The 25 February 2021 military strikes and the ‘armed attack’ requirement of self-defence: from ‘sina qua non’ to the point of vanishing?

ABSTRACT Following a rocket attack that occurred at Erbil airport in Iraq, President Biden authorised the first use of military force since becoming President on 25 February 2021. This was legally justified on the basis of self-defence. On the face of it this seemed an innocuous justification. Yet, this article argues that through both downplaying the treaty source of the right of self-defence and its express requirement for an armed attack, as well as promoting a contextual and enabling form of necessity, the Biden administration’s military action and ensuing strategy of legal justification place question marks over the meaning of, and even the requirement for, an armed attack. However, seeing the 25 February incident in the context of broader US and other state practice, while various attempts at diluting this requirement and the interpretation provided to it by the International Court of Justice have been sustained, others have clearly not.

militiamen. 3 Unlike the Trump administration's targeted killing of Qassem Soleimani just over a year earlier, 4 the Biden administration was quick to justify the action, amongst other things informing Congress and submitting a letter to the United Nations Security Council (UNSC), 5 as it is required to do in these circumstances when it takes action under the legal justification of self-defence. Indeed, in its letter to the UNSC the administration stated that '[t]he United States took this action pursuant to the United States' inherent right of self-defense as reflected in Article 51 of the United Nations Charter.' 6 This provision of the Charter provides: Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations, until the Security Council has taken measures necessary to maintain international peace and security. Measures taken by Members in the exercise of this right of self-defence shall be immediately reported to the Security Council and shall not in any way affect the authority and responsibility of the Security Council under the present Charter to take at any time such action as it deems necessary in order to maintain or restore international peace and security. Importantly, this right of self-defence exists as an exception to the prohibition of the use of inter-state force which is contained within Article 2(4) of the UN Charter, as well as in customary international law. 7 The Biden administration was also keen, however, to point out that the strikes were 'necessary and proportionate', two criteria for lawfulness that are not contained within Article 51 or the UN Charter in general, but which today exist within customary international law. 8 Indeed, President Biden stated that '[t]he United States always stands ready to take necessary and proportionate action in self-defense'. 9 While this seemed like a relatively straightforward legal justification, invoking as it does Article 51 and claiming to have taken necessary and proportionate military action, the action and associated legal justification raise a number of important questions and issues, some of which have been 3 Ibid. 4 See 'Qasem Soleimani: US kills top Iranian general in Baghdad airstrike', BBC News (3 January 2020) www.bbc.co.uk/news/world-middle-east-50979463. 5 The White House, 'A Letter to the Speaker of the House and President Pro Tempore of the Senate Consistent with the War Powers Resolution' (27 February 2021) www.whitehouse.gov/briefing-room/ statements-releases/2021/02/27/a-letter-to-the-speaker-of-the-house-and-president-pro-tempore-ofthe-senate-consistent-with-the-war-powers-resolution/; Letter dated 27 February 2021 from the Permanent Representative of the United States of America to the United Nations addressed to the President of the Security Council, UN Doc S/2021/202 (3 March 2021). 6 Ibid. addressed in the other contributions to this symposium by O'Connell and O'Meara. This particular contribution seeks to focus on the requirement for an 'armed attack' before the right of self-defence may be invoked. In particular, the International Court of Justice (ICJ) has stated this to be the 'sine qua non' of the right of self-defence and provided its view on certain aspects of the concept, 10 and an emphasis has been placed on the importance of this particular requirement in the work of scholars. 11 However, while the 25 February strikes were justified as self-defence, the Biden administration's justification placed question marks over the meaning of, and even the requirement for, an 'armed attack'. It did so primarily by downplaying the treaty source of the right where this requirement is expressly stipulated and, instead, moved towards the utilisation of a more enabling form of the necessity criterion of self-defence. However, the strategy adopted on this occasion by the Biden administration should be seen in the broader context of both US practice as well as the practice of other states. What we see is that while various criticisms of the armed attack requirement have at least opened up a degree of ambiguity as to where the law stands which the US has sought to exploit, the US stands more isolated in respect to other attempted reinterpretations.

2.
A targeted downplaying of the relevance of the treaty source of the right of self-defence?
In invoking the right of self-defence, the Biden administration was keen to point out in its various justifications that the right that it was relying on was an 'inherent right', 12 thereby arguably alluding to the customary source of the right of self-defence which began to emerge prior to the adoption of the UN Charter and with it the stipulation for an armed attack expressly contained within Article 51. There is nothing inherently controversial about this; Article 51 itself expressly refers to the right as an 'inherent right of self-defence' that '[n]othing in the present Charter shall impair'. Furthermore, the ICJ has stated that in its view the employment of the word 'inherent' here by the drafters of the Charter is a reference to the right in its customary form. 13 Yet, the Biden administration's choice of phrasing is arguably significant when viewed in light of the fact that it also referred to the right as 'reflected in Article 51 of the United Nations Charter', not, as it might perhaps more intuitively be referred to, 'as contained in' or 'in accordance with' this 10 Nicaragua (merits) (n 7) para 237. 11  provision. This way of phrasing the nature of the right was repeated by virtually all the different members of the Biden administration speaking on the attacks, indicating, perhaps, that it was a predetermined and planned strategy in presenting the right of self-defence on this occasion.
