Introduction

The common methodological shortcomings, i.e. the empiricist fallacy and methodological individualism, have generated two theoretical problems in the existing research on the subject that this book aims to address. The first problem can be encountered in a tendency to reduction of explanation. The empiricist fallacy follows the political realist worldview in that the usual suspects, namely ‘those betraying great powers’,12 are indicted, in the manner of a typical Hobbesian image of world politics. This assessment would not be problematic if the method of falsification was used, eliminating other possible or alternative causes of the UN’s collapse in Rwanda and leaving the lack of political will on the part of member states as the only plausible explanation. Such method has not been applied. Subsequent records take institutional factors into account, but nevertheless tend to restrict argumentation to the actions and omissions of individual UN officials. Thus, the explanatory factors of the failure are reduced to the behaviour of either states or of individual UN officials, with no considerable (read: ‘causally effective’) role given to the wider socio-historical context within which these behaviours were situated. This book views a contextual analysis of the early 1990s as an essential step in understanding the Rwandan case. No evidence has been produced to counter the presumption that both the causes of the UN’s breakdown in Rwanda and the potential to avert the disaster lay in the socio-historical factors of UN conflict management specific to that time. This was the era during which the Security Council, in co-operation with the Secretariat, demonstrated an unprecedented willingness to monitor international conflicts with the aim of preventing their outbreak in the first place. The early 1990s saw a fleeting moment of optimism in the Council’s long history of impasse: it was an interregnum between the Cold War paralysis, caused by the bipolarity of world politics in the 1980s, and the late 1990s paralysis, caused by the crushing disappointments of UN peacekeeping in the early and mid-1990s.13 During that brief period of optimism, the Council invoked mechanisms by which to control international conflicts, the term ‘mechanism’ here meaning ‘social control’ by monitoring institutions. According to this definition, as Colin Wight describes, ‘A mechanism, even a social mechanism, is a process or technique for achieving a desired end state or outcome’ (Wight 2004: 288). In the case of an early warning mechanism, for example, the desired outcome was the early detection of potential genocides. In the Security Council, such control mechanisms usually emerge through interaction between member states and the Secretariat. One example of this type of interaction took place in January 1992, when the President of the Council asked the Secretary-General to come up with recommendations for strengthening the conflict management capacity of the UN, and Secretary-General Boutros Boutros-Ghali responded to this by writing An Agenda for Peace. The Agenda, published in May 1992, envisaged the establishment of early warning mechanisms, which were supposed to move the focus of UN conflict management from previous reactive and ex post facto measures to more proactive and preventive strategies (Boutros-Ghali 1999: 23-6). It is essential to appreciate that the early warning mechanism was invoked by the Secretary-General only two years prior

to the genocide, and ask how it is possible that this mechanism failed so disastrously at the moment when it was so desperately needed. Reductive explanation underlies the most serious ‘blind spot’ in existing accounts, namely the lack of emancipatory vision. Such a vision would inform how agents could have liberated themselves from unwanted and suppressive constraints, such as those structural factors of the UN system which impeded the ability of individual diplomats and officers to raise concerns regarding Rwanda. However, most of the studies are underpinned by the political realist assumption that the Security Council could not have risen above the constraints set by the selfish great powers of the Council.14 Other accounts view the Rwandan case through a neo-liberal prism which portrays UN institutions as merely the instruments or powerless servants of states. These studies therefore argue that, irrespective of whether the Secretary-General could have recognised the genocide at an early stage and lobbied for intervention, such efforts would have made little difference vis-à-vis the member states of the Council once they had made up their minds to freeze all humanitarian interventions, following their failure in Somalia in 1993.15 Hence, the Hobbesian and neo-liberal theoretical underpinnings effectively nullify the possibility of finding a means of emancipation in the UN’s conduct of the genocide: the Organisation is viewed either as a victim of the ‘irresistible forces’ of great powers or as a powerless servant of states. It is argued in this book that an exploration of mechanisms might be the key to providing an emancipatory vision. An Agenda for Peace proves that there was a meaningful effort underway in the Security Council to monitor international conflicts at the beginning of the 1990s. If such monitoring mechanisms had worked properly, the Rwandan genocide might have been prevented. The final chapter of this book will take this emancipatory vision even further; it will examine the hypothetical, or counterfactual, impact of post-2005 conflict management mechanisms on the prevention of the Rwandan genocide. It will be claimed that the UN peacebuilding architecture and the mainstreaming of the responsibility to protect, both of which were initiated at the 2005 UN World Summit, could have played a significant role in preventing the Rwandan genocide, if they had already been in operation in the early 1990s. Widening the ontological horizon of research from states to control mechanisms yields a more complex view of the UN and the Security Council, but, as a reward, it may reveal unrealised possibilities of emancipation inherent in those mechanisms. It could thus be said that freedom lies within complexity. However, it is equally important to expose those dysfunctions and constraints that hampered the working of control mechanisms. These possible defects may, in turn, encompass not only the lack of political will on the part of member states and institutional malfunctions, but also factors such as human limitations in predicting events in complex civil wars; cognitive dissonance; the rigidity of bureaucratic procedures; the loose coupling of modern organisations; the ensuing fragmentation of worldview; and the reproduction of stereotypical images of African conflicts. Any research question usually hinges on the philosophical foundation of research, which in previous studies of Rwanda has been positivism, the domi-

