When the Dead are Not Silent: The Investigation of Cultural Perspectives Concerning Improper Burials in Northern Uganda A Thesis Presented for the Master of Arts Degree The University of Tennessee, Knoxville Adrianne S Kembel August 2015 ii Copyright (©) Adrianne S Kembel All rights reserved. iii Acknowledgements I would like to thank my family for their continuous love and support throughout my entire academic career. I could not have achieved so much without them. I also owe a great deal of gratitude to my friends and colleagues, especially Jaymelee and Kristen, who have been sound voices of reason over the years. I would like to give a special thanks to Cheryl Heykoop for providing invaluable counsel and companionship while I was collecting data in Uganda. Thanks to my committee for their instrumental suggestions and feedback. I would also like to give additional recognition to Trish, for her generous support, guidance, and above all patience. I could not have asked for a better advisor. Lastly, I would like to thank the Refugee Law Project, the staff at the National Memory and Peace Documentation Centre, and especially the participants whose willingness to share their thoughts, opinions, advice, and time made this work possible. iv Abstract This thesis presents the findings of a qualitative examination of the effects of improper burials and the associated cultural impacts on the Acholi population of northern Uganda. Since independence in 1962 Uganda has experienced several internal conflicts, including the notorious struggle between the Lord’s Resistance Army and the Ugandan government. This conflict, which disproportionately affected the Acholi ethnic group, resulted in tens of thousands of deaths and culturally inappropriate burials. These burials are particularity problematic because the Acholi maintain continual ties to the dead through ancestor veneration with proper burial being one of the most important conciliatory rites. In the negotiation between the living and the dead, the social infrastructure that is built around and on mortuary practices is part of the active creation of social relationships and ways of being. Improper burials are, therefore, an obstacle to post-war recovery efforts, as these interments violate beliefs within the Acholi religious system as to how the dead should be treated. As a consequence of this mistreatment these spirits, known as cen, become angry and cause disease, death, possession, nightmares, and other misfortunes. This research utilizes a psychosocial framework based in grounded theory to explore the impact of improper burials within Acholi communities, the cultural consequences of these interments, local perceptions of what should be done about this issue, and how such endeavors should be accomplished. Additionally, this study also assessed the appropriateness of forensic excavation and DNA identification as a potential solution to improper burials. Though the findings revealed that improper burials and the associated cen are a significant problem, the desired approach for addressing this issue depends upon whether communities practices tradition Acholi cosmology or Christianity. Based on these results I argue that religion and cosmology are vital coping strategies for addressing cen on local and institutional levels within northern Uganda. Furthermore, future psychosocial interventions aimed at post-conflict recovery would benefit from a more comprehensive understanding and inclusion of religious beliefs and their effects on local perceptions and needs. v Table of Contents CHAPTER ONE: INTRODUCTION…………………………………………………………………………..1 Improper Burial and Cen in Northern Uganda……………………………………………..1 Situating the Research within the Literature………………………………………………2 Research Questions and Summary of Findings………………………………………….11 Thesis Layout………………………………………………………………………………………….16 CHAPTER TWO: HISTORY OF THE NORTHERN UGANDA REGION………………………..18 The Importance of A Contextualized History…………………………………………….18 Uganda as a British Protectorate………………………………………………………………18 Political Movements Since Independence…………………………………………………22 Ontological Insecurity and Social Anxiety…………………………………………………25 The Government of Uganda and the LRA…………………………………………………..29 International Influences…………………………………………………………………………..33 National Development Efforts…………………………………………………………………..36 CHAPTER THREE: RELIGIOUS AND COSMOLOGICAL HISTORY OF THE REGION…..38 Understanding Cen within Religious Discourses……………………………………….38 Luo Cosmology………………………………………………………………………………………..38 The Spread of Christianity………………………………………………………………………..41 CHAPTER FOUR: METHODS………………………………………………………………………………..46 Laying the Foundation……………………………………………………………………………..46 Participant Descriptions…………………………………………………………………………..47 Research Locations………………………………………………………………………………….48 Gaining Access to Participants………………………………………………………………….52 vi The Research Scheme………………………………………………………………………………55 Locating the Researcher…………………………………………………………………………..57 Developing Interview Guides…………………………………………………………………...58 Informed Consent……………………………………………………………………………………63 CHAPTER FIVE: ANALYSIS………………………………………………………………………………….66 Qualitative Coding…………………………………………………………………………………...66 Methodology…………………………………………………………………………………………...67 First Cycle Coding……………………………………………………………………………………70 Second Cycle Coding………………………………………………………………………………..72 CHAPTER SIX: FINDINGS AND DISCUSSION………………………………………………………...76 Understanding Proper Burials………………………………………………………………….76 Improper Burials as Moral Violations……………………………………………………….78 The Multifaceted Notion of “Improper”…………………..………………………………...81 Traditional Acholi Cosmology…………………………………………………………………..82 Christian Denominations…………………………………………………………………………87 Necropolitics…………………………………………………………………………………………...90 Forensic Exhumations and DNA Identification………………………………………….93 Capacity within Communities…………………………………………………………………..99 Necropolitics as an Economic Tool…………………………………………………………100 Prioritizing Cen in Everyday Life…………………………………………………………….102 Research Limitations……………………………………………………………………………..105 CHAPTER SEVEN: CONCLUSION……………………………………………………………………….108 REFERENCES CITED…………………………………………………………………………………………111 vii APPENDICES……………………………………………………………………………………………………121 Vita………………………………………………………………………………………………………………….132 viii List of Figures Figure 1. Two household Abilas………………………………………………………………………..…40 Figure 2. Village Abila………………………………………………………………………..………………..40 Figure 3. Map of districts in Uganda…………………………………………………………………….49 Figure 4. Image of Kitgum District and its constituent sub-counties…………………….50 Figure 5. Map of research locations…………………………………………………………………….51 Figure 6. Government hierarchy and associated titles……………………………………….53 Figure 7. Screenshot of data and coding strips within NVivo 10………………...………...68 Figure 8. Chart displaying the evolution of participant statements to themes………..77 Figure 9. Locating improper burials and cen amongst other problems………………..107 Figure 10. Sacrifice word tree…………………………………………………………………………132 1 Chapter One Introduction Improper Burial and Cen in Northern Uganda This thesis examines the effects of improper burials from numerous conflicts and the associated cultural impacts and implications for the Acholi population of northern Uganda. Since receiving independence from Britain in 1962, Uganda has experienced several rebel movements and dramatic regime changes. The most recent and protracted conflict, from 1986-2006, involved the Lord’s Resistance Army (LRA) and the Ugandan government. During this twenty-year conflict, the LRA inflicted brutal acts of violence against civilians, including murder, mutilation, intimidation, rape, and the kidnapping of children and adolescents to be trained as child soldiers, while the government’s army frequently failed to adequately protect civilians (Branch 2010; Villa-Vicencio et al. 2005). In addition, both the LRA and government soldiers participated in looting and the destruction of non-combatant property. Though this conflict touched all of northern Uganda (West Nile, Acholi, Teso, and Lango regions), the Acholi ethnic group within Acholiland was the most heavily affected (RLP 2004). It is estimated that tens of thousands were killed and, due to insecurity, were left unburied, hastily buried, and/or buried in mass graves (JRP 2007a, 2009, 2010; Proctor 2013; OHCHR 2007). The presence of these remains in what is locally considered to be improper burials is proving an obstacle to post-war recovery in the region (Proctor 2013). This problem arises because Acholi in northern Uganda maintain strong and continual ties between the living and the dead through kinship reckoning and ancestor veneration (Harlacher and 2 CARITAS 2006; P’Bitek 1971). The dead play an active role in people’s lives and there is interest in how remains and memorials are treated (Petrig 2009: 342). For the Acholi this relationship is sustained through ritualistic maintenance, with a proper burial being one of the most important conciliatory rites that occur between the living and the dead (Proctor 2013). In the negotiation between the living and the deceased the social infrastructure that is built around and on death and mortuary practices is part of the active creation of social relationships and ways of being (Crossland 2000: 147). In general, proper funerals or a culturally accepted equivalent can serve to fulfill religious and communal obligations to the dead, strengthen community ties, encourage personal closure (Stover et al. 2002; Stover and Shigekane 2003: 860), and promote cosmological balance (Baines 2010; Eppel 2007, 2014). As such, Acholi communities feel that the failure to conduct proper funeral rites and fulfill obligations to the dead will result in serious consequences (Amone-P’Olak 2007). This belief renders the existence of family or community members who have been improperly buried in one of the many mass or individual graves particularly problematic for the psychological and psychosocial health of recovering individuals and communities (JRP 2007b, 2007c). Situating the Research within the Literature Drawing from several sources in the existing literature (Baines 2010; Neuner et al. 2012; Proctor 2013), psychosocial health is defined in this thesis as the summed interaction of individual emotions, behavior and mental health, community 3 relations, economic development, as well as traditions and culture, all of which may be affected by the presence of improper depositions. In order to include and, in fact, emphasize the relationship between psychological aspects of individual experience and wider social factors a psychosocial framework was used in this research. There is a broad spectrum of approaches that are encompassed under the umbrella of this empirical framework but general goals include restoring, maintaining, or enhancing personal and social capacity through the mobilizations of local strengths and coping strategies, needs assessments, linking people or communities to resources, and addressing environmental stress (Woods and Hollis 1999). Additionally, psychosocial approaches may pertain to the investigation of any of these aspects. Though this framework was originally used for clinical social work it has expanded to cover a wide variety of topics including forced migration (Ahearn 2000), post- disaster/post-conflict well-being (World Health Organization 2005), social reintegration (Betancourt et al. 2009), mental health (Miller and Rasmussen 2010), and social reconstruction from violence, trauma, or loss sustained in post- disaster/conflict situations (Michael and Monteiro 2001). Many of these topics are also components of transitional justice (TJ) which in the years following the 2006 ceasefire in Northern Uganda, became the dominating post-conflict framework through which international, national, and local actors have attempted to promote peacebuilding, development, and to address human rights abuses within Uganda. This top-down