Читать книгу Remediation in Rwanda - Kristin Conner Doughty - Страница 8

Introduction

Harmony Legal Models and the Architecture of Social Repair

This book tells the story of how the Rwandan government aimed to remediate the country’s social ills in the aftermath of the 1994 genocide, and to rebuild the social fabric through revitalizing a traditional practice of mediation and embedding it in grassroots legal forums. Government officials hoped to use local courts as part of a broader architecture of social repair to help re-create families, repair communities, and rebuild a unified nation, consistent with the global rise in transitional justice. The three legal forums at the heart of this book—genocide courts called inkiko gacaca, mediation committees called comite y’abunzi, and a legal aid clinic—used mediation by third-party actors with the power of punishment to advocate compromise solutions across a wide range of disputes. This book describes how people’s lives were shaped by law-based mediation. It illustrates how people contested collective belonging through the micro-disputes that formed the warp and woof of daily life, and how people negotiated moral community and imagined alternative futures through debates in these new legal forums. By looking across a range of legal forums, it connects the exceptionalism of disputes about genocide to the ordinary disputes people faced living in its aftermath.

Mediation-based models are by no means limited to Rwanda; they are increasingly implemented by state and international actors, especially in postconflict contexts, often in the name of customary law, local solutions, or restorative justice. This is therefore a story about rebuilding in Rwanda, but also a broader story of how the global embrace of law, particularly in postconflict contexts, shapes people’s lives. Though law-based mediation is framed as benign, and is often promoted on the basis of local culture—by domestic governments such as Rwanda as well as by the transitional justice movement more broadly—I show that its implementation involves coercion and accompanying resistance. Yet, perhaps unexpectedly, in grassroots legal forums that are deeply contextualized, processes based on promoting unity can open up spaces in which people reconstruct moral orders, negotiating what Kimberly Theidon has called the “micropolitics of reconciliation” (Theidon 2012).

Legal Disputes in the Aftermath of Genocide

This book adds another strand to the complex story of how Rwandans rebuilt their social lives in the wake of genocide.¹ From April to July 1994, devastating state-led violence designed to exterminate the country’s minority Tutsi population spread from the capital, Kigali, throughout the country.² The extraordinary violence was perpetrated by the military and ordinary civilians, and it extended deep into everyday landscapes, touching all Rwandans. It was framed by a civil war and subsequent violent political and military struggles, ultimately claiming up to a million lives. Since the moment the worst of the violence abated, victims, perpetrators, bystanders, their families, and the recently returned diaspora population have lived side by side across the country’s tightly packed hillsides.

In the mid-1990s, as bodies lay unburied, and refugees streamed out of Rwanda fleeing retaliatory violence, equal numbers crossed into Rwanda after years in exile. These diaspora people replaced, and eventually exceeded, the dead. Within little more than a decade later, this cosmopolitan mixture of people mingled in markets and manually tilled their adjacent fields, they rode pressed hip to hip and shoulder to shoulder in minibus taxis, and their children sat together in schoolrooms. Reports and documents produced by the Rwandan government and international nongovernmental organizations (NGOs) commonly diagnosed postgenocide Rwanda as suffering from a “devastated” social fabric among these people, including the “destruction of family bonds” (Biruta 2006:155; Ministry of Finance 2000; Ndangiza 2007).

In an effort to create a break from the past, and consolidate power, the Rwandan government focused on reconciliation as the foundation of its governance strategies to restore peace and rebuild a prosperous nation-state. In 1999 the government established a National Unity and Reconciliation Commission, which worked at the national level in collaboration with ongoing government programs, as well as at the grassroots level educating the population.³ The purported goal was to “bridge the deep rifts in society and heal the wounds inflicted by the genocide,” creating cohesion out of division and exclusion (Ndangiza 2007:1). In that vein, Rwanda’s new 2003 constitution underscored the importance of “eradicat[ing] ethnic, regional and any other form of divisions,” “emphasizing the necessity to strengthen and promote national unity and reconciliation,” and recognizing that “peace and unity of Rwandans constitute the essential basis for national economic development and social progress” (2003: Preamble, Articles 2, 4, 5). National unity was justified on the basis that, as the constitution enumerated, “we enjoy the privilege of having one country, a common language, a common culture, and a long shared history which ought to lead to a common vision of our destiny” (2003: Article 7). The public reconciliation discourse was ubiquitous, from political speeches to NGO conferences to sporting and music events, and was taught in schools (King 2014) and solidarity camps (Thomson 2011b).⁴ Rwanda’s policy of unity and reconciliation formed what Sharon Abramowitz (2014) has described in Liberia as a “total environment of post-conflict transformation,” and it was emblematic of how peace building and reconciliation emerged as global master narratives of the late twentieth century (Nader 1999a:304, 2002b:38; Scheper-Hughes and Bourgois 2004).

