Abstract Contemporary conflict prevention depends on information gathering and knowledge production about developments within the borders of a state, whose internal affairs have been deemed precarious by external actors. The international community, especially the United Nations (UN), calls this early warning and early action. However, for governments whose affairs are considered in need of monitoring, preventive endeavours—and the knowledge production they entail—can be seen as ‘early aggression’. In this article, we argue that seeing knowledge production as having power effects reveals contemporary conflict prevention as an interventionary practice. Through an analysis of the international community's preventive diplomacy vis-à-vis Burundi (2015–2016) we highlight three unintended power effects: privileging the UN's knowledge production created resistance to international involvement from the Government of Burundi, it led to a change in patterns of violence and to a backlash against the institutionalization of international monitoring beyond Burundi, and it enabled arguments for further, more forceful, intervention possibilities. This framing enables us to understand the recent return to conflict prevention not as a retreat from liberal interventionism, but as a pragmatic response to its purported crisis. Crucially, although conflict prevention falls short of military intervention, it nonetheless leaves important interventionist footprints. The framing of interventions as ‘saving strangers’, or as responses to a shared responsibility to protect populations from mass atrocities, has become dominant, especially with ‘the expansion of liberal interventionism throughout the 1990s’.1 Many important dimensions of the contemporary intervention landscape have been explored in debates about liberal interventionism's response to crisis, and the optimistic views of the 1990s have been challenged. With reference to how ‘Iraq and Afghanistan have created a crisis of both confidence and credibility for proponents of intervention’, liberal interventionism is (again) said to be in crisis.2 One central theme in current debates about the ‘crisis’ of liberal interventionism is a disagreement over the nature of this crisis. In response to what he refers to as a ‘hyper-critical’ school of scholars, Roland Paris sets out to save liberal interventionism not from its crisis but from its critics.3 To him, the question becomes whether the focus should be on ‘explor[ing] alternatives within liberal peacebuilding’ or alternatives that go beyond liberal interventionism.4 These disagreements are reflected in different answers to the question of whether the current crisis in liberal interventionism is best offset by pragmatic, incremental improvements, or whether more radical changes are needed.5 Another central theme is the question of how to interpret liberal interventionism's response to this crisis of confidence and credibility. While some scholars suggest that we are seeing the return of proxy interventions,6 or that we are witnessing an ‘impasse’, as liberal interveners ‘are increasingly unable to purposefully and instrumentally engage with international governance problems’,7 others argue that the crisis has sparked a retreat from liberal interventionism. Neil Cooper, for example, suggests that we are seeing a ‘retreat from democracy in liberal interventionism’,8 while David Chandler describes what he see as ‘a retreat from the consequences of this liberal internationalist imaginary’,9 and Mark Duffield argues that this retreat—in his analysis, a retreat of international aid workers into fortified compounds—is not antithetical to a form of interventionism that at the same time appears ‘expansive’.10 The present article adds to these debates, developing the idea that what may appear as a retreat is not necessarily antithetical to expansion. We do so by showing why attention should be given to one important, yet under-theorized, form of intervention—conflict prevention—when interpreting liberal interventionism's response to its double crisis. Key institutions of liberal intervention, notably the UN, are placing renewed emphasis on conflict prevention, and have made monitoring and other forms of knowledge production central elements of preventive diplomacy.11 In this article, we analyse what this renewed focus on conflict prevention says about the crisis in liberal interventionism. We argue that while conflict prevention is sometimes presented as a pragmatic retreat from ‘intervening’, it is better understood as a different mode of intervention. Importantly, focusing on conflict prevention allows us to demonstrate how, although ‘policy debates have shifted away from intrusive forms of coercive international governance’, this trend is not necessarily indicative of retreat or disengagement.12 Indeed, we offer an alternative reading of prevention as an interventionary practice which is inseparable from power/knowledge effects unintended by interveners.13 Once conflict prevention is recognized as a practice of intervention, the recent return to (and renewed attempts at) this approach can then be understood as part of a broader trend wherein liberal interventionism is finding new forms of engagement, including ‘light footprint’, technology-enabled forms of intervention with fewer ‘boots on the ground’, and/or ‘earlier’ forms of intervention.14 We argue that the knowledge production inherent in contemporary conflict preventions is interventionary in and of itself. It not only paves the way for future interventions, but produces changes in conflict dynamics on the ground. The overarching research question that we address in this article is as follows: how does understanding knowledge production for conflict prevention as an interventionary practice offer new analytical perspectives on the 2015 crisis in Burundi? To explore this question, the article proceeds as follows. First we situate the turn to prevention in the context of broader reforms of UN peace operations, with prevention becoming a continuous undertaking aimed at ‘sustaining’ rather than ‘building’ a (liberal) peace. Next, we situate our conceptualization of prevention as intervention in relation to current calls for a broader understanding of what should be considered an interventionary practice. We go on to argue that Michel Foucault's notions of power/knowledge and resistance are useful analytical lenses through which to show that liberal interventionism's preventive endeavours are interventionary. The significance of conceptualizing preventive knowledge production as interventionary is then demonstrated in an analysis, based on interviews in Bujumbura and New York, of the UN's preventive endeavours in relation to the political crisis in Burundi in 2015–2016.15 A pragmatic (re)turn to conflict prevention Humanitarian interventions in which international forces took it upon themselves to ‘save strangers’ were arguably the most visible aspect of the 1990s surge in interventionism. Yet, in response to the failures of the UN in Rwanda and elsewhere, the 1990s also witnessed debates about conflict prevention. Conflict prevention is not a new ambition in the UN system.16 Recognition of the importance of preventive diplomacy, as part of broader conflict prevention efforts, dates back to the days of Dag Hammarskjöld in the 1950s.17 More recently, the Brahimi Report of 2000 mentioned fact-finding missions and monitoring as being at the very heart of preventive efforts, while at the same time noticing ‘the understandable and legitimate concern of Member States, especially the small and weak among them, about sovereignty’.18 As conflict prevention has returned to the agenda, the case of Burundi illustrates how monitoring has taken priority over concern for the sovereignty of this small state. Micah Zenko and Rebecca Friedman have made a related argument about UN early warning systems. Writing in 2011, they found a ‘general aversion of member states to being monitored by any outside organization for activities that occur within their sovereign territory’.19 The 2015 reports by the High-level Panel on Peace Operations (HIPPO) and the