and bottom-up post-conflict framework, which has previously been implemented by a number of countries, includes both judicial and non-judicial actions. As the issues that can be addressed 4 through the application of transitional justice measures expands, this approach is being applied to an ever-growing number of conflicts in countries all over the globe (Hinton 2010; Teitel 2000). Currently, there are a wide array of TJ mechanisms in existence including international, national, local or hybrid criminal prosecution, truth commissions, amnesty acts, reparations, institutional reforms, memorialization, and reconciliation (International Center for Transitional Justice 2015). The last component is of particular importance to this project because it ultimately affects issues relating to the dead. Broadly stated reconciliation includes the effort to address past grievances and associated contemporary issues. For example, within northern Uganda one issue that could be categorized under the purview of this transitional justice mechanism is the relationship between the living and the dead. As such, using a psychosocial framework to tackle the issue of improper depositions and their consequences aligns directly with transitional justice and has the potential to influence post-conflict mechanisms, interventions, and development being implemented by the nation-state. However, despite the potential for TJ to bring about positive change, this approach is not without flaws. A growing number of critiques point out that many of the founding ideologies that serve as the core of TJ strategies are based on Western or external notions of what is needed in post-conflict situations and how these should be achieved (Hinton 2010; JRP 2007b; Shaw 2005). The “tendency to exclude local communities as active participants in transitional justice measures is a primary flaw” and raises pressing questions about whose interests are being served (Lundy and McGovern 5 2008: 266). Perhaps not surprisingly, researchers have found that if communities are not consulted and involved as prominent stakeholders, any action taken becomes disconnected from the needs and perspectives of those affected by conflict and may have little positive impact (Aronson 2011; JRP 2011, 2007b; OHCHR 2007). Top-down processes that are produced by elites and institutions are often too far removed from on the ground realities (Robins 2011: 75). Additionally, researchers have suggested that if TJ efforts are to produce sustainable, long-term results it is important to document the opinions, perceptions, and expectations of the different stakeholders continually throughout the project, from beginning to end (Aronson 2011; Arriaza and Arriaza 2008; Doretti and Burrell 2007; Eppel 2006; Fonderbrider 2002; Quinn 2007). If such an endeavor is not undertaken and external discourse is privileged, it further disempowers those with the greatest need and can potentially perpetrate a neocolonial cycle of violence (Robins 2011: 78). Furthermore, as every conflict has unique contexts and circumstances, effective intervention measures need to be firmly grounded in the lived realities of survivors and account for their perceptions and needs (Aronson 2011; Doretti and Burrell 2007, Eppel 2006; JRP 2011, 2007b; Quinn 2007; OHCHR, 2007). A psychosocial model is appropriate for this research because it recognizes that situations, communities, and individuals present with a unique constellation of factors that influence their needs and perceptions and, therefore, advocates for community centered methods that recognize local practices and the agency of various actors. Examples of how this approach has been used in northern Uganda include issues pertaining to displacement (Bolton et al. 2007), demobilization and 6 reintegration of former child-soldiers (McKay and Mazurana 2004; Annan et al. 2008; Chrobok and Akutu 2008; Amone-P’Olak 2007), general reintegration (Liu Institute for Global Issues 2005, JRP 2008c), perceptions of accountability and reconciliation (Office of the United Nations High Commissioner for Human Rights 2007), reparations and rehabilitation (JRP-IRJ 2011b, JRP 2012a), truth-telling (JRP 2007b; JRP-IRJ 2011a), and peace building (JRP 2009). Many of these works state that matters pertaining to the dead and improper burial are sociocultural components of broader issues such as reconciliation and positive peace building. According to Acholi cosmology, the spirits of the murdered and those who are not given a proper burial become angry and cause harm to the living (JRP 2006b). This “ghostly vengeance” or cen can cause disease, sickness, death, infertility, possession, nightmares, poor crops, and bad luck (JRP 2006b, 2008a; P’Bitek 1971). In addition, transgressions against moral order known as kiir (“abomination”), which is a phenomenon related to cen, will cause misfortune to befall the children of both the victims and the perpetrators until the situation is remedied or a ceremony is performed to cleanse the area. Examples of kiir often pertain to the desecration of the dead, including burying several bodies together in a mass grave or walking over human remains remaining on the grounds surface (JRP 2007a). The consequences of these violations are attributed to a disruption in the close and continual relationship the Acholi believe exists between the living and the dead (Baines 2010; JRP 2006b; P’Bitek 1971). Cen and the disturbances brought by these vengeful spirits can also be seen as a form of social suffering. 7 The concept of social suffering recognizes that pain, trauma, injury, and loss are rooted in and experienced through past and present sociocultural conditions that dictate how distress is experienced and expressed (Wilkinson 2005; Kleinman et al. 1997). In consequence “social worlds are inscribed upon the embodied experience of pain and there are many occasions where an individual’s suffering should be taken as a manifestation of social structural oppression and/or collective experience of cultural trauma” (Wilkinson academia.edu: 1; also see Wilkinson 2005). Disturbances then “give expression to wider social anxieties and threats experienced by a community” (Baines 2010: 419). Put another way, individual and social apprehension to moral transgressions (improper depositions), is a form of social suffering among Acholi populations that is expressed in the phenomenon of cen and embodied though possessions, haunting, nightmares and other disturbances. Physical, psychological, moral, spiritual, social, economic and political dimensions of disturbances are intricately interwoven (Kleinman et al. 1997; Victoria 2011). Given these interconnections, if we are to better understand how cen is experienced it is imperative to explore how social, economic, political, and institutional powers entwine with ghostly vengeance and how communities react to this occurrence. One approach to this endeavor is to ask the target population how they address the problem of cen, or more broadly, to engage in the study of culturally specific coping mechanisms. There are three types of improper burial considered in this study: individual graves, mass graves, and those who are presumed dead but whose physical location 8 is unknown.1 Improper individual graves that may lead to cen include incidences where the family knows where their loved one is interred but the body’s current location is either temporary or culturally inappropriate, or where the grave’s location is acceptable but culturally mandated funeral rites have yet to be done. Mass graves, which the Acholi view as taboo, are defined as internments that “contains the remains of more than one victim who share some common trait connected with the cause and manner of death” (Schmitt 2001: 279; also see JRP: 2007a) including both organized and unorganized graves. Though other definitions of mass graves exist (International Criminal Tribunal for The Former Yugoslavia 1996; Mant 1987; Skinner 1987), this definition was chosen because it mirrored most research participants’ conceptions and use of the term “mass grave.” Lastly, when a family member is known to have died or there is strong evidence supporting their death, but the location of the remains is unknown, the spirit may eventually feel forgotten and abandoned by their family and become cen. This scenario is included in the category of improper burial because the cause of the cen and the consequences for the family and the community are the same as improper individual burials and mass graves. 1 The three types of depositions discussed in this thesis is not an exhaustive list of burial types. In the absence of certain cultural actions, such as placing leaves on remains, unburied remains left on the grounds surface is another improper form of burial within Acholiland that can cause disturbances. However, participants within the research area did not report that this type of deposition was bothering them because the cultural institution Ker Kwaro Acholi had previously removed a large number of these surface scatters from the area. This is not to say that surface depositions are not problematic within Acholiland, and future research may find it appropriate to replace the phrase improper burials with the more expansive term improper depositions. However, as unburied human remains were only mentioned in passing, its effect on local communities was not explored. 9 Because conducting proper burials and maintaining good relations with the living-dead (the spirits of family members) are of such importance, Acholi families and community members often utilize any means available to them to conduct rituals or prayer ceremonies for human remains in an attempt to satisfy the spirits of the deceased. Such attempts can even be made when the remains of the deceased cannot be located. Calling-of-the-spirit ceremonies symbolically beckon the spirit back home and serve as indication to the dead that they are loved, remembered, and recognized as being part of the family. However, this endeavor, as well as any other ritual or memorial prayer activity, is often hindered by practical limitations. In the past, insecurity due to numerous conflicts prevented recovery and reburial efforts; currently, poverty arising from the economic and structural underdevelopment of the north, as well as from property loss during the LRA conflict, often impedes attempts to address the issue. Instead, scarce funds are often directed toward the more immediate needs of the living, such as food, education, and other health issues (Finnström 2008; Proctor 2013). This inability of people to respond to the needs of the dead because of a lack of capacity is a constant reminder of looting, destruction, and trauma sustained during the conflict, which exacerbate other harms individuals and communities have endured (Eppel 2010; Finnström 2001; JRP 2008b; OHCHR 2007). Though traditional rituals have the potential to alleviate cen and prevent further disturbances, these ceremonies require valuable resources that many families either do not have or cannot spare. The Justice and Reconciliation Project has reported that this issue is further complicated through the prohibition born- again Christianity places on engaging in traditional rituals (2008a: 12). As such, 10 even if capacity were not an issue certain portions of the Acholi population would not support traditional rituals. Against this background, and in consideration of current debates in transitional justice and the literature in related fields, I consider how the problem of improper burials in northern Uganda can be resolved within a psychosocial context. This thesis contributes to the associated body of literature by continuing to investigate the consequences of improper burials and the disturbances caused by cen. However, this work also moves beyond identification of the problem and examines community perceptions of coping strategies used to appease