Grassroots legal forums in Rwanda were a central site for promoting national unity and restoring the social fabric. Specifically, Rwandan government authorities identified local mediation-based courts as an antidote for the divisive politicized understandings of ethnicity at the heart of the genocide ideology. As the president Pasteur Bizimungu explained in 1999, law was intended not just to punish or to make people “be afraid” of committing a crime but rather to change the “bad ideology” that caused the divisive violence and to rebuild the nation (Bizimungu 1999:6). These forums were designed to create the rules, values, and norms among the participants that would “become the basis of collaboration and unity”⁵—since, as anthropologist Danielle de Lame wrote in 1996, “the old representations that unified Rwandan culture have become meaningless” in the wake of such catastrophic violence and displacement (2005:491). As one Rwandan legal administrator explained to me in 2008, “It would be better for the country if people could change their mindset to accept solutions built on unity and compromise rather than litigation.” The government justified the principles of unity and compromise at the heart of Rwanda’s mediation-based forums as deriving from the shared historical-cultural Rwandan identity undergirding the policy of national unity and reconciliation.

Using law to bridge divides from the past was perhaps most notable in the creation of the now widely known gacaca courts (inkiko gacaca), community courts in which suspects of the 1994 genocide were tried among their neighbors before locally elected judges. I sat through one such trial with a few hundred people on a rural hillside in southern Rwanda in December 2007. The defendant, a man I call Alphonse who had been incarcerated since 1995, was a fifty-five-year-old widower with six children, educated through primary school, and recently converted to the Seventh Day Adventist Church. During his trial for acts allegedly committed thirteen years earlier, Alphonse had apologized for his role in several murders, confessing in detail by providing names of co-perpetrators and describing how and where several specific victims were killed. The eighteen witnesses and victims who testified claimed to forgive him despite the severity of his crimes, and many even spoke positively about ways he had helped some people.

In the closing moments of his trial, Alphonse knelt before the five judges, his bare knees pressing into the packed dirt. He clasped his hands behind his back and bowed his head toward his chest, choking back tears as he asked forgiveness of the judges, his neighbors, and all Rwandans. As Alphonse stood to take his seat on a low bench at the base of the tree under which his trial had taken place, he drew a tissue from his jacket pocket and wiped his eyes. The prisoners on either side of him shifted uncomfortably, and looked away.

The judges rewarded Alphonse’s rather dramatic performance by, after deliberating, reducing his sentence based on his confession, which allowed him to return home within weeks to live among the people who had attended his trial. Over the next six months, I saw him regularly around town, chatting with former prisoners as well as genocide survivors, listening to the radio or having a beer with other men. In subsequent trials in the following months, the judges repeatedly invoked Alphonse’s name as a model for other defendants to follow. He testified regularly from the general assembly, sometimes confirming defendants’ versions of events, other times refuting them. Five months after his trial, Alphonse described what it was like returning home:

People reacted better than I expected them to. When I was first released, I thought some people would not talk to me, I thought I would have to isolate myself, that there would be places I could not go, people I could not embrace. I came and stayed at my house for two weeks. When my children would talk, I would caution them to keep their voices low. I would even draw the curtains and keep the radio volume low, so no one would know I was home. Then people came to visit, making noise and singing. They asked me why I wasn’t coming out, and told me not to be afraid, to come and join the others. It’s good so far.⁶

Alphonse’s case stood out as a particularly dramatic example of how gacaca was intended to address the seemingly intractable problem of restoring the social fabric devastated by the violence of the early 1990s. Rwanda’s gacaca courts, created by national law in 2001, were designed, according to government sensitization propaganda, to “allow the population of the same Cell, the same Sector to work together” and therefore for gacaca to “become the basis of collaboration and unity.”⁷ Ostensibly, the courts were intended to provide a process to facilitate the reintegration of perpetrators into quotidian life, out of prison. Suspects were tried publicly onsite at the location of their alleged crimes, by a panel of locally elected judges called inyangamugayo (Kinyarwanda for “people of integrity”). Mandatory participation and the deliberate avoidance of lawyers or professional jurists were intended, at least in theory, to enable all Rwandans to participate actively as witnesses, prosecutors, and defenders with limited hierarchy. The gacaca process hinged on mediating compromises prioritizing the collective over the individual good: as Alphonse’s case demonstrated, defendants received reduced penalty for confessions, and victims forwent demands for harsher punishment in return for information about their loved ones. At the same time, the gacaca process established criminal accountability for genocide, including fines and prison sentences of up to thirty years. The inyangamugayo, backed by police, armed prison guards, and even soldiers, could issue subpoenas and could bring charges against case participants for refusing to testify, for perjury, or for blackmail (Organic Law N.16/2004, Articles 29, 30).