Peacebuilding Commission (PBC) Advisory Group of Experts laid the analytical ground for a renewed focus on conflict prevention, especially by introducing to the UN system the concept of sustaining peace, which makes peacebuilding a continuous enterprise and therefore necessitates constant knowledge production.20 Indeed, a key element of the contemporary re-emergence of conflict prevention is its focus on the need for central actors, notably the UN Security Council (UNSC), to be informed about developments on the ground much earlier and more extensively than has hitherto been the case. Since taking office on 1 January 2017, the new Secretary-General, António Guterres, has made prevention his main platform; for example, in his first address to the UNSC in his new capacity, he said: ‘We must rebalance our approach to peace and security. For decades, this has been dominated by responding to conflict. For the future, we need to do far more to prevent war and sustain peace.’21 This renewed focus on prevention is often justified with reference to the preferability of preventing conflict over managing conflict. Arguments in favour of preventing conflicts revolve around the lower human and economic costs incurred where the need for forceful intervention is avoided. It is commonly accepted that ‘many—both within the UN and outside of it—believe that an increased focus on conflict prevention would be far cheaper than a continued narrow focus on ongoing or recurring conflicts’.22 The Swedish Minister of Foreign Affairs, as President of the UNSC, exemplified this belief in her statement of 10 January 2017, in which she coupled the UN's resource constraints with the urgency of improving conflict prevention: The United Nations humanitarian and peacekeeping instruments have come under immense pressure, with $22.2 billion in humanitarian appeals and over 100,000 United Nations peacekeepers. Meanwhile, research shows that measures to peacefully prevent conflict cost, on average, just a tenth of post-conflict recovery efforts. Investing in prevention is not only morally right; it is the smart, economically sound and sustainable thing to do.23 The hope is that, by preventing conflict, the need to deploy UN peacekeepers can be reduced, which would mean less intervention and fewer expenses. This viewpoint is echoed by research. Malcolm Chambers, for example, has calculated that, ‘even on conservative assumptions, investments in conflict prevention can be highly cost-effective for the international community’.24 One result of the distinction between preventing and managing conflict is that conflict prevention is seen as separate from intervention. Instead of analysing the power effects of conflict prevention, the focus tends to be on its technical aspects. Alice Ackermann, for example, surveys the literature and finds that while much attention is given to the ‘viability, legality, and effectiveness of conflict prevention’, its political effects are never questioned, and so it is not seen as interventionary.25 In contrast to this prevailing perspective, we argue that preventive endeavours have a number of important effects that are de facto interventionary. On the basis of this argument, we offer an alternative perspective on the UN's conflict prevention endeavours by showing how the knowledge production inherent in contemporary preventive diplomacy has become an element in intense struggles over claims to define the ‘true’ account of the contested situation in Burundi. Analytical framework: interventionary practices and power/knowledge Traditionally, and especially in the liberal peace tradition, intervention means forceful intervention. Michael Doyle, for example, contrasts this with non-forceful intervention, which he refers to as ‘interference’.26 Unfortunately, the issue of conflict prevention has been ‘masked by’ the much stronger focus on ‘forceful interventions’.27 Nonetheless, some scholars have argued that to understand the contemporary landscape of intervention, we need to look not only at practices that qualify as forceful interventions, but at interventionist practices more broadly conceived, including capacity-building missions, crisis management and ‘neoliberal reforms’.28 These scholars have called for a broader understanding of what practices are conceived of as interventionary. They have debated what this broader set of interventionary practices tells us about how power operates in the context of contemporary interventionism—and not only when intervention takes the form of forceful, military operations.29 This literature frames our analysis of the UN's preventive activities in Burundi in 2015 and 2016 as an interventionary practice. Like the scholars cited above, we broaden the notion of intervention. We argue that it should include elements of contemporary conflict prevention, with a specific focus on preventive knowledge production.30 To frame our analysis of the preventive endeavours of the UN as interventionary, we draw on Foucault's concept of power/knowledge and the idea that ‘there is no point in dreaming of a time when knowledge will cease to depend on power … It is not possible for power to be exercised without knowledge, it is impossible for knowledge not to engender power.’31 In drawing upon this aspect of Foucault's work, we engage with and build on an important post-structuralist legacy as well as emerging debates about knowledge production in intervention.32 Concerning the former, early post-structuralists demonstrated how Foucauldian insights could help expand our appreciation of the effects of practices of intervention: for example, by calling attention to the role of representational practices in shaping possibilities for intervention in the case of Bosnia,33 and to knowledge as ‘a constitutive part of the world of meaning and practice’,34 rather than as something neutral and unprejudiced, distinct from the exercise of power. While these insights form a crucial backdrop for the analysis offered in this article, we also seek to extend them by suggesting that conflict prevention and related knowledge production are themselves to be understood as interventionary—that is, not as important only because of how they shape possibilities for future—or rather further—intervention. From this starting-point, our analysis demonstrates how the UN's preventive endeavours, notably the knowledge produced and made available to external actors, can be said to ‘engender power’, and to affect conflict dynamics and patterns of violence in ways that one cannot see if prevention is conceptualized as something that precedes intervention. In using power/knowledge to frame our analysis of UN practices of conflict prevention, we also build on a more recent body of literature that draws upon—or examines the limits of35—Foucault's notion of power/knowledge. Examples include work by Merlingen and Ostrauskaite,36 who have used a power/knowledge approach to examine critically various EU peacebuilding activities; Solà-Martín,37 who has used the notion to examine the UN's conflict resolution endeavours in the Western Sahara; and Barry Ryan,38 who has operationalized the notion to unpack an otherwise overlooked effect of international actors' involvement in police reform, namely how ‘police reform is situated at the confluence of the production of power/knowledge and the assumptions of liberal peace’.39 Grayson, too, has applied Foucault's notion of power/knowledge in his analysis of ‘biopolitical intervention’.40 Collectively, these analyses have delivered valuable insights about, for example, power effects of seemingly neutral knowledge production. Even so, the question of power/knowledge effects in the context of the UN's efforts at conflict prevention remains under-theorized. The UN's renewed emphasis on conflict prevention represents a ripe moment for extending our power/knowledge analysis to the realm of monitoring and other forms of preventive knowledge production. When, for preventive purposes, the UN expands its monitoring activities and increasingly produces knowledge to be made available to external actors, this should be expected to entail various power effects. To study these in