nant epistemology in IR theory. Although social scientists tend not to explicitly state their adherence to positivism, they often implicitly base their propositions upon positivist assumptions, namely the belief that facts are theory-neutral and phenomenalism, i.e. reliance on observed events.16 This is also the case with previous investigations of the Rwandan crisis, which generally prefer simple observation of events to any particular theory as a means of illuminating why the UN failed. Hence, existing accounts either do not define any theory or research question at all, or pose a straightforward question implicitly, along the lines of: ‘Why did the Security Council not stop the genocide?’ Such an approach resembles in many respects a Sherlock Holmes type of detective story, as it strives to investigate meticulously and thoroughly all events and persons involved in the UN debacle and to scrutinise all pieces of information surrounding the case. This common method typically results in accurate chronological reportages, which produce evidence of the ‘crime of indifference’ committed by certain members of the Council. A further feature of positivism is atomism, which means looking for the smallest observable units that cannot be broken down any further, such as individuals in society (Smith 1998: 76). In this manner, the objective of researchers is to find those individual culprits inside the UN system whose actions and omissions led to a betrayal of Rwanda. However, quantity possesses no quality of its own: the power of pure information-gathering or fact-finding cannot provide a substitute for the power of theoretical thinking, simply because all facts are theory-laden to some degree. A theory is a set of concepts used to explain some phenomenon (Silverman 1998: 103), so taking stock of phenomena as such is not sufficient to explain anything. Explanation requires theory. As most of the previous studies tend to disregard or undervalue theory and, instead, rush to describe the events surrounding the Rwandan shambles, they often imply a simplified explanation in which individual ‘betrayers’ and ‘betraying’ are taken as given facts. This study invokes theory in order to open up all possible causes of the malfunctioning of the UN without taking any apparent causes, such as betraying states, for granted, on the grounds that there might also be other ‘troublemaking’ factors in the UN system that merit attention. The question can therefore be formulated as follows: ‘How is it possible that the Security Council failed to prevent the Rwandan genocide?’17 Such an approach resembles not so much detective work as criminal psychology, in which it is not sufficient to ask, ‘Why did actor X commit the crime?’, but is also necessary to enquire, ‘How is it possible that actor X could commit the crime?’ Sometimes the motive for a crime is easier to solve than to trace the factors that enabled its perpetrator(s) to commit it in the first place. The motives may be rooted in greed, incitement, or self-defence, but a different question is how an ordinary person could possibly commit a cold-blooded murder without feeling a bit of sympathy, or witness a brutal crime without making a sound to stop or prevent it. The latter question widens the circle of investigation from the motives of an actor to the circumstances and context of the crime and to the ‘nature’ or ‘state of mind’ of the agent. In the same way, the motive of the Security Council in

allowing Rwandans to die is quite obvious: it was the political refusal of member states to put their soldiers in harm’s way to ‘save strangers’.18 However, a much wider question to ask is how it was possible for ordinary UN diplomats and bureaucrats to disregard the most efficient mass killing in history. What are the ethics of the Security Council such that it could continue with ‘business-asusual’ while a genocide was happening? It would be understandable if the alarm bells failed to ring when an ‘ordinary routine abuse of human rights’ was happening, but it is much harder to comprehend how the Council could be so indifferent when a ‘supreme humanitarian emergency’19 was occurring. In sum, what structural conditions enable decent actors to commit such horrendous acts year after year, decade after decade, and persevere uninterrupted by devastating failures, while faces change in bureaucracies? Whilst the motive behind the UN’s failure obviously lay in member states (which answers the Why-question), an understanding of the factors that enabled the UN to act with such indifference requires taking into consideration not only states, but also its institutions and mechanisms (addressing the How-question).20 The most obvious alternative to the empiricist approach of positivism is critical realism, because the latter initially emerged as a critique of positivism in the philosophy of science (see, for example, Bhaskar 1997). Thus, critical realism must be strictly separated from political realism encompassing classical and neorealism in IR theory. While the latter contains substantial – and often fallible – axioms concerning the universal nature of human beings and the state-system, critical realism allows a socio-historical case study to reveal its particular nature. As opposed to positivism, critical realism deems that explanation based on events, phenomena and appearances is unsatisfactory. One should always go beyond events to explain and understand the structures and mechanisms that produced them (Collier 1994: 49). What this means in terms of the Rwandan case is that one should not restrict analysis to the sequence of events surrounding the tragedy. Instead, one should explore the ‘troublemaking’ structures and mechanisms operative within the UN system at the beginning of the 1990s. The crux of the critical realist critique is directed against the empiricist fallacy of positivism, which erroneously assumes that a constantly occurring correlation between events A and B implies that the former is the cause of the latter in every instance (Bhaskar 1997: 34). In existing accounts, A refers to the selfishness of great powers while B denotes the UN’s failure, and they are seen as bound together in the constantly occurring paralysis of the Security Council. Critical realism, however, deems that real causes cannot be found in regularities but in particular socio-historical contexts. Selfish great powers do not always constitute the sole cause of the malfunctioning of the UN, though this was certainly often the case during the Cold War. There may be other troublemaking factors within the UN system which also merit attention. Reductionism tends to limit its view to a single explanatory factor emphasised by a particular theory. This reduction, in turn, leads to an overall fragmentation of explanation. By emphasising the material powers of states, most of the studies characterised by political realism tend to disregard or underestimate