With the Biden administration consistently highlighting both the 'inherent' nature of the right and the fact that the right is 'reflected' within Article 51, one gets the sense that the administration was attempting to place a greater emphasis on the customary source of the right, with the treaty source merely codifyingor 'reflecting'this. There exists no hierarchy between the sources and such a phrasingand arguable downplaying of the treaty source of the rightis not one that is generally seen in state practice. Indeed, when reporting their actions in self-defence to the UNSC, states often begin their letter by stating that they are writing 'in accordance with Article 51', thereby meeting the reporting requirement. By contrast, however, they rarely offer any viewsimplicitly or otherwiseregarding the relationship between the treaty and customary forms of the right of self-defence. Even the US itself has not traditionally done so until relatively recently, with its invocation of the right of self-defence in 1986 in response to the Berlin disco bombing, 14 in 1993 in response to the assassination attempt on former president George H.W. Bush, 15 in 1998 in response to the African embassy bombings, 16 and in 2001 in response to the attacks of 11 September 2001 17all without any similar possible attempts to place greater emphasis upon the customary source of the right over its treaty source.
Yet, we can arguably see a division beginning to emerge between states on this issue during the interventions in Syria in the fight against the so-called Islamic State that took place from 2014. On that occasion, the US, as with the strikes in February 2021, made the claim that the right of self-defence was 'reflected' in Article 51, 18  Admittedly, while this is only an inference based upon a somewhat cursory survey of state practice, 25 it does nonetheless flag the Biden administration's use of language on this occasion as being potentially significant, or, at least, as something of an outlier, and raises questions as to the possible reasons for it, which are addressed subsequently in this article. 3. An express downplaying of the requirement for an 'armed attack'?
If we are to accept that the Biden administration was consciously downplaying the Article 51 source of the right, a possible reason for this might be Article 51's specific requirement that before the right of self-defence may be invoked there must be the occurrence of an 'armed attack'. Indeed, Article 51 essentially contains three specific stipulations: the occurrence of an 'armed attack', that self-defence is a right which exists until the UNSC has taken measures necessary to maintain international peace and security, and that an invocation of the right needs to be reported to the UNSC. The US is a member of the UNSC and so could prevent it from taking action that might call into question the continuing necessity of any action taken.
In addition, and as noted previously the Biden administration reported its actions swiftly to the UNSC. As such, this strategy might arguably be seen as one taken to cast doubt on the first of these elements, that is, the requirement for the occurrence of an 'armed attack' as a prerequisite for the invocation of the right of self-defence. That being said, the ICJ has consistently 'held not only that an action taken under Article 51 must be in response to an armed attack (above all else), but that this also holds true in customary international law.' 26 The question thus arises as to why the US might seek to downplay the treaty source of the right if the armed attack requirement also exists as a sina qua non element of the right in customary international law?
In general, treaties and the obligations contained within them might, on the one hand, be perceived as somewhat rigid, formal and textual, while customary international law, on the other hand, is seen as vague, unwritten, nebulous, and difficult to pin down. 27 Furthermore, the UN Charter as a treaty is something of a constitutional document and attached to a global inter-governmental organisation. Consequently, on the one hand, obtaining any formal modifications to its provisions faces serious obstacles. 28 On the other hand, however, the interpretation of them through 'subsequent practice in the application of the treaty' would require establishing 'the agreement of the parties regarding its interpretation', 29 that is, the agreement of all the parties to the UN Charter, or in other words virtually every state. Modifications to customary international law, however, can take place through the rather less stringent and arguably vaguer concept of a generality of state practice and opinio juris as part of a process which is not easily and uniformly understood and which is, as a consequence, open to often vast differences of opinion and debate. 30 This is not something which arises, at least in the same way, in the context of treaty interpretation. Focusing on the customary source of the right of self-defence may, therefore, be seen to provide both greater flexibility and possible breadth of action, in this case in respect to the armed attack requirement. 31

Gravity issues
While a reading of the UN Charter provides for a prohibition of the 'threat or use of force' and a right of self-defence 'if an armed attack occurs', the reader is not informed as to the differenceif anybetween a use of 'force' and an 'armed attack'. The ICJ, in addressing this point, has declared that there is, in its view, a difference, and framed this in terms of the gravity of the forcible act concerned. Indeed, an 'armed attack' is distinguishable by its particular 'scale and effects'. 32 This gravity threshold between the two concepts is something that a large proportion of scholarsarguably a majoritysubscribe to, 33 although there remain many problems in attaching such a threshold to the concept of an armed attack, in particular how 'scale and effects' is to be interpreted and measured and where the threshold is to be drawn, or whether it is to beindeed, can only bedetermined based upon the individual circumstances of a particular situation. 34 The rocket attack on the airport in Erbil on 15 February killed a Filipino contractor, wounded four US contractors, and wounded a US soldier. 35 Whilst this was certainly a 'forcible' actionor an 'attack' simpliciterit might be hard to depict this as being of sufficient gravity to constitute an 'armed attack' for the purposes of the ICJ's conception of such forcible acts, thus permitting the invocation of the international right of 219-223. This is arguably exemplified in the jurisprudence of the ICJ. In the Nicaragua case the Court distinguished armed attacks from 'mere frontier incidents', although did not elaborate on the distinction between the two: Nicaragua (merits) (n 7) para 105. In the Oil Platforms case the Court, on the one hand, ruled out the possibility that a single missile strike was of sufficient gravity to trigger the right of self-defence, whilst, on the other hand, could not 'exclude the possibility that the mining of a single military vessel might be sufficient': Oil Platforms (Islamic Republic of Iran v United States of America) (judgment) [2003] ICJ Rep 161, paras 64, 72. 35 While the Erbil airport attack was the only one expressly identified by the Biden administration, it also justified its invocation of the right of self-defence on the basis of other 'attacks', both those that had occurred in the past and those that were predicted to take place in the future. See section 3.2.