the case of knowledge production for the purpose of conflict prevention in Burundi, we also draw upon Foucault's notion of resistance. For Foucault, resistance is integral to the exercise of power: ‘There are no relations of power without resistances; the latter are all the more real and effective because they are formed right at the point where relations of power are exercised.’41 What this means for our analysis is that we pay careful attention to how the Government of Burundi (GoB), in various ways, resists the UN's preventive endeavours (monitoring, reporting and other forms of knowledge production) through a number of counter-strategies. Within this framing, the GoB's refusal to accept the preventive presence of the UN becomes an important indicator of resistance to the UN's exercise of power. This analysis challenges the representation of preventive monitoring and reporting as neutral and objective practices—even if they are understood as such by those who implement them. Conflict prevention: the case of Burundi Our analysis of how power/knowledge effects played out in the context of the international community's conflict prevention efforts in relation to the situation in Burundi in 2015–2016 is based on interviews as well as analysis of official documents and news sources. We conducted interviews in Bujumbura in July 2016 and in New York during the autumn of that year. In Bujumbura, we interviewed Burundian government officials, including politicians and senior civil servants, whose statements form the basis of our analysis of the actions and positions of the GoB. We did not conduct interviews with opposition figures, relying instead on their public statements, both for security reasons and because the analysis is mainly about the relation between the GoB and the international community. Also in Bujumbura we interviewed independent Burundian academics, who were immensely helpful in offering perspectives on historical and institutional dynamics. In both Bujumbura and New York, we interviewed UN staff members and foreign diplomats covering Burundi about their roles and positions.42 All statements from the interviews are treated either as the views of interested parties or as a prompt to triangulate information with other sources. The additional sources consist mainly of official documents that express the decisions and standpoints of the parties, supplemented with news sources when necessary.43 The international community has been involved in developments and crises in Burundi for decades.44 The analysis presented below is, however, limited to a discussion of selected events that occurred between the spring of 2015 and early 2017, when Burundi became the focus of an intensive campaign of preventive diplomacy by the international community, notably the UN.45 On 25 April 2015, the ruling party in Burundi announced that the sitting President, Pierre Nkurunziza, would stand in the forthcoming elections scheduled for 26 June 2015. This was controversial, because the constitution limits presidential tenure to two terms of five years, and Nkurunziza had been president for ten years; the opposition accordingly warned that his renewed candidature was unconstitutional and could send Burundi into chaos.46 In turn, Nkurunziza argued that because the parliament had appointed him for the first five years, that period of his presidency should not be included in the constitutionally defined two-term limit. On 5 May 2015, the Constitutional Court ruled in favour of the President,47 albeit amid considerable controversy (the vice-president of the court fled the country).48 In the lead-up to the presidential election, demonstrations intensified. Security forces responded with force, and accusations of responsibility for the violence went back and forth.49 On 13 May, while Nkurunziza was in Tanzania for an East African Community summit about the crisis in Burundi, the protests against the President culminated in an attempted coup d'état. This was quickly defeated by government forces, and the following day Nkurunziza returned to Burundi to declare victory.50 Following pressure from the opposition and the international community, the elections were postponed to 21 July 2015. The opposition boycotted the poll, but the electoral commission kept their names on the ballots, as they had already announced their candidatures. Nkurunziza won with 69.41 per cent of the votes.51 At this point, the violence began to take on the form of an armed insurrection. Grenade attacks, including on civilian targets and police forces, created an atmosphere of uncertainty and fear. On 11 December 2015, the violence reached a peak with attacks on army barracks.52 The government crackdown that followed reportedly left 34 people dead in the streets of Bujumbura, in addition to those killed in the attacks.53 In 2016, overt general violence diminished once more, but targeted assassinations of political figures began to become a dominant motif.54 By the end of the year, the situation was seemingly stable, but remained very tense. More than 300,000 refugees remained outside the country and, as we will show below, disagreements over both how to define the situation and how to respond to it ran deep.55 During the events of 2015–2016, the international community made significant efforts to prevent the precarious situation from escalating into civil war or even genocide. These efforts included that the UNSC adopted three resolutions, three presidential statements and conducted two missions to Burundi. In short, the UN significantly increased its preventive endeavours in Burundi. Below, we elaborate on these UN efforts at preventive diplomacy by looking at three types of power/knowledge effects. Whereas external actors mostly argue that by intensifying their monitoring they simply increase the possibility of knowing the truth about affairs in a given country, Foucault's power/knowledge concept helps us recognize how such endeavours (monitoring the situation inside Burundi ever more closely in the name of delivering truthful information) are inseparable from important power effects. We therefore analyse conflict prevention in Burundi by asking not simply whether it succeeded, but what effects it had. First, we discuss how the UN's claim to know the truth about the situation in Burundi was accepted, but also how the GoB resisted the UN's position as sole provider of ‘neutral’ knowledge and questioned its role. Second, we analyse how the GoB started arguing in a different register in response to what it found to be unwarranted knowledge production about Burundi. Third, we show that there are competing narratives about Burundi as an object of knowledge, with views ranging from the situation being calm and under control to the country being on the brink of a genocide—variations that entail very different perspectives on the need, or otherwise, for (further) intervention. Defining Burundi as in crisis The renewed emphasis on conflict prevention also entailed a new emphasis on information gathering, analysis and reporting by UN agencies. Indeed, the UN increased its monitoring presence in Burundi to an extent that the GoB considered entirely excessive. Even voices within the UN Secretariat saw a risk of heavy-handedness: ‘Burundi receives a special envoy every two weeks. The fact that there is almost an army of us going to Burundi is too much. All they see is the UN, AU [African Union], and EU being against them.’56 Not only did members of the UNSC visit Burundi to see the situation for themselves, the Council increasingly requested that it should be made party to observations by various other ‘outside’ actors (from the AU and other parts of the UN). For example, the Secretary-General's Special Adviser, Jamal Benomar, was asked on a number of occasions (e.g. in February, April, October and November 2016) to share his observations from Burundi.57 In these ways, the UNSC privileged the knowledge being produced by various UN and AU actors, thus constituting these external actors as objective sources of knowledge and truth. Other actors, too, referred to the UN as ‘a neutral source of information’, even if they acknowledged