the role that UN institutions and the Secretariat could have played in lobbying for intervention in Rwanda (see, for example, Feil 1998: 4; Kuperman 2001: viii). Certain investigations underpinned by neo-liberal assumptions do take UN institutions into consideration but fail to conceive of how the Secretariat could have made a difference in rescuing Rwandans vis-à-vis states. Critical realism can serve as an ‘umbrella meta-theory’ or ‘safety net’ to prevent the occurrence of such reductionism, and therefore has the potential to provide a consistent, balanced and holistic view of the causes of the Rwandan drama. This expansion of worldview, in turn, paves the way for an emancipatory vision: if the efforts of individual states and UN officials would have been insufficient to save Rwandans, could states, institutions and mechanisms have made a difference by working together? Although IR research boasts of being inter-disciplinary in nature, compartmentalisation is still prevalent in the discipline and only a few IR approaches actually succeed in supplying a consistent, multi-theoretical methodology to a research topic. Critical realism can respond to that challenge. Martin Hollis and Steve Smith argue that IR theory and social science in general have been characterised by a tension between two methodological positions. The ‘explaining’ position (Erklären) views the social world as composed of external structures, whereas the ‘understanding’ viewpoint (Verstehen) relies on hermeneutics and historical methods to explicate meanings and collective rules (Hollis and Smith 1990: 4-5; 1991: 408-9). Despite an impressive range of studies on Rwanda, all previous books on the subject, except for Michael Barnett’s study Eyewitness to a Genocide (2002), have provided an explanatory approach. They have meticulously explained on whose desks and on which dates reports of genocide landed, and how desk officers responded to, or, in most cases, ignored these crucial pieces of information. But most of the analyses have not gone further to try to understand those reasons and mechanisms inside experts’ minds, in inter-subjective rules and in organisational conventions that hampered the processing of such information. Critical realism deems this explanation-driven ‘story-telling’ insufficient. Understanding is also necessary in any piece of research simply because reasons are causes (Patomäki 1996a: 108). The reasons why UN officials saw the genocide stereotypically as ‘just another instance of mad tribal violence in Africa’ and why it did not deserve attention within the UN bureaucracy were causes of the UN’s subsequent withdrawal from Rwanda. Critical realism and hermeneutics are similar in the sense that they both ask the post-Kantian question of what the necessary preconditions are for knowledge to be possible (Patomäki 2002a: 9). However, critical realism and hermeneutics have different interests in knowledge in the sense that the latter views the communication of knowledge as an intriguing research object, whereas critical realism strives to take a step further and explore the production of knowledge. Critical realism draws upon the work of the Frankfurt School by emphasising that knowledge is produced in order to reinforce power structures in society and to serve hidden interests, which function through certain mechanisms. Communication and language games are set in motion and regulated by these underlying mechanisms. Thus, going beyond communication to explore them is more

revealing than researching the language games as such. What this means in terms of the Rwandan case is that the hermeneutical approach would concentrate on the Security Council’s language games through which Rwanda was excluded from the concerns and sympathy of the international community, whereas critical realism would go further and aim to expose those mechanisms that started this draconian language game in the first place and kept it in motion throughout the genocide. So what are the fundamental principles constitutive of the Lakatosian ‘hard core’ of critical realist philosophy (Lakatos 1978: 4)? The term ‘critical realism’ emerged from the elision of two main philosophical principles, namely scientific realism and critical naturalism (Collier 1994: x-xi). As Copernicus removed the human being from the self-imposed centre-stage of natural sciences, critical realism strives to launch a similar Copernican revolution in the social sciences. Scientific realism views research based on individual human actions as insufficient and hence seeks to examine the structures and mechanisms that underlie and affect actors’ decisions (Bhaskar 1997: 19). What this means in terms of research on the cases of Rwanda and Darfur is that less attention should be paid to the omissions and failures of individual UN officials. Instead, greater emphasis should be placed on the dysfunctions of the wider structural framework within which those individuals were situated, such as a bureaucratic culture breeding indifference. This non-anthropocentric approach has a smaller chance of allocating individual responsibilities and indicting those individual UN diplomats and officials who should be blamed for the debacle. Instead, the scientific realist approach possesses a greater potential of revealing ‘troublemaking’ factors in the structures and mechanisms of the UN. Critical naturalism brings an important counter-balance to the naturalscientific sound of the terms ‘mechanisms’ and ‘structures’ by claiming that social objects, unlike natural ones, are concept-and activity-dependent and do not uphold the distinction between fact and value (Bhaskar 1998b: 38). Hence, a social scientist should always maintain a critical stance towards research objects and never take speeches, statements and narratives as ‘objective truths’, because they usually serve particular political purposes. The significance of this to understanding the Rwandan drama is that one should critically evaluate certain statements made by UN officials, according to which the UN’s failure had nothing to do with its knowledge-production processes in general, or early warning in particular, but was caused by the political unwillingness of member states. Kofi Annan, the Under-Secretary-General for Peacekeeping at the time, formulated this claim as follows: ‘If there was a problem, it was not one of information or intelligence. The problem was lack of political will.’21 Similarly, the then Secretary-General Boutros-Ghali asserted to the author that,

Member states were opposed to intervention in Rwanda, with early warning and without early warning. So the real problem is this: if there is no political will among the major actors in the Security Council, any [UN] system which we try to improve will be useless.22