cen and assess the appropriateness of forensic excavation and DNA identification as a potential solution to improper burials. Lastly, this research explores the dynamics that exist among traditional Acholi cosmology and various Christian denominations as well as their strategies and tactics for managing the vengeful spirits that plague the living. This work has applications to a broad array of disciplines and practitioners interested in transitional justice theory and practice, religious studies, human rights, anthropology, international relations, political science, psychology, and peace and conflict studies. Before turning to an exploration of this research study, it is important to consider my use of the terms “traditional” and “tradition.” These are terms that are often perceived as static in nature and rooted in the distant past, which potentially limits their analytical effectiveness in present settings. However, for traditions to remain useful they must be able to meet the needs of current problems and demands. Thus, they must - and often do - evolve and change. All tradition was at 11 one point a creation, made manifest in the context of a particular social setting at a given point in time. This “making and remaking of meaning follows the changing conditions of everyday life, and is in a constant state of becoming; as it is made, contested, reinterpreted, and remade” to fit the current situation (Finnström 2008: 46). Based on this logic, the terms tradition and traditional and the processes associated with these words should be understood as what people (the interviewees) believe them to be, both in memory and imagination. These terms and their current uses are based in perceptions of the world as it was and an interpretation of the world as it is. Another writing style choice I would like to highlight is my use of faith-based accounts. At times throughout this thesis, I write as if I am making faith statements instead of scholarly observations. This approach was purposefully taken in order to reflect the very real beliefs that the participants hold; namely that spirits exist and that they can, and do, affect their lives. Such an endeavor follows the reasoning that “emotive practices and experiences should be described in terms that can plausibly be attributed to the subject…terms that would be familiar to, incorporating beliefs that would be acknowledged by, the subject” (Proudfoot 1985: 195-196; also see Rex and Stepick 2013: 30). It was not this thesis objective to ascertain if, in fact, cen exists. Whether this phenomenon is real or imagined is irrelevant because if the participants believe in its existence then it is real in its consequences. Therefore, as I have endeavored to keep this research grounded in the perspectives of the participants, it was appropriate to describe Acholi religion and cosmology in a manner that reflects their beliefs. 12 Research Questions and Summary of Findings This project investigated four fundamental questions: 1) Are mass graves and improperly buried individual graves viewed as problematic for communities within Acholiland? 2) What are the consequences of improper burial? 3) What, if anything, should be done about these burials? 4) How can this be accomplished methodologically? The latter question also includes attention to what might be necessary to accomplish this task, and who should be involved. My main findings revealed that mass graves and improperly buried individual graves are seen as problematic by survivors because these internments violate cultural notions of how the dead should be buried. In response to these violations, the spirits of the deceased come to feel angry and abandoned and eventually become cen, or vengeful spirits that cause haunting, possession, nightmares, and sickness. This belief seemed to hold true for a wide variety of individuals regardless of age, sex, or socio- economic status. Despite the consistent belief in cen there was marked disagreement between how this issue should be addressed which hinged upon religious or cosmological belief systems. This prioritization of certain religious practices produces political implications for who is involved in the process of addressing the dead and how such an endeavor is, could, or “should” be accomplished. The religious composition of northern Uganda primarily includes both traditional Acholi beliefs and Christian religious doctrines, such as Catholicism, Anglicanism, and Evangelicalism or born-again Christianity (Proctor 2013). 13 Denominations and individuals combine, balance and utilize beliefs in various ways, leading to different interpretations of how to interact with cen. For example, evangelical Christians expressed that only prayers are necessary to satisfy angry spirits, while those who embrace more traditional Acholi beliefs, even if they also identify as Catholic or Anglican, stated that human remains need to be excavated and specific ceremonies, which involve ritual animal sacrifices, need to be performed in conjunction with the reburial process to appease the spirits of the dead. 2 In instances of mass graves, the inability to individually identify the deceased often prevents reburials and the associated ceremonies that could alleviate cen for those who follow traditional Acholi practices (Proctor 2013). Interestingly, mainline Christian groups in Acholiland - the less conservative Anglicans and Catholics - and many general participants were willing to tread the path that fell between Acholi ritual and Christian prayer. Religious leaders from both the Anglican and Catholic Churches expressed that their primary belief was that only prayers were necessary to placate cen, but if a traditional Acholi ritual was important to the populace, then it could be tolerated. Many of the interviewees were actively negotiating a path between Acholi cosmology and Christian beliefs, which rendered the inclusion of both prayers and rituals less problematic. One potential solution to the cen caused by unidentified or commingled individuals is forensic excavation and processing in conjunction with DNA 2 I recognize the critique that using terms like sacrifice portrays the culture under discussion as “primitive” and exotic. However, I have chosen to keep this word because it was the term used by the participants and I have endeavored to remain as true to their voices as possible. For examples of the context in which this term was used see the word tree located in Appendix F). 14 identification. Thus, the desirability and suitability of foreign anthropologists assisting with excavations were examined in the later stages of this research. It was found that the appropriateness of this option mirrored earlier findings, which depended upon the personal experience and belief system of the participant and, at times, popular beliefs in the area. Evangelical Christians felt excavations were not desirable because they would require animal sacrifices, which are associated with traditional cultural practices, and as such are rejected as unnecessary by born-again practitioners. However, many participants expressed that using foreign anthropologists would be an agreeable situation because they (anthropologists) have the means of accomplishing excavations and identifications while the local population currently does not. These responses could be summed as: “If I want to see it done, I would allow and agree to it.” In summary, there are three different options put forth to address the improper individual and mass graves that are causing cen: traditional rituals that involve the use of animals, commemorations or events that only include prayers, or a combination of the two. But why is there no consensus between and amongst Christian denominations and traditional cosmology? How can these different opinions be interpreted and explained? One component is the faith-based or spiritual dimensions that either require or prohibit animal sacrifices. However, I argue that another part of the equation on how improper graves are to be handled and managed is concerned with the political uses of the dead, which I refer to as 15 “necropolitics.”3 A historical review of political and military movements involving northern Uganda, provided in Chapter Two, demonstrates that the dead can have a significant influence on the living, and as such, any person or institution that can control the dead can wield considerable power over communities. If the dead are viewed as part of the community then reorganizing relationships with the dead - or how the dead are interacted with - also reconstitutes the living (Verdery 1999: 109). As researchers working in other cultural contexts have shown, this (re)organizing of the dead is often done in very purposeful ways and because of this human remains have become a vigorously contested site for various constructions that promote specific agendas (Crossland 2000). The relationship between the living and the dead can be turned into a political symbol for how the dead are used. This negotiation can then be used to a variety of ends, such as who owns the dead, who controls the past, who can legitimately direct the future (Crossland 2000: 146; McEvoy and Conway 2004: 545). Different opinions about how improper burials should be treated may have more to do with who is included, or excluded, from influencing this process. For example, when evangelical leaders claim that only prayers are necessary to address cen, they are not just denying Acholi rituals; they are also excluding the cultural leaders who would be directing traditional rituals and attempting to reduce their social power. By limiting the breadth of social actors who can legitimately manage 3 To my knowledge the term necropolotics as it is defined in this thesis has not been coined by other sources within the literature. However, the concept has been explored and elaborated upon by a number of researchers working within other post-conflict contexts, including Crossland 2000, 2009, Verdery 1999, and Wagner 2008 to name only a few. 16 the dead, evangelical leaders are attempting to strengthen their own power over the living. As such, the dead are “sites and sources of important political and ideological conflicts” (McEvoy and Conway 2004: 546). In every community affected by improper depositions, and especially by mass graves, there are undoubtedly tensions between the use of bodies as political platforms, as economic opportunities, and as a site of interaction between the living and the dead, both individually and socially. In instances of necropolitics it is not only human remains that are powerful as “graves are often as potent as the bodies themselves in the creation of political narratives” (Crossland 2009: 291). This is especially true in northern Uganda because graves are continually present on the landscape and serve as a visceral indication of events. Furthermore, all of these components impact efforts of post-conflict reconstruction. Based on the findings of this research, I argue that religion and cosmology are vital coping strategies for addressing cen on a community and institutional level within northern Uganda. Additionally, future psychosocial interventions, which may ultimately extend to forensic excavations and DNA identifications in order to mediate the relationship between the living and the dead would benefit from a more comprehensive understanding and inclusion of religion and its effects on local perceptions and needs. Thesis Layout In order to illustrate how I arrived at this conclusion I first provide a contextualized political, cosmological and religious background of Acholiland, which 17 frames and helps to explain the opinions, perceptions, and needs expressed by the participants in this study. Next, I detail the methods used in the field while collecting the data. This section includes not only the approaches I found to be successful, but also those that were less effective, and my recommendations for how such problems could be avoided in future research. This section is immediately followed by discussion of how and why the data was coded, analyzed and interpreted through the qualitative data analysis software NVivo. From here, I present the research findings and discuss their implications before finally concluding with recommendations based on these results. 18 Chapter Two History of the Northern Uganda Region The Importance of A Contextualized History In this chapter I provide a historical background to several relevant conflicts and political movements that have affected northern Uganda, as contemporary trends have their roots in these past events. For example, answers to questions on how mass graves should be addressed and who should be involved are affected by both current actions and past circumstances, such as struggles over authority and power or the rise and fall of political and religious groups. Interpretations of the past and assertions made about Acholi society influence responses to conflict and perceptions of appropriate future directions amongst religious, cultural, and political leaders (Allen and Vlassenroot 2010: 4). To more fully comprehend the perceptions shared by participants during research, it is necessary to understand the historical development of these religious, cultural and political dynamics. Uganda as a British Protectorate Uganda became a British protectorate in 1896 and remained so until independence in 1962. This period saw, among other things, the creation of administrative units and districts in order to facilitate indirect rule by the colonial power (Allen and Vlassenroot 2010: 4). However, it should be recognized that units of social organization existed before British occupation and were not solely a colonial invention (Finnström 2008: 31). Prior to colonization, the socio-political structure was characterized by decentralized, lineage-based chiefdoms that were 19 based on clans in northern Uganda, and centralized kingdoms (Bunyoro, Buganda, Busoga, Ankole and Toro) in southern Uganda (Branch 2010: 26; Villa-Vicencio et al. 2005: 110-111). The British were first interested in Uganda for its geographic significance and only became politically invested after the rulers of the Buganda kingdom requested the presence of missionaries in order to learn about Christian religion and obtain “Western knowledge” (Mutibwa 1992: 1). Buganda, at the time of this request, was under threat from both Egypt and the Ugandan kingdom of Bunyoro; therefore, British investment in the region was potentially politically advantageous for the ruling Bagandan elite (Mutibwa 1992: 1-3). It is important to consider colonial governance and the policies that the British employed because these actions are the foundation for several of the national divides that have ensued since independence. Additionally, both British policy and subsequent national divisions have affected how the Acholi perceive and define themselves as well as how they are defined by others. Colonialism is also part of the impetus for the Acholi formation as one ethno-political unit in order to facilitate broader representation at the national level. Ultimately these divisions and struggles over political participation became key components in many of the armed conflicts. It is necessary to understand the history of these armed movements as it is from these conflicts that improper individual and mass graves were created, which in turn give rise to and are directly linked to social uncertainty and anxiety associated with improper burials, which is expressed socially as cen. After the British became established in Buganda, colonial rule later spread throughout the region by employing Baganda as imperial agents, a political policy 20 that suited the expansionist goals of both parties. In the remaining kingdoms, the British aligned themselves with rulers who already reigned over centralized state- like regions with established hierarchies in order to enable indirect control over large numbers of people though minimum effort (Mutibwa 1992: 1). To facilitate the same type of rule in the north, where centralized kingdoms did not exist, the colonial administration, in concert with fledgling Christian ministries, appointed new chiefs to represent their interests (Allen and Vlassenroot 2010: 4). The colonial government ensured the success of these locally illegitimate appointees in lieu of the traditionally elected chiefs through the provision of resources such as access to rifles, allowing British colonial appointees to maintain power through force if necessary (Behrend 1999: 17). The institutionalization of British-made units over clan organization also actively contributed to the creation of territorially-defined ethnic identities that had previously been more geographically fluid (Allen and Vlassenroot 2010: 4; Finnström 2008: 31, 36). A segment that had previously been part of a broader Luo group found in several nations, such as South Sudan, Ethiopia, and Kenya, began to define itself more concretely as one Acholi entity. It should be noted that the British did not create a domain of ethno-regional belonging out of nothing. Acholi collective identity had been forming since the eighteenth century, but colonial practices had a significant impact on Acholi identity as it is perceived today (Finnström 2008: 52). This decidedly political order became naturalized and reified over the years as administrative units changed to reflect ethno-political identities. For example, in order to facilitate greater participation within the centralized system created by the 21 British, and in competition with other ethnic groups, the Acholi began to delineate themselves as a cohesive cultural unit (Behrend 1999:18). Eventually boundary- making, ethnic codification and a demand for hierarchies would prompt the Acholi to create a paramount office of the Acholi, a position which is still in contention as some consider this individual to be “a leading Acholi chief and not chief of the Acholi” (Finnström 2008: 45-46). The British also endeavored to unite the Acholi, often through the utilization of Christianity and the promotion of chiefs that were willing to cooperate with the colonial administration. However their efforts were aimed at controlling and managing the protectorate and not uniting the people for Acholi benefit (Finnström 2008: 41). The British also emphasized and eventually institutionalized a regional division between Nilotic language groups living in the north and Bantu language groups in the south. In part, this divide manifested itself through the exploitation of well-developed hierarchy systems in the south and the preferential development of this region to serve colonial ends, while largely neglecting the nomadic pastoralist groups in the north (Branch 2010: 27; Villa-Vicencio et al. 2005: 110-111). People in the south were, in general, treated as intellectuals and civil servants associated with government rule while groups from the north, especially the Acholi, were heavily recruited into the army and came to be seen as a warrior class. This perception of the north as a military/police recruitment zone not only disproportionately marshaled northerners into specific professions but promoted the perception of the military as an attractive career, as it provided one of the few opportunities for the undereducated to rise socially and participate in the national 22 sector (Behrend 1999: 18). Thus, when Uganda achieved self-governance, these two centers of power (the north and the south) were at times ethnicized as postcolonial manifestations of collective identity in the pursuit of political agendas (Branch 2010: 29; Finnström 2008: 31; Villa-Vicencio et al. 2005: 115). An increased disunity between the north and the south was not the only rift created by colonialism. In general the "political, educational, economic and social policies that were pursued by the British not only sharpened existing differences, but also introduced new class formations, stratifications, and cleavages" (Mutibwa 1992: xiii). These separations would later be used by various ruling parties to drum up support by vilifying the opposing group and initiated a legacy of division that to this day “impedes the development of an integrated Ugandan nation” (Allen and Vlassenroot 2010: 6; Mutibwa 1992: 6). It is necessary to briefly discuss the various presidents and some of their actions to better understand the fluctuations in political representation and the ontological insecurity faced by the Acholi today and their existential efforts to control their lives. All of these aspects have ramifications for how improper burials are perceived and potentially addressed today. Political Movements Since Independence In 1962, Milton Obote, a northerner from the Langi ethnic group, became president of an independent Uganda. During his first presidency, Obote worked to fill positions of the central state with northerners while continuing to entrench and expand the colonial policy of using northern populations to staff the armed forces. By the late 1960s one-third of the national army was made up of Acholi (Branch 23 2010: 29). This militarization of politics became standard practice for how domestic affairs would be conducted for years to follow. In 1971, Idi Amin, the Army Chief of Staff from the West Nile region, staged a successful coup supported by the United States and Britain, who were increasingly worried about Obote’s anticapitalist rhetoric (Finnström 2008: 66-67). After replacing Acholi and Langi troops with soldiers from West Nile, South Sudan, and the Democratic Republic of the Congo, Amin immediately began using violence to correct what he saw as Obote's ethno-regional favoritism, which itself was a division based in colonial politics. First political representatives, then appointed chiefs, and finally, Acholi and Langi civilians who opposed his government were either killed or forced into exile by Amin (Finnström 2008: 65; Mutibwa 1992; Villa-Vicencio et al. 2005: 112). Ultimately tens of thousands were killed, and this purge left the people within Acholiland destabilized and without political representation at the national level (Branch 2010). After this brutal displacement, Amin turned south and ordered invasions into Tanzania, who countered by joining forces with the rebel members of the Uganda National Liberation Army (UNLA) loyal to Obote. The subsequent collaboration between Tanzania and the UNLA succeeded in the overthrowing Amin in 1979 (Branch 2010: 29). The 1980 elections saw Obote beat contender Yoweri Museveni to regain power in a campaign plagued with allegations of corruption (Behrend 1999: 23). The Acholi and Langi populations from the north were once again heavily recruited to fill the ranks of the army. However, unlike before, there was no wide-scale political rehabilitation, as only a small portion of the Acholi elite was incorporated 24 into the new government (Mutibwa 1992). Unfortunately, these newly instated northern soldiers took revenge on individuals from the West Nile region for the violence that had been inflicted under Amin (Finnström 2008: 65). This cycle of revenge-seeking, manifested in militarized postcolonial politics, served to further entrench and polarize ethnic divisions, and served as a precursor for additional violence to come. Following his defeat, Museveni and his followers responded to what they viewed as northern military tyranny by creating the National Resistance Movement (NRM), with its military division called the National Resistance Army (NRA), and waged a guerrilla warfare campaign against Obote for roughly six years, from 1981- 1986. Unlike Idi Amin, who had rooted his prosecutions in ethnic and tribal differences, Museveni's NRM/A was composed of a heterogeneous mix of southern ethno-linguistic groups. Solidarity and support was achieved by articulating the campaign as a regional grievance between the north and the south, effectively reframing the movement from a war against Obote to a war against the north with the Acholi and their strong military presence being assigned most of the blame (Branch 2010: 30-31). These actions illustrate the potential malleability of ethno- political identity, as different leaders muted or emphasized different aspects of the population as a means to achieve a desired end. In 1985, a portion of the UNLA led by Bazilio and Tito Okello, both from Acholiland, broke away from the main army and ousted Obote. However, only six months later, control of the country was lost to the NRA despite the peace accord that was signed between Museveni and Okello (Finnström 2008: 68). Museveni 25 immediately suspended multi-party political activity and sent the NRA to pursue what remained of the UNLA as they fled north to Acholiland. In part because the NRM had successfully employed regional divisions as a means to create group cohesion and demonize the north, it was assumed by the NRM/A that the returning UNLA had support from the rural Acholi who carried the colonial stereotype of being warriors (Finnström 2008: 75). This assumption painted the picture of a massive force poised to threaten the south, resulting in the NRA launching an insurgency against civilians and the UNLA, while denying the Acholi political representation at the national level (Branch 2010: 33-34; Finnström 2008: 71). Ontological Insecurity and Social Anxiety When thousands of soldiers fled north to escape Museveni's NRA, they brought with them an increased threat of cen (Behrend 1999: 29). Historically, when soldiers returned home they were considered to have developed an “impure heart” or cwiny marac, because of the violence they had wreaked. Violence committed against enemies during armed conflict was not problematic in and of itself nor was it viewed as a moral violation. The threat to the community came from the spirits of the fallen. According to local beliefs, a soldier remained in a liminal state in the eyes of the community until he was ritually cleansed and the spirits of the dead were appeased. However with the growing anonymity of warfare it became increasingly difficult to identify which individual had been killed by whom. This rendered communities unable to reconcile with the spirits of the dead 26 and by extension the associated perpetrator remained “impure” (Behrend 1999: 29). Problems associated with modern warfare were not the only stumbling block when it came to addressing cen. As the power of elders declined, so too did the willingness of the soldiers to submit to their authority, and without ritual cleansing the abundance of vengeful spirits multiplied.4 These spirits and those viewed as responsible for their creation were assigned blame for the misfortunes the region was experiencing, further straining inter-societal cohesion. Despite the tensions between the returned soldiers and elders, the indiscriminate harassment and mounting human rights abuses including murder, torture, rape and general plunder committed by the NRA led to the development of a union between local Acholi elders, inactive ex-UNLA soldiers in the area, and the Uganda People's Democratic Army (UPDA), which was comprised of ex-UNLA who had formed a NRA resistance movement in South Sudan (Branch 2010: 34-36; Behrend 1999: 24-25). Part of this cooperation was due to the lineage-based elders’ claim that returning UNLA soldiers carried cen, which only the elders had the power to dispel. By inserting themselves as one of the only local entities capable of addressing moral and spiritual violations within communities the elders created a monopoly over control of the dead (Behrend 1999: 29; Branch 2010: 33). Though ex-UNLA soldiers had originally been 4 The declining power of elders is a side effect of past colonial endeavors and present Christian religious attempts to delegitimize their influence and power within local populations. This enterprise was in part achieved through missionary efforts at delegitimizing Acholi ancestral beliefs, which are spirits traditionally associated with and interacted with through elders (Behrend 1999; Finnström 2008; Harlacher and Caritas Gulu Archdiocese 2006). 27 reluctant to bow to the wishes of these elders and receive ritual cleansing (Behrend 1999: 25), as part of the UPDA they gained the support of elders, and tensions between these two groups were temporarily stayed due to the presence of a common enemy. Initially, this collaboration was successful; rebel soldiers stopped harming local communities and began outreach efforts, explaining their actions as an attempt to stem NRA violence and increase northern political inclusion. By 1987 this alliance began to fracture due to intense NRA counter- insurgencies, and eventually UPDA soldiers turned on the Acholi communities in order to obtain the supplies and manpower needed to sustain the rebel effort (Branch 2010: 35). This coercion led to diminished support for the UPDA, which was increasingly unable to defend the local population from the NRA. This internal crisis within Acholiland encompassed a loss of state power, a lack of political representation, the need to address NRA and UPDA violence, and inter-societal rifts between elders and an inflated number of young soldiers that were thought to carry cen. All of these dynamics set the stage for a new type of rebel movement; one that incorporated religion as a rallying point. A previously unknown Acholi spirit medium named Alice Auma, who was known as Lakwena, or “Messenger,” emerged as the leader of a movement that was initially concerned with ritual healing and cleansing former Acholi soldiers in an attempt to rehabilitate and reintegrate those who had become internal strangers within their own culture (Behrend 1999: 26). However, violent encounters with NRA soldiers prompted Lakwena to incorporate military actions, and the organization became known as the Holy Spirit Movement (HSM) (Allen and 28 Vlassenroot 2010: 8; Finnström 2008: 75). By incorporating aspects of Christianity, Lakwena was able to trump the power of male Acholi elders who had previously claimed to possess the sole authority to cleanse cen, as Christian divination was not included in the historic authority of the traditionally elected elders (Allen 1991). Another contributing factor to Lakwena’s popularity was the heightened insecurity in the late 1970's to the early 1990's that caused many Christian religious leaders to live in exile or internal displacement. This collapse in institutional church life and reduced religious leadership allowed Alice Lakwena to receive additional support, as her movement was a functioning outlet of Christian expression (Ward 2001: 195- 196). With the advantage of having seen the success and failures of previous rebel movements, Lakwena and the spirits (or jogi, singular jok) she claimed to channel proceeded to address conflict within Acholiland by using the NRA as a common enemy.5 She offered purification for past deeds and inclusion of soldiers while emphasizing regional commonality through a discourse of spiritual healing and redemption that blended traditional beliefs with Christianity (Behrend 1999: 43- 45). This last aspect allowed the HSM to become extremely successful in a short period of time - one year - as Lakwena's message of redemption and unity, regardless of ethnicity, appealed to a much wider audience, including not only soldiers from the former UNLA and the UPDA, but also groups outside of Acholiland (Behrend 1999: 26; Finnström 2008: 76). In addition, it has been argued that the 5 Jogi can also be spelled with two g’s. After consulting several Lwo dictionaries I have chosen to spell the word with only one g. 29 HSM gained acceptance because the spiritual beliefs of the movement became a mechanism for making sense of dramatic social changes in a time of lessened political representation and a means through which social discontent in northern Uganda could be expressed (Allen and Vlassenroot 2010: 8; Omara-Otunnu 1992: 458). Ultimately this allowed Alice Lakwena to successfully unite most of northern Uganda. However, once her army crossed the Nile River, they were seen as northern invaders, and the HSM was defeated roughly 100km from Kampala (Branch 2010: 37; Finnström 2008: 76). The Government of Uganda and the LRA In the wake of Alice Lakwena's defeat, the northern region was once again without widespread, systematic leadership, and the civilian population was being threatened by the remaining factions of several failed rebel movements and government troops. It was out of this turbulent assemblage that Joseph Kony and the Lord's Resistance Army (LRA) would emerge. Kony, like Alice, also claimed to be a medium that was possessed by spirits, and believed that all Acholiland had to be purged of internal enemies (Branch 2010: 37). At the time of the LRA's infancy a significant portion of the UPDA had accepted the government's Pece Peace Agreement or joined the NRA (Finnström 2008: 89).6 With only a limited number of 6 The Pece Peace Accord of 1988 was an agreement to ceases hostilities between the UPDA and the government of Uganda. Notable components or the agreement included releasing prisoners of war, reintegration of UPDA soldiers into the NRA if they so choose, and rehabilitation of war affected areas (Untied Nations Peacemaker). 30 volunteers joining his cause, Kony increasingly resorted to violence and coercion in order to secure soldiers and supplies (Branch 2010: 38). As the LRA was emerging so too were several government policies and actions that would later influence how the LRA treated civilians. First, the NRM developed a policy of becoming more invested in local government structures, and local councils (LCs) were instituted by the government at the village, parish, sub- county, county or municipality, and district levels to act as indirect representation (Finnström 2008: 93). Secondly, there were efforts to form local defense units (LDUs) to help protect the civilian population. Lastly, beginning in 1996, the government forced thousands of rural inhabitants into internally displaced person (IDP) camps in order for the NRA to more effectively target and eradicate the LRA (Harlacher and Gulu Archdiocese 2006: 3). All of these actions led Kony to believe that the enemy had become internalized, effectively transforming NRA collaborators, whether real or imagined, into traitors within the civilian population and thus targets for eradication (Branch 2010: 41). The LRA utilized a fluctuating methodology for discerning between "good" and "bad" Acholi and began a campaign of inflicting brutal violence against noncombatants that included murder, mutilation, destruction, intimidation, rape, and the kidnapping of children and adolescents to be trained as child soldiers (Branch 2010: 42, Villa-Vicencio et al. 2005: 116). This distinction between good/pure or bad/impure did not resonate with noncombatants who were often labeled "supporters of the enemy" by both the LRA and the NRA (Finnström 2008: 86). For example, part of the NRA's method of protecting people was to move them into IDP camps. However, by concentrating the 31 population into camps and restricting their movement, the NRA subjected noncombatants to increased violence from the LRA, who viewed those who moved to IDP camps as supporters of the government and subsequently acceptable targets to be cleansed. Noncombatants had little choice in the matter and were often left underprotected by the government after being forced to relocate (Branch 2010: 42). Between the NRA, which was renamed the Ugandan Peoples Defense Force (UPDF) in the early 1990's, and the LRA, the civilian population was trapped and terrorized by both groups, with neither one protecting the populace from the other. Additionally the looting and losses in IDP camps decimated economic productivity and later influenced the (in)ability of families and communities to conduct funeral rites for proper burial, which by extension exacerbates the presence of cen. Not only were IDPs restricted in their movements and economic options, but the prolonged conflict and extended camp displacement caused a severe disruption in social reproduction (Finnström 2008; JRP 2007a, 2009, 2010; Proctor 2013). For example, strict curfews at night often prevented the practice of wang-oo, or storytelling around a campfire, during which community or clan elders teach children norms and values through stories of heritage (Finnström 2008: 146). Additionally, the physical restrictions on movements over the landscape prevented elders, many of whom died during the conflict, from passing along important customary knowledge, such as information on land boundaries and ownership (Harlacher and Gulu Archdiocese 2006: 5). Declining social reproduction not only stunted traditional knowledge transmission, it also contributed to a further decline 32 in the power and authority of the elders. Conversely, the dynamics of the conflict and camp life led to an increase in power for women and youth. During these two decades of conflict, numerous peace attempts failed due to mistrust and unmet demands on both sides. An early illustration of this concerns the government's actions during the Pece Peace Agreement in the 1980's where several high-ranking UPDA rebels were imprisoned and killed "under unfortunate conditions" after surrendering (Finnström 2010a: 79). This mistrust has carried over to the blanket Amnesty Act, passed in 2000 by the Ugandan parliament. Rebel fighters were required to surrender their weapons and renounce any continued violence in the name of the rebellion. In return, all those who surrendered were either assimilated into the NRM or demobilized for reintegration into the civilian community without having to confess to any crimes they may have committed (Villa-Vicencio et al. 2005: 125). For those still within the LRA, relations with the NRM only grew more tense when those who accepted the 2000 amnesty were portrayed as surrendering to the will of the government instead of pursuing reconciliation between the two parties (Finnström 2010b: 137). LRA rebels also saw this amnesty as reinforcing a hierarchical order in which the superior government forgave the inferior upstarts (Finnström 2008: 229). In addition, Museveni undercut the stability of these peace measures by publically referring to the rebels as "hyenas," who were objectified as wild, sub-human creatures that needed to be managed through proper guidance (Finnström 2008: 114). The denigrating language denied the LRA any political credence and caricatured people in the north as animals. This classification not only allowed the violence to be 33 perpetuated, but also denied the legitimacy of any ideas for social change - good or bad - proposed by the rebels (Finnström 2010b: 151). According to this perspective, amnesty became yet another tool for persecution in the repertoire of measures used by the government. This theme of overt action developed into a common concern that leading members of the LRA repeatedly brought up during various peace initiatives over the years (Atkinson 2010). Yet the actions taken during this conflict arose from not only local agents, but were also affected by the contributions of external actors from beyond the borders of Uganda or even the Great Lakes Region. International Influences In 1986, Museveni inherited a nation-state in economic crisis, with foreign aid being the only available option to stem inflation. Thus, after bowing to the requests of the World Bank and the International Monetary Fund, the government began trading economic reform for substantial amounts of foreign aid, which was in turn primarily invested in developing the south while fighting the rebels in the north (Mwenda 2010: 47). In the early 1990s the United States, with its concerns about stopping the spread of Islamic extremism, began providing extra funding to the NRM/A (who had been renamed as the Ugandan Peoples Defense Force or UPDF) to combat the LRA (who were receiving funding from the National Islamic Front based in Sudan). This relationship provided Museveni with the perfect opportunity to request more and more funds in the name of supporting US geopolitical interests, as well as make the moral appeal that additional military 34 spending was necessary to protect civilians in the north (Blattman and Annan 2010: 155; Mwenda 2010: 49-55). In addition, by labeling the LRA as terrorists, the government of Uganda was able to dismiss negotiations demanded during peace talks by claiming that it would be immoral to negotiate with such a group (Finnström 2010a: 81). It has been suggested that this opportunistic manipulation severely hindered any incentive to end the rebellion, with the war providing the means for both regional solidarity and economic stability in the south. In 2003, the Ugandan government referred the crisis in the north to the International Criminal Court, which issued arrest warrants for Kony and his top four commanders two years later in 2005. Several of these individuals have subsequently died, while one is currently being tried at the ICC.7 Despite these efforts, the presence of the ICC has been contested on several fronts. To begin with, the ICC has been seen among northerners as a tool of the government, given that its investigation and prosecution is only geared toward the LRA despite the human rights abuses committed by the NRA/UPDF. Instead, the NRA successfully requested that any human rights violations brought against government officials be tried in national courts. The ICC explained its acceptance of this request because "the alleged crimes committed by the Ugandan government were not severe enough to breach the court's gravity threshold" (Finnström 2010b: 137). This one-sided approach illustrates bias, as government troops have been committing crimes 7 Dominic Ongwen was captured/surrendered in January of 2015 and made his first appearance at the ICC on January 26th 2015. He is being tried for three counts of crimes against humanity, and four counts of war crimes, including murder and the cruel treatment of civilians. 35 against and neglecting civilians in the north with almost complete impunity since 1986. In addition, as part of the ICC's mandate, it cannot prosecute crimes perpetrated before 2002.8 This arbitrary period (July 1st, 2002 to the present) seems wholly inadequate, as seventeen years of abuse that are part of the region's memory are not being acknowledged (Finnström 2010b: 137). Consequentially, the ICC warrants strained peace negotiations, especially during the 2006-2008 Juba peace talks, which required negotiation concerning the accountability of the LRA leaders (Allen 2010: 242). There has, however, been a distinctly positive aspect of ICC involvement, namely the greater awareness overall of human rights violations. With international scrutiny focused on the area, it was increasingly necessary for the Ugandan army to put forth a more successful effort when it came to protecting the populations in the north (Allen 2010: 261). The LRA has been pushed out of Uganda since 2006, and it is unlikely that this group will return. Yet some people still live in fear that the depleted LRA forces that operate out of the Democratic Republic of the Congo, the Central African Republic, and South Sudan will return to northern Uganda. Despite their reduced capacity, the rebels still cause havoc, use violent force against civilian populations, and continually manage to avoid capture by the contingents of UPDF soldiers that 8 This time period is based on the temporal jurisdiction of the Rome Statute, which is the treaty that established the ICC and details among other things, the crimes that fall under the ICC’s jurisdiction, its rules of procedure, and mechanisms for cooperation between the ICC and signatory States Parties. Specifically, the ICC “has jurisdiction only with respect to crimes committed after the entry into force of this Statute” (UN General Assembly, Rome Statute of the International Criminal Court 1998: 10). As such, the court only has the power to try crimes committed after July 1st 2002. 36 have been sent to pursue them. Meanwhile, Ugandans face the immense task of healing after 20 years of extreme violence as well as underdevelopment in the north. National Development Efforts Still under the leadership of Museveni, the current government has implemented two development frameworks to aid in facilitating sustainable peace in northern Uganda. The first Peace Recovery and Development Plan (PRDP), which was initiated in 2007, was plagued with insufficient oversight and funding, understaffing, and corruption. The second phase of this project, PRDP II, has been carried out from 2012 to mid-2015 and aims to ensure the functionality of implemented developments and improve on past efforts. A recent report from the Advisory Consortium on Conflict Sensitivity (ACCS) has found that although PRDP II made improvements in reconstructing physical security, health and education facilities, as well as water and public works sectors, there were a number of issues of importance to northern communities that were either not addressed or failed to attain saliency at a local level (Advisory Consortium on Conflict Sensitivity 2015: 2- 3). These included an increasing number of conflicts over land and other natural resources, youth underemployment, sexual and gender-based violence issues, and a “lack of appropriate transitional justice mechanisms,” all of which amounted to the finding that northern Uganda was still in a state of underdevelopment when compared to the rest of the country (Advisory Consortium on Conflict Sensitivity 2015: 3). Given the perceived inadequacies of the ICC and the shortcomings of the 37 PRDP plans, local populations in Acholiland have often turned to some of the mechanisms that have served to orient their lives in the past - religion and cosmology - because these strategies previously afforded communities with the means of coping with situations beyond their control. As soldiers returned to their homes, both during and after the conflict, they faced some of the same dilemmas that had afflicted the returning UNLA troops; their actions created cen and the vengeful spirits had come home with them. As it had in the past, cen and, by association, returnees, were blamed for the misfortunes that befell the populace (Harlacher and Gulu Archdiocese 2006: 5). This increased psychosocial tension only added to the distress felt by the community and hindered both reintegration and reconciliation attempts. Additionally, the capacity for elders to dispel cen through traditional ceremonies was even lower than it was before as fewer people knew the proper rituals, possessed the means or ability to fulfill them, believed in their efficacy, or were willing to submit to the power of traditional authorities. Thus, though the presence and effects of cen have multiplied, the ability to address these spirits has been reduced (Behrend 1999). However, as seen above, traditional cosmology was not the only religious influence operating within Acholiland. Christianity has played an influential role in the trajectory of the country’s development since colonization. In order to better understand how cen is presently experienced and perceived, it is important to not only understand it as a concept, but to consider the context that led to its current expression. 38 Chapter Three Religious and Cosmological History of the Region Understanding Cen within Religious Discourses The discourse in rural areas is inextricably tied to religion and spiritual forces that give expression to catastrophic experiences; thus, current expressions and perceptions of religion and cosmology can only be “understood in the context of the history of various religious discourses” (Behrend 1999: 100). Cen, specifically, can be more effectively comprehended by taking note of the complexities of the concept within its context, including its historical underpinnings. It is not enough to simply study cen on its own as its current expression is the culmination of traditional practices and beliefs, Christianity, and political maneuverings both during colonialism and after independence. Luo Cosmology As mentioned in Chapter Two, the socio-political organization in northern Uganda during the pre-colonial era was organized into chiefdoms, which in turn were composed of clans. From the level of clan, social units are separated into villages, then hamlets, and finally to households (Harlacher and Caritas Gulu Archdiocese 2006: 24-25). Cosmology amongst these units are utilized to explain and cope with misfortune and to guide behavior (Behrend 1999: 107; P’Bitek 1971). Both hardships and blessings were seen as the result of spiritual forces including chiefdom and clan jogi, ancestors, and free jogi (explained below), all of which could 39 display anthropomorphic characteristics such as feeling cold, hungry, thirsty, abandoned, or loved (Harlacher and Caritas Gulu Archdiocese 2006: 42-43). Each chiefdom and clan has its own jogi or spirit who is responsible for enforcing moral order within a specific geographic area (Behrend 1999: 108; Harlacher and Caritas Gulu Archdiocese 2006: 44). Violations of moral order can cause an individual to become impure of heart and force the jogi to send punishment or misfortune to the perpetrators (Behrend 1999: 107). During yearly celebrations, inhabitants would make sacrifices to the clan or chiefdom jogi to ask for protection, but only those who were clean of heart and free of guilt could do so without incurring the wrath of the jogi. This in turn encouraged individuals and families to settle their disputes before attending the ceremony. Ancestors consist of clan members who have died and are responsible for enforcing social rules and promoting clan unity (Harlacher and Caritas Gulu Archdiocese 2006: 43). These living-dead are capable of helping or hurting people according to how the living behave. An abila, or ancestral shrine, can be found in many homes and is the location where people would go to offer sacrifices for blessings or for thanks to the ancestors for general good fortune, successful harvests or hunts, or the birth of children (See Figures 1 and 2). Both the ancestors and the clan or chiefdom jogi were thought to only send just punishments. As such, severe punishments and unexplainable sickness are attributed to the work of free jogi (Harlacher and Caritas Gulu Archdiocese 2006: 46). Unlike the other spirits, free jogi are not tied to a specific location and can move over large distances. Though these spirits can be used to heal or to harm, 40 Figure 1. This picture displays two household abilas. Photo by A. Kembel 2012. Figure 2. This photo displays a village abila. Many of these shrines were purposely demolished by the LRA in order to destroy the power they symbolized (Finnström 2008). Photo by A. Kembel 2012. 41 especially when harnessed by an ajwaka, or witchdoctor, they can also strike without cause. These free jogi frequently bring illnesses with specific symptoms such as anxiety (P’Bitek 1971: 114). In instances of these “spiritualized” ailments, treatment serves to address both the physical, psychological, and religious cause of the condition (P’Bitek 1971: 114). As colonial control spread and new forms of political organization were implemented, the influence of clan and chiefdom jogi was greatly reduced as traditional offices were replaced or dispensed with altogether (Harlacher and Caritas Gulu Archdiocese 2006: 24, 48; Russell 1966: 29). This same time period, which corresponds to the introduction of Christianity, also saw the proliferation of free jogi, a category under which cen is subsumed (Behrend 1999: 109-110; P’Bitek 1971: 108). A number of these new spirits had foreign names, demonstrating that “the Acholi were using their ‘cultural archive’ to interpret this new threat to their worldview” (Behrend 1999: 106; Foucault 1972). The Spread of Christianity Christian missionaries first arrived in Uganda in the late 1870’s, with an evangelical wing of the Anglican (Protestant) church, the Church Missionary Society (CMS) appearing in 1877, and the Catholic White Fathers Mission arriving in 1879 (Pirouet 1978: 1-2). These Christian groups, being confined to Buganda, initially worked together to align themselves with the politically-active younger generations and were soon set in opposition to a traditional political party and a Muslim party (Pirouet 1978: 3). In 1887, the new ruler of Buganda, Mwanga, elevated Christian chiefs in an attempt to offset the traditional powers. This tactic proved so successful 42 that Mwanga soon attempted to turn against the Christians to curb their growing power only to be deposed by a Muslim-Christian alliance. Rivalry and suspicion quickly disrupted this newly formed coalition, and the stronger Muslim party drove the Christians into exile (Gifford 1998: 112). In 1888, the Christians were able to oust the Muslims and reinstate Mwanga, though his powers were severely curtailed (Pirouet 1978: 5). Being English, the CMS missionaries also led the campaign for the British declaration of Uganda as a protectorate, as it would cement their power over the Catholics who were French evangelists (Gifford 1998: 112). As a direct result of this power change, any individual with political ambition had to be a member of one of the Christian groups, successfully uniting political and religious motives within the region for the foreseeable future (Pirouet 1978: 7). Despite early Anglican success, this evangelical brand of Christianity was at times less popular than Catholicism because the evangelical Protestants expected a strict adherence to Christian morals. Conversely, the Catholics did not require the same level of religious fervor and realized that Christianity in the region had become indigenized; therefore, they did not expect the population to give up worldly items or practice monogamy (Pirouet 1978: 10-11). Despite these different approaches both religions remained popular, in part, because adherents were taught to read and write, a skill that became mandatory for political advancement (Pirouet 1978: 29). This was especially the case with the expanding colonial power filling political positions with literate individuals who would support their cause. The initial Christian expansion into Acholiland beginning in 1904 was at the invitation of a rwot (chief) named Awic who had been imprisoned in Kampala for 43 resisting British rule (Pirouet 1978: 149, 152-153). During his incarceration in the south, Awic observed the expansive changes the British were bringing and the potential advantages conferred upon those who could read and write. After returning to the north, a request was extended to the CMS church to send catechists, as Christian churches were the sole source of education at the time (Russell 1966: 3). Despite several false starts, Christianity successfully expanded into and throughout Acholiland because of spiritual motivation, a local demand for education, and because missionaries had become the handmaidens of the British regime (Pirouet 1978: 35; Russell 1966: 33). As British expansion took place in the north, local chiefs who were antagonistic to the new rule were deposed and replaced by more acquiescent candidates who had been taught in missionary schools (Marchetti 1999: 21). These new pupils were comprised of the traditional chief’s sons, who worked to serve a variety of purposes. The rwot would have educated offspring who could handle affairs with the government, which was a requirement that was becoming necessary for political advancement and the betterment of the clan. In addition to this, the colonial power would gain new chiefs who would support their endeavors and retain the appearance of traditional legitimacy within communities, while the churches gained the opportunity to spread their faith and their power base. These different motivations culminated in churches becoming “an essential part of the social fabric, and have become deeply infused into political, social, and cultural life.” (Gifford 1998: 116). Though the centrality of the Anglican and Catholic churches 44 has remained strong, other denominations have crept into prominence; perhaps the most notable of which is Pentecostalism. In the mid-1980’s Pentecostalism started gaining popularity throughout Uganda. Though this movement is independent of the Anglican balokole revival, which took place in Uganda from the 1930’s to the 1980’s, the Pentecostal movement has continued to build upon some of the same stances and roles (Kalu 2008: 95). Such characteristics included being “saved” and an “uncompromising rejection of assimilation between Christianity and African customs” (Gifford 1998: 152-153; Kalu 2008: 95). With church founders originating from both within Uganda and abroad, the born-again experience emphasizes or exploits a need for enthusiasm and participation in services that create a collective atmosphere; attempts to promote a perception that adherents can breaking through traditional or economic boundaries associated with poverty; and that emphasizes that believers can control their life circumstances and be victorious over the various struggles they are grappling with (Gifford 1998: 156-157, 169-170). This viewpoint was at times bolstered by the state, which required new missionary groups to become involved in development. In the south, this obligation painted Pentecostalism as a new dispensation with the potential to provide greater opportunities than the established Anglican and Catholic churches (Gifford 1998: 170-171). Despite the popularity of evangelical churches, the north remains predominantly Catholic, in part due to the Catholic Church’s greater presence in terms of numbers, external funding, and infrastructure investment (Gifford 1998; 147; Pirouet 1978: 144). 45 The partition of cosmology and religion into different sections within this chapter was for organizational purposes and should not be seen as a suggestion that cosmology and religion in Acholiland are entirely separate entities. Though most Acholi have converted to some form of Christianity, traditional practices and Acholi cosmology have infused Christianity and vice versa (Behrend 1999). For example, the literature addressing cen (JRP 2012b, 2011, 2008a, 2007a, 2006b; Neuner et al. 2012) demonstrates that it is a significant problem in Acholiland, affecting a wide variety of individuals regardless of age, sex, or socio-economic status. This research also found that the perception of a vengeful spirit that could cause harm and whose genesis is rooted in traditional cosmological beliefs was internalized within the belief system of Christians. As will be discussed in Chapter Six, disagreements over how spirits should be appeased arise between participants because both religious institutions and individuals combine and balance beliefs according to diverse social, political, and economic needs, which leads to different understandings of cen. 46 Chapter Four Methods Laying the Foundation This thesis is based on research undertaken in Uganda during the summer of 2012, though networking and several of the founding ideas for this project were the product of two previous visits in 2011. For ten weeks I conducted ethnographic research on Acholi cultural practices, needs, expectations, and perspectives in order to answer the question, what are the culturally appropriate solutions for addressing improper individual and mass graves in northern Uganda? This research, which was funded in part through the generosity of the W. K. McClure Scholarship for the Study of World Affairs, was conducted under the auspices of the Refugee Law Project (RLP), a Uganda-based non-governmental organization (NGO) that focuses broadly on human rights issues within the country. With several locations in Uganda, the organization strives to address topics related to conflict, transitional justice, governance, access to justice, gender and sexuality, displacement, and mental health and psychosocial well-being. I specifically worked in the RLP Kitgum field office, the National Memory and Peace Documentation Centre (NMPDC). Their activities focus on establishing a national memory narrative by documenting all of the conflicts that have affected northern Uganda. The office also conducts community outreach, community-based action research, and documentation that fall under the umbrella of its parent organization RLP. One example of this research is the NMPDC’s massacre scoping and conflict mapping project, which, as its name implies, records events associated with conflict 47 throughout Acholiland. The decision to work with this NGO was based on the previous networking I had done with several of the staff and administrators, their invaluable expertise in and familiarity with the area, and their established status as a reputable organization. All original research material, including informed consent forms and audio recordings, remained with RLP upon my departure. Additionally, the government of Uganda requires all foreign researchers to be associated with an organization already operating in Uganda (See Appendix A for Letter of Association with the Refugee Law Project). This requirement is coordinated through the Ugandan National Council for Science and Technology (UNCST) which reviews all proposals based on many of the same guidelines as those considered by American Institutional Review Boards (IRB), with particular concern for ethics (See Appendix B for UNCST approval). Participant Descriptions The 52 participants in this study originated from war-affected communities with known occurrences of improper individual burials, reburials, and/or mass graves, or they had conducted ceremonies on behalf of deceased and missing relatives. Because the landscape contains burials from several conflicts, selected participants were not limited to those who had lost loved ones during the struggle between the LRA and the government of Uganda. It should be noted that the number of listed participants only reflects the people who contributed in semi- structured interviews and focus groups. The research was also influenced by 48 numerous daily informal exchanges, which were recorded field notes but not counted as part of the official tally. The 52 participants, whose responses served as my primary source of data, included individuals who either lived in close proximity to a mass grave(s), were believed to have a loved one located in a mass grave, had participated in a single grave reburial, desired to conduct a reburial, had participated in a calling-of-the- spirit ceremony, or some combination thereof. In addition, key informants that held distinct positions, such as elders, rwodi (chiefs), and various religious leaders of different denominations were consulted because their perspectives could shed light on how the research questions relate to communities on a macro-level. Research Locations The research locations were identified through initial field visits and in consultation with Refugee Law Project directors. The identification of potential field sites was based on past massacre scoping conducted by NMPDC that had identified the presence of mass graves and other war-related losses within communities. Labongo Akwang sub-county was selected as the primary geographical area of exploration because reburials were known to have occurred in the area and because transportation to and from the area was logistically feasible during the rainy seasons (See Figures 3 and 4). Naamokora and Omiya Anyima sub-counties were included later in the research, based upon the presence of two mass graves, the characteristics of the site and the ability to access participants (See Figure 5). 49 Figure 3. Map of districts in Uganda. Map Source: http://www.artofanderson.com/map-of-uganda-districts-2012/ 50 Figure 4. Image of Kitgum District and its constituent Sub-counties. Map Source: http://reliefweb.int/map/uganda/uganda- greater-north-planning-map-details-14-oct-2010 51 Figure 5. Map of research locations. Labongo Akwang Sub-county is located to the left of Kitgum town while both Omiraanyima and Naam Okora town are on the right. Map Source: http://reliefweb.int/map/uganda/uganda-planning-map- details-22-dec-2010 52 Gaining Access to Participants Before any interviews took place, accessing participants also required obtaining consent from the Local Council Member 3’s (LC3’s). An LC3 is a government official and gatekeeper that presides over a specified geographic area - a sub-county - and manages the activities within his/her respective region. After an LC3 had granted permission to conduct research she or he would introduce me to a Local Council Member 1 (LC1) or a member of their chair personnel, who helped mobilize community members for participation. An LC1 is the government official that supervises at the village level, while the chair personnel under a LC3 serve as consultants and help the LC3 with his or her duties (For a hierarchical listing of these local government positions see Figure 6). After initial interviews took place within a community, snowball sampling was used to locate other affected families in the area. Accessing participants through introductions is necessary on a formal level because it is a cultural norm for outsiders, both foreign and domestic, to work through existing hierarchies of authority. Any researcher, including other Acholi, wishing to speak with community members would be expected to follow this protocol. Additionally, land wrangling (disputes over land boundaries or ownership and even attempts to swindle another out of land ownership) is rampant in the north and communities are suspicious that outsiders may try to steal their property, as land is one of the most profitable resources in the region. Because of these fears, a researcher must approach communities through a trusted individual, lest their intentions be misunderstood. However, there are potential biases that arise from 53 Figure 6. Government Hierarchy and Associated Titles. 54 meeting participants in this manner. It is possible that the LC3 and other officials steered us toward personal family and friends or people who agreed with their viewpoints concerning the study. For example, in one research location the first people we met were members of that individual’s clan who had very coordinated opinions on how to address the mass grave in their area because the future management of that particular grave had economic implications for them. This selective directing is a factor within the research that should be recognized because it implies that the results should be cautiously applied to the wider Acholi population. However, this possible bias should not be considered a limitation because these people were contending with mass and individual graves and thus their opinions and perspectives are still relevant to the research questions. Another related dilemma arose when an LC1 remained for the interview after making introductions to a family. At times, questions posed to the participants were answered by the LC1 and his presence may have affected the diversity of opinions the participants were willing to voice. However, LC1’s are elected members from the village level and are often elders within their communities. Because of this close connection to their village, these individuals have a vested interest in their community’s perceptions and needs. Additionally, social pressure makes it less likely that an LC1 would deliberately push an agenda that was not supported by the community as he or she may be related to, live close to, work with, and depend upon this network of individuals. It is also important to note that Acholi society is collectivist in nature and important decisions often require input from an 55 individual’s family and clan. Because kin-based decisions are a cultural norm a researcher in this region should be cautious about overemphasizing and prioritizing individual opinions and interviews to the exclusion of collective responses. However, if the researchers goal is to hold interviews without the presence of an LC1 I would suggest interviewing the official in a separate session before the families so that he or she is no longer curious about the type of questions that the researcher will be asking and less likely to stick around for another interview process. The Research Scheme The methodology used in this research was qualitative and the study exploratory in nature. Primary data collection methods included informal interviews and discussions, thirteen semi-structured interviews, and two focus group discussions. The interviews and focus group discussions were conducted at a location of the participant’s choosing. It was usually in their home and in the presence of several family members from the same household or compound. As interviewing can be used to investigate motivations, opinions, experiences, and differences in meaning (Dunn 2010: 102) this method effectively allowed me to investigate the validity of the project’s objectives. Interviewing also encourages responses in the participants’ own words (Dunn 2010: 103), and permits the informant to convey the information she prioritizes. Furthermore, semi-structured interviews were utilized instead of structured or unstructured styles because they facilitated conversations that generally focused on improper burial while still 56 permitting participants to introduce aspects of this topic and directions that had not previously occurred to me (Dunn 2010: 110-111). On several occasions more than one family was present, and the interview became an impromptu focus group. This transition to a different research method was necessary to ensure that the different families were not offended as they had often taken time away from vital daily activities, such as digging (farming), in order to be present. Utilizing focus groups was beneficial in several ways. First, these groups allowed me to test my interpretation of emerging conceptual categories that were generated during individual interviews and refine or redirect future inquiries accordingly. Secondly, focus groups can be useful for understanding the socially- constructed nature of knowledge (Cameron 2010: 154), as well as observing the transmission of information between individuals. This method allowed me to observe group dynamics as “relevant constitutive forces in the construction of meaning and the practice of social life” (Kamberelis and Dimitriadis 2005: 902). Both of the groups were comprised of a mixed-demographic, including men, women, young adults, and the elderly. The inclusion of both young adults and elders allowed me to observe the dynamic that may exist between these two groups. Several sources within the literature pertaining to northern Uganda (e.g., Allen and Vlassenroot 2010; Behrend 1999, and Finnström 2008) suggest that the power of the elders has been greatly diminished. Because of these reports, I was interested in gauging how eager the youth were to listen to elders and their perspectives. Though elders may possess less power and authority than they did in the past, the youth that I observed seemed willing to consider the responses provided by elders or at 57 the very least did not openly reject them. Importantly, young adults reported that traditional rituals were an important aspect of addressing improper burials, but many of them were unfamiliar with how to conduct such ceremonies or the rationale for why certain objects were used. If this observation is representative of the broader population, it may have impl