the existence of contradictory narratives about the situation in Burundi.58 Furthermore, supported by the UN, the AU decided to send and later augment a number of military observers and human rights observers. The first round of observers arrived in Bujumbura on 22 July 2015. The GoB was allowed only a peripheral role in discussions about Burundi in the UN and other international forums, and it was absent in informal UNSC meetings, where most of the deliberations took place. NGOs working in Burundi were another potential source of information, which was not prioritized in the UN system. Reports from human rights NGOs were referred to selectively from time to time, but never in official documents. One western diplomat explained his view of a particular report, saying: I do not want to judge its quality, but I know that some advocates feel it is a bit off the mark. You do not have that with OHCHR [Office of the UN High Commissioner for Human Rights]. Since they are in the Secretariat, I have the feeling they are more cautious and neutral.59 In making this comment, the diplomat indicated a hierarchy of knowledge providers, with the UN at the top, and NGOs and the GoB further down. The UN dream of objective knowledge The starting-point for the UN's conflict prevention ambitions in respect of Burundi is that it is possible, through knowledge-producing activities, to objectively access and convey the truth about the situation in the country. In January 2017, António Guterres formulated this view as follows: ‘Preventive action is essential to avert mass atrocities or grave abuses of human rights. And we can achieve this only through reasoned discussion, based on facts and the pursuit of truth.’60 Examples of such knowledge-producing activities include monitoring carried out by the OHCHR, reporting by the Special Adviser of the Secretary-General on Conflict Prevention, including in Burundi, and field visits made by representatives of the UNSC and other high-level envoys. Underwriting all these activities is an assumption that each of these UN bodies will add new details to what will eventually amount to an objective assessment of the situation in Burundi. This ‘dreaming of a time when knowledge will cease to depend on power’ is part of the explanation why the UN cannot see its knowledge production as interventionary, let alone take into account its power effects. The UN's assumption that it is possible to know the truth about an objective reality in Burundi clashes with the existence of radically different accounts of events in the country. At least three descriptions of the situation were presented by different actors. Broadly speaking, positions differed over whether the situation was stable, whether there was a risk of conflict escalation and whether Burundi was on the brink of genocide. Tellingly, in July 2016 we interviewed a government official and a UN staff member in their respective offices in Bujumbura. The government official pointed out of the window to a quiet street, saying that everything was under control and inviting us to visit the interior of the country. The UN staff member pointed to a security installation which it had been necessary to set up because of fear of grenade attacks. The most controversial narrative was the one focusing on the danger of genocide, because it evoked the greatest potential for international action. Against the backdrop of a history of civil war and ethnic strife in Burundi, fear mounted in the international community that the conflict could acquire an ethnic dimension.61 Officially, the UN was careful not to accentuate the ethnic dimension of the conflict. For example, while the UN Special Adviser on the Prevention of Genocide, Adama Dieng, warned about ethnic violence and the risk of atrocity crimes during his visit to Burundi on 29–30 May 2015, he stopped short of discussing genocide.62 Perhaps the strongest criticism of the GoB by a UN source came from the UN Independent Investigation on Burundi (UNIIB), which was mandated by the Human Rights Council (HRC). Its report stated that ‘experts find that gross human rights violations have and are taking place, committed primarily by state agents and those linked to them’, and concluded that: Without determined interventions by the Government of Burundi and a renewed robust engagement by the international community, including the United Nations and the African Union, the country's downward spiral is unlikely to be reversed, endangering not only the rights of individuals concerned but also the overall security of the region.63 Resistance from the government of Burundi The GoB consistently sought to counter the logic that more monitoring is a neutral means of getting closer to the true situation in Burundi. One of the ways in which it sought to do this was by discrediting the UN's attempt to position itself a an anchor of objective knowledge about the situation in Burundi. From the perspective of the GoB, the knowledge produced by the UN is not objective truth but biased criticism. One example of such an attempt at discrediting the UN was made in March 2016, when members of the UNSC came on mission to Burundi to observe the situation. Disagreements among the P5 meant that no clear statement was issued afterwards.64 The GoB used this dissent in the UNSC to argue that there was no crisis in Burundi that would justify continued UN scrutiny, portraying the UNSC's silence as an indication that all was well.65 Thus, the visiting mission can be seen as counterproductive to the Council's preventionist ambition. Similarly, an interviewee from the Foreign Ministry in Burundi pointed out: ‘African ambassadors see one situation, Europeans see another’, thereby highlighting the coexistence of competing accounts of events while simultaneously attempting to structure the divide in a specific way, portraying preventionist ambitions as a European agenda. The GoB also used divisions in the UNSC to discredit and challenge the idea that the UN held privileged and objective knowledge about the situation, arguing for example that the decision to issue Resolution 2303 was illegitimate as it was based on ill-founded and biased information about the situation in Burundi. In other words, the GoB did not accept that the UN—or indeed any other outside actor—should be categorized as neutral or trustworthy. In denying, first, that outsiders were neutral, the GoB was not making an epistemological point, but rather asserting that the outsiders were politically biased against them. We were told that ‘there are numerous signs that the international community is on the side of the opposition’.66 One interviewee from the ruling party in Burundi, CNDD-FDD, asked why the countries opposing Nkurunziza's candidature did not oppose leaders of other countries in the region who flouted term limits or stayed in power, for example in the Democratic Republic of Congo, Rwanda and Uganda. In an official statement of July 2016, the ruling party attacked the opposition for being against democracy and international organizations for acting on their behalf.67 In other words, from the perspective of the GoB, the international community was discriminatory in its approach to Burundi, which resulted in unreasonable overreaction. Second, the GoB maintained that the UN and other international actors could not be trusted. The evidence that almost all of the government representatives we interviewed would cite revolved around two specific stories. The first was that the United States, and consequently the UNSC, was biased because the US Ambassador to the UN at the time, Samantha Power, was a close friend of Alexis Sinduhije, leader of the Burundian opposition party Mouvement pour la Solidarité et le Développement (MSD). They made sure we knew that in 2000 Samantha Power co-founded an opposition radio station (Radio Publique Africaine) with Sinduhije, including by showing us a summary of the founding meeting of the station.68 They did not mention that in late 2015 the United States instituted targeted sanctions against Sinduhije and three other named persons in Burundi for being a threat to peace and security.69 The other recurrent story