DETENTION IN NIAC
CASE OF ISIL DETAINEES IN SDF CUSTODY
Franca Lucretia Eckstein franca.eckstein@bluewin.ch Student No.: 12546925
International and European Law: Public International Law
Supervisor: Prof. Dr. Terry D. Gill
Abstract
This thesis conducts a case study and investigates whether OAGs have legal authority to detain members of another OAG in NIAC. The case study concerns the countless ISIL fighters which are held in SDF custody. The armed conflict that led to the captivity of ISIL fighters is classified as a NIAC between the SDF and the US-led coalition on one side and ISIL on the other. Syria, the territorial state, is not party to the conflict; therefore, AP II does not apply. IHL does not explicitly address the issue of legal authority to detain in NIAC. The prevailing opinion, which favours the existence of such an authority, argues that an inherent power to detain pursuant to CA3 exists. Furthermore, it suggests that legal authority of OAGs to detain individuals can also be established under customary IHL. This thesis, however, argues against the existence of such legal authority and, thus, considers the detention of ISIL fighters as unlawful. In this context, the author introduces a new legal concept distinguishing between the jus ad captivitatem and the jus in captivitate. However, the author recognises that a progressive development regarding detention by state-like OAGs is taking place. Lastly, several options how the issue of the afore-mentioned detainees can be dealt with are proposed, such as extraterritorial detention by states, handing over the detainees to a government, transferring them to their state of nationality, or installing an international or hybrid tribunal.
Table of Contents
Table of Abbreviations ... III
I. Introduction ... 1
II. Applicability of IHL ... 3
1. Applicability ratione materiae ... 3
a) Armed Conflict between the SDF and ISIL ... 3
b) Involvement of the US-led International Coalition ... 5
c) Absence of Syria ... 6
d) Conclusion ... 7
2. Applicability ratione personae ... 7
3. Applicability ratione loci ... 7
4. Applicability ratione temporis ... 8
5. Conclusion ... 8
III. Legality of Detention ... 9
1. Requirement of a Legal Basis ... 9
2. Treaty Law ... 10
a) Application of CA3 ... 10
b) Interpretation of CA3 ... 10
i) Textual Interpretation ... 10
ii) Application of the Law of IAC per analogiam ... 11
iii) Inherent Legal Authority to Detain ... 12
iv) Legal Authority to Detain derived from Legal Authority to Kill ... 16
v) Killing: Only Alternative to Detention ... 17
c) Conclusion ... 17
3. Customary International Law ... 17
a) NIACs among OAGs ... 18
b) Traditional NIACs per analogiam ... 20
4. Progressive Development? ... 22
5. Conclusion ... 24
IV. Consequences of Unlawful Detention ... 25
V. Final Remarks ... 27
Table of Abbreviations
AP I Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts (adopted 8 June 1977, entered into force 7 De-cember 1978) 1125 UNTS 3
AP II Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of Non-Interna-tional Armed Conflicts (adopted 8 June 1977, entered into force 7 December 1987) 1125 UNTS 609
CA3 Common Article 3 GC
CIHL Customary International Humanitarian Law
CJTF-OIR Combined Joint Task Force – Operation Inherent Resolve ECHR Protocol 4 Protocol 4 to the European Convention for the Protection of
Hu-man Rights and Fundamental Freedoms, securing certain Rights and Freedoms other than those already included in the Conven-tion and in the First Protocol thereto (adopted 16 September 1963, entered into force 2 May 1968) ETS No. 46
GC I Geneva Convention for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 31 GC II Geneva Convention for the Amelioration of the Condition of
Wounded, Sick and Shipwrecked Members of Armed Forces at Sea (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 85
GC III Geneva Convention Relative to the Treatment of Prisoners of War (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 135
GC IV Geneva Convention Relative to the Protection of Civilian Per-sons in Times of War (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 287
IAC International Armed Conflict
ICC International Criminal Court
ICCPR International Covenant on Civil and Political Rights (adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171
ICJ International Court of Justice
ICJ-Statute Statute of the International Court of Justice (Annex to the UN Charter)
ICRC International Committee of the Red Cross
ICTY International Criminal Tribunal for the former Yugoslavia
IHL International Humanitarian Law
IHRL International Human Rights Law
ILC-Statute Statute of the International Law Commission 1947 (adopted by the General Assembly in resolution 174 (II) of 21 November 1947, as amended by resolutions 485 (V) of 12 December 1950, 984 (X) of 3 December 1955, 985 (X) of 3 December 1955 and 36/39 of 18 November 1981)
ISIL Islamic States of Iraq and the Levant
NES Autonomous Administration of North and East Syria
Rome Statute Rome Statute of the International Criminal Court (adopted 17 July 1998, entered into force 1 July 2002) 2187 UNTS 90
SDF Syrian Democratic Forces
UN Charter Charter of the United Nations (adopted 26 June 1945, entered into force 24 October 1945) 1 UNTS XVI
UNSC United Nations Security Council
US United States of America
VCLT Vienna Convention on the Law of Treaties (adopted 22 May 1969, entered into force 27 January 1980) 1155 UNTS 331
I. Introduction
The armed conflict in Syria has not come to an end yet. While the beginning of the conflict in 2011 primarily concerned the Syrian government and its allies combatting the insurgent oppo-sition, it shifted and enlarged during the progress of the conflict due to the strengthening of ISIL1 in Syria and internal and international actors trying to combat it. Today, as far as we
know, ISIL does not directly control territory in Syria anymore. The SDF played a crucial role in defeating the ISIL group with significant support from its ally, the international coalition led by the US.2
Although ISIL fighters were supposedly trained to fight to the death, the SDF captured and imprisoned many of them in detention sites located in the north-east of Syria, which is under the control of the NES (also known as Rojava). However, not much is known about these de-tention facilities.3 According to various sources, approximately 10,000 ISIL fighters are
cur-rently imprisoned.4 A considerable number of foreign ISIL fighters are also among the
detain-ees. Their national states do not want to take them back5 in which case they could face trial in
Syria. However, the SDF do not have the required resources to try all ISIL fighters before a court which meets international standards. Nonetheless, they intend to do so, since they have not been repatriated by their home countries to face justice and there has been no progress in setting up an internationally recognised tribunal.6
A major problem the SDF are facing is the lack of necessary infrastructure and the resources to detain the captured ISIL fighters over an extended period of time. Most “prisons” are repur-posed schools or public buildings. These informal internment facilities are overcrowded and
1 ISIL goes under a variety of names, including ISIS, IS or Daesh. For our purposes, ISIL will be used. 2 see Annyssa Bellal, The War Report: Armed Conflicts in 2016 (Geneva Academy of International
Humanitar-ian Law and Human Rights 2017) 33ff. < https://www.geneva-academy.ch/joomlatools-files/docman-files/The%20War%20Report%202016.pdf> accessed 13 June 2020.
3 Miriam Berger, ‘Here’s what we know about the ISIS prisons controlled by the Syrian Kurds’ The Washington
Post (Washington, D.C., 14 October 2019) < https://www.washingtonpost.com/world/2019/10/12/inside-isis-prisons-controlled-by-syrian-kurds/> accessed 13 June 2020.
4 Charlie Savage, ‘The Kurds’ Prisons and Detention Camps for ISIS Members, Explained’ The New York Times
(New York City, 22 October 2019) < https://www.nytimes.com/2019/10/13/us/politics/isis-prisoners-kurds.html> accessed 13 June 2020; UNGA ‘Report of the Independent International Commission of Inquiry on the Syrian Arab Republic’ (28 January 2020) UN Doc A/HRC/43/57, para 64.
