Relational R2P? Civilian-Led Prevention and Protection against Atrocity Crimes

In: Global Responsibility to Protect
Felicity Gray PhD Scholar, School of Regulation and Global Governance, Australian National University, Canberra, ACT, Australia,

Search for other papers by Felicity Gray in
Current site
Google Scholar
Open Access


The protection of civilians in the context of the Responsibility to Protect (R2P) is often understood as a practice of ‘saving strangers’. In this article, I argue that such an understanding overlooks the ways that close relational webs can be used as a form of prevention of, and protection from, atrocity crimes. Through an examination of unarmed protection practices in the context of the civil war in South Sudan, I highlight how practices of unarmed, civilian-led protection – performed both by non-governmental organisations (ngo s), and by communities themselves – hinge on the creation and sustaining of close relational webs. Rather than ‘saving strangers’, this ‘relational R2P’ hinges on familiarity as a means of protecting one’s friends, families, and neighbours. Through a lens of relational connectivity, fresh perspectives and opportunities for re-imagining ‘intervention’ in the face of atrocity crimes arise. The article explores potential opportunities, challenges, and limitations for the implementation of unarmed, civilian-based approaches in the context of atrocity crimes.

1 Introduction

In July 2016, 27,000 residents of the United Nations (UN) Protection of Civilians (PoC) internal displacement site in Juba, South Sudan, were subjected to three days of violence perpetrated by government forces.* In the official UN report on the crisis, the fighting around the site is described as ‘unrestrained violence’ in which ‘participating fighters left a trail of destruction and suffering in their wake’.1 This attack, and many others like it, demonstrate the endemic nature of atrocity crimes in this civil war, as documented by the UN Commission for Human Rights and others.2

The Cammaert report, quoted above, was commissioned after UN peacekeepers – those explicitly mandated to ‘save strangers’ – failed to intervene to protect civilians. Residents recounted how peacekeepers watched from behind locked gates as women were raped outside the UN compound.3 ‘Some women were taken to imc [the on-site medical clinic], they cannot walk,’ said one man. ‘unmiss [United Nations Mission in South Sudan] was just watching’.4 While peacekeepers waited behind the gate, civilians – the direct targets of the attack – attempted to protect their families, friends, and neighbours. Some fled, following routes they believed to be safest, to escape the immediate violence. Some lay down alongside their families in the drainage trenches that line the PoC site, below the trajectories of bullets. National staff of unarmed civilian protection organisations staffed the pedestrian gates, which had been abandoned by the private security personnel tasked with this role. Civilian-led, community-based protection mechanisms also mobilised, directly communicating information and strategy among community members.5

These civilian-led responses to violence – and the withdrawal of armed military response behind closed gates – raise questions about the ‘saving strangers’ narratives and assumptions that have characterised efforts to implement the protection of civilians norm, particularly in the context of Responsibility to Protect (R2P). While the most controversial aspect, the potential and actual use of military intervention remains a core pillar of the R2P doctrine.6 Advocates of this vision of R2P argue that this approach to protection is ‘the only show in town for those serious about preventing future Kosovos and future Rwandas’.7 Critics of this position have argued that the elevation of third-party military intervention erases the agency and self-protection work of those directly experiencing violence.8

In this article, I build on these critiques of R2P as a practice of ‘saving strangers’. Building on critiques advanced by Mégret, Brigg, Luck and Luck, and others, I explore how R2P can be mobilised in practice through civilian-led ‘relational R2P’. I use the example of unarmed civilian protection, a form of civilian-led protection that uses nonviolent strategies to protect civilians and prevent further violence, to demonstrate how relationships can and are used as a form of civilian protection. Unarmed civilian protection practices, performed both by international non-governmental organisations (ingo s) and communities themselves, hinge on the creation and sustaining of close relational webs. Rather than ‘saving strangers’, familiarity can be used as a means of protecting one’s friends, family, and neighbours. This article explains how this is already occurring in South Sudan and explores the possibilities, challenges, and limitations presented by such an approach to atrocity prevention.

This analysis stems from over 18 months of multi-sited ethnographic fieldwork on civilian protection carried out between February 2018 and December 2019, including in contexts like South Sudan and Myanmar that have been at the centre of contemporary discussions on the risk of atrocity crimes. I draw on over 140 open-ended interviews with community members and practitioners engaged in different forms of civilian protection. Interviews and participant observation were conducted in Europe (Germany, the Netherlands, Poland), the United States, Australia, Myanmar, Lebanon, and South Sudan. Interviews are anonymised at request of the interviewee, or for security reasons. In addition to three months of research fieldwork in South Sudan in 2019, I draw on understanding gleaned from a year (2020–2021) based in Bentiu, South Sudan, working as a protection practitioner with an unarmed civilian protection ingo.

This article first canvasses existing critiques of R2P narratives and practice, highlighting the ways that ‘saving strangers’ rhetoric is a flawed narrative that results in potentially damaging practices. I canvas how key scholars have sought to critique this perspective and articulate alternative approaches that elevate civilian agency. I use this foundational work to advance an argument for a relational R2P, highlighting how the protection and prevention of atrocity crimes can occur through close relations and familiarity. I then provide key practical examples from South Sudan that demonstrate how this is already occurring in the context of the civil war, focusing on three key strategy areas: community-led protection and sexual violence, mediation of intercommunal conflict, and direct protection from violence. Finally, I conclude by discussing the role of international actors in such an approach, and caution against romanticisation of local protection mechanisms. The article contributes a timely overview of R2P literature as it relates to civilian agency and uses this to assess the potential role for ‘relational R2P’ approaches in future conflicts.

2 ‘Saving Strangers’

In the two decades since the landmark International Commission on Intervention and State Sovereignty (iciss) report first conceptualised R2P,9 an understanding of R2P as a practice focused on strangers – far away, unknowable, would-be victims of mass atrocities – has persisted. Nicholas Wheeler’s foundational text on humanitarian military intervention, Saving Strangers, frames R2P as a necessary solution to ‘the dilemma of what to do about strangers who are subjected to appalling cruelty by their governments’.10 This is familiar R2P rhetoric, reflective of a broader ‘salvation paradigm’11 in humanitarianism, and rooted in racist, colonial logics12 that encourage an undervaluing of civilian agency.13

The idea of ‘saving strangers’ hinges on an invocation of difference. This difference is ‘a distance marked by self and other’ which then ‘structures relations between humanitarian actor and humanitarian subject in wider humanitarian practice’.14 An extension of colonial logics and praxis, this envisions international interveners as active and powerful agents working to ‘save’ the stranger, who is figured as vulnerable and passive.15 Sara Ahmed’s definition of the stranger as ‘a fetish, acquiring a life of its own by being cut off from histories of determination’16 provides a useful definition here. In the context of R2P, the particularities of those under threat – as well as the complex character of those that threaten – are erased, and entire communities are objectified instead by their potential and actual victimhood in the face of atrocity crimes. In addition to further embedding harmful, racist assumptions about conflict-affected communities in the Global South, this deepens the pervasive assumption that those facing violent threats have no role to play in prevention or protection in the face of atrocity crimes.17

