The permanency of mass atrocities: The fallacy of ‘never again’?

Abstract

The terminology of ‘never again’ has been studied in Sociology, Cultural Studies and History, yet remains neglected in International Relations. This is despite its centrality in debates over humanitarian intervention, the Responsibility to Protect and mass atrocity prevention. To foster a conversation over the term’s use, this article uses an interdisciplinary approach to show how the meaning of ‘never again’ has changed over time. Building on this, we analyse five real-world problems: (1) the quantitative problem, (2) the nuclear problem, (3) the regime change problem, (4) the weak state problem and (5) the P5 problem. We find the blanket call of ‘never again’ oversimplifies the complexity of mass atrocity prevention and creates an unrealistic goal. Going forward, we call on those invoking the phrase to explain what they mean by it and why they are using it as part of a broader reassessment of the term’s use in International Relations.

Introduction

Mass atrocities (genocide, crimes against humanity, ethnic cleansing and war crimes) around the world have brought the debate over mass atrocity prevention in an era of shifting power balances to the forefront of International Relations (IR) (Gallagher and Wheeler, 2021; Pattison, 2021; Peak, 2023; Welsh, 2019). Reflecting on this tragic reality, the then Special Adviser of the United Nations (UN) Secretary-General on the Responsibility to Protect (RtoP), George Okoth-Obbo proclaimed ‘“never again” needs to be preserved and echoed’ (Special Adviser of the Secretary-General on the Responsibility to Protect, 2022). The statement forms part of an everyday discourse in which political elites, academics, non-governmental organisations (NGOs) and journalists proclaim that we should ‘never again’ let mass atrocities occur. In sharp contrast, speaking at the Global Network of RtoP Focal Points, the Director of the Simon-Skjodt Center for the Prevention of Genocide at United States Holocaust Memorial Museum, Naomi Kikoler (2022), stated that she finds the ‘never again’ label ‘unhelpful’. What we see, therefore, are two opposing views expressed by people equally committed to the prevention of mass atrocity crimes. The former represents the mainstream view while the latter provides a fleeting insight into what very few people ever explicitly acknowledge. Although there has been research on ‘never again’ in relation to genocide in Sociology Baer and Sznaider, 2019 (2016), Cultural Studies (Popescu and Schult, 2019) and History (Kansteiner, 2017; Kellner, 1994), there has been very little research in IR. ¹ This is surprising because the phrase is commonly invoked in relation to prominent debates in IR over humanitarian intervention, the Responsibility to Protect (RtoP) and mass atrocity prevention which underlines the pressing need for research.

December 2023 marked the 75th anniversary of the 1948 UN Convention on the Prevention and Punishment of the Crime of Genocide. The Holocaust was so shocking that humans have struggled ever since to capture the unimaginable horrors that occurred. Martin Gilbert (1987: 419) explains that neither words, nor statistics, nor examples, can adequately convey the suffering involved. We are reminded of Primo Levi’s (2009: 32–33) discussion of the limitations of language:

Then for the first time we became aware that our language lacks words to express this offence, the demolition of man. In a moment, with almost prophetic intuition, the reality was revealed to us: we had reached the bottom. It is not possible to sink lower than this; no human condition is more miserable that this nor could it be conceivably so.

Against this backdrop, we understand the plea of ‘never again’ as a shorthand to convey a rallying cry to the world: humans should ‘never again’ let such crimes be perpetrated. From this perspective, ‘never again’ is meant to mean something. Problematically, however, this something has changed significantly over time which raises a series of questions that are difficult to answer precisely because ‘never again’ is rarely studied or debated. First, what does ‘never again’ mean? As ‘The expanding parameters of never again’ section explains, the remit of ‘never again’ expanded from (1) no more Holocausts, to (2) no more genocides, to (3) no more mass killing, to (4) no more ‘mass atrocities’ which refers to four crimes under the RtoP, genocide, crimes against humanity, war crimes and ethnic cleansing. These are different objectives and highlight that the term should not be invoked as though its meaning is fixed. Second, what are the implications of this expansion? Although the need for enlarging the remit of [inter]national responsibility beyond that of ‘just’ preventing genocide has been discussed as part of the RtoP literature (Doyle, 2016), there has been a silent creep in the ‘never again’ discourse as the term is linked to more and more crises and contexts, yet this is very rarely, if ever mentioned.

This brings us onto the third question, ‘How is “never again” used?’ Does it represent galvanising rhetoric, a realistic goal, an [unachievable] ideal type, an amalgamation of the above or something else? Regarding the former, it may be that the label is used to try and ignite political will in a world where studies show that this remains a fundamental barrier to effective mass atrocity prevention (McLoughlin et al., 2023). If this is the case, research on this issue is urgently needed. For example, there are studies on the RtoP which question its value as a ‘rallying call’ yet argue that it is important as a ‘habit former’ (Bellamy, 2013); however, there is no such analysis in relation to ‘never again’. In terms of whether it is a realistic goal, it may be that one’s answer to this question hinges on one’s answer to the first question; ‘just’ preventing genocide is presumably more realistic than preventing all mass atrocity crimes, but we find even this narrow definition to be unachievable. Conversely, some may consider ‘never again’ to be an inspirational aim even if it is currently perceived as unobtainable. To quote Browning (2007: 144), ‘Ah, but a [hu]man’s [sic] reach should exceed his grasp, or what’s a heaven for?’ Those who advocate such use may argue that throughout history there are countless examples of humans achieving what previous generations would have viewed as unachievable and even unimaginable. Finally, it could be that the phrase is valued for other reasons such as symbolic power, as people hold up signs which state ‘never again’.

