The international legal framework on the recourse to armed force consists of the prohibition on the inter-State use of force, existing in customary law and enshrined in Article 2(4) of the UN Charter, and two exceptions to this primary rule: authorization of force by the UN Security Council under Chapter VII of the UN Charter, and self-defence against armed attacks under customary law and enshrined in Article 51 of the Charter.
These components of the normative framework are tightly interwoven.
For the purposes of this article, we focus on the right to use force in self-defence. The existence of the right is uncontested. However, the precise parameters of self-defence have been subject to long-standing, vigorous debate, much of which has centred on what constitutes an ‘armed attack’ that would trigger the right.
We engage with only one aspect of this debate: whether and when attacks by non-State actors trigger the right to self-defence.
We examine this question in relation to self-defence measures against IS in Syria.
The circumstances under which the actions of non-State fighters can be attributed to a State such that it would be exposed to lawful response action by the victim State have long been questioned. The questions assumed heightened importance after the terrorist attacks of 11 September 2001, and have taken centre stage since the rise of the IS network in Iraq and Syria.
We will illustrate that repeated cycles of legal challenge have resulted in a loosening of the attribution standards and, since 2001, in widespread acceptance that the right to self-defence can be exercised not only against armed attacks that are undertaken by, or attributable to a State, but also against terrorist attacks without State nexus. In the latter context, however, the circumstances under which armed response action can be taken in the territory of another State remain unsettled. While a number of States have advanced the proposition that such action is lawful when the relevant State is unwilling or unable to avert the threat emanating from terrorists operating in its territory, this approach is contested. Many States explicitly resist the formula, others are silent or ambiguous in their reactions, and a handful have advanced an alternative, narrower set of preconditions for lawful response action.
A. Evolution of Legal Practices in the Charter Era
The UN Charter's use of force framework was a response to World War II and, therefore, enshrined rules that were directed to inter-State conflict.
However, the text of Article 51 does not make reference to ‘States’ as the origin of the ‘armed attack’ that gives rise to the right to self-defence, and commentators have suggested that the pre-Charter history does not justify insistence on a strictly inter-State right to self-defence.
Some States and commentators began, as early as the 1960s, to assert that the ‘inherent’ right of individual or collective self-defence could be exercised against non-State actor attacks, and in States that were harbouring the attackers.
Israel was the first State to assert, in the 1970s, that it could take self-defence measures in a State (Lebanon) that was unwilling or unable to prevent cross-border attacks on Israel emanating from its territory. These interventions, however, were denounced by the Security Council and rejected by many States, albeit for a variety of reasons unrelated to the non-State actor dimension.
The controversy surrounding attacks by non-State actors, and any right of response, was highlighted in the course of the UN General Assembly's efforts to provide authoritative definitions of the use of force. Article 3(g) of the Definition of Aggression qualified the ‘sending by or on behalf of a State’ of non-State forces as an act of aggression, along with situations involving a State's ‘substantial involvement’ in the non-State attack.
It is significant that even those States that had urged an expansive definition accepted the premise that only attacks directly supported by States could amount to aggression.
The International Court of Justice (ICJ), in its landmark ruling in the Nicaragua case,
built on the Definition of Aggression to underscore that only attacks by a State, or attributable to a State, could trigger the right to use force in self-defence. The Court took Article 3(g) of the Definition of Aggression as a statement of customary international law, but focused on the ‘sending’ of irregular forces, requiring that the State had effective control over the armed operation.
In noting that various kinds of assistance to irregular forces might amount to illegal intervention or illegal use of force, but not an ‘armed attack’, the Court read the ‘substantial involvement’ aspect of Article 3(g) of the Definition of Aggression narrowly.
The ICJ reaffirmed the narrow construction of the right to self-defence against non-State actor attacks in several subsequent decisions.
However, with the rise of international terrorism in the 1980s, the United States began to assert a broader right to self-defence. Initially it relied primarily on wider tests for State involvement, but, beginning in the 1990s, it advanced the unwilling or unable standard
—a standard that plainly did not meet even the most expansive reading of the ‘substantial involvement’ criterion in the Definition of Aggression.