5 Charlie Savage, ‘As ISIS Fighters Fill Prisons in Syria, Their Home Nations Look Away’ The New York Times
(New York City 18 July 2018) < https://www.nytimes.com/2018/07/18/world/middleeast/islamic-state-detain-ees-syria-prisons.html?action=click> accessed 13 June 2020.
6 see UNGA, ‘Report of the Independent International Commission of Inquiry on the Syrian Arab Republic’ (1
February 2018) UN Doc A/HRC/37/72, para 68; Dan Sabbagh, ‘Syrian Kurds to put Isis fighters from dozens of countries on trial’ The Guardian (6 February 2020) < https://www.theguardian.com/world/2020/feb/06/syr-ian-kurds-to-put-isis-fighters-from-dozens-of-countries-on-trial> accessed 22 June 2020.
lead to harrowing conditions.7 With the withdrawal of the US from the conflict and the advance
of Turkey into Syria in late 2019, the situation worsened.8 Accordingly, Human Rights Watch
stated, that “any authority that effectively controls these informal prisons is legally bound to urgently improve conditions and ensure that each and every detainee is held lawfully”.9
However, the detention of ISIL fighters by the SDF is in a legal grey area.10 Under international
law, it is uncertain whether the SDF have a legal authority to detain ISIL fighters. Therefore, the armed conflict which led to the captivity of the ISIL fighters will be classified in order to determine the applicable legal framework (II.). Subsequently, this thesis examines the legality of detaining captured fighters of one OAG by another OAG in a NIAC: in this case, the deten-tion of ISIL fighters by the SDF, assuming the conflict between the SDF and ISIL is a NIAC (III.). Lastly, the legal consequences of a potentially unlawful detention are researched: What can be done with the captives (IV.)? The thesis concludes with final remarks (V.).
The first part of the thesis is descriptive in nature, as it describes the (legal) situation of the conflict in question and the detention of ISIL fighters. The main research revolves around a legal analysis examining whether detention by an OAG in a NIAC has a legal basis or not. This culminates in the proposition of new legal concepts through a normative approach. The aim is to find a way under international law to keep ISIL fighters in lawful detention and give them a proper trial.
7 Berger (n 3). 8 Savage (n 4).
9 Human Rights Watch, ‘Northeast Syria: Boys, Men Held in Inhumane Conditions’ (8 October 2019)
<https://www.hrw.org/news/2019/10/08/northeast-syria-boys-men-held-inhumane-conditions> accessed 13 June 2020.
10 see Alessandra Spadaro, ‘ISIS members detained by Kurdish forces in Syria: Operational and legal challenges’
Armed Groups and International Law (13 March 2019) < https://armedgroups-international- law.org/2019/03/13/guest-post-by-alessandra-spadaro-isis-members-detained-by-kurdish-forces-in-syria-op-erational-and-legal-challenges/> accessed 13 June 2020.
II. Applicability of IHL
To examine the legality of detention11 of ISIL fighters in detention facilities run by the SDF,
the applicable legal framework needs to be set out. Regarding classification of armed conflicts, the law differentiates between legal rules applicable in situations of IAC and of NIAC. Accord-ing to the GCs, an IAC is a conflict “between two or more of the High ContractAccord-ing Parties, even if the state of war is not recognised by one of them”.12 A NIAC, however, is defined in CA3 as
an armed conflict, taking place on the territory of a party to the GCs, in which one or more OAGs are participating. The latter is subject to a more limited and also more contentious legal regime; especially within the ambit of detention.13
1. Applicability ratione materiae
Classifying the conflict(s) in Syria is difficult due to the large number of different actors in-volved.14 Scholars widely agree that there is not just one armed conflict, but that there are
sev-eral parallel armed conflicts taking place.15 However, the number of armed conflicts in Syria,
past or underway, is debated.16 For the case in question, the hostilities between the SDF and
ISIL require classification.
a) Armed Conflict between the SDF and ISIL
As both, the SDF and ISIL, are non-state actors, their conflict can only be qualified as a NIAC. Its beginnings were in 2013, when Kurdish forces captured the city of Ras-al-Ayn. The conflict reached a peak in 2016 with the offensive on Raqqa and continues today with the SDF having significantly expanded their territory in north-eastern Syria. ISIL has gradually but steadily lost the territory which it formerly had under its control, mostly to the SDF.17 Hence, this conflict
11 The terms “detention”, “internment” and “deprivation of liberty” are used interchangeably. 12 Common Art. 2 GC.
13 In IAC, captured fighters are qualified as “prisoners of war” (cf Art. 4 GC III). This status does not exist in
NIAC.
14 see Tilman Rodenhäuser, Organizing Rebellion. Non-State Armed Groups under International Humanitarian
Law, Human Rights Law, and International Criminal Law (OUP 2018) 7.
15 cf David Wallace, ‘Tying to Make Sense of the Senseless: Classifying the Syrian War Under the Law of Armed
Conflict’ (2017) 25 Michigan State International Law Review 555, 590. The existence of several parallel armed conflict is recognised, see Case Concerning Military and Paramilitary Activities in and Against Nicaragua
(Nicaragua v. United States of America) (Merits) [1986] ICJ Rep 14, para 219 (hereinafter: Nicaragua case).
16 see Terry D Gill, ‘Classifying the Conflict in Syria’ (2016) 92 International Law Studies (Naval War College)
353.
17 Annyssa Bellal, The War Report: Armed Conflicts in 2018 (Geneva Academy of International Humanitarian
Law and Human Rights 2019) 128ff. < https://www.geneva-academy.ch/joomlatools-files/docman-files/The%20War%20Report%202018.pdf> accessed 13 June 2020.
is not just a short-lived insurrection and goes beyond a mere terrorist activity.18 Furthermore,
the “intensity” criterion required to establish its definition as a NIAC is fulfilled:19 since 2014,
countless confrontations between the SDF and ISIL resulted in thousands of fighters and some civilians of both parties losing their lives or getting gravely injured. About 100,000 civilians fled the combat zone, most of them escaped to Turkey. Although the UNSC addressed the sit-uation in Syria on various occasions,20 its active involvement keeps getting put off not because
the conflict lacked intensity, but because of opposing interests and positions among the P-5.21
Likewise, both the SDF and ISIL fulfil the “organisation” criterion.22 According to art. 15 of
the constitution of the NES, the People’s Protection Units (YPG) are its official defence force.23
They constitute the SDF together with smaller and less important Arab militias. The SDF were founded in late 2015, although the YPG has existed considerably longer, and have tens of thou-sands of fighters at their disposal,24 with a vast territory under their control and administration.
Their access to weapons and other military equipment as well as their military training is mainly provided by the US, its ally.25 ISIL, on the other hand, considered a terrorist organisation, was
created in 2006 and strengthened its power over Iraq and northern Syria with the help of more than 10,000 fighters.26 Raqqa became its de facto capital.27 ISIL funded itself through various
sources, inter alia by selling oil.28 Through its quite developed military structure, it was able to
effectively control and administer the occupied territory.29 In conclusion, the threshold of a
NIAC is met.
18 cf Prosecutor v Tadić (Trial Judgment) ICTY-94-1-T (7 May 1997) para 562.
19 see the indicative factors set out in Prosecutor v Haradinaj (Trial Judgment) ICTY-04-84-T (3 April 2008)
para 49.
20 see, eg, UNSC Res 2139 (22 February 2014) UN Doc S/RES/2139.
21 cf, eg, UNGA Res 66/253 (7 August 2012) UN Doc A/RES/66/253 B, that expressed its deep concern about
“the failure of the Security Council to agree on measures to ensure the compliance of Syrian authorities with its decisions” and “encourages the Security Council to consider appropriate measures in this regard”.
22 see the indicative factors set out in Prosecutor v Haradinaj (n 19) para 60.
23 Social Contract of the Democratic Federation of Northern Syria < https://internationalistcommune.com/social-contract/> accessed 13 June 2020.
24 On 12 February 2016, the Economist considered the SDF as “essentially a subsidiary of the Kurdish YPG”
<https://country.eiu.com/article.aspx?articleid=1363937520&Country=Syria&topic=Politics> accessed 13 June 2020. Tom Gal, ‘Legal Classification of the Conflict(s) in Syria’ in Hilly Moodrick-Even Khen et al. (eds), The Syrian War. Between Justice and Political Reality (CUP 2020) 29, 51 sticks in his entire article to “YPG” than to “SDF”. The ICRC argues in International Humanitarian Law and the Challenges of
Contem-porary Armed Conflicts. Recommitting to Protection in Armed Conflict on the 70th Anniversary of the Geneva
Conventions (October 2019) 51 correctly that when several armed groups are cooperating against another OAG
or against a state, “it might be more realistic to examine the intensity criterion collectively by considering the sum of the military actions carried out by all of them fighting together”.
25 Bellal (n 17) 129f. 26 Bellal (n 2) 37. 27 Bellal (n 17) 128. 28 Ibid.
29 Alex Bilger, ‘ISIS Annual Reports Reveals Metrics-Driven Military Command’ Institute for the Study of War
(22 May 2014) < http://www.understandingwar.org/backgrounder/ISIS-Annual-Reports-Reveal-Military-Or-ganization> accessed 13 June 2020.
b) Involvement of the US-led International Coalition
The detention facilities in which ISIL fighters are held captive are run by the SDF and are located within their controlled territory.30 Therefore, the involvement of the US-led
interna-tional coalition (hereinafter coalition) is only of importance for the conflict’s classification if the conflict between the SDF and ISIL as well as the involvement of the coalition are considered one single armed conflict. If the coalition’s attacks on ISIL are viewed as an armed conflict occurring in parallel to the SDF and ISIL’s, its classification is irrelevant because then the co-alition would not be a party to the conflict which resulted in the detention of ISIL fighters. Considering the amount of military equipment and training the international coalition provided the SDF in its conflict against ISIL and the recognition as allies, they are in fact combatting ISIL together.31
The US founded the coalition under its leadership in late 2014 to combat the territorial advance of ISIL. Various states are partaking in the task force.32 Since 2014, thousands of airstrikes
targeting ISIL have been conducted.33 In contrast to all other member states in the coalition, the
US, the UK and France have additionally deployed ground forces assisting and supporting the SDF, their allies.34 Scholars disagree on how to classify a conflict predominantly possessing a
non-international character involving international elements. It is unclear whether the involve-ment of the US-led international coalition transformed the NIAC between the SDF and ISIL into an IAC or, at least, provoked the applicability of the law of IAC, since they intervened in a NIAC taking place in another state’s territory.
The coalition or even just the US alone exercising such control over the SDF could amount to “internationalising” this NIAC. The ICJ and the ICTY set up different thresholds for third states 30 see, eg, Berger (n 3).
31 Many authors classify armed conflicts much broader. Cf the ICRC’s support based-approach in Tristan Ferraro,
‘The ICRC’s Legal Position on the Notion of Armed Conflict Involving Foreign Intervention and on Deter-mining the IHL Applicable to this Type of Conflict’ (2015) 97 International Review of the Red Cross 1227.
32 cf the official webpage of the Combined Joint Task Force – Operation Inherent Resolve (CJTF–OIR):
<https://www.inherentresolve.mil/About-CJTF-OIR/> accessed 13 June 2020.
33 Helene Cooper and Eric Schmitt, ‘Airstrikes by U.S. and Allies Hits ISIS Targets in Syria’ The New York
Times (Washington, 22 September 2014) < https://www.nytimes.com/2014/09/23/world/middleeast/us-and-al-lies-hit-isis-targets-in-syria.html> accessed 13 June 2020.
34 In relation to the U.S. ground forces see Eric Schmitt, ‘U.S. Resumes Large-Scale Operations Against ISIS in
Northern Syria’ The New York Times (Manama, 25 November 2019) < https://www.ny-times.com/2019/11/25/us/politics/us-syria-isis.html> accessed 13 June 2020; in relation to the British and French ground forces see Lara Seligman, ‘Britain, France Agree to Send Additional Troops to Syria’ Foreign Policy (Washington, 9 July 2019) < https://foreignpolicy.com/2019/07/09/britain-france-agree-to-send-addi-tional-troops-increase-syria-us-withdrawal-uk/> accessed 13 June 2020.
exercising control over an OAG.35 However, only the Appeals Chamber of the ICTY
contextu-alised the threshold in the case of the classification of an armed conflict.36 Thus, the overall
control test introduced by the ICTY is the only relevant one in this case.37 The test expects third
states to exercise overall control over the OAG, which goes “beyond the mere financing and equipping of such forces and involving also participation in the planning and supervision of military operations”.38 The SDF is mostly acting independently from the US and the coalition.39
Thus, their actions do not suffice to meet the threshold of overall control. The only known case in which it was almost met was during the Raqqa offensive, when US Special Operation Forces did not only train the SDF but also gave them advice.40 Based on what is known, neither the
US nor the coalition exercised overall control over the SDF, as they only gave advice, did not participate more deeply in the planning of the operation, and did not repeat the incident.41
Therefore, this NIAC was not “internationalised”.
c) Absence of Syria
The conflict set out in the previous sub-chapters did not involve Syria, the territorial state. Syria did not directly participate in this specific conflict at any time. At the time of the intervention, the Syrian government had lost its northern territory to ISIL and the SDF. Over the course of the conflict, the SDF wrested control over most of the territory in northern and northeast Syria, which had formerly been under ISIL control and came to dispose of considerable territorial autonomy bordering on quasi-independence. The SDF, ISIL as well as the coalition have non-interconnected hostile relations with Syria. This dissertation’s author opines that they all con-stitute parallel armed conflicts. Consequently, Syria is not considered a party to the conflict in
35 cf Nicaragua case (n 15) para 105–115; Application of the Convention on the Prevention and Punishment of
the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro) (Judgment) [2007] ICJ Rep 43,
para 398–407 (hereinafter: Bosnian Genocide case) and Prosecutor v Tadić (Appeal Judgment) ICTY-94-1-A (15 July 1999) para 115ff. (hereinafter: Tadić case).
36 Tadić case (n 35) para 162.
37 cf the convincing argumentation by Marko Milanovic, ‘The Applicability of the Conventions to ‘Transnational’
and ‘Mixed’ Conflicts’’ in Andrew Clapham et al. (eds), The 1949 Geneva Conventions: A Commentary (OUP 2016) para 31–34, who argues that overall control suffices for the internationalisation of a conflict.
38 Tadić case, para 145. This threshold is considerably high. However, Yoram Dinstein, ‘Concluding Remarks
on Non-International Armed Conflicts’ (2012) 88 International Law Studies (Naval War College) 399, 406f., 411 goes too far arguing that the OAG has to be a de facto organ of the intervening state.
39 cf Wallace (n 15) 592.
40 Alissa J. Rubin, Karam Shoumali and Eric Schmitt, ‘American Is Killed in First Casualty for U.S. Forces in
Syria Combat’ The New York Times (New York City, 24 November 2016) < https://www.ny-times.com/2016/11/24/world/middleeast/syria-warplanes-turkey.html?&moduleDetail=section-news-3> ac-cessed 13 June 2020.
41 Gal (n 24) 53 and Wallace (n 15) 592f. came to the same conclusion. However, the former applied the strict
control test introduced by the ICJ in the Nicaragua case, whereas the latter applied the overall control test introduced by the Appeals Chamber in the Tadić case (see n 35).
question. Therefore, the requirements of a NIAC under AP II are not met, because the territorial state is not a party to the conflict.42 Nevertheless, Syria plays a role in the respective NIAC, as
the area occupied would lawfully be its territory. Syria is not an ally of either party to the con-flict and not only disapproves of the territorial claims made by the SDF (or, in fact, by the NES) and ISIL but also of the coalition’s involvement.
d) Conclusion
The conflict between ISIL, the SDF and the international coalition classifies as a NIAC accord-ing to CA3. Therefore, CA3 and customary IHL are applicable, but not AP II.
2. Applicability ratione personae
In principle, IHL binds parties to an armed conflict.43 In the case of a NIAC, the general
con-sensus is that both conventional and customary IHL is binding for both state and OAGs.44
Con-sequently, IHL is equally applicable ratione personae to ISIL, the SDF and the coalition.
3. Applicability ratione loci
In the Tadić case, the ICTY ruled that “international humanitarian law continues to apply […], in the case of internal conflicts, [in] the whole territory under the control of a party, whether or not actual combat takes place there”.45 Thus, IHL will bind an OAG to the territory in which it
is active or exercising effective control. So, IHL is applicable ratione loci not only where ISIL fighters were captured but also where they are currently detained. In accordance with the Tadić practice, the scope of IHL is also met in the case of some detention facilities situated outside of Syria that are still within the boundaries of the NES.
42 cf Art. 1(1) AP II.
43 Jean-Marie Henckaerts and Louise Doswald-Beck, Customary International Humanitarian Law. Volume I:
Rules (CUP 2005) Rule 139 (hereinafter: CIHL).
44 Knut Dörmann et al. (eds), Commentary on the First Geneva Convention: Convention (I) for the Amelioration
of the Condition of the Wounded and Sick in Armed Forces in the Field (2nd edn, ICRC 2016) para 393f.
(hereinafter: ICRC, Commentary GC I).
45 Prosecutor v Tadić (Decision on the Defence Motion for Interlocutory Appeal on Jurisdiction) (2 October
4. Applicability ratione temporis
Recently, no general hostilities between the SDF and ISIL have been recorded. However, IHL is still applicable ratione temporis.46 It extends for fighters imprisoned for reasons related to
the respective NIAC beyond the end of hostilities and lasts until their release.47
5. Conclusion
In conclusion, the law of NIAC is generally applicable. The armed conflict between the SDF, the US-led international coalition, and ISIL classifies as a NIAC, which is not “international-ised”. Syria is not a party to the conflict. Therefore, only CA3 and customary IHL apply, since AP II requires the territorial state to be a party to the conflict. The SDF, the coalition, and ISIL are bound by IHL ratione personae. Furthermore, IHL applies to the detention facilities where ISIL fighters are held captive. Lastly, IHL is still applicable to detainees, even though active hostilities have ceased (at the moment).
46 Different views regarding the cease of the applicability of IHL exist. Cf Ibid, para 70 and ICTY in the case
Prosecutor v. Ramush Haradinaj, Idriz Balaj, Lahi Brahimaj (Trial Judgment) ICTY-04-84bis-T (29
No-vember 2012) para 396.
III. Legality of Detention
Detaining prisoners of war under the legal regime of IAC is well established.48 However, the
legality of detaining captured fighters in NIAC is controversial, even for states. There is no explicit legal basis49 under IHL, and IHRL is not regarded as providing a legal basis for OAGs.50
States usually rely on domestic law as a legal basis for detention in NIACs.51 OAGs
participat-ing in a NIAC, on the other hand, cannot rely on domestic law for their actions; OAGs are not addressed by domestic law.52 In the case of one OAG detaining members of another OAG, as
seen by the imprisonment of ISIL fighters by the SDF, the legal situation is even less clear.53
Scholars widely discuss the legality of detention in a NIAC between the territorial state and an opposing OAG, but no attention is given to the same question between two or more OAGs. Therefore, we need to answer whether OAGs have legal authority to detain members of an opposing OAG in a NIAC.
1. Requirement of a Legal Basis
Detention unprescribed by law is arbitrary,54 but international law provides for the absolute
prohibition of arbitrary detention. According to one view, the principle of non-arbitrariness of detention in armed conflicts is found in customary IHL;55 according to another, it is based on
IHRL56. Whatever legal regime may base this case’s prohibition on arbitrary detention,57 we
48 see Art. 4 GC III and GC III in general; Art. 43 and 44(1) AP I. However, if IHL provides a legal authority to
detain in IACs or if the authority must be found elsewhere is still open to some debate, see Ryan Goodman, ‘Authorization versus Regulation of Detention in Non-International Armed Conflicts’ (2015) 91 International War Studies (Naval War College) 155, 159.
49 The terms “legal basis”, “authorisation” and “power” to detain are used interchangeably. See ICRC, Internment
in Armed Conflict: Basic Rules and Challenges, Opinion Paper (November 2014) 7.
50 For IHL cf CA3 and AP II; for IHRL see Joshua J Niyo, ‘Non-State Armed Groups and the Power to Detain
in Non-International Armed Conflict’ (2020) 53 Israel Law Review 3, 14 who convincingly argues that the basis as such has to be provided in IHL (like for detention in IAC), but the legal regime of IHRL sets up a frame in which this detention has to occur.
51 Serdar Mohammed v Ministry of Defence [2014] EWHC 1369 (QB) para 417ff.; Lawrence Hill-Cawthorne,
Detention in Non-International Armed Conflict (OUP 2016) 70.
52 cf, eg, Ezequiel Heffes, ‘Detention by Armed Opposition Groups in Non-International Armed Conflicts:
To-wards a New Characterization of International Humanitarian Law’ (2015) 20 Journal of Conflict & Security Law 229, 230.
53 see, eg, UNGA, ‘Report of the Independent International Commission of Inquiry on the Syrian Arab Republic’
(1 February 2018) UN Doc A/HRC/37/72, para 69, that left the lawfulness of detention unanswered.
54 cf Art. 9(1) ICCPR; Human Rights Committee, ‘General Comment No 35. Article 9 (Liberty and Security of
Person)’ (7–31 October 2014) UN Doc CCPR/C/GC/35, para 11.
55 CIHL (n 43) Rule 99; ICRC, Commentary GC I (n 44) para 719; Goodman (n 48) 156.
56 Human Rights Committee, ‘General Comment No 35. Article 9 (Liberty and Security of Person)’ (7–31
Octo-ber 2014) UN Doc CCPR/C/GC/35, para 10ff.; Kevin J Heller, ‘What Exactly Is the ICRC’s Position on De-tention in NIAC?’ Opinio Juris (6 February 2015) < http://opiniojuris.org/2015/02/06/exactly-icrcs-position-detention-niac/> accessed 14 June 2020.
57 Hill-Cawthorne (n 51) 67, however, argues rightly that the legal basis must be found in IHL, because IHL also
provides the legal basis for detention in IAC. Moreover, with this approach, discussion about applicability of IHRL to OAGs can be avoided.
need a legal basis allowing OAGs to detain members of the enemy armed forces in a NIAC, so that such detention is based on legal ground. Otherwise, detention is not lawful. A UNSC Res-olution or a special agreement between the parties to the NIAC could provide legal authority to detain members of opposing parties for specific cases.58 However, neither was set up for the
detention of ISIL fighters by the SDF. Therefore, international law needs to be looked at, treaty and customary international law,59 in order to conclude a legal basis that provides the expected
authority to detain in general.
2. Treaty Law
a) Application of CA3
In a NIAC between OAGs without the involvement of the territorial state, AP II is not applica-ble.60 Therefore, only CA3 is applicable besides customary IHL. It protects individuals who are
not actively partaking in hostilities (anymore). CA3 states in paragraph 1 relating to detention: “Persons (…) placed hors de combat by (…) detention shall in all circumstances be treated humanely (…).” Furthermore, it prohibits cruel treatment and torture,61 humiliating and
degrad-ing treatment62 and “the passing of sentences and the carrying out of executions without
previ-ous judgment pronounced by a regularly constituted court, affording all the judicial guarantees which are recognized as indispensable by civilized peoples”63.
b) Interpretation of CA3
i) Textual Interpretation
The wording of CA3 only recognises that detention may occur and regulates its conditions, but it does not expressly authorise the parties to the conflict to detain; thus, it does not provide an explicit legal basis64 and, therefore, OAGs cannot imprison anyone in accordance with the
“or-dinary meaning (…) of the treaty”65. However, it is argued that treaty law of IHL rarely
explic-itly provides legal authority, but that it is more about prohibiting certain acts. Therefore,
58 Eg, Hill-Cawthorne (n 51) 66 in relation to UNSC resolutions and Heffes (n 52) 233 in relation to special
agreements.
59 According to the view represented here, it must be clearly distinguished between a possible legal basis in treaty
and/or customary IHL.
60 see above II.1.c). 61 CA3(1)(a). 62 CA3(1)(c). 63 CA3(1)(d).
64 Hill-Cawthorne (n 51) 69. 65 Art. 31(1) VCLT.
regulating conditions of detention could easily be construed as an indirect recognition of legal authority to detain, especially since deprivation of liberty is not prohibited by CA3.66 But it
should be noted that the law of IAC explicitly provides legal authority to detain.67 Thus,
ac-cording to the view advanced here, an indirect recognition should not be read into the ordinary meaning of the treaty provision itself but rather should be construed, if appropriate, through a systematic or teleological interpretation.68
ii) Application of the Law of IAC per analogiam
Considering a more contextual and systematic approach, legal authority to detain provided in the law of IAC may be applied per analogiam, establishing a type of “quasi-POW” deten-tion69.70 States are reluctant to grant legal authority to detain to OAGs vis à vis members of
their own armed forces. However, the situation is less clear vis à vis members of other OAGs. Considering that OAGs are actually equal (as states in an IAC) and not just regarded as equal parties, like in the case of an armed conflict between a territorial state and an OAG, applying legal authority to detain granted to states under the law of IAC may be suitable by analogy.
In the last two decades, jurisprudence and the development of customary IHL brought the law of IAC and the law of NIAC closer.71 Some scholars even argue, that the differences between
these two legal branches of IHL have been diminished so significantly, that such an analogy is feasible.72 Moreover, in relation to internment, Sassòli and Olsen argue that the same provisions
can be applied for both IACs and NIACs even if there are differences between the two regimes, “as long as the rules are applied according to the person’s function rather than status”.73
The law of IAC contains the category of “protected persons”. They encompass, inter alia, pris-oners of war and civilians. Art. 21(1) GC III provides parties to an IAC with legal authority to detain prisoners of war; Art. 42 GC IV provides the same for civilians under certain conditions. The law of NIAC does not have this categorisation. However, based on Sassòli’s and Olsen’s
66 ICRC, Internment in Armed Conflict: Basic Rules and Challenges, Opinion Paper (November 2014) 7. 67 Art. 21(1) GC III; see the following sub-chapter, II.2.b)ii.
68 see below II.2.b)ii) and II.2.b)iii).
69 Deborah Casalin, ‘Taking Prisoners: Reviewing the International Humanitarian Law Gorunds for Deprivation
of Liberty by Armed Opposition Groups, (2011) 93 International Review of the Red Cross 743, 754.
70 Called by Heffes (n 52) 231 as “traditional approach”. 71 cf CIHL (n 43); Tadić case (n 35) para 96–127.
72 see Marco Sassòli and Laura M Olsen, ‘The Relationship between International Humanitarian and Human
Rights Law where it matters: Admissible Killing and Internment of Fighters in Non-International Armed Con-flicts’ (2008) 90 International Review of the Red Cross 599, 601f.
opinion, the individual’s status does not matter; their function is decisive. Id est, this categori-sation in IACs is less relevant to decide whether the legal basis set out in Art. 21(1) GC III is applicable to detentions in NIACs by analogy.
Following this argument, the law of IAC cannot be applied to detention in NIAC per analogiam. Although the law of IAC and the law of NIAC has converged on certain points, certain differ-ences, legal and not just factual, still remain and need to be considered. One of these widely argued differences is detention:74 as mentioned before, there is no evidence that states want
OAGs to have legal authority to detain; neither members of their own armed forces nor mem-bers of other OAGs. Furthermore, a general application of the law of IAC by analogy would make the distinction between IAC and NIAC redundant. Secondly, also beyond international law, legal authority to detain individuals is inherent to states and not to non-state actors. It is nearly impossible to argue why a non-state actor would have legal authority to detain another non-state actor. Therefore, OAGs detaining individuals should not be equated with states doing the same. Consequently, the law of IAC in relation to detention should not be applied by anal-ogy. Lastly, Sassòli and Olsen admit that in this case, the application of the law of IAC per
analogiam would be welcomed in practice but would not make it legally binding.75
iii) Inherent Legal Authority to Detain
It is uncertain, whether a teleological approach of CA3 can provide inherent legal authority to detain under CA3 and, if confirmed, whether such an authority is sufficient; id est, whether there is no need of an explicit legal basis,76 because there is no prohibition on restricting the
liberty of individuals under CA3.
Scholars widely discuss whether there is an implicit legal basis to detain under the law of NIAC.77 Analysing the present situation, the predominant view favours its existence and, thus,
74 Hill-Cawthrone (n 51) 66; Lindsay Moir, ‘The Concept of Non-International Armed Conflict’ in Andrew
Clap-ham et al. (eds), The 1949 Geneva Conventions: A Commentary (OUP 2016) 414. Ryan Goodman, ‘The De-tention of Civilians in Armed Conflict’ (2009) 103 AJIL 48, 50, however, argues that “whatever is permitted in international armed conflict is permitted in noninternational armed conflict” because–according to him–IHL is less restrictive in a NIAC.
75 Sassòli/Olsen (n 72) 626.
76 Many authors do not make a clear distinction between treaty and customary IHL, especially when discussing
the question of an inherent legal authority to detain in NIAC. Nevertheless, it is evident that the legal basis would–in case it exists–fall under CA3 and, thus, be treaty law.
affirms legal authority of OAGs to detain individuals;78 although hesitantly.79 According to this
view, the object and purpose of regulating detention requires the respective legal authority to detain. Detention is regularly happening in NIACs. Therefore, legal authority is inherent to CA3. However, scholars expressing this opinion only refer to NIACs between an OAG and a state.80 They do not discuss the legal basis to detain in NIACs between two or more OAGs
without the involvement of the territorial state.
The base of these arguments are real circumstances: several OAGs detain individuals.81
More-over, they argue that, since territorial states can base detention on domestic law, the members of the OAGs would be discouraged to respect other norms of IHL because they have less rights under the law of NIAC compared to territorial states if they have no legal authority to detain individuals.82 In any case, we shall leave the motivations aside and turn to the rationales:
Among various rationales in favour of inherent legal authority to detain, two are considered here as the strongest, since they are widely quoted by scholars representing the predominant opinion.
Firstly, the principle of equality of belligerents requires the application of the same rights and obligations to all parties to the NIAC. Although the parties themselves may not be equal in all respects (a state is a state and an OAG is a non-state actor), their rights and obligations are.83
Therefore, many scholars argue that OAGs have the power to detain individuals in NIAC, as well. Otherwise there would be an asymmetry, which would “threaten the idea that the laws of war provide for a ‘level playing field’ with mutual rights and obligations”.84 But, on what
ground do they base the authority of states to detain individuals?85 Some argue that states have
78 See, eg, ICRC, Commentary GC I (n 44) para 720; Sean Aughey and Aurel Sari, ‘Targeting and Detention in
Non-International Armed Conflict: Serdar Mohammed and the Limits of Human Rights Convergence’ (2015) 91 International Law Studies (Naval War College) 60, 103ff.; Daragh Murray, ‘Non-State Armed Groups, Detention Authority in Non-International Armed Conflict, and the Coherence of International Law: Searching for a Way Forward’ (2017) 30 Leiden Journal of International Law 435, 436ff. Cf in favour of the opposite view, eg, Lawrence Hill-Cawthorne and Dapo Akande, ‘Does IHL Provide a Legal Basis for Detention in Non- International Armed Conflicts?’ EJIL-Talk! (7 May 2014) < https://www.ejiltalk.org/does-ihl-provide-a-legal-basis-for-detention-in-non-international-armed-conflicts/> accessed 14 June 2020; Heller (n 56).
79 cf Niyo (n 50) 16 referring to the position expressed by the ICRC. 80 Which fall in the scope of application of AP II.
81 cf, eg, ICRC, Commentary GC I (n 44) para 717ff. or Murray (n 78) 436 giving examples of OAGs detaining
individuals.
82 see Casalin (n 69) 750; Heffes (n 52) 231. 83 Heffes (n 52) 238f.
84 Andrew Clapham, ‘Detention by Armed Groups under International Law’ (2017) 93 International Law Studies
(Naval War College) 1, 2.
a right to detain in NIACs under IHL, others argue that IHL does not provide such a basis.86
States themselves, however, usually base such detentions on domestic law and, consequently, do not accept a legal basis of detention under IHL. Detention by OAGs will always be prohib-ited under the domestic law of the state involved in the conflict.87 Therefore, at first glance, a
legal basis under domestic law would not be compatible with the principle of equality of bel-ligerents and would require a legal basis for both under IHL. However, this view misses the point that the principle of equality does not exist under domestic law. Consequently, according to the view represented here, in a NIAC between a state and an OAG, an inherent legal power of OAGs to detain based on the application of the principle of equality cannot be established. Following this argument, it is hard to proclaim that OAGs fighting each other in a NIAC have legal authority to detain members of another OAG if there is no legal authority for OAGs to detain members of the governmental armed forces, particularly because the principle of equality applies among them, as well. Especially in the case of various interrelated NIACs between a territorial state and several OAGs, it would not be compatible with the principle of equality if an OAG has inherent legal authority to detain only members of some adversary parties, namely of other OAGs, but not of the governmental armed forces.
The second rationale, which is widely argued to establish inherent legal authority to detain individuals in NIAC, is Art. 4–6 AP II. Art. 5(1) AP II refers explicitly to “persons deprived of their liberty for reasons related to the armed conflict”. Taking both criminal and non-criminal, security-related detention, into account,88 it is argued that at least the latter must lead to the
conclusion that OAGs can detain members of the enemy armed forces.89 Therefore, an inherent
legal basis is regarded as existent. Unlike CA3, AP II states the material detention environment and contains more advanced protections for individuals, especially in relation to armed con-flicts.90 Consequently, it goes much further than CA3.91 This is often held up as a strong
argu-ment, that there must be legal authority under IHL to detain. However, the ICRC Commentary on AP II reads that the purpose of Art. 5 AP II is “to ensure that conditions of detention for persons whose liberty has been restricted for reasons related to the conflict will be reasonable”
86 see the discussion in Murray (n 78) 446ff. 87 see, eg, Casalin (n 69) 744.
88 Yves Sandoz, Christophe Swinarski and Bruno Zimmerman (eds), Commentary on the Additional Protocols
of 8 June 1977 to the Geneva Conventions of 12 August 1949 (ICRC 1987), 1386 (hereinafter: ICRC, Com-mentary AP).
89 Heffes (n 52) 240. 90 see Art. 5 AP II.
91 Ramin Mahnad, ‘Beyond Process: The Material Framework for Detention and the Particularities of
and is silent on matters of legal authority.92 It has to be taken into consideration that AP II has
not been ratified universally. Thus, deducting a general authorisation to detain in NIAC with the reasons set out above, seems too far-fetched. In addition, AP II does not apply in NIACs between two or more OAGs.93 So, AP II is of no importance for the question whether there is
inherent legal authority for OAGs participating in a NIAC where the territorial state is not in-volved. To sum up, both arguments of the predominant sides are according to the view advanced here strong yet not sufficiently persuasive in relation to a NIAC among OAGs.
Lastly, according to IHRL, inherent legal authority clearly does not suffice the requirement of a legal basis.94 Although criticised in scholarship to be a strict formalist approach to the law,95
the author opines that one cannot interpret the missing authorisation or regulation into a provi-sion, regardless of the desirability of such authorisation or regulation.
According to the view advanced here, the object and purpose of regulating detention in CA3 is to create substantive treatment standards to which parties to a NIAC must adhere, taking into consideration that detention occurs.96 A legal basis simply does not exist.97 So, there is a
dif-ference between regulating detention and authorising it.98 Therefore, the author suggests to
in-troduce a new approach, distinguishing between a jus ad captivitatem and a jus in captivitate in armed conflict. The former decides whether an entity has the authority to detain members of the adversary armed force, whereas the latter provides the conditions of detention and the sub-stantive treatment standards the detaining party must comply with99. Although both are included
under the umbrella term “detention in armed conflict”, they regard a completely different set of rules and, therefore, must be strictly separated. Consequently, a detention which is lawful under the jus in captivitate, does not have to be lawful under the jus ad captivitatem.
92 ICRC, Commentary AP (n 88) 4565. 93 Art. 1(1) AP II e contrario.
94 Or IHL, depending on which legal regime the principle of non-arbitrariness of detention in armed conflicts is
based, see above II.2.
95 Niyo (n 50) 12.
96 see Hill-Cawthorne (n 51) 69.
97 Niyo (n 50) 15 disagrees and–unconvincingly–calls for an implicit authority “to make practical sense”.
How-ever, he ignores that although something might be “real” and “practical”, it does not have to be lawful.
98 However, according to Niyo (n 50) 13 the line between regulation versus authorisation is rather blurred. This
view is not shared by the author.
To sum up, scholars have not settled their disagreement around inherent legal authority for OAGs to detain individuals, especially in the case of a NIAC among OAGs. The author sees such an authority to not be inherent in CA3, which only consists of jus in captivitate.
iv) Legal Authority to Detain derived from Legal Authority to Kill A conclusive argument against the opinion that IHL does not provide legal authority to detain in NIAC is that legal authority to kill in a NIAC could, at least implicitly, include legal authority to detain.100 It is argued that because parties to a NIAC can kill enemy fighters in hostilities, it
must be assumed that these fighters could also be captured.101 Id est, the lesser power of
deten-tion must be included in the greater power of killing.102 Then, detention will be preferable over
killing in an armed conflict.103 However, various scholars have brought forward different
argu-ments that question the persuasiveness of this option. Clapham states that, firstly, states do not accept OAGs’ right to kill in NIACs and, secondly, detention by OAGs can include groups of persons that go beyond those who can be targeted, or even killed, without violating IHL.104
Especially the second argument is widely upheld: the lesser power of detention goes ratione
personae further than the greater power of killing. It also extends to civilians who indirectly
participate in hostilities, for instance by posing a security threat.105 Those individuals are not
lawful targets and cannot be killed in a NIAC.106 Therefore, at least in relation to these civilians,
legal authority to detain cannot be derived from legal authority to kill in NIAC. Another legal basis is required. Additionally, Goodman rightly points out that if the authority to kill included the authority to detain, belligerents would be encouraged “to excessively expand the defini-tional scope of lawful targets to justify detention policies”.107 Based on this evidence, the author
is not convinced by the initial argument that legal authority to detain is included in legal au-thority to kill.
100 see Hill-Cawthorne/Akande (n 78).
101 Chatham House and International Committee of the Red Cross, ‘Expert meeting on procedural safeguards for
security detention in non-international armed conflict’ (London, 22–23 September 2008), (2009) 91 Interna-tional Review of the Red Cross 859, 863.
102 Argumentum a maiore ad minus, see Aughey/Sari (n 78) 104. 103 Heffes (n 52) 248.
104 Clapham (n 84) 3.
105 cf Art. 5, 27, 41–43 and 78 GC IV.
106 Ryan Goodman, ‘The Limits of the Logic that the Power to Kill includes the Power to Detain’ Just Security
(16 May 2014) <https://www.justsecurity.org/10485/power-kill-includes-power-detain-limits/> accessed 30 June 2020.
107 Goodman (n 48) 169. The British government, for instance, suggested such an expansive approach in the
Mohammed case before the High Court of England and Wales, see Serdar Mohammed v Ministry of Defence [2014] EWHC 1369 (QB) para 237.
v) Killing: Only Alternative to Detention
Lastly, it is argued that if no legal authority to detain can be established, and, hence, the deten-tion by OAGs in a NIAC is illegal, it is likely, that the captives will be killed instead.108
There-fore, it seems reasonable that under IHL, OAGs can capture soldiers instead of killing them.109
This gives effect to the principle of humanity which is inherent in IHL.110 However, this
argu-ment assumes that IHL provides legal authority to kill fighters in NIAC.111 But, the existence
of legal authority to kill in NIAC is under debate. To the contrary, in IAC, legal authority to use lethal force against the adversary party is established.112 Furthermore, states are entitled to
prosecute members of OAGs for killing members of other OAGs or of the armed forces of the state.113 So, the question of legal authority of OAGs to use lethal force in NIAC still remains
unsettled. Consequently, trying to establish legal authority to detain by connecting it to legal authority to kill is not persuasive. In addition, according to the view represented here, a distinc-tion should be made between the quesdistinc-tion of the existence of legal authority to detain and the question of the alternatives. Killing members of the adversary party is widely regarded as the only alternative to the missing legal authority to detain, but–whether it is an alternative or not– it does not lead eo ipso to the establishment of legal authority to detain. From a strictly legal point of view, no argument can be made deriving from killing legal authority of OAGs to detain.
c) Conclusion
According to the view represented here, there is no legal basis under treaty law that would authorise OAGs to detain individuals in a NIAC among OAGs. Therefore, it can only be ex-amined if customary IHL establishes such authorisation.
3. Customary International Law
In scholarship, an authorisation under customary IHL is just as widely affirmed as one under treaty law (relating to a NIAC between a state and an OAG). However, treaty law and custom-ary international law are mixed up often and it is not free of ambiguity why customcustom-ary IHL provides this legal basis. For instance, it is considered that “[i]t could be argued that such a
108 cf Clapham (n 84) 3; Casalin (n 69) 750. 109 Heffes (n 52) 247. 110 Aughey/Sari (n 78) 106. 111 Hill-Cawthorne/Akande (n 78). 112 Art. 43(2) AP I. 113 Hill-Cawthorne/Akande (n 78).
legal basis exists in customary IHL”,114 without further examining the jus ad captivitatem under
custom in more detail.115 Alternatively, the implied power to detain is argued to not only be
found in treaty law but also in customary IHL,116 which is unconvincing because an implied
power in international law is a power not expressly provided in a treaty and, thus, it is deducted by teleological interpretation from a written norm (which customary international law is not)117.
Therefore, it will now be examined whether there is a legal basis to detain members of an OAG by another OAG in the circumstances of a NIAC under customary IHL.
Art. 38(1)(b) ICJ-Statute describes international custom as “evidence of a general practice ac-cepted as law”. It is generally acac-cepted that customary international law requires two condi-tions: state practice and opinio juris. State practice must be settled, and it must include espe-cially the practice of states whose interests are speespe-cially affected.118 Opinio juris is a “belief
that this practice is rendered obligatory by the existence of a rule of law requiring it”.119
Non-state actors, such as OAGs, cannot contribute to the emergence of customary international law.120
a) NIACs among OAGs
Detention of members of an OAG by another OAG in a NIAC is rather rare, since both OAGs have to meet the organisation criterion and the hostilities need to reach the required level of intensity to amount to a NIAC. Besides the conflict between the SDF and ISIL, there are and have been just a few NIACs among OAGs taking place. Among them, regular detention of
114 Chatham House and International Committee of the Red Cross, ‘Expert meeting on procedural safeguards for
security detention in non-international armed conflict’ (London, 22–23 September 2008), (2009) 91 Interna-tional Review of the Red Cross 859, 866.
115 In modern customary international law, new rules are often introduced without being really founded in state
practice, see David J Bedman, Custom as a Source of Law (CUP 2010), 145.
116 Nelleke van Amstel, ‘In Search of Legal Grounds to Detain for Armed Groups’ (2012) 3 International
Human-itarian Legal Studies 160, 168; Jelena Pejic, ‘Conflict Classification and the Law Applicable to Detention and the Use of Force’ in Elizabeth Wilmshurst (ed), International Law and the Classification of Conflicts (OUP 2012) 94; ICRC, Commentary GC I (n 44) para 728. Two years earlier, the ICRC limited the inherent power to detain under treaty and customary IHL for extraterritorial NIACs, see ICRC, Internment in Armed Conflict:
Basic Rules and Challenges, Opinion Paper (November 2014) 7. Since the Commentary of 2016 did not specify
about internment in traditional and extraterritorial NIAC, it is not clear what position the ICRC currently holds.
117 According to Reparation of Injuries suffered in the Service of the United Nations (Advisory Opinion) [1949]
ICJ Rep 174, 180, the power must be “implied in its constituent documents” (in relation to international organ-isations), id est in a treaty.
118 North Sea Continental Shelf (Federal Republic of Germany v Denmark/Federal Republic of Germany v the
Netherlands) (Judgment) [1969] ICJ Rep 43, para 74 (hereinafter: North Sea Continental Shelf case).
119 North Sea Continental Shelf case (n 118) para 77.
120 cf CIHL (n 43) xlii: “The practice of armed opposition groups, such as codes of conduct, commitments made
to observe certain rules of international humanitarian law and other statements, does not constitute State prac-tice as such. While such pracprac-tice may contain evidence of the acceptance of certain rules in noninternational armed conflicts, its legal significance is unclear (…)”.
individuals is not very widespread. In Colombia, next to the NIACs involving the state, NIACs are occurring between OAGs that have notable territorial control. In particular, the ELN (Ejér-cito de Liberación Nacional (National Liberation Army)) are fighting against other rebel groups.121 ELN hold a few “detainees”, but they are considered more as hostages than actual
detainees and they are members of the governmental armed forces, not of the adversary OAG;122 the FARC-EP (Fuerzas Armadas Revolucionarias de Colombia – Ejército del Pueblo
(Revolutionary Armed Forces of Colombia – People’s Army)) used to do and have the same123.
During the war in Yugoslavia, detention by non-state actors took place, as well. However, sev-eral detainees belonged to the state armed forces.124 For instance, the Kosovo Liberation Army,
which was widely considered as a terrorist group,125 ran several detention camps. The ICTY
convicted various persons of crimes which were committed in these camps.126
The only comparable situation to the detention of ISIL fighters by the SDF is detention by rival militias in the Yugoslav wars. Also, there are some similarities with the situation in Colombia. Taking hostages seems to be more common than detention and when OAGs hold captives, they rather hold members of the government’s armed forces than of the adversary OAG. But the act of taking hostages is unlawful and cannot contribute to the emergence of a customary interna-tional norm which creates legal authority for OAGs to detain in NIAC.127 Moreover, no state
acquiesced to the hostage-taking of these OAGs and qualified it as internment. Therefore, ac-cording to the view represented here, there is no settled state practice which considers detention by OAGs of members of other OAGs as lawful. Such occurrences have been too infrequent to affirm the emergence of legal authority under customary international law; there is no evidence. Therefore, a legal basis under customary IHL, which could be applied by analogy to the situa-tion in quessitua-tion, has to be found elsewhere.
121 Geneva Academy of International Humanitarian Law and Human Rights, The Rule of Law in Armed Conflicts
(RULAC), Non-International Armed Conflicts in Colombia < http://www.rulac.org/browse/conflicts/non-inter-national-armed-conflicts-in-colombia> accessed 14 June 2020.
122 Julia Symmes Cobb, ‘Colombia's ELN rebels free three helicopter hostages’ Reuters (Bogota, 3 February 2019)
< https://www.reuters.com/article/us-colombia-rebels/colombias-eln-rebels-free-three-helicopter-hostages-idUSKCN1PS0LD> accessed 14 June 2020.
123 Adriaan Alsema, ‘Kidnapping victims reject FARC’s ‘accommodated’ testimony sent to war crimes tribunal’
Colombia Reports (3 March 2020) < https://colombiareports.com/kidnapping-victims-reject-farcs-accommo-dated-contribution-to-war-crimes-tribunal/> accessed 14 June 2020.
124 Human Rights Watch, ‘Evidence of KLA Secret Prisons in Kosovo and Albania’ (9 April 2009)
<https://www.hrw.org/news/2009/04/09/evidence-kla-secret-prisons-kosovo-and-albania> accessed 30 June 2020.
125 see, eg, UNSC Res 1160 (1998) (31 March 1998) UN Doc S/RES/1160 (1998).
126 Eg, Prosecutor v Zlatko Aleksovski (Appeals Chamber Judgment) ICTY-95-14/1-A (24 March 2000). 127 Niyo (n 50) 7.
b) Traditional NIACs per analogiam
The Customary International Humanitarian Law study by the ICRC focuses on the jus in
cap-tivitate, without providing a legal basis to detain under customary IHL.128 It states the
prohibi-tion of arbitrary detenprohibi-tion129 or rules in relation to the conditions of internment, such as the
provision of basic necessities to130 or the correspondence of persons deprived of their liberty131,
but no rule providing the jus ad captivitatem.
CIHL clearly addresses states with the rules regarding detention but not OAGs, although OAGs are bound by the rules set forth in CIHL, as well, provided that they are parties to the conflict. For instance, under Rule 99, in relation to the grounds for detention in NIAC, CIHL refers to “military manuals, national legislation and official statements, as well as (…) international hu-man rights law” establishing the prohibition of arbitrary deprivation by state practice.132
Do-mestic legislation only provides legal authority to detain for the state itself. The prohibition of arbitrary detention outlined in military manuals is addressed to governmental armed forces and not OAGs. States may call on OAGs to not detain individuals arbitrarily in NIACs with the release of an official statement. But firstly, this does not authorise the detention as such and, secondly, such an official statement may address the international community to emphasise, that the state adheres to the prohibition. Lastly, CIHL refers to IHRL as a state practice. It is not undebated whether IHRL is also applicable for OAGs that are parties to a NIAC.133 CIHL
refers to human rights treaties even more in the context of detention; for instance, to the Con-vention on the Rights of the Child,134 which is applicable only for states, parties to the
Conven-tion135. Interestingly, most of these sources outlined under Rule 99 apply for both IACs and
NIACs.136 Therefore, the view advanced here clarifies that CIHL is considerate of states when
speaking of deprivation of liberty.
128 This position is consistent with the ICRC’s view in general that bases legal authority to detain on treaty law
(inherent power), see ICRC, Commentary GC I (n 44) para 720.
129 CIHL (n 43) Rule 99. 130 CIHL (n 43) Rule 118. 131 CIHL (n 43) Rule 125. 132 CIHL (n 43) Rule 99, 347.
133 see, eg, Sandesh Sivakumaran, The Law of Non-International Armed Conflict (OUP 2012) 95. 134 CIHL (n 43) Rule 99, 348 and Rule 126, 448.
135 Art. 2(1) Convention on the Rights of the Child. 136 CIHL (n 43) Rule 99, CIHL, 347 with n 290.
CIHL contains a broad range of practice, but always (implicitly) considers the state as the de-taining power.137 Nevertheless, there are various examples of OAGs detaining members of the
governmental armed forces as well as civilians. However, in comparison to the detention by the SDF, these detainees are considerably fewer and, moreover, are usually interned for a shorter amount of time. For instance, in 2012 in the NIAC taking place in Mali, the Tuareg National Movement for the Liberation of Azawad (MNLA) detained about 100 people, most of them not being part of the opposing governmental armed forces but they were suspected to be spies of the government. Soon after the detention occurred, all but ten were released.138 Another
exam-ple is the Liberation Tigers of Tamil Eelam (LTTE) during the NIAC in Sri Lanka. They ran their own prisons and detention centres that were located within the territory under their con-trol.139 Both of these detentions by OAGs had in common that they did not provide the
condi-tions of detention and humane treatment, which is required under CA3, the jus in captivitate. This is criticised regularly by various actors, such as states, but without claiming that these detentions were either lawful or unlawful–that is to say, without reference to the jus ad
captiv-itatem.140 This is not sufficient to be considered as a settled state practice.
The Copenhagen Process Principles and Guidelines, which constitute state practice, are com-parable; they regard detention as well-established and accepted in NIAC. It is a multilateral approach to state the conditions and the treatment of individuals detained in NIACs.141
How-ever, neither do they affirm nor deny the existence of a legal authority of OAGs to detain in NIACs. Furthermore, they seem to have a rather state-centric approach.
Lastly, what speaks against the existence of customary IHL in this matter is the practice of states to rely on domestic law to detain members of the adversary party in NIAC.142 According
137 see Jean-Marie Henckaerts and Louise Doswald-Beck, Customary International Humanitarian Law. Volume
II: Practice (CUP 2005) Rule 99, para 2516ff.
138 Human Rights Watch, Mali: New Abuses by Tuareg Rebels, Soldiers. Kidal Offensive Heightens Need for
Discipline, Accountability by All Sides (7 June 2013) < https://www.hrw.org/news/2013/06/07/mali-new-abuses-tuareg-rebels-soldiers> accessed 14 June 2020.
139 International Crisis Group, ‘Sri Lanka’s Return to War: Limiting the Damage’ Asia Report N° 146 (20
Febru-ary 2008) 13
< https://reliefweb.int/sites/reliefweb.int/files/resources/60A40B7D476FF2ACC12573F5004F7CC7-Full_Re-port.pdf> accessed 14 June 2020.
140 see, eg, n 138 and 139. 141 Aughey/Sari (n 78) 107.
142 cf Hill-Cawthorne/Akande (n 78), who refer to Nepal’s Terrorism and Disruptive Activities Act and Ordinance
and Sri Lanka’s Prevention of Terrorism Act. The former is extensively discussed in UNSC, ‘Letter dated 28 December 2001 from the Chairman of the Security Council Committee established pursuant to resolution 1373 (2001) concerning counterterrorism addressed to the President of the Security Council’ (18 January 2002) UN Doc S/2001/1326.
