home       biography       publications       speeches       organisations       images       @contact

Responding to Mass Atrocity Crimes: The Responsibility to Protect (R2P) After Libya and Syria

Public Lecture by Professor the Hon Gareth Evans, Co-Chair, Global Centre for the Responsibility to Protect, President Emeritus of the International Crisis Group, and former Foreign Minister of Australia, Central European University School of Public Policy, Budapest, 24 October 2012

As we look back at not only what has been happening in recent months in Syria, but over recent decades, and indeed over the whole course of human history, one of the most depressing and distressing realities we have to acknowledge has been our inability to prevent or halt the apparently endlessly recurring horror of mass atrocity crimes – the murder, torture, rape, starvation, expulsion, destruction of property and life opportunities of others for no other reason than their race, ethnicity, religion,  nationality, cast,  class,  ideology or opinion.

What is in some ways hardest of all to believe is how little changed in the decades after World War II. One might have thought that Hitler’s atrocities within Germany and in the states under Nazi occupation would have laid to rest once and for all the notion – predominant in international law and practice since the emergence of modern nation states in the 17th century – that what happens within state borders is nobody else’s business: to put it starkly, that sovereignty is essentially a license to kill.

But even with all the developments in international human rights law and international humanitarian law which followed the War – even with the Nuremberg Tribunal Charter and its recognition of “crimes against humanity” which could be committed by a government against its own people; even with the recognition of individual and group rights in the UN Charter, and more grandly in the Universal Declaration of Human Rights and the subsequent International Covenants; even with the new Geneva Conventions taking forward international humanitarian law on the protection of civilians; and even after the Genocide Convention signed in 1948 – aimed at preventing and punishing the worst of all crimes against humanity, attempting to destroy whole groups simply on the basis of their race, ethnicity, religion or nationality – the killing still went on.

Why didn’t things fundamentally change? Essentially because the overwhelming preoccupation of those who founded the UN was not in fact human rights but the problem of states waging aggressive war against each other. What actually captured the mood of the time, and that which prevailed right through the Cold War years, was, more than any of the human rights provisions, Article 2(7) of the UN Charter: "Nothing should authorize intervention in matters essentially within the domestic jurisdiction of any State".

The state of mind that even massive atrocity crimes like those of the Cambodian killing fields were just not the rest of the world's business was dominant throughout the UN's first half-century of existence: Vietnam's invasion of Cambodia in 1978, which stopped the Khmer Rouge in its tracks, was universally attacked as a violation of state sovereignty, not applauded. And Tanzania had to justify its overthrow of Uganda's Idi Amin in 1979 by invoking “self-defence”, not any larger human rights justification. The same had been true of India’s intervention in East Pakistan in 1971.

With the arrival of the 1990s, and the end of the Cold War, the prevailing complacent assumptions about non-intervention did at last come under challenge as never before. The quintessential peace and security problem – before 9/11 came along to change the focus to terrorism – became not interstate war, but civil war and internal violence perpetrated on a massive scale. With the break-up of various Cold War state structures, and the removal of some superpower constraints, conscience-shocking situations repeatedly arose, above all in the former Yugoslavia and in Africa.

But old habits of non-intervention died very hard. Even when situations cried out for some kind of response, and the international community did react through the UN, it was too often erratically, incompletely or counter-productively, as in the debacle of Somalia in 1993, the catastrophe of Rwandan genocide in 1994, and the almost unbelievable default in Srebrenica in Bosnia just a year later, in 1995.

Then the killing and ethnic cleansing started all over again in Kosovo in 1999. Not everyone, but certainly most people, and governments, accepted quite rapidly that external military intervention was the only way to stop it. But again the Security Council failed to act, this time in the face of a threatened veto by Russia (an unhappily familiar story again over the last year, in the context of Syria, as I will come back to below). The action that needed to be taken was eventually taken, by a coalition of the willing, but without the authority of the Security Council, thus challenging the integrity of the whole international security system (just as did the invasion of Iraq four years later in far less defensible circumstances).

