World War, particularly with resistance movements fighting against occupation regimes.Footnote 115 Nevertheless, this does not mean that other categories, such as armed groups engaged in NIACs, are to be excluded from being considered as potentially ‘belonging to a Party to the conflict’. The Commentary does not explain the meaning of ‘belonging to a party to the conflict’. It speaks solely of the fact that resistance movements must be fighting on behalf of a party to the conflict and that:Footnote 116
it is essential that there should be a ‘de facto’ relationship between the resistance organization and the party to international law which is in a state of war, but the existence of this relationship is sufficient. It may find expression merely by tacit agreement, if the operations are such as to indicate clearly for which side the resistance organization is fighting.
There are at least three approaches to interpreting ‘belonging to a party to the conflict’. For Del Mar the ‘belonging’ requirement ‘demands nothing more than a form of acceptance, either express or tacit, on the part of the state and individuals concerned that the latter are fighting on behalf of the state’.Footnote 117 For Melzer ‘it appears essential that [the armed groups] conduct hostilities on behalf and with the agreement of’Footnote 118 the party to the conflict. The US Department of Defense Law of War Manual (2015) considers that a state's formal acknowledgement that an armed group ‘belongs’ to it is sufficient to make this armed group enjoy the privileges of combatant status, yet a mere claim by the armed group that it acts on behalf of a state would be insufficient for its members to enjoy such status.Footnote 119
The second position appears preferable, because ‘belonging’ goes beyond mere tacit agreement, which seems too little. In turn, demanding a formal acknowledgement by a state that an armed group belongs to it seems too high a threshold to require, and one that states may easily avoid if they so wish. This leads to a further question: if the armed group ‘belongs’ to the state, does it really need to be proved that it is actually controlled by it? In a way one would not need to be controlled because one already belongs to the state in question.
Yet an armed group that belongs to a foreign state is not only fighting in the interests of that state; it is also under a certain degree of control. For example, the state may instruct the group on what and how to perform certain actions. Further, when an armed group ‘belongs to a party to the conflict’, arguably it may be bound to this state by ties of dependence or allegiance, and may follow the state in its ideology.Footnote 120 Based on such a relationship, the armed group can thus be considered a de facto organ of the state.Footnote 121
In sum, the standard of control contained in ‘belonging to a party to the conflict’, based on the reading of the 1960 Commentary to GC III, is perceived here as the group receiving general directions from the state and, based on these, opposing the adversary state. The same reading of the standard of control is detected in the ICRC's recent commentary on GC III. A de facto relationship, while it signifies a factual existing relationship between the state and the armed group, indicates that it might not amount to day-to-day control from the side of the state, nor to control exercised specifically over each detail of the operation/task, but rather to control exercised in a general manner.Footnote 122 Such an interpretation of the threshold of control is preferable based on the view that generating facts that would substantiate that a state is exercising overall control over an armed group is more realistic than proving total dependency or effective control.
3.3.2. GC IV Article 29
A second possible basis in IHL for turning a foreign state into a party to an armed conflict through an armed group engaged in a conflict with another state can be found in GC IV Article 29:
The Party to the conflict in whose hands protected persons may be, is responsible for the treatment accorded to them by its agents, irrespective of any individual responsibility which may be incurred.Footnote 123
While again dealing with a very specific aspect of IAC, the focus here is on the formulation ‘its agents’. According to the travaux préparatoires, this wording refers to any employee of a state.Footnote 124 In her dissenting opinion in the Tadić case, Judge Mcdonald held Article 29 to be useful both for classifying a prima facie NIAC as an IAC, and also for establishing situations of military occupation through an armed group.Footnote 125 For Judge Mcdonald, this article barred puppet regimes from absolving their sponsoring states.Footnote 126 According to the judge, it was important to disregard the formal criteria of the military structure and focus on establishing whether the VRS was dependent on and controlled by the FRY.Footnote 127 Pursuant to the 1958 Commentary to GC IV, by means of ‘a general formula’ Article 29 defines the persons who by their acts give rise to the responsibility of a certain state.Footnote 128
The word ‘agent’ in Article 29 is worthy of further consideration. It may include anyone in the service of a state, including an armed group. Again, what has to be examined is the relationship between that outside state and the agent, in this case the armed group. The 1958 Commentary, while clarifying that the term ‘agent’ embraces a wider circle than the definition in the Fourth Hague Convention of ‘persons forming part of its armed forces’, also provides an essential reservation:Footnote 129
the word ‘agent’ limits the scope of the provision to those persons alone who owe allegiance to the Power concerned. The nationality of the agents does not affect the issue … It does not matter whether the person guilty of treatment contrary to the Convention is an agent of the Occupying Power or in the service of the occupied State; what is important is to know where the decision leading to the unlawful act was made, where the intention was formed and the order given.