Such concerns have not only been voiced by critics of R2P. Even the most enthusiastic advocates concede that a failure to engage with the role of those directly affected by conflict has undermined R2P.18 Former UN Special Adviser for R2P, the late Edward Luck reflected that ‘the 2005 conception of R2P left no room for agency by those threatened by potential atrocity crimes. Vulnerable populations were treated as objects, not actors. Their fate was to wait for governments and intergovernmental institutions to act on their behalf’.19 In so doing, R2P has played a role in reproducing harmful iniquities of global politics, reinforcing narratives told from the perspective of a powerful subject – an ‘us’ constituted by states of the Global North – preoccupied with ‘saving’ ‘them’ – the objectified and passive strangers of the Global South.20

These narratives cause harm in practice, particularly in settings where third-party interveners interface with civilians.21 In South Sudan, a sense of estrangement reflective of the ‘saving strangers’ approach was evident among both UN peacekeepers and civilians. In this context, it was clear that many UN peacekeepers mandated with the protection of civilians neither see themselves, nor are seen by the community, as ‘protectors’. ‘I didn’t come here to die’ is a common refrain from deployed military personnel in the country. Though this is by no means characteristic of all personnel that work with the mission, it remains a concern because of the way such attitudes seed mistrust among the civilians whom they are committed to in fulfilling their protection of civilians mandate. In Bentiu PoC site, residents reflected on Ghanaian troops stationed there dismissively, lamenting their perceived lack of willingness to protect civilians because of their lack of embeddedness in the community: ‘The ghanbatt s [Ghana Battalion], when they got their independence, they got it peacefully, so now they won’t risk their lives or injuries here in a country like South Sudan. So they just prefer to stay away from the issues.’22

In the view of this individual, it is not possible for the Ghanaian personnel to truly be protective, because their preference is to ‘stay away’ from the circumstances at hand; they do not have social embeddedness or intimacy with the community required to provide that care. They are, in other words, strangers – unknown and disconnected from those they are mandated to protect. Similar distrust was levelled at peacekeepers who failed to intervene in the attack on Juba PoC site in 2016.23 This does not mean that third-party peacekeepers cannot or will not protect in general – there are many examples where ‘strangers’ to a context have intervened to protect and risked their lives in the process. But it does suggest that conventional practice does not engender trust from civilians, and contributes to an overarching structural dynamic that ultimately risks damaging the capacity of missions to accomplish protection objectives.24

In the context of conventional R2P missions, seeing civilians as strangers to be saved also re-embeds and reinforces disempowering, colonial and racist dynamics between civilians and interveners. The power dynamic that is produced on a global level through the ‘saving strangers’ narrative – in which a powerful North ‘saves’ a passive South – also has the potential to be reproduced at ground level in violent interactions – both structural and direct – between international interveners and civilians. In an act framed as rescue, there can be an expectation of a ‘debt of gratitude’, and an attendant foreclosure of complaint, on the part of the person ‘saved’.25 This power inequality underscores longstanding concerns about the potential for exploitation and abuse by members of intervention missions.26 In response to such concerns, there has been a recent and influential trend in the peacekeeping and peacebuilding literature advocating for relationship building between international and local actors.27 Though it is tempting to suggest that the solution to the failings of ‘saving strangers’ modalities of protection is to strengthen connections between communities themselves and third-party interveners, this still does little to disrupt the militarised, racialised, and gendered hierarchies inherent in conventional protection architectures.28

A second implication of the ‘saving strangers’ narrative is to consider the ongoing conceptualisation of R2P as a state-focused doctrine in which protection is understood as military intervention by the ‘international community’. This view forecloses understanding and progress towards protection from mass atrocities from non-state – and specifically, civilian-led – perspectives. Studies and discourse around R2P continue to focus on the role of the state, and specifically on military intervention.29 For example, in a recent special issue of the primary journal dedicated to studies of R2P on genocide risk in Myanmar,30 the edition was framed around the role of the ‘international community’, with featured articles considering the role of regional blocs like asean, individual states including Australia, Norway, the United Kingdom and Indonesia, and state-based institutions such as the International Criminal Court (icc). Obviously, such analyses remain essential to understanding the international political landscape of mass atrocity prevention, and the authors of these pieces advance this scholarship admirably. However, when explorations of R2P are concentrated on one unit of analysis – the state – we lose opportunities to deepen both understanding of and capacity to respond to mass atrocity risks.

3 Re-orienting R2P: Suggested Pathways in the Existing Literature

In response to the outlined concerns, some authors have sought to broaden or reorient how R2P is conceived. For some, a potential response has been to explore the role of individuals. Luck and Luck, as well as Hindawi, call for an ‘Individual Responsibility to Protect’ (IR2P), to be integrated alongside existing national and international approaches.31 This analysis emphasises the role of individuals in responses to atrocity crimes, and harnessing individual actors to further legitimise the larger R2P project32 Luck and Luck argue that R2P ‘will only succeed to the extent that it persuades individuals, as well as governments and institutions, that they have choices and responsibilities’33 in the face of actual and potential atrocity crimes. Their argument focuses on the potential of individuals to further strengthen ‘the larger enterprise’34 of R2P as it currently exists.

Others have argued for a more wholesale recasting of R2P doctrine that focuses on the role of local actors, including individuals but also civil resistance movements, local civil society groups, and perhaps more controversially, non-state armed groups.35 These critiques dovetail with the growing literature and evidence on civilian self-protection which demonstrates the role of conflict-affected communities in their own protection.36 For example, Mégret is more circumspect than Luck and Luck and Hindawi when it comes to the current R2P project, and argues for a reorientation that privileges local ‘resistance’ actors broadly defined:

the argument is call to start from ‘what is here’, the huge, powerful force of human resilience in the face of atrocities, instead of starting from ‘what might be’ – some hypothetical vision of a cosmopolitan community of mankind dictating its will to states.37

Brigg follows a similar line of argument, calling for would-be interveners to ‘hold back’38 to enable space for more organic and grassroots agents to contribute to prevention and protection responses. Unlike critiques that emphasise individual and other groups as primary units of analysis, however, Brigg begins to explore what it might be like to consider the role of affect and relationships in R2P. He calls for ‘reconfiguring and extending the intersubjective moral impulse underpinning R2P’,39 noting the ways in which existing ‘common-sense foundations of R2P risk participating in a problematic moral economy of symbolic exchange between would-be rescuers and victims’.40 It is the role of intersubjectivity – of the ways we are co-constituted by one another, and how these interconnections can be used as a strategy to advance R2P objectives – to which I now turn.