was that French media, Radio France Internationale and France 3 in particular, had used images in their coverage of Burundi that were misleading; specifically, they had published a picture, saying it was from Burundi, although it had actually been taken in Nigeria and at a different time. Both outlets later denied the allegation, but our respondents saw it as a deliberate attempt to smear Burundi, orchestrated by the French government.70 Perhaps the most important means of resistance used by the GoB was its ability to deny international observers access, thereby limiting their ability to monitor. The UN was, for example, forbidden to go into the streets of Bujumbura to undertake monitoring following the attacks and raids that took place in December 2015. Amnesty International later reported that it had identified mass graves filled with the bodies of those killed during the incidents. In the same vein, the GoB responded particularly strongly to the UNIIB report of September 2016: on 11 October it issued an official statement saying that it had suspended cooperation with the the OHCHR, which meant that current OHCHR officers were no longer allowed to observe and monitor the situation in Burundi, and that their visas would not be renewed. According to the GoB, OHCHR was ‘complicit’ in writing the ‘lying and controversial report’. The statement continued by naming three experts in UNIIB as persona non grata.71 Around the same time, the GoB requested that Benomar no longer cover Burundi for the UN. That letter was couched in diplomatic niceties, but nonetheless amounted to throwing him out of the country, which had also been the fate of his predecessors as UN representatives during the crisis.72 Arguing in a different register: effects of power/knowledge on patterns of violence Together with the GoB's resistance, the UN's preventionist knowledge production had important effects on the situation in Burundi. Monitoring the situation in order to produce supposedly superior knowledge contributed to a shift in (rather than simply a lessening of) violence. Specifically, violence shifted away from the visible public forms it had taken at the beginning of the crisis, such as the piles of burning tyres seen in media images, the coup d'état, and crackdowns on demonstrations in the streets of Bujumbura. These were generally replaced with more individualized forms such as targeted assassinations and extra-judicial killings, allegedly perpetrated by both sides. The GoB is also said to have responded by changing its tactics to contain the opposition through less observable, but no less repressive, means. The more covert forms of repression reportedly included moving the violence to secret facilities, so that demonstrators were no longer beaten in the streets but removed to hidden torture chambers. The International Federation for Human Rights (FIDH) found that, in place of the earlier mass arrests, from the beginning of 2016 numbers of enforced disappearances increased. ‘This practice aims in particular to conceal evidence of abuses by security forces,’ it notes.73 If the FIDH is correct, this development fits well with the overall picture of the GoB as being increasingly opposed to scrutiny. Unlike the UN, the GoB perceives knowledge production as interventionary. The pressure on Burundi also led the government to perceive itself as isolated in international forums. It saw less interest in constructive engagement in these forums, which it perceived to be used as platforms for international criticism of the GoB. A clear consequence with potential fallout for the international legal order more broadly was the announcement that Burundi would leave the International Criminal Court (ICC), being the first ever country to do so since the court's creation. In this way, the GoB would ‘hide’ itself from outside legal scrutiny. Also, according to diplomats covering the Third Committee of the UN General Assembly, which deals with human rights issues, Burundi took an unusually hard line during the 71st session in autumn 2016, ostensibly in a backlash against outside human rights monitoring of individual sovereign states. As an example, another diplomat pointed out that Burundi had traditionally voted in favour of the General Assembly's resolutions on human rights in the Democratic People's Republic of Korea (DPRK), but reversed its position to vote against in 2015.74 Underlying the switch, according to the same diplomat, was the thought that Burundi might be the next object of a critical resolution in the Third Committee.75 In other words, this is an example of Burundi working to delegitimize outside knowledge production about the ‘internal’ affairs of states, and attempting to decrease international visibility and the productive effects of power/knowledge. By early 2017, the obstruction of outside monitoring activities could be said to have had some success. Diplomats in New York told us that they were preoccupied with other situations in the region. Since little was happening in Burundi, or at least little they knew of, there was not much they could do. After the adoption of Resolution 2303 on 29 July 2016, the UNSC only had Burundi on its agenda twice during the rest of the year—a considerable slowdown for the period—and mostly to be briefed by Benomar on the (lack of) implementation of the resolution. From ‘truth’ to intervention (or not) Since the UN, through its preventive endeavours, had become the key source of information about Burundi, it mattered a great deal what knowledge was circulating in the UN system. The accepted knowledge about the situation has a strong influence on the perceived need for further intervention, and thus shapes the policy options that international actors see as relevant. For example, in Resolution 2303 the UNSC reiterated its ‘deep concern about the persistence of violence in Burundi’ and—despite an explicit lack of consent from the GoB—authorized the deployment of 228 police officers.76 As of October 2017, the lack of consent had made it impossible to implement the resolution.77 In taking its decision, the Council relied significantly on information from UN agencies' monitoring activities and briefings by UN officials. As such, the choice to report on Burundi as a country in crisis, possibly even at risk of genocide, enabled centrally placed actors to explore the possibilities for intervention in the more traditional sense of the word, that is, with foreign troops on the ground. Such intervention had almost materialized at an earlier moment, when in December 2015 the AU Peace and Security Commission decided to deploy an African Prevention and Protection Mission in Burundi (MAPROBU), consisting of 5,000 peacekeepers, whose first mandated task included further knowledge production in the form of monitoring: ‘MAPROBU is mandated to: (a) prevent any deterioration of the security situation, monitor its evolution and report developments on the ground.’78 In the end, deployment was cancelled at a subsequent AU summit. This account offers an illustration of how the knowledge production carried out in the name of conflict prevention can have important impacts on the framing of intervention possibilities. Crucially, the privileging of observations made by various UN bodies about detailed developments ‘inside’ Burundi made it possible, in the name of prevention, to render Burundi open to new forms of intervention by external, knowledge-producing actors. This in turn enabled other external actors (notably the AU, East African Community and EU) to enter the picture as producers of valid knowledge about the situation in Burundi—a situation whose ‘truth’ was thus increasingly produced by external actors. The power to define Burundi was taken away from the GoB and vested in outside entities. Case findings The definition of Burundi as in crisis, the change in patterns of violence, and the framing of Burundi as ‘intervenable’ are three examples of effects of UN knowledge production activities—activities that are integral to contemporary forms of conflict prevention. For now, the international community, and the UN in particular, does not recognize that its knowledge