Two Responsibilities to Protect

Abstract

The purpose of this paper is to re-theorize the evolution of the Responsibility to Protect (RtoP) in the UN through to 2011, the apogee of liberal interventionism in the post-Cold War period. Contrary to a common argument in existing literature, and notwithstanding the adoption of the concept as an annual agenda item of the General Assembly, international contestation is not about implementation as neatly separated from meaning, but rather definition or interpretation. To better understand the boundaries of intergovernmental understanding, we need to interrogate the language or terms of the debate, particularly the ways in which those terms have been practiced. There have been two Responsibilities to Protect in international society. A discursive practice called Southern RtoP, traced through UN-based political dialogue, contests a meaning that has been prevalent for 20 years at least: that of Northern RtoP. This article shows evaluative nuance and data from the perspective of the Global South and provides a discursive history of an ongoing non-aligned protest against a NATO-associated theory of defeasible sovereignty.

Deux responsabilités de protéger

2021 marked 20 years since the publication of The Responsibility to Protect. ¹ Not much discussed but surely very relevant to this anniversary was the meeting that occurred in the Fifth Committee of the General Assembly on 30 December 2020. There Cuba proposed to defund and effectively cancel the UN Office of the Special Adviser on the Responsibility to Protect – from one view, to delete RtoP on the eve of its 20th anniversary.

‘There was no intergovernmental agreement, negotiated by Member States, to define that concept’, explained Ms Yaima de Armas Bonchang. ‘In line with her Government’s position of principle against genocide, her delegation fully supported the functions of the Office of the Special Adviser on the Prevention of Genocide’, ‘the oral amendment it wished to propose was not intended to undermine that Office’. ² But, as Cuba had pointed out ‘for more than a decade’, the General Assembly has ‘not decided on the concept of the responsibility to protect, its scope, implications and possible ways of implementation’. ³

In the end there were only 18 votes in favour, 80 votes against and 55 abstentions. ⁴ The results might be compared with those of a resolution introduced later that year, which proposed to include RtoP as an annual agenda item of the General Assembly; the latter passed by an overwhelming majority of 115 voting for, with 28 abstaining, and 15 against. ⁵ The failed amendment, like the successful resolution, might seem to indicate the degree to which the concept in fact continues to have a wide appeal and shared meaning among states. Yet there are compelling reasons to think otherwise, which pose a substantive and methodological challenge to prevailing accounts of the global humanitarian protection debate.

The purpose of this article is to respond to the challenge posed by Cuba and at least some of its supporters: to what extent has a common meaning of RtoP been reached in the practice or practices of states? What are the most significant limits of intergovernmental understanding? And do such discursive contestations amount either to the neglect or rejection of ‘responsible sovereignty’ and ‘mass atrocity prevention’, or might competing approaches to the content of this concept rather reveal something more positive about its multiple functions in the present? The answers I have in mind are intended to recast the ways in which we narrate the origins, development and institutionalization of RtoP in world politics, and how we ought to go about assessing and reforming humanitarian protection policies accordingly.

To better explain actual diplomatic debate in the 2020s, it is suggested that we start to think in the following sort of way: there have been two RtoPs in international society. I do not mean ‘RtoP’ and ‘RtoP Lite’ – RtoP as formulated in the 2001 report of the International Commission of Intervention and State Sovereignty (ICISS) and RtoP as it appears in the Outcome Document of the 2005 World Summit. ⁶ I mean Northern and Southern RtoP. To historicize these competing meanings would be to make significant revisions to the ways in which we interpret dissent and help to resituate the humanitarian protection debate in a much more complex, hybrid and de-centred context of international relations than the one in which the ICISS report was written. It would also expose, by reference to the deliberations of statespeople, certain ideological tendencies in recent depictions of the international debate around sovereignty, human rights, and the use of force. Considering the central role played by these ideas in ordering ongoing crises, and in the resurgence of military competition and alliance structures that may threaten longer-term peace, retrieving a more critical and non-aligned perspective on the rise and fall of liberal interventionism seems especially appropriate.

The main argument is that, contrary to a common argument in existing literature, there is not and has not been an ever-increasing convergence among states on the meaning of RtoP. There is of course a broad consensus on the existence of an international duty for the protection of civilians from atrocity crimes, but disagreement is not confined to matters of case-specific implementation. Northern and Southern meanings diverge in relation to the suspension, as distinct from the qualification, of sovereignty. Two RtoPs help us to see that the central dilemma has never been whether to intervene for humanitarian reasons, but whether to intervene in a way that invalidates sovereignty or countenances its invalidation. A theory of “indefeasible sovereignty,” in place of the often-conflated but different ideas of ‘qualified’, ‘limited’ or ‘conditional’ sovereignty, rejects the dualism of human rights versus sovereignty, but not for the reasons usually cited - not just because ‘sovereignty implies responsibility’, as all states agree and have long agreed, but because responsibility, as shared by the international community, need not imply control. ⁷ Emergency collective action for the protection of civilians with the presence of coercion and taken through the Security Council, in accordance with Chapter VII, yet without interference in local-political or essentially domestic affairs, is also a coherent expression of RtoP.

Is There a Distinctive Southern Practice of RtoP?

RtoP, writes Ramesh Thakur, ‘is not, and should not be, a North-South issue’. ⁸ Thakur concedes that in the late 1990s, a debate did clearly emerge ‘between the Global North, as represented by NATO, and the Global South, as articulated by the Non-Aligned Movement’, but a move from ‘humanitarian intervention’ to the ‘Responsibility to Protect’ allegedly ‘re-established an international consensus on the legitimate ends of the use of military power’. ⁹ Part of the reason, he argues, had to do with the resonance of RtoP with long-standing and cross-regional local traditions: ‘[m]any non-Western societies have a historical tradition of reciprocal rights and obligations that bind sovereigns and subjects’, ‘RtoP is rooted as firmly in their own indigenous values and traditions as in abstract notions of sovereignty derived from European thought and practice’. ¹⁰ Thakur points to modern and ancient examples: the 3rd-century rock edicts of Ashoka, the Mauryan emperor (‘this is my rule: government by the law, administration according to the law, gratification of my subjects under the law, and protection through the law’) and the constitution of modern India, particularly the ‘RtoP-type responsibility on federal and state government in its chapters on fundamental rights and directive principles of state policy’. ¹¹

Ashoka’s rock edicts and the like might suggest that practices of sovereignty as responsibility are very widespread across history and the world, but are these also practices of intervention? If we were to take RtoP to mean very generally that sovereigns have a duty to protect their citizens, no doubt all states would agree, but are we satisfied with saying that the meaning of RtoP stops there? Do statespeople themselves consider the implications of sovereignty as responsibility to be settled and widely agreed upon?

