R2P, Terrorism, and the Protection of Civilians – ‘Are All Humans Human? Or Are Some More Human than Others?’

Terrorism is often cited in the justifications of state perpetrators of mass atrocities. The reality behind these claims runs the gamut from thin pretext to genuine security threats. Irrespective of this reality, the discursive abuse of counter-terrorism to perpetrate atrocities is a key challenge to the Responsibility to Protect. Perpetrators employing the language of counter-terrorism to justify their actions, in an attempt to pre-empt objections or interventions, disincentivises external action by actors unwilling to incur the risk that they may inadvertently protect terrorists. This risks limiting the application and successful operationalisation of the Responsibility to Protect to relatively simple or ideal cases. This article provides a comparative analysis of two crises often described as successes or failures for R2P, Kenya (2007–08) and Sri Lanka (2009) respectively, to demonstrate this challenge for the operationalisation of R2P even in cases with complex conflict dynamics.


Introduction
The Responsibility to Protect (R2P) and counter-terrorism are thematic norms that have emerged concurrently in the early years of the twenty-first century.1 R2P sets out a normative framework to prevent mass atrocities, emphasising sovereign responsibility. It established the international community's responsibility to prevent, react, and rebuild -interventions scaling in severity with the use of force as a last resort.2 Counter-terrorism came to dominate the lives of much of the world following the 9/11 attacks on the United States and subsequent War on Terror. Counter-terrorism prioritises the degrading and destruction of terrorist groups and networks to prevent future attacks, sometimes at the expense of civil liberties. These two norms share a focus on state responsibility with transnational coordination and cooperation. While both effectively focus on protection of civilians, counter-terrorism has complicated the operationalisation of R2P. This article examines the challenge that has, and will likely continue to, hamper the operationalisation of R2P. That is, how does the international community respond to situations, such as those in Sri Lanka, Syria, China, Myanmar or Ethiopia, where civilians are at risk of mass atrocities, but perpetrators employ the language of counter-terrorism to justify their actions and pre-empt objections or interventions? While in some cases this language is employed to create a securitised fig leaf of legitimacy for repression, the more bedevilling problem is when there are terrorist groups embedded within vulnerable civilian populations. Perpetrators are too often able to justify atrocities under the guise of counter-terrorism. This in turn disincentivises external action by actors unwilling to incur the risk that they may inadvertently protect terrorists. The mixing of mass atrocities with legitimate counter-terrorism also adds complexity as to whether and how R2P applies. Debates over legitimacy, sovereignty, and proportionality can slow or stall the authorisation and operationalisation of external interventions. In effect, employing the terrorism card has become a licence to commit atrocities. This research does not seek to discredit or undermine the value and importance of R2P in advancing the prevention of mass atrocities; however, this problem speaks to the limits of R2P's universality and is reflective of Romeo Dallaire's question: 'are all humans human? Or are some more human than others?'3 I argue that the oversimplification of conflict dynamics in mass atrocity cases, and the framing that is often necessary to mobilise sufficient political will, inhibits the application and operationalisation of R2P in cases where bad actors are embedded within populations. The inability to parse the legitimate use of force in counter-terror or counter-insurgency operations and deliberate campaigns of atrocities allows international actors to obfuscate and delay action, or simply do nothing. In retrospect, these failures then manifest in claims of partiality or weakness.
It is imperative to interrogate the potential weaknesses and failures to protect for two reasons. First, these failures often hold vital lessons for future attempts. Second, they drive discussion and debate that is critical to the normative power of R2P. This much is made abundantly clear by the overt focus on the failures of the Libyan intervention. Yet the intervention in Ivory Coast was contemporaneous to Libya, and although not perfect, was far more successful. Often, the debate over R2P is driven far more by discussions of Libya, Syria, and Myanmar than it is Kenya, Guinea, or Kyrgyzstan. Advancing human protection via the R2P requires engaging with critiques from actors of both good and bad faith, with the latter eager to hobble progress where their self-interest is threatened.