self-defence. 36 Indeed, it was neither of a particularly significant scalebeing a rocket attack by a non-state group against an ostensibly non-military target or had particularly serious effectswhich, from the US' perspective involved the wounding of several US nationals. It was therefore perhaps notable that the Biden administration refrained from depicting at any point that what it was responding to was an 'armed attack', but rather made reference exclusively to this as an 'attack'. 37 This may not seem particularly surprising given that the US infamously does not distinguish between 'armed attacks' and 'uses of force', and sees any forcible act as potentially giving rise to the right of self-defence. 38 Yet, while, as discussed previously other states generally appear to give more emphasis than the Biden administration did on this occasion to Article 51 and the UN Charter as the basis of the contemporary right to self-defence, they similarly often omit to speak of 'armed attacks' in the abstract or depict attacks that they are responding toor that other states have responded toas 'armed attacks'. Furthermore, while the US is often painted as an outsider in terms of its views regarding the absence of a distinction between a 'use of force' and an 'armed attack', 39 the difference between its views and the position of other states can appear marginal. Indeed, 'customary evidence … makes clear that the gravity threshold should not be set too high and that even smallscale attacks involving the use of (possibly) lethal force may trigger Article 51.' 40 As an example of this general trend, in response to the military action taken by the US in response to the assassination attempt on former president George H.W. Bush in 1993, while other states did not expressly characterise the assassination attempt as an armed attack, the invocation of Article 51 by the US on this occasion was widely accepted. 41 Ultimately, this draws into question the continued significance of the 'sina qua non' status of the requirement for an 'armed attack'distinguishable by its 'scale and effects'which was provided by the ICJ, and adds to the ambiguity of where the gravity threshold, if any, is positioned. 36 See Green (n 26) 38-41. 37 As a point of comparison, the Trump administration, while far from clear and consistent in its approach, did at least attempt to partly justify its killing of General Qassem Soleimani the previous year on the basis that, in its view, an 'armed attack' had occurred. See Letter dated 8 January 2020 from the Permanent Representative of the United States of America to the United Nations addressed to the President of the Security Council, UN Doc S/2020/20 (9 January 2020). However, it was noticeable that while it made reference to 'armed attacks' perpetrated by the state of the Islamic Republic of Iran, when discussing the actions of Qods Force-backed militia groups it referred merely to these as 'a series of attacks' and 'a series of indirect fire attacks' (emphasis added). 38  However, it was noticeable that while the Erbil attack took centre stage in the justification of the Biden administration, there was also seemingly a perceived need by the administration to increase the perceived gravity of what it was responding to. Indeed, in its report to Congress, the Biden administration was keen to stress that the non-state militia groups behind the Eribil attack were also involved in other 'recent attacks against United States and Coalition personnel in Iraq', 42 a strategy that was similarly employed by the Trump administration at the time of the strike against General Soleimani. 43 Given that the Biden administration did not reference the 'armed attack' requirement of self-defence in its justificatory discourse, it is not surprising that it did not attempt to explain how this series of nonspecified past attacks might fit within the 'armed attack' paradigm of Article 51. Scholars, however, have attempted to rationalise states' invocation of the right of self-defence in response to such a series of past attacks to fit them within the 'armed attack' paradigm of Article 51 through the socalled 'accumulation of events doctrine'. 44 That is, while one of the attacks by itself would be insufficient to meet the gravity threshold of an armed attack, the accumulation of more minor attacks potentially could.