Because agents show allegiance to the state, the state controls them, at least to some extent. Allegiance itself can be interpreted to mean a demonstration of devotion and of voluntary subordination to a certain entity. This amounts to a general form of control, which need not require the state to control each step of the agent. Such a type of relationship resembles overall control, as proposed by the ICTY Appeals Chamber in Tadić. The latter part of the quote above indicates the level of the foreign state's control over the armed group. Allegiance is therefore the mechanism which allows the state to wield power over the group in generally planning its actions.Footnote 130
One could object that Article 29 deals with protected persons only in times of IAC, particularly in situations of military occupation. However, it also applies in all contexts as per Common Article 2 of the Geneva Conventions. Hence, Article 29 can also be used to turn a prima facie NIAC into an IAC because it holds a possible mechanism of control, thereby linking an armed group to an outside state. Through this, a state is less likely to escape its responsibilities under IHL when acting through armed groups for which it does not wish to account but which it controls.
3.3.3. Some Possible Concerns
Deriving control tests from GC III Article 4 and GC IV Article 29 may appear to be unsuitable and problematic. First, both articles assume that an IAC exists, and employing them as the reference for the classification of a conflict may seem circular. However, this concern misses a crucial point: while these articles primarily concern straightforward scenarios of IAC, they recognise categories of persons – forming non-formal armed forces – who may be linked to a foreign state. Such persons may then be engaged in a prima facie NIAC despite not having official status under the internal law of that state. A link that determines their relationship with the state concerned is proved either in belonging to the state or in their allegiance. Both terms imply a certain level of control. Instances of belonging and allegiance need not denote manifestations of effective control, but both may very well meet the overall control test, as argued above.
Second, AP I Article 43 is used in this article to identify a party to the conflict based on its armed forces. Some would possibly argue that Article 43 does the converse – it identifies the armed forces of a given party. In other words, Article 43 would assume an IAC already to exist. This article, however, should not be viewed as only espousing such a stance. Article 43(1) could be employed to bridge the gap between situations where armed groups are fighting in a prima facie NIAC against a state, and those where armed groups are controlled by a foreign state. That state would ultimately be hindered in avoiding its status as party to an armed conflict: ‘In essence, this definition of armed forces covers all persons who fight on behalf of a party to a conflict and who subordinate themselves to its command’.Footnote 131 This definition reads as an example of foreign state control over an armed group that is fighting in a prima facie NIAC. If a state controls such an armed group, then it also acts through it, and the existence of an IAC can be established based on such a connection.
Third, it was advanced that IHL is a self-contained system. Referring to ARSIWA may thus seem incoherent for our analysis. However, being a self-contained system, IHL does not exclude the fact that there is a dialogue with other branches of international law, especially with the law of state responsibility. The existence of such a dialogue was demonstrated by IHL referencing the laws of state responsibility in rule 149 (Section 3.2 above). Hence, instead of looking at the relationship between these two branches as one of tension and differences, perhaps it is worth identifying grounds for considering them as mutually reinforcing.
4. Conclusion
In times of armed conflict states act through either their armed forces or de facto agents. Either means is foundational in establishing that a state is a party to an armed conflict. This article advocates that evaluating a state's actions in times of a prima facie NIAC on another state's territory is best carried out via the overall control test as developed by the ICTY Appeals Chamber in the Tadić case. The advantages of this approach are twofold. First, it is a more realistic benchmark for generating the required evidence than the complete dependency or effective control tests developed by the ICJ. Second, this test makes it less likely for a state to escape its responsibilities under IHL when it acts through armed groups in a prima facie NIAC on the territory of another state.
Some problems remain, however. It is highly unlikely that the ICJ, the principal judicial organ of the United Nations, will abandon its own approach to interstate disputes using the complete dependency and effective control tests for determining state responsibility for internationally wrongful acts of its de facto organs. Hence, ambiguities and contradictions in classifying conflicts and establishing state responsibility during armed conflicts will most likely remain in the future, unless grounds for their cooperation and mutual reinforcement are pursued. Consequently, one test for both state responsibility and conflict classification should exist,Footnote 132 as the responsibility of a state in an armed conflict is determined by its actions. For determining the nature of armed conflicts in which a foreign state is involved through exercising control over an armed group, the complete dependency and effective versus overall control test dichotomy should be assessed against the backdrop and within the context of IHL. This should not ignore the fact that IHL itself echoes ARSIWA. Indeed, GC III Article 4 and GC IV Article 29 themselves contain a threshold of control which may be helpful in determining the nature of armed conflicts with foreign state participation.