This article offers a comparative case analysis to demonstrate the problem of how narratives of counter-terrorism affect the operationalisation of R2P across temporally comparable cases of international responses to mass atrocities. I compare the case of Kenya following the 2007 presidential election with Sri Lanka in 2009 during the final months of the civil war. Both (otherwise significantly different) cases have been discussed as successes or failures of R2P. I argue that the issue of terrorism wrapped in geopolitical imperatives in both cases augmented the international incentives that likely had an effect on the response to each respective crisis. This analysis expands our understanding of the variation of outcomes in these cases, comparing the rapid and relatively effective response to Kenya's post-election violence with the manifest failure to protect Tamil civilians in areas under the control of the Liberation Tigers of Tamil Eelam (ltte). While there are many dissimilarities between these cases, their temporal similarity allows some control for the international community's attitude to R2P. nevertheless used to justify disproportionate repression.4 Most prominent among external justifications for intervention is the US 'War on Terror' , waged predominantly under the second Bush administration and to varying degrees thereafter. The United States and state actors often used the label of terrorism as an instrument to legitimise overt and covert intervention. US support for the Contras in Nicaragua was premised on their fight against 'terrorist' Sandinistas -the act of labelling a critical element of the broader strategy.5 Before and after the onset of the war on terror, Russia used the language of terrorism to describe the actions of Chechen rebels, sometimes with justification, implicitly justifying the atrocities of their counter-insurgency wars.6 Terrorism has also been shown to reduce the likelihood of external intervention in favour of those actors who use terror attacks as a strategy.7 Whether the likelihood of intervention influences the use or framing of opposition actors as terrorists is uncertain; it does, however, fit a logic that would seek to dissuade support for opposition groups. There is no accepted definition of terrorism, for the purposes of this article I adopt that of the Report of the High-Level Panel on Threats, Challenges and Change: any action, in addition to actions already specified by the existing conventions on aspects of terrorism, the Geneva Conventions and Security Council resolution 1566 (2004), that is intended to cause death or serious bodily harm to civilians or non-combatants, when the purpose of such an act, by its nature or context, is to intimidate a population, or to compel a Government or an international organization to do or to abstain from doing any act.8 How terrorism is used and abused as a label clearly affects the likelihood and nature of R2P being applied. For the most part, this is often extremely challenging in cases where some truth exists around the use of terrorism. Real and significant terror threats do exist in particular cases and states hold the legitimate authority to address those threats. The problem arises where the methods and strategies of counter-terrorism are instead applied across an entire group, as is common in counter-insurgency. The process in which counter-terrorism campaigns can morph into mass atrocities is similar, and at times indistinguishable, from the process Ernesto Verdeja describes of counter-insurgency, Counterinsurgency campaigns … may become genocidal if state forces move from combating rebel forces to targeting for destruction the civilian groups that ostensibly support the rebellion. These campaigns may begin by killing individual political opponents, forced relocations, and widespread and indiscriminate killings of their supporters; and eventually coalesce into policies of systematic and deliberate destruction of the group.9 There are cases where minor threats are magnified as part of an effort to justify atrocities against a larger ethnic group. In 2017, the Tatmadaw campaign that killed at least 10,000 Rohingya civilians and expelled 700,000 into Bangladesh was framed and justified as a response to attacks by the Arakan Rakhine Salvation Army (arsa).10 While arsa committed terror attacks, they were a relatively weak and small group whose violence was a thin justification for a disproportionate military campaign that deliberately killed and targeted civilians on the basis of their ethnicity and religion.11 Terrorism has been invoked in justifications of repression and systematic destruction of Uighur culture by Chinese authorities in Xinxiang.12 In Ethiopia in 2020 and 2021, descriptions of Tigrayans as terrorists were commonly used in justifications of atrocities committed by Ethiopian and Eritrean forces.13 Terrorism framing and strategies that can be justified as counter-terrorism easily meld into the toxification of target groups as security necessitates action, under what Jonathan Leader Maynard describes as 'atrocity-justifying ideologies' .14 Such justifications are often thin, yet the responsibility to protect civilians remains even in cases where terrorism is ongoing and atrocities occur amidst legitimate military or policing operations. Yet there is insufficient discussion of how to sufficiently manage this tension. Alan Kuperman has clearly demonstrated the vexing issue of terrorism and intervention where an Islamic terrorist uprising coincided with other moderate demonstrations in Libya.15 His argument illustrates why there are risks in intervening where conflict dynamics are complex and there are risks of adverse consequences. Indeed, it is only in the Libyan case where such concerns were largely put aside by intervening nato forces. However, as Alex Bellamy and Stephen McLoughlin describe it, the confluence of international dynamics that allowed the use of force in Libya was extremely rare and is unlikely to manifest on regular occasions.16 The post-intervention fall-out in Libya has been at the core of much of the criticism by those sceptical of R2P.

R2P's Dilemma: Political Will vs. Conflict Complexity
The problem of the protection of civilians in cases involving bad actors, terrorists in particular, is perhaps driven by the original narratives that informed efforts to protect them. The adoption of the responsibility to protect was a monumental achievement for its proponents. The 2005 United Nations World Summit adopted the R2P with the following tenets: states have the responsibility to protect their people; the international community must engage to empower states in their responsibility; and should a state 'manifestly fail' in its responsibility, the international community has the responsibility to protect civilians.17 While both sovereign and international responsibility has a long history, this was a major development.18 Yet the need for R2P was overwhelmingly framed in light of two major examples of genocide and mass atrocities that necessitated such a radical shift in the international system, the genocides in Rwanda and Bosnia. In both cases there were clear, unambiguous examples of genocide that required an immediate response, yet that response was inadequate or not forthcoming. The failure to respond was key to the debates regarding sovereignty and intervention in the galvanising of the international community to support the concept at the UN World Summit. The framing was arguably essential to drive the political will required for R2P, emphasising the innocence of victims as a better case for action -especially where that action is as significant as the last resort mechanism of the use of force. This framing extends to domestic justifications for interventions. But what little support may be generated for supporting an intervention will certainly be hampered by recognising any inherent complexity and difficulty in application. A Manichean framing is needed, and in some cases does come closer to reflecting a reality. This framing, however, creates challenges for cases where ambiguity exists. Where bad actors are embedded within, or form part of a target group, it becomes simpler to do nothing to avoid the risk of being seen to or actually providing material support to terror groups or insurgencies in the name of humanitarian action. Rather, the framing for conflict and response falls into counter-insurgency or counter-terrorism; where multiple frames could coexist there is confusion and space for obfuscation, disincentivising effective action. The stakes are high for intervention, as Bellamy and McLoughlin contend, and the cost of inaction relatively meagre. Few votes are clearly to be won intervening in foreign humanitarian crises -and even where successful, the political costs of failure or casualties are far clearer.19 In such a context the incidental support of terror groups would present a serious political risk, at least for democratic interveners. As such the framing of events as terrorism and counter-terrorism, rather than some other form of political violence, is perhaps more important than the violence itself. An honest cogent assessment of the complex conflict dynamics and the real potential cost and actual consequences are often missing. Where ambiguity exists, the likelihood of inaction, late or insufficient action increases. We are too often left with perfect or ideal cases for R2P -that is, clear cases of mass atrocities with simple or simplified conflict dynamics or narratives, where external interests incentivise action, and the risks of adverse consequences are low. This is a fundamental problem for the progress of the R2P given cases involving a high risk of mass atrocities that include ambiguity and bad actors on both sides of a conflict. Casting groups only as perpetrator and target, evil and innocent, may suit short-term appeals to mobilise political will but are often thin narratives, easily deconstructed. Conflicts are generally complex, and the existence of bad actors within a target population should not reduce the expectations of protection afforded to vulnerable civilians.
To illustrate this issue I draw upon two case studies, a year apart, roughly within the same international context: Kenya and the post-election crisis of 2007-2008 and the final stages of the Sri Lankan civil war between January and May of 2009. The starkly different responses -swift and effective in Kenya versus the manifest failure to protect vulnerable Tamil civilians in Sri Lanka -lay bare the problem presented in this article. Kenya was an ideal case for R2P in which compromise between the conflicting parties remained possible and the international community's interests clearly incentivised the forceful response. This compared to the broadly insufficient actions of the international community in Sri Lanka. In that case, the prospect of protecting civilians may have also entailed the possibility of incidentally supporting a terror group and acted against counter-terrorism. This, seemingly, dissuaded strong action from Western actors. Sri Lanka rejected attempts to intervene on the grounds that by fighting terrorism they were protecting civilians, invoking Pillar One of R2P, which enshrines responsibility to protect civilians with the sovereign state.20 Responsibility was deferred to India, who in turn were hamstrung by geopolitical concerns, which combined with their hatred of the ltte informed a permissive attitude to the slaughter of civilians. Ambiguity drove ambivalence, and even after the conflict ended, action to hold perpetrators accountable has been largely ineffective or absent. In the debates regarding R2P, Kenya is often held up as a success story; Sri Lanka, on the other hand, has been often neglected or forgotten for the ambiguity and multiple frames that applied to the case. The effect of terrorism and counter-terrorism does not provide a causal explanation for the variation of outcomes across these two cases. Nevertheless, I argue terrorism framing is an important factor that can aid explanation and demonstrates a broader problem with the operationalisation of R2P.

Kenya
In the wake of disputed elections in December 2007, at least 1,133 people were killed, thousands wounded, and 350,000 displaced.21 The violence, predicated on a conflict between ethno-nationalist political parties, was quelled by a mediation team of eminent African personalities who brokered a political solution for a power-sharing government.
Mass atrocities do not occur in a vacuum. The 2007 Kenyan presidential election results sparked a latent conflict with a significant risk of mass atrocities. Yet the broad support for peace from Kenyan society, the lack of involvement from Kenya's armed forces, and the absence of difficult actors, like terror groups, made it far easier to resolve than many other crises to which the R2P label is attached.22 Resolution was also made easier by virtue of geopolitical interests that incentivised Kenyan stability.
Despite public commitments to peaceful campaigning, the 2007 election campaign affirmed divisive tribal politics. Fought between Raila Odinga's Orange Democratic Movement (odm) and President Mwai Kibaki's People's National Union (pnu), both sought to generate fear of the other to provoke mass turnout. Elections 'heightened expectations of Kenyans by promising to redress these issues overnight' , thereby raising the stakes of electoral victory.23 The odm framed the Kibaki administration and ethnic Kikuyus as one and the same, using decolonisation-era populism through its 'Kenya against the Kikuyu' rhetoric to invent a class from an ethnic group.24 Their campaign targeted prominent Kikuyus 'in banking, government, trade, outmigration, education, and commercial farming' .25 Hate speech was broadcast in which 'Kikuyus were identified … as madoadoa [a Kiswahili word for stains or spots] translated in this specific context as spotted hyenas to be violently expelled from the Rift Valley' .26 Such language is a sign of toxification, a process using language to construct a group as a threat which is synonymous with impending anti-Kikuyu leaflets were circulated in the Rift Valley … it was alleged that if the odm won the elections it planned to carry out genocide against no less than one million Kikuyu in order to change Kenya's electoral demography permanently to ensure that the odm ruled in perpetuity.28 The intended effect of this propaganda was to strike fear in Kikuyu voters ahead of the vote. Scapegoating and fear-mongering on both sides built a self-perpetuating fear of the outcome of the election. the church, using mattresses to block escape, and set it alight killing 28 and injuring 54.34 Kiambaa garnered significant media coverage both inside Kenya and internationally.

The Election and Post-election Violence
Kenyan police also bear responsibility -405 deaths were attributed to police.35 While some occurred amidst riots, Human Rights Watch found that 'the majority of people shot in Kisumu died in the residential slum areas far from the shops in the commercial center' where demonstrations took place.36 Retaliatory violence began as displaced Kikuyus arrived in Central Province with Kikuyu militias attacking Luos in Nakuru and Naivasha.37 In one incident, Kikuyu youths blocked the highway to Nakuru and demanded to see the identity cards of anyone stopped. While Kikuyus passed unharmed, Luo or Kalenjin were murdered.38 Justifications for atrocity were also seemingly proffered at lower levels of Kikuyu leadership as Raila Odinga and Luos were branded as terrorists and devil worshippers planning a revolution.39 While it may not be possible to know what would have occurred in Kenya without internal and external responses to de-escalate the crisis, there can be relative confidence that Kenya was at a significant risk of further, more extensive atrocities.

4.2
Kenya's Strategic Importance As violence intensified, the international community responded with speed and coordination. This may have been due to the economic and strategic importance of Kenya in trade and counter-terrorism. The unified response of the international community was led by former UN Secretary-General Kofi Annan and was underpinned by an implicit commitment to the responsibility to protect. As Annan later reflected, I saw the crisis in the R2P prism with a Kenyan government unable to contain the situation or protect its people … I knew if the international community did not intervene things would go hopelessly wrong … Kenya is a successful example of R2P at work.40 We are very concerned about the violence which is continuing to convulse Kenya … We fear lest these excesses plunge Kenya into a deadly conflict of an ethnic character. In the name of the Responsibility to Protect, it is urgent to help the people of Kenya … The United Nations Security Council must take up this question and act.52 By combining a multitude of pressures from Kenyan civil society and the international community, Annan presented the parties with no choice but to compromise or cease to enjoy the benefits of power. After nearly a month of intense mediation the parties signed the Agreement on the Principles of Partnership of the Coalition Government, establishing a power-sharing arrangement that brought the crisis to an end.
Kenya has, for R2P advocates, represented a prominent success story. The violence, as clearly outlined, was an R2P situation. Conscience-shocking acts and the clear potential for escalation focused the minds of the international community who 'did not simply defer to the shibboleth of national sovereignty, but acted' .53 The case of Kenya does not, in itself, present a complete success of R2P, however. The core tenet of R2P is prevention, and the failure to heed the warning signs of Kenya represents a failure to prevent.54 Yet the geopolitical imperatives, paired with a crisis in its early stages, represented an ideal case for the R2P. There was broad support for peace from the grassroots through to the international community, no involvement of armed forces, no significant divisions at the international level, and political leaders willing to de-escalate and compromise.

Sri Lanka
The end of the  63 This major strategic victory propelled the rapid advance of the sla as the ltte, and hundreds of thousands of Tamil civilians compelled to follow them, retreated to their last stronghold of Mullaitivu. At this point the ltte had been all but defeated, protecting itself by holding hostage up to 330,000 Tamil civilians that remained with them.
It must be recognised that the ltte was a particularly egregious non-state actor. The ltte's use of suicide bombers, chemical weapons attacks, civilian targeting, assassinations, and forced conscription of child soldiers place them in a company of truly repugnant groups. 64 In the final phase of the civil war, the ltte turned civilians into human shields and shot many of those caught attempting to escape the rebel-held zone. 65 The crimes of the ltte, however, do not justify the means used by the sla to bring about their end. There is no justification for the deliberate slaughter of tens of thousands of civilians. There were multiple instances in the final phase of the conflict between January and May 2009 where options of negotiation provided viable alternatives for ending the conflict. The Sri Lankan army persisted nevertheless, tasked with achieving their singular and deadly task: the complete and final destruction of the ltte. The sla designated areas in the Wanni as 'No Fire Zones' (nfz s), and directed civilians to gather there for protection. Hundreds of thousands of civilians moved into these zones, alongside humanitarian organisations and UN staff, under the assumption that they would be protected from the conflict. The sla then proceeded to specifically target the nfz s with a bombardment of mortars and other heavy artillery. As the UN Panel of Experts report describes, From as early as 6 February 2009, the sla continuously shelled within the area that became the second nfz, from all directions, including land, air and sea. It is estimated that there were between 300,000 and 330,000 civilians in that small area. The sla assault employed aerial bombardment, long-range artillery, howitzers and mbrl s [multi-barrel rocket launchers] as well as small mortars, rpg s [rocket-propelled grenades] and small arms fire, some of it fired from a close range. mbrl s are unguided missile systems designed to shell large areas, but if used in densely populated areas, are indiscriminate in their effect and likely to cause large numbers of casualties.66 The sla were repeatedly informed, by UN agencies, that they were bombing civilians, including schools and hospitals. As Tameshnie Deane writes, 'In line with its military counterinsurgency strategy, the government forces continued to repeatedly bomb and shell the increasingly constricted populated areas, even during the periods of "safe-passage", and in most cases, with the knowledge that they were attacking civilians' .67 These actions meet the criteria for targeted mass killing, as civilians of a specific ethnic and religious group were deliberately and disproportionately targeted in order to expel and affect the political activity of the group.68 Between January and May 2009, the sla slowly gained control over more territory, altering the location and size of nfz s multiple times and continuing to target civilians. The sla also restricted humanitarian aid and bombed convoys to deliberately limit the capacity of the remaining 66 ibid., p. 28. 67 Tameshnie  adapted to a post-9/11 world. As Shannon Zimmerman articulates, from the 9/11 attacks onwards, a consistent narrative (often supported by ltte attacks) framed the eventual response as one of counter-terrorism.77 Gerrit Kurtz and Madhan Mohan Jaganathan clearly illustrate the way in which Sri Lanka was able to successfully adopt a counter-terrorism narrative as a shield for its own crimes.78 For at least the period prior to the final phase of the civil war they received support from the US and United Kingdom, including explicit approval of their resolve to militarily defeat the ltte.79 They further argue, The international debate was considerably biased in favour of the Sri Lankan government. The ltte's track record of violence against civilians, forced recruitment of children, and suicide attacks seem to have legitimised the government's counterinsurgency operation in the eyes of most governments.80 The global war on terror augmented the narrative, and, in turn, the boundaries of acceptable conduct to include the deliberate killing of civilians in the name of combating terrorism. For this strategy to be successful, however, there needed to be sufficient support for its use internationally. It is true that China and Russia, along with others, stymied effective action in the UN Security Council. It was, however, the inaction of India, Sri Lanka's closest neighbour who has often dominated its politics, that best encapsulates the challenge of responding to a case like Sri Lanka. Intervention without Indian support was unviable. The role of India, as the main regional power, in driving the insufficient, often ambivalent response to the crisis is perhaps best summed up by their opposition to US-led efforts to evacuate civilians from ltte held areas. In March 2009, the proposal from the US to evacuate civilians using 'military assets from the US Pacific Command, based in Hawaii failed because of resistance from the ltte and India's preference for a UN effort (the Indian opposition reportedly convinced the Sri Lankan government to reject the offer as well)' .81 Eelam War iv coincided with increased geostrategic competition for influence in Sri Lanka between India and China. As Destradi contends, 'the competition between India and China induced New Delhi to support Colombo's military efforts against the ltte in order to avoid losing its influence on Sri Lanka to an extra-regional actor' .82 While the sla purchased arms from a variety of sources, the bulk during the period of interest came from China. Throughout the fourth Eelam War, Chinese assistance was believed to be worth US $1 billion annually.83 In March 2007, Sri Lanka and China signed eight agreements on closer economic cooperation, investment, and other issues.84 The joint communique explicitly set the stage for Chinese support for Sri Lanka's counter-insurgency. It affirmed both states were 'resolved to fight tirelessly against the three evil forces of terrorism, separatism and extremism and will step up consultation and coordinating on regional and international counter-terrorism action' .85 This signal of support for Sri Lanka's civil war strategy coincided with the significant increase in military support from China to Sri Lanka. During the previous three Eelam wars, India had either been on the side of the Tamil rebels or had taken a 'hands off' approach. It was only with its acquiescence that such an extreme military solution was seemingly attempted.86 India's inability or unwillingness to intervene to prevent or stop the atrocities speaks to the problem of ambiguity. In the absence of a clear narrative around which political will can coalesce, geopolitical interests and realpolitik will fill the vacuum.
Final defeat of the ltte occurred after the most significant Western powers had designated the ltte as a terrorist organisation. While the end of the civil war came at a period where the 'war on terror' rhetoric of the Bush administration in the United States had been swept away by the incoming Obama administration, there was little sympathy for the ltte in the West. Eelam War iv had begun in 2006 and was significantly tied to a counter-terrorism narrative. Many Western powers were otherwise preoccupied with the fallout from the global financial crisis and the war in Gaza. The Obama administration made a series of efforts from naming and shaming through to threats of prosecution and failed attempts to evacuate civilians. With China, India and Russia in their corner, however, the administration appeared unwilling to spend significant political capital where it held no real interests. As such, the Government of Sri Lanka could effectively ignore any appeals for restraint from the West.87 At the UN Security Council, the crisis in Sri Lanka never made it onto the agenda, as China and Russia both blocked resolutions on the crisis.88 With the obvious political cost of being seen to be protecting a designated terrorist organisation, many Western powers were hesitant to use too much political capital. The failure to act is reflected in the statement made on 15 May 2009 by Francis Deng, then UN Special Adviser to the Secretary-General on the Prevention of Genocide, calling on both sides of the conflict to show respect for international humanitarian law and seek a peaceful solution.89 Though an uncontentious statement, its relative ambiguity in relation to the Sri Lankan government's actions while specifically noting the ltte's horrific use of human shields is revealing. It reflects a permissive attitude whereby the international community had decided to look the other way until it was all over. The framing of the conflict as a counter-terrorism war allowed the Sri Lankan Army to commit mass atrocities with impunity. Despite the atrocities of the Sri Lankan Army along with those of the ltte being one of the first true tests of the R2P, the international community manifestly failed in their obligation. This can, at least in part, be explained by the ambiguity of the conflict dynamics that blended counter-insurgency, counter-terrorism, and mass atrocities into a mire that disincentivised the operationalisation of R2P to effectively protect civilians.

Discussion
These two cases illustrate the way terrorism and counter-terrorism narratives can augment and complicate the application and operationalisation of the R2P. Terrorism was an implicit factor that shaped external responses to imminent and ongoing mass atrocities in both cases. The effect was clear, disincentivising the effective operationalisation of R2P in Sri Lanka while incentivising its rapid and successful use in Kenya. In Sri Lanka, the government was able to co-opt the language of the R2P, wrapped in war on terror era counter-terrorism doctrine to implement their final solution to the Tamil Tiger problem and with it, send a message to any other groups that may seek to challenge the Sinhala nationalist regime. By using the language of counter-terrorism over many years, but especially post-9/11, there was a clear attempt to set the agenda, initially during peace negotiations and later in the conduct of Eelam War iv. In a global atmosphere that was still overwhelmingly willing to accept extreme means to justify particular ends, what was portrayed as strong counter-terrorism strategy was largely accepted. Indeed, the failure of peace process, which eventually led to the failure of R2P was in part significantly undermined by a lack of support from the United States for the Norwegian-led mediation. This, as Nissen argues, was partly due to the adopted framing of the ltte as a terror group, and later the acceptance of the Rajapaksa regime's military solution to the conflict.90 This co-option to their own ends appears to have been relatively easy, as Richmond and Tellidis argue, 'orthodox terrorism approaches undermined liberal peacebuilding and evacuated local alterity in conformity with the nationalist Sinhala strategy of the post-independence era' .91 That the ltte was a terrorist organisation does not justify the disproportionate targeting of civilians. The failure to effectively disentangle counter-terrorism from an ethno-nationalist campaign of targeted mass killing signals a genuine failure of the international community to operationalise the R2P when tested. In Kenya, terrorism was a factor in the international response, albeit from the standpoint that Kenya played a critical role in the war on terror in Africa. A weaker or theoretically collapsed state could have proven dangerous for the US and other states involved in the war on terror. During the mediation talks, a congressional subcommittee on Africa noted that 'Kenya was too important in the region and the world to be allowed to go the way of Rwanda and Somalia' .92 The level of international coordination on Kenya is exemplified by US collaboration on official statements with lead mediator Annan. US Secretary of State Rice's lockstep approach on Kenya was unusual for Bush administration efforts around multilateralism; as a member of the mediation team noted, 'this was unheard of! I would give the statements to Annan and he would approve them before they went out' .93 This was not necessarily the only possible framing, but it was the prevailing one. As discussed, some Kikuyu politicians and political leaders had described Luos as terrorists in a deluge of hate speech during the post-election violence.94 Indeed, attacks like that at the Kiambaa church in Eldoret reasonably fit under definitions of terrorism. The targeting of civilians designed to achieve a political objective is relatively clear in this case. Whether the audience was the government or Kikuyu civilians is less certain. It would not, however, have been a significant stretch from parts of the international community or media to adopt a terrorism framing in siding with the government against the odm had such a path been chosen. It was perhaps the competing predominance of the anti-Islamic fundamentalist war on terror and Kenya's role in fighting it in the Horn of Africa that determined such a framing would not be adopted, even where it conceivably could have been. The stark contrast with how the US support, or lack thereof, for peace processes was affected by terrorism framings and impacted subsequent outcomes in Kenya compared to Sri Lanka is clear. An unambiguous choice of framing as terrorism versus ethnic conflict drastically affected responses between the two crises. While we cannot be entirely certain of the extent to which strategic interests directly augmented the international response, it is not difficult to see why the confluence of direct interests made Kenya an ideal case for R2P. The conflict was at a much earlier stage. There was no mass militarisation and no involvement of armed forces. There was broad support for peace across Kenyan society. The level of external unity was commendable between Kenya's neighbours, the African Union, United Nations, and the broader international community. As such the odds of a positive outcome were good. Operationalising R2P in Kenya was easier than it would have been in Sri Lanka, but the general lack of any attempts to apply R2P in the latter case is rooted in the prioritisation of a counter-terrorism narrative and, from a civilian protection perspective, the ambiguity of its conflict narrative and the lack of a coherent, cohesive set of interests in preventing atrocities.
This comparative variation of framing does to an extent reflect the varied magnitude and level of difficulty in resolving the cases. It also, however, demonstrates the ability of the international community to distinguish genuine terrorism and counter-terror responses from illegitimate ones that act as part of the justificatory mechanism for atrocities. It also demonstrates that complex conflict dynamics will often be framed to fit overly simplistic narratives, and that the international community can choose which narrative they are willing to accept. It is often the interests of powerful states that underpin that choice. To paraphrase the oft-cited adage, it seems one state's R2P emergency is another's counter-terrorism operation. This dynamic is clear from observing other cases of intervention and non-intervention. nato was willing and able to overlook the human rights abuses committed by the Kosovo Liberation Army in 1999 both before and following its intervention, due at least in part to a clear narrative that stemmed from the genocide in Bosnia four years earlier.95 A similar willingness to overlook crimes committed by anti-Gaddafi rebels also prevailed in Libya following the 2011 intervention.96 Meanwhile, perhaps emboldened by the success for impunity and non-interference in Sri Lanka, Asia's great powers China and India acted in unison with their descriptions of the 'Clearance Operations' against the Rohingya as counter-terrorism.97 This is despite the clear evidence that the violence unleashed was grossly disproportionate compared to the threat arsa legitimately represented. This is not to downplay the crimes of the ltte or overstate Kenya as a simple and easy case of conflict resolution. Rather it is an illustration of how complexity breeds ambiguity, and from ambiguity, states, and at times the broader international community, can pick the narrative that suits, according to their own interests and priorities. In some cases, this will be grounded in a clear narrative of terrorism and counter-terrorism, at other times it will act merely as a fig leaf to justify atrocities. The framing of conflicts and events as terrorism and counter-terrorism or ethnic conflict, electoral violence, tribal conflict, intercommunal violence, etc has significant effects on how the international community responds. These forms of political violence may manifest in very similar forms, yet it is often how they are labelled that informs external responses. In this way, framing may be more important than the ground truth of events in how R2P can be operationalised. Ultimately, the international community must be more aware and better able to respond to the use of counter-terrorism as a cover for the disproportionate use of force that often results in mass atrocities.
This article prompts important questions about how the R2P is applied and operationalised. There is broad recognition of its imperfections, but it remains an important normative banner under which the international community can rally in an effort to prevent and mitigate mass atrocities. There are, however, clearly critical weaknesses that imperil its future success. In the case studies presented in this article, one state, whose stability was at risk, was a critical partner to international efforts in combating terrorism; in the other, the designated target of the violence was a terrorist organisation, a fact used to justify mass atrocities against the population within which the group operated. These characteristics were intertwined with a variety of other geopolitical dynamics; however, it augmented the international response to atrocities and demonstrates the gap between promise and practice for the R2P. The ability of states to carry out atrocities with relative impunity after labelling the targets 'terrorists' is one such challenge, and one likely to persist. There are cases where such a claim is clearly illegitimate, though even in these cases it remains 96  a useful rhetorical device for perpetrating states or their external supporters to hide behind or muddy the waters to stymie intervention. This problem is particularly challenging where there is some degree of truth in such a label and atrocities can be set aside as permissible collateral damage; in effect, where the ends justify the means. It is a matter of genuine difficulty but one that needs to be grappled with to make progress towards the promise of the R2P. Situations in which there is a significant risk of mass atrocities are rarely black and white, vulnerable civilian populations will often include bad actors who use them as cover -creating grey zones. This problem does not negate the value of R2P as an evolving norm. Indeed, expecting the perfect application of a norm that is designed to grapple with the one of humanity's most intractable issues is akin to building a strawman for a bonfire. Seeking to manage this issue, to disaggregate vulnerable populations from bad actors that operate amongst them is a critical issue for the future of civilian protection. Challenging the cloak-and-dagger justifications for operational strategies technically designed to counter terrorist activities but often used as cover for wholescale repression or atrocities is a first step. In practice, this could involve states, organisations, and advocates directly challenging the narrative of counter-terrorism, an approach akin to naming and shaming. Another key measure is challenging those states that support perpetrators and provide legitimacy at the international level by blocking discussions or resolutions at the UN Security Council. Assuming bad faith from states that protect or enable perpetrators only guarantees such behaviour will continue to manifest. It is exceedingly difficult to create costs for external actors that are permissive of atrocities. Seeking, however, to raise any cost for that permissiveness is urgent. Raising costs for permissiveness begins simply from not assuming that states who protect perpetrators don't care for their international reputation. To do so is to let them off the hook. Tighter controls or sanctions on military assistance for these operations are other urgent and necessary steps. The Arms Trade Treaty and mechanisms like the Global Magnitsky Act are potentially significant measures but are relatively recent initiatives, the full effects of which are difficult to measure.
There are a number of potential avenues for future research that stem from this article. I focus on the desire to avoid the appearance of supporting terrorist actors that is important to explaining the success of the terrorism card in some cases but this is just one possible mechanism and further research should consider in what other ways might terrorism affect international responses? From the very small sample of cases it appears that the potency of terrorism framing is very much case and context dependent. It would, however, be an important 98 Butcher et al., 'Introducing the Targeted Mass Killing Dataset' . step to interrogate whether such framing is more or less likely to be adopted across a broader range of cases.

Conclusion
This paper has argued that the use of counter-terrorism narratives can be co-opted to justify and implement mass atrocity campaigns with impunity. This presents a unique challenge to the operationalisation of the R2P. The ability of perpetrating states, and their external supporters, to craft and reinforce a message of legitimate use of force, shielding gross abuses of human rights, obviates the application of R2P. I have demonstrated how the cases of Kenya and Sri Lanka are a temporally comparative example of how the international community struggles to operationalise R2P in cases where bad actors such as terror groups are operating amid large civilian populations. There are large dissimilarities between the cases presented in this research and there are many confounding factors besides those presented that make Sri Lanka a difficult R2P case and Kenya an ideal one. Yet the fact that both are cited in discussions of R2P is reflective of the wide gamut of conflicts the norm is expected to be able to address. The efforts to address the failure to prevent atrocities has incidentally created problems where success in Kenya or Ivory Coast should by some natural process of momentum lead to success in Syria or Myanmar despite wholly differing conflict dynamics and geopolitical contexts. The answer to Romeo Dallaire's question that forms the title to this article remains uncertain. There have been tremendous advances in efforts to prevent mass atrocities that continue a long tradition of humanitarianism. While still catastrophic, the severity of mass atrocities has waned significantly from the twentieth century's worst calamities.98 These successes are, however, always relative so long as atrocities persist. An appreciation of the complexity and challenges of conflict dynamics where atrocities are likely or ongoing is essential to grappling with efforts to undo the progress made to date. I argue that one key challenge is managing the use of counter-terrorism discourses that build justifications for mass atrocities. Citing terrorism cannot continue to be a licence to commit atrocity crimes. At present R2P is shown to be effective in managing cases at an early stage of escalation to atrocities. As in Kenya, while the conflict roots were deep the dynamics were not such that the parties, with the help of the international community could not step back from the cliff. In nanlohy Global Responsibility to Protect 14 (2022) 178-203 Sri Lanka, however, a long established conflict that had developed into a dirty war lacked a clear or simple narrative around which a cohesive protection effort could coalesce. The starkness of outcomes is plain to see. Atrocity prevention, like war, is messy; embracing complexity is a necessary step to more effective prevention. Without such efforts history is bound to repeat and some humans will remain less human than others.