While this approach attempts to bring the attacks and the response to them within the 'armed attack' paradigm, it might also be seen as a dilution of itgiven that 'armed attack' is seemingly expressed in Article 51 in the singular and subsequently provided with a particular gravity by the ICJor even a complete displacement of it. 45 Yet, in determining whether the gravity threshold had been met in the Nicaragua case, the ICJ was open to the possibility that incursions 'may be treated for legal purposes as amounting, singly or collectively, to an "armed attack".' 46 As such, while neither the UN Charter nor subsequent instruments of the UN provide any elaboration on the armed attack requirement, there is some support by scholars, the ICJ, and, most importantly, by states, 47 for the possibility of satisfying the requirement of an armed attack in this way. However, and importantly, the doctrine 'does not relieve the defending State of its duty to demonstrate that it has actually been the victim of the use of armed force … Merely hiding behind a general hostile environment or behind unsubstantiated assertions of territorial intrusions is certainly insufficient.' 48 In this respect, for the Biden administration not to at least identify and provide some elaboration on the nature of the other recent attacks is problematic due to the absence of any possibility for objective verification and assessment of the legitimacy of the claim in self-defence implicitly made on this basis. 49

Temporal issues
Aside from the gravity aspects of the 'if an armed attack occurs' stipulation of Article 51, there are also temporal aspects to it. Phrased as it is, Article 51 clearly provides for the invocation of the right if an armed attack is actually physically occurring or its effects are still being felt or manifesting themselves, for example in the situation of an invasion and subsequent occupation of territory, meaning that there is a clear necessity to act in self-defence to secure immediate protection to the state concerned. Yet, the situation presented in the 25 February strikes clearly did not fit within this particular paradigm, and raised several key questions regarding the contemporary right of self-defence and, importantly for the present inquiry, the specific nature of the armed attack requirement.
There are various issues to consider with the accumulation of events doctrine outlined previously, including the quantity and quality of the prior attacks necessary to cross the 'armed attack' threshold, or, at least, give rise to the right of self-defence, and the timing and proportionality of the response in self-defence. 50 There is also the questionin light of the fact that the doctrine is essentially a justification for responding to prior attacks 51of what purpose is served by the launching of military action? In particular, does a defensive necessity exist or is the action more appropriately characterised as an act of reprisal, punishment, or just being seen to be doing something, all of which would by themselves be unlawful? In this respect, the Biden administration was particularly keen on this occasion to stress the continuing nature of the attacks it was responding to. In addition to the attack on Erbil airport and other recent attacks on which the Biden administration was basing its claim to self-defence, the administration stated that the groups were 'also engaged in ongoing planning for future such attacks', 52 while in its report to the UNSC it stated that '[i]n recent weeks, U.S. and Coalition partner forces in Iraq have been the target of an escalating series of threats and attacks by such non-State militia groups … ' 53 These both highlight the plurality of recent attacks, and not just the one that took place at Erbil airport, but also the fact that further attacks were expected. In regards to the invocation of self-defence against future attacks, while the notion of 'first-strike' in pre-emptive self-defence in response to the threat of a possible temporally remote armed attackof the 'Bush doctrine' ilkis not one that has received widespread acceptance, 54 the possibility of the invocation of self-defence in response to an 'imminent' armed attack is one that has gained a notable degree of approval amongst states and scholars. 55 This is by no means universal; indeed, some, for example, remain of the view that this does violence to the requirement of the occurrence of an armed attack in Article 51, and nothing short of the physical manifestation of an armed attack will suffice. 56 Yet, others are of the view that if an actor has irrevocably committed itself to the physical commission of an armed attack, then the concept of imminence is an integral, woven-in, part of the occurrence of an armed attack. 57 Indeed, anything else would mean that in signing up to the UN Charter states were in effect signing up to a suicide pact. For its part, the ICJ has not expressed a clear position. In the Nicaragua case this was because 'reliance [was] placed by the Parties only on the right of self-defence in the case of an armed attack which ha[d] already occurred, and the issue of the lawfulness of a response to the imminent threat of armed attack ha[d] not been raised', 58 although the Court did not rule out the idea that such an imminent threat could give rise to the right of selfdefence.
However, it could be legitimately questioned whether any requirement that an armed attack be at least 'imminent' before self-defence may be resorted to is only of relevance if we are talking about a standalone future attack, or whether it is also relevant in the context of a series of attacks, both that have occurred in the past and with a belief that more are forthcoming? Some would argue that it continues to be of relevance in these latter situations, 59 so that before an action in self-defence becomes justifiable it would need to be demonstrated that a further attack in the series of attacks occurring is imminent. Others, however, have argued that in these sorts of situations imminence is something of a 'red herring'. 60 Indeed, a 'campaign' of attacks can be equated to a single continuing attack providing for a defensive necessity so long as another attack is anticipated at some point. 61 'In such cases', the argument goes, 'the question is not whether a subsequent attack is imminent but whether the campaign is still underway.' 62 And in the context of the 25 February strikes, '[s]o long as the United States reasonably concluded in good faith that the groups would continue to mount attacks, the February 25 operation fulfilled the temporal condition.' 63 Some have gone further, however, and argued that given the prior stream of attacks '[t]he attack on the Erbil airport merely confirmed that the campaign was still underway, thereby satisfying the temporal element of self-defense' with no need for any evidence suggesting that further attacks were expected. 64 Indeed, the Trump administration hovered on this line of thinking in its evolving justification for the killing of General Qassem Soleimani the year before. 65 Whether or not one is of the view that a perceived 'campaign' of attacks can, in and of itself, give rise to the right of self-defenceand reasonable actors may take different views on thisreference nonetheless continues to be often paid to the requirement for an imminent armed attack at least before self-defence becomes an option. 66 Even the Trump administration attempted, initially at least, to justify the military action against Qassem Soleimani in January 2020 on the basis that there was a threat of further imminent attack. 67 A handful of states, 68 including the US, have expressly referenced and adopted Sir Daniel Bethlehem's 'Principles', which appear to require an imminent armed attack prior to the invocation of self-defence. 69 However, while paying lip service to the need for imminence, something which 'is traditionally used as a temporal description, pointing to a specific impending attack', 70 these 'Principles' instead represent a clear attempt to shift the meaning of imminence from a concept which hinges on the temporality of an actual attack to one based on a number of contextual factors: Whether an armed attack may be regarded as 'imminent' will fall to be assessed by reference to all relevant circumstances, including (a) the nature and immediacy of the threat, (b) the probability of an attack, (c) whether the anticipated attack is part of a concerted pattern of continuing armed activity, (d) the likely scale of the attack and the injury, loss, or damage likely to result therefrom in the absence of mitigating action, and (e) the likelihood that there will be other opportunities to undertake effective action in self-defense that may be expected to cause less serious collateral injury, loss, or damage. The absence of specific evidence of where an attack will take place or of the precise nature of an attack does not preclude a conclusion that an armed attack is imminent for purposes of the exercise of a right of self-defense, provided that there is a reasonable and objective basis for concluding that an armed attack is imminent. 71 This view of 'imminent armed attack' raises several key issues. First, despite employing the language of imminent armed attack, in actual fact the focus is on whether there is a perceived necessity by the acting state to resort to military action given the circumstances. While this might not seem problematic, given that necessity is an important criterion for lawful self-defence found within customary international law, moving from a single consideration of temporality in the assessment of imminence to the consideration of a host of contextual factors not only broadens the possibilities to take military action but to such a degree that necessity arguably shifts from being a specific requirement within the jus ad bellum that exists to limit the possibilities for the resort to military action to a more general principle that in fact enables it. Indeed, depending upon the circumstances, more or less focus can be placed on each of the relevant factors to justify the taking of military action in a broad range of circumstances.
In the context of the 25 February strikes, while the Biden administration did not make any attempt to claim that a further armed attack was 'imminent' in the restrictive temporal sense, it also refrained from making any reference to the expansive contextual interpretation of imminence. The fact that the Biden administration appeared to consciously avoid framing pending attacks as in any way 'imminent' may have been a result of the tangles in which the Trump administration found itself following its depiction of imminent attacks as a key partinitially, at leastof its legal justification for the invocation of self-defence in the killing of Soleimani the previous year. 72 When pressed, the Trump administration was simply unable to defend or explain this aspect of its justification.
Yet, the Biden administration's reference to taking 'necessary' action, given the circumstances in which it was taken, represents a clear nod to necessity being seen in its enabling form rather than as a means of limiting when, and the extent to which, military action is permitted. This can arguably be seen not only in the stated purposes for the military action but also its timing and evidential basis. Two distinguishing features of this enabling form of necessity are that it provides a generous discretion to states both when to act and for what purpose. While necessity was once seen in terms of whether there was an immediate need to take action, or a need to take action in self-defence which was 'instant, overwhelming, leaving no choice of means and no moment for deliberation', 73 the expansive enabling form of necessity, as opposed to limiting a state's actions to those necessary in an emergency, rather provides states with a significant degree of discretion as to when to act. This was encapsulated in the remarks by White House Press Secretary Psaki in February 2021 that 'when threats are posed, [the President] has the right to take an action at the time and in the manner of his choosing', 74 sweeping away any remanence of urgency or lack of alternative options encapsulated in the traditional form of necessity. In addition, the Biden administration seemed to suggest that the purpose of its action was 'deterring' future attacks, 75 sending 'an unambiguous message', 76  whitehouse.gov/briefing-room/statements-releases/2021/02/26/press-gaggle-by-press-secretary-jenpsaki-and-homeland-security-advisor-and-deputy-national-security-advisor-dr-elizabeth-sherwoodrandall/. The State Department also said about the rocket attack in Erbil that 'we will respond in a way that's calculated within our own timetable and using a mix of tools at a time and place of our choosing.' See 'U.S. voices outrage over rocket attacks in Iraq but will not "lash out"', Reuters (22 February 2021) www.reuters.com/article/us-iraq-security-usa/u-s-voices-outrage-over-rocket-attacks-in-iraqbut-will-not-lash-out-idUSKBN2AM2F3. 75  perhaps most controversially of all, 'in furtherance of United States national security and foreign policy interests'. 77 Ultimately, whether these can be seen to be in any sense a form of defending the US, they further entrench the view that the Biden administration was dismissing the need for any real threat of an imminent armed attack in the temporal sense and in support of basing decisions to resort to military action upon subjective determinations that a necessity exists to do so, a necessity that is of a general and enabling nature rather than the more specific and restrictively defined concept of necessity found within the jus ad bellum. 78 Lastly, while under the Bethlehem Principleswhich, as noted previously the US has adopteda state may take into account what it perceives to be 'all relevant circumstances', there is also an apparent need to be able to demonstrate that a 'reasonable and objective basis' exists for concluding that an armed attack is 'imminent'. On the one hand, this injects a significant degree of subjectivity into decisions as to when to resort to force, while on the other, it seems to require at least a form of objective verification of these decisions. The reality is, however, that in practice evidence and intelligence upon which decisions are based will, as was the case here, be kept from scrutiny and verification. This means that any possibility for others to determine that there was indeed a 'reasonable and objective basis' for acting is, in reality, small if not completely excluded, again promoting a shift away from the 'if an armed attack occurs' requirement.

Perpetrator issues
Whilst the gravity and temporal issues are important, the issue of who may perpetrate an armed attack has also been prominent in the debates on the 'armed attack' element of self-defence. Indeed, there has been, and continues to be, extensive debate not just between international lawyers but also between states on the question of whether the right of self-defence extends to armed attacks by non-state actors, or whether it is restricted to armed attacks by states. 79 This issue can essentially be divided it into three distinct questions, all of which are relevant to the 25 February strikes. 3.3(a). Can non-state actors perpetrate armed attacks?
The first question is whether the perpetration of armed attacks is restricted to states or whether non-state actors are also able to carry out armed attacks so as to trigger the right of self-defence. In regard to the military action of 25 February, we might of course first ask the question whether the non-state groups were, in fact, 'non-state' actors? Or, in other words, were they either an organ of a state or is it possible to attribute the attacks perpetrated by them to a state? While the claim was made that the groups targeted in the military action were 'supported' by Iran, 80 and at least some of them had been incorporated into the armed forces of Iraq, 81 there was no suggestion that the groups were either an official organ of, or sufficiently controlled by, either state to be able to legally attribute the attacks to them. Indeed, there was nothing within the justifications proffered by the Biden administration to suggest that, in its view, the armed attacks perpetrated by the groups were to be legally attributed to a particular state, whether that be Iraq, Iran, or even Syria. The traditional position in legal doctrine was that armed attacks were restricted to being perpetrated by states, or those sent on their behalf, a position that was entrenched further by the ICJ. 82 Some have also made the argument that given that Article 51 is an exception to the prohibition of force contained in Article 2(4), the applicability of which is restricted to an interstate basis, then this is also the case with the armed attack requirement of Article 51 and the invocation of the right of self-defence, 83 although there is nothing expressly within the text of Article 51 restricting it in this way.
However, there has been a reconsideration of this position, particularly in light of events following the attacks of 9/11. While there are some states and scholars who continue to maintain the traditional position, 84 arguably the more prevalent position today is that the actions of non-state actors can constitute an armed attack in the absence of any state involvement, and this is something which can be seen to be generally supported in state practice. 85 In that sense, while the 25 February military action did not by itself represent a break away from the traditional state-centric position regarding the possible perpetrators of an armed attack, it certainly confirmed the dilution of it.

3.3(b). The location of the response in self-defence
Traditionally, if self-defence were to be invoked it was to take place within the territory of a stateor against manifestations of the stateto which an armed attack could be attributed. 86 As Tams notes, in the jurisprudence of the ICJ self-defence 'depended on complex, and typically fact-dependent, questions of attribution, and required responding states to show a substantial involvement of the territorial state in the very attacks of a terrorist organization against which the response was directed (referred to as "effective control" test)' and '[b]y adopting a restrictive approach to attribution the Court effectively restricted self-defence to the inter-state context.' 87 While, as discussed previously, there has been a dilution of the state-centric view regarding the possible perpetrators of an armed attack, a view particularly attributed to the ICJ, a more pressing question has been, and continues to be, where a response can lawfully take place if a state finds itself the victim of an armed attack perpetrated by non-state actors in the absence of any substantial involvement by the host state and the right of self-defence is invoked? Indeed, if we take, for the sake of argument, that non-state actors are able to perpetrate armed attacks, the fact that any response in self-defence will invariably take place in the territory of a third state raises the question of how the victim state is able to justify the launching of military action upon the territory of that state given that this would implicate the legal prohibition of inter-state force. In this respect, the 25 February strike in response to the attacks by the non-state groups took place not in Iraq or Iran, but on Syrian territory. Given that the rights of the state upon which military action is takenin particular, its sovereignty and territorial integrityare brought into question, some have made the argument that armed attacks perpetrated by non-state actors must still be attributable to the territorial state. 88 In this sense, the claim is made that we have witnessed in state practice a lowering of the traditional standard for the attribution of an armed attack to a statethat is, that the state effectively controlled or sent the non-state actors in the perpetration of the armed attack 89to one of the state 'aiding and abetting' the non-state actors in them carrying out the armed attacks, 90 the state being 'complicit' in the attacks, 91 the state having harboured the non-state actors on its territory, 92 or that the state has asserted a level of control over them of an 'overall', rather than an 'effective', nature. 93 Indeed, following the attacks of 11 September 2001, the US itself infamously made the claim that it would make 'no distinction between the terrorists who committed these acts and those who harbor them.' 94 Overall, these attempts at lowering the threshold for attribution maintain the focus on the need to attribute responsibility for an armed attack to a state so as to justify an intervention into that state's territory, albeit representing a dilution of the required threshold.
However, an alternative view shifts the focus away from the armed attack requirement entirely and instead places the justifying element firmly within the realm of the customary criterion of necessity. 95 In other words, regardless as to whether the armed attacks of the non-state actors can be attributed to the host state under the traditional standard of effective control, if there is a perceived necessity to take action in self-defence then this justifies an incursion into the host state's territory for the purposes of targeting the non-state actor perpetrators of the armed attack. The US has expressed this position in terms of the 'unable or unwilling' doctrine, that is, while the armed attack may not be attributable to the host state, if the state is not able to take the necessary action to prevent the attacks or is unwilling to do so then a military incursion in self-defence can be justified on the basis of it being legally necessary. 96 Although the US has relied upon the 'unable or unwilling' doctrine previously, 97 the doctrine was again relied upon in the context of the 25 February strikes in that the Biden administration claimed that the use of force in self-defence is permissible when, 'as is the case here, the government of the state where the threat is located [Syria] is unwilling or unable to prevent the use of its territory by non-state militia groups responsible for such attacks.' 98 Although this doctrine is far from broadly accepted, it is arguable that it has at least a degree of support amongst other states. 99 In this respect, the US is not exactly an outlier in adopting this view. Yet, the significance of this position for the purposes of the present inquiry is that it represents not just a dilution of the relevance of the armed attack requirement for the purposes of determining the location of any action in self-defence but rather a positive shift away from it. Indeed, the American view again prioritises the principle of necessity and views it not as one which places a restriction upon the breadth and scope of any action taken, which is the function this principle is traditionally perceived as providing, but rather as a principle which acts as an enabler of military action upon the territory of host states, in this case Syria. This conclusion is supported by the fact that evidence to support a claim that a host state is either 'unable' or 'unwilling' to take the required action is, as was the case here, rarely advanced by the acting state for objective scrutiny, broadening the 'freedom' which states have in invoking the right of self-defence.

3.3(c). The specific target of the response
In invoking its right of self-defence, the US was noticeably vague in its various justifications as to whom, exactly, was the target of its military action, and what they might be responsible for. For example, it claimed on the one hand that it had carried out a 'targeted military strike against infrastructure in eastern Syria used by Iran-supported non-state militia groups' who 'were involved in recent attacks against United States and Coalition personnel in Iraq, including the February 15, 2021', 100 but who were not necessarily the direct perpetrators of the attacks. Furthermore, it claimed that 'the strikes destroyed multiple facilities located at a border control point used by a number of Iranian-backed militant groups, including Kait'ib Hezbollah and Kait'ib Sayyid al-Shuhada', 101 but did not expressly mention the group Awliya al-Dam which had claimed responsibility for the rocket attack in Erbil. Yet, in other justifications the US was not even clear as to which groups were using the targeted infrastructure. In its letter to the UNSC, for example, after having stated that the US had 'undertaken a targeted military strike in eastern Syria against infrastructure used by Iran-supported non-State Militia groups … ' it was claimed that '[i]n recent weeks, U.S. and Coalition partner forces in Iraq have been the target of an escalating series of threats and attacks by such non-State militia groups … '. 102 It is questionable whether vague identifications of the target of the strikes are sufficient to justify the invocation of the right of self-defence and the launching of military action, or whether the target of any military action in self-defence would need to be expressly identified as either the perpetrator of an armed attack or infrastructure belonging to it, or alternatively actors directly controlling the perpetrator. Indeed, the traditional position would suggest that such specificity is required, thereby keeping the decision regarding the targets of any action in self-defence firmly linked to those directly involved in the armed attack.
There have been previous attempts at broadening the range of permissible targets of military action taken in self-defence. As a prominent example of this, the Obama administration included 'associated forces' of al-Qaida as permissible targets under the 2001 Authorization for Use of Military Force. 103 Furthermore, in his 'Principles', Daniel Bethlehem stated that in the context of a 'concerted pattern of continuing armed activity' it was permissible to invoke self-defence against individuals and groups 'acting in concert', 104 that is, 'those planning, threatening, and perpetrating armed attacks and those providing material support essential to those attacks, such that they can be said to be taking a direct part in those attacks'. 105 This position was subject to criticism, 106 as not only does it move away from the traditional position being that self-defence can only be invoked against the perpetrators of an armed attack, or, as the case may be, those directly controlling them, but also simply on the basis of the wide breadth of actors who may ultimately find themselves the subject of a response in self-defence under it.
It was, however, notable that even Bethlehem's stated threshold for determining that an entity or group is 'acting in concert' was not one which the Biden administration made any real attempt to meet in respect to the 25 February action in self-defence. Indeed, the accusation that those targeted in the strikes were 'involved' in attacks against the US is somewhat weaker than Bethlehem's standard of 'taking a direct part in those attacks'. Yet, apart from the vague reference to the targeted infrastructure being used by groups 'involved' in attacks against the US, the only trait supposedly shared by those groups who had perpetrated the attacks against the US and the groups who owned the infrastructure targeted was that they were 'Iran-backed'. Such a relaxed standard for the targeting of specific groups represents not just a dilution of the armed attack requirement, but arguably a complete severance from it andas with the justification for the location of action in self-defencean acceptance of the permissibility of basing such targeting decisions on a general and enabling interpretation of the customary criterion of necessity.

Conclusion
This article has argued that the Biden administration, and the US in general, has attempted to lean more heavily on the customary source of the right of self-defence over the treaty source of the right and has done so due to the perceived greater flexibility provided by this source, in particular adopting a broad and enabling interpretation of the customary requirement of necessity. The question remains whether this strategy has been, or could be, successful in providing legal cover for such military operations and, ultimately, in altering the right of self-defence?
In general, in situations such as this, where state behaviour is regulated both through a treaty and by customary international law, even if we are able to identify changes in the customary source this would not necessarily have an impact on the rights and obligations contained in the treaty source. The two sources exist independently and possess separate and different processes of change and interpretation. And even if there was the possibility for inter-source impact, if we are to accept that the prohibition of force as it exists in both its treaty and customary forms is a jus cogens norm, then any asserted changes that might be seen as broadening the possibilities for the use of force would face a particularly high threshold in impacting upon the lex lata.
However, Akande and Johnston have argued that Article 51's express affirmation of the 'inherent right' of self-defence means that any changes to the customary right are directly accommodated within the treaty right 'since the Charter rule on self-defence is given content by the customary law on self-defence.' 107 This is significant in that it maintains a level of coherence and consistency between the two sources of the right of self-defence, but also because any changes to the breadth of the right would not require any reconsideration of the breadth and scope of the prohibition of force contained within either customary international law or in Article 2(4). 108 Yet, this overlooks the fact that the 'inherent right' of self-defence is expressly qualified by the stipulation contained within Article 51 of the right only existing 'if an armed attack occurs'. While the ICJ proclaimed in 1986 that this constitutes a 'sina qua non' element of the right in both sources, it is an indelible aspect of the treaty sourceunlike its customary counterpart meaning that any attempts at modifying the right of self-defence are limited by its existence. There have been no attempts to formally modify the 'if an armed attack occurs' requirement in Article 51. Nonetheless, the phrasing of the requirement is broad enough to permit some leeway in its interpretation. As demonstrated in this article, while the Charter is silent on the nature of an armed attack, and how it is to be distinguished from a use of force, the US has been particularly keen to downplay the existence of any gravity threshold separating them, and other states have also not necessarily been adherents to a threshold distinction as set out by the ICJ. In addition, while the Charter is silent on the issue of the nature of the perpetrators of an armed attack, the US and many other states have been more open to non-state actors as perpetrators than the ICJ has been. Although in neither case is it possible to claim that the proposed interpretations have found the agreement of all parties to the UN Charter, 109 there is at least a discernible level of ambiguity for the US to operate within.
Yet, any perceptible agreement quickly dissipates when it comes to the question of where, and under what circumstances, a response in selfdefence may be undertaken in the context of attacks perpetrated by nonstate actors. Furthermore, it is an issue on which the views of the US become difficult to square with a textual reading of the 'if an armed attack occurs' requirement. First, the ICJ, along with certain states and scholars, is of the view that self-defence can only be undertaken in the territory of states to which the armed attack perpetrated by the non-state actors can be attributed, thereby maintaining a direct link between the armed attack and the location of any response taken. Others, including the US amongst a handful of other states, have sought to downplay this direct link to an armed attack and rather see decisions regarding the location of an action in self-defence as determined through a broad and enabling form of the criterion of jus ad bellum necessity. This approach not only removes any need for attribution of the armed attack to the state in which military action is taken but also, and perhaps more controversially, seemingly removes the need to provide any evidence of the existence of such necessity. Second, from the text of the Charter it is clear that an armed attack actually has to occur. While the ICJ and some states and scholars have been willing to go as far as to view an imminent threat of attack as within the realms of the 'occurrence' of an armed attack, the US in particular has attempted to push the limits beyond what the text of the Charter is able to legitimately accommodate. In particular, while continuing to often pay lip service to the requirement of an 'imminent' armed attack, it has instead attempted to displace the need for any armed attack at alleven of an imminent kindwith, again, a contextual and enabling form of necessity and thus providing significant flexibility to the timing of the military action, its purpose, and the evidence upon which it is based.
Ultimately, while the ICJ's 'sine qua non' interpretation of the requirement of 'armed attack' has by no means reached the point of vanishing, it has clearly been subject to a significant and sustained attack on multiple fronts, with the 25 February military action and ensuing legal justification proving to be the one of the most recent.

Disclosure statement
No potential conflict of interest was reported by the author(s).