After the terrorist attacks of 11 September 2001, the issue of self-defence against non-State actor attacks rose to the top of the international agenda. The Security Council, in resolutions 1368 and 1373,
recognized ‘the inherent right of individual or collective self-defence’. However, in their letters to the Security Council indicating that they had begun self-defence operations in Afghanistan, the United States and the United Kingdom did not invoke the unwilling or unable standard. Instead they highlighted the support of Al-Qaeda by the Taliban regime,
effectively deploying the ‘substantial involvement’ standard. The international response to these operations was generally supportive, but international practice remained ambiguous, fitting narrower self-defence criteria or being accompanied by alternative justifications.
The rise of the IS, a sophisticated, transnational terror network that at one point exercised a substantial degree of territorial control over parts of existing States, prompted other States to take military actions in Iraq and Syria. The interventions by Western States in Syria are particularly relevant for present purposes, as they were undertaken under the banner of self-defence against IS, and in the absence of a link between the IS operations and the Syrian State. Indeed, Syria, with its ally Russia, was fighting its own battles against IS.
Since military action is accompanied by explicit legal justifications, the operations against IS in Syria provide an opportunity to assess both material and rhetorical practices.
In September 2014, the United States and a coalition of several other States began to strike IS in Syria,
from where the network was planning and carrying out cross-border attacks in Iraq along with other terrorist attacks. To find reliable evidence of the legal opinions of the intervening States, we use the official letters they sent to the Security Council, in keeping with their obligation under Article 51 of the UN Charter, to report that they had initiated self-defence measures. From a normative perspective these statements help us understand the legal meaning of these concrete actions.
In its letter of September 2014 to the Security Council, the United States asserted that States ‘must be able to defend themselves, in accordance with the inherent right to individual and collective self-defence … when … the government of the State where the threat is located is unwilling or unable to prevent the use of its territory for … [terrorist] attacks’.
With this statement, the United States unequivocally went beyond the Nicaragua case and the Definition of Aggression, asserting that the right to self-defence against terrorist actors could be exercised in another State absent ‘substantial,’ or for that matter any, involvement of the State with the attacks. Canada sent a similarly worded letter in March 2015,
and Australia followed suit in September 2015.
Some coalition members that participated in operations in Syria, including Bahrain, Jordan, Qatar, Saudi Arabia and the United Arab Emirates, refrained from making explicit legality claims.
Other coalition States, including Belgium, Denmark, and the Netherlands, confined their participation to Iraq, which had consented to strikes in its territory.
The November 2014 letter sent by the United Kingdom, for example, stated only that it was ‘taking measures in support of the collective self-defence of Iraq as part of international efforts led by the United States’ and that it fully supported international strikes on IS ‘sites and military strongholds in Syria’.
France, which commenced air strikes in Syria in September 2015, offered only a general Article 51 justification, noting that it had ‘taken actions involving the participation of military aircraft in response to attacks carried out by ISIL from the territory of the Syrian Arab Republic’.
Similarly, several Arab States, while supporting or participating in air strikes against IS, have steered clear of invoking the unwilling or unable standard.
A large number of States remained altogether silent on the commencement of strikes against IS in Syria, while a few expressed various kinds of objections.
The situation shifted again, however, after the IS attacks in Paris of 13 November 2015. Shortly after the attacks, the Security Council adopted resolution 2249, which determined that the IS constitutes a ‘global and unprecedented threat to international peace and security’, in part because of ‘its control over significant parts and natural resources across Iraq and Syria’.
Resolution 2249 went on to call upon States ‘to take all necessary measures, in compliance with international law, … on the territory under the control of ISIL … in Syria and Iraq’.
In the Security Council debates after the adoption of the resolution, the French representative stated that the French military action in Syria, which France had previously justified as collective self-defence, ‘can now also be characterized as individual self-defence, in accordance with Article 51’.
France did not, however, invoke the unwilling or unable standard advanced in the US Article 51 letter from 2014. The same is true for the UK, which began to extend its military strikes to IS sites in Syria in December 2015. Although Prime Minister David Cameron invoked the unwilling or unable doctrine in two public pronouncements,
the official statement of the British government, contained in the letter of 3 December 2015 to the Security Council, declared only that the UK was ‘taking necessary and proportionate measures against ISIL/Daesh in Syria, as called for by the Council in resolution 2249’.
Resolution 2249 was ambiguous in that it neither authorized the use of force under Chapter VII of the UN Charter nor offered an explicit endorsement of various States’ assertions of self-defence,
much less of the unwilling or unable standard. Nonetheless, it played an important role in the continuing process of contestation over the scope of the right to self-defence. Particularly significant in this context was the justification offered by Germany for joining in the collective self-defence operations against the IS in Syria. Germany's letter to the Security Council of 10 December 2015 refers to resolution 2249 and goes on to state that ‘ISIL has occupied a certain part of Syrian territory over which the [Syrian] Government … does not at this time exercise effective control’ so that States are ‘justified under Article 51 … to take necessary measures of self-defence, even without the consent of the [Syrian] Government’.
Belgium sent a similarly phrased Article 51 letter in June 2016.
Two aspects of this approach to justification stand out. First, Germany and Belgium emphasized the IS control over parts of Syrian territory and the resultant lack of effective control by Syria of the relevant areas. Second, like a number of other countries, Germany and Belgium linked their self-defence arguments to resolution 2249. In other words, the German and Belgian Article 51 letters provided a narrower justification than the unwilling or unable standard advanced by the United States, a justification that some commentators have likened to a ‘lex-ISIL’, a special set of rules not applicable to other terror networks who lack ‘a territorial basis and military forces capable of making and stabilizing territorial gains’.
What is more, by connecting this argument to a Security Council resolution that had confirmed IS control over parts of Syria, the German and Belgian letters appear to suggest that a determination by the Council of the preconditions for self-defence (with the multilateral deliberation that Council involvement entails) is required in the case of non-State actor attacks that cannot be attributed to a State.
Four other States (Denmark, the Netherlands, Norway and the United Kingdom) connected their Article 51 justifications to resolution 2249,
albeit not in the specific way that Germany and Belgium had done.
The legal picture at the end of May 2017, then, was that some shifts in the material and rhetorical practices concerning the right to self-defence have occurred. Whereas before 11 September 2001, and even up until the recent events involving the IS, there was still considerable support for the idea that only attacks by a State or attributable to a State give rise to a right to use force in self-defence,
it now seems much more widely accepted that non-State actors can commit armed attacks.
The threshold for attributing non-State actor attacks to a State seems to have been lowered, with a more expansive understanding of ‘substantial involvement’ having gained ground.
However, the preconditions for the exercise of the right to self-defence against a non-State actor attack undertaken from a third State, but not attributable to that State, remain contested.
According to the authors of an October 2016 article surveying State attitudes towards the unwilling or unable standard, ten States (US, UK, Germany, the Netherlands, Czech Republic, Canada, Australia, Russia, Turkey and Israel) explicitly endorsed the test, while three States (Belgium, Iran, South Africa) were doing so implicitly; nine States’ positions were classified as ambiguous, and four States (Cuba, Ecuador, Syria and Venezuela) were found to object.
A closer look suggests, however, that the survey overstates the support for the unwilling or unable standard and that a consistent pattern of practice concerning the right to self-defence against non-State actors is in fact difficult to discern, even among the States that have participated in military actions against IS in Syria.
Germany and Belgium's positions arguably should not be classified as endorsements, explicit or implicit, of the unwilling or unable test, but as advancing a more specific, narrower standard.
Perhaps surprisingly, even the United Kingdom's position is not unequivocal. As noted above, then-Prime Minister Cameron invoked the ‘unwilling and/or unable’ standard in public statements on two occasions in November 2015.
Yet, in three separate letters of the UK government to the Security Council explaining the use of force in Syria and Iraq, the UK government never uses the phrase ‘unwilling or unable’.
Even the United States, in a December 2016 ‘legal and policy framework’ document, uses examples to support the unwilling or unable standard that are the narrowest cases, rooted strictly in necessity as traditionally understood. It allows that ‘inability perhaps can be demonstrated most plainly where, for example, the State has lost or abandoned effective control over the portion of its territory’ from which the non-State actor operates. This articulation is actually close to the German standard, which requires the non-State actor to have effective territorial control. In turn, ‘unwillingness might be demonstrated where, for example, a State is colluding with or harboring a terrorist organization operating from within its territory and refuses to address the threat posed by the group’.
This statement approximates the old ‘substantial involvement’ trigger. The use of narrow justifications aligns with the broad assertion in the document that there is no intention to undermine State sovereignty.