There was at least real debate about these issues in the ‘90s, but it was fierce, doctrinal and essentially ideological argument, producing nothing remotely resembling consensus. On the one hand, there were advocates, mostly in the global North, of "humanitarian intervention" – the doctrine that there was a "right to intervene" militarily, against the will of the government of the country in question, in these cases (‘droit d’ingerence’ in the words of Bernard Kouchner, its primary advocate). On the other hand there were defenders of the traditional prerogatives of state sovereignty, who made the familiar case that internal events were none of the rest of the world's business.

It was very much a North-South debate, with the many new states born out of decolonization being very proud of their new won sovereignty, very conscious of their fragility, and all too conscious of the way in which they had been on the receiving end in the past of not very benign interventions from the imperial and colonial powers, and not very keen to acknowledge their right to do so again, whatever the circumstances. Hardly anyone talked about prevention or less extreme forms of engagement and intervention, and there was no system of international criminal justice to which anyone could resort.  The options were ‘Send in the Marines’ or do nothing. This was the environment which led Kofi Annan to issue his now famous challenge to the General Assembly in 2000:

If humanitarian intervention is indeed an unacceptable assault on sovereignty, how should we respond to a Rwanda, to a Srebrenica – to gross and systematic violations of human rights that offend every precept of our common humanity?


So the world had both an institutional problem, in particular the absence of international courts and tribunals with the jurisdiction, and resources, to try and punish those accused of major war crimes and crimes against humanity; and a political and normative problem, the absence of any agreed principles for addressing mass atrocity situations. 

The first piece of good news is that a major part of the institutional problem has been remedied in recent years. There has been the development of a number of specialist national courts with international assistance, like the Special Courts for Sierra Leone and Cambodia. There has been the establishment (following the example of the International Military Tribunal set up Nuremberg in 1945) of specialist tribunals to deal with war crimes committed in specific conflicts – in particular for the former Yugoslavia and Rwanda (in the establishment of which CEU Rector John Shattuck played an absolutely central role). And, most importantly, there has been the establishment by treaty, the Rome Statute of 1998, of the International Criminal Court — setting up a permanent court to hear cases of genocide, crimes against humanity, and war crimes, with no time limitation on its ability to prosecute.

But all international law – as much as it pains international lawyers to confront this reality – is ultimately politics. Which means that while the international courts and tribunals, and other legal strategies, are important elements in the mass atrocity prevention and reaction toolbox, whether these tools are actually applied depends on political will – on international consensus about the relevant norms, and international cooperation in applying them. And it is that element of political will, and the practical cooperation which it makes possible, which has been profoundly lacking, not just for decades but for centuries, in the case of major war crimes and other mass atrocity crimes.

That brings us to the second piece of good news – on which I want to focus for the rest of this lecture – that we have in the last decade in fact taken a giant stride forward in addressing the normative problem, that question of political will, with the birth and evolution, of the new principle of “the responsibility to protect” (R2P for short). [1] The concept was initiated in the 2001 report of that name by the International Commission on Intervention and State Sovereignty (ICISS), which I co-chaired with the African diplomat Mohammed Sahnoun, and which was established specifically to respond to the challenge laid down by Kofi Annan, and to find a way past the agonizing consensus-free zone of the 1990s.

I do not pretend for a moment that we have yet solved the problem of mass atrocity crimes once and for all – how could I in the face of the totally unresolved mess in Syria? And there are innumerable implementation problems that will continue to arise every time the principle is invoked. But the reality is that we are closer to consensus now on the nature and extent of the international responsibility to respond to these crimes than we have ever been.

The Commission report took the whole debate in a new, and what is now acknowledged to be much more productive, direction, in three main ways:

First, presentationally, by changing the language of the debate:  turning “the right to intervene” into “the responsibility to protect”, and re-characterizing the issue as not being about the “right” of any states, particularly large and powerful ones, to throw their weight around militarily, but rather the “responsibility” of all states to act to protect their own and other peoples at risk of suffering from mass atrocity crimes.

Secondly, by broadening the range of actors in the frame. Whereas “the right to intervene” focused just on the international response – and by those capable and willing to apply military force – R2P involves three distinct levels of state responsibility. The primary responsibility is that of the sovereign state itself to its own people – one that is absolute, unconditional, and continuing – not to perpetrate or allow atrocity crimes on its territory (what we now call ‘Pillar I’). The second responsibility is that of others in the international community – including other states and intergovernmental organizations – to assist states to discharge that primary responsibility, if they are willing to be so assisted (‘Pillar II’). The third responsibility is that of others – if prevention fails, and a state is manifestly failing to protect its own people – to then provide that protection by every means prescribed, and circumscribed, by the United Nations Charter (‘Pillar III’).

The third dimension of R2P involved dramatically broadening the range of responses. Whereas the right to intervene, or humanitarian intervention, focused one-dimensionally on military reaction, the responsibility to protect involves multiple elements in the response continuum: preventive action, both long and short term; reaction when prevention fails; and post-crisis rebuilding aimed again at prevention, this time of recurrence of the harm in question. The ‘reaction’ element, moreover, was itself a nuanced continuum, beginning with persuasion, moving from there to non-military forms of coercion of varying degrees of intensity (like sanctions, or threat of international criminal prosecution), and only as an absolute last resort – after multiple criteria were satisfied – contemplating coercive military force.

Articulated this way, R2P had an extraordinarily rapid take-up, almost unprecedented in the history of ideas. Within four years, in 2005, after some intense and sustained, and often cantankerous, diplomacy in multiple forums, the core elements of the concept were unanimously endorsed by the more than 150 heads of state and government meeting as the UN General Assembly at the World Summit celebrating the UN’s 60th anniversary: it was made clear beyond doubt that genocide, war crimes, ethnic cleansing and crimes against humanity were not no-one else's business but everyone’s. And within another year concept had been embraced in a formal Security Council resolution. These formal statements were in themselves rather breathtaking achievements, since what was involved here conceptually was nothing less than, as the British historian Martin Gilbert has put it, “the most significant adjustment to national sovereignty in 360 years”.

But words on UN paper are one thing, implementation something else. The next five years – from 2005 through to the end of 2010, which I think of as R2P’s adolescence period – saw not a great deal of effective action, but much tortured argument about R2P’s scope and limits: whether and how it should apply in cases like Darfur, the Congo, Sri Lanka, the response to the cyclone in Myanmar, and Russia’s invasion of South Ossetia. That said, efforts made in the General Assembly during this period by a number of spoiler-countries – notably Nicaragua, Venezuela, Sudan and Cuba – to turn back the clock on the 2005 consensus were rebuffed, and  there were some clear-cut practical success stories, most of all in Kenya in early 2008, where in the context of a Rwanda-like explosion of ethnic-focused violence in the aftermath of a contested election, and with a major genocide feared, a diplomatic mission led by Kofi Annan under the auspices of the UN and African Union and explicitly invoking R2P successfully defused the situation – demonstrating in the process that even in the most extreme cases, R2P was not just about military intervention. 

It was not until 2011 that the UN Security Council itself took coercive action explicitly invoking R2P. But when it did so, in the cases of Cote d’Ivoire and Libya, this was widely heralded – including by me – as the coming of age of the responsibility to protect.  Libya especially, at least in February and March last year, was a textbook example of how R2P is supposed to work in the face of a rapidly unfolding mass atrocity situation during which early-stage prevention measures no longer have any relevance. In February, Gaddafi’s forces responded to the initial peaceful protests against the excesses of his regime, inspired by the Arab Spring revolutions in Tunisia and Egypt, by massacring at least several hundred of his own people. That led to the unanimous UN Security Council Resolution 1970, which specifically invoked ‘the Libyan authorities’ responsibility to protect its population”, condemned its violence against civilians, demanded that this stop and sought to concentrate Gaddafi’s mind by applying targeted sanctions, an arms embargo and the threat of ICC prosecution for crimes against humanity.

Then, as it became apparent that Gaddafi was not only ignoring that resolution but planning a major assault on Benghazi in which “no mercy or pity” would be shown to perceived opponents, armed or otherwise—his reference to “cockroaches” having a special resonance for those who remembered how Tutsis were being described before the 1994 genocide in Rwanda—the Security Council followed up with Resolution 1973, also invoking R2P, which, by majority vote with no Russian or Chinese veto or other dissenting voices, explicitly authorized “all necessary measures”, i.e. military intervention by member states,  “to protect civilians and civilian populated areas under threat of attack”.  Acting under this authorization, NATO-led forces took immediate action, and the feared massacres did not eventuate. If the Security Council had acted equally decisively and robustly in the 1990s, the 8,000 murdered in Srebrenica and 800,000 in Rwanda might still be alive today.

But with the apparent maturity of R2P also came a mid-life crisis. As the weeks and months wore on, as I will describe further in a moment, the Western-led intervention came under fierce attack by the BRICS countries – Brazil, Russia, India, China and South Africa – for exceeding its narrow civilian protection mandate, and being content with nothing less than regime change. For present purposes the most important result of this continuing dispute, and all the distrust it has engendered, has been its impact on the Security Council’s response to Syria, where the one-sided violence by the regime was by mid-2011 manifestly worse even than that which had triggered the Libyan intervention. In the face of threatened vetoes from Russia and China, and continuing unhappiness by the other BRICS members, the Council found itself long unable to agree even on a formal condemnatory statement, and is now as far away as ever from agreeing on more robust measures like sanctions, an arms embargo, or the threat of ICC prosecution.

That said, just as any celebration about the triumph of the R2P principle would have been premature after the Libyan resolutions last year, so too would be despair now about its future.  What we do know, on the evidence of the major debates on R2P that have taken place in UN General Assembly annually since 2009 is that, even last year at the height of the concern last year about claimed over-reach of the Security Council’s mandate in Libya, and even this year, with that concern still resonating, the overwhelming majority of member states remain fully supportive of the new norm. And the Security Council itself, for all its divisions over Libya and Syria, continues to use R2P language where appropriate, as it has in resolutions over the last year on Yemen and the Congo, and a presidential statement on South Sudan. Secretary-General Ban Ki-Moon was not exaggerating when he said last September, ‘Our debates are about how, not whether, to implement the Responsibility to Protect. No government questions the principle’.


But, and it’s a very big ‘but’ indeed, we have to acknowledge that for all the normative consensus that does now exist at the level of high principle, the debate about how to implement R2P in practice – at least at the sharp end, in the hardest cases, when prevention has manifestly failed and violence is actually occurring, and the question arises whether coercive military force should be used – has   become very fierce and very divisive. From the high point of consensus and cooperation we reached in the Security Council in February and March last year, with the resolutions over Libya (and Cote d’Ivoire as well) we have now plunged to a very low point indeed.

So what went wrong, and how do we fix it?

Part of the reason for the paralysis over Syria, as compared with the speedy and effective response to Libya, is that the geopolitics of the Syrian crisis are very different, with potentially explosive regional sectarian divisions, no Arab League unanimity in favour of tough action, a long Russian commitment to the Assad regime, and a strong Syrian army meaning that any conceivable intervention would be difficult and bloody.

But there’s more to it than that. Security Council consensus about when and how to apply R2P in sharp-end cases of human rights catastrophe, so evident in February and March 2011, has simply evaporated in a welter of recrimination about how the NATO-led implementation of the Council’s Libya mandate “to protect civilians and civilian populated areas under threat of attack” was actually carried out.

Leading the critical charge, as I have noted, have been the ‘BRICS’ . Their complaints are not about the initial military response – destroying Libyan air force infrastructure, and air attacks on the ground forces advancing on Benghazi – but what came after, when it became rapidly apparent that the three permanent member states driving the intervention (the US, UK and France, or ‘P3’) would settle for nothing less than regime change, and do whatever it took to achieve that.

Particular concerns are that the interveners rejected ceasefire offers that may have been serious, struck fleeing personnel that posed no immediate risk to civilians and locations that had no obvious military significance (like the compound in which Gaddafi relatives were killed) and, more generally, comprehensively supported the rebel side in what rapidly became a civil war, ignoring the very explicit arms embargo in the process.

The P3 is not without some answers to these charges. If civilians were to be protected house-to-house in areas like Tripoli under Gaddafi’s direct control, they say, that could only be by overturning his whole regime. If one side was taken in a civil war, it was because one-sided regime killing sometimes leads (as now in Syria) to civilians acquiring arms to fight back and recruiting army defectors.  Military operations cannot micromanaged with a ’1,000 mile screwdriver’. And a more limited ‘monitor and swoop’ concept of operations would have led to longer and messier conflict, politically impossible to sustain in the US and Europe, and likely to have produced many more civilian casualties.