4 Relationality and the Responsibility to Protect

A relational understanding – one that foregrounds the ways that people and institutions are shaped by their social relationships and environments in a dynamic, co-created process – 41 enables us to understand protection as a product of these relations. This is an approach that, following many Indigenous philosophical frameworks, turns away from a focus on individuated entities toward interconnectedness of all beings, human and otherwise.42 If we start from this premise and privilege understanding the ways that selfhood and actions are contingent on relations, we are prompted to ask different questions, and to look for different explanations for atrocities and how they unfold: what protective power, we can ask, emerges from this interconnectedness? What kinds of relations are conducive to community safety, and atrocity prevention? On this basis, we can explore potential for a ‘relational R2P’: how can and are relations used to prevent atrocity crimes and protect civilians? What can be done to encourage these kinds of protective relational webs? In utilising a relational lens we can consider ‘protection’ not as a ‘thing’ that is embodied in a particular actor or outcome, but rather as a dynamic formation of relationships. Such an approach has natural synergies with conceptual approaches such as assemblage theory in which discrete bodies, actors, and concepts are de-privileged in favour of understanding liminal events, the coming together of different relations at different times.43

Though relational approaches have not been commonly noted within the R2P literature, a strong foundation of work exists in research on peace, conflict, and peacebuilding.44 As Hunt, de Leon, Brigg, and others have argued, such analyses move beyond a fixed identitarian analytic, and enable understanding of dynamic, complex relationships in peace and conflict settings.45 One example of bringing a relational lens to R2P is Brigg’s analysis noted above, in which he notes the importance of ‘a more expansive conception of politics than that usually entertained in R2P advocacy, in order to recognise diverse political orders, including their support for prevention and protection’ – to move beyond the state, or any individual institution, to ‘politics as grounded in interactions among people’.46

What implications does a relational understanding of protection have for R2P in practice? By understanding the ways that different actors – and more broadly, protection institutions and practices – are co-constituted by relations, we can better understand how relations shape both perpetration of atrocities and protection practice. In this article, I focus on relational webs endemic to conflict settings, and how these can be used for protection against and prevention of atrocity crimes. In so doing, it becomes possible to better appreciate variations, tensions, and challenges within particular atrocity-risk contexts, and to begin to identify opportunities for R2P that go beyond predetermined actors and strategies. Centring relations and agency of conflict- affected populations in the face of mass atrocity crimes has several potential benefits, including sharper contextual understanding of needs and desires of those most impacted and the potential to be more efficient in responding to on-the-ground needs.47

This does not mean that all R2P eggs should be in one proverbial basket, nor that relational approaches are without their own challenges. It is important to flag upfront that relational webs are not inherently protective or preventative. Just as easily as they can be used to prevent and protect, they can be used to prey and perpetrate. In Lee-Ann Fujii’s research on the Rwandan genocide, she demonstrates in incisive detail how and why people sometimes chose to protect, but also brutalise one another: ‘despite long-standing, mostly amicable, relations with their victims … killers went about their task in determined fashion. They trapped victims at road-blocks, lured them to public buildings, and descended on their homes and hiding places’.48 Concentrating on relationships is therefore not a call to romanticise them, but to be aware of their empirical significance, and how they can be both used and abused in the context of prevention and protection against atrocity crimes. Though I concentrate on the potential of relational webs as a form of prevention and protection in this article, others have demonstrated the ways social connections can act to facilitate violence. Understanding why and how different outcomes are enabled, and the nature of these encounters, is a critical area of continuing research. The remainder of this article builds on the theoretical foundations provided by Mégret and others by exploring empirically how relations may be used for protective outcomes through the practice of unarmed civilian protection. Through exploration of a practical example of ‘relational R2P’ – unarmed civilian protection in South Sudan – we can better understand the strengths and challenges associated with a relational approach to protection.

5 Relational R2P in Practice: Unarmed Civilian Protection in South Sudan

For much of South Sudan’s contemporary history, civilians have lived with the threat of atrocity crimes. Throughout periods of slavery and colonisation, the war with Northern Sudan, before independence in 2011, and during the civil war, violence against civilians has been an ongoing concern.49 Since the civil war began in 2013, sparked by a political dispute between President Salva Kiir and his deputy, Vice President Riek Machar, civilians have faced widespread violence. Though there is clear evidence and consensus that atrocity crimes have been committed since 2013,50 whether actions have constituted ‘genocide’ has been a topic of debate. As Pinaud notes, international actors have withheld use of the term, despite compelling evidence to the contrary.51 Overall it is estimated that almost 400,000 people have died due to the conflict52 and many more displaced, injured and harmed. Despite a political agreement reached in late 2018, followed by the formation of a transitional government in February 2020, there remains a high risk of violence. Many remain concerned about ongoing risk of atrocity crimes.53

Amid this violence, civilians have acted in a range of ways to protect themselves that are contingent on their relational networks.54 To run away, to find safe haven, to be trafficked to safety, or to organise new forms of community governance – all of these acts are a form of protection shaped by embeddedness in social networks. As noted in the introduction, this has often occurred in the face of serious protection gaps, including failures to protect by external actors, including UN peacekeepers. This civilian self-protection hinges on connections between civilians themselves, their embeddedness in local context, and ability to use this network and knowledge to navigate threats of violence. Some civilians have done so through a framework of unarmed civilian protection, a methodology that has parallels with civilian self-protection. Civilian self-protection is a broader concept, however, that can encompass actions including armed self-defence and resistance.55 Though such actions undoubtedly also utilise relationships in some form, they are also utilise force.

In contrast, unarmed civilian protection refers to a practice that emphasises nonviolent, civilian-led strategies for protection and prevention. It can still be ad hoc or organised, draw on a range of strategies, and be carried out and influenced by a range of actors. Fundamentally, however, the practice works by leveraging one’s relations as a form of protection – a form of relational R2P. The term ‘unarmed civilian protection’ itself has been most commonly used in relation to ingo-aligned forms of nonviolent protection, encompassing strategies like protective accompaniment or presence carried out by organisations such as Nonviolent Peaceforce or Peace Brigades International.56 This usage is evident in the few mentions of unarmed civilian protection in R2P literature, where it is used to refer to third-party ingo deployments of unarmed civilians, particularly in situations of pre-and post-violence.57 However, this narrow use of the term overlooks unarmed civilian protection practice as it is carried out with and/or by communities themselves, often in highly sophisticated, organised, and effective forms, and often amid ongoing violence and atrocities. As such, I use the term unarmed civilian protection to refer to practices of civilian-led nonviolent protection that encompass a broader history and community of practice. This can include ingo s dedicated to unarmed civilian protection but is not contingent on their presence.

In South Sudan, unarmed civilian protection is being used to prevent escalations of violence, to protect civilians amid ongoing violence, and to rebuild and repair in its aftermath. In this section, I canvass some of the practical ways that civilians are contributing to the prevention and protection against atrocity crimes through the relational strategies that underpin unarmed civilian protection. These examples are not blanket solutions to the threat of atrocity crimes. They may not always be appropriate to every context or potential threat, and even when they are, they may not always work to prevent violence or protect civilians. But in many cases, relational strategies have meant that threats have been reduced, and lives saved. Understanding what ‘relational R2P’ may look like in practice is a first step in understanding how these kinds of approaches can be better recognised and supported in prevention and protection against atrocity crimes.

5.1 Example 1: Community-Led Protection and Sexual Violence

One of the widespread atrocity crimes that has characterised the conflict in South Sudan has been rape and sexual violence.58 In mid-2014, women in Bentiu PoC site raised concerns about these risks when leaving the camp to collect essential resources like firewood. With over 200,000 people living in Bentiu PoC site at the time, resources close to the camp perimeter were quickly depleted, and women were spending significant time – five, six, eight hours – walking to areas where firewood was still available. While outside, women were at grave risk of attacks by armed actors. They were harassed, forced to pay a ‘tax’ to pass, assaulted, raped, abducted, and killed. ‘We go to the bush to collect firewood, and lots of women disappear,’59 explained one woman in Bentiu PoC.

In the face of the risk of assault by armed actors, women in the camp mobilised into Women’s Protection Teams (wpt), community-based protection groups comprised of local women that use unarmed civilian protection strategies. One wpt member I interviewed explained how they started standing at the gate to count the women going out, and then count them back as they returned, so they would know if someone was missing, and if so, raise an alarm.60 They began to organise together to travel in groups to collect firewood, in an attempt to deter would-be attackers. Through their network, they worked together to identify the most at-risk routes and advocated within community forums and church gatherings for women to utilise alternatives. With embedded local knowledge, these risk assessments were highly specific and dynamic, accounting for factors often overlooked by other protection actors who tended to patrol the same areas without responding to changes in context. With embedded relational connections, the way this knowledge could be shared was more efficient, dynamic, and contextual. Working alongside ingo Nonviolent Peaceforce, the wpt s successfully lobbied unmiss to commence armed foot patrols of key risk areas they had identified, leveraging available protection resources. The women also reached out to local authorities, building relationships with commanders and other troops, and using interpersonal and familial connections to discourage perpetration of crimes. In other words, they sought to strengthen and use their relational webs, and relationship with the broader context, to prevent violence and protect women moving outside of the site. Women reported a significant decrease in rates of assault in the aftermath of these mobilisations.61

This is not a panacea. In a now infamous report, in November 2018 medical ingo Médecins Sans Frontières (msf) released data citing that, in one week alone, 125 survivors had been treated for the impacts of rape, sexual, and physical abuse in their clinics in Rubkona County, Unity State.62 This number was indicative of a widespread increase in rape and sexual assault of women around Rubkona and the Bentiu PoC site. Though community members and protection ingo s present in the area had been implementing unarmed civilian protection strategies, the reality is that community-based mechanisms operate with next to no resources, with those leading efforts often living through precarious personal situations of displacement and poverty. This is not to say that their work was unsuccessful – evidence suggests the opposite. However, without legitimacy and resourcing, the extent to which these efforts can be sustained and scaled up is limited.

A common response to these efforts from international protection policymakers is to question why they are necessary, particularly given the presence of an international peacekeeping mission:

I don’t really understand why it makes such a big difference whether the accompaniment is armed or unarmed. I think either way, it’s trying to accomplish the same objective which is deterring actors from using violence against civilians by the presence of these international peacekeepers. I think whether or not they have weapons is a little bit immaterial.63

This kind of response overlooks the ways that unarmed civilian protection is practised by local community networks, not just international actors, and therefore also overlooks the importance of local relational networks to the efficacy of these efforts. In addition, if we conceive of R2P in a more holistic sense that imagines life in South Sudan beyond the presence of an international peacekeeping mission, civilian leadership and involvement in protection and violence prevention is essential to fostering sustainable peace and should be prioritised. As one practitioner noted in Myanmar:

you can note down all the protection issues of women, under 132564 and under this code, and under this code, and under this code, and under this code. When it comes to [unarmed civilian protection] you not only note what they are, but you also ask the civilians: is there anything they can do within their own structures that can reduce, that can prevent, that can mitigate, that can respond, to those violations?65

5.2 Example 2: Intercommunal Violence Mediation

Of concern throughout the civil war and currently are the risks associated with intercommunal violence, a major factor in ongoing threats of atrocity crimes.66 This is an area widely recognised as key to building sustainable peace in South Sudan, and something in which unmiss has also invested significant focus from a civil affairs perspective.67 Though international actors can have a role to play in these processes, there are key examples of mediations pursued and successfully facilitated by unarmed civilian protection organisations that hinge on the role of national staff and/or community-run protection teams and their positionality as locally embedded actors. For example, a South Sudanese staff member working with Nonviolent Peaceforce in the Mundri area of Western Equatoria State recalled a major intervention that resulted in successful cessation of hostilities after Mundari cattle-keepers – the movements of whom are a central source of intercommunal violence in South Sudan – crossed over the state border and raided cattle:

they raided cattle from our [Moru] communities, those in the border areas who are also keeping cattle due to intermarriage. They came from Central Equatoria State. They also have some people from Lakes State. They raided 370 cattle. They fought with some youth who tried to resist them. Consequently the youth used bow and arrows, one of them was shot. Two were arrested and put in prison.68

This is the kind of violence that can easily escalate into mass intercommunal violence, revenge killings, and ultimately atrocity crimes, something that has been observed time and time again in South Sudan.69 To prevent this kind of escalation requires speed in response, which is contingent on pre-existing relationships, and understanding how to leverage those relationships in the complex social and political context.70 Unarmed civilian protection focuses on equipping people who are already situated in these relational webs to enact prevention and protection solutions efficiently, and with context sensitivity. In this example, the protection officers – born and raised in Mundri, and relationally embedded in the local community – were immediately alerted to the issue by local youth, who came to them directly for assistance.

In response, the staff mobilised and began to engage local leaders, starting with the Commissioner:

We met with the Commissioner and he said ‘this one, you know what, no solution. Because the cattlekeepers have already stabbed our communities first, and the youth have vowed they are going to fight. They are going to show the other community that we are also fighters, and they are going to restore the cattle.’

Following engagement with the Commissioner, the team engaged with paramount chiefs, both government and opposition authorities, and community elders. The overall response was again, scepticism: ‘for them it sounds very contrary to what they believe and what they know that works best for them. They ask “how?”’ When the team suggested a formal letter communication requesting dialogue with the cattle keepers, the community leaders were again dismissive: ‘This is not good! How can we write a letter when they are the ones that violated, the ones that took things from us! How can we go and take them for dialogue?’. Knowing who to contact, in what order, and in what way to coordinate these efforts is highly sensitive work, and only made possible through relationships local staff have within their own communities.

Ultimately, the practitioners convinced community leaders to write the letter, and so began an unusual series of events. The Mundari chief was surprised to receive the letter, knowing that Moru youth had already mobilised and were determined to fight to restore the cattle that had been stolen. Scepticism about goodwill persisted, and dialogue continued to be seen as an unlikely solution. Despite this, over the course of a two-day engagement, sharing meals and discussion with both parties to the conflict, beset by difficult logistics and politics, the protection team successfully held space for dialogue. Discussion went well beyond the cattle-raiding incident that sparked the dialogue, delving into drivers of the raiding and violence such as economic security and access to markets, inter-state border politics, and interclan marriage. Ultimately, the communities agreed on procedures to resolve these underlying challenges and drivers of conflict, as well as returning the stolen cattle.71

This might sound like a small task and victory, but it is not. Successful engagement and dialogue require intimate knowledge of communities, familial connections, and underlying political, economic and historical relations. In the words of the national protection officer who led the effort, ‘if you are not keen enough to see what is missing, the plan that you have, it will fade’.72 That ‘keen-ness’, as he termed it, is knowledge borne of being relationally embedded, an intuitive form of diplomacy borne of being in and of the community itself. Without being situated in this way, the historical and relational knowledge and political skills required would not have been available.

This case study demonstrates how relational webs can be utilised as a protection resource. The extent to which someone understands the intricacies of these webs shapes their capacity to de-escalate, interrupt, and respond to violent dynamics as they reverberate through the community. National personnel coordinating local protection groups in other locations such as Myanmar echoed the same: ‘Because I am living in that tribe, I know everyone, and everyone knows us … I use existing relationships, and my father’s relationships,’73 said one female leader from northern Shan State with an impressive record of unarmed civilian protection implementation, such as negotiating release of child soldiers. Relational intimacy supports effective unarmed responses, enabling risk assessment, identification of resolution strategies, and enough trust to be present without the threat of violent enforcement. Amid crisis, it also enables unarmed civilian actors to make these strategic calculations with more speed and specificity than other external actors. ‘A live threat can’t wait on a response from New York,’ explained one South Sudanese protection leader in Juba, critical of the bureaucratic approvals processes that characterise unmiss responses to security incidents. ‘The UN is a kind of government that I cannot describe because they have a very complex role. They don’t take a quick decision. Slowly, very slowly.’74 In contrast, the relational webs of local actors are ‘always already there’,75 ready to be mobilised in pursuit of protection goals.

This is not to say that international actors cannot play a role in some circumstances.76 The presence of international staff from ingo s can prove useful in some circumstances, particularly if they have been present for long enough to cultivate and understand their own relational networks. However, without the relational networks and expertise of those local to a conflict setting, chances for success are limited. Even worse, ill-thought-out or misguided approaches to mediation can have deadly results, with many examples of dialogues devolving into armed violence, including deaths of participants.77 This example demonstrates the invaluable contribution that relational, civilian-led approaches can have on identifying, preventing, and responding to violent conflict.

5.3 Example 3: Direct Protection

The previous two examples speak largely to the value of unarmed civilian protection as it relates to prevention of atrocity crimes. Over time, these kinds of actions shape deeper connections and resilience to social and political conflict within communities, and thus contribute to building longer term, sustainable peace. More controversially, perhaps, unarmed civilian protection has also been used in situations of direct protection from atrocity crimes. Contrary to some misguided understandings, unarmed civilian protection is not a practice that encourages placing civilians in the middle of active armed conflict and using them as human shields (though early iterations of third-party nonviolent intervention were certainly envisioned in this way).78 Rather, by seeking relationships with would-be perpetrators, even in the heat of immediate violence, there are examples in which the presence of unarmed civilians has resulted in positive protection outcomes.

In South Sudan, there are a number of accounts of civilians deterring active violent threats in this manner. In one high profile instance in April 2014, during an attack on Bor PoC site by gunmen, two international protection officers directly prevented the deaths of civilians. A UN report on the attack recounts that ‘attackers targeted victims from close range. Attackers went from tent to tent, pulling idp s [Internally Displaced Persons] out … Upon entering the tents, they would harass, threaten, and beat the occupants … those who resisted or refused were shot by other men waiting outside with guns’.79 Witnesses asserted that the clear aim was to target Nuer civilians: ‘the attackers were targeting idp s based on their Nuer facial markings … If idp s had no markings, the attackers asked them in the Dinka language which tribe they were from, and if the idp s could not respond in Dinka, they were killed or beaten’.80

Two international protection officers employed by Nonviolent Peaceforce were present in the site when the attack started and moved to a shelter with a group of civilians – five women and nine children. Each protection officer stood at a one of the two entrances to the shelter, with the civilians inside.81 As the attack wore on, gunmen armed with AK47s passed and threatened them on three separate occasions. As the gunmen approached, they stood their ground, making it clear they were unarmed, and seeking to connect with the aggressors:

I think if we had had a gun in that situation we would have been shot immediately, because we’re a threat. Because we were unarmed, I think it changes. It’s still there, but it can be managed in a different way other than violence, and it opens the door to look for solutions. If we were armed peacekeepers, the solution is you shoot back. And you do force protection. Because we are not armed, then we can find other ways. It worked in this context because we did have the opportunity to engage in some sort of dialogue with the aggressors. And we did have something to say that did change their minds.82

The unarmed civilian protection actions taken by the protection officers were successful, and both they and the 14 civilians inside the tukul were protected.

In this case, the fact that both civilians present were of foreign appearance cannot be overlooked. Indeed, the majority of literature on third-party nonviolent intervention (a historical precursor to unarmed civilian protection) focuses on the use of whiteness, and the greater power afforded to white bodies, as a form of protection. The opportunity to build relationship in the moment arises, in part, because of the pause this unexpected foreign presence gave the aggressors. Per Mahrouse, ‘it is precisely the ways in which certain bodies are classified within a global racial hierarchy that make these interventions effective’.83 Certainly, in the example described above, the leveraging of race in this manner (and the way this risks reifying racial iniquity and structural violence) is a concern.84

At the same time, similar examples of direct protection that do not rely on foreign presence in this way were recounted throughout my research, both in South Sudan and elsewhere. In Bentiu, local practitioners negotiated with armed gunmen for the release of both themselves and other civilians during an attempted abduction in 2015 outside of Bentiu PoC site.85 Another example was given by a local protection leader in northern Shan State in Myanmar, in which she used pre-existing relationships with government soldiers to access and negotiate a short ceasefire to enable the escape of civilians trapped in between two fast advancing frontlines.86 In the United States, where similar work is underway in response to gun violence, direct protection and de-escalation is performed by individuals within their own neighbourhoods.87 In the words of one organiser in Brooklyn, New York, the success of these interventions hinges on intervening actors who ‘have relationships. People that know somebody that know somebody.’88

The circumstances described here are not ideal. Prevention of violence in the first instance should be the priority. The risks faced by civilians who protect themselves and others in the face of potential atrocity crimes, whether they are trained to do so or not, can be significant. There are examples of civilians intervening in similar ways on an ad hoc basis (as opposed to those who identify within a more institutionalised unarmed civilian protection framework), with mixed outcomes – in some cases, including death.89 Like armed intervention, ‘relational R2P’ is not without risk to those who intervene. With that said, the individuals that dealt with the circumstances described above – both international and national actors – were clear in their skill sets and trained to use relationships as a form of protection.

Relational protection skills are not generally emphasised in R2P as it is conventionally conceived. In considering current and future interventions, what kind of role does a relational approach have to play? What implications does this have for existing international approaches to R2P? A relational reorientation is not about abandoning international involvement or existing atrocity prevention tools we know have a positive and material impact. The UN Security Council arms embargo on South Sudan, which prohibits the supply, sale, and transfer of weapons to the country,90 is a good example of an international protection response that contributes to improving atrocity risk conditions on the ground. Targeted individual financial and travel sanctions on those known to be involved in atrocity crimes is another important measure, and in the case of South Sudan should be expanded, as per advocacy by the UN Special Adviser on the Prevention of Genocide.91 High-level political advocacy, such as leveraging the influence of the international community to encourage accountability and leadership from the South Sudanese government, is another example of how international actors may contribute to protection outcomes. However, South Sudan is also a cautionary tale when it comes to the involvement of international actors. Some argue that rather than reducing violence, the long-term investment and attention from international actors in relation to South Sudan has ‘contributed to shape the ideology of ethnic supremacy’92 that underpins mass atrocities and genocide. As Kydd and Straus have argued, these risks underscore the importance of protection approaches that are multidimensional and highly sensitive to context and power dynamics.93 In considering this multidimensional approach, it is imperative that civilian-led, relational R2P strategies – like unarmed civilian protection – are part of the conversation.