production ‘induces effects of power’.79 When international actors attempt to move the goalposts of sovereignty, as they have done with preventive diplomacy in Burundi, it is regarded as a progressive move ‘from early warning to early action’;80 but, seen from the perspective of the GoB, the risk is that it becomes ‘early aggression’.81 Thus, conflict prevention is not an alternative to intervention, but itself an interventionary practice. We have shown the importance of taking note of the new and sometimes subtler forms that interventionary practices are taking—and of the potential for even subtle forms of intervention to have important constitutive effects. More specifically, we argue that to appreciate the scope and effects of contemporary practices of intervention, knowledge production for the purpose of early warning and preventive diplomacy should be included among these practices. We thus challenge the idea that by aiming to prevent violent conflict, intervention can be avoided. Conclusion: conflict prevention as a pragmatic response to a twofold crisis—in Burundi and of liberal interventionism Some observers have deemed UN endeavours at conflict prevention in Burundi an outright failure. This assessment is based partly on how the GoB has responded to the UN's initiatives, and partly on the continuance—and shifting patterns—of violence in Burundi. Others argue that conflict prevention has been successful, in that the situation would have deteriorated further had the UN not been monitoring the situation. They argue that UN engagement ‘led to relatively peaceful elections’ and that ‘things could have been much worse otherwise’.82 Whether conflict prevention was a failure or success is likely to continue to be disputed. However, irrespective of how that question is answered, the insight remains that the UN's knowledge production for purposes of preventive diplomacy had important constitutive effects on the situation. In the context of a changing interventionist landscape, the case of Burundi illustrates a mode of engagement that avoids traditional large-scale interventions, but ends up being highly interventionist nonetheless. Thus this analysis of conflict prevention as intervention is linked to broader debates about the expansion or retreat of liberal interventionism. Contributing to these debates, our analysis suggests that although preventive intervention falls short of large-scale interventionism and might be assumed to overcome some of the associated challenges, paying attention to the power/knowledge effects of preventive modes of engagement allows us to better appreciate the important effects of current forms of preventive intervention, and their capacity to produce some of the same challenges. By portraying conflict prevention as an alternative to intervention, most discussions short-circuit considerations about the pros and cons of intervention. As such, the aspiration to prevent conflict risks blinding the international community to the unintended consequences that follow from such interventions. Obviously, we agree that conflict prevention is preferable to conflict. What we wish to highlight is the point that seeing prevention as intervention will allow important lessons to be learned and usefully applied, including the risk of unintended consequences. The implementation of preventive diplomacy is part of a pragmatic turn in the sense that it is indicative of a shift away from a ‘blueprint’ or ‘template’ approach to conflict prevention, in line with the HIPPO Report's assessment that ‘too often, mandates and missions are produced on the basis of templates instead of tailored to support situation-specific political strategies’.83 That said, preventive diplomacy, and the monitoring activities it necessitates, risks becoming a new template if applied uncritically to all situations. There are many good reasons for making prevention the preferred solution, including reasons laid out in the HIPPO and PBC reports; but the recommendation that responses be tailored to each individual situation certainly also remains valid. As the concept of sustaining peace is still new to the UN system and the crisis in Burundi remains unresolved, it is too early to evaluate the implications for broader policy developments. If, however, knowledge production for preventive diplomacy comes to be seen as a best practice, it is likely to be replicated by the international community in other settings. In this way it could become a mechanism for the expansion of interventionary practices. It is possible to argue that this form of intervention is already projecting power in more countries than were previously considered accessible to outside interference. For example, since 2014 the UNSC has held annual debates about the human rights situation in the DPRK, notwithstanding strong opposition from some member states, including the Russian Federation and China. This means that, disregarding whether the government in Pyongyang consents to the definition, the UNSC defines the situation in the DPRK as a threat to international peace and security—a claim explicitly based on monitoring and reporting conducted by UN officials.84 As such, the approach is very similar to that pursued in the case of Burundi. To summarize, conflict prevention in the form of monitoring and reporting from a supposedly superior position of knowledge production is inseparable from important power effects; and by appreciating this, it becomes possible to see conflict prevention as a pragmatic interventionary practice in response to a dual crisis. The UN is testing new ways of preventing the escalation of a particular crisis in Burundi. This, however, is also part of a broader response to ongoing debates about a ‘crisis’ of liberal interventionism—a response that is not simply a retreat but rather a testing of new modes of engagement. There may well be a shift away from large-scale interventions; but conflict prevention is increasingly regarded as an alternative, that is, as another way of (rather than a retreat from) intervening, with a lighter footprint. From this perspective, the framing of conflict prevention as an interventionary practice leads us to conclude that liberal interventionism is still active. On the surface, there is a retreat from liberal interventionism because of a lack of political will to make large-scale forceful interventions; but we are also seeing an expansion through increased efforts aimed at monitoring strangers. 1 Nicholas Wheeler, Saving strangers: humanitarian intervention in international society (New York: Oxford University Press, 2000); Edward C. Luck, ‘Sovereignty, choice, and the Responsibility to Protect’, Global Responsibility to Protect 1: 1, 2009, pp. 10–21; Philippa Atkinson, ‘Liberal interventionism in Liberia: towards a tentatively just approach? Analysis’, Conflict, Security and Development 8: 1, 2008, pp. 15–45. 2 Neil Cooper, ‘Review article: on the crisis of the liberal peace’, Conflict, Security and Development 7: 4, 2007, pp. 605–16. See also Stanley Hoffmann, ‘The crisis of liberal internationalism’, Foreign Policy, no. 98 (Spring 1995), pp. 159–77; Georg S⊘rensen, A liberal world order in crisis: choosing between imposition and restraint (Ithaca, NY: Cornell University Press, 2011). 3 Roland Paris, ‘Saving liberal peacebuilding’, Review of International Studies 36: 1, 2010, p. 338. 4 Paris, ‘Saving liberal peacebuilding’, p. 339 (emphasis in original); Neil Cooper, Mandy Turner and Michael Pugh, ‘The end of history and the last liberal peacebuilder: a reply to Roland Paris’, Review of International Studies 37: 4, 2011, pp. 1995–2006. 5 Roland Paris, ‘Peacebuilding and the limits of liberal internationalism’, International Security 22: 2, Autumn 1997, pp. 54–89. 6 See e.g. Florian P. Kühn and Nicolas Lemay-Hébert, ‘Letter from the editors: revisiting ten years of intervention and statebuilding’, Journal of Intervention and Statebuilding 10: 1, 2016, pp. 1–2. 7 Peter Finkenbusch, ‘“Post-liberal” peacebuilding and the crisis of international authority’, Peacebuilding 4: 3, 2016, p. 248. 8 Cooper, ‘Review article: on the crisis of the liberal peace’. 