As noted by Sean and Aurel, the appeals proceedings in Serdar Mohammed v Ministry of Defence have sparked a renewed debate about detention in non-international armed conflict (NIAC). They have set out their arguments in an interesting article and in summary form in this post. I am not convinced by their arguments though, and despite the fact that certain provisions of the law of NIAC address the restriction of liberty or otherwise recognize that on occasion persons will be held by a party to the conflict, I do not see any authorisation for detention in the black-letter, or customary, law of NIAC. In this reply, I address some of the arguments made in favour of finding such authorisation and put forward an opposing view, in support of Leggatt J’s judgment.
Sean and Aurel, and others claiming that authorisation to detain must exist because it is (partially) regulated, fail to acknowledge that the entire body of post-WW2 IHL shows that the regulation of a situation (or behaviour) does not make the occurrence of that situation legal or authorised. The pragmatism of the ICRC and the recognition that conflicts would continue to occur and regulation of the behaviour of warring parties would continue to be necessary, despite the UN’s insistence that no further need for regulation was necessary after the adoption of the UN Charter that outlawed aggression, does not make it legal to wage war. The fact that rules were adopted for NIAC, did not give armed groups any authorisation to fights their governments (or each other). Nor did it authorise governments to take action against such armed groups. Instead, IHL explicitly recognises that sovereign States had that right, independent of IHL.
To construe a legal basis to detain under IHL as a corollary from a non-explicit (allegedly “implicit”) authority to kill is not convincing. Especially when this legal basis is supposedly found in the rationale and principles of IHL, yet said to only apply to one side of the conflict, thereby turning up-side-down the entire rationale of IHL (parity of the parties). A principle of IHL has to apply equally to all sides; otherwise it cannot be a principle. As a Brit, like Leggatt J, would say: “You can’t have the cake and eat it”.
To follow the order of the arguments as they are made in the post, I will start with a comment on Additional Protocol II (AP II). It is no secret that AP II is a partial adoption of the IHL applicable to international armed conflicts (law of IAC) into IHL applicable during times of NIAC (law of NIAC). It started out as a comprehensive set, but during the drafting more and more was cut. Rather than a wholesale transplant, AP II ended up being a collection of core protective rules, on which all States could agree. Furthermore, it is, of course, very clear that the word “persons” used in Article 4 (Fundamental guarantees) of AP II, derived from the article addressing fundamental guarantees (Art. 75) in Additional Protocol I (AP I). Article 4 was the proposed NIAC-variation of Art. 75. Similarly, the AP II provisions dealing with the protection of the civilian population (Art. 13 and further) reflect the language of their IAC-counter parts in AP I. However, not all IAC-rules were incorporated in AP II, such as, significantly, the provision on combatant privilege. It is therefore hard to see how the existence of these (and other) protective rules, would necessitate the existence of a legal basis for (status-based) targeting, on the basis of that same set of rules. Such a line of reasoning fails to appreciate the realism that underlies IHL: that also for unwanted situations, it is useful to aim for regulation of certain behaviour. Moreover, it disregards the historical development of the international regulation of NIACs. As I set out here and here (and as also has been highlighted by Lawrence and Dapo, Kevin and Jonathan, and in published form by Els and Gabor) the development of the law of NIAC shows that authorisation to detain was not included in IHL. Initially, international law of war only became applicable to those internal situations where the fighting had reached such a scale that it either effected other States, or the government facing a rebellion had lost such control over its territory that is was beneficial to have the opposition become bound by certain rules regulating the fighting. In those situations, third States or the relevant government could opt for recognition of the rebels as belligerents. When in the twentieth century humanitarian considerations came into play as a reason for international regulation of civil wars (the threshold that Common Article 3 was originally envisaged for), only protective rules were adopted (see, e.g., here at pp 69-72).
And of course, all of this makes perfect sense, for in internal situations, i.e. in classic civil wars, the government obviously wants to be the only party that is allowed to detain. Indeed, the government can detain rebels on the basis of national criminal law, but it would be undesirable for States to allow rebels to legally hold captured members of the government’s armed forces. Moreover, the law of NIAC does not recognise such a thing as a “license to kill”, as it does not include the combatant immunity/privilege. Again, this makes perfect sense in a classic NIAC, where the government, on the basis of the State’s national laws has, and wants to have, the monopoly on the use of force. It alone wishes be entitled to use lethal force against those rising up against the government, yet it wants to be able to prosecute and punish members of the opposition when they kill government soldiers. States have not and will not grant armed groups such a right/authorisation, so it will never become part of customary IHL applicable to NIAC. Aurel mentions in one of the comments that domestic law does not provide “sufficient basis for status-based operations”. Yet, even if that is the case, all it shows is that status-based targeting (and relatedly the Interpretative Guidance) may be problematic in a NIAC, not that for lack of legal basis in domestic law it must therefore be considered to be authorised under the law of NIAC. In quashing a rebellion, States may wish to go beyond what their domestic law provides for, but that does not mean that the law of NIAC should therefore be seen to authorise such actions. The interference by international law that States accepted was purely a certain number of limitations to the manner in which they would handle their internal matters.