Caution should also be exercised when exploring new avenues of support for civilian-led mechanisms. As scholars of peacebuilding and civil society have argued, and as anyone who has worked in community politics has experienced, ‘the local’ is also a space of contentious politics.94 The impetus for atrocity crimes often stems from the interests, actions, and ideologies of local actors, and such actors cannot be assumed to be intrinsically good.95 Relational R2P strategies do not circumvent these politics, but they do seek to account for and understand them in a way that is contextually sensitive and locally responsive. The more these are understood and navigated, rather than overlooked and avoided, the better protection interventions can be tailored to respond to these dynamics.

6 Conclusion: Reshaping the R2P Toolkit

This article does not intend to suggest that atrocity crimes are easily prevented or halted, through any means – unarmed or armed, civilian or military.96 Humility is essential when considering any form of response, given the complexities, ambiguities, and high stakes of mass atrocities.97 However, scepticism and unwillingness to consider anything other than international military intervention in the face of mass atrocities neglects, delegitimises, and forecloses serious considerations of alternatives, including the role of civilian-led, relational approaches. This does not mean we abandon rule- and state-based initiatives against atrocity crimes. As Baines and Paddon reflect, ‘civilian innovation to self-protect is not the solution, but it is one that exists within the complex web of violence and survival’.98 However, so long as narrow understandings of R2P as ‘saving strangers’ by force of arms persist, the framework will continue to reinforce racist, colonial logics of intervention that undermine civilian agency and the role of non-military strategies for averting atrocity crimes.

Examples of unarmed, relational protection like those outlined in this article, and the way people understand this protection work at a community level, deserve to be taken seriously. In the same way that scholars scour Security Council debates, interview diplomats, and review foreign policy documents to analyse state R2P approaches, there must be investment in understanding the work of communities themselves. So far, this has occurred only to a limited extent in mass atrocity-related literature and requires further dedication from researchers and practitioners. This kind of understanding will provide a foundation on which to refashion the prevention and protection toolkit, one that accounts for the experiences of those most affected by, and most often first responders to, atrocity crimes – communities themselves.

Targeted support for ngo s and community mobilised groups practising unarmed civilian protection has established the value of their work, and the utility of a civilian-led, relational approach to R2P. The examples given above demonstrate that there is value in these protection practices that deserve further recognition, research, legitimacy, and support. This reorientation is not about adding more practices to the R2P toolkit, but considering the boundaries and assumptions of the toolkit itself: who decides what practices are included? Who has access to the tools? Who decides when and how they are used? For too long, discussions around atrocity prevention have utilised the selfsame tools and frameworks that often spur atrocities in the first place. This is a call for redefinition of the proverbial R2P house itself, requiring a reimagining of what it means to intervene – ‘for the master’s tools will never dismantle the master’s house’.99 Rather than a top-down blanket solution, a relational R2P seeks to interrupt violence from within the very relational webs in which it is produced.


I thank the community of scholars at the 2021 IR2P Symposium on Intervention and Responsibility to Protect: Past, Present, & Futures for the opportunity to present and discuss this article in their company. I also thank those mentors, colleagues, and friends who provided comments and critiques on drafts, including John Braithwaite, Charles Hunt, Claire Guinta, and the two anonymous reviewers who were generous and insightful with their feedback. I am grateful to Nonviolent Peaceforce South Sudan for the support provided while undertaking this research. Finally, I acknowledge the many people who contributed their time, expertise, and generosity to this research – most of all those protecting themselves and their communities in South Sudan.


Patrick Cammaert, ‘Executive Summary of the Independent Special Investigation into the Violence in Juba in 2016 and the Response by the United Nations Mission in South Sudan’, United Nations Security Council, S/2016/924, 1 November 2016.


Clémence Pinaud, War and Genocide in South Sudan (Ithaca: Cornell University Press, 2021); UN Human Rights Council, Report of the Commission on Human Rights in South Sudan, a/hrc/43/56, 20 March 2020.


Interview 100 – South Sudanese protection officer, Juba, September 2019; Interview 105 – Community member, Juba, September 2019; Interview 108 – Community member, Juba, September 2019.


Interview 105.


Interview 100.


Charles T. Hunt and Phil Orchard, ‘Introduction: Contestation and Consolidation of the Responsibility to Protect’ in Hunt and Orchard (eds.), Constructing the Responsibility to Protect: Contestation and Consolidation (Abingdon: Routledge, 2020), pp. 1–27.


Alex Bellamy, ‘The Responsibility to Protect after the 2005 World Summit’, Policy Brief No. 1, Carnegie Council on Ethics and International Affairs, 2006, p. 15.


Erin Baines and Emily Paddon, ‘“This Is How We Survived”: Civilian Agency and Humanitarian Protection’, Security Dialogue, 43(3) 231–247 (2012).


Gareth J. Evans and Mohamed Sahnoun, The Responsibility to Protect: Report of the International Commission on Intervention and State Sovereignty (Ottawa: idrc, 2001).


Nicholas Wheeler, Saving Strangers: Humanitarian Intervention in International Society (Oxford: Oxford University Press, 2002), p. 1.


Frédéric Mégret, ‘Beyond the “Salvation” Paradigm: Responsibility to Protect (Others) vs the Power of Protecting Oneself’, Security Dialogue, 40(6) 575–595 (2009); see also on ‘politics of pity’ Michael Barnett, Empire of Humanity: A History of Humanitarianism (Ithaca: Cornell University Press, 2011), pp. 50–51; Polly Pallister-Wilkins, ‘Hotspots and the Geographies of Humanitarianism’, Environment and Planning D: Society and Space, 38(6) 997–998 (2018).


Sherene Razack, Dark Threats and White Knights: The Somalia Affairs, Peacekeeping, and the New Imperialism (Toronto: University of Toronto Press, 2004), p. 11; Alexander Anievas, Nivi Manchanda, and Robbie Shilliam, Race and Racism in International Relations: Confronting the Global Colour Line (New York: Taylor and Francis, 2014).


On R2P and lack of recognition of civilian agency, see Mégret, ‘Beyond the “Salvation” Paradigm, p. 580; Anne Orford, ‘Muscular Humanitarianism: Reading the Narratives of the New Interventionism’, European Journal of International Law, 10(4) 679–711 (1999), p. 695; Morgan Brigg, ‘Humanitarian Symbolic Exchange: Extending Responsibility to Protect Through Individual and Local Engagement’, Third World Quarterly, 39(5) 838–853 (2018), p. 844; Barnett, Empire of Humanity, pp. 34, 229, 235–236.


Pallister-Wilkins, ‘Hotspots and the Geographies of Humanitarianism’, p. 998.


Michael Barnett, ‘International Paternalism and Humanitarian Governance’, Global Constitutionalism, 1(3) 485–521 (2012), pp. 486–487; Pallister-Wilkins, ‘Hotspots and the Geographies of Humanitarianism’, p. 998; Mégret, ‘Beyond the “Salvation” Paradigm’, p. 580; Daniel Levine, ‘Some Considerations for Civilian-Peacekeeper Protection Alliances’, Ethics and Global Politics, 6(1) 1–23 (2013), p. 1; Baines and Paddon, ‘“This Is How We Survived”’, p. 234.


Sara Ahmed, ‘Travelling with Strangers’, Journal of Intercultural Studies, 42(1) 8–23 (2021), p. 8; see also Sara Ahmed, Strange Encounters: Embodied Others in Post-Coloniality (London and New York: Routledge, 2000), p. 5.


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 580.


Rama Mani and Thomas G. Weiss, ‘R2P’s Missing Link, Culture’, Global Responsibility to Protect, 3(4) 451–472 (2011), pp. 456–457.


Edward C. Luck, ‘The Adolescent: R2P at Fifteen’, Global Responsibility to Protect, 12(4) 381–383 (2020), pp. 382–383; see also Edward C. Luck and Dana Zaret Luck, ‘The Individual Responsibility to Protect’ in Sheri Rosenberg, Tibi Galis, and Alex Zucker (eds.), Reconstructing Atrocity Prevention (New York: Cambridge University Press, 2015), pp. 207–248.


Brigg, ‘Humanitarian Symbolic Exchange’, p. 839; Mahmood Mamdani, ‘Responsibility to Protect or Right to Punish?’, Journal of Intervention and Statebuilding, 4(1) 53–67 (2010), p. 59.


Brigg, ‘Humanitarian Symbolic Exchange’, 847; Mégret, ‘Beyond the “Salvation” Paradigm’, pp. 585–586.


Interview 123 – Community member, Bentiu, December 2019.


Interview 100; Interview 108; Interview 123; Interview 111 – Community member, Juba, September 2019.


This has been discussed more broadly by Severine Autesserre, Peaceland: Conflict Resolution and the Everyday Politics of International Intervention (Cambridge: Cambridge University Press, 2014), and in relation to prevention of sexual exploitation and abuse by Jasmine-Kim Westerndorf, Violating Peace: Sex, Aid and Peacekeeping (Ithaca: Cornell University Press, 2020).


Nora J. Kenworthy, ‘Global Health: The Debts of Gratitude’, WSQ: Women’s Studies Quarterly, 42(1 & 2) 69–85 (2014).


Sam R. Bell, Michael E. Flynn, and Carla Martinez Machain, ‘U.N. Peacekeeping Forces and the Demand for Sex Trafficking’, International Studies Quarterly, 62(3) 643–655 (2018); see also Sherene Razack, Dark Threats and White Knights; Westerndorf, Violating Peace.


For example, Autesserre, Peaceland.


I develop this argument further in Felicity Gray, ‘Protection as Connection: Feminist Relational Theory and Protecting Civilians from Violence in South Sudan’, Journal of Global Ethics (2022) (forthcoming).


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 578; Coralie Pison Hindawi, ‘What if R2P Was–Truly–Everyone’s Business? Exploring the Individual Responsibility to Protect’, Alternatives: Global, Local, Political, 41(1) 29–48 (2016), pp. 30–31.


Martin Mennecke and Ellen E. Stensrud (eds.), ‘Special Issue: Myanmar and (the Failure of) Atrocity Prevention’, Global Responsibility to Protect 13(2–3) (2021).


Luck and Luck, ‘The Individual Responsibility to Protect’; Hindawi, ‘What if R2P Was–Truly–Everyone’s Business?’.


Kenneth Waltz, Man, the State, and War (New York: Columbia University Press, 1959).


Luck and Luck, ‘The Individual Responsibility to Protect’, p. 248.


ibid., p. 248.


Mégret, ‘Beyond the “Salvation” Paradigm’, pp. 589–590.


On South Sudan, see Baines and Paddon, ‘“This Is How We Survived’’’; Emily Paddon Rhoads and Rebecca Sutton, ‘The (Self) Protection of Civilians in South Sudan: Popular and Community Justice Practices’, African Affairs, 119(476) 370–394 (2020); On community self-protection strategies more broadly, see Oliver Kaplan, Resisting War: How Communities Protect Themselves (Cambridge: Cambridge University Press, 2017); Jana Krause, Resilient Communities: Non-Violence and Civilian Agency in Communal War (Cambridge: Cambridge University Press, 2018); Betcy Jose and Peace Medie, ‘Understanding Why and How Civilians Resort to Self-protection in Armed Conflict’, International Studies Review, 17(4) 515–535 (2015).


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 592.


Brigg, ‘Humanitarian Symbolic Exchange’, p. 849.


ibid., p. 845.


ibid., p. 849.


Lauren Tynan, ‘What Is Relationality?’, Cultural Geographies, 28(4) 597–610 (2021).


ibid., p. 598. Note: In this article, I focus on relations between people, despite this being only one component of relational approaches. Other elements of relationality in mass atrocity contexts – such as relations with space, time, and material objects – are essential avenues for further research on relational approaches to R2P.


Jasbir Puar, ‘“I Would Rather be a Cyborg than a Goddess” Becoming-Intersectional in Assemblage Theory’, philoSOPHIA, 2(1) 57–59 (2011).


Marysia Zalewski, ‘Forget(ting) Feminism? Investigating Relationality in International Relations’, Cambridge Review of International Affairs, 32 (5) 615–635 (2019); Justin de Leon, ‘Lakota Experiences of (In)security: Cosmology and Ontological Security’, International Feminist Journal of Politics, 22 (1) 33–62 (2020).


Morgan Brigg, ‘Relational and Essential: Theorizing Difference for Peacebuilding’, Journal of Intervention and Statebuilding, 12 (3) 352–366 (2018); Charles T. Hunt, ‘Beyond the Binaries: Towards a Relational Approach to Peacebuilding’, Global Change, Peace and Security, 29(3) 209–227 (2017); de Leon, ‘Lakota Experiences of (In)security’.


Brigg, ‘Humanitarian Symbolic Exchange’, p. 846.


Mégret, ‘Beyond the “Salvation” Paradigm’, pp. 583–585.


Lee Ann Fujii, Killing Neighbours: Webs of Violence in Rwanda (Ithaca: Cornell University Press, 2009), pp. 2–3.


For overviews of violence across these periods, see Peter Adwok Nyaba, South Sudan: Elites, Ethnicity, Endless Wars and the Stunted State (Dar es Salaam: Mkuki na Nyota, 2019); Pinaud, War and Genocide in South Sudan.


UN Human Rights Council, Report of the Commission on Human Rights in South Sudan, a/hrc/37/71, 23 March 2018; African Union Commission of Inquiry on South Sudan, Final Report of the AU Commission of Inquiry on South Sudan, 15 October 2014, available at


Pinaud, War and Genocide in South Sudan, p. 10.


Francesco Checchi, Adrienne Testa, Abdihamid Warsame, Le Quach, and Rachel Burns, Estimates of Crisis-Attributable Mortality in South Sudan, December 2013–April 2018 (London: London School of Hygiene and Tropical Medicine, 2018).


Sarah McIntosh, Immediate Risks of Mass Atrocity in South Sudan, Policy Brief (Washington DC: Simon-Skjodt Center for the Prevention of Genocide, 2022),, accessed 17 February 2022; Global Centre for the Responsibility to Protect, ‘South Sudan’, 31 May 2021,, accessed 10 August 2021; UN Human Rights Council, Report of the Commission on Human Rights in South Sudan, a/hrc/37/71, 23 March 2018; Alice Wairimu Nderitu, ‘Remarks by Ms. Alice Wairimu Nderitu Special Adviser to the Secretary-General on the Prevention of Genocide’, 28 June 2021,, accessed 10 August 2021.


See footnote 35.


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 589.


Rachel Julian, ‘The Transformative Impact of Unarmed Civilian Peacekeeping’, Global Society, 34(1) 99–111 (2020).


Alex Bellamy, The First Response: Peaceful Means in the Third Pillar of the Responsibility to Protect (Muscatine: Stanley Foundation, 2015), pp. 50–53.


Clémence Pinaud, ‘Genocidal Rape in South Sudan: Organization, Function, and Effects’, Human Rights Quarterly, 42(3) 667–694 (2020).


Interview 132 – Community member, Bentiu, November 2019.


Interview 132.


Interview 132; Interview 29 – Unarmed civilian protection practitioner (formerly in South Sudan), Beirut, June 2018; Interview 31 – Unarmed civilian protection practitioner (formerly in South Sudan), Beirut, June 2018; Interview 109 – Community member, Juba, September 2019.


Médecins Sans Frontières, ‘125 Women and Girls Seek Emergency Assistance in Bentiu after Horrific Sexual Violence’, 30 November 2018,, accessed 10 August 2021.


Interview 76 – Senior ngo official on peacekeeping and civilian protection, Washington DC, December 2018.


unsc Res. 1325 [on women and peace and security], 31 October 2000, available at, accessed 20 December 2021.


Interview 80 – Senior unarmed civilian protection practitioner, Yangon, Myanmar, April 2019.


Global Centre for the Responsibility to Protect, ‘Atrocity Alert No. 196: Yemen, South Sudan and Mali’, 25 March 2020,, accessed 10 August 2021. See also Jana Krause, ‘Stabilization and Local Conflicts: Communal and Civil War in South Sudan’, Ethnopolitics, 18(5) 478–493 (2019).


Human Rights Division – United Nations Mission in South Sudan, ‘Quarterly Brief on Violence Affecting Civilians’, April–June 2020,, accessed 10 August 2021; Interview 74; Interview 125; Interview 89 – Senior peacebuilding official, New York City, July 2019; Interview 79 – UN official, New York City, March 2019.


Interview 58 – National protection officer with Nonviolent Peaceforce, teleconference, October 2018. All quotes that follow in this section are from this interview.


United Nations, ‘Victims of Intercommunal Violence in South Sudan Deserve “Justice, Truth and Reparations”: Bachelet’, 22 May 2020,, accessed 10 August 2021.


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 584.


Interview 58.


Interview 58.


Interview 82 – Community protection leader, Lashio, Myanmar, April 2019; also Interview 83 – Community protection leader, Lashio, Myanmar, April 2019; Interview 85 – Community protection leader, Lashio Myanmar, April 2019; Interview 86 – Community protection leader, Lashio, Myanmar, April 2019; Interview 87 – Unarmed civilian protection practitioner, Lashio, Myanmar, April 2019.


Interview 100; also Interview 88 – Unarmed civilian protection practitioner, teleconference, May 2019; Interview 96 – International unarmed civilian protection practitioner, New York City, July 2019; Interview 92 – International unarmed civilian protection practitioner, teleconference, July 2019.


Mégret, ‘Beyond the “Salvation” Paradigm’, p. 584.


Jonathan Wilkenfeld, Kathleen Young, David Quinn, and Victor Asal, Mediating International Crises (Abingdon: Routledge, 2005); William Nomikos, ‘Peacekeeping and the Enforcement of Intergroup Cooperation: Evidence from Mali’, esoc Working Paper No. 20, 2020.


See Zach Vertin, A Poisoned Well: Lessons in Mediation from South Sudan’s Troubled Peace Process (New York: International Peace Institute, 2018).


Thomas Weber, ‘From Maude Royden’s Peace Army to the Gulf Peace Team: An Assessment of Unarmed Interpositionary Peace Forces’, Journal of Peace Research, 30(1) 45–64 (1993).


unmiss, ‘Attacks on Civilians in Bentiu & Bor April 2014’, 9 January 2015, p. 21,, accessed 10 August 2021.




Testimony from Andres Gutierrez, International Protection Officer with Nonviolent Peaceforce, 12 May 2014,, accessed 10 August 2021; Interview 47 – Unarmed civilian protection practitioner, teleconference, September 2018.


Public testimony from Andres Gutierrez, International Protection Officer with Nonviolent Peaceforce, 12 May 2014,, accessed 10 August 2021.


Gada Mahrouse, Conflicted Commitments: Race, Privilege and Power in Solidarity Activism (Montreal and Kingston: McGill-Queen’s University Press, 2014), p. 11.


Mahrouse, Conflicted Commitments, p. 146; Interview 47; Interview 41 – International protection practitioner, Berlin, July 2018; Interview 45 – International protection practitioner, Beirut, July 2018.


Interview 138 – National protection practitioner, Bentiu, December 2019.


Interview 82; Interview 85.


There are a range of examples across the United States (both government and non-government supported), such as ManUp! (Brooklyn), Take Back Our Hoods (North Carolina), Cure the Streets (Washington DC), Dallas Cred (Dallas), and member groups of the Los Angeles Intervention Coalition (laic) (Los Angeles): Interview 97 –Unarmed civilian protection practitioner, teleconference, July 2019.


Interview 139 – Unarmed civilian protection practitioner, Brooklyn, February 2019.


See examples relating to interposition in Alberto L’Abate, ‘Nonviolent Interposition in Armed Conflicts’, Peace and Conflict Studies, 4(1) article 3 (1997), DOI: 10.46743/1082–7307/1997.1180.


unsc Res. 2577, 28 May 2021, available at, accessed 10 August 2021.


Adam Dieng, Statement by the UN Special Adviser on the Prevention of Genocide to the UN Security Council on his visit to South Sudan, 17 November 2016. Available at–17.AD.Statement%20to%20SC.South%20Sudan%20-%20final.pdf, accessed 10 August 2021.


Pinaud, War and Genocide in South Sudan, p. 17.


Andrew H. Kydd and Scott Straus, ‘The Road to Hell? Third-Party Intervention to Prevent Atrocities’, American Journal of Political Science, 57(3) 673–684 (2013), p. 683.


Shahar Hameiri and Lee Jones, ‘Beyond Hybridity to the Politics of Scale: International Intervention and “Local” Politics’, Development and Change, 48(1) 54–77 (2017); Joanne Wallis, ‘Is “Good Enough” Peacebuilding Good Enough? The Potential and Pitfalls of the Local Turn in Peacebuilding in Timor-Leste’, The Pacific Review, 30(2) 251–269 (2017).


Brigg, ‘Humanitarian Symbolic Exchange’, p. 847.


Though I do share the scepticism of critics of R2P who have highlighted the dangers of reactive military intervention. An overview of critiques is provided in ‘Is R2P Still Controversial? Continuity and Change in the Debate on “Humanitarian Intervention”’, Cambridge Review of International Affairs, 31(5) 415–436 (2018). On risks of self-protection in the face of atrocity crimes, see Deborah Mayersen, ‘“Is Help Coming?” Communal Self-Protection during Genocide’, Stability, 9(1) 1–17 (2020).


Brigg, ‘Humanitarian Symbolic Exchange’, p. 845.


Baines and Paddon, ‘“This Is How We Survived”’, p. 243.


Audre Lorde, Sister Outsider (Berkeley: Crossing Press, 1984).

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 2541 802 69
PDF Views & Downloads 2693 1048 89