And yet. These arguments all have force, but the P3 resisted debate on them at any stage in the Security Council itself, and other Council members were never given sufficient information to enable them to be evaluated. Maybe not all the BRICS are to be believed when they say that, had better process been followed, more common ground could have been achieved.  But they can be when they say they feel bruised by the P3’s dismissiveness during the Libyan campaign -- and that those bruises will have to heal before any consensus can be expected on tough responses to such situations in the future.

The good news is that there is a way forward, and it is Brazil that has shown the way, in the argument it has been promoting over the last year that the R2P concept, as it has evolved so far, needs supplementing by a new set of principles and procedures  which it calls “responsibility while protecting” (“RWP”). When stripped down to its bare essentials, RWP involves three core propositions:

 –  First, R2P, as endorsed in 2005 and as refined since, remains a valuable normative advance, not least in its strong focus on prevention of, as well as reaction to mass atrocity crimes, and whatever the issues involved in its  practical implementation, particularly in relation to the use of military force, the baby should not be thrown out with the bathwater: RWP is designed to complement R2P, not replace it.

–  Second, before acting under Pillar III of R2P, and under Chapter VII of the UN Charter, to endorse any use of coercive military force, more formal and systematic attention needs to be paid by the Security Council to relevant prudential criteria or guidelines (including in particular ‘last resort’, ‘proportionality’ and ‘balance of consequences’)

 –  Third, after such action has been taken, there should be enhanced UN Security Council procedures to monitor and assess the manner in which such mandates are interpreted and implemented.

Although the initial reaction by the P3 powers to the Brazilian RWP proposal when it was first articulated at last year’s UN General Assembly  –  ‘these countries would want all those delaying and spoiling options, wouldn’t they’ – this has begun to soften, as it must.  They have begun to realize, as they must, that if an un-vetoed majority vote is ever going to be secured again for tough action in a hard mass atrocity case, even action falling considerably short of military action, the issues at the heart of the backlash that has accompanied the implementation of the Libyan mandate, and the concerns of the BRICS states in particular – voicing as they do the concerns of a much wider swathe of the developing world – simply have to be taken seriously.


All that said, renewed consensus on how to implement R2P in the hardest of cases in future is going to be hard to achieve, and will certainly take time to achieve: it is very difficult to believe that it could come in time to be very helpful in solving the present crisis in Syria.  Which leaves us with a real problem about what to do right now with Syria, given that the situation continues to deterioriate, that mass atrocity crimes continue to occur, and with an increasing tendency for them to be perpetrated now not only by the Assad regime but on the rebel side as well.

The problem for policymakers is that none of the familiar tools in the atrocity prevention and conflict resolution toolbox are readily applicable to the current situation:

--  Condemnation, targeted sanctions, the threat of ICC prosecution and an arms embargo might have made a difference had the Security Council been willing to apply them in the middle of last year: but the combination of Russian realpolitik cynicism and post-Libya backlash meant that not only did Assad escape that pressure then, but was reinforced in his sense of impunity. Even if the Security Council were now to change its position – or sanctions outside the UN framework to intensify – it is difficult to believe, given the extent to which the conflict has extended and intensified, this kind of pressure could now make any real difference.

--  Diplomatic negotiation, of the kind attempted by Kofi Annan and now being continued by Lakhdar Brahimi, has been blocked by three separate obstacles, none of which show signs at the moment any sign of being overcome. First, the extent to which the Assad regime is now dug in, with its own atrocious behaviour making a negotiated retention of some portion of its own power – which might have been possible at the outset – ever more impossible. Second, the lack of international confidence in the opposition SNC or any other leadership group as a strong, united and moderate negotiating partner. And third, the continuing lack of unity among the key international players as to the form any negotiated outcome might take.

--  As to military options, on which attention is inevitably refocusing as the body count from atrocity crimes continues to grow and with it the sense that something more substantial must be done, the trouble is that there are immense problems with every possible course of action:

  • Supplying arms to the rebels – the simplest option and one already happening from some of the neighbours but not the major Western powers –  runs the real continuing risk, of which there is plenty of evidence already, of them being diverted into the hands of exrremist Islamist groups. 
  • External military intervention on the ground with the notionally limited objective of carving out buffer zones or safe havens or food and medical supply routes would – except in the wholly unlikely event of acquiescence by the Syrian military – rapidly lead rapidly lead to full-scale war, with all the even further escalation in suffering this would inevitably involve.
  • And enforcement of a no-fly zone over the whole, or even just parts, of the country, which would on the face of it neutralize the Syrian regime’s greatest single advantage over the rebels, runs into the acute problem that the US or any other military power conceivably likely to be able and willing to implement this would almost certainly insist on this being accompanied by the destruction of Syrian airfields, air defence systems and command and control infrastructure – which would mean again full-scale war.

No direct military intervention option, however notionally limited, could possibly survive a Russian veto in the present environment, which raises all the old agonizing problems about bypassing the Security Council and challenging the ground-rules of international order that the Council is there to protect. Acute sectarian tensions in the region, mirroring those in Syria itself, raise the prospect of spillover war flowing from any intervention there. And any Western military intervention without the kind of support from the Arab/Islamic world that accompanied the initial Libyan intervention would be bound to generate a backlash around the region, with all the heightened risk of terrorist attacks that go with it.

 I am conscious that all this adds up to something like a counsel of despair, and – while I’ve been around long enough to know, as Brent Scowcroft said recently in this context “Just because there’s a problem doesn't mean that there is a solution” – I certainly do feel an acute sense of despair and frustration at the absence of any obvious solution right now.

It may be that the diplomatic route is the only possible one – with the object being to build maximum possible institutional credibility for the rebels,  and a believable political transition plan from them, with this then putting new pressure on Russia to persuade the Syrian leadership to accept a settlement process.  This is a very slender reed, but it may be the only one we have.

The worst possible outcome would be for Syria, through frustration and familiarity, to over time slide out of public consciousness into the realm of forgotten conflicts and forgotten atrocity crimes. It is absolutely crucial that those of us committed to R2P, the norm we have worked so hard to create over the last ten years, do everything we can to keep global attention focused on what is happening within the country and not becoming desensitized or immune to new shocks – maintaining our passion and maintaining our shame at our inability to stop this carnage, and not resting until we have done so.


The completely effective implementation of R2P is going to be work in progress for some time yet. There are bound to be acute frustrations and disappointments and occasions for despair along the way, but that should not for a moment lead us to conclude that the whole enterprise has been misconceived.

Despite what has gone wrong over the last year, I think the basic consensus on the Security Council of February-March 2011 can be re-established, with Brazil’s RWP proposal, further refined and developed, playing a critical role as a circuit-breaker. I say this because I don’t think there is any policymaker in the world who fails to understand that if the Security Council does not find a way of genuinely cooperating to resolve these cases, working within the nuanced and multidimensional framework of the R2P principle, the alternative is a return to the bad old days of Rwanda, Srebrenica and Kosovo.

That means either total, disastrous, inaction in the face of mass atrocity crimes, or action being taken to stop them without authorization by the Security Council, in defiance of the UN Charter and every principle of a rule based international order. After all that has been achieved over the last decade, that would be heartbreaking. Being the congenital optimist that I am, I believe that won’t happen, and that sanity and decency will prevail. But all those of us who care are going to have to work very hard to ensure that it does.


[1] More or less parallel with the birth and evolution of R2P, we have had another important and closely related normative development, the emergence through the UN since 1999 of a set of principles and strategies addressing the Protection of Civilians in Armed Conflict (POC). The scope of this concept is wide ranging, including e.g. attention to the humanitarian impact of sanctions, and problems arising from mixing combatants and civilians in camps for refugees and internally displaced persons.  But probably its most important practical application has been in widening the scope of military peacekeeping mandates to ensure that there is capacity to deal forcefully with those who would violently disruptive – and to ensure in particular that there will be no more debacles in the future like Srebrenica in 1995, when 8000 men and boys were taken from under the unprotesting noses of UN peacekeepers and led to their slaughter.

The R2P and POC norms are very much sister concepts. They differ in just two respects, neither significant for present purposes. POC is broader than R2P to the extent that the rights and needs of populations caught up in armed conflict go well beyond protection from mass atrocities. But in one major respect the scope of R2P goes well beyond POC, in that it is concerned with preventing and halting mass atrocity crimes regardless of whether they occur in times of armed conflict. Cambodia in the mid-1970s, Rwanda in 1994, Kenya in 2008 and Libya at least at the time of the initial UN intervention in February /March 2011 are major examples of such one-sided violence, non-war situations.