Relying on the IAC version of military necessity to distil permissive rules of NIAC is tricky. Not only would States refuse to recognise that “the complete submission of the enemy” would be a legitimate aim for armed opposition, States do not need IHL in order to find legal grounding to pursue that aim in a NIAC. A government can call upon other forms of necessity, such as the need to ensure stability and safety. Transplanting the concept of military necessity as set out by the IMT Nuremberg wholesale to NIAC, and distilling permissive rules from it, would necessarily mean that armed groups would also have these permissive rules. In any case, the view that the law of NIAC would contain permissive rules is very recent and only relates to extraterritorial situations, where (some) States needed a justification for targeting certain persons. History has shown that in internal situations States are reluctant to admit that a NIAC exists and that IHL applies. In a classic NIAC, States do neither need the law of NIAC to include any permissive rules, nor do they want it to.
It is true that case law of international court and tribunals establishes that in a NIAC, people can be held internationally criminally liable only for violations of IHL. But that legal limitation by no means supports the premise that a right to kill or detain exists in the law of NIAC. Participation in hostilities is not an international crime, but it is a crime under municipal criminal law, at least for those fighting against the government. In fact, it is generally recognised that such participation is a crime that warrants extradition (for example, of alleged PKK members to Turkey, or alleged members of armed groups fighting in Iraq to the US). The requirement of double criminality is considered to be fulfilled in case of such extradition requests, because it relates to a NIAC and the alleged crime also qualifies as a crime under the national law of the extraditing State. IHL does not form any barrier to such extraditions.
One of the main arguments made in favour of an authorisation to detain builds upon a “continued authority to kill”. It does sound interesting, but it appears to be a partial and incorrect transplantation of the law of IAC to NIAC. In an IAC, a combatant does not become a civilian when he falls hors de combat. There is, however, explicitly no continued authority to kill, e.g., a POW (nor is such authority suspended). A person hors de combat is protected against attack. Period. Only when a POW voluntarily relinquishes his protection by escaping, he can be fired at again.
To say, as Sean and Aurel do in their article, that there must be a right to detain, because otherwise a ‘fighter’ could be killed immediately upon release fails to acknowledge that this is not allowed with respect to POWs either, and ignores the language of Article 5(4) of AP II. The argument also misrepresents the significantly different situation between an IAC and a NIAC.
Furthermore, the argument does not hold (sufficient) water, as it cannot work in practice: if the authority to detain is derived solely from the “continued authority to kill”, supposedly based on a continuous combat function (CCF), what then happens if the CCF ceases? If we are to follow the “continued authority to kill” argument, when can we consider the CCF ended for that purpose? On the basis of the ICRC Interpretative Guidance, the CCF would end once the person is unable to continue “to repeatedly participate in hostilities”. As the membership is not based on national law or an ID card but on “function”, it would cease to be continuous when it is impossible to retain the said function (which, for example, happens when someone is detained and [physically] unable to participate in hostilities). Even if a more lenient approach were taken, the CCF would still have to cease once the group that the person belonged to is no longer organised (which happens regularly, see for example the situation in Syria), or when the group is defeated. In such a case, would the government then be obliged to immediately release the person? Neither will this happen in practice, nor does it have to. Such an obligation does not exist, precisely because the situation is different from an IAC, where a POW is held, with IHL as the legal basis, to prevent him from taking part in the continuing “active hostilities” (and cannot be punished for taking part in hostilities because of the combatant privilege). The ‘fighter’ in a NIAC, however, can be detained under national law, as he fought the government and violated national law. And that same national law continues to provide justification for detention, also after the NIAC ends (of course, charges should be brought etc, but that is a different matter).
I acknowledge that La Tablada uses language that would support Sean and Aurel’s argument, although it is debatable whether this judgment is the best source to rely on. The Korbely case, however, is merely based on findings of the Hungarian national courts and the ECtHR’s does not really seem to find anything more than the Hungarian courts had erroneously qualified the victim as a civilian (for the purposes of crimes against humanity). The ECtHR’s finding that the victim was not entitled to protection was based on the fact that – in the Court’s view – he was directly participating in hostilities at the time he was shot, not on any perceived CCF (which anyway could not appear to have been the case, based on the factual findings of the Hungarian courts). All that the case appears to confirm is the existence of the protective rules of the law of NIAC, but not of any permissive rules. Moreover, a case that fundamentally misunderstood some essential aspects of IHL, as well as of crimes against humanity (see, e.g., the analysis of the case in this book), and talks about “combatants” and “non-combatants” in the context of a NIAC, can hardly be called a convincing authority.
Of course, I have to recognise that my understanding of the law of NIAC is premised on the understanding that IHL applies equally to both sides of the conflict, whilst Aurel is of the view that customary IHL for NIAC is asymmetrical. It is an intriguing proposition, but so long as it has not convincingly been shown that this is the case, I continue to operate on the basis of the classic view, where IHL is based on parity of the belligerents (as already clearly set out by Dapo and Lawrence on an earlier occasion. See also here and here). Of course, some States would be all too happy to have such an asymmetrical law applicable to their military operations against armed groups. Just like certain States would welcome a version of IHL (applicable to IAC) that distinguishes between the “good guys” and the “bad guys”, with stricter rules applicable to the latter. But can such a body of asymmetrical rules still be called IHL?