Clearly some do not, and the position is not only Cuban: in proposing the inclusion of RtoP in the official agenda of the General Assembly, for example, the argument was also explicitly Ghanaian and Australian. ¹² States voting for and against observed ‘deep differences’ on its ‘definition or implementation’, that ‘the very concept of the responsibility to protect is still being analysed and discussed’, that there was no consensus “on the conceptual framework of the notion’, or ‘on its interpretation or mechanisms for implementing it’. ¹³ ‘It is evident’, argued India,

that we need to reflect on the gaps in our understanding of the concepts behind the responsibility to protect and ensure that the quest for a more just global order is conducted in a way that does not undermine international order itself. ¹⁴

Jennifer Welsh shows there has been both procedural and substantive ways of contesting RtoP in the UN. ¹⁵ And she rightly reminds us that for all the recent emphasis on contested ‘implementation’ as opposed to ‘meaning and scope’, a great deal of literature has shown how patterns of implementation are meaning, too. ¹⁶ We can speak of agreement or convergence at a very high level of abstraction, but to define RtoP as a singular norm, even an unfinished or evolving one, is very often to conceal a set of differences which seem crucial to making sense of its actual international functions. ¹⁷ If the content of RtoP cannot be divorced from its contested use – from the diverse ways in which multiple actors both ‘implement’ and describe the concept of the responsibility to protect – then to know its place in international society, it seems necessary to better map out not just whether states subscribe to ‘the RtoP’, but how they have variously asserted its meaning. ¹⁸ An influential argument articulated by Alex Bellamy holds that ‘the key debates are now about how best to implement RtoP, not about whether to accept the principle’, and that accordingly there is ‘no longer serious principled disagreement between states on the merits of RtoP itself’ or ‘its meaning and scope’. ¹⁹ Yet to say so is to presuppose first, that meaning sits comfortably outside of practice, and second, a substantive and international conception of ‘RtoP proper’.

What is, or ought to be, RtoP proper? There are prescriptive and explanatory ways of approaching this question. But in accounting for the content of RtoP in global order and the UN, it is often with a particular sort of empirical assessment that we are concerned – one that accounts for ‘international society’ or the ‘diplomatic community’ of states. ²⁰ When analysing concepts at this particular level of normative meaning, it would not be enough to rely on conventional understandings adopted by IR scholars, transnational civil society networks, or originators or ‘entrepreneurs of the RtoP vocabulary. ²¹ To know RtoP in actual international society, and in actual global-political debate, it would instead be necessary to refer to the evolving, institutionalized practices of the statespeople to whom the ICISS proposal, World Summit agenda and related reports of the Secretary-General, for instance, have always been addressed. ²² And these practices would need to be traced and compared discursively: they would need to be located within contexts that provide insight into the terms within which practitioners have comprehended or explained their own actions.

The account that follows reflects a view of global ordering that is interpretivist and grounded by a specific set of hermeneutic commitments. ²³ It is concerned with the beliefs of statespeople and the practices that arise from official debates, speeches and formal agreements expressive of those beliefs. Political texts including meeting transcripts and reports are used as means of retrieving interpretations that actors have of the RtoP concept and that, in their recurrence, stability, and institutionalisation become part of the discursive fabric of the society of states. ²⁴ These practices might hence usefully be called discursive practices. UN verbatim and summary records are particularly relevant in tracing RtoP as discursive practice. As we will see, the process of deciding ‘what is RtoP’ does not precede debate in places like the General Assembly, in important, self-conscious, and explicit ways it is the debate itself. In this way I am reaffirming a reading both of the contingency of major concepts of liberal global governance and of contestation and change through the self-descriptions of situated agents. ²⁵ It is through interpretation of these historical self-descriptions offered by states and their representatives that South/North practices of RtoP may be discerned at politically significant degrees of abstraction.

This is not to say that two RtoPs are exhaustive of all impressions on the matter or that there are no exceptions; it is of course the case that some smaller and non-aligned states take positions that either fit the Northern construction or deviate from the Southern one as I understand it, and of course some larger states do play quite non-Northern roles (Russia and arguably China being the most significant and well-known examples). Conceivably there are also internal disagreements at higher levels of specificity. For example, we could identify additional layers of difference and emphasis among groups of practitioners of Northern and Southern RtoP. ²⁶ But the key point is, in pulling together and reconciling regular viewpoints, these rival discursive forms help us to better understand the nature of the intervention problem since the end of the Cold War. Southern RtoP is “Southern” not only because its practitioners have tended to hail from Asia, Africa, and Latin America, but because its substance is consistent with the historical values of Southern intergovernmental projects like the Non-Aligned Movement. Far from being merely regional or regionalist, it relates to the Global South as an international political identity that tells us something about the global. ²⁷

So it is true that, in 2005, at the largest gathering of Heads of State and Government in history, UN member states adopted the language of ‘the responsibility to protect’, which soon became commonplace within the UN Secretariat as well. ²⁸ RtoP did come to be accepted by both ‘the Global North, as represented by NATO, and the Global South, as articulated by the Non-Aligned Movement’. ²⁹ At this point all did become interventionists, at least in the general sense of the word that Kofi Annan had in mind in his pivotal 1999 speech. ³⁰ However the global adoption of shared vocabularies did not imply the emergence of a single substantive meaning or true and deep consensus. Rather there co-existed and co-exists today multiple conceptions of sovereignty as responsibility, two of which seem particularly influential and should be laid bare in the records of the General Assembly and Security Council.

Northern Practice Encounters Southern Practice

The Northern formulation is more well-known. Emerging in the late 1990s, it could be summarized by the assertion that: ‘the last few years of this century have disproved the notion that people and human freedoms take second place to state sovereignty’. ³¹ Sovereignty ought to yield to higher imperatives or obligations, and in legitimate-interventionist scenarios, there is a culpability to be associated with sovereigns – both for having failed to protect civilians, and for having obstructed international action by raising sovereignty ‘as a shield’:

There is one lesson to be learned from these years of turmoil: actions to prevent and repress the most serious violations of human rights may take precedence over respect for national sovereignty. No government can hide behind the shield of its own borders. ³²

The claim rests, and rested in 1999, for instance on an assumption about which problems were located within the domestic and the international. ‘The Kosovo crisis . . . obliges us to ask a sensitive question about the limits to the right of the international community to intervene in the internal affairs of a state’. ³³ Intervention protects human rights, and especially the right to life, in the internal domain as if it were already and exclusively there. To say so was to adopt earlier strands of thinking about the droit d’ingérence:

If there is one lesson for our Organization to learn from the twentieth century, it is that for no state can the massacre of its own people be considered an ‘internal affair’, under any pretext. This legal formalism would ultimately amount to admitting that, as the head of the UN Interim Administration in Kosovo, Bernard Kouchner has said, it would be ‘legitimate, although not elegant, to massacre one’s own people’. ³⁴

This retelling of the twentieth century was historically misleading – genocide, crimes against humanity and large-scale human rights abuses were not widely considered ‘internal’ or purely ‘domestic affairs’, as the international and UN-based struggle against apartheid in South Africa for instance, made very clear. Yet the feat of Northern RtoP was never about accurately recalling the past. ³⁵ Rather what was important was its establishment of a particular set of terms of debate: ‘we have come to understand that absolute sovereignty and total non-interference are no longer tenable’ (as if they had once been), ‘this implies a rethinking of the principles . . . that have governed the community of nations for over three centuries’; the new international order would abolish the ‘sovereign right to ethnic cleansing and genocide’; it was time to declare ‘sovereignty cannot mean impunity for genocide and human rights abuses’. ³⁶

Such a broad Western position was compatible with a classical theory of humanitarian intervention, according to which, in emergency, intervention might derogate from the right to sovereignty. RtoP is of course often distinguished from humanitarian intervention, by both states and academics alike. But sovereignty as responsibility, or at least a prominent strain of its practice, does trace its origins to the humanitarian intervention debates of the 1980s and 1990s. What links these conceivably distinct concepts over time is a particular understanding of the relationship between sovereignty and human rights.

A Northern understanding of the sovereignty-human rights question, cutting across classical humanitarian intervention and some formulations of RtoP, had three major points. First, that humanitarian disasters were caused by malevolent or negligent agents in the form of state leaders – ‘dictators and murderers’ whose sovereignty, ‘a shield’, provided a kind of ‘license to kill’. ³⁷ Second, that affairs historically understood to belong to the international were to be shifted into the domestic: it was legitimate to enter the domestic because it was there that we could find the problem of humanitarian protection. And hence third, that to effectively solve that problem and to prevent large-scale human rights abuses, it was legitimate and even necessary to infringe upon the sovereign equality of states, including through the use of force. The other, more commonly discussed changes of the period – unilateral versus multilateral authority, coalitions of the willing, veto restraints – were also important, but the density of agreement upon them varied (consider for instance Germany’s scepticism of unilateralism), and all mattered less than the logic not of who intervenes, but what are the limits of intervention. ³⁸

And if this exception to sovereignty was essential to Northern practice, then it was at the heart of a Southern counter-practice. This also appeared in the 1990s. Again, other problems were being described too – about for instance just cause (which human rights, which atrocities), the role of international law, of regional organizations – but these were less significant sites of contestation. The concern was less about a legitimate justification to intervene than about the character of legitimate intervening; an unacceptable politicization of intervention occurred when the objectives of the protectors moved from the international-humanitarian into the internal-political. The moment at which its application became non-objective, selective and partial was the moment at which sovereign equality was violated. All the while the suspension of the right to sovereignty could be distinguished from the qualification of sovereign privilege, and from the argument – very old – that all states and the UN had a shared duty to protect human rights.

In its classical sense, ‘humanitarian intervention’ has long been opposed by states belonging to the Non-Aligned Movement. This much is commonly acknowledged. ³⁹ At the 13th Ministerial Conference in Cartagena, 8–9 April 2000: ‘We reject the so-called “right” of humanitarian intervention, which has no legal basis in the UN Charter or in the general principles of international law’. ⁴⁰ A few days later, at the South Summit in Havana, the member countries of the Group of 77 and China re-adopted this rejection verbatim and additionally requested that the Chairman of the Group of 77 and Chairman of the Non-Aligned Movement jointly coordinate their consideration of the concept. ⁴¹

But to reject the ‘right of humanitarian intervention’ was not necessarily to reject international duties to protect civilians. ⁴² This point is less commonly acknowledged, particularly outside of the context of subsequent support for RtoP. ⁴³ In the late 1990s, the problem of genocide prevention was, from a non-aligned perspective, largely a story of political will and enforcement, not about an overcoming of sovereignty. To states like Indonesia, it was too easy to exonerate international actors from their liabilities in previous disasters:

. . . to blame this principle [sovereignty] for the inability of the Organization to come to the aid of suffering humanity anywhere is to distort the truth. To extend such assistance is a solemn obligation. Indeed it is imperative. However, there are many reasons why the UN often finds itself ineffective in situations that require it to act swiftly and decisively. These include resource constraints, lack of political will, selectivity, misplaced media attention, and disfunction in the working of bodies as the Security Council and in implementing mechanisms. These have nothing to do with the principle of sovereignty. ⁴⁴

For all the period’s references to Rwanda, for instance, quickly forgotten was that Nigeria, on behalf of NAM, had proposed a draft resolution under Chapter VII that would have empowered UNAMIR to enforce ‘law and order and the establishment of transitional institutions’ while protecting civilians. ⁴⁵ This was a draft to ‘increase the strength of UNAMIR and to revise its mandate’ – precisely what Roméo Dallaire considered necessary to stop the killing. ⁴⁶ What was needed to avert another Rwanda was not a Northern doctrine – one that suspended, in an unequal world, the rights and freedoms guaranteed by the principle of sovereign equality – but the will to uphold commitments already made. ⁴⁷ In 1998, at the Durban Summit, the Non-Aligned Movement declared:

In four and a half decades, the world has changed vastly from the days of the Bandung meeting. Yet the principles laid down by the founders of the NAM remain valid, and the ideas, goals, and vision articulated then, continue to guide our movement. As we mark the 50th anniversary of the Universal Declaration of Human Rights, nothing should be used as a convenient mask to hide genocide, gross violations of human rights and crimes against humanity, nor should human rights be used as a political instrument for interference in internal affairs. ⁴⁸

Three overlapping points were coming into view: 1) a rejection of the human right versus sovereignty dualism, 2) a re-assertion of the duties of states and the international community both to protect universal human rights (including, if necessary, through collective action) and to refrain from interference in internal affairs and 3) a belief that these duties were consistent with each other because internal or essentially domestic affairs could be distinguished from affairs shared by the international community. Similarly, consider the arguments of the Chinese delegation during the 1999 debates:

. . . putting an end to these conflicts and crises and eliminating their root causes is the ardent desire of the peoples of the countries concerned, as well as a legitimate concern of the international community. However, such arguments as ‘humanitarian intervention’ and ‘human rights over sovereignty’ that cropped up recently set up human rights against sovereignty. ⁴⁹

Emphasis on the ‘ardent desire of the peoples of the countries concerned’ in some ways corresponded to logics of popular sovereignty that informed earlier anti-colonial organizing, but notably the vocabulary of self-determination was not being used here. ⁵⁰ Still the sovereignty and will of the peoples of the countries concerned remained a guiding principle: legitimate intervention was international involvement in the protection of civilians, but never involvement that casted into doubt the equal right of all nations and peoples to freely determine their own political, economic and cultural futures. ‘The history of China and other developing countries shows that a country’s sovereignty is the prerequisite for and the basis of the human rights that the people of that country can enjoy’. ⁵¹

Six months later, at the previously mentioned 13th Ministerial Conference of the Non-Aligned Movement in Cartagena, a new but never flaunted strategy seemed to shine through: if ‘the interpretation of the Charter is broadened to make room for humanitarian intervention’, then ‘the modes of that intervention should also be broadened’. ⁵² The goal was to ensure that ‘any new orientation of humanitarian assistance should be based on unconditional respect for the Charter of the United Nations and international law’ and ‘must be inspired by the basic criteria of neutrality and impartiality’. ⁵³ It was not that renewed Western activism for international humanitarian protection ought to be blocked outright, but that the South ought to seek to promote a way of conceiving of intervention, or at least of protecting civilians abroad from atrocities that looked very different from Western practice. As South African Foreign Affairs Minister Nkosazana Dlamini-Zuma, then-Chair of the Movement, explained: ‘[w]e need to define humanitarian intervention, otherwise others will define it for us’. ⁵⁴

Re-Evaluating the ICISS Report: Responsibility and Control

From a Southern perspective, a few key themes were hence missing from the ICISS report: a distinction between sovereignty’s suspension and its qualification (between first, sovereignty’s invalidation and, second, curtailments on the range of privileges afforded to sovereigns); sensitivity to a line between domestic including local-political affairs and international affairs (including the prevention of large-scale human rights abuses); and the desirability of interventions for humanitarian protection, including through collective action if necessary, in a manner that was described as neutral and impartial. Related ideas, such as consent and the nature and sources of national consent in large-scale humanitarian emergencies, were explicitly outside of the ICISS purview: ‘the kind of intervention with which we are concerned in this report is action taken against a state or its leaders, without its or their consent, for purposes which are claimed to be humanitarian or protective’. ⁵⁵

Sponsored by the Canadian government in September 2000, with a membership intended ‘to fairly reflect developed and developing country perspectives’, the ICISS sought to ‘build a broader understanding of the problem of reconciling intervention for human protection purposes and sovereignty’ and ‘develop global political consensus’. ⁵⁶ And it set out to do so by ‘shifting the terms of the debate’ such that its ‘proposed change in terminology is also a change in perspective’: to move from ‘humanitarian intervention’ to ‘a responsibility to protect’ was to move from an understanding of ‘sovereignty as control’ to one of ‘sovereignty as responsibility’. ⁵⁷ But what should be confessed is the possibility that its recommendations sanctioned something else, as well: what might have been called, ‘responsibility as control’.

The central theme of the ICISS report comprehended the following:

sovereign states have a responsibility to protect their citizens from avoiding catastrophe – from mass murder and rape, from starvation – but when they are unwilling or unable to do so, that responsibility must be borne by the broader community of states. ⁵⁸

Still, there was some equivocation as to the nature of what it meant for the international community to bear responsibility: yes, three specific responsibilities were implied by the RtoP (to prevent the crisis, to react to its outbreak, to rebuild afterward), but if, when a state was unable or unwilling to protect, ‘it becomes the responsibility of the international community to act in its place’, exactly what sort of transfer of authority was involved in a transfer of humanitarian responsibility? ⁵⁹ While the Commission acknowledged that ‘the primary responsibility to protect’ rested with the state, might the international community, in fulfilling the residual RtoP in extreme circumstances, take on still more national- or local-political responsibilities?

The substance of the RtoP was described by the Commission as ‘the provision of life-supporting protection and assistance to populations at risk’. ⁶⁰ Its relationship with territorial integrity, self-determination and choice or replacement of government was discussed in the report’s section on ‘right intention’:

The primary purpose of the intervention must be to halt or avert human suffering. Any use of military force that aims from the outset, for example, for the alteration of borders or the advancement of a particular combatant group’s claim to self-determination, cannot be justified. Overthrow of regimes is not, as such, a legitimate objective, although disabling that regime’s capacity to harm its own people may be essential to discharging the mandate of protection – and what is necessary to achieve that disabling will vary from case to case. ⁶¹

Yet just as the ordinary meaning of guardianship has at least two connotations, so too does this seem true of the international RtoP. ⁶² Might keeping guard of people, through the ‘provision of life-supporting protection’, also confer the position or office of becoming more generally responsible for the care of those people? ⁶³ The latter meaning seemed to blur into a logic of trusteeship whereby ‘taking responsibility’ or fulfilling the RtoP granted – for a transitional period – a right to control. So, by inheriting an RtoP from threats of large-scale loss of life, might guardians inherit a tacit responsibility over the affairs of the guarded? To say so would be to suspend competing claims to or sources of supreme authority.

In fact, in the 2001 report, this was explicitly admitted: ‘Intervention suspends sovereignty claims to the extent that good governance – as well as peace and stability – cannot be promoted or restored unless the intervener has authority over a territory’. ⁶⁴ The Commission did stress that ‘the suspension of the exercise of sovereignty is only de facto for the period of the intervention and follow-up, and not de jure’. ⁶⁵ But not only was there a concern here about slipping away from a logic of the preservation of life to one of ‘good governance’ as well as ‘peace and stability’, there were also good reasons to doubt the relevance of this particular distinction between the de facto and de jure. For example, cited by the Commissioners was the case of Yugoslavia: ‘Yugoslavia could be said to have temporarily had its sovereignty over Kosovo suspended, though it has not lost it de jure’. ⁶⁶ This was not clearly true as of 2001, and it became much less probably true after Kosovo’s declaration of independence from Serbia in 2008. And one very central concern, for the South, was always about intent as distinguished from motive; about whether the intent to control internal or local-political affairs, even as a means to effect a humanitarian result, was ever legitimate.

This process of suspending sovereignty, in the course of transferring RtoP from the state to the international community, is what is meant by responsibility as control. The language of defeasibility – expressive of a right that is open to forfeiture or annulment in the fulfilment of some condition – is equally useful. While qualified sovereignty refers to any strain of sovereignty as responsibility, only defeasible sovereignty is compatible with responsibility as control – with a particular process by which sovereignty is thrown off or temporarily suspended. ⁶⁷ And it is in light of responsibility as control that we can understand the rejection, in the South, of the first of two RtoPs: it is basically in reference to responsibility as control that states later allege that the RtoP project is ‘full of traps’ concealing ‘unmentionable motives’. ⁶⁸

Intervention was in all cases to endeavour ‘to sustain forms of government compatible with the sovereignty of the state’ but, from the view of Northern practice, such an undertaking never represented a hard limit: one sub-section of the ICISS report was devoted to trusteeship and international administration (paragraph 5.22: ‘useful guidelines for the behaviour of intervening authorities during a military intervention in failed states, and in the follow-up period, might be found in a constructive adaptation of Chapter XII of the UN Charter’; paragraph 5.24: ‘[t]here is always likely in the UN to be a generalized resistance to any resurrection of the “trusteeship” concept . . .’), and elsewhere it was acknowledged that ‘occupation of territory may not be able to be avoided’ (as long as there was a commitment to ‘returning the territory to its sovereign owner at the conclusion of hostilities or, if that is not possible, administering it on an interim basis under UN auspices’). ⁶⁹ In the penultimate chapter of the report – ‘Responsibility to Rebuild’ – the promise to return the RtoP to the state was described not as a returning of the obligation to defend civilians from atrocity or threats to life, but ‘returning the society in question to those who live in it, and who, in the last instance, must take responsibility together for its future destiny’. ⁷⁰

RtoP was no doubt a ‘linking concept that bridges the divide between intervention and sovereignty’, but the way in which those concepts were being linked suggested the possibility of a mechanism by which sovereignty became defeasible, according to which it might be come legitimate in the course of protecting lives – its overarching objective – to overthrow governments or regimes, or to approve of foreign occupation such that it was conceivable to return ‘the society in question to those who live in it’ at some later date. ⁷¹ Hence from a Southern view: to what extent had the report really been authored by an ‘International Commission on Humanitarian Intervention’ (as ICISS was in fact initially called by Canadian officials)? ⁷² If a logic of trusteeship was a defining feature of classical humanitarian intervention, it seemed latent in a Northern RtoP. ⁷³

The basic transformations from the ICISS report to the World Summit Outcome Document (WSOD) have been much discussed: the narrowing of threshold criteria or just cause to four ‘mass atrocity crimes’; the disappearance of post-intervention obligations and precautionary criteria; the move, as construed in the Secretary-General’s 2009 report, from three responsibilities (prevent, react, rebuild) to three pillars. ⁷⁴ Scholars also frequently note, correctly and importantly, that the WSOD did not sanction recourse to ‘coalitions of the willing’ when ‘the Security Council fails to act’. ⁷⁵ Still, the question of authority in the evolution of RtoP has on the whole been poorly understood in actual discursive contexts of states. ⁷⁶ Authority refers not simply to sources of authorization, as in the ‘who will intervene’ or ‘who will approve of intervention’, but rather: what sort of authority might RtoP grant to interventionists over the intervened upon. Too often conflated with ‘whether RtoP’ is the separate question: whether ‘to take responsibility’ means to take control, in the course of keeping guard.

Constitution and Contestation in the UN

In 2005 there was adopted a ‘responsibility to protect’ in the UN. But if we look closely, it was not a Southern formulation at the expense of a Northern one, or vice versa. Nor was it a fusion of two viewpoints into one, without their preservation. Despite the global adoption of paragraphs 138 and 139, incompatible South/North practices were read into the same document, and persevered. That the World Summit did not take an explicit stance on responsibility as control is among the most important reasons why the WSOD was accepted by both sides. ⁷⁷

Accordingly, it was not as if, in the patterned interpretations of states, the WSOD necessarily excluded the most controversial features of classical humanitarian-interventionist practice. At the Millennium Summit of 2000, during his unveiling of ICISS, Canadian Foreign Affairs Minister Lloyd Axworthy explained the Commission’s objectives in the context of a dualistic choice between sovereignty or intervention:

Nothing so threatens the United Nations’ very future as this apparent contradiction between principle and power; between people’s security and governments’ interests; between, in short, humanitarian intervention and state sovereignty. ⁷⁸

At the World Summit of 2005, the Icelandic delegation argued similarly: while ‘sovereign equality’ and ‘the promise of living in peace as good neighbours’ were ‘valuable principles’, ‘Iceland has never looked on the United Nations as a mechanism solely for safeguarding sovereignty’. ⁷⁹ For decades ‘key commitments to the peoples enshrined in the UN Charter have not been given their due weight’. ⁸⁰ The WSOD ‘makes significant strides towards redressing the imbalance. In particular, we have established the concept of the responsibility to protect’. ⁸¹

Patterns of Southern contestation at the Millennium Summit also resembled those of the 1999 debates (‘we disagree with the assumption that the principles of sovereignty and humanitarian international law are at variance’; where intervention did occur, ‘[a]ction must be confined to the saving of lives, not the overthrowing of governments’, ‘intervention . . . must never be used as a guise for unwarranted interference in internal state governance’). ⁸²

Later, at the World Summit appeared again the South’s assertion that while sovereignty as responsibility did imply qualified sovereignty, it did not imply defeasible sovereignty; yes, an international responsibility to protect, but never responsibility as control. In this way it was always a consistent argument, not some muddle or confusion, to say, in reference to RtoP, that ‘any intervention must give due recognition to the Charter principles pertaining to sovereignty, territorial integrity and non-interference in internal affairs’. ⁸³ A comparable formulation was notably adopted in the text of the AU’s Ezulwini Consensus earlier that year. ⁸⁴

From one view, then, the NAM strategy had not really changed; quietly the aim was to re-define humanitarian protection in a context in which meaning had not been settled, in which Northern activism for its RtoP was predominant, but also one in which the WSOD had multiple interpretations. For example, when the NAM ‘observed similarities between the new expression “responsibility to protect” and “humanitarian intervention”’ at its 2003 summit in Kuala Lumpur (as well as elsewhere), it also ‘requested the Co-ordinating Bureau [of the NAM] to carefully study and consider the expression “the responsibility to protect,”’ particularly on the basis of ‘territorial integrity and national sovereignty of States’. ⁸⁵ And in Havana, 2006, rather than clearly opposing RtoP, the Movement agreed to ‘remain seized of the matter’ while bearing in mind ‘the principles of the UN Charter and international law’ including ‘respect for the sovereignty and territorial integrity of states’. ⁸⁶

Likewise, uncertain meaning was an important basis upon which the NAM opposed the creation in 2008 of a UN Office on RtoP and UN Special Adviser on RtoP at the Undersecretary-General level. On 7 February 2008, the Cuban delegation in the Fifth Committee had protested that: ‘[i]n a letter dated 20 June 2007, the Chairman of the Coordinating Bureau of the Movement of Non-Aligned Countries had requested the Secretary-General to take into account, in any decision on that very sensitive matter’, the sensitive matter of RtoP, ‘the fact that Member States had not completed their deliberations’. ⁸⁷ If there were to be a Special Adviser, their role would be to ‘assist the General Assembly in its consideration of the issue rather than to promote the idea of the RtoP’ as if it had been adequately or wholly formed. ⁸⁸

It was precisely because of what the Secretary-General’s Office itself had described as the ‘manifold concerns regarding the conceptual range and application, and the practical implications’ of RtoP that its ‘crystallized’ and more ‘concrete form’ was understood as a goal or priority of the UN, not yet an international-political fact. ⁸⁹ During meetings between ‘senior officials of the Executive Office of the Secretary-General’ and several permanent representatives of non-aligned countries, as well as ‘discussions attended by the Chairman of the Coordinating Bureau of the Non-Aligned Movement’, ‘permanent representatives had expressed reticence over the “concept” of the “responsibility to protect”, arguing that there was no consensus on the matter’. ⁹⁰ Hence, on ‘the question of mandates’, the Special Adviser’s work was always to encompass an ‘examination of how to further the ideas contained in paragraphs 138 and 139’ – paragraphs that included a call for the UNGA to ‘continue consideration’ of the RtoP while ‘bearing in mind the principles of the Charter and international law’. ⁹¹

And again, the need to ‘further the ideas’ of RtoP was particularly true of the place of sovereignty and its suspension. As the Secretary-General put it in his 2009 report on the subject, ‘the responsibility to protect is an ally of sovereignty, not an adversary’. ⁹² If the ‘ally of sovereignty’ argument referred to the fact that according to Pillar One, the primary RtoP belongs to states and that sovereignty entails rights but also responsibilities’ then of course all could agree. The ground became less certain as explanation went further.

‘Neither concerns about sovereignty nor the understanding that sovereignty implies responsibility’, noted the SG’s report, ‘are confined to one part of the world’. ⁹³ But the precise contents of different sets of concerns did vary across regions. Their substance lay behind or underneath the paragraphs of this and subsequent SG reports, in one sense literally – as was the case with the footnotes of paragraph 8, containing writings by then-Special Adviser on RtoP Edward Luck. ⁹⁴ These are writings that recall ‘the conditions under which rulers surrender the right to govern their people and other nations become free to intervene’; Kofi Annan’s comments (post-World Summit, notably later retracted) that RtoP had produced ‘a new equation, with “human life, human dignity, human rights raised above even the entrenched concept of state sovereignty”’; the arguments of Francis Deng et al. (reaffirmed in the SG report’s seventh paragraph) that ‘living up to the responsibilities of sovereignty becomes in effect the best guarantee of sovereignty’; and readings of Hobbes to the effect that when a sovereign can ‘no longer fulfill the function for which he or she was given power’, that sovereign is ‘no longer owed obedience, is no longer indeed a sovereign’. ⁹⁵

The emergence of a kind of neo-Hobbesian logic to justify a transferring outwards of authority – from domestic subjects towards international subjects – seemed to realise some of the very worst fears of the developing countries. And their suspicions of RtoP were mounted further by the report’s suggestion that

Member States may want to consider the principles, rules, and doctrine that should guide the application of coercive force in extreme situations relating to the responsibility to protect. This issue was addressed in the 2001 report of the ICISS and by my predecessor, Kofi Annan, in his 2005 report ‘In larger freedom’ . . . ⁹⁶

The problem for developing states always had to do with the fuller meaning of Pillar Three: ‘we must continue to think in a concerted way about the third pillar’ so as to affirm that ‘no juridical norm can legally justify humanitarian intervention by the Security Council under Chapter VII of the Charter’. ⁹⁷ Coercive intervention could be acceptable but had to be ‘fully consistent with the Charter. That means, inter alia, recognizing that a State’s responsibility to protect does not qualify State sovereignty’ – in the sense that sovereign equality was never to be suspended – but also that sovereignty does ‘not exempt the state from its obligation to protect its population. On the contrary, it is from that very attribute that such obligation derives’. ⁹⁸ RtoP ‘must be implemented pursuant to premises that do not undermine the guarantees and sovereignty of states’, meaning there were purposive limits to the privileges available to foreign protectors even in crisis zones: ‘the implementation of the responsibility to protect should not contravene . . . the principle of non-interference in the internal affairs of states’. ⁹⁹

Hence ‘it is critical that our discussions not be reduced to the simplistic dichotomy of states on the one side insisting on absolute sovereignty and, on the other side, RtoP proponents demanding that states surrender absolute sovereignty’. ¹⁰⁰ Clearly ‘we are all united behind our fundamental desire to protect innocent people and to prevent another Rwanda’. ¹⁰¹

But if RtoP embraced responsible as distinct from absolute sovereignty, what was left, for the non-aligned or Southern group, was to insist upon differences between responsible sovereignty as qualified and as defeasible. To assume otherwise might be to yield, surreptitiously, to the worst excesses of the old humanitarian-interventionist form. The Serbian representative could not restrain himself from quoting Martti Ahtisaari, the former President of Finland and subsequent UN Special Envoy for the Kosovo Status Process, who:

. . . in an interview with CNN on 10 December 2008, gave his view of the Responsibility to Protect. After acknowledging the fact that the General Assembly had ‘accepted the principle of RtoP in 2005’, he went on to justify it by saying that ‘if a dictatorial leadership in any country behaves the way Milosevic and company did vis-à-vis the Albanians in Kosovo, they lose the right to control them anymore’. We wonder if such interpretations of the noble concept of the Responsibility to Protect truly lead us away from the supposedly defunct concept of humanitarian intervention. ¹⁰²

Responsibilities While Protecting

So far, the story has focused on two divergent practices – two RtoPs – without denying the reality of secondary disagreements (concerning double standards, voluntary veto restraint, early warning systems and rapid response, development goals including food security, etc.) present in informal interactive dialogues of the UNGA, open debates in the UNSC on the Protection of Civilians in Armed Conflict, annual reports of the UNSG etc. Primarily, I have argued, international contestation had to do primarily with a theory of defeasible sovereignty, as a kind of guardianship that was also a trusteeship. That theory reached its high point in 2011, a year with which scholars of intervention are very familiar and which should only be contextualized briefly.

It is in Libya above all that Southern suspicions of the Northern approach to the RtoP concept (whereby, if a state is unwilling or unable to protect its people, ‘they lose the right to control them anymore’) are vindicated, and it is from this point forward that the conversation takes a new turn. ¹⁰³ It would be wrong to suggest that RtoP ‘died’ in Libya – clearly, for most states, it did not, and while RtoP language has since declined in usage, it is embedded in a broader normative and institutionalized structure that makes it unlikely to disappear completely. ¹⁰⁴ Rather what was eroding or decaying was the political fortune, in the UN and the global diplomatic community, of its Northern conception specifically. Arguably the rejection of the Libya model was a rejection of responsibility as control.

In the South, that is more or less how the intervention came to be understood: NATO had gone ‘far beyond the letter and spirit of resolution 1973’ by ‘[a]busing the authorisation granted by the Council to advance a political regime-change agenda’. ¹⁰⁵ All agreed that intervention he purposes of protecting innocents from large-scale violence was desirable and approved – if we mean intervention as international involvement, even coercion, but not as control over the internal affairs of Libya. However, from the Southern view, NATO’s protective measures became outright participation in civil war: a decisive and politicized attack – particularly against loyalist forces and strongholds – as opposed to an impartial humanitarian defence of civilians supporting either the TNA or the government. ¹⁰⁶

A Northern reply to this Southern view was basically, to quote the ICISS report, that while ‘the overthrow of regimes’ ought never to be a core interventionist motive, the removal of an atrocious regime might legitimately become ‘essential to discharging the mandate of protection’. ¹⁰⁷ ‘That is what we did in Libya . . . By striking Colonel Al-Gaddafi’s forces before they entered Benghazi, France and its partners helped to prevent a massacre there . . ..’ ¹⁰⁸ And even after Benghazi had been saved, humanitarian risks remained; Gaddafi’s complicity in violence against his own population was clear, he could not be trusted to refrain from further atrocities, and he appeared unlikely to abide by a truce. The next question was whether he had therefore lost the legitimacy to rule, perhaps even in the eyes of the Libyan people. To this many governments answered, as early as 15 April, with an unmistakable yes. ¹⁰⁹

Brazil’s Responsibility while Protecting (RwP) proposal could be said to have arrived in the context of this clash of Northern and Southern RtoPs. ¹¹⁰ True, RwP was not just a response to Libya, but it contained within it the makings of a strong critique of NATO’s intervention and the ‘Libya model’, and we might see it as a kind of channel through which to advance Southern-RtoP arguments. ¹¹¹ At the same time, it was written in such a way so as not to expressly exclude a Northern reading of RtoP. This was its political feat: to create space for a confluence of Northern and Southern practices, much like the WSOD had done, while also pressing upon themes of interest to the South especially, including jurisdiction and authority.

There were important components of Brazil’s concept paper that highlighted long-term prevention and non-aggravation of existing conflicts (‘our collective point of departure should resemble the Hippocratic principle of primum non nocere – first, do not harm’) but, viewed through the lens of two RtoPs, it was RwP’s emphasis upon accountability to Security Council mandates that mattered most (‘enhanced Council procedures are needed to monitor and assess the manner in which resolutions are interpreted and implemented to ensure responsibility while protecting’). ¹¹² According to this interpretation, by drawing attention to accountability in implementation, and to the monitoring of mandates under RtoP’s third pillar, RwP proposed a procedural path through which to protect the sovereign equality of states.

To be clear the RwP document did not explicitly mention the indefeasibility of sovereignty – perhaps this is the one reason why RwP was not immediately embraced by the BRICS as a whole. ¹¹³ But it was arguably it was also because of this reticence that RwP became tolerable to at least some adherents to Northern RtoP as a potentially constructive contribution. There emerged two key readings of the precise import of RwP’s idea of accountability in implementation. On the one hand, accountability in implementation resembled traditional jus ad bellum provisions that defined ‘other precautionary criteria’ in the ICISS report’s chapter on Responsibility to React. ¹¹⁴ On the other hand, RwP seemed to entail a responsibility to respect the line between domestic and international affairs. While protecting, there was a duty to stay out – to stay out of local political affairs while ensuring the safety of innocents or, more provocatively, to respect self-determined futures.

The latter reading was often the sort of RwP that we saw adopted used among emerging powers. ‘[R]egime change and the arming and harming of civilians cannot be justified in the name of protecting civilians’, explained South Africa before the Security Council in November 2011, ‘those entrusted with such responsibilities must uphold them while protecting civilians, as clearly stated by the representative of Brazil’. ¹¹⁵ The Chinese delegation argued similarly:

The protection of civilians falls within the realm of humanitarianism. There should be no political motives or purposes involved, including regime change. For that reason, how to strictly and effectively monitor the implementation of Security Council resolutions has become an important and urgent issue. China welcomes and will carefully study the concept paper proposed by Brazil and actively supports discussions considering that paper. ¹¹⁶

In the same meeting India refers to themes of accountability and ‘monitoring of the manner in which the Council’s mandates are interpreted’. ¹¹⁷ ‘The Council’s recent actions have brought to the fore a considerable sense of unease about the manner in which the humanitarian imperative of protecting civilians has been interpreted for actual action on the ground’. ¹¹⁸

The actions of the Council and the international community should facilitate engagement among warring factions in a conflict situation in a nationally-owned and inclusive political process and not complicate the situation by threats of sanctions, regime change, et cetera. This inclusive approach to national reconciliation, anchored in state sovereignty, is the only way to move forward and ensure the protection of civilians in an effective, pragmatic and enduring manner. ¹¹⁹

An illegitimate application of RtoP seemed to be intervention into the territory of a state, with the agreement either of the receiving state or Security Council, that contravened the conditions provided for in the agreement. And the expectation was that the conditions provided for in an appropriate Protection of Civilians mandate would also be conditions that reflected, to use the words of Resolution 1973, a ‘strong commitment to . . . sovereignty, independence, territorial integrity, and national unity. . .’. ¹²⁰ According to some, such a mechanism might enable the sabotage of RtoP. To others, it would mark the concept’s progressive institutionalisation in line with the principles of the UN Charter.

Conclusion: Re-Interpreting the Global Debate on humanitarian protection

The argument has been that the extent to which RtoP is or ever was the subject of international cultural convergence must be understood within the context of politically significant discursive practices. A shared meaning or norm of RtoP cannot be said to exist insofar as it assumes the content of one of these practices but not the other. While all agree on the qualified nature of sovereign rule, particularly in the sense that sovereignty implies a responsibility to protect civilians, the emergency transfer of the responsibility to protect from the state to the international community need not imply a de facto or de jure suspension of sovereignty. Nor does insistence upon sovereign equality necessarily defy sovereign responsibility. From the New-Interventionist organizing of the 1980s and 1990s, to the high point of liberal interventionism in 2011, there re-emerged distinctly Northern and Southern interpretations of international humanitarian protection.

The post-Cold War period did witness a bid for a sovereignty that was responsible, but this bid was also, and much more controversially, for a sovereignty that was defeasible. And while the latter form of responsible sovereignty seems relatively novel in the history of international society, its appeal never globally diffused. It is not that Southern RtoP abrogates the use of force for humanitarian purposes under Chapter VII. In fact, for supporters in the South, more is to be done on precisely the level of emergency collective action, though in a different chord: ‘in this context it is pertinent to mention that we find several member states all too willing to expend considerable resources for regime change in the name of protection of civilians’ yet ‘unwilling to provide minimal resources, like military helicopters, to the United Nations peacekeeping missions’. ¹²¹ Where Northern RtoP appeared to accept a process by which, if a state is unwilling or unable to protect its people, ‘they lose the right to control them anymore’, Southern RtoP never derogated from the inviolability of sovereign equality as a tradition of the Non-Aligned Movement and G77, groups for which sovereignty has long been more a prerequisite than adversary of human rights. ¹²²

This re-reading comes at a time both when faith in the model of sovereignty as responsibility, has steadily eroded in the study of International Relations and when the contestation and interpretation of RtoP, especially its third pillar, has gradually become the focus of more issue-specific literature. ¹²³ A story of two RtoPs shares this interest in the contingency and evolving content of intervention norms, reaffirms that ‘non-Western powers are increasingly shaping the international normative discourse on humanitarianism’, and theorizes further the paternalist complicities of particular forms of responsibility. ¹²⁴ But it also proposes a major correction: there is an RtoP that is practiced by veritable ‘non-Western stalwarts of sovereignty’. ¹²⁵ Through the turn of the century, what we see is a loosely Southern coalition that rejects ‘absolute non-interventionism’, while asserting (as it had before) a sovereignty that implies humanitarian responsibilities – including in relation to prevention of gross violations of human rights – without dropping a commitment to its non-suspension, even in emergency circumstances. Southern practice suggests: yes, sovereignty as responsibility, and yes, an international responsibility to protect, but never responsibility as control. In this view a more responsible form of protection, or intervenors’ RwP, is about accountability and prevention, but as conduits through which to preserve the self-determined political futures of the rescued.

Whither then the Responsibility to Protect? No doubt the overall political promise of the RtoP agenda has waned since 2011. This is exemplified in part by the Security Council’s inability to act in timely and decisive ways in the face of multiple significant humanitarian crises and by a more comprehensive and complicated backlash against Liberal Global Order. ¹²⁶ Post-Libya practice encompasses much more than practice in relation to Syria, but it is also true that official statements made in relation to Syria crisis illustrate the enduring heuristic value of Northern and Southern RtoPs. Consider the reappearance of Southern logics in related voting explanations of the BRICS countries, particularly in references to the principle of a ‘Syrian-led’ (and UN-facilitated) process ‘leading to effective political reform’ in ‘accordance with the will of the Syrian people’ and the ‘aspirations of their people’, and in the view that the crisis cannot ‘be considered in the Council separately from the Libyan experience’, for if ‘Libya in the NATO interpretation is a model for the future actions of NATO in implementing the responsibility to protect’, then RtoP is either undesirable or must not adhere to the practice of NATO member states. ¹²⁷ Furthermore, while it is true that the General Assembly has included RtoP as an annual agenda item, it should be remembered that the proposed aim of initial plenary meetings was never to celebrate a concept already formed, nor to adopt a resolution, but ‘to build consensus’ and ‘common ground’ because ‘we are not all on the same page’. ¹²⁸ In fact when we look at much post-2011 UN debate, what we continue to see is a clash of, first, sovereignty’s clear qualification but indefeasibility, such that ‘RtoP is not a justification for intervention by external actors in the domestic affairs of sovereign states’, and, second, of its suspension for higher purposes, such that while ‘national ownership is important’ there is a ‘fundamental tension’ when states fail to uphold their primary RtoP. ¹²⁹

Still, nothing suggests a discursive retreat from the legitimacy of emergency collective action for humanitarian purposes, including by coercive means, such as UN peace operations with robust mandates to protect civilians, in accordance with the Charter including Chapter VII. The challenge rather has to do with the implications and translations of this theory for effective policy and with its plausible applications in a period of resurgent geopolitics. Can intervention that stops short of defeasible sovereignty successfully perform its protective function? If so, in which contexts, across which range of cases, and how, exactly, to draw the line? What resources might be necessary for such operations to succeed and, in the first place, be shouldered and approved by governments? These are not easy questions to answer. They also recall historical and ongoing debates among statespeople. ¹³⁰ In particular the feasibility of neutral protection surely narrows when national leaders are widely recognized and yet constitute the perpetrators of one-sided violence. Where crimes of genocide occur in such scenarios, atrocity prevention must act in favour of the victims and, at least in some sense, against the sovereign. ¹³¹

But the key point is that a recovery of Southern protest shows how these questions are now much more politically consequential questions than the choices most usually offered by interventionist accounts of the post-Cold War era: of variations on humanitarian intervention and the suspension of sovereignty on the one hand, and of absolute sovereignty and standing by on the other. A normative tension between state sovereignty and universal human rights protection should no longer be simply assumed in international society. In contrast to prevailing explanations of RtoP as a singular norm, discursive practices of RtoP expose paths ignored and misconstrued at a time of profound contestation of global governance structures and existing security orders. Engagement with these underlying differences in interpretation is non-optional in a persuasive understanding of the global humanitarian protection debate.