What is more, in attempting to anchor the unwilling or unable standard in State practice, the US government cites three governmental letters to the Security Council, attributing one of those letters to France that was actually sent by the Canadian government.
As we noted above, France in fact has never clearly supported the US legal position.
It must also be noted that although the Canadian government of Stephen Harper did support the unwilling or unable standard, the new Justin Trudeau government ceased its air strike operations in Syria and Iraq.
It is not clear that this largely political move can be read as implying a change in legal position. Canada's past statement was evidence of the close alignment of the then-government with the US government's security policies and legal positions. However, the current US and Canadian governments seem to be moving in different directions on several issues. Perhaps the relatively new Canadian government may not yet have had an opportunity to fully review its position on the unwilling or unable standard.
The Czech Republic expressed uncertainty as to whether the unwilling or unable standard was ‘part of the customary law international law requirement of necessity’,
as the United States appears to argue.
Russia has implicitly retreated from its clear endorsement of the unwilling or unable test, at least in the context of Syria.
Finally, the Non-Aligned Movement (NAM), a loose grouping of 120 States, at its September 2016 meeting, reiterated its view that ‘consistent with the practice of the UN and international law, as pronounced by the ICJ, Article 51 of the UN Charter is restrictive and should not be re-written or re-interpreted’.
Although this statement does not specifically refer to the unwilling or unable standard, the NAM has long resisted any attempts to expand the right of self-defence, or any other unilateral use of force against sovereign States.
There is no doubt that some Western States, led by the the USA and arguably, the UK, are attempting openly to shift the law. They are supported most clearly by Australia and Israel. We salute the willingness of these States to offer explicit legal justifications for their material practice. Frank justification is preferable to simply accreting practices that are never related to transparent legal arguments, leaving commentators and other States to read tea leaves. However, it turns out that the legal statements by important governmental actors such as the UK Attorney General and the US President are more equivocal than they seem to be at first glance. Moreover, there is a curious interplay amongst State officials, former officials writing in their personal capacity and some academic commentators, whereby a small group tries to expand its influence by constantly cross-referencing each other.
Curiously, though, they tend to pass the ball by suggesting that others have supported the new norm, without saying that they themselves support the norm. For example, in a major speech from January 2017, specifically intended to lay out ‘when it is lawful to use force in self-defence,’ the UK Attorney General did not directly align the United Kingdom with the unwilling or unable test. Instead, he observed that:
A number of [S]tates have also confirmed their view that self-defence is available as a legal basis where the State from whose territory the actual or imminent armed attack emanates is unable or unwilling to prevent the attack or is not in effective control of the relevant part of its territory.
In support of this proposition, he cited the October 2016 survey of State practice that counts the UK among supporters of the unwilling or unable standard.
That survey, however, quotes only the Prime Minister's statements, while acknowledging that the UK letters to the Security Council make no mention of the test. This approach is at best a backhanded endorsement of the ‘unwilling or unable’ standard.
It would seem that all the advocates of the unwilling or unable standard are conscious of their fragile norm entrepreneurship. Even those States that invoke the standard in their letters to the Security Council preface their justification by saying that States ‘must’ be able to invoke the unwilling or unable’ standard (a policy argument), not that they ‘can’ (a legal argument).
In our view, the rather mixed, and largely self-referential, practice of a small number of primarily Western States cannot suffice to shift customary law in the face of the silence of a majority of States on the operations against IS in Syria
and the explicit preference of 120 States that the self-defence norm remain a narrow exception. To be more specific, on our reading of the available material and rhetorical State practice, the unwilling or unable standard is currently supported by only five countries: the USA, Israel, Australia and Turkey unequivocally, and the UK ambiguously. Russia and Canada's positions are also ambiguous, with Russia withdrawing its earlier support of the standard in the Syrian case and Canada retreating from its involvement in air strikes. In both cases, these decisions could be seen as politically opportunistic, without any legal intention. Yet these decisions suggest that support for the unwilling or unable standard expressed by each State previously might now be equivocal.
No State from Africa, central or east Asia, or Latin America has even engaged in the debate, other than through support for the general statements on self-defence of the Non-Aligned Movement.
The NAM statements, however, should be accorded considerable weight, given the sheer number of States endorsing them. Meanwhile, Germany and Belgium are advocating for a more restrictive standard which may be specific to IS's activities in Syria. This constellation of State practice falls far short of the ‘widespread and consistent’ practice required for the formation of customary international law.