9 David Chandler, ‘Resilience and human security: the post-interventionist paradigm’, Security Dialogue 43: 3, 2012, pp. 213–29 at p. 18. On ‘retreat’, see also Colleen Bell and Brad Evans, ‘Post-interventionary societies: an introduction’, Journal of Intervention and Statebuilding 4: 4, 2010, pp. 363–70; Robin Niblett, ‘Liberalism in retreat: the demise of a dream’, Foreign Affairs 96: 1, January/February 2017, pp. 17–24. 10 Mark Duffield, ‘The liberal way of development and the development–security impasse: exploring the global life-chance divide’, Security Dialogue 41: 1, 2010, pp. 53–76. See also Michael Aaronson, Ariel I. Ahram, Mark Duffield, Amitai Etzioni, Jack Holland, Roger Mac Ginty, Laura McLeod, Sukanya Podder, Oliver P. Richmond and David Roberts, ‘The Journal of Intervention and Statebuilding ten years on: critical reflections and stimulating ideas on an evolving scholarship’, Journal of Intervention and Statebuilding 10: 1, 2016, pp. 3–24. 11 Adrian Gallagher, ‘The promise of pillar II: analysing international assistance under the Responsibility to Protect’, International Affairs 91: 6, Nov. 2015, pp. 1259–76; Alex J. Bellamy and Christopher T. Hunt, ‘Twenty-first century peace operations: protection, force and the changing security environment’, International Affairs 91: 6, Nov. 2015, pp. 1277–98. 12 David Chandler, ‘Rethinking the conflict–poverty nexus: from securitising intervention to resilience’, Stability: International Journal of Security and Development 4: 1, 2015, pp. 1–14 at p. 2. 13 Gëzim Visoka, ‘Peace is what we make of it?’, Peacebuilding 4: 1, 2016, pp. 54–70. This is true not only for ‘liberal peace actions’ undertaken by the UN. See e.g. Berit Bliesemann de Guevara's analysis of knowledge production and notions of ‘authenticity’ in German policy-making vis-à-vis conflict spaces in ‘Intervention theatre: performance, authenticity and expert knowledge in politicians' travel to post-/conflict spaces’, Journal of Intervention and Statebuilding 11: 1, 2017, pp. 58–80. 14 Duffield, ‘The liberal way of development’; Louise Wiuff Moe and Markus-Michael Müller, eds, Reconfiguring intervention: complexity, resilience and the ‘local turn’ in counterinsurgent warfare (Basingstoke: Palgrave Macmillan, 2017). 15 We wish to make it clear to the reader that this assessment should not be seen as a moral judgement by the authors. In our view, interventions can be necessary under certain circumstances. We do not take a stance on whether those circumstances are present in the case of Burundi or not. Rather, we wish to show the analytical value of understanding conflict prevention as an interventionary practice. 16 Despite the desirability of conflict prevention, it is an agenda which has proved difficult to implement—and which has arguably ‘remained peripheral to the study of peace operations’. See Alex J. Bellamy and Paul D. Williams, Peace operations and global order (Abingdon: Routledge, 2014), p. 12; also Eli Stamnes, ‘Critical security studies and the United Nations preventive deployment in Macedonia’, International Peacekeeping 11: 1, 2004, pp. 161–81. 17 UN Secretary-General, Preventive diplomacy: delivering results, 26 Aug. 2001, http://www.un.org/undpa/sites/www.un.org.undpa/files/SG%20Report%20on%20Preventive%20Diplomacy.pdf. (Unless otherwise noted at point of citation, all URLs cited in this article were accessible on 29 Oct. 2017.) 18 Report of the Panel on United Nations Peace Operations, S/2000/809 (the Brahimi Report) (New York: UN, 2000), p. 5. 19 Micah Zenko and Rebecca R. Friedman, ‘UN early warning for preventing conflict’, International Peacekeeping 18: 1, 2011, pp. 29–30. 20 Uniting our strengths for peace: politics, partnership and people, report of the UN High-level Independent Panel on UN Peace Operations, S/2015/446 (New York: UN, 2015; hereafter HIPPO Report); Advisory Group of Experts on the Review of the Peacebuilding Architecture, Challenge of sustaining peace, S/2015/490 (New York: UN, 2015). 21 Secretary-General's remarks to the Security Council open debate on ‘Maintenance of international peace and security: conflict prevention and sustaining peace’ (as delivered), 10 Jan. 2017, https://www.un.org/sg/en/content/sg/statement/2017-01-10/secretary-generals-remarks-security-council-open-debate-maintenance; Thomas G. Weiss and Tatiana Carayannis, ‘Windows of opportunity for UN reform: historical insights for the next Secretary-General’, International Affairs 93: 2, March 2017, pp. 309–26. 22 Gustavo de Carvalho and Adriana Erthal Abdenur, Getting clear about conflict prevention at the UN (Pretoria, Addis Ababa, Dakar and Nairobi: Institute for Security Studies, 7 July 2016), https://issafrica.org/iss-today/getting-clear-about-conflict-prevention-at-the-un. 23 Margot Wallström, ‘Remarks by Minister for Foreign Affairs Margot Wallström at ministerial debate in the Security Council on conflict prevention and sustaining peace’, Government of Sweden, 10 Jan. 2017, http://www.government.se/speeches/2017/01/remarks-by-minister-for-foreign-affairs-margot-wallstrom-at-ministerial-debate-in-the-security-council-om-conflict-prevention-and-sustaining-peace/. 24 Malcolm Chambers, ‘Spending to save? The cost-effectiveness of conflict prevention’, Defence and Peace Economics 18: 1, 2007, pp. 1–23 at p. 20. 25 Alice Ackermann, ‘The idea and practice of conflict prevention’, Journal of Peace Research 40: 3, 2003, pp. 339–47. 26 Discussing intervention through the lens of an essay on the topic by John Stuart Mill, Doyle finds that sovereignty is the norm, but that it is sometimes justified to override or disregard sovereignty in favour of intervention for the purpose of protection. See Michael W. Doyle, The question of intervention: John Stuart Mill and the Responsibility to Protect (New Haven, CT: Yale University Press, 2015). 27 Michael Merlingen and Rasa Ostrauskaite, ‘Power/knowledge in international peacebuilding: the case of the EU police mission in Bosnia’, Alternatives 30: 3, 2005, pp. 297–323. 28 Julien Jeandesboz, ‘Intervention and subversion: the EU Border Assistance Mission to Moldova and Ukraine’, Journal of Intervention and Statebuilding 9: 4, 2015, pp. 442–70; Katja Lindskov Jacobsen, ‘Maritime security and capacity building in the Gulf of Guinea: on comprehensiveness, gaps, and security priorities’, African Security Review 26: 3, 2017, pp. 237–56; Ruben Andersson and Florian Weigand, ‘Intervention at risk: the vicious cycle of distance and danger in Mali and Afghanistan’, Journal of Intervention and Statebuilding 9: 4, 2015, pp. 491–541; Graham Harrison, ‘Practices of intervention: repertoires, habits, and conduct in neoliberal Africa’, Journal of Intervention and Statebuilding 4: 4, 2010, pp. 433–52. 29 Marc G. Doucet, ‘Global assemblages of security governance and contemporary international intervention’, Journal of Intervention and Statebuilding 10: 1, 2016, pp. 116–32; Christian Olsson, ‘Interventionism as practice: on “ordinary transgressions” and their routinization’, Journal of Intervention and Statebuilding 9: 4, 2015, pp. 425–41; Harrison, ‘Practices of intervention’; Mark Duffield, Development, security and unending war: governing the world of peoples (Cambridge: Polity, 2007); Rita Abrahamsen and Michael C. Williams, ‘Introduction: the privatisation and globalisation of security in Africa’, International Relations 21: 2, 2007, pp. 131–41; Jan Bachmann, Colleen Bell and Caroline Holmqvist, eds, War, police and assemblages of intervention (New York: Routledge, 2015); Aaronson et al., ‘The Journal of Intervention and Statebuilding ten years on’. 30 Accordingly, we do not attempt to set definitional boundaries for intervention. 31 Michel Foucault, Power/knowledge: selected interviews and other writings 1972–1977 (New York: Pantheon, 1980), p. 52. 32 See e.g. Journal of Intervention and Statebuilding 11: 1, 2017, special issue on ‘Knowledge production in conflict and intervention: finding “facts”, telling “truth”’. 33 David Campbell, National deconstruction: violence, identity, and justice in Bosnia (Minneapolis: University of Minnesota Press, 1998); Lene Hansen, Security as practice: discourse analysis and the Bosnian War (London: Routledge, 2006). 