A Responsibility to Prevent? A Norm's Political and Legal Effects

This article studies the preventive pillar of the responsibility to protect (R2P) norm and focuses on its normative development, robustness, normative fit and its chances of becoming a legal norm, based on the sources of international law and the theory of interactional law. This article argues that prevention is the pillar of R2P with the widest level of international support and is based on increasingly deeper and more precise shared understandings. The institutionalisation process and a norm cascade - possibly leading to internalisation - are materialising simultaneously while the interpretation and application of the norm are still being contested. To be more robust and more effective in practice, the norm ‘responsibility to prevent’ has to gain in specificity. Prevention of mass atrocities finds increasing support in the sources of international law. For an interactional legal norm to emerge, the exact content and division of responsibilities has to become clearer, officials have to act in line with the norm and practices of legality have to emerge.


Introduction
The failures of the international community to prevent the genocide in Rwanda in 1994 and in Srebrenica in 1995 as well as the 1999 NATO-intervention in Kosovo without United Nations Security Council (UNSC) authorisation triggered a debate on humanitarian intervention and state sovereignty.In reaction to this debate of the 1990s, the norm responsibility to protect (R2P) gained prominence as a concept developed by the International Commission on Intervention and State Sovereignty (ICISS) in 2001.The ICISS aimed to reframe the debate by focusing on the responsibility a state has to protect its own population and, if the state is unable or unwilling to do so, the responsibility for the international community to provide protection to the population.Since then, R2P has been endorsed by the 2005 World Summit and has been explicitly referred to in UNSC resolutions on Libya 12 and Ivory Coast. 34Frank Huisingh (fhuisingh@gmail.com) completed the LLM Law and Politics of International Security at the Vrije Universiteit Amsterdam in 2012and now studies for a Master of Letters in Peace and Conflict Studies at the University of St Andrews.This article is based on his final dissertation for the LLM, supervised by Geoffrey Gordon LLM. 1 In Resolution 1970 the UNSC stated in one of the preambles: "Recalling the Libyan authorities' responsibility to protect its population" (2011 United Nations Security Council Resolution, Peace and Security in Africa, S/RES/1970(2011)); Resolution 1973 states in its preambles "Reiterating the responsibility of the Libyan authorities to protect the Libyan population" (2011 United Nations Security Council Resolution, Libya, S/RES/1973(2011)) The ICISS defines R2P as consisting of three pillars: the responsibility to prevent, the responsibility to react and the responsibility to rebuild. 5 In the 2005 World Summit Outcome, the emphasis was placed on the responsibility of states to protect their own population and prevent mass atrocities 6 and the responsibility of other states in supporting the execution of these tasks.Prevention is deemed essential and military intervention only the option of last resort.
Nevertheless, the political and academic debates have mostly focused on the military reaction side of R2P, 7 neglecting the responsibility to prevent (from now on called R2Prev). 8However, in light of the emphasis in the ICISS Report and the World Summit on R2Prev, as well as the General Assembly debates in 2009 in which the widest degree of consensus was reached on R2Prev, academic attention for R2Prev is warranted.This article will study the norm R2Prev from a political and a legal perspective.From a political perspective, it will focus on its political normative development, its robustness and normative fit.From a legal perspective, its value will be assessed by looking at the traditional sources of international law as well as the theory of interactional law.The main research question of the article will be: Why and to what extent will the norm R2Prev have political and legal effects?
R2Prev is successful as a norm if the normative pull of its prescriptions are strong; if actors consider those prescriptions as the most appropriate course of action. 9The goal for R2Prev norm advocates is that actors (states, the UN and regional organisations) internalise the norm, which means that actors act habitually according to its prescriptions.Internalisation is the final stage of a norm development process.
In order to shed light on all elements involved in answering the question posed above, the article to follow will consist of a distinct political and legal part.To analyse the emergence of R2Prev, the first part will analyse the political normative trajectory upon which R2Prev has embarked.The second part of this article will shift focus and centre upon the legal analysis of the norm.
As for the first part of the article, the life cycle approach as developed by Finnemore and Sikkink proves a useful analytical tool for analysing the norm development process of R2Prev from norm emergence to norm cascade and norm internalisation.
However, this framework is insufficient of itself to explain the effectiveness of the norm in the daily practice of national and international politics.Therefore, this article will analyse the robustness and concurring effectiveness of the norm.The robustness of a norm is based on the specificity of the norm, its concordance in diplomatic discussions and its durability when dealing with violations.Furthermore, the content of the norm and its normative fit affect the likelihood of its effect in practice.This paper argues that the life cycle and the robustness approach supplement each other.The further norms are internalised and the more robust they are, the more effect they will have in practice.This leads to the following two sub-questions: 1) How did the norm R2Prev develop as seen from the 'life cycle' approach?2) How much effect is the norm R2Prev expected to have, based on its robustness and normative fit?As for the second part, it is essential to stress that R2Prev unequivocally implies responsibilities.R2Prev is not only an appropriate option of behaviour among other options, but implies certain legal responsibilities and obligations.To what extent does such an obligation exist, and if so, will actors feel this sense of obligation and live up to it?If this is indeed an obligation, is it also a legal obligation?In short, what is the legal status of the norm R2Prev?This leads to the following two sub-questions: 3) To what extent does the norm R2Prev find support in the sources of international law? 4) To what extent does the norm R2Prev create a sense of legal obligation?
The third question will be answered by focusing on the legal value of R2Prev based on the sources of international law as defined in Article 38 of the Statute of the International Court of Justice (ICJ).However, solely considering the sources is insufficient to study the effectiveness of R2Prev as a legal norm.International law is not backed up by a higher authority with coercive powers to enforce compliance.The question is whether the compliance pull of a norm is so strong that actors experience a sense of legal obligation to live up to the norm.The theory of interactional international law developed by Brunnée and Toope10 provides the step between political norms as studied by constructivism and international legal norms.The theory of interactional law argues that norms based on shared understandings, fulfilling the internal requirements of legality and supported by a practice of legality, will lead to a sense of legal obligation.
All these approaches will be applied to R2Prev in an empirical analysis.On the basis of this analysis, this article will argue, among other arguments found in the conclusion, the following.Firstly, the norm R2Prev is the pillar of R2P that is based on the widest shared understandings, which deepened and became more robust throughout the period under study.Secondly, the processes of institutionalisation, norm cascade, norm contestation and the start of the internalisation process are taking place simultaneously.So, while shared understandings are still being discussed and recreated, the norm cascade and the start of the internalisation process take place.This demonstrates that norm development is not a straightforward unidirectional process.Thirdly, R2Prev's normative development shows that while widely shared understandings can emerge on a broadly defined norm, it needs to become specific in its prescriptions and demonstrate durability in order to become and remain effective in practice.Finally, when it comes to the norm's chance of becoming interactional international law, it can be argued that it is at least based on increasingly deeper and wider shared understandings.More clarity is needed on the exact responsibilities of actors and division of tasks between states, the UN and regional organisations.More clarity will reflect positively on the other criteria of legality.Until this happens, it will remain hard to assess whether congruence exists between the norm and the actions of officials and whether a practice of legality is present.

Ι.1 The Life-Cycle, Robustness and Normative Fit of the Norm Responsibility to Prevent
The life cycle approach analyses the development of norms 11 and the different stages of the process of norm development.It can shed light on how the norm R2Prev emerged, the extent to which R2Prev has cascaded around the international community and is possibly internalised into their behaviour; the approach thus provides a tool to map the development of the influence of R2Prev on state's actions.The norm life cycle consists of three stages: norm emergence; norm cascade; and internalisation.Norms emerge through the persuasion of a critical mass of actors by so-called 'norm entrepreneurs', actors with strong convictions about what constitutes appropriate behaviour. 12It is critical for norm entrepreneurs to act from an organisational platform, which provides them with information and access to important audiences. 13The norms already promoted by these international organisations, NGOs or transnational advocacy networks influence the way the new norm is promoted.In order for new norms to successfully compete with old norms, norm entrepreneurs use language to reframe a debate and create a new way of understanding an issue. 14The indeterminacy of an issue as well as events that trigger a response can open the way for a new norm to emerge.
If a norm has emerged as a candidate norm in international politics, it can only have effects if it is accepted by many actors in the international system.In the second phase, after a tipping point has been reached, a norm cascade takes place.This means that the norm gets adopted by many more actors through a process of international socialisation by norm leaders. 15This 'tipping point' is reached when a critical mass of relevant state actors adopts the norm, an argument that Finnemore and Sikkink argue is supported by quantitative evidence. 1617Some states can be considered critical for a norm cascade to take place, because without them "the achievement of the substantive norm goal is compromised", or because they have a certain moral stature. 18After the tipping point, the international norm cascade -instead of domestic norm entrepreneurs -leads other states to adopt the norm. 19Institutionalisation, the adoption of a norm in international rules and by organisations, can take place prior to or during the norm cascade and can contribute to the step from norm emergence to broad norm acceptance.Institutionalisation can clarify the norm and define procedures by which norm leaders coordinate disapproval and sanction normbreaking behaviour.
If a norm gets adopted by a state and it becomes part of a state's habit, is subsequently taken for granted and not a matter of public debate, then the norm is internalised. 20he norm is extremely powerful, because no one questions an actor for behaving according to the norm.This also implies that internalised norms are hard to discern, since actors do not "seriously consider or discuss whether to conform". 21

Robustness of norms
In order to conduct a proper analysis of the likely political effects of the norm R2Prev, it is essential to clarify what stage of norm development the norm has reached.It is, however, also important to consider the norm's strength, or in other words, its robustness.Norms come in varying strengths 22 and more robust norms can be expected to have more impact 23 and are more likely to generate compliance. 24esearchers have the tendency to focus on explaining successful norms 25 while it is insightful to study ineffective and new norms too.Therefore, studying R2Prev, whose success on the long term is yet uncertain, can be a contribution to the academic debate.Specificity, durability and concordance are all argued to contribute to the robustness of a norm, and will be briefly addressed below. 26

Specificity
A more specific norm is better defined 27 and thus more intelligible for the actors to whom it is applicable.Simple, clear and straightforward norms lead to less disagreement about implementation than unclear and ambiguous norms and will thus be more effective. 28The notion of specificity is related to the notion of precision and the 'determinacy' criterion identified by Franck 29 in the international law literature.Precise norms or rules contain a lot of detail and are clear as to what the norm requires, prohibits or authorises actors to do or refrain from doing.They are suited for a consistent interpretation and application across different cases and this makes its effects more predictable.However, a tension can exist between the need for specificthat is, simple and straightforward -norms and the call for precise norms with more detail.If details improve the clarity of a norm, there is no inconsistency between specificity and precision.However, in many cases there will be a trade-off between the simplicity and intelligibility of a norm and the degree of detail in its prescriptions.
Determinacy is one of the factors which according to Franck influence whether a norm is seen as legitimate and can gain compliance. 30The degree to which a norm is determinate rests on the clarity and transparency of the content of a norm. 31Regular interpretation and application of the norm across different cases -by legitimate institutions -can strengthen the determinacy of a written formulation of a norm. 32

Durability
Norms that have been effective longer are obviously more robust. 33For a norm to be durable, it has to effectively deal with challenges, such as violations of the norm.If norm violations and violators are penalised, the norm is reinforced and reproduced.It is therefore important to study what constitutes a violation of a norm and how this violation is interpreted and reacted to by relevant actors.The notion of durability is similar to Franck's notion of pedigree, which also studies the historical roots of a norm and the history of its consistent application. 34

Concordance
A norm is also more robust if there is more concordance as regards the norm: its acceptance and repetition in diplomatic discussions and treaties. 35Public reaffirmations of a norm are not always an indication of a norm's strength, but may be used to try to save a weakening norm.This article argues that new or evolving norms will be reinforced and become more robust when affirmed.If rules and norms are not explicitly supported in public, but states do not even consider violating them and seem to take them for granted, one can identify a situation of clear internalisation. 36

Content of the norm
The content of the norm affects its legitimacy, effectiveness and compliance. 37If a norm fits better with the normative framework that is in place, it will most likely have more effect in practice and actors are more likely to comply with it. 38This finding implies that ground-breaking norms or norms that do not (yet) fit well within the range of existing norms will have difficulty gaining acceptance.It implies that normative frameworks prefer stasis to change.In order to influence or change the debate, norms should resonate with existing norms and take a step-by-step approach in order to push the normative framework within which it operates in the desired direction.
This idea of normative fit is very similar to Franck's idea of coherence. 39Coherence here means that norms do not operate alone, but complement each other in a broader normative framework. 40Coherence will contribute to legitimacy.Furthermore, if a specific norm or rule adheres to one of the core principles and goals of the international system, it is even more likely to gain legitimacy and generate compliance. 41

Ι.2 An Interactional Account of International Law
After having addressed the theoretical framework to be employed in the first and political-historical part, this article now turns to the framework to be used for the second and legal part of this investigation into the expected effects of the R2Prev norm.The interactional theory of international law, as developed by Brunnée and Toope, distances itself from the formalistic interpretation of international law which is unable to explain compliance, parts of the positivist doctrine, as well as a rationalist reading of international relations.The interactional theory of international law studies under what conditions a rule or a norm leads to a sense of legal obligation, which they argue is the distinctive aspect of law. 42According to Brunnée and Toope only the law created through the interactional framework can be described as legitimate. 43The interactional theory of international law adopts the insight from social constructivism that the perceived self-interest of states is not based on a rational cost-benefit analysis, but on social norms.Interaction and communication create shared understandings and shared knowledge and in this process the social norms guiding the behaviour of actors emerge.
Legal obligation is created via interaction.In contrast to what neo-realists and positivists argue, interactional law argues that law is not only made by states.Law is made through the interactions of a broad variety of actors such as states, elites, the media, NGOs and 'ordinary citizens'. 44Interactional law is created through interaction and reciprocity, which has to be maintained by actors collectively. 45To create a sense of legal obligation, three conditions have to be fulfilled. 46Firstly, law should be congruent with shared understandings.Secondly, the internal qualities of law should be fulfilled.These qualities are specified in the eight criteria of legality as laid down by Fuller, the legal scholar who influenced the theory of interactional law.Thirdly, the norms have to be enmeshed in a practice of legality. of practice, which consist of actors dealing with the same issue.In communities of practice, actors participate to create, maintain or shift shared understandings and influence states and other actors. 49Via these interactions and collective learning processes, which have to be quite dense and specific in order to have an effect, shared understandings are produced and reproduced. 50reading norms is not a unidirectional process, in which norm entrepreneurs teach other actors.If norm entrepreneurs want to diffuse their norms to other actors, they have to expand the community of practice with those actors.In this interactional process, the norm can be reshaped again. 51This illustrates that within these communities of practice, conflicts between actors are possible.Furthermore, since a plurality of communities of practice exists, they also often have overlapping priorities and membership. 52ared understandings can represent deep moral commitments, but the minimum for interactional law creation is a commitment to communication and autonomy, as well as a shared understanding of what constitutes legality. 53Increased interaction may lead to more interconnected communities and richer shared understandings.Different opinions and cultures around the world explain the scarcity of deep shared understandings and underline the fact that law can only be created within specific times and places where actors agree at least on basic objectives to achieve together. 54cial norms -shared understandings-and legal norms exist on a continuum.The difference between social and legal norms is the fulfilment of the internal qualities of law and a practice of legality concerning the norm. 55Only then a sense of legal obligation is created. 56

Internal qualities of law
The eight criteria of legality are generality, promulgation, non-retroactivity, clarity, non-contradiction, not asking the impossible, constancy, and congruence between rules and official action. 57The generality of rules means that they must require or prohibit a certain conduct. 58Secondly, legal norms must be promulgated, accessible to the public, so actors know what the law requires.Thirdly, law should be prospective, so actors can take the law into account in their decision-making.Fourthly, the law must be clear and understandable.Fifthly, law should avoid contradiction.Sixthly, law should be realistic and not demand the impossible.Seventhly, the law should be constant over time.Finally, the actions of officials operating under the law should be congruent with the law.
To create interactional law, the application of the legal norm should also satisfy the legality requirements; a practice of legality should exist. 59States and other actors will only feel a continuing legal obligation if other actors in the communities of practice behave according to the rules. 60In contrast to customary law, interactional law requires inclusive participation by many different actors, not only states. 61teractional law is a theory of legal obligation.A sense of legal obligation does not necessarily mean that actors will also comply with the legal norm.This theory explains an attitude, a commitment to a legal system and certain rules that is part of actors' identities. 62Ongoing interactions -for which a suitable setting is needed 63 -in the communities of practice are needed to maintain a compliance pull.However, considerations of interest and power can still lead actors to choose not to comply with rules to which they feel a sense of obligation. 64

ΙΙ.1 Life Cycle, Robustness and Normative Fit
This article studies the reports that spurred the interaction which created and developed the norm R2Prev, as well as the decision-making processes that provided the forums in which this interaction and norm-creation took place.This article argues that the emergence of R2Prev as a norm is not only part of the normative development of redefining sovereignty as responsibility towards one's own population, but also part of a normative development focused on preventing conflict and mass atrocities rather than reacting to it.
The first major point of reference in the debate about prevention, focused on conflict, was the Carnegie Commission on Preventing Deadly Conflict 1997.The Commission acted as a norm entrepreneur aiming to shift the norms and shared understanding from reacting to conflict to conflict prevention. 65It argued that direct and structural prevention would be cheaper than intervention and rebuilding and noted that UN efforts on prevention had been ad hoc, advocating a shift to a system of conflict prevention.The Commission argued that effective prevention rests on three principles that address both the root causes and direct causes of conflict: "early reaction to signs of trouble; a comprehensive, balanced approach to alleviate the pressures, or risk factors, that trigger violent conflict; and an extended effort to resolve the underlying root causes of violence". 66The commission thus identified the means of conflict prevention subsequently used in the 2001 ICISS Report in which R2Prev was introduced. 6759 Brunnée & Toope 2010a, supra note 10, pp.7-8. 60Ibid., pp.35 & 47. 61 Ibid., p. 54. 62Ibid., p. 92. 63Ibid., p. 107. 64Ibid., p. 110. 65 The reframing of the debate got support from other norm entrepreneurs, strengthening the emerging norm.Kofi Annan argued for a "culture of prevention", 68 and the UNSC adopted Resolution 1366 in 2001, affirming the centrality of conflict prevention to the work of the UNSC. 69Intersubjective agreement and concordance from powerful states around the prevention norm was further strengthened with the 2001 G8 Rome Initiative on Conflict Prevention.Actions were taken in the form of two conflict prevention pools created by the United Kingdom in 2001: the Global Conflict Prevention Pool and the Africa Conflict Prevention Pool. 70Next to the shift from reaction to prevention, the concept of sovereignty as responsibility came up. 71he first norm entrepreneur, Francis Deng, was Representative of the UN Secretary-General on Internally Displaced persons when he presented the idea in 1996 together with Roberta Cohen. 7273tersubjective agreement around the norm was strengthened even further by the African Union (AU) Constitutive Act.This Act, adopted in 2000, stressed that the organisation would replace its doctrine of non-intervention with a doctrine of nonindifference if an AU member state would fail to protect its population from genocide, war crimes or crimes against humanity. 74,75This endorsement of the concept of sovereignty as responsibility shows it is not a Western concept, but that it is shared by African states. 7668 Bellamy 2008, supra note 8, p. 136; 2001 Report of the Secretary-General, Prevention of Armed Conflict, A/55/985-S/2001/574, p. 1. 69 Bellamy 2008, supra note 8, p. 136; 2001 United Nations Security Council Resolution, 4366 th meeting, S/RES/1366.The resolution also welcomed the creation of a central fund for the development of regional conflict prevention capacity. 70Bellamy 2008, supra note 8, p. 137; Foreign & Commonwealth Office United Kingdom, Conflict Pool, 2001, at: http://www.fco.gov.uk/en/about-us/what-we-do/spend-our-budget/funding-programmes1/conflict-funding/conflict-pool(accessed on 15 May 2012). 71Sovereignty as responsibility rests on the inalienable human rights of individuals and the idea that governments are responsible for protecting the human rights of their population (A.J.Bellamy, Global Responsibility to Protect.From words to deeds, London and New York: Routledge 2011, p. 13) 72 F.M. Deng, 'Divided Nations: The Paradox of National Protection', The Annals of the American Academy of Political and Social Science 2006-1, pp.217-225.Deng argued about internally displaced persons: "The internally displaced are paradoxically assumed to be under the care of their own government despite the fact that their displacement is often caused by the same state authorities" (p.214). 73Mani & Weiss 2011, supra note 7, p. 3.; 2009 Report of the Secretary-General, Implementing the Responsibility to Protect, A/63/677, p. 6. 74 Global Centre for the Responsibility to Protect, "Implementing the responsibility to protect".Responding to the UN Secretary-General's report on the responsibility to protect populations from genocide, war crimes, ethnic cleansing and crimes against humanity, 2009b, p. 1; Heads of State and Government of the Member States of the Organization of African Unity, Constitutive Act.African Union, 2000, at :http://www.africaunion.org/root/au/aboutau/constitutive_act_en.htm(accessed on 12 March 2012). 75Article 4h states: "the right of the Union to intervene in a Member State pursuant to a decision of the Assembly in respect of grave circumstances, namely: war crimes, genocide and crimes against humanity" (Heads of State and Government of the Member States of the Organization of African Unity, 2000).The question is often raised to what extent this doctrine is in line with the role of the UNSC as the only body which can authorize an armed intervention.Since the Constitutive Act took effect, the three missions undertaken by the AU since that time (AMIB in Burundi; AMIS in Darfur and AMISOM in Somalia) have all taken place with the approval of the UNSC and host government consent (Bellamy 2011, supra note 71, p. 14). 76Bellamy 2011, supra note 8, p.13; P.D. Williams, ''The "Responsibility to Protect', Norm Localisation, and African International Society', Global Responsibility to Protect 2009-1, pp.392-416.
The ICISS Report "… 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 affect every precept of our common humanity?" 77is challenge, identified by Kofi Annan addressing the GA in 1999 and again in 2000, was still not settled.The sovereignty versus humanitarian intervention problem was indeterminate, which opened space for a new norm.The ICISS combined the two developments identified before in its report.The Canadian Government established the ICISS in 2000 and thus established a platform from which the authors of the report could get widespread attention from states for their new norms.Co-chair Evans and Commissioner Thakur, among others, became active norm entrepreneurs.The ICISS was structured in a way to support successful norm creation based on shared understandings by building consensus among a worldwide team of commissioners and rounds of international consultations.
The ICISS published its report called 'The Responsibility to Protect' in December 2001.The ICISS redefined sovereignty from a norm giving absolute protection to a sovereign authority from outside intervention, to the idea of sovereignty as responsibility. 78The old frame of a possible right to humanitarian intervention was replaced by a frame of duty, of a responsibility to act.The ICISS fundamentally changed the debate. 79The ICISS called R2Prev the first priority for the international community. 80ough this article has noted that the rhetorical support for prevention was mounting, the Commission argued that real commitment to conflict prevention was far less and they proposed a redistribution of funding towards conflict prevention.
For R2Prev, the norm was similar to R2P in general: the first responsibility for conflict prevention was that of the state, though for effective prevention assistance from the international community could be necessary.The international consultations identified that there was a need to exhaust prevention efforts before outside actors should resort to intervention, and that outside coercive efforts would receive greater support if local prevention had been supported first. 81tfalls of prevention strategies were also identified: if prevention strategies by outside actors mean a conflict becomes internationalised, the fear is that this will end with military intervention; other pitfalls are the danger of breaching sovereignty and unintentionally magnifying internal conflict.The ICISS advises policy makers to be sensitive to these concerns, respect sovereignty and ensure that good intentions do not worsen the situation.In general, less intrusive and less coercive measures can alleviate "the fears, prevalent in the non-western world, that there were neo-colonialist or neoimperialist motives underlying such humanitarian action". 82hat is the proposed norm, the proposed standard of appropriate behaviour?The ICISS recommendations relate to the goal of preventing "deadly conflict and other forms of manmade catastrophe". 83The 2005 World Summit would specify the goal to "preventing mass atrocities".
In order to reach this goal, the ICISS proposes to address both root causes and direct causes of internal conflict and other manmade crises that put populations at risk. 84he ICISS argues for a combination of early warning efforts; 85 the 'preventive toolbox', the policy measures that can be applied to prevent deadly conflict; the political will to apply the measures necessary.Though the ICISS argues that the problem is often not early warning but timely response, they advocate a centralised early warning capacity within the UN Secretariat.On the 'preventive toolbox', the ICISS reiterates the shared understandings about root cause prevention and direct prevention, as used by the Carnegie Commission and in the literature on conflict prevention. 86The ICISS summarises the political, economic and legal means for structural and direct prevention and argues specifically for a pool of unrestricted development funding, which can be used to employ quick-impact development projects to back up diplomatic initiatives.The UN should take the lead in working together with regional or sub-regional organisations, the World Bank and the IMF to support fragile states in good conflict prevention behaviour.
The measures proposed for outside states to assist come from existing literature on conflict prevention.This lack of innovation contrasts with the innovation shown in the development of the concept of sovereignty as responsibility.The ICISS is unsuccessful in reframing the debate on prevention.Since the ICISS argued that prevention is the single most important element of R2P, this lack of innovation is a clear shortcoming.
It is not clear what will exactly trigger outside actors to take preventive measures, and it is not obvious who should act.The Commission did not offer any guidelines for prevention, as it did with regard to applying military measures in reaction to humanitarian catastrophes. 87This inhibits the specificity of the norm.In that sense, R2Prev is not conceptually integrated with the other two components of R2P.Furthermore, the ICISS did not explicitly discuss how to ensure that early warning signs are translated into political action, 88 even though they have established the problem of generating political will.
In defence of the ICISS, three arguments should be given.First of all, their primary goal was to solve the Gordian knot of humanitarian intervention and sovereignty, an issue they took up successfully with innovative ideas.Secondly, crisis situations will always differ and it is thus hard to come up with a clear guideline about the different roles of states, neighbouring states, regional organisations and the UN that will suit all the different possible contexts.However, when it comes to long term root cause prevention this should be easier.Thirdly, it is not intrinsically wrong to copy the ideas that earlier work on conflict prevention has provided.
83 ICISS 2001, supra note 5, p. 19. 84Ibid., p. XI. 85  The norm R2Prev as defined by the ICISS is still very broad, both in the sense of what to prevent and of how to prevent it.However, the ICISS Report signifies the emergence of the norm, and through institutionalisation, interpretation and application the norm can and will become more determinate and precise.
The authors of the ICISS Report tried to fit the norm R2Prev with the normative framework of humanitarian intervention as well as the framework of absolute sovereignty by arguing that outside assistance should always respect the sovereignty of a state.Furthermore, the primary responsibility to protect lies with the state itself, which means that sovereignty entails responsibility and not, therefore, the total freedom to act as the leadership pleases; however, it is clear that if a state acts in line with R2Prev itself, no outside states will interfere with it.
Though many norm entrepreneurs feared that R2P and R2Prev had already lost its momentum due to the War on Terror, 89 Secretary-General Kofi Annan prepared the 2005 World Summit and appointed a High-Level Panel to discuss security issues ahead of the Summit, which took up R2P.ICISS Co-Chair Evans (at the time president of the International Crisis Group) participated in this Panel and he played an important role as a norm entrepreneur for R2P by pushing for its inclusion in the Report.Evans continuously published about R2P, advocating the norm and discussing its content, thereby acting as a strong individual entrepreneur. 90The Report 91 shows increased concordance and intersubjective agreement on shifting the debate on international security from reaction to prevention as well as on redefining sovereignty as responsibility.This act of norm entrepreneurship by the High Level Panel created a chance for the norm to emerge at the international political level at the World Summit.
The international composition of the High Level Panel helped to create shared understandings, and the (passive) support of the former vice premier and foreign minister of China and member of the Panel, Qian Qichen, indicates emerging support for R2P from one of the staunchest defenders of sovereignty. 92The norm was still in need of further clarification and specification and it was the Secretary-General of the UN, Kofi Annan, who, in the March 2005 report "In Larger Freedom", took this step.His endorsement of R2P 93 can be seen as a first step towards the institutionalisation of the norm within the UN system.From his position of power and influence, Annan acted as a norm entrepreneur within the UN and within the international state system. 94Annan specified the norm by making it applicable to the four mass atrocity crimes defined in international law: genocide, ethnic cleansing, war crimes and crimes against humanity.This specification made the norm more robust, since it clarified what the crisis situations were that states and the international community had a responsibility to prevent.
Both the report by the Secretary-General and the High-Level Panel Report were meant to open the way for new norms to emerge and build shared understandings which could then be endorsed by the 2005 World Summit.

The World Summit 2005 Outcome Document
The World Summit 2005 was a gathering of the representatives of all Member States of the United Nations, 'celebrating' the UN's sixtieth anniversary and a follow up to the 2000 Millennium Summit.The Summit made R2P and R2Prev the subject of an international agreement.The debates before the Summit could be seen as a process of international socialisation, but not in the one-directional and straightforward way perceived by Finnemore and Sikkink.It was a community of practice -consisting of individual norm entrepreneurs and states -who discussed the norm before the Summit.During the Summit, the community of states continued negotiations on the norm, sometimes influenced by norm entrepreneurs.The norm R2Prev changed and developed into the shared understanding we can find in the Outcome Document.
The World Summit Outcome Document affirms most of the basic principles of R2P as developed by the ICISS: the idea of sovereignty as responsibility and assertion of the threefold responsibility of the international community -to prevent, to react and to rebuild -in case a state is unable or unwilling to protect its population.Furthermore, the World Summit reaffirms the threshold of the four mass atrocity crimes.
With regard to prevention, the Document is clear that the goal of the norm R2Prev for states is the prevention of -the incitement of -the four mass atrocity crimes.
How should this goal be achieved?The responsibility to protect one's own population is unambiguous in the World Summit Outcome Document.Mass atrocities -and their incitement -within states have to be prevented by all necessary and appropriate measures, with the word "necessary" indicating that a state cannot choose to protect its population when it suits them, but rather that a state has to (always) use all necessary means to do so.
The international community should, "as appropriate, encourage and help states to exercise this responsibility". 95Therefore, this is a norm of outside assistance in preventing mass atrocities.This language is weaker, with no absolute responsibility for outside states to take action.Furthermore, in times of crisis the international community, through the UN, has a "responsibility" to use all peaceful meansdiplomatic, humanitarian and other means -to protect populations.This is a stronger obligation.The international community furthermore should "support" the UN in establishing an early warning capacity.Moreover, states intend to commit themselves to help others build capacity to protect their populations from the four mass atrocity crimes and assist them if they are under stress before crises and conflicts break out "as necessary and appropriate".The intention of states to commit themselves is not a strong obligation, but the words "necessary and appropriate" mean that in the event of states committing themselves, their obligations are quite far-reaching.prevention in comparison to the other pillars. 96However, the lack of specific proposals -apart from strengthening the UN's early warning system and endorsing already existing measures -is also recognised. 97The scope and exact circumstances of the responsibility for outside actors is yet unclear and this makes the norm less precise and thus less robust.Subsequent application and interpretation by states -in their policies, in times of mass atrocity crises and in the deliberation in the GA that is argued for in the Outcome Document -will demonstrate to what extent states considered the norm to be precise enough to apply it and can help to make the norm more precise.The planned discussions on R2P within the GA are a further step of institutionalisation, as well as an indication of a shared understanding of procedural legality, essential in interactional law creation.
The endorsement of the Summit of R2Prev indicates impressive concordance concerning the new norm.The World Summit presents a clear rhetorical commitment by the international community.Though the endorsement of the World Summit Outcome Document was unconditional, the text of the Outcome makes it clear that the acceptance of a responsibility for outside actors to prevent mass atrocities was not.Some UN Member States were still cautious about the principle.It might be that they only rhetorically supported the concept because they considered it as watered down anyway, or because they were just not planning on implementing it.Some states later argued that the World Summit only committed the international community to further consider R2P. 98 procedures were laid down by which norm leaders coordinate disapproval and sanction norm-breaking behaviour.One could argue that the sanction for a state being unwilling to protect its own population is the option of collective military enforcement if all other measures fail.But for states who fail to provide assistance to states under stress or who fail to provide help to build capacity to protect the population of another country, no sanctions are identified.This could also influence the durability of the norm: since sanctions are unclear, it is unpredictable how the norm will deal with violations.
Though institutionalisation is often the trigger for a norm cascade, I would argue that the process of the norm spreading is different for R2Prev from what was theorised by Finnemore and Sikkink.They argue that in the stage of norm emergence, before a tipping point is reached, domestic actors have to convince their governments to adhere to a certain norm.However, in this case the norm entrepreneurs were primarily international norm entrepreneurs, who -together with some statesapproached specific countries leading up to the World Summit, 99 partly because the responsibility to assist other countries is an international norm and the negotiation process took place on an international level.Furthermore, Finnemore and Sikkink argue that a norm cascade will mostly take place as soon as 1/3 of the countries endorse the norm.This process went differently with R2Prev.The World Summit was the moment at which all states officially and 96 R. Gerber, 'Prevention: Core to the Responsibility to Protect', e-International Relations 2011, p. 29. 97Bellamy 2008, supra note 8, p. 141. 98Bellamy 2008, supra note 8, p. 25.Those states were India, Cuba, Sudan, Venezuela, Pakistan and Nicaragua. 99Evans 2009, supra note 89, p. 50.
unanimously endorsed the new norm.A process in which each state, one after the other, accepts the norm, a process which can be found concerning issues such as women's suffrage, did thus not happen concerning R2Prev.This can be partly explained by the content of the norm.R2Prev assigns responsibilities on the international community to act.It is a norm that would only have serious effects if a great majority of states lives up to it.An exception could be if certain regions are in agreement about the norm and in dealing with regional crises all act in accordance with it.So the norm entrepreneurs and states tried to spread the norm by making it a prominent new idea on the world stage and through the persuasion of many states on the way to and during the World Summit, it was given a place in a document that was endorsed by all state leaders at once.How likely is it that states are active in pursuing the prescriptions on R2Prev?This is the question of the robustness of the norm, first of all the specificity.For a norm to be effective it needs to cascade around, be internalised in the actions of states and be robust.Since the norm is yet quite imprecise, it is possible that many states only rhetorically supported the norm and were not planning to change their behaviourespecially when assisting as outside states.The behavioural prescriptions were unclear and they thought they would not be held accountable by others in the international community.If many fail in implementing R2Prev -especially critical states which experienced mass atrocities in their own territory and those outside states with the power to act to prevent them -even an overwhelming majority of states supporting the norm is not sufficient to argue that the norm has cascaded around in a way that changes the effects on the ground.
Though clear prescriptions might lead to a habit of acting in a certain way -a robust norm cascading might lead to effective internalisation -this is not necessarily positive for the success of R2Prev.This takes us back to the earlier debate on specificity.On the one hand, specific rules create expectations for all actors -those taking action and those facing action from the international community -and the patterns of action will be more predictable, which will in the longer term strengthen the norm.On the other hand, too specific prescriptions that become a habit can also mean that the wrong measures will be taken which are not suitable for the situation at hand.
In conclusion, in order to strengthen the R2Prev norm, states and other actors have to interpret and apply the norm and implement specific actions for every situation that deserves action, whether it be long term or short term operations.Experience with different situations can lead to clearer -an internal quality of law in interactional lawand more specific prescriptions.R2Prev requires pro-activeness on the side of states, regional organisations and the international community in general.If this is achieved, the norm demonstrates that it can deal with challenges and become durable and thus more robust.If successful, this will also lead to congruence between the norm and the actions of officials, an important criterion of legality in the interactional theory of law.Furthermore, specific prescriptions will make it clear whether the expectations of the norm R2Prev are realistic for outside actors.

After the World Summit
Though the World Summit agreed on a certain interpretation of R2Prev, two different interpretations of R2P and R2Prev were used in the subsequent debates.The first strand is primarily supported by states, especially from the global South.It focuses on the responsibility of states to protect their own population and provide mutual assistance on a consensual basis. 100R2P is primarily meant to enable and encourage states and non-state actors to prevent mass atrocities.Armed intervention can be the most extreme measure to be applied, but it is not the focus of R2P.The second strand sees R2P as a response to the dilemmas of humanitarian intervention.They see R2P as a way to mobilise political support and generating political will to avoid 'new Rwandas'.This is the way most academics apply the concept, the interpretation some would have liked states to have committed to, as well as the interpretation used by those who criticise R2P, for example with the argument that it legitimises breaches of sovereignty.Of course, sovereignty as responsibility is essential in both interpretations, and both would acknowledge the influence that the debate on humanitarian intervention had on R2P.But they clearly differ on the centrality of armed intervention, the scope of prevention and the function of R2P.Though the second interpretation has been prominent in the debate outside the UN and has some influence in constructing the norm, the norm under study is an international norm, primarily agreed to by states.
In the immediate aftermath of the Summit, in 2006 and 2007, mostly Arab and Asian states were wary about the consequences of R2P for state sovereignty. 101They protested against the norm in order to block its implementation inside the UN.Western states and many sub-Saharan African and Latin American states defended the norm. 102The new Secretary-General Ban Ki-moon and his Special Adviser on the Responsibility to Protect Edward Luck, who started in February 2008, 103 worked on a renewed consensus and conceptual refinement for R2P. 104e appointment of a Special Adviser on the Responsibility to Protect anchored the development of R2P structurally within the UN system.This appointment ensured that R2P would be further institutionalised within the UN system.Ban Ki-moon aimed and succeeded at making the shared understanding reached at the World Summit deeper, more specific and more precise, also searching for more concordance around the norm and a better normative fit with the normative frameworks of different countries. 105However, he focused on a concrete implementation agenda for R2Prev instead of focusing too much on creating new concepts.Acting as a norm entrepreneur, he realised that norms have to be precise in order to be effective and induce compliance.

UN Secretary-General -Implementing the Responsibility to Protect (2009)
The process leading up to the January 2009 report of the Secretary-General saw a renewed momentum around R2P. 106 The report proposes a three pillar approach for implementing R2P, consisting of the protection responsibilities of individual states, international assistance and capacity building as well as timely and decisive response 100 Bellamy 2011, supra note 71, p. 8. 101 Ibid., pp.27-31. 102Ibid, pp.27-31. 103Ban Ki-moon wanted the Special Adviser for working out this implementation agenda, but did not get funding for the position by the UN Budget Committee.However, he decided to appoint the Special Adviser anyway in February 2008, without funding (Bellamy 2011, supra note p. 32).This case shows the reluctance, in 2006 and 2007 up to the beginning of 2008, of some member states to work on the further implementation agenda of R2P. 104Bellamy 2011, supra note 71, p. 31. 105Ibid., p. 32. 106 from the international community.107Ban Ki-moon stresses the value of prevention in combination with, if prevention fails, an "early and flexible response tailored to the specific circumstances of each case". 108This approach supplements the existing three pillar approach of prevention, reaction and rebuilding.The new approach specifies when national responsibilities to prevent and protect transform into international responsibilities.International assistance and capacity building are part of a strategy to prevent mass atrocities, while the 'timely and decisive response' is clearly a reaction strategy.
Ban Ki-moon includes structural prevention in his definition of R2Prev. 109Part of his proposed prevention strategy is a focus on preventing and stopping the incitement of mass atrocities. 110Ban Ki-moon stresses the legal obligation to abstain from inciting one of the four specified crimes, 111 since cases of inciting hatred can be referred to the International Criminal Court. 112He identifies a wide range of actions that states can take to prevent mass atrocities and that others can take to assist.This implementation strategy helps to specify the norm.It makes R2Prev more detailed, but it is possible that it might make it so detailed as to be less intelligible for member states.Furthermore, a clear division of responsibilities between the UN, states and regional organisations should still be specified.This narrow but deep approach connects R2Prev to many other policy agendas 113 and strengthens the normative fit and coherence with other policy agendas of member states and the UN.This can enhance R2Prev's compliance pull.
The Report emphasises that paragraphs 138 and 139 are part of conventional and customary international law, according to which states have obligations to prevent and punish genocide, war crimes and crimes against humanity.Though ethnic cleansing is not a crime in its own right in international law, it may constitute one of the other three crimes.Ban Ki-moon refers to the Independent Inquiry into the action of the United Nations during the 1994 Rwandan Genocide which states that the Convention on the Prevention and Punishment of the Crime of Genocide imposes a responsibility for states as part of the international community to act in the face of mass atrocities. 114The Secretary-General stresses that the role of R2P is to reinforce these existing obligations for member states. 115 order to strengthen early warning, Ban Ki-moon calls for setting up a joint office of Special Advisers on the prevention of genocide and on the responsibility to protect.In addition to providing early warning, this proposal is also part of an institutionalisation strategy.This joint office could help the SG with making proposals for his implementation agenda and it could be the central point from which R2P and R2Prev could spread around the UN system.This institutionalisation could lead to the internalisation of R2P and R2Prev concerns within the UN system.Such a focal point could prevent member states from silently killing the debate on the implementation of the R2P if the UN itself did not get any further because of institutional inertia. 116n Ki-moon thus calls for structural prevention.Those seeing R2P as a concept which is used to mobilise political support for a prevention or reaction strategy 117 argue that using the terminology R2P and R2Prev for structural measures would weaken its call in times of real crisis.R2P might lose its attraction and power as a mobilisation strategy to spur the international community into action.This way of perceiving R2P means comparing "speaking R2P" with "speaking security", an analogy based on the theory of securitization developed by the Copenhagen School. 118The Copenhagen School argues that issues that are securitised result in setting aside normal politics and allowing for extraordinary measures to be taken.Speaking R2P can work as speaking security.However, invoking R2P in all cases in which preventive action is necessary, thus also including structural prevention, would make the appeal of the call of R2P weaker. 119Therefore, Stamnes proposes to try to include structural R2Prev prevention measures into normal politics, i.e. politics as usual. 120This is in line with the Copenhagen School, who argue that is generally better to de-securitise an issue and make it a normal political issue.
Though a sympathetic proposal, it is based on the unconvincing assumption that it is possible to ensure sufficient attention to long term, mass atrocity prevention without using the concept R2Prev as a label.However, structural measures are often painful and costly for powerful states, and it is not foreseeable that they will be consistently implemented and sufficiently funded without the link made to R2P and the promises made in 2005.
The Secretary-General's Report was followed by a GA debate.Before this debate, norm entrepreneurs lobbied to create a wider and deeper shared understanding on R2P.State members of the Group of Friends of R2P in New York, the UN Special Adviser on R2P and civil society actors managed to create, unexpectedly, 121 the wider consensus that became visible during the debate. 122One could argue that this demonstrates a norm cascade of R2Prev, since many states were convincing other states into supporting the norm.However, as long as R2Prev is insufficiently specified, it is hard for the norm to cascade and be subsequently internalised into the actions of states.
powers and members of the Non-Aligned Movement that had been reticent or hostile to R2P in the past. 123Only Cuba, Nicaragua, Sudan and Venezuela sought to roll back the conclusions of the 2005 World Summit.
The consensus was especially strong on strengthening prevention efforts, even among critics of the reaction side of R2P. 124The support of a critical mass of states seemed to be reached on this broad principle.Agreement emerged on specific proposals on prevention, such as national obligations under R2P, ratifying human rights treaties, adopting accountability measures, education and awareness, international and regional mediation, stand-by abilities, the use of the Peacebuilding Commission and, albeit with some concerns, early warning mechanisms were advocated. 125The proposal to establish a joint office for the Special Representative for the Prevention of Genocide and the Special Adviser for the Responsibility to Protect in order to strengthen UN capacity now got support from many more states than before, including those from the global South. 126 structural prevention, which was advocated in the SG's report, the viewsespecially on whether or not to include economic development -diverged.Many Western advocates for R2P, including NGOs such as the International Coalition for the Responsibility to Protect, resist including economic development as part of an implementation strategy of R2P and R2Prev.This seems to be partly inspired by their fear of broadening R2P to such an extent that it becomes unhelpful in generating political will, referring back to their interpretation of R2P.
On capacity building, another element of a structural approach addressing root causes, a higher degree of consensus between the 'global North' and the 'global South' exists.General consensus emerged on measures to build civilian prevention capacity and strengthen the rule of law. 127On the mode of assistance, at least there was broad agreement that this should only be done if the state concerned consents and cooperates. 128Furthermore, many states identified the need for increased cooperation between the UN and regional organisations, especially the AU but also ECOWAS, in building prevention capacities. 129More detailed thinking and more specification of the norm seems necessary on this issue.
Though agreement on the first pillar identified in the report is expressed, states "generally shied away from commenting on how they planned to exercise this responsibility". 130The overall consensus is promising for R2Prev, as it shows an increased concordance on the issue and it seems it could withstand the attacks on the norm of 2006 and 2007.Since the norm R2Prev is not yet internalised, we should see its repetition and public affirmation in diplomatic discussions as strengthening the norm.This optimism should, however, be cautioned.Although public denunciations of the concept are becoming less likely, states can use many ways to stop the implementation of R2P, e.g. by ensuring it will not get enough funding, which is very possible indeed seeing the UN's continuing budget crisis, or by pointing to competing priorities. 131e GA resolution following the debate took note of the Secretary-General's report and of the timely and constructive debate, and decided to continue its consideration of R2P. 132Only the removal of the words "with appreciation" -in the clause about taking note of the Secretary-General's report -could ensure that the resolution was adopted with consensus. 133This shows that states were not in agreement on all of the proposals of the Secretary-General.However, most scholars are positive about the value of this GA resolution in the normative development of R2P, 134 since it reaffirms the fact that the GA will take up the role of further discussing and therefore implementing R2P.
The shared understanding that was found in 2005 was certainly moved forward through the 2009 debate.The general focus shifted from the immediate reaction to mass atrocities to a long-term approach of protecting civilians and preventing mass atrocities.Bellamy asserts that if the agenda that emerged from the 2009 SG's report and the GA debate were implemented, it "would almost certainly help prevent genocide and mass atrocities and strengthen the international community's response". 135e debate on early warning continued in 2010.At that time, the Department of Political Affairs played an important role in collecting the information on risky situations, but it lacked the capacity to conduct early warning analysis.The Special Adviser on the Prevention of Genocide, appointed in 2004, collects information that already exists within the UN system and developed an analysis framework to assess the risk of genocide, making a step in developing UN early warning capacity.Furthermore, the current Special Adviser Francis Deng is increasingly cooperating with the Special Adviser on R2P, Edward Luck.However, opposition still exists among UN member states to improve early warning capacity within the UN System, caused by concerns about sharing intelligence information and state sovereignty.
In his Report on Early Warning, Assessment and the Responsibility to Protect the SG urges for more information sharing within the UN system.Furthermore, information should be assessed better and through an R2P lens.This could ensure efficiency and system-wide coherence when dealing with R2P situations. 136onvening mechanism and the OSAPG will now incorporate all four mass atrocity crimes in its work.These are all steps on the way to an integrated joint office.
The question of the role of regional organisations received attention in a July 2011 Report137 by the SG -his third report on R2P -and a subsequent informal dialogue between member states, the SG and his Special Advisers on the Prevention of Genocide and on the Responsibility to Protect, regional organisations and civil society representatives.The SG argued that R2P now needs to be institutionalised by states and internalised by societies, though he acknowledged the need for refining the concept. 138 emphasised that the role of (sub-)regional organisations is widely recognised in operational, direct prevention efforts, and that it is closely related to the third pillar as identified in his 2009 Report.This implies that direct prevention and reaction are not always substantially different, but are all actions on an ongoing scale.The SG calls the improvement of operational prevention and the cooperation of the UN with regional partners a priority.Though the SG sees the role of regional organisations as less obvious when it comes to structural prevention, they can contribute to the development of norms, standards and institutions promoting tolerance, transparency, accountability and a constructive management of diversity. 139Furthermore, regional organisations can work on planning in order for regions to be prepared for the effects of (manmade) disasters.They can provide the communication and information link between local and national actors and global actors on what is needed in international assistance to countries under stress. 140The SG says improvements have been made in the cooperation with (sub-) regional organisations on prevention, through mediation, facilitation and dialogue.He expects deeper global-regional partnerships for implementing R2Prev in the years ahead. 141The amount of regional initiatives set up in the last years related to R2Prev to which the SG refers demonstrates an institutionalisation of the principle on the regional level, though to differing degrees. 142his can be seen as part of a norm cascade of the norm R2Prev around many regions in the world, and institutionalisation may be a step towards internalising R2Prev into the actions of regional organisations.
In his report on regional organisations the SG repeats a call that was made by the ICISS 10 years earlier: prevention is under-resourced locally, nationally, regionally and globally. 143Though we have noted that the shared understandings around the importance of prevention seem to be substantial and deep, this supposed change in attitude -internalising prevention strategies -needs to be supported by financial resources to have effect.States need to 'put their money where their mouth is' in order for R2Prev to be successful.
During the 'interactive dialogue' following the SG's report between UN officials, specialists and member states, structural prevention efforts received less attention from member states than direct prevention. 144Brazil and some other states called for the inclusion of social and economic development in structural prevention strategies, because of their link with political stability.Switzerland, the EU and Japan argued that the most important structural efforts are good governance, the rule of law and respect for human rights.This demonstrates again some norm contestation about what it required to prevent mass atrocities in the long term and thus currently some lack of clarity about what the norm requires.
Concluding, the debate showed that progress on regional cooperation on prevention is underway and that the shared understandings and the willingness of states are present.However, the need for more funding at all levels -global, regional, national and local -had not been addressed.Furthermore, the question about the division of responsibilities for taking preventive action between the UN and regional organisations remains.

ΙΙ.2 R2Prev and the Sources of International Law
To grasp the extent to which R2Prev is or could become a legal norm, this article will address its basis in the sources of international law.Because the 2005 World Summit Outcome is the statement of R2Prev endorsed by member states of the UN, this formulation will be used.The World Summit is not a treaty, and R2P and R2Prev have never been adopted in a treaty. 145Therefore, one has to consider other legally binding sources of international law which support R2PRev.R2Prev applies to the four mass atrocities which have been, to a different degree, defined in international law: genocide, war crimes, ethnic cleansing and crimes against humanity. 146The responsibilities of a state towards one's own population are found in international human rights law, international humanitarian law and the growing body of international criminal law.These responsibilities can be considered jus cogens. 147The 1948 Genocide Convention and the Rome Statute establishing the International Criminal Court (ICC) are both considered customary international law and they forbid a state to commit mass atrocities against its own population.
Of these four crimes, war crimes and genocide are the better defined and the scope of a state's responsibility to prevent them and protect the population from these crimes is clearer than in relation to crimes against humanity and ethnic cleansing.The 1948 Convention on the Prevention and Punishment of the Crime of Genocide states in Article 1 that the crime of genocide itself is prohibited and that states have a duty to actively prevent genocide and punish the perpetrators.The definition of genocide in Article 2 is adopted by the International Criminal Tribunal for the former Yugoslavia (ICTY), the International Criminal Tribunal for Rwanda (ICTR) and the ICC.The exact scope of an outside state's responsibility to prevent genocide and punish the perpetrators is contested in the genocide regime.In the Bosnia v. Serbia case, the ICJ stated that the Article 1 obligation requires states to "employ all means which are 144 International Coalition for the Responsibility to Protect, 'Interactive dialogue of the UN General Assembly on the role of regional and sub-regional arrangements in implementing the Responsibility to Protect.ICRtoP Report.' 145  reasonably available to them" so as to prevent genocide as far as possible. 148States that have influence and information have a legal obligation to take positive action to prevent genocide. 149Arbour argued that the ICJ ruling implied an emerging legal duty to prevent genocide, 150 which would thus be supportive of the R2Prev regime.
The four Geneva Conventions entail a responsibility for states to refrain from committing war crimes and Article 1, common to the four Geneva Conventions, provides the obligation for states to undertake to "ensure respect for the present Convention". 151Article 8(2) of the Rome Statute 152 is a further codification of war crimes, and is considered to reflect customary international law.The obligation for states to 'ensure respect' for the law can be implemented via diplomatic and other non-military means aimed at encouraging compliance with the law.This includes preventing and halting the commission of war crimes by non-military means as well as bringing the perpetrators to justice.This implies an obligation to help other states in preventing mass atrocities, and thus strengthens the legal underpinnings of the R2Prev regime.
On ethnic cleansing, a definite legal definition is not yet found, but ethnic cleansing in a broad sense -the forcible deportation of a population -is regarded a crime against humanity under the Statutes of the ICC 153 and the ICTY. 154When one employs stricter definitions of ethnic cleansing, which includes mass-scale killing of a designated population, the Statutes of the ICTY and ICC would define themdepending on the facts -under the heading of either war crimes or genocide.The legal regimes for those crimes then apply.
When it comes to crimes against humanity, the Rome Statute clearly provides that parties to the ICC Statute have a legal duty to investigate alleged crimes against humanity and punish the perpetrators.However, there is still no specialised convention which establishes a duty to prevent crimes against humanity and punish the perpetrators. 155The definition of crimes against humanity in Article 7 of the Rome Statute is considered customary international law and jus cogens.Article 7 prohibits many different acts, but only when they are committed as "part of a widespread or systematic attack directed against any civilian population, with knowledge of the attack". 156 the committing of the four mass atrocity crimes is clearly forbidden in international law, whereas a responsibility to prevent is emerging in the legal regime on genocide and a duty of further compliance with the law is present in the regime on war crimes.Acts under the heading of ethnic cleansing are dealt with by the legal regimes of the other crimes.
Apart from the specialised conventions, how is the issue of responsibility -an essential element of R2Prev -dealt with by the International Law Commission's (ILC) Draft Articles on State Responsibility for Internationally Wrongful Acts, which were adopted by the ILC in 2001 and have been widely endorsed, most importantly by the ICJ?In Article 41 of the Draft Articles, 157 the ILC argued that in relation to grave breaches of international law, other states have a certain obligation to foster compliance with the law. 158These grave breaches were defined by the ILC as a "serious breach" 159 ("a gross or systematic failure by the responsible state") 160 of a "peremptory norm of general international law". 161Of the four mass atrocities, the ILC in its commentaries refers to the prohibition of genocide, the basic rules of international humanitarian law, seen as intransgressible by the ICJ in its ruling on the Legality of the Threat or Use of Nuclear Weapons, 162 as peremptory norms of general international law and stresses that the list is not exhaustive and that new peremptory norms can develop.
Article 41(1) entails a positive obligation of states "to cooperate to bring the serious breach to an end through lawful means".This idea comes close to the idea of a collective responsibility as defined in R2P, 163 mostly in its reaction element.The Commission stated that this obligation for states to cooperate applies irrespective of whether they are individually affected by the serious breach or not. 164Two forms of action can be taken to fulfil this obligation.First of all, the ILC mentions "a joint and coordinated effort by all states to counteract the effects of serious breaches" 165 of peremptory norms of international law.Secondly, international cooperation could be a solution, which would be "organised in the framework of a competent international organisation, in particular the United Nations". 166The ICL acknowledged that this positive duty of cooperation "may reflect the progressive development of international law". 167ahn argues that the Outcome Document, by extending the responsibility to the four mass atrocity crimes, goes further than the ILC, which referred to the serious breaches of peremptory norms.In the case of genocide, by stating that the Outcome VOL 5:1 Document goes further than the ILC, Stahn thus implicitly makes a distinction between genocide and a serious genocide. 168I would have to argue that this distinction is an impossible one, since to define a crime as genocide already means passing a high threshold.The ILC agrees on this. 169Furthermore, the reference by the ILC to the basic rules of international humanitarian law as peremptory norms means that a war crime -in itself a serious breach of international humanitarian law -could well fall within the definition of serious breaches of peremptory norms.Furthermore, the legal norms on crimes against humanity and ethnic cleansing could develop into peremptory norms of international law, as the ILC leaves this possibility explicitly open.Though the World Summit goes some steps further in relation to crimes against humanity and ethnic cleansing, I would have to disagree with Stahn that the Summit went further than what was agreed by the ILC concerning genocide and war crimes.
Article 48 of the Draft Articles of the ILC entails that when a mass atrocity is already taking place and an internationally wrongful act is being committed, all states may ask for the "cessation of the internationally wrongful act, and assurances and guarantees of non-repetition" when the "obligation breached is owed to the international community as a whole". 170These obligations owed to the international community are similar to the erga omnes obligations, which include outlawing genocide.The Court does not use the term erga omnes since it recognises the possibility of the development of new obligations towards the international community as a whole.Though Article 48 deals with diplomatic means of stopping an ongoing mass atrocity, and not with preventive action before a mass atrocity, at least it gives states a right to ask for the cessation of atrocities.This can, if action is taken quickly enough, be part of an R2Prev strategy.
In sum, the Draft Articles do not provide an explicit responsibility for states to prevent a possible mass atrocity.But the ruling in Bosnia v. Serbia combined with the explicit obligations of states in the ILC's Draft Articles to cooperate to end serious breaches of peremptory norms of international law imply that R2Prev is part of a progressive development in international law in which the next stage could be a general obligation for states with the means and the information to prevent a mass atrocity in another state to do so.

ΙΙ.3 R2Prev and Interactional Law
R2Prev is created and spread by the hard work of norm entrepreneurs. 171Member states, NGOs such as GCR2P and ICR2P, academics (the journal Global Responsibility to Protect) and legal practitioners participate in dense and specific interactions and form a community of practice, as seen in the three GA debates.
As noted, not all participants in this community of practice have the same opinion about R2Prev and not all members of the community of practice interact in the same forums, which goes some way towards explaining why different interpretations of R2P and R2Prev still co-exist.However, interaction takes place between those, for example, that see R2P purely as a concept to deal with reaction and those that want to focus more on structural prevention; in other words, actors are influenced by one another. 172The growing consensus shows a "commitment to communication, a willingness to listen, and openness to reciprocal modification of outlook". 173 the moment, a continuous and inclusive negotiation about collective meanings is going on, proving the point made by Brunnée and Toope that norm diffusion is not unidirectional, but interactional.A process of deepening shared understandings has taken place and is still taking place.Member states agree on the importance of prevention as an aspect of R2P policy making.However, if R2Prev is to be based on a deeper and more specific shared understanding including divisive issues such as structural prevention, interaction needs to continue.
How deep are the current shared understandings?The R2P and R2Prev norms are, for many, about deep moral commitments; this is the case for norm entrepreneurs such as Gareth Evans,174 for the NGOs working to foster the norm, and for some states who experienced mass atrocities or feel morally compelled to act to prevent mass atrocities.However, some states do not share these deep moral commitments, since R2P and R2Prev were sometimes seen to go against their deep moral commitments to sovereignty.Therefore, communication as a substantive commitment 175 is essential to come to the required substantive commitments which can then create substantive interactional law.This communication is ongoing through the debates organised in the GA and whenever cases are discussed that might require action according to R2Prev.Though deep shared understandings are not (yet) shared around the globe, a basic shared objective in the debate needed to create interactional law is present, namely the objective to prevent mass atrocities.Interactional law lends support for acknowledging differences in the means used to allow for regional variation in the action taken, while sharing the same goals. 176e political shared understandings among states were supported by legal developments identified earlier, such as the ICJ's ruling in Bosnia v. Serbia on the prevention of genocide and the ILC's Articles on State Responsibility.This demonstrates R2Prev's basis in the interaction with legal norms.

Eight criteria of legality
Though some criteria of legality have been discussed throughout the article already, it is good to analyse all the criteria together.One cautionary remark needs to be made.Brunnée and Toope argue that one needs a shared understanding that law is needed in a given context.Proof in that direction is mixed.On the one hand, the building of more specific norms in the World Summit and the GA could be an indication of the willingness of states to create law and an indication of a growing sense of procedural legality. 177On the other hand, the possibility cannot be ruled out that many states would like to set a political framework for action, but that they are not interested in developing a legal norm.Nevertheless, it is of importance to analyse the chances of R2Prev developing into a legal norm.
First of all, the criterion of generality: to be legal, a norm must prohibit or require a certain conduct.R2P and R2Prev prohibit mass atrocities by member states against their own population.Furthermore, it requires states to protect their citizens to the best of their abilities against such mass atrocities.To what extent it implies a responsibility to ask for outside assistance if this is needed in order to prevent mass atrocities is not yet clear.R2Prev furthermore requires outside states, if asked by another state, to assist in preventing mass atrocities.However, there is insufficient clarity about who has to act when.To whom should a state turn if in need of assistance?Should the state first ask a (powerful) neighbour, or a regional organisation, or the UN?The debate until now seems to focus on how all these actors can work together in a better way, but when it comes to responsibilities, it needs to be clear to what extent different actors have a responsibility to take action.
Secondly, the R2Prev was promulgated at the 2005 World Summit.However, since then the debate on the implementation of R2Prev has been going on inside and outside the UN and different interpretations of the norm have been given, and it has been applied in some cases.So the norm development process is still ongoing, and the World Summit should not be seen as the only promulgation of R2Prev, 178 though it is the one with the most widespread political support.
Thirdly, the norm is indeed prospective.Though disagreement could exist on the extent to which R2Prev is applicable to situations that were already ongoing at the time of the World Summit, it is clear that it cannot be applied retrospectively.
Fourthly, the law must be clear about what it requires.As explained under generality, it is not always clear what R2Prev permits, prohibits or requires.Bellamy asserts that "the demands imposed by pillars two and three of RtoP are indeterminate". 179This indeterminacy is also due to the fact that actors have competing judgments about the facts of the case -whether R2Prev action should be taken -and then have different assessments about what the best course of action is.Add a lack of clarity about what needs to be done by whom in certain circumstances and it becomes clear that the sense of legal obligation is weakened. 180fthly, the norm should avoid contradiction.The confirmation that the norm only applies to the four mass atrocity crimes does help to avoid contradiction.Though the exact requirements of the norm are not yet clear, it is clear that it does not prohibit and require the same action at the same time.But it is not unforeseeable that some actors will argue that a situation requires preventive action, while others would argue that the time for prevention is over and states have to apply measures from the 177 Evans 2009, supra note 89, p. 351. 178Brunnée and Toope argue that the World Summit Outcome Document should not be seen as the "culmination of the norm building process, but rather as a platform for further normative interaction and deliberation" (Brunnée and Toope 2010b, supra note 58, p. 204). 179Bellamy 2011, supra note 71, p. 86. 180Bellamy argues that "the indeterminacy of pillars two and three severely restricts R2P compliance pull, and hence its ability to pull states to agree as to what needs to be done and who should contribute" (Ibid., p. 86).
reaction pillar of R2P.181This is more a question of when different aspects of the norm should be applied.Furthermore, it is possible that some will argue for preventive action in the R2Prev framework, while others argue that the seriousness of the expected atrocities does not yet warrant the application of the R2Prev framework.This is a problem for R2Prev, since one can never know for sure whether assessing something through an R2Prev lens was justified if mass atrocities have not yet taken place.However, though people will argue about the label to be applied, few will argue against preventive measures being taken.Continuing application and interpretation might provide the answers to these questions.
The sixth requirement for a norm to conform to the eight criteria of legality is to be realistic; to not demand the impossible.The fact that R2Prev is now clearly only focused on the four mass atrocity crimes makes its -especially concerning outside assistance -more realistic than a broader norm.It is problematic that it is not yet settled what the norm exactly demands.Let's consider the different aspects of the norm.It is realistic to expect a state to protect its own population from mass atrocities.If assistance is needed, it is politically painful but perfectly possible to demand that states ask assistance from other states.If this were unrealistic, the whole three-pillar structure from the 2009 Report would be unrealistic.For other states, it is realistic that if asked to assist a state in protecting its own population, a state will do so by sending civil service -be it to reform or assist the state's security sector, or to negotiate among different parties.However, if states ask other states to provide military assistance to protect a particular group, this is not a straightforward and easy request and the answer depends on the political circumstances.
Further specification and more clarity could help lessen these problems.If it were clearer who has to assist in certain circumstances, and the burden is more or less shared among different states, regional organisations and the UN, it might be that those responsible feel a sense of obligation to take action.
Concerning the seventh criterion, constancy about what the law requires, it seems too early to draw definite conclusions.Again, because R2Prev only applies to the four mass atrocity crimes, discussion about other human rights violations falling under the R2Prev framework will probably take place less often.These calls for a broader application are morally understandable, but they can harm the shared consensus.In relation to what has to be done in preventing mass atrocity situations, much is still unclear, so constancy about the action that the law requires -provided R2Prev becomes a legal norm -is not yet reached.If a division of tasks is carried out properly, and the norm has become clearer, the norm will become more constant as to what it requires.
Concerning the eighth criterion, congruence between the norms and the actions of officials, the reactions of the international community to crises in which mass atrocity crimes might have been impending have seen mixed results. 182Bellamy states that "the mutual recognition of a positive duty to exercise pillars two and three is inconsistent at best". 183The case of Kenya in 2008 indicates that official action in line with R2Prev has successfully been taken.
Since the norm is still in need of further clarification and the implementation is underway, it is also hard to judge whether certain behaviour was in line with R2Prev.The claim that that R2P's reaction side is applied selectively is understandable. 184Prev therefore needs more clarity and specificity.Continuing interpretation and application might provide this, but further codification based on inclusive shared understandings and fulfilling the criteria of legality can also be a useful next step in clarifying and specifying R2Prev.If fulfilling these conditions, codification can be achieved through GA resolutions, a treaty and/or legitimate judicial decisions.

Practices of legality
Since R2Prev cannot properly be considered a legal norm, it might be too early to consider the practices of legality.However, some aspects of the norm are clear.As regards the primary responsibility of states to prevent mass atrocities, one can see that those states that opposed or did not take part in the emerging shared understandings -such as Sudan and North Korea -are also the ones who do not protect their population against mass atrocities.Although a shared understanding among most member states is evident in relation to the first pillar of R2P as defined in 2009 and while this pillar conforms most strongly to the criteria of legality, practice against the emerging legal norm still exists.Those states that have failed to protect their own population have been criticised severely by most other states, demonstrating that sovereignty is at least no longer considered a shield from criticism in case of mass atrocities.Though the interpretations of the world's reaction to the 2008 post-election crisis in Kenya differ, it is widely seen as an effective application of the R2Prev norm to build diplomatic pressure and prevent escalation. 185e most important point is that if one judges a norm by the immense expectations some actors have, it will often fail.If we only look at the material interests that states have, we will fail to see normative developments.If we only consider the short term, we might feel the need to dismiss a norm when it first fails, while overly celebrating it when it first seems to be working.Looking at the long term is important; even more so for the preventive side of R2P, since only in the long term we can see whether the action taken in line with R2Prev has prevented mass atrocities.Then one can consider whether a practice of legality emerged around the norm R2Prev.prevent mass atrocities.Creating and maintaining law is hard work, which needs to be done continuously if one aims to prevent mass atrocities.

Conclusion
A wide consensus exists about R2Prev, though states differ on the issue of structural prevention.A process of institutionalisation within the UN system is slowly taking place.
As is argued by Finnemore and Sikkink, a process of institutionalisation can help to clarify and specify the norm, while the norm cascade starts mostly after the institutionalisation.In the case of R2Prev, institutionalisation, the norm cascade and the further clarification and specification of the norm, are happening at the same time.Norm entrepreneurs, UN member states and academics are still arguing about what R2Prev exactly means.At the same time, the norm is cascading and starting to internalise, by its implementation in the UN system, regional organisations and slowly in member states.This finding is interesting for Finnemore and Sikkink's theory.As has been demonstrated, the norm contestation, institutionalisation and cascade are processes that cannot be clearly separated.
The norm development of R2Prev teaches us that broad norms -norms still in need of specification -might be the best way to convince states at the stage of norm entrepreneurship, as can be seen in the consensus at the 2005 Summit.This however implied that when that consensus has been reached, it has to be regained every time the norm sets a next step towards specification and implementation.For a norm such as R2Prev, with such a broad consensus in the World Summit, but with so many specifications still needed, it is no wonder that the implementation process is not without hurdles and that it takes time.
Finnemore's and Sikkink's framework of a life cycle with clearly separated stages probably works better for simple and straightforward norms which forbid certain behaviour, i.e. constraining norms.R2Prev is partly prescriptive and this leaves open many possibilities for contestation.Furthermore, R2Prev as an international norm, also steering the behaviour among states, needs consensus every time it wants to take a new step within the UN system.It seems that it is both the prescriptive and the international character of R2Prev that explains why its development differs from the model foreseen by Finnemore and Sikkink.
How much effect will the current norm have, how robust is it and what is its normative fit?Though the prescriptions about what can be done to implement R2Prev become more specific and concrete, the question on the locus of responsibility for different aspects of assistance is yet unsettled.
The norm has so far proven durable and experienced strengthened concordance.R2Prev is being used in many diplomatic discussions, in the SG's reports, the GA debates and it played a role in a few cases, such as in Kenya in 2008.The reaction pillar has been applied in Libya and Ivory Coast.An unfortunate by product of those UNSC applications of the norm is that while they might enhance concordance around the norm, the flexible interpretation of the 1973 UNSC resolution by NATO made Russia and China more hostile to the R2P principle and maybe also less willing to implement the consensus existing on R2Prev.The test came in July 2011 when in the GA debate both Russia and China proved supportive of the existing consensus on R2Prev. 187This shows that R2Prev is durable and is resilient to challenges to other pillars of the R2P-norm.
Different states consider R2Prev within their different normative fits.It is possible that states will only effectively implement the measures that fit within their normative framework.For a successful implementation of R2Prev, continuing discussion on the right course of prevention is needed to strengthen the consensus and come to a widely supported implementation.However, this implementation process needs to be supported by sufficient funding, which is currently still lacking at the global, regional and national levels.R2Prev can only be effective if it is not only based on shared understandings, not only robust, but also sufficiently funded.
Different aspects of R2Prev find different degrees of support in the sources of in international law.On R2Prev as an interactional legal norm, continuing interpretation and application over the long term could lead to the improvements needed for an interactional legal norm: increased clarity on what needs to be done by whom in certain circumstances.
Further research on R2Prev should apply the same theoretical frameworks in a few years' time and see how the norm R2Prev developed and what its legal status is.Furthermore, it is interesting to see whether studies will be done on the internalisation of the norm in specific states and (regional) organisations.In this vein, one could study whether normative development on R2Prev has led to more funding or a reallocation of funding within states, regional organisations and the UN.This could also be compared to other norms that require a reallocation of funding for their implementation: to what extent and under what circumstances did this happen?Finally, it is worthwhile to keep studying how the norm R2Prev has functioned in different cases.
How much normative development can one infer from the Outcome Document?It is claimed that the balance of expected global efforts leans more heavily towards 95 High-Level Plenary Meeting of the 60th Session of the General Assembly, 2005 World Summit Outcome, A/RES/60/1.
Bellamy 2008, supra note 8, p. 136; Carnegie Commission on Preventing Deadly Conflict, Final Report with Executive Summary, Washington DC: Carnegie Corporation of New York 1997.