To return to the studies in History (Kansteiner, 2017; Kellner, 1994), Sociology (Baer and Sznaider, 2016), Cultural Studies (Peschel and Sikes, 2020; Popescu and Schult, 2019) and IR (Schiff, 2008), notably, these situate ‘never again’ within the legacy of the Holocaust and as a result, use the phrase in a narrow sense. Although written from different disciplinary perspectives, they cover the related themes of historical representation, memory narratives and performativity. The key findings include, first, the complexity of the term, meaning that ‘unpacking the semantic contents of “Never Again” would be an enormous task’ (Kellner, 1994: 127). Second, the blurred boundary between ‘never forget’ and ‘never again’ as the term has been invoked to shine a light on the horrors of the past and prevent these from occurring again. Accordingly, Baer and Sznaider (2019 (2016): 4) raise the need to ‘[r]emember atrocity, honor the victims, learn for the future’. This aligns with Schiff (2008) as he argues political elites use the term as part of a ‘redemptive politics’ that looks to the past while vowing to never let it occur in the future. Third, they find a tension between a universal commitment embodied in the broad parameters of ‘never again’, as opposed to the narrower conception of ‘never again us’ (Baer and Sznaider, 2019 (2016): 5, 13). Fourth, the authors acknowledge the failure of ‘never again’ in terms of preventing genocide yet continue to champion the phrase. The exception being Schiff (2008: 48) who argues ‘Never Again’ should be replaced by ‘Not this time! Not here, not now, not today’ but this too embodies the idea that this alternative terminology can be used to galvanise action. Committing to ‘Never Again’, Peschel and Sikes (2020: 259) explain that they are inspired by Samuel Beckett’s writing, ‘Ever tried. Ever failed. No matter. Try again. Fail again. Fail better’, while claiming ‘the emotional force behind it (never again) can still have the power to motivate’ (Peschel and Sikes, 2020: 260). Yet, this clearly raises the question, ‘If it does have the power to motivate, why does it keep failing?’ It is here that the value of IR comes to the fore, as the studies focus on groups and do not engage in the anarchical realm.

Going forward, we hope that those invoking the phrase ‘never again’ explain what they mean by it and why they are using it. To shape this conversation, the article is structured in six parts. ‘The expanding parameters of never again’ section explains the historical evolution of the term and how its meaning has changed over time; however, the implications of this are rarely, if ever, discussed. The article then analyses five problems which are chosen because they highlight different issues that need to be considered carefully in future reassessments of ‘never again’. ² First, the quantitative problem demonstrates that the increase in mass atrocities around the world has seen ‘never again’ linked to more and more cases and contexts, which we argue makes it unrealistic, raises questions over its power as a rallying cry and dictates that even as an ideational goal it is now further from reach. Second, the nuclear problem: if a nuclear state is ideologically committed to perpetrating genocide, we struggle to see how states can prevent these atrocities from occurring. This is important because even if someone rejects the expanding parameters of ‘never again’ and argues that it refers to a commitment to ‘just’ preventing major genocides, they would still be faced with ‘the nuclear problem’. Third, the ‘regime change’ problem reflects on the disastrous consequences associated with this practice which has fuelled a political unwillingness to use force in the future which puts the goal of ‘never again’ even further out of range. Fourth, the weak state problem exposes that even if the state in question requests international assistance, the complexity of these cases cannot be overstated, thus challenging the idea that ‘never again’ is realistic. Fifth, the permanent five (P5) problem: by conflating ‘never again’ with the RtoP, the former has become bound by the institutional barriers embodied in latter which means those invoking the phrase need to either de-couple it from the RtoP or address the challenges that arise from this relationship. Overall, we find that the blanket call of ‘never again’ creates an unrealistic goal which oversimplifies the complexity of mass atrocity prevention in IR.

The expanding parameters of never again

The phrase ‘never again’ when invoked in relation to mass violence is often traced back to the 1927 poem by Yitzhak Lamdan titled ‘Masada’, but became synonymous with the Holocaust as survivors of Buchenwald concentration camp held signs in different languages which read ‘never again’ (Baer and Sznaider, 2019 (2016): 9). Even then there were different meanings at play as political prisoners used the phrase against fascism, whereas Jewish survivors were stressing the need to ‘never again’ let such acts occur and also ‘never forget’ the victims. It was in the aftermath of the Holocaust that the phrase became universalised with different meanings becoming increasingly blurred (Baer and Sznaider, 2019 (2016): 11; Popescu and Schult, 2019: 135–136). During this period, ‘never again’ became synonymous with the 1948 UN Convention on the Prevention and Punishment of the Crime of Genocide, even though the words do not appear in it.

To understand how ‘never again’ became mainstreamed, it is important to factor in the politicisation of the phrase as political elites reflected on cases of genocide while vowing to ‘never again’ let them occur. For instance, Power (2002, xxvii) explains how in the United States a new culture of Holocaust awareness saw US Presidents Carter, Reagan, Bush and Clinton all vow to ‘never again’ let genocide occur while doing little to prevent it. It was not actually until 1988 that the United States ratified the 1948 Genocide Convention. President Reagan proclaimed, ‘And I want you to know that we intend to use the [genocide] convention to expand human freedom and right human rights abuses around the world. Like you, I say in a forthright voice, “Never Again!”’ (cited in Ronayne, 2001: 37). The statement captures both the politicisation and the expansion of the term as Reagan came out in support of ratification just 8 weeks before the general election and did so in a way that suggested it could be used as a foreign policy tool to prevent a wider spectrum of human rights abuses than genocide alone. Since then, the relationship between remembrance, ‘never forget’, and ‘never again’, has continued to raise its head in US politics as Presidents Obama (News24, 2011), Trump (Rhodan, 2017) and Biden have reflected on Rwanda and the Holocaust while reiterating the ‘sacred vow’, to use Biden’s words, of ‘never again’ (White House, 2022). Ultimately, the globalisation of the term saw more and more actors invoke the phrase which in turn saw it appropriated with an ever-increasing array of problems, crises and contexts.

What was initially rooted, therefore, in a commitment to ‘never again’ let another Holocaust occur, expanded over time to include all victims of genocide. As Baer and Sznaider’s (2019 (2016)) study evidences, the ‘globalized Holocaust culture’ saw ‘never again’ used in Argentina, Spain and Eastern Europe to draw attention to the human rights violations going on. In the post-Cold War era, the genocides in Rwanda and Bosnia led to claims that ‘never again’ was a ‘hollow’ slogan (Gourevitch, 2000). What we see therefore is that very different cases of genocide were placed under the same banner of ‘never again’. For example, the Ambassador of Bosnia to the United States acknowledged that the genocide in Bosnia did not equate to that of the Holocaust but expressed his frustration because ‘[t]he Genocide Convention was adopted by the UN as an instrument of international law that would prevent genocide from happening ever again’ (Alkalaj, 1999: 358). He goes on to explain that what was shocking about what occurred in Bosnia was not just the crimes perpetrated but that ‘the West’s’ belief that they had put genocide behind them, by vowing ‘never again’ let genocide occur, had been shattered (Alkalaj, 1999: 358). Regarding the question of whether ‘never again’ represents a realistic goal or galvanising rhetoric, he seemingly upholds the former position, ‘we must ask ourselves what we should do to ensure that “never again” really means never again’ (Alkalaj, 1999: 362). Yet at the same time it is important to recall that his definitional parameters are narrow as he is just taking about genocide.

The expansion ‘never again’ beyond that of ‘just’ genocide was driven by a two-fold problem. On the one hand, genocide is notoriously difficult to prove because of its requirement of ‘intent’ which is not included in the legal definitions of crimes against humanity and war crimes. On the other hand, there may be grave human rights violations that do not meet the threshold of genocide but still ‘shock the conscience of humankind’, which was the threshold used for contemporary debates over the legitimacy of humanitarian intervention to ‘save strangers’ (Wheeler, 2000: 50). The crises in Iraq, Somalia, Rwanda and Kosovo represented different scales and types of mass violence yet raised a common concern that more needed to be done in order to protect civilians. Against this backdrop, the then UN Secretary-General, Kofi Annan, stated:

Of all my aims as Secretary-General, there is none to which I feel more deeply committed than that of enabling the United Nations never again to fail in protecting a civilian population from genocide or mass slaughter. (UN, 1999, emphasis added)

The statement underlines both the prioritisation and expansion of ‘never again’ and aligns with the post-Cold War expansion beyond that of ‘just’ genocide. Of course, this followed the Rome Statute of the International Criminal Court (ICC) in 1998 which identified genocide, crimes against humanity and war crimes as international crimes under the jurisdiction of the ICC.

The most significant expansion of the phrase ‘never again’ occurred as it became synonymous with the 2005 World Summit agreement on the RtoP which, like the 1948 Genocide Convention, does not include the terminology ‘never again’ but the phrase is often used interchangeably with it. For example, a journalist at the time reflected on the RtoP agreement and entitled his piece ‘UN must “never again” be found wanting on genocide’ (Turner, 2005). While framed in terms of genocide prevention, it illustrates how the promise of ‘never again’ and the RtoP became conflated with academics going as far as to claim ‘R2P was supposedly our answer to “never again”’ (Kersavage, 2014: 36). According to Roland Paris (2014: 579), ‘R2P simultaneously symbolizes something larger than the 2005 agreement: it is the embodiment of the pledge to “never again” allow genocide to occur, a commitment born out of the Holocaust’. The rhetoric has continued to be invoked as the failure to prevent mass atrocities in cases such as Darfur and Myanmar and has been explicitly framed as a missed opportunity to make ‘never again’ a ‘reality’ (Jentleson, 2007: 27; Šimonović, 2021: 390). Although Jentleson (2007: 22) acknowledges that ‘the prevention of all mass killing is unrealistic’, he goes on to claim that turning ‘never again’ from ‘rhetoric to reality’ is needed so that we can become a ‘global community’ (Jentleson, 2007: 23). Evans (2009) adopts a different stance, as he does not believe that the RtoP can end all conflict (Evans, 2007), but he does think that it can bring an end to all mass atrocities (Evans, 2009: 3) and maintains his position that the RtoP can ‘realise’ the ‘dream of making “Never Again” more than a slogan’ (Evans, 2021). In a similar vein, Adams (2021: np) claims his book offers a ‘unique perspective regarding how we make “never again” a living principle rather than a cliché, and how we end the politics of impunity, indifference and inaction once and for all’. To give a final example, a leading NGO released a report to commemorate the 10th anniversary of the RtoP titled ‘Never Again? Guide to the Responsibility to Protect and the Prevention of Atrocity Crimes’ (UNA-UK, 2015). The report is just one of many such guides which set out to explain how policymakers can make good on the commitment of ‘never again’.

Evidently, the conflation of ‘never again’ and the RtoP has seen the parameters of the former extend far beyond preventing genocide; however, there are critical voices in the discourse and here two strands emerge. First, it has been argued that ‘to keep the promise of never again the world must recognise genocide as a crime distinct from crimes against humanity and war crimes’ (Ochab, 2016: 42). From this perspective, it is possible to fulfil the promise of never again allowing genocide, but to do this we need to prioritise genocide prevention over other forms of mass atrocity prevention. Moreover, this is presented as the right thing to do. We are reminded of debates over whether genocide should be viewed as worse than crimes against humanity, ³ but such thinking has been overlooked in contemporary uses of ‘never again’. Second, there are those who find the conflation of RtoP and ‘never again’ problematic (Badescu, 2010: 166; Gallagher, 2015: 268) as it creates an unrealistic goal that the norm cannot fulfil. This gives critics ammunition to reject the norm because, in part, it does not fulfil the goal of ‘never again’ even though this was not feasible anyway and not written into the World Summit agreement on the RtoP.

This is not to suggest that the expansion of ‘never again’ has been solely driven by its connection with the RtoP and as discussed above, the globalisation of the Holocaust discourse began in the Cold War. The point is that in contemporary IR we see political elites invoke ‘never again’ in many different contexts. For example, when reflecting on the Holocaust, Lithuania’s Prime Minister Ingrida Šimonytė stated, ‘“never again”’ must become the air we breathe every day’ (The Office of the Government of the Republic of Lithuania, 2022). Such use aligns with earlier meanings of the phrase as the focus is very much on preventing another Holocaust. Capturing a broader understanding, when speaking about ‘crimes against humanity’ against the Rohingya, the Bangladesh Prime Minister, Sheikh Hasina (Dhaka Tribune, 2020), claimed, ‘On this day, let us reaffirm our commitment to ‘never again’ to genocide for the sake of humanity and global peace’. Denouncing Russian airstrikes, Ukrainian President Volodymyr Zelenskyy, asked ‘To the world: What is the point of saying never again for 80 years, if the world stays silent when a bomb drops on the same site of Babyn Yar?’ (Treisman, 2022). To give one last example, discussing the migrant crisis in Europe, the European Union’s foreign policy chief Federica Mogherini, declared, ‘We have said too many times “never again”. Now is time for the European Union as such to tackle these tragedies without delay’ (Emmott, 2015). This is obviously very different to preventing a major genocide and further illustrates that the meaning of ‘never again’ has changed considerably, as its remit has expanded and it has been linked to more and more cases.

Overall, there appears to be a silent creep in the ‘never again’ discourse as its meaning has changed over time, yet this is not discussed and debated in the discourse. The world now seems very different to that of 1948, and here it is important to juxtapose the expanding meaning of ‘never again’ with real-world developments that have significant implications for it. To illustrate this, we turn out attention to five key problems which exposes that ‘never again’ is unrealistic and that even a narrow interpretation that focuses on genocide prevention is also unachievable, at least in a world with nuclear weapons.

The quantitative problem

We share the sentiment expressed by (Ignatieff, 2021: 178) the era in which the RtoP was forged feels like a very different world to the one we live in now. Whereas he focuses on the changes that have occurred between the United States, China and Russia, we focus here on the increase in mass atrocities around the world.

The 21st century began somewhat peacefully. As Harff (2003: 61) explains:

although the last half-century has seen a long-term increase in numbers of genocide/politicide, by the end of 2001 mass killings were on-going in only two countries, Sudan and Angola, fewer than at any other time in the previous 30 years.

While one could challenge this conclusion on the grounds that other cases should have been included, the extensive study (1955–2001) captured the downwards trend in mass violence that accompanied the start of the new century. Against this backdrop, one could be forgiven for thinking that the ambition of ‘ending mass atrocity crimes once and for all’ (Evans, 2008) was realistic. After all, a decade on from Harff’s (2003) data endpoint, Pinker (2011) argued we are living in the most peaceful period of human history while acknowledging this is contingent. The Arab Spring saw uprisings and repercussions challenge this perception of peace. Mass atrocities in the Democratic Republic of Congo (DRC), the Central African Republic (CAR), South Sudan and Syria led analysts to claim ‘never again’ was in fact ‘ever again’ (Caryl, 2015). Yet, many continued to defend the idea of progress which embodied a two-fold argument. First, the world is really not as bad as many think it is. Yes, there are atrocities around the world but ‘the frequency and intensity of these terrible episodes has diminished significantly over the past half-century’ (Ulfelder, 2015). To put this another way, the death toll in the very worst crises (Syria) are in the hundreds of thousands rather than the tens of millions as they were in the 20th century (Rummel, 1994). Second, despite the shortcomings of the international response, particularly of the UN Security Council over crises such as Syria and Myanmar, it is important to recognise the efforts made in cases such as CAR, DRC, Mali and South Sudan. The latter reflects that we live in an era of ‘routine international responses’ that sets ‘early twenty-first-century politics of human protection apart from earlier periods’ (Glanville, 2021: 166).

Yet, even if we accept the claim that international human protection has become more common in the 21st century, the sheer number of mass atrocities outstrips our capacity to prevent them. For instance, the Political Instability Task Force (PITF) State Failure Problem Set 1955–2018 includes data sets on ‘genocide and politicide’ between 2000 and 2018 (Centre for Systemic Peace, 2020a) and ‘ethnic wars’ between 2000 and 2018 (Centre for Systemic Peace, 2020b). The former identifies six cases, ⁴ while the latter includes 28 cases. ⁵ Although the definitional debates that surround each individual mass atrocity crime dictate that we should tread with caution when incorporating related terms such as politicide and ethnic wars, the data suggest that 34 cases of potential mass atrocity took place in the first 18 years of this century. Gallagher (2021: 1029) goes further as he identifies 37 countries that experienced mass atrocities or serious concerns of them between 2000 and 2020. Speaking at the time, the Director of Policy and Research at the Global Centre for the Responsibility to Protect (GCR2P), Jaclyn Streitfeld-Hall (2020) explained that in the previous 9 years, GCR2P had covered 30 countries in its monthly monitor and another 20 in its atrocity alert. This equates to over a quarter of the membership of the UN and this does not include the cases of mass atrocities that have occurred in countries since. In 2022 alone, Genocide Watch identified a ‘genocide emergency’ in India, Afghanistan, Ukraine, Syria and Iraq. Although the scale of mass atrocities in terms of the numbers killed in each conflict remains much lower than in the 20th century, ⁶ the sheer number of cases juxtaposed with the expansion of ‘never again’ to include all mass atrocities dictates that it is simply unrealistic to think all mass atrocities can be prevented.

We will return to the question of whether ‘never again’ represents galvanising rhetoric, a realistic goal, an unachievable ideal, or an amalgamation of the above. For those who believe that the term’s power lies in its ability to stir action, surely they need to reassess whether it has failed to do so, if so why, and what they can do in future. It may be that expanding the parameters of ‘never again’ (‘The expanding parameters of never again’ section) was detrimental in this regard as it became linked with so many crises that cries of ‘never again’ drown each other out. In part, this may add weight to the idea that the parameters of ‘never again’ need to be reduced. To link this to the second use, this would certainly make it more realistic as preventing ‘just’ genocide is easier than preventing all cases of genocide, crimes against humanity, ethnic cleansing and war crimes, yet this should not be seen as an easy fix because, as the five cases raised by Genocide Watch in 2022 illustrate, there are still profound challenges facing this ‘lesser’ goal and furthermore, as to be discussed, the nuclear problem remains unresolved. Finally, for those who see ‘never again’ as something we should aspire to even if it is unachievable, we are of the view that the significant number of cases has undoubtedly placed this goal even further out of reach (to use Browning’s rhetoric). Accordingly, we question the value of ‘never again’ if it merely acts as a blanket call that places too many different crises under the same slogan.

The nuclear problem

If a nuclear weapons state perpetuates mass atrocities against their own people and is wholly committed to the cause, it is very difficult to see how states can prevent these atrocities from occurring. Of the nine states that currently possess nuclear weapons, at least three are currently perpetrating mass atrocities, China, Russia and North Korea and two more, Pakistan and India, are identified as ‘high-risk’ (Simon-Skjodt Center for the Prevention of Genocide, 2021). There have also been allegations of mass atrocities regarding US treatment of migrants under Trump (Cronin-Furman, 2019), the United Kingdom’s treatment of the Chagos Islanders (Human Rights Watch, 2022) and Israeli policies towards Gaza with nearly 800 lawyers, scholars and practitioners warning of a potential genocide as we write (al Attar, 2023). While the problem is clear, the solution is anything but.

To illustrate this further, let us consider the case of the Democratic People’s Republic of Korea (DPRK). The history of systematic atrocities is detailed in the 2014 report of the UN Commission of Inquiry on Human Rights in the DPRK, which ‘reveal a State that does not have any parallel in the contemporary world’ (UN Human Rights Council, 2014: 15). The chair of the report spoke of the ‘many parallels’ with the crimes of the Nazis and their allies in the Second World War, given the ‘strikingly similar’ reports of atrocities committed by the DPRK against its own citizens (Walker, 2014). Echoing this judgement, a former judge at the International Court of Justice, himself a survivor of Auschwitz and Sachsenhausen, claimed conditions in the DPRK’s political prison camps were ‘as terrible, or even worse, than those I saw and experienced in my youth in [the] Nazi camps and in my long professional career in the human rights field’ (quoted in Fifield, 2017). The point here is not to draw parallels with Nazi Germany as such but to highlight that the crimes perpetrated in the DPRK may be the gravest on planet earth for a long time. Yet, the DPRK remains an enduring case partly because it is such a hard case, one that confronts policymakers and advocates with a series of unavoidable political constraints and difficult moral choices which limit the extent to which mass atrocity prevention strategies can usefully address the problem at hand. This has led to an explicit RtoP debate. Walt argued the absence of military intervention – given the nuclear threat posed by the DPRK – means that ‘R2P does not apply to North Korea’ (quoted in Song, 2015), whereas Liberals such as Adams claim it remains an ‘an essential lens through which to view and address the DPRK situation’ (Global Centre for the Responsibility to Protect, 2014: 4). Although one can rightly point out that Walt mistakenly conflates RtoP with the use of force and therefore fails to acknowledge the other coercive and non-coercive measures, the reality is that no one has come up with a mass atrocity prevention strategy for the DPRK.

This brings us onto the horrifying dilemma that pursuing a strategy of mass atrocity prevention may increase the likelihood of nuclear war. To put this another way, policymakers need to weigh up the threat of omnicide, the destruction of humankind, versus the threat of crimes such as genocide, the destruction of a group (Gallagher, 2013: 105). Although the merits of nuclear weapons continue to be debated, this debate centres on their role in creating security within the anarchical realm, but this does not address whether they exacerbate the problems of mass atrocity prevention. This is despite the fact that nuclear weapons were framed as a mass atrocity issue at the outset with scientific advisors at the time warning they could be used as a ‘weapon of genocide’ (Kuznick, 2007: 420). Although no one has used the nuclear option as a ‘final solution’, the dilemma of nuclear states perpetrating atrocities is ever present. Both the United States and the DPRK have made threats of genocidal nuclear violence (Kiernan and Simon, 2017). To give another example, the mass atrocities perpetrated by Russian forces in Bucha led to claims that ‘the West has decided the price of countering atrocities is simply too high . . . so much for “never again”’ (Herszenhorn, 2022). Meanwhile, Lieberman (2022) argued that the United States should enforce a no-fly zone as part of its ‘responsibility to protect’. Such analysis is problematic because it calls for ratcheting up response measures in order to fulfil the promise of ‘never again’, but fails to consider that preventing a potential nuclear war between ‘the West’ and Russia is a form of mass atrocity prevention in and of itself (Ralph, 2022). This is not to say states should do nothing, but that they should focus on positive incentives or lower risk options (sanctions and naming and shaming) rather than higher risk tactics (no-fly zone) as the latter increases the chances of starting a nuclear war which threatens the destruction of humankind.

We will return to the question of whether ‘never again’ represents galvanising rhetoric, a realistic goal, an unachievable ideal or an amalgamation of the above. The nuclear problem highlights that ‘never again’ could be counterproductive as there may be cases, as Ralph’s analysis of the Ukraine crisis begins to illustrate, that using it to stir action may escalate the threat of a nuclear war. Although those invoking the term may counter this by claiming that they use it to galvanise action and that this action could take many forms, it seems fair to ask that in the future these academics need to come up with nuclear-specific policies so when they do call for action, they can better guide policymakers. We find that the grave threat posed by nuclear states dictates that ultimately they can get away with mass atrocity crimes. This is important because even if someone rejects the expanding parameters of ‘never again’ and argues that it refers to a commitment to ‘just’ preventing major genocides (Ochab, 2016), how can they solve ‘the nuclear problem’? We do not find ‘never again’ to be a realistic goal, at least in a world where there are nuclear powers and we do not see this changing in our lifetime. Finally, although we are sympathetic to the claim that ‘never again’ is an aspirational ideal, more consideration needs to be given to explicitly acknowledging that this is not realistic, with careful thought given to whether it should be used in relation to nuclear states.

The regime change problem

Our concern here is not whether forcible regime change is an appropriate atrocity prevention policy as such, but rather that the negative consequences of regime change, especially Afghanistan (2001), Iraq (2003) and Libya (2011), have catalysed an unwillingness of states to use force in the name of atrocity prevention. Although the wars in Afghanistan and Iraq did not set out to prevent mass atrocities as such, the long-term implications are important for this topic because the unwillingness to act has decreased the chances of ‘never again’ being fulfilled but also increased the risk of mass atrocities as perpetrators know they are less likely to be held to account. The result is that ‘never again’ is even further from reach.

Whether it is Afghanistan, Iraq or Libya, the lesson to be learnt seems to be the same: getting rid of a regime is just the beginning. Realists question why political elites are not learning this lesson (Kissinger, 1976) and warn that occupation fuels resistance (Mearsheimer, 2009). In Gordon’s (2020: 257) aptly titled Losing the Long Game, he argues that ‘the costs [of regime change] will be higher than expected, unintended consequences will emerge, and results will leave much to be desired’. The point here is that whether the states involved intervene for the right reasons or not, the costs are far more complex than initially anticipated. In Afghanistan, many felt the only hope of success hinged upon the West leaving shortly after they intervened. This would help prevent the resistance realists warn of. Notably, the Taliban did disband after the initial invasion in 2001 (Gopal, 2014) and the decision to stay, for critics, laid the foundations of an unwinnable war. The cost was significant. Farrell’s (2017: 2) study explains that the United States spent $686 billion and the United Kingdom £37 billion while the former saw 2352 service personnel killed and the latter 456, with many more wounded. The true financial cost could be in the trillions. Of course, such statistics do not capture the grave losses experienced by Afghanistan citizens with an estimated 25,000 killed since 2007 alone (Farrell, 2017: 3). Sadly, Iraq bore similar lessons. Writing in 2020, Neta Crawford estimated US expenditure to be $1.922 trillion which she explains only includes official war funding and not that spent by the State Department (Crawford, 2020). Meanwhile, the Iraq Body Count (2022) claims over 185,000 civilians have been killed in Iraq.

The regime problem came to the fore once again over Libya in 2011. The fact that the Security Council authorised the use of force to prevent mass atrocities without the consent of the government in question led to claims that ‘we’ had entered a new era of human protection (Bellamy and Williams, 2011). The optimism was short lived (Welsh, 2011). The mandate creep from human protection to regime change (Luck, 2011) created a political fallout that had broader implications for international society (Reike, 2015; Ralph and Gallagher, 2015). For critics, this was another intervention driven by humanitarian motives which had catastrophic consequences, not least, for the Libyan people (Kuperman, 2015). The reality was that an intervention to prevent mass atrocities created a power vacuum in which further mass violence then took place. Whether an intervention is authorised by the Council (Afghanistan and Libya) or not (Iraq), the regime change problem remains the same. While one could rightly point out that there has always been a political unwillingness (Power, 2002), it is evident that at least in part, the dreadful consequences of interventions in the 21st century have accentuated this. Gallagher and Wheeler (2021: 192) claim that it is highly unlikely that the use of force under pillar III of the RtoP will be authorised anytime soon. Meanwhile, Ignatieff (2021: 178) claims that Libya ‘solidified the conviction, among politicians and their publics alike, that protecting civilian populations was an enterprise fraught with hazardous and unmanageable consequences’.

At this point, the reader may rightly point out that the focus on regime change fails to capture cases where there is use of force with the consent of the host state, such as the peacekeeping missions in countries such as DRC and CAR, which evidently did prevent mass atrocities from occurring. For example, Adams (2021: 119) argues, ‘most future mass atrocity situations will in all likelihood continue to look more like the CAR than like Libya. That is to say, they will continue to require elements of both coercion and consent. In a similar vein, Glanville claims that the focus of human protection should move towards the recognition that the international community has to ‘work with host states rather than against them’ (Adams, 2021: 172, emphasis in the original). There are two points of concern here. First, implicit within this shift away from regime change is an acknowledgement that regimes perpetrating mass atrocities, who do not consent to an external intervention, will get away with it. Second, this shift overlooks the fact that consenting states can be just as complex as non-consenting states (problem 4).

Let us return to the question of whether ‘never again’ represents galvanising rhetoric, a realistic goal, an unachievable ideal, or an amalgamation of the above. Those who champion its galvanising power need to better understand why elites are unwilling to act without a more informed understanding of what is fuelling political unwillingness and how they can hope to transform it. As part of which, it is necessary to consider the negative consequences of regime change as well as the shift from the use of force without consent to the use of force with consent for the pledge of ‘never again’. If the phrase is invoked to ignite action, it seems to us that real-world developments have blown its fuse. In terms of whether it is realistic or an aspirational ideal, we think this goal is now further away than ever before precisely because of the consequences from regime change in the 21st century. The paradox being that as the parameters of ‘never again’ expanded (‘The expanding parameters of never again’ section) which saw it linked with more and more crises (‘The quantitative problem’ section), the empirical realities unfolding in Afghanistan, Iraq and Libya actually made this more unachievable and as discussed, the shift away from non-consenting to consenting states, does not resolve this problem which brings us onto the next issue.

The weak state problem

Even if we accept that the future of human protection will look more like CAR than Libya, this leads us onto the next problem, as consenting states such as CAR are just as complex as non-consenting states such as Libya. In the past, realists have argued that liberals have underplayed the complexity of mass atrocity prevention and our concern here is that ‘never again’ falls into the trap of oversimplifying mass atrocity prevention. To illustrate the problem at hand, we turn our attention to CAR and the mass violence that occurred in the aftermath of the Séléka coup in 2012 (UN, 2014). Following the coup in December 2012 and the anti-Balaka backlash that followed, an estimated 3000–6000 people were killed, 825,000 internally displaced and 423,000 fled CAR between 2013 and 2015 (UN, 2014: 25, 92). The UN (2014: 26) investigation acknowledged the numbers are a ‘radical under-estimate’. Since 2015, there have been warnings of genocide in 2017, 2021 and 2022. The warnings and on-going violence add empirical weight to claims that the underlying conditions within CAR have not changed (Cinq-Mars, 2015; Glawion and De Vries, 2018; Lamarche, 2021).

When trying to explain why the mass atrocities occurred, a wide range of issues have been raised, including colonial legacies (Smith, 2015), a historical culture of violence (Glawion and De Vries, 2018; Lombard and Batianga-Kinzi, 2015), regional dynamics (Marchal, 2015a), non-state armed groups and the spread of arms (Conflict Armament Research, 2015), inequality and fear (Marchal, 2015b), peripheral neglect, outsourcing and the uniqueness of the coup (Glawion and De Vries, 2018) and the failures of international interventions (Lombard, 2016; Picco, 2015). The complexity involved is neatly captured in Lombard’s reflections, who it is worth noting is one of the world’s preeminent researchers on CAR:

I can pinpoint to historical and ethnographical dynamics, A, B, C, D, and E that have helped bring CAR to their current predicament. And I can cite grievances X, Y, and Z that likely motive the fighters. But the addition of all those factors does not somehow ‘add up’ to the violence over the last year and a half. There remains an excess, beyond which is explainable through reasons – even reasons related to symbolism and performance. (Written in 2014, cited in Lombard, 2016, 184)

The statement captures the two-fold problem. First, returning to the long list of causes above, how can mass atrocity prevention strategies address all these aspects? If we take just one of these, regional dynamics, it seems that in order to try and prevent violence in CAR, UN Member States need to address violence in Chad, Sudan, South Sudan, the Democratic Republic of Congo and even Cameroon, which until recently was viewed as a more peaceful state in this region. The porous borders, movement of armed groups, resource flows and regional politics begin to highlight the profound challenge. Moreover, the atrocities in CAR are said to have created an even ‘more divided region’ (Marchal, 2015a: 187), thus only accentuating the idea that this is mission impossible. But let us assume, for one moment, that these issues could be addressed, this brings us on to the second problem. Even if we add all these factors up, they still do not explain why the atrocities occurred. To return to Lombard, how can even willing actors prevent an elusive cause?

Furthermore, identifying causes is only one side of the coin, establishing the right strategy for prevention is the other. Despite the UN deployed 13,000 personnel compared to 200–700 previously in CAR (Glawion and De Vries, 2018: 433), it was widely criticised for failing to act quickly enough, being too risk averse and implementing quick fixes at the expense of long-term structural reform (Cinq-Mars, 2015; Lombard, 2016; Picco, 2015). For Liberals, the answer is to increase resources and long-term commitment, but questions have to be asked over whether this is realistic. For example, it is estimated that the Africa-led support mission to CAR (MISCA) cost €800,000 per day (Lombard, 2016: 20); here is the real issue, as CAR is just one of many cases as illustrated above. At times, it seems that academics discuss crises with blinkers on as they call for more and more in case A, without considering how this will impact cases B, C, D, E, F, G and so forth. Even if the answer were a bigger and bolder form of Liberal interventionism, it is extremely difficult to imagine who could fund multiple unending missions. And of course, many reject the idea that more robust Liberal interventionism is the answer. In a sobering assessment, Lombard (2016: 245) concludes that identifying problems is easier than solutions and claims that despite the efforts made in CAR, it remains a ‘work in regress’. The way forward is to break free of state-centric approaches, but here details become thin on the ground. What is clear from Lombard’s studies is that this requires not just different tools in the toolbox, but a completely different way of thinking. The UN and its Member States cannot build states where they do not exist and have to accept different forms of ordering. A sentiment which is shared in Glawion’s (2021) study of CAR, South Sudan and Somaliland begins to underline that this is not an issue exclusive to CAR.

Where does this leave us? If the causes of mass atrocities in these countries are so far ranging and ‘the state’, in the Weberian sense, within these countries remains elusive; the implication of this is that long-term mass atrocity prevention requires an unquantifiable amount of resources, time, commitment and, more worryingly, the right strategy amid claims and counter-claims that one strategy over another are [in]correct. If we factor in the two problems already outlined, (1) the large number of mass atrocities around the world, and (2) the amount of resources already spent fighting long wars, it is difficult to see who is able or willing to address the problem at hand, especially as the ‘right’ strategy remains so contested. This is important because it also sheds new light on why leaders are reluctant to act in the first place. For example, when discussing the ongoing mass atrocities in Sudan, the former UN Special-Advisor on the Prevention of Genocide, Adama Dieng (2023), rightly pointed out the horrors involved but claimed they evidence that ‘humanity has learned nothing from Rwanda, Kosovo and elsewhere’. We question this judgement as we view the inaction in Sudan to be driven by the lessons learnt from action in DRC, CAR, Mali and South Sudan (and one could add Libya), where peace remains elusive despite significant efforts being made, rather than inaction over Rwanda. In other words, are these missions this generation’s Somalia syndrome? ⁷ When faced with impending mass atrocities, there is very little appetite to respond because of recent action rather than inaction with the profound complexities of so-called weak states looming large. Although one could point to the recent Security Council authorisation of an intervention force in Haiti (23 October 2023) to challenge this narrative, it is important to bear in mind that this is far smaller than previous interventions and does little to give us confidence that UN Member States are eager and willing to even try and address the challenges embodied in the cry of ‘never again’.

Let us return again to the question of whether ‘never again’ represents galvanising rhetoric, a realistic goal, an unachievable ideal or an amalgamation of the above? There appears to be a pressing need to reassess the power of ‘never again’ to stir action, especially in the wake of the post-Arab spring interventions that did not bring about the desired results. The fact that interventions in consenting states have been so problematic adds further weight to our position that ‘never again’ is unrealistic and that mass atrocities are a permanent feature of IR. Although there may be those who hold on to the idea that it should be invoked as an ideal type, they need to reassess the negative implications which arise for as this section demonstrates, even when the target state welcomes international assistance, the challenges involved are potentially insurmountable. The blanket call of ‘never again’ downplays complexities such as the regional dynamics involved while ratcheting up expectations of what can be achieved.

The P5 problem

The problem here refers to the conflation of ‘never again’ with the RtoP which was raised in ‘The expanding parameters of never again’ section. We view this as a dangerous liaison as the moral cry of ‘never again’ is constrained by the political structures embedded within the RtoP. As a result, those invoking ‘never again’ need to either de-couple it from the RtoP or address the challenges that arise from this relationship.

The RtoP norm, with its emphasis on acting through the UN Security Council, effectively places preponderance of decision-making with the P5 veto-wielding states of the UN Security Council (the United States, the United Kingdom, France, Russia and China). This means that mass atrocity response efforts are primarily reliant on responsible action and co-operation between these five states. If there was any hope in 2005 that the P5 would lead by example, this is now gone as these very states often act in flagrant violation of their RtoP responsibilities, for example, Xinjiang, Ukraine and Syria. To re-cap, paragraph 139 of the World Summit Outcome makes clear that, on a case-by-case basis, the international community should consider acting to address mass atrocity situations ‘through the Security Council’ (UN, 2005). Yet, the P5’s power to veto any draft resolution was left untouched at the adoption of RtoP in 2005, and veto restraint initiatives have remained unheeded by several of the P5 (Illingworth, 2020). The result is that the RtoP’s close connection with the P5 veto creates a two-fold problem for responding effectively to mass atrocities.

First, the veto power enables P5 members to shield their allies from international reprobation through the Security Council. Syria is the most pertinent example of this: 18 separate UN Security Council draft resolutions have been vetoed by Russia since 2011, with 10 of these draft resolutions seeing a joint veto exercised by both Russia and China. These vetoes have been cast to protect Bashar al-Assad’s regime despite perpetrating well-documented mass atrocity crimes, including wide scale use of torture, siege and starvation tactics, and the deployment of chemical weapons (Bellamy, 2022). We are also acutely aware, however, that the ‘P5 problem’ is not just shorthand for a ‘Russia and China problem’. The United Kingdom and France may not have exercised their veto power since 1989, but the United States continues to actively deploy its veto power when it suits its interests. The United States has vetoed 34 draft resolutions over Israel and the occupied Palestinian territories, with the most recent coming on 18 October 2023 (Dag Hammarskjöld Library). In addition, the P5’s ability to shield other states is evident in the so-called ‘hidden veto’, as issues can fail to even make it to a formal resolution vote if Security Council members know a veto would likely be forthcoming (Security Council Report, 2015: 3). For instance, genocidal violence committed against the Rohingya in Myanmar – as well as crimes committed by the Tatmadaw since its military coup in March 2021 – did not lead to quick Security Council action, with China acting to protect its ally. A year and a half on from the coup and the UN Security Council had done nothing more than meet in private and issue press statements. It was not until 22 months after the coup that the Security Council passed a resolution.

Second, the veto power helps enable P5 culpability in atrocities themselves, either directly or indirectly, and get away with it. Essentially, the P5 members can carry out grave human rights violations in the knowledge that their veto power, or threat of their veto power, can prevent any action being taken against them in the Security Council. This is clearly reflected in the Ukraine case with Russia’s violations of international law, as well as in the Xinjiang case, where the Chinese state has arbitrarily detained one million ethnic Uyghur Muslims (Uyghur Human Rights Project, 2020: 1). Despite allegations, and the determination by an independent panel, that China is guilty of crimes against humanity and genocide (Uyghur Tribunal, 2021), the matter has not even made it onto the Security Council’s agenda. Much like the nuclear problem discussed above, what this shows is that ‘never again’ is an especially unrealistic goal when major power states are concerned, made even more utopian by the fact it has been tied to the RtoP norm which allows the P5 to perpetrate mass atrocities and use their veto as a shield to protect themselves.

At first glance, the obvious solution appears to be to distance ‘never again’ from the RtoP norm as a case could be made that the former should not be constrained by the UN Security Council. Although this is not discussed explicitly in the literature, it aligns with the view that the likelihood of Security Council paralysis should lead us to champion the potential of unilateral action under the remit of ‘coalitions of the willing’ (Erskine, 2014). Problematically, however, such moral theorising seems somewhat detached from post-Libya developments. It seems fair to say that when it comes to military intervention without the consent of the Security Council or the target state, there are few, if any, coalitions of the willing. This stems in part from the track record of failed intervention discussed above. Of course, it should be acknowledged that states have more options at their disposal than just use of force (Barber, 2021; Bellamy and Luck, 2018: 141–43; Glanville, 2017; Pattison, 2018). These ‘alternatives to war’ (to use Pattison’s terminology) should not be discounted and not all of these need the authorisation of the Security Council. For example, some have argued for a greater role for the UN general Assembly in actioning response to mass atrocity situations (Barber, 2021; Butchard, 2020; Illingworth, 2024). Therefore, some may argue that such tools can and should be used to help make ‘never again’ a reality. Although we do not oppose such measures being taken, we remain sceptical that they will affect the P5 members given their material and social position in world politics. For example, when faced with mass atrocities in China, the UK’s Foreign Affairs Committee released a report titled Never Again: The UK’s Responsibility to Act on Atrocities in Xinjiang and Beyond (Foreign Affairs Committee, 2021), but this fails to put forward a strategy that could change President Xi Jinping’s approach.

We will return to the question of whether ‘never again’ represents galvanising rhetoric, a realistic goal, an unachievable ideal or an amalgamation of the above. The P5 problem highlights that ‘never again’ is unable to galvanise political actors consistently across time and space – even actors who would proclaim themselves ‘champions’ of human rights and atrocity prevention. This really should not be of any surprise, but what makes this so problematic in the context of the P5 is that these states have been so closely associated with the implementation of the RtoP despite the institutional flaws of the international peace and security architecture that the RtoP does not address. While this architecture remains in place, ‘never again’ is unrealisable at least under the RtoP framework. Furthermore, as the other problems discussed in this article at length show, even if the veto power was reformed, or ‘never again’ was uncoupled from the RtoP, it is doubtful that it could be achieved.

Conclusion

When people were liberated from Buchenwald concentration camp, they held signs saying ‘never again’, which they hoped would mean something but as ‘The expanding parameters of never again’ section explained, this something has changed over time as the parameters of ‘never again’ expanded from (1) no more Holocausts, to (2) no more genocides, to (3) no more mass killing, to (4) no more genocide, crimes against humanity, war crimes and ethnic cleansing. This expansion has profound implications for its meaning and intended goal, yet this is rarely, if ever, discussed. Although there has been research on ‘never again’ in Cultural Studies, Sociology and History, there is far less from an IR perspective. This is needed for two reasons. First, the phrase is routinely invoked by academics, political elites, journalists and NGOs in relation to prominent IR debates over humanitarian intervention, the Responsibility to Protect and mass atrocity prevention. Second, since the term became mainstreamed in the 1940s there have been real-world developments that have implications for ‘never again’ and here the discipline of IR can help inform the debates over the term’s meaning and use. The authors identified and analysed five real-world problems: (1) the quantitative problem, (2) the nuclear problem, (3) regime change problem, (4) the weak state problem and (5) the P5 problem. Although this list is not exhaustive and some of these problems are more insurmountable than others, collectively, they underline the magnitude of the task at hand when it comes to mass atrocity prevention.

The expansion of the term’s meaning (‘The expanding parameters of never again’ section) has exacerbated what we view as unrealistic expectations because of the sheer number of mass atrocities it is now linked to (‘The quantitative problem’ section). This cannot be resolved by simply reducing the parameters of ‘never again’ because it does not address the nuclear problem (‘The nuclear problem’ section). Furthermore, the consequences of regime change have entrenched political unwillingness which dictates that ‘never again’ is even further out of reach (‘The regime change problem’ section). At the same time, the complexity of weak states highlights that even with the consent of the state in question the challenges are potentially insurmountable because of issues such as regional dynamics (‘The weak state problem’ section). Finally, the conflation of the RtoP and ‘never again’ has created new problems for the latter, namely, the institutional barriers embodied in the norm (‘The P5 problem’ section). This set of empirical realities makes it impossible to fulfil the objective of ‘never again’ at least for the foreseeable future.

While we are sympathetic to the idea that the parameters of ‘never again’ should be reduced to focus on just preventing genocide and let related ideas such as the RtoP norm cover the broader remit of mass atrocity crimes, two problems arise. First, it may be too late to put the genie back in the bottle, as its meaning has been expanded so much and that recapturing its earlier meaning may be illusive. This is not to suggest that it is impossible but just that reconstructing the meaning of the phrase is no easy task given that it has become commonly associated with many different crimes. This has been made even more difficult by the term being linked with movements and causes that we simply did not have scope to go into here. Second, although reducing the parameters of ‘never again’ would make it more feasible, without major global reform, it still seems unrealistic. For instance, how do ‘we’ prevent a nuclear state from perpetrating mass atrocities? Rather than uphold the commitment of ‘never again’, we conclude that it would be better to acknowledge and adapt to the permanency of mass atrocities in IR.