34 Vivienne Jabri, Discourses on violence: conflict analysis reconsidered (Manchester: Manchester University Press, 1996), p. 23; see also Cynthia Weber, Simulating sovereignty; intervention, the state and symbolic exchange (Cambridge: Cambridge University Press, 1995). 35 Part of this revival of Foucauldian analysis is also a debate about the limits of key terms such as power/knowledge and discourse (see David Lewis, ‘The myopic Foucauldian gaze: discourse, knowledge and the authoritarian peace’, Journal of Intervention and Statebuilding 11: 1, 2017, pp. 21–41). 36 Merlingen and Ostrauskaite, ‘Power/knowledge in international peacebuilding’. 37 Andreu Solà-Martín, ‘Lessons from MINURSO: a contribution to new thinking’, International Peacekeeping 13: 3, 2006, pp. 366–80. 38 Barry Ryan, Statebuilding and police reform: the freedom of security (London: Routledge, 2011). 39 Ryan, Statebuilding, p. 52. 40 Kyle Grayson, ‘Human security as power/knowledge: the biopolitics of a definitional debate’, Cambridge Review of International Affairs 21: 3, 2008, pp. 383–401. 41 Foucault, Power/knowledge, p. 142. 42 Because the situation in Burundi was still precarious at the time of writing, we have accepted the interviewees' wish to remain anonymous. 43 It should be stressed that although we focus on the role of the UN, we do not wish to suggest a homogeneous idea of ‘interveners’. Indeed, in the case of Burundi in 2015–16, the AU, the EAC, individual states and others also played important roles. Had this preventive endeavour been a response to a precarious situation in a different country, the configuration of intervening actors would probably have taken a different form. 44 For more historical background, see e.g. Walter Lotze and Alexandra Martins, ‘The responsibility to prevent atrocity crimes’, in Serena K. Sharma and Jennifer M. Welsh, eds, The responsibility to prevent: overcoming the challenges of atrocity prevention (Oxford: Oxford University Press, 2015), pp. 250–79. See also René Lemarchand, Burundi: ethnic conflict and genocide (New York: Cambridge University Press, 1996). 45 Since the situation on the ground has not been settled at the time of writing, we do not attempt to evaluate the events or make predictions about their outcome(s). 46 BBC Afrique, ‘Le CNDD-FDD investit Nkurunziza’, 25 April 2015, http://www.bbc.com/afrique/region/2015/04/150425_burundi-politics. 47 Cour Constitutionelle de Burundi, ‘RCCB 303’, 5 May 2015. 48 BBC News, ‘Burundi court backs President Nkurunziza on third-term’, 5 May 2015, http://www.bbc.com/news/world-africa-32588658. 49 See e.g. ‘Burundi: nouvelles attaques à la grenade à une semaine des élections’, Africantime.com, 23 June 2015, http://fr.africatime.com/burundi/db/burundi-nouvelles-attaques-la-grenade-une-semaine-des-elections. 50 Goran Tomasevic, ‘Burundi's president urges end to protest, coup leader at large’, Reuters, 15 May 2015, http://www.reuters.com/article/us-burundi-politics-idUSKBN0NY0O020150515. 51 Commission Electorale Nationale Independante, ‘Préesentation des résultats provisoires de l'election présidentielle 2015’, 31 July 2015, http://www.ceniburundi.bi/presentation-des-resultats-provisoires-de-l-election-presidentielle-2015. 52 BBC News, ‘Burundi crisis: military bases attacked in Bujumbura’, 11 Dec. 2015, http://www.bbc.com/news/world-africa-35070154. 53 According the BBC, government sources stated that a total of 87 people were killed that day, but it was unclear whether the 34 persons mentioned here were included: BBC News, ‘Burundi crisis: army says 87 killed in day of violence’, 12 Dec. 2015, http://www.bbc.com/news/world-africa-35083823. 54 For example, during our visit in Burundi, Hafsa Mossi was shot by unknown gunmen. See AFP, ‘Former Burundi minister Hafsa Mossi shot dead’, MailOnline, 13 July 2013, http://www.dailymail.co.uk/wires/afp/article-3688204/Former-Burundi-minister-Hafsa-Mossi-shot-dead.html. 55 On 18 Dec. 2016, the UN High Commissioner for Refugees reported 327,576 refugees from the current situation. See Refugees Operational Portal, ‘Refugees from Burundi: current situation’, 23 March 2017, https://data2.unhcr.org/en/situations/burundi. 56 Authors' interview, New York, 5 Oct. 2016. 57 Benomar was Special Adviser from Nov. 2015: Security Council Report, ‘Burundi’, http://www.securitycouncilreport.org/burundi/. 58 Authors' interviews, Bujumbura, 14 July 2016. 59 Authors'; interview, Bujumbura, 13–14 July 2016. 60 Secretary-General's remarks to the Security Council open debate on ‘Maintenance of international peace and security: conflict prevention and sustaining peace’ (as delivered), 10 Jan. 2017. 61 Following the coup attempt, US officials, for example, warned of ethnic tensions. See Nahal Toosi, ‘US fears Burundi on cusp of ethnic violence’, Politico, 18 April 2015. 62 UN, ‘Statement by Adama Dieng, Special Adviser of the Secretary-General on the Prevention of Genocide on mission to Burundi’, press release, 30 May 2015, http://www.un.org/en/preventgenocide/adviser/pdf/2015-05-30.SAPG%20statement%20on%20the%20sitations%20in%20Burundi.pdf. 63 Report of the UNIIB established pursuant to HRC Resolution S-24/1, HRC document A/HRC/33/37, 20 Sept. 2016. 64 Among our interviewees, Russia and China were seen as opposing a strengthened UN engagement in Burundi, while France and the US were seen as pushing for a greater international presence. 65 Authors' interviews, Bujumbura, 13 July 2016. 66 Authors' interviews, Bujumbura, 13 July 2016. 67 Conseil National pour la Défense de la Démocratie–Forces pour la Défense de la Démocratie (CNDD-FDD), communiqué no. 013/2016, 26 July 2016, http://cndd-fdd.org/2016/08/03/communique-no-0132016-du-parti-cndd-fdd-du-26-juillet-2016/. 68 Radio Publique Africaine, ‘Procès-verbal de l'assemblée générale de l'association Radio Publique Africaine en sigle “R.D.A”’. 7 Jan. 2000. 69 US Department of the Treasury, ‘Treasury sanctions four Burundian individuals’, press release, 18 Dec. 2015, https://www.treasury.gov/press-center/press-releases/Pages/jl0310.aspx. 70 Authors' interviews, Bujumbura, 14 July 2016. 71 Gouvernement du Burundi, Déclaration du gouvernement burundais sur la collaboration et la coopération avec l'Office du Haut Commissariat des Droits de l'Homme au Burundi, 11 Oct. 2016, http://www.burundi.gov.bi/spip.php?article1538. 72 Two previous envoys had been forced out during 2015. 73 FIDH, Repression and genocidal dynamics in Burundi, Nov. 2016, https://www.fidh.org/IMG/pdf/burundi_report_english-2.pdf. 74 Authors' interviews, New York, 4 Oct. 2016. For voting records on the resolutions, see A/C.3/70/L.35, http://www.un.org/en/ga/third/70/docs/voting_sheets/L.35.pdf (for 2015) and A/C.3/69/L.28/Rev.1∗, http://www.un.org/en/ga/third/69/docs/voting_sheets/L.28.Rev.1.pdf (for 2014). 75 Authors' interview, New York, 4 Oct. 2016. 76 UNSC Resolution 2303, 29 July 2016, http://www.un.org/en/ga/search/view_doc.asp?symbol=S/RES/2303(2016). 77 The UNSC expressed deep concern about the lack of implementation in its Presidential Statement of 2 August 2017, https://www.un.org/en/ga/search/view_doc.asp?symbol=S/PRST/2017/13. 78 Peace and Security Council of the AU, ‘Communiqué of the 565th meeting of the PSC on the situation in Burundi’, 17 Dec. 2015, http://www.peaceau.org/en/article/communique-of-the-565th-meeting-of-the-psc-on-the-situation-in-burundi. 79 Foucault, Power/knowledge, p. 51. 80 International Crisis Group, Seizing the moment: from early warning to early action, Special Report no. 2/Global, 22 June 2016, https://www.crisisgroup.org/global/seizing-moment-early-warning-early-action. 81 The phrase ‘early aggression’ came up in the authors' interview with a Burundian official, Bujumbura, 14 July 2016. 82 Authors' interview, Bujumbura, 15 July 2016. 83 HIPPO Report, p. vii. 84 UNSC, 7830th meeting, 9 Dec. 2016, http://www.un.org/en/ga/search/view_doc.asp?symbol=S/PV.7830. © The Author(s) 2018. Published by Oxford University Press on behalf of The Royal Institute of International Affairs. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact email@example.com
International Affairs – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera