Characterization of the violence between Türkiye and the PKK

Abstract

This article addresses the question of whether the violence between Türkiye and the Kurdistan Worker’s Party (PKK) constitutes an armed conflict within the meaning of international humanitarian law. The article first explains the different non-international armed conflict descriptions provided by (i) the ICTY’s famous Tadić decision, (ii) Additional Protocol II, and (iii) the Rome Statute of the ICC and discusses the different applicability thresholds set by these sources. After noting that the terrorist nature of the acts of violence in a situation will not prevent them from being characterized as an armed conflict and that the application of the international humanitarian law norms would not in fact affect the legal status of the PKK, this article examines whether Türkiye’s struggle against the PKK could be classified as an armed conflict subject to international humanitarian law norms.

1. Introduction

Armed conflicts, whether international or non-international, are governed by international humanitarian law (IHL). The Geneva Conventions and their Additional Protocols (four treaties and three Additional Protocols), which regulate the conduct of armed conflicts and seek to ameliorate the effects of armed conflicts on civilians and soldiers, form the core of IHL. Türkiye, though a party to the four Geneva Conventions of 1949, has not ratified or acceded to any of the three Additional Protocols. ¹ Thus, Additional Protocol II ² (AP II), which aims at protecting the victims of non-international armed conflicts (NIAC), would not apply to a NIAC in Türkiye except for its provisions that overlap with customary international law. Article 3 common to each of the four Geneva Conventions of 1949 (common Article 3 or CA3) would be the only Geneva Conventions-based law applicable to a NIAC in Türkiye.

Common Article 3 introduced the phrase ‘armed conflict not of an international character’ into the legal lexicon without proffering a definition of it. ³ The elucidation of requirements for the applicability of the provisions governing NIACs is thus ‘among the most difficult and controversial issues in international humanitarian law’. ⁴ While the International Committee of the Red Cross (ICRC) seeks broader recognition for the lower thresholds of the armed conflict characterization, states usually tend to refuse the applicability of IHL to the conflicts to which they are a party.

Due to the lack of a definition of a NIAC in CA3, the descriptions provided by the two conventional instruments (Additional Protocol II and the Rome Statute of the International Criminal Court) and by the subsidiary means for the determination of rules of international law (judicial decisions and teachings of publicists ⁵ ) have gained prominence. These sources, however, have set different thresholds for the applicability of the provisions governing NIACs, all of which will be addressed in this article.

The International Criminal Tribunal for the Former Yugoslavia’s (ICTY) famous Tadić decision on Interlocutory Appeal of Jurisdiction (1995) established that a NIAC can be identified by hostilities above a certain level of intensity which take place either between state forces and organized non-state armed groups or solely between organized non-state armed groups. ⁶ Accordingly, two elements need to be assessed when ascertaining the NIAC nature of a situation: a certain level of intensity of the violence produced and a certain level of the organization of the armed group(s) concerned. Subsequent ICTY jurisprudence has provided the indicative factors that demonstrate the existence of the requisite levels of intensity and organization.

The threshold set for the application of AP II, on the other hand, is considerably higher than that of CA3 as additional requirements, such as territorial control by the armed group(s) concerned, need to be met for the Protocol to apply. The scope of application of AP II, therefore, is much more limited than that of CA3. One can see the conflicts, in which AP II is applicable, as a different type (Additional Protocol II-type) of NIAC.

This article will discuss the conditions for the existence of a NIAC under CA3, AP II, and the ICC Statute, and will attempt to ascertain whether the violence between Türkiye and the Kurdistan Worker’s Party (Partîya Karkerên Kurdistanê or PKK) rises to the level of an armed conflict subject to IHL. It will offer a case study of the violence between Türkiye and the PKK from the perspective of the elements required for an armed conflict characterization and conclude that the situation in Türkiye constitutes a NIAC subject to CA3, but not to Additional Protocol II.

While some scholars opined that ‘it is at least arguable that the situation in south-eastern Türkiye does not amount to an internal armed conflict within the meaning of the law of armed conflict’, ⁷ the predominant view is that the situation in Türkiye does constitute a non-international armed conflict subject to IHL. The views against or on behalf of an armed conflict characterization, however, usually fail to engage in a careful and objective analysis of the elements of a NIAC. Most of the views on behalf of an armed conflict characterization, for example, fail to recognize the fact that the PKK has never been able to control a part of Türkiye’s territory, a sine qua non requirement for the applicability of AP II.

There exists an international court judgment that mentioned IHL as one of the relevant laws to be argued to apply to the situation in Türkiye: in the case of Benzer and Others v. Turkey in 2014, the ECtHR cited the common Article 3 of the 1949 Geneva Conventions as one of the international materials ‘relevant’ for the acts committed as part of the violence between the Turkish security forces and the PKK. ⁸ This reference, however, was made without any analysis as to whether the violence in question could be characterized as an armed conflict subject to IHL; CA3 was simply referred to as one of the potential grounds which could potentially be presented by the parties in the case as a law regulating Türkiye’s counter-terrorism efforts against the PKK. Moreover, the reference to CA3 (as one of the relevant laws to be argued to apply to the situation) has not been repeated in the ECtHR’s other judgments on Türkiye’s counter-terrorism measures.

Following this introduction, the second section explores the constituent elements of a NIAC by setting out the conditions for the applicability of CA3 in light of the Tadić decision. The third section sets out the requirements for the applicability of AP II by highlighting the additional requirements for its applicability. The fourth section explains the conditions for the applicability of the Rome Statute’s provisions governing NIACs. After stressing that the Rome Statute contains two separate lists of war crimes committed as part of NIACs, the section discusses and contributes to the two contending views relating to the question of whether the Statute brings only one common (shared) applicability threshold, which is equivalent to that of CA3.

The fifth section provides an overview of Türkiye’s struggle against the PKK and a description of its concerns regarding the armed conflict characterization of the situation thereof. The sixth section examines whether Türkiye’s struggle against the PKK could be classified as an armed conflict under CA3, AP II, and the Rome Statute by applying the relevant criteria to the factual situation in Türkiye. The final section draws a number of conclusions.

2. The ICTY’s Tadić decision and the lower (common Article 3) threshold for an armed conflict characterization

The ICTY’s Appeals Chamber famously held, in Prosecutor v. Tadić in 1995, that the requirement for the existence of a NIAC is met when there is a ‘protracted armed violence between governmental authorities and organized armed groups or between such groups within a State’. ⁹ The Chamber noted that ‘international humanitarian law applies from the initiations of such armed conflicts and extends beyond the cessation of hostilities until … a peaceful settlement is achieved’. ¹⁰ It was subsequently clarified by the ICTY that the term ‘protracted’ should be interpreted to mean ‘intense’ rather than its ordinary meaning ‘prolonged’. ¹¹ The Trial Chamber of the ICTY elaborated:

The test applied by the Appeals Chamber to the existence of an armed conflict for the purposes of the rules contained in Common Article 3 focuses on two aspects of a conflict; the intensity of the conflict and the organization of the parties to the conflict. In an armed conflict of an internal or mixed character, these closely related criteria are used solely for the purpose, as a minimum, of distinguishing an armed conflict from banditry, unorganized and short-lived insurrections, or terrorist activities, which are not subject to international humanitarian law. ¹²

According to this test, the intensity of the hostilities and the organization of the parties are the only requirements for the determination of the existence of an armed conflict. Clearly, the threshold set by the Appeals Chamber of the ICTY is considerably lower than that of AP II as the ‘territorial control’ or ‘involvement of government forces’ ¹³ are not required. This threshold is referred to as ‘the lower threshold of common Article 3’ as opposed to the ‘higher threshold of Additional Protocol II’. Accordingly, factors other than the ‘intensity of the conflict’ and ‘organization of the parties’ can only be indicative criteria rather than prerequisites.

As for the requisite degree of intensity of the hostilities, AP II’s restriction that the Protocol will only apply in situations distinguishable from ‘internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature’ ¹⁴ is also valid for CA3. ¹⁵ In this regard, the terrorist nature of the acts of violence in question will not prevent them from being characterized as an armed conflict, as long as they are not ‘single acts of terrorism’ ¹⁶ and intensity and organization elements are met. As affirmed by an ICTY Trial Chamber, ‘it is immaterial whether the acts of violence perpetrated may or may not be characterized as terrorist in nature’. ¹⁷

The low threshold for armed conflict characterization set by the Tadić decision was reaffirmed by subsequent decisions of the ICTY ¹⁸ and subsequently found its way to the Rome Statute (through a proposal from Sierra Leone). ¹⁹ The Rome Statute used the same language as the ICTY’s Tadić decision for addressing NIACs, except that the term ‘protracted armed violence’ was replaced by ‘protracted armed conflict’: ‘armed conflicts that take place in the territory of a State when there is protracted armed conflict between governmental authorities and organized armed groups or between such groups’. ²⁰ Whether this discrepancy (replacement of the word ‘violence’ with ‘conflict’) aimed at bringing an additional threshold or the new sentence merely represents reformulation of the existing threshold has been debated extensively. ²¹ In order to understand and contribute to these debates, the fourth section will analyse the articulation of Article 8 of the Rome Statute, which deals with war crimes.

3. Additional Protocol II threshold for an armed conflict characterization

As discussed above, identification of a NIAC requires the violence concerned to cross a certain level of intensity and the armed group(s) concerned to be sufficiently organized. Additional Protocol II, however, has introduced additional requirements for its applicability. Article 1(1) of AP II sets out that the protocol applies to NIACs:

which take place in the territory of a High Contracting Party between its armed forces and dissident armed forces or other organized armed groups which, under responsible command, exercise such control over a part of its territory as to enable them to carry out sustained and concerted military operations and to implement this Protocol. ²²

Accordingly, in addition to the intensity of the hostilities and organization of the armed group concerned, AP II requires (i) that the conflict takes place in the territory of a state party; (ii) that the conflict is between state forces and non-state armed group(s) (that state forces be involved in the conflict); (iii) that the non-state armed group concerned is under responsible command; and (iv) that the non-state armed group concerned exercises such control over a part of the territory of the state party concerned as to enable the group ‘to carry out sustained and concerted military operations and to implement’ the Protocol. ²³

Here, it should be noted that the ‘sustained and concerted’ nature of the conflict and actual implementation of AP II (respect for IHL) by the armed group concerned are not required for the Protocol to be applicable. ²⁴ Having the territorial control sufficient to allow the armed group ‘to carry out sustained and concerted military operations and to implement’ the Protocol is what is required by the language of Article 1(1). ²⁵

That the violence concerned must reach a degree of intensity that rises above sporadic acts of violence, on the other hand, is still a requirement as it is one of the constituent elements of a NIAC. That is to say, the ability of the armed group to engage in protracted armed violence is not enough; there should actually be a conflict with a certain degree of intensity. In this regard, Article 1(2) of AP II expressly warns that the ‘Protocol shall not apply to situations of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature, as not being armed conflicts.’ ²⁶

4. The Rome Statute threshold for an armed conflict characterization

Article 8 of the Rome Statute contains two separate provisions that enumerate a list of offences committed during armed conflicts not of an international character and each is followed by another section that sets out the scope of application. ²⁷ While Article 8(2)(c) provide a list of serious violations of CA3 (and is followed by Article 8(2)(d) which sets out the scope of application), Article 8(2)(e) enumerates a list of ‘other serious violations of the laws and customs applicable in armed conflicts not of an international character’ (and is followed by Article 8(2)(f) which sets out the scope of application).

Both Article 8(2)(c) (serious violations of CA3) and Article 8(2)(e) (other serious violations of the laws and customs applicable in NIACs) are explicitly stated not to apply ‘to situations of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence or other acts of a similar nature’. ²⁸ This negative formulation is taken directly from Article 1(2) of AP II. ²⁹ However, AP II’s other restrictive elements such as territorial control or involvement of state forces have not been imported.

It is important to note that Article 8(2)(f), which defines the scope of application of Article 8(2)(e), contains an additional sentence that states that Article 8(2)(e) will apply to ‘armed conflicts that take place in the territory of a State when there is protracted armed conflict between governmental authorities and organized armed groups or between such groups’. ³⁰ This sentence is taken directly from the ICTY’s Tadić decision except that the term ‘protracted armed violence’ was replaced by ‘protracted armed conflict’. ³¹

The said additional sentence gives rise to the same question asked at the end of the second section: Does this additional sentence and the replacement of the word ‘violence’ with ‘conflict’ represent an additional requirement that would raise the threshold of applicability of Article 8(2)(e) to a level above that of CA3 but below that of AP II, thereby creating a new (intermediate) category of NIAC? Or does it merely rephrase (and clarify) Article 8(2)(d)?

The facts that (i) the Rome Statute included two separate lists of war crimes relating to the situations of armed conflicts not of an international character, ³² (ii) the phrase ‘protracted armed conflict’ was preferred instead of ‘protracted armed violence’, and (iii) the ‘protracted armed conflict’ requirement referred only to the second list (Article 8(2)(e)) exacerbate the doubts about the uniformity of applicability in 8(2)(d) and 8(2)(f) of the Rome Statute. While some scholars, including Marco Sassoli, Antoine Bouvier, and René Provost suggested that Article 8(2)(f) of the Rome Statute represents a new (additional) threshold, some scholars including Theodor Meron, Michael Bothe, and Claus Kress opposed such understanding. ³³ The opponents claim that Article 8 of the Rome Statute brings only one common (shared) threshold, which is equivalent to that of CA3 and applies to both Article 8(2)(c) and (e).

The ICC’s jurisprudence too implicitly admitted the lack of clarity on the matter without drawing a clear conclusion:

The Chamber is also mindful that the wording of Article 8(2)(f) of the Statute differs from that of Article 8(2)(d) of the Statute, which requires the existence of a ‘protracted armed conflict’ and thus may be seen to require a higher or additional threshold to be met- a necessity which is not set out in article 8(2)(d) of the Statute. The argument can be raised as to whether this requirement may nevertheless be applied also in the context of Article 8(2)(d) of the Statute. However, irrespective of such a possible interpretative approach, the Chamber does not deem it necessary to address this argument, as the period in question covers approximately five months and is therefore to be regarded as ‘protracted’ in any event. ³⁴

Apart from the possibility of establishment of such extension by the jurisprudence of the ICC in the future, this author is of the view that the difference in the wording of Article 8(2)(d) and (f) suggests that the violence that is sufficiently intense but not protracted would fail to satisfy Article 8(2)(f) but could still cross the threshold of the 8(2)(d), and thus CA3 protections of Article 8(2)(c) would still apply. The reason why the chambers of the ICC have not so far categorized NIACs as those that meet the lower threshold of CA3 and those that cross the so-called intermediate ³⁵ threshold of Article 8(2)(f) may be that all conflicts that have been so far found sufficiently intense by the Court were also protracted in nature anyway. While it is true that the protracted nature of a conflict is one of the indicative factors that may assist in the determination of the intensity of the conflict, ³⁶ such nature is, at the same time, a sine qua non requirement for the applicability of the Article 8(2)(e) provisions.

The preparatory work of the Rome Statute and the circumstances of its conclusion, which are ‘supplementary means of interpretation’ as specified by the Vienna Convention on the Law of Treaties, ³⁷ too suggest that referral to ‘protracted armed conflict’ was made in order to satisfy the states that favoured a higher threshold for the applicability of Article 8(2)(e), and thus to secure their support in the adoption of the Rome Statute. In the early drafts of the Rome Statute, ‘serious violations of Article 3 common to the four Geneva Conventions of 12 August 1949’ and ‘[o]ther serious violations of the laws and customs applicable in armed conflicts not of an international character’ was included in a single provision. Some states present at the Rome Conference (Egypt, Sudan, and Bahrain), ³⁸ however, insisted that as opposed to the lower threshold of applicability for the serious violations of common Article 3, the higher AP II threshold of applicability should be set for the ‘other serious violations’. ³⁹

This suggestion, which was in turn introduced as a proposal by the Conference Bureau, ⁴⁰ would, if adopted, exclude conflicts fought between non-state armed groups (without the involvement of government forces) and the conflicts in which the anti-government armed group did not exercise territorial control from the jurisdiction of the Court. Thus, a number of states were strongly opposed to the proposal.

In an attempt to balance the expectations of the states favouring broader applicability (without any distinction between the serious violations of CA3 and the other serious violations) with the concerns of the states that preferred the higher AP II threshold for the latter, Sierra Leone introduced a proposal which has subsequently become the final version of Article 8(2)(f). ⁴¹ Accordingly, Rome Statute provisions on serious violations other than that of common Article 3, namely Article 8(2)(e), apply to ‘armed conflicts that take place in the territory of a State when there is protracted armed conflict between governmental authorities and organized armed groups or between such groups’. ⁴²

In this regard, it should first be noted that the ICC’s jurisprudence on the identification and characterization of armed conflicts has not diverged from the ICTY. Chambers of the ICC and the Office of the Prosecutor have consistently cited and relied on the jurisprudence from the ICTY. ⁴³ According to a comprehensive definition adopted and affirmed by the chambers of the ICC, an armed conflict not of an international character is characterized:

by the outbreak of armed hostilities of a certain level of intensity, exceeding that of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence or other acts of a similar nature, and which takes place within the confines of a State territory. The hostilities may break out (i) between government authorities and organized dissident armed groups or (ii) between such groups. ⁴⁴

In line with the jurisprudence of the ICTY and the definition set out above, the jurisprudence of the ICC has made it clear that the Court focuses on two aspects of the hostilities in order to distinguish a NIAC from less serious forms of violence: intensity of the confrontations and the organization of the non-state armed group(s) party to the hostilities. ⁴⁵ Other factors, such as territorial control or involvement of government forces are not determinative but are usually relevant and important factors.

While the determination of the existence of an armed conflict does not require the violence in question to be protracted in nature, such nature is still a requirement for the applicability of the provisions contained in Article 8(2)(e) (as it is expressly envisaged by Article 8(2)(f)). The Pre-Trial Chamber’s observation in the Prosecutor v. Jean-Pierre Bemba Gombo that ‘the period in question covers approximately five months and is, therefore, to be regarded as “protracted” in any event’ ⁴⁶ indicates that the duration needed to fulfil the ‘protracted armed conflict’ requirement for the purposes of the applicability of the Article 8(2)(e) provisions can be as short as five months or even shorter than that depending on other factors.

5. Türkiye’s struggle against the PKK and concerns about the armed conflict characterization

The PKK is an armed group that is designated as a terrorist organization by much of the international community, including the European Union, ⁴⁷ NATO, ⁴⁸ Türkiye, ⁴⁹ the United States, ⁵⁰ Canada, ⁵¹ Japan, ⁵² Australia, ⁵³ New Zealand, ⁵⁴ and so on. Its request seeking removal from the EU terrorist list has recently been rejected by the European Court of Justice. ⁵⁵

Abdullah Öcalan, the founding leader of the PKK who is now serving life in prison in Türkiye, acknowledged that:

Turkish authorities had correctly estimated, or even under-estimated, the number of people killed and wounded in the PKK’s armed actions since 1978 – with the deaths amounting to 4,472 civilians, 3,874 servicemen, 247 police officers and 1,225 village guards in 6,036 armed attacks, 3,071 bombings, 388 armed robberies and 1,046 abductions. ⁵⁶

Today, the number of civilians killed by the PKK exceeds 7,000. ⁵⁷

Türkiye does not characterize its operations against the PKK as an armed conflict subject to the IHL norms but as ‘domestic counter-terrorism’ subject only to international human rights law, on such grounds as that the PKK does not exercise territorial control needed to carry out sustained and concerted military operations and it is a terrorist organization that does not respect the rules of war. ⁵⁸ In this regard, Türkiye is concerned that its acknowledgment of the applicability of IHL to its operations against the PKK would entail admitting that Türkiye is unable to stop large-scale violence within its borders and that the PKK wields control over a part of Türkiye’s territory. ⁵⁹

Türkiye is also concerned that characterization of the violence between Türkiye and the PKK as an armed conflict within the meaning of IHL could offer some advantages to the PKK and impose additional burdens on Turkish security forces. Such characterization could, Türkiye fears, bestow the PKK with moral legitimacy, and facilitate the PKK’s claims that it has the requisite legal ‘personality to carry on diplomacy and participate in peace conferences’ ⁶⁰ or it could confer combatant status on the PKK members.

Contrary to the concerns of Türkiye, however, the application of IHL would not, at least legally speaking, affect the legal status of the PKK (and its terrorist activities), as emphasized by the last sentence of CA3. ⁶¹ Unlike IHL of IACs, which confers on a combatant the right to target an enemy combatant and be immune from prosecution for such action, under IHL of NIACs, a person who engages in hostilities against the state forces would not be accorded the status of combatant. ⁶² Consequently, he would not, if captured, enjoy the rights of prisoner of war status and would not be immune from criminal prosecution for attacking state forces or properties. Therefore, acknowledging the applicability of IHL of NIACs would not in fact require Türkiye to treat the captured PKK militants as prisoners of war nor to accord them immunity from domestic legislation for attacking security forces. Concordantly, AP II warns that the Protocol will not affect the responsibility of a government ‘to defend the national unity and territorial integrity of the State’. ⁶³

Some states, such as Colombia, refused, for years, to join the states that ratified AP II ‘for fear that to do so would be to recognize the armed groups as belligerents and grant them status’. ⁶⁴ Here it might be useful to briefly explain the concept of belligerency. Traditionally, recognition of a situation as belligerency was tantamount to recognition of civil war and brought responsibilities for the recognizing states. If a situation meeting certain factual preconditions – such as the existence of hostilities beyond the scope of mere local unrest or occupation of a substantial part of the territory of the state by the insurgents – was recognized as belligerency by the parent state, it was regulated by the full body of the laws of war and the belligerents were treated as combatants, not as traitors. ⁶⁵ Third-party states recognizing belligerency, on the other hand, were bound by the law of neutrality and prohibited from offering assistance to belligerents or the state. ⁶⁶ Although recognition of belligerency has fallen into disuse, ⁶⁷ ‘it still exists in international law’ ⁶⁸ and ‘a situation may still be recognized as belligerency’. ⁶⁹

Recognition of the applicability of AP II, however, ‘is in no way tantamount to recognition of belligerency’. ⁷⁰ The Diplomatic Conference that negotiated and adopted the four 1949 Geneva Conventions passed a resolution which provided that conditions of recognition of an armed group as belligerent ‘are governed by the general rules of international law on the subject and are in no way modified by the Geneva Conventions’, demonstrating that rules governing NIACs are ‘not intended to apply only in situations in which belligerency could be recognized’. ⁷¹

It should be noted that the application of CA3 would not, in fact, impose additional burdens on Türkiye as it is also party to both the European Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR) and the International Covenant on Civil and Political Rights (ICCPR). As stated by the Inter-American Commission on Human Rights, ‘the provisions of common Article 3 are essentially pure human rights law’. ⁷² In this regard, human rights instruments binding on Türkiye are even more extensive than CA3. ⁷³

Finally, Türkiye is deeply concerned that the classification of its internal hostilities as an internal armed conflict and recognition of the applicability of IHL could lead to allegations of war crimes against the Turkish security forces. In this regard, Türkiye is especially concerned that individual abuses by the security personnel will be ‘turned into gross generalizations of violations of human rights by the PKK propaganda machine’ and that Türkiye’s counter-terrorism operations, which are already under the scrutiny of the ECtHR, could be undermined by the politically motivated allegations of war crimes. ⁷⁴ This concern also constitutes one of the biggest obstacles to Türkiye’s accession to the International Criminal Court (ICC). ⁷⁵

Türkiye’s refusal to characterize its operations against the PKK as an armed conflict, nevertheless, is not determinative. The following section will deal with the issue of characterization of the violence between Türkiye and the PKK.

6. Characterization of the situation in Türkiye

The second section of this article explained that two requirements need to be satisfied for the determination of the existence of an armed conflict: the sufficient degree of organization of the non-state armed group concerned and the sufficient degree of intensity of the violence. The following headings will address, in light of the indicative factors of the constituent elements of a NIAC, the question of whether the violence between Türkiye and the PKK satisfies the two constituent elements; and if so, whether the AP II would be applicable if Türkiye ratifies it.

6.1 Organization requirement

The ICC, whose jurisprudence on the identification of armed conflicts has not diverged from the ICTY, provides an up-to-date list of factors that may be taken into account when assessing the two constituent elements of a NIAC. ⁷⁶ Accordingly, whether an armed group is organized may be assessed by taking into account, inter alia, the following indicators:

i) the existence of a command structure, the existence of headquarters, the issuing of political statements, and the use of official spokespersons; (ii) the military (operational) capacity of the armed group, which may be shown by, for example, the ability to define a unified military strategy, the use of military tactics, the ability to carry out (large scale or coordinated) operations, the control of territory, and having a territorial division into zones of responsibility; (iii) the logistical capacity of the armed group, indicated, among others, by the existence of a supply chain for military equipment, as well as by the group’s ability to move troops around and to recruit and train personnel; (iv) the existence of an internal disciplinary system and the ability to implement IHL; and (v) the group’s ability to speak with one voice, indicated, for example, by the capacity of the leadership to act on behalf of its members in political negotiations and to conclude agreements, such as cease-fire or peace agreements. ⁷⁷

The PKK has a significant number of members. The number of active PKK militants between 1991 and 1993, when the PKK was at the summit of its power, is estimated to range from 10,000 to 20,000. ⁷⁸ The current number of the PKK’s armed members, according to the US Department of State’s most recent ‘Country Reports on Terrorism’, is estimated to be 4,000 to 5,000. ⁷⁹ The PKK’s ‘capacity to declare and hold ceasefires’ with the instruction of its captured leader suggests that it is ‘under responsible command’. ⁸⁰ It has effective control over its members and an effective hierarchical structure with identifiable ranks and positions. The PKK also has the capacity to acquire weapons, recruit and train new members. Its attacks are usually carried out as part of an organizational plan and policy.

Though these qualifications together seem to suggest that the PKK is an organized armed group and thus meets one of the two constituent elements of a non-international armed conflict, its failure to meet some of the crucial indicative factors of the organization criterion makes it difficult to determine whether the PKK could be considered ‘sufficiently’ organized to be a party to an armed conflict against Türkiye. Indeed, the PKK’s inability to control territory, ⁸¹ to implement IHL, ⁸² or to carry out sustained large-scale operations to oust the state’s authority, ⁸³ as well as that the organization routinely carries out terrorist attacks that target civilians, ⁸⁴ increase the doubts about the sufficiency of its organization.

Nevertheless, this author agrees with the generally accepted view that the PKK is sufficiently organized and under responsible command, as suggested by its significant number of members, effective hierarchical structure, capacity to recruit and train new members, and equip them with weapons and uniforms. The group satisfies the majority of the indicative factors emphasized by the jurisprudence of the ICTY and ICC for the determination of the organization of a non-state armed group, such as ‘the existence of a command structure’, ‘ability … to recruit and train personnel’ and ‘ability to speak with one voice, indicated, for example, by the capacity of the leadership to act on behalf of its members in political negotiations’. ⁸⁵ The PKK can therefore be considered sufficiently organized to be a party to an armed conflict, thereby satisfying one of the two criteria for the non-international armed conflict characterization.

6.2 Intensity requirement

Whether the degree of intensity required for the non-international armed conflict characterization is crossed in a situation of violence may be determined by taking into account, inter alia, the following factors:

(i) the seriousness and frequency of attacks and armed clashes; (ii) the spread of clashes over territory and the group’s ability to control territory over a period of time; (iii) whether any ceasefire orders had been issued or ceasefires agreed to; (iv) the type and number of armed forces deployed, including any involvement of the government; (v) the type of weapons used; (vi) whether the situation had attracted the attention of the UN Security Council, or involvement of other international organisations; (vii) whether those fighting considered themselves bound by IHL; and (viii) the effects of the violence on the civilian population, including the extent to which civilians left the relevant area, the extent of destruction, and the number of persons killed. ⁸⁶

The PKK mostly avoids pitched battles with state forces. ⁸⁷ It usually employs hit-and-run guerrilla warfare strategy. Its attacks against the security forces take place chaotically, usually as hit-and-run ambushes, suicide or car-bomb attacks, sabotage, road-side attacks, abductions, and mine traps, each time in different areas. ⁸⁸ Moreover, the PKK does not usually show considerable resistance to Türkiye’s operations in the areas where it hides, except for its ambushes, land mines, or booby traps.

Though the types of these attacks may conduce one to consider that there are no serious direct armed clashes that reach the requisite degree of intensity, there have been some periods and several clashes that were intense enough to trigger the applicability of IHL. Especially, the early 1990s and the period between August 2015 and March 2016 witnessed the most violent engagements between the PKK members and security forces. These periods will be detailed under the following two headings. Before moving to discuss the intensity of the violence in these two periods in detail, it should be noted that once triggered, the international humanitarian law will apply from the initiation of the armed conflict concerned and will extend ‘beyond the cessation of hostilities until … a peaceful settlement is achieved’. ⁸⁹ In other words, once the threshold of intensity is crossed, IHL will continue to apply even if the violence falls below a certain level of intensity or in the absence of active hostilities until ‘a lasting cessation of armed confrontations without real risk of resumption’. ⁹⁰

6.2.1 Intensity of the violence produced in the 1990s

The PKK was founded in 1978. Its first attack directly targeting the Turkish security forces was carried out on the night of 15 August 1984. ⁹¹ Its biggest and most violent attacks were carried out in the early 1990s. On 15 May 1992, for instance, a group of 600 PKK militants attacked a gendarmerie outpost in Taşdelen village of Şırnak province, as a result of which 27 soldiers and 40 PKK militants were killed. ⁹² Only a few months later, on 29 September 1992, a group of 500 PKK militants attacked a gendarmerie border outpost in the Derecik district of Hakkari province, as a result of which 28 soldiers and 174 PKK militants were killed. ⁹³ This PKK attack is the one with the highest number of soldiers killed since the beginning of the conflict ⁹⁴ (except for the execution of 33 unarmed soldiers, on 24 May 1993, who were traveling by a civilian bus to surrender to their military units ⁹⁵ ). Super Cobra helicopters and war jets were employed to attack the militants, ⁹⁶ ‘many of whom apparently escaped into Iraq after the raid’. ⁹⁷

Two days later, on 1 October 1992, a group of 100 PKK militants raided Cevizdalı village of Bitlis province and ‘killed twenty-nine people, including many women and children’. ⁹⁸ The PKK’s adoption of the ‘murder of civilians as open policy’ exacerbated the effects and scope of the conflict. ⁹⁹ The civilian population of certain southeastern districts was, therefore, profoundly affected by the conflict, especially throughout the 1980s and early 1990s. In 1987 alone, the PKK attacked 14 villages, killing 242 villagers and burning down 137 schools. ¹⁰⁰

The state’s forcible evacuation of the more than 2,500 villages and hamlets that were under the threat and/or influence of the PKK during the 1980s and 1990s (especially in the first half of the 1990s) is another indicator of the intensity and civilian cost of the violence produced in that period. ¹⁰¹ Likewise, the imposition of emergency rule from 1987 to 2002 in eastern and southeastern provinces most affected by the conflict is among indicators of the seriousness of the situation in the 1980s and 1990s. ¹⁰²

The PKK carried out 4,198 attacks in 1993, ¹⁰³ 3,538 attacks in 1994, ¹⁰⁴ and 2,200 attacks in 1995. ¹⁰⁵ 11,546 PKK militants, 3,621 security personnel, and 5,014 civilians had already been killed by the first half of the 1990s as a result of PKK attacks. ¹⁰⁶ Some of Türkiye’s operations, too, have resulted in intense confrontations. Though the PKK has mostly avoided pitched battles in favour of hit-and-run attacks and ambushes, ¹⁰⁷ there have been rare cases where the PKK units responded to Türkiye’s military operations against the PKK hideouts with a conventional type of combat ‘and in turn suffered heavy losses’. ¹⁰⁸

On 12 October 1992, only two weeks after the PKK’s Derecik attack, for example, 15,000 Turkish ground troops backed by tanks, helicopters, and fighter jets, ¹⁰⁹ entered northern Iraq ‘where some PKK holdouts initially refused to put down their weapons’. ¹¹⁰ 1,452 PKK militants were killed or wounded, according to Turkish General Staff, as a result of this cross-border operation that lasted 20 days. ¹¹¹ The PKK acknowledged only 161 casualties. ¹¹² Turkish military destroyed much of the PKK’s infrastructure and ammunition depots in northern Iraq ¹¹³ and ‘hoisted the nation’s flag in Haftanin’, ¹¹⁴ a PKK holdout nestled in the mountains of northern Iraq which ‘had grown into an important gathering and training camp for the PKK’. ¹¹⁵ The operation was carried out with the apparent support of the Iraqi Kurdish forces (PUK and KDP) that fought alongside the Turkish military. ¹¹⁶

The PKK, which suffered severe losses in its attempt to fight a conventional type of combat in October 1992, later ‘returned to areas [in northern Iraq] vacated by Turkish troops’, ¹¹⁷ but ‘never again tried to defend territory and instead relocated fighters as necessary’ ¹¹⁸ (except for the clashes in 2015 and 2016 which will be covered in the following heading). The 1990s, nevertheless, saw larger scale cross-border operations targeting the PKK hideouts in northern Iraq. On 20 March 1995, for instance, as many as 35,000 Turkish soldiers entered northern Iraq, launching a new major anti-PKK operation that lasted 43 days ¹¹⁹ in hopes of finally ending the PKK’s ‘ability to use the border strip to launch attacks’. ¹²⁰ It was reported that 568 PKK militants were killed as a result of the operation. ¹²¹ Clearly, Türkiye’s cross-border operations into the PKK bases in northern Iraq have greatly expanded the geographical scope as well as the human and material costs of the conflict.

6.2.2 Fierce clashes between August 2015 and March 2016

The PKK attempted to prevent policing functions by Turkish security forces in several neighbourhoods of four southeastern provinces in Türkiye by digging trenches and setting up barricades at the doorsteps of the civilian population between August 2015 and March 2016, following the end of a period called ‘solution process’. ¹²² The solution process, which lasted from March 2013 to July 2015, was a dialogue process that aimed at persuading the PKK, by peaceful means, to lay down its arms. The government’s policy of overlooking the PKK mobilization during the solution process encouraged the PKK to spread the conflict to urban areas by erecting barricades on the streets of several Kurdish neighbourhoods, planting explosives and booby traps in these barricades, and using civilian houses as shelters. The PKK’s attempts to keep security forces out of several neighbourhoods by building barricades and trenches following the collapse of the solution process were, as will be detailed below, all short-lived and far from successful but they led to fierce clashes in several neighbourhoods.

Curfews were imposed to prevent civilian casualties in the neighbourhoods concerned during the operations to remove the trenches and to clear these neighbourhoods of PKK militants. While the operations in some neighbourhoods lasted for several days, operations in certain neighbourhoods lasted for weeks and the curfews were held even longer to clear leftover explosives. The said operations took place in different neighbourhoods of four southeastern provinces (Diyarbakır, Mardin, Şırnak, and Hakkari), with the 103-day-long operations in the Sur district of the Diyarbakir province being the longest-lasting. ¹²³ Operations in the Derik, Dargeçit, and Nusaybin districts of the Mardin province lasted for six, 18, and 82 days, respectively. ¹²⁴ Operations in the İdil, Silopi, and Cizre districts of the Şırnak province, on the other hand, lasted for 20, 37, and 78 days, respectively. ¹²⁵ Finally, operations in the Yüksekova district of the Hakkari province lasted for 37 days. ¹²⁶

The PKK’s attempts to spread the conflict to urban areas by erecting and defending the barricades on the streets of several Kurdish neighbourhoods between August 2015 and March 2016 claimed the lives of 859 security personnel and 335 civilians. ¹²⁷ 4,571 PKK militants were killed during the operations to remove the barricades and trenches. ¹²⁸ More than 355,000 civilians left their homes in the conflict area. ¹²⁹ These figures support this section’s argument that there were intense confrontations between August 2015 and March 2016. A great number of buildings had to be demolished after the end of the clashes ‘because PKK militants had planted bombs in them and dug tunnels beneath them’. ¹³⁰ ‘During the clean-up of rubble … security forces were detonating around 70 explosives a day planted in the houses. This was one of the main reasons why the whole process took so long in Nusaybin.’ ¹³¹

The above-described (i) human and material costs, (ii) temporal and geographical scope of the conflict, (iii) effects of the conflict on civilians, including the extent to which civilians left the conflict areas and extent of the destruction, (iv) as well as the nature of the state’s response and the type of weapons used all suggest that the violence between Türkiye and the PKK is sufficiently intense to constitute an armed conflict within the meaning of IHL. The conclusion for this section is that both the requisite degree of organization criterion and the requisite degree of intensity criterion seem to be fulfilled in the context of Türkiye-PKK confrontations. The violence between Türkiye and the PKK, therefore, constitutes a NIAC within the meaning of IHL and is subject to CA3.

As for the applicability of the Rome Statute, Section 4 explained that the Statute contains two separate lists of war crimes committed as part of NIACs – Article 8(2)(c) and Article 8(2)(e) – and that the protracted nature of a conflict is a sine qua non requirement for the applicability of the latter. While the determination of the existence of an armed conflict does not require the violence in question to be protracted in nature, such nature is still a requirement for the applicability of the provisions contained in Article 8(2)(e). Given that the duration needed to fulfil the ‘protracted armed conflict’ requirement for the purposes of the applicability of Article 8(2)(e) can be as short as five months or even less, ¹³² it is not difficult to conclude that nearly 40 years of violence between Türkiye and the PKK is protracted in nature. Both Article 8(2)(c) and Article 8(2)(e) of the Rome Statute would, therefore, be applicable to the situation in Türkiye if the preconditions to the exercise of jurisdiction of the ICC were to be satisfied (e.g., if Türkiye was a state party).

Additional Protocol II, however, would not be applicable to the situation in Türkiye even if Türkiye was a party to it, because the PKK has never been able to exercise the territorial control needed to implement the Protocol, which will be the subject of the next heading. As there are conflicting views on whether or not the PKK ever exerted territorial control, a separate heading is devoted to the applicability of AP II to the situation in Türkiye.

6.3 Applicability of Additional Protocol II

As detailed in Section 3, territorial control by the armed group concerned is a sine qua non requirement for the applicability of AP II. In order for AP II to be applicable to a non-international armed conflict, the armed group concerned must exercise such control over a part of the territory of the state concerned as to enable the group ‘to carry out sustained and concerted military operations and to implement’ the Protocol. ¹³³ Territorial control, on the other hand, can be said to exist only if the armed group in question is able ‘to move freely, access information and resources, and prevent its enemies’ movement and access in a particular place and time’. ¹³⁴

In the late 1980s and early 1990s, the PKK’s main goal was to establish a ‘liberated zone’ it had dreamed of since its foundation in an area encompassing three southeastern provinces (Hakkari, Siirt, and Van) where it often carried out ambushes and attacked gendarmerie stations. ¹³⁵ The Turkish security forces, however, usually responded quickly to the sporadic attempts by the PKK to exercise territorial control. The group contested the state for control of several southeastern districts in March 1992 but the state forces managed to maintain control ‘after a few days of heavy clashes’. ¹³⁶ As will be detailed below, the PKK was never able to take and hold territory, ¹³⁷ or exert complete territorial control over even rural areas, not even in the early 1990s, the period of its greatest success. ¹³⁸

Though the PKK carried out large-scale attacks against the gendarmerie border outposts in the early 1990s, the two largest of which mentioned above, it was never able to take over the control of outposts or check-points of the Turkish security forces even in the most rural areas. ¹³⁹ The reason the PKK has relied solely on hit-and-run ¹⁴⁰ and terrorist ¹⁴¹ attacks in Türkiye was the fact that it was never able to take and hold the territory needed ‘to carry out sustained and concerted military operations’. ¹⁴²

One of the PKK’s main aims in 1990 was to be able to hold its fourth congress in Türkiye. ¹⁴³ However, since it could not find a safe haven even in the rough, mountainous terrains of southeastern Türkiye, the fourth congress was held in northern Iraq. ¹⁴⁴ As a matter of fact, the PKK could not hold any of its congresses in Türkiye, despite its great desire.

Some argued that the PKK exerted control over some districts in southeastern Türkiye in the early 1990s. Aliza Marcus, for example, stated in her book that ‘to break the PKK’s control over parts of the southeast, Turkish soldiers conducted frequent house raids, looking for signs of PKK support and weapons’. ¹⁴⁵ However, the PKK control Marcus mentioned, in which the PKK militants were so invisible that house raids were necessary to find them and their weapons, is not the type of control that would trigger the applicability of AP II.

Those who referred to the PKK’s ‘control’ in the early 1990s were, in fact, generally referring to the group’s ability to terrorize parts of the southeast and to the Turkish forces’ inability to ‘protect each village’. ¹⁴⁶ In this regard, the PKK’s burning of schools and killing of teachers in rural villages in the late 1980s and early 1990s cannot be considered as territorial control within the meaning of IHL; they can only show the state’s failure to take efficient measures to protect those areas from the PKK attacks. Most objective observers do not dispute that the PKK was never able to take and hold territory or gain exclusive territorial control over a part of Türkiye, not even in the early 1990s.

The fact that the PKK is hiding and headquartered in the ‘uncontrolled’, ¹⁴⁷ extremely rugged, and remote mountains of northern Iraq since the 1990s, is a proof of its inability to build a base or hide within Türkiye, let alone to control a territory therein. Even the most rural areas and the mountainous terrain in east and southeast Türkiye, where the majority of the PKK’s attacks took place, are tightly controlled and patrolled by the security forces; and the caves that are used by the PKK members as shelters are destroyed during these routine patrols. ¹⁴⁸ Even in the early 1990s when the PKK lived its ‘golden years’, ¹⁴⁹ the Turkish armed forces routinely patrolled the rural areas of southeastern Türkiye to destroy the caves and dugouts that served as shelters for the PKK militants.

If the PKK had apparent and permanent bases in the area and was patrolling the area in search of Turkish soldiers, then it could be argued that the area was under the PKK’s control within the meaning of the applicability of AP II. Clearly, this has never been the case in Türkiye. It has always been Türkiye that patrolled the rural areas of the southeast in search of PKK militants, their hideouts, and weapons. The PKK, on the other hand, was the party trying to ‘infiltrate’ the Turkish territory from its bases in the ‘uncontrolled’ mountainous terrain in northern Iraq to conduct hit-and-run attacks against the Turkish state forces. ¹⁵⁰

Some observers refer to the PKK’s ability to manage smuggling and drug trafficking activities in the border area as ‘PKK control in the region’, ¹⁵¹ without discussing whether the group ever gained exclusive territorial control. Türkiye or Iraq’s failure to protect these areas from such PKK activities as managing smuggling activities, however, does not indicate the existence of territorial control by the PKK within the meaning of applicability of AP II. As explained by Marcus, ‘it was impossible to fully secure the high mountains and deep ravines that ran along the border’. ¹⁵² The PKK was far from gaining exclusive and complete territorial control over even rural areas in northern Iraq.

Türkiye has carried out countless military operations into northern Iraq to eliminate the PKK bases which are mostly underground and hidden in the Qandil Mountains. The PKK once tried to defend territory in northern Iraq in October 1992, in the face of the above-mentioned major cross-border operation into PKK bases there. However, due to the heavy losses, the PKK ‘never again tried to defend territory and instead relocated fighters as necessary’. ¹⁵³ It should nevertheless be noted that even if the PKK had territorial control in northern Iraq and Türkiye was a party to AP II, that would not be sufficient to trigger the applicability of AP II to the conflict between Türkiye and the PKK because the Protocol requires the non-state armed group concerned to exercise control over a part of the territory of the state party concerned (in this case Türkiye). ¹⁵⁴

As detailed in the previous heading, the PKK also attempted to control several neighbourhoods in Türkiye’s four southeastern provinces between August 2015 and March 2016, by digging trenches, building roadblocks and booby traps to keep security forces out. These efforts, however, were all short-lived ¹⁵⁵ and were far from successful, as evidenced by the fact that even the entries and exits of the neighbourhoods concerned were fully controlled by the security forces and the PKK militants were only able to move through the tunnels. ¹⁵⁶

It is clear that the PKK’s short-lived efforts to protect the barricades and trenches by such methods as using civilian houses as shelters and planting mines or other types of explosives cannot be considered as ‘territorial control’ within the meaning of applicability of AP II. In fact, these efforts were not even based on a realistic evaluation of whether they could actually result in territorial control, but were a result of the government’s orders to the security forces and local authorities to ignore the PKK mobilization in the region for the purpose of maintaining the solution process. ¹⁵⁷

The reason barricades were not destroyed within minutes (by, for example, an aerial bombardment) was not the PKK’s ability to defend the barricades but the fact that the group was using civilian houses as shelters. The security forces’ efforts not to endanger civilian locals are the main reason why the PKK has not managed to get Türkiye’s Kurds (who are estimated to account for 15–18 percent of Türkiye’s population of about 85 million ¹⁵⁸ ) to support its separatist violence and why Türkiye has not experienced a civil war or a large-scale prolonged armed violence within its borders as neighbouring Syria. Due to the endangered civilian locals’ refusal of the PKK’s calls to support its efforts to build and stand behind the barricades ¹⁵⁹ and the PKK’s inability to engage in sustained armed confrontations with the state forces, the PKK’s efforts to establish territorial control over certain neighbourhoods and to incite a civil war were far from successful.

As for the applicability of AP II, the language of the Protocol requires that the level of territorial control achieved by the armed group concerned be sufficient to allow it to implement the Protocol. The PKK’s efforts to exercise territorial control were clearly far from enabling it to implement AP II (to house prisoners and convene courts to guarantee their due process and fair trial rights; ¹⁶⁰ to take care of the wounded and sick; ¹⁶¹ to provide children with proper education ‘including religious and moral education’; ¹⁶² or to remove children ‘to a safer area within the country and ensure that they are accompanied by persons responsible for their safety and well-being’ ¹⁶³ ). The recent execution of 13 Turkish citizens, whom the PKK kidnapped as hostages under the guise of prisoners of war and dragged from one cave to another, in a cave in northern Iraq in the face of Türkiye’s rescue operation ¹⁶⁴ is an example that demonstrates that the PKK is far from having achieved the territorial control needed to implement the norms of IHL.

In sum, as the discussion above clearly suggests, the PKK does not exercise the territorial control needed ‘to carry out sustained and concerted military operations and to implement’ AP II. ¹⁶⁵ In fact, the PKK has never been able to exercise territorial control over a part of Türkiye’s territory even for a short period of time. The territorial control criterion for the applicability of AP II is not, therefore, met by the PKK and the Protocol would not apply to the violence between Türkiye and the PKK even if Türkiye was a party to it.

7. Conclusion

This article first explained that, under IHL of NIACs, a person who engages in hostilities against the state forces would not be accorded the status of combatant. Acknowledging the applicability of IHL to the violence between Türkiye and the PKK would not, therefore, require Türkiye to treat the captured PKK militants as prisoners of law or to grant them immunity from domestic legislation for attacking state forces. The article also clarified that the application of IHL would not, contrary to the concerns of Türkiye, affect the legal status of the PKK.

The article has argued that both the requisite degree of organization criterion and the requisite degree of intensity criterion for the armed conflict characterization seem to be fulfilled in the context of Türkiye-PKK confrontations. First, the PKK satisfies the majority of the indicative factors for the determination of the organization of a non-state armed group, such as the existence of a command structure and the ability to recruit and train personnel. Second, human and material costs, the temporal and geographical scope of the conflict, as well as the nature of the state’s response and the type of weapons used all suggest that the violence between Türkiye and the PKK is sufficiently intense to constitute an armed conflict within the meaning of IHL. The violence in question, therefore, constitutes a NIAC within the meaning of IHL and is subject to CA3.

This article has also demonstrated that the Rome Statute provisions on NIACs (Article 8(2)(c) and Article 8(2)(e)) would be applicable to the situation in Türkiye if it was a state party. It finally submitted that AP II would not be applicable to Türkiye’s fight against the PKK even if Türkiye was a state party, because the PKK has never been able to exercise such control over a part of Türkiye’s territory as to enable it to carry out sustained military operations and to implement the Protocol. The article’s analysis regarding the fulfilment of AP II-standard is also a challenge to the alternative narrative by other scholars that AP II-standard is also satisfied. In this sense, the Türkiye-PKK conflict provides an excellent practical example of how the two different legal standards for NIACs can indeed produce two different conclusions on the nature of a given situation of violence.