We see efforts to articulate an opinio juris around the unwilling or unable standard, but here too the examples are highly limited and not entirely consistent.
B. Processes of Legal Change
World events evidently play a role in prompting and shaping change in law. The rise of the IS, a global terror network with extraordinary resources and capacity to inflict harm, marked a key moment in the long-standing debate around the scope of the right to self-defence. Furthermore, as one commentator put it, IS ‘barbarism seems to have opened the floodgates’ for acceptance of the unwilling or unable standard.
However, change in material circumstances is not itself sufficient to prompt a shift in a legal framework. Indeed, it may be that unique requirements of law militate against a rapid or radical shift in rules. The debate around the legal status of the unwilling or unable standard for self defence against non-State actors illustrates this dynamic.
We use the interactional law approach that we have developed over the last 20 years as a useful lens through which to examine why and how change occurs in international law.
It reveals why the various arguments for and against the unwilling or unable standard raise not only policy questions, but fundamental issues of the rule of law. It also offers practical insights into the processes and mechanisms of legal change, providing guidance to norm proponents and resisters. Our framework posits that all legal norms are embedded in, and must be broadly in line with, social norms that arise from the practices and understandings of the society in which they operate.
However, we stress that these shared understandings alone do not make law. What distinguishes legal norms from other types of social norms is not widespread social acceptance alone; nor is it form or pedigree, but adherence to criteria of legality. The most commonly referenced legality criteria were proposed by Lon Fuller,
and we use them as shorthand for ‘legality’ because of their clarity and cogency. They are central to every effort to define the rule of law and they capture distinctive traits of legal practice.
Finally, norms that are based in shared understandings and that largely meet the criteria of legality must continually be upheld through norm implementation, application and interpretation that also meet the criteria of legality, what we call ‘practices of legality’. As we will show, this framework provides a unique perspective on the interplay between the State practice and opinio juris required to form and uphold customary international law.
It is important to remember the context for the debate over the unwilling or unable standard. The prohibition on the use of force, set out now in Article 2(4) of the UN Charter, is a widely shared understanding. It is the primary rule in the field; the right to self-defence is an exception. This rule-exception relationship plays out in all attempts to widen the right to self-defence which have been entangled in persistent contestation, framed by insistence on narrow, verifiable parameters for the exception. As we examine the unwilling or unable standard against the criteria of legality that distinguish law from other forms of social norms, we will focus on the ability of the standard's proponents to reshape the law.
The unwilling or unable standard is subject to serious questions in relation to the criteria of legality that legal norms be ‘general’—that they be ‘rules’, rather than ad hoc decisions. While framed in apparently general terms, in practice the unwilling and unable standard singles out a certain category of State. Several commentators have observed that the test has been applied only to States in the Global South
and, we would add, is being promoted primarily by relatively powerful Western States. The test establishes a ‘legal framework for … the semi-periphery’
harkening back to the standard of ‘civilization’ in nineteenth century international law
that enabled European States to accord non-European States a lesser legal status because of how they were ‘internally organized’.
This critique is powerful to the extent that the effect of the unwilling or unable standard can be to single out even ‘willing’ States on the basis of relative incapacity.
The sovereignty of such States would be compromised, for they would be presented with the choice of either consenting to operations in their territories or suffering intervention under the banner of self-defence.
The effect of the German and Belgian attempts to limit, or displace, the nascent unwilling or unable standard is to mitigate the standard's impact on the criterion of generality. The focus on territorial control of the non-State actor, affirmed by a Security Council finding, serves to render highly exceptional the circumstances in which a State that is not actively supporting the non-State actor can be exposed to self-defence measures.
On the standard account, customary law is generated and maintained by a consistent, widespread and representative practice, accompanied by States’ conviction that they are legally bound to conduct themselves in accordance with the norm (opinio juris).
Norm-shifts, then, occur when a new standard is being embraced by consistent and widespread practice, accompanied by opinio juris understood, we have argued, as continuing practices of legality. Since much of the content of Article 51 of the UN Charter is provided by customary law, the treaty provision is subject to the inherent dynamism of this process. Although customary law nonetheless tends to be relatively stable, shifting only incrementally, the diffuse, fluid nature of the process makes it harder to identify the precise point at which a new law has been promulgated.
Promulgation matters because legal rules must be known; they cannot be secret. Uncertainty as to whether or not a rule actually exists undermines its legality. In the current context it is safe to say that the scope of the right to self-defence against non-State actor attacks remains highly contested, as admitted even by proponents.
Therefore, the unwilling or unable standard has not been promulgated as law, nothwithstanding its invocation by some States to justify their actions in Syria.
Clarity of a rule is another criterion of legality. Much of the struggle over the scope of the right to self-defence revolves around efforts to maintain parameters that are capable of objective assessment.
To be sure, the formulation of the right to self-defence must be sufficiently open-meshed to accommodate a diversity of situations and allow for evolution. Nonetheless, the capacity for objective assessment has been at issue in virtually all the discussions about expanding the right, in contexts that go beyond the case of non-State actors. For example, in his famous account of the Cuban Missile Crisis, the former US State Department Legal Adviser, Abram Chayes, explained why the United States was not relying on the right to self-defence in the following terms: ‘… to expand … [the right to anticipatory] self-defence to include threatening deployments that do not have imminent attack … as their probable outcome … is to make the occasion for forceful response essentially a question for unilateral decision that would not only be formally unreviewable, but not subject to intelligent criticism either’.
Similarly, in the current debate about self-defence against non-attributable attacks by non-State actors, it is precisely the fact that the lack of clarity of the unwilling or unable test provides greater room for self-serving claims that has been of concern.
Even strong proponents of the unwilling or unable standard have stressed the importance of objective assessment in the law of self-defence.
The criterion of non-contradiction shines the spotlight on the rule-exception relationship between the rules contained in the UN Charter's use of force provisions. Chayes’ observations in the context of anticipatory self-defence apply with equal force in the debate about the scope for self-defence against non-State actor attacks: a wide ‘construction of “armed attack” could not help but weaken … normative checks’ on how [S]tates behave.
We would add that an overbroad reading of the self-defence exception would threaten to eviscerate the primary rule, in fact contradicting it. This point has been made elsewhere in powerful terms, with one commentator asking whether embracing the unwilling or unable standard would entail removing the legal constraints on inter-State violence, and whether it would produce ‘a creature that can still be called law’.
The protracted contestation surrounding the unwilling or unable standard tends to confirm that these concerns are not merely those of academic commentators, but are shared by a great number of States. The inherent contradiction between an expansive right to self-defence and the prohibition on the use of force is among the legality considerations holding back the acceptance of the unwilling or unable standard as a rule of customary international law. Arguably, the same is true for the tension between this standard and the foundational legal principles of sovereignty and non-intervention.
Legal norms should make reasonable demands, not ask the impossible of those to whom they are addressed. For example, in the current context, a sweeping and absolute prohibition on the use of inter-State force would be practically (and politically) unthinkable without appropriate exceptions. This practical point has legal implications. An overly narrow construction of the right to self-defence might be seen as asking States to do the impossible: to suffer attacks and tolerate urgent threats.
For example, the ICJ in the Nicaragua case requires that a State exercise effective control over the armed actions of non-State actors for it to be liable to self-defence action on its territory. Increasingly, this view is seen to be too narrow, demanding an impossibly high proof of linkage before necessary self-defence action can be taken.
Here we have an example of a pull towards legal change resulting from what has become an arguably unreasonable legal constaint.
In the context of the unwilling or unable standard, the need to avoid impossible demands actually pulls in the opposite direction, resisting change to current limitation on the lawful invocation of self-defence. As currently phrased, the standard would make it impossible for States that are willing but unable to take sufficient measures against terrorist attacks from their territory to avoid foreign military intervention.
Either they consent to intervention, or have it forced upon them.
A final criterion of legality is congruence between the purported demands of a rule and its operation in practice. Although the unwilling or unable standard has been invoked by a very small group of States for some time, as we have demonstrated, there does not currently exist a widespread, consistent and representative practice of legality around the standard. We venture to suggest that such a practice is unlikely to develop unless the failures to instantiate the criteria of legality that we have traced out above are sufficiently addressed, and concerted efforts are made to build wider and more representative support for the unwilling or unable standard. In the conclusion, we will suggest what might be done to open up avenues for inclusive support for this purported legal norm.