This emphasis on collaborative, unity-based justice in Rwanda was not limited to genocide crimes. In the days and weeks around Alphonse’s trial, I sat on the same hillside as people brought cases before newly implemented mediation committees (comite y’abunzi) that had jurisdiction over low-level civil and criminal disputes. Comite y’abunzi were established by national law around the same time as gacaca courts based on the same general procedure, in order to “provid[e] a framework of obligatory mediation” for ordinary civil and criminal disputes, where locally elected abunzi (Kinyarwanda for “mediators” or “reconcilers”) would “seek first to conciliate the two parties” (Organic Law N.31/2006, Articles 3, 20). Even while the abunzi aimed to find compromise solutions with which parties would voluntarily comply, state-backed punishment remained an integral component. Based on comite y’abunzi judgments, authorities and the judicial police could seize land, livestock, crops, and other personal property (Organic Law N.31/2006, Article 24).

In one example, from May 2008, a woman named Joselyne had filed a case against a neighbor named Beata for saying insulting things about her in public. Both women wore the kitenge wrap skirt over a T-shirt common among rural women, their slim and strong frames testament to their lives of subsistence farming, their bare feet covered in light dust from the walk. The women explained they had previously taken their dispute to a local women’s committee, which reprimanded Beata and told her to pay a fine of three thousand Rwandan francs (six U.S. dollars). Unable to pay, Beata had agreed to farm for Joselyne, and she spent one afternoon fetching water and tilling Joselyne’s garden. Beata did not go again, claiming it was too much hardship after her child fell ill, especially since her husband was doing twice-weekly community service work as part of his gacaca sentence. Joselyne felt the debt had not been paid and filed a case with the comite y’abunzi. The abunzi had summoned Beata repeatedly before she finally complied. During the case discussion, the abunzi concluded that because Beata had already acknowledged guilt in Joselyne’s defamation accusation, the issue at stake was the unpaid fine.

The president of the mediators proposed that Beata should simply pay the fine imposed by the women’s committee. Beata agreed. Joselyne repeatedly rejected this solution. She complained it was unfair that Beata had been so insulting to her, and then had shown more contempt by repeatedly refusing to pay and most recently by not responding to the abunzi summons, which had wasted Joselyne’s time and prevented her from farming. She wanted Beata to have to pay more than the original fine. Joselyne asserted stridently, “I do not want to forgive Beata, I want you to punish her.” The abunzi who were hearing the case responded by explaining that their job was not to punish but to mediate and reconcile. They spent more than an hour reasoning with Joselyne, emphasizing a range of compromise solutions aimed at avoiding escalation. One mediator tried to convince Joselyne of the benefits of Beata’s willingness to farm for her. Another likened Joselyne to a parent and Beata to an errant child who should be forgiven. This case hearing concluded not with a performance of repentance and reintegration as in Alphonse’s situation, but rather with the disputants walking purposively in opposite directions, deliberately avoiding each other over the coming weeks.

I encountered law-based mediation in a third forum located a few hours’ walk away, this one a legal aid clinic run by the Faculty of Law of the National University of Rwanda. The clinic’s staff members, who were university lecturers and law students, provided pro bono legal advice to all clients who came to them, with cases ranging from divorce or land disputes to employment concerns or insurance claims. I was initially drawn to the clinic by an interest in exploring another legal context that addressed quotidian disputes, though one based in universal human rights and Western-style law rather than the alleged customary law focus at the heart of gacaca and comite y’abunzi. Yet I was struck to find that in the months I spent there, the majority of clients who came to the clinic requesting help with filing a case or defending themselves against a claim ended up participating in mediation. The clinic’s staff members were explicit that mediation was at the core of their approach. As the head of the clinic indicated in a presentation to his funders in 2008, “Very often clients are advised about the option of mediation,” and “then it is the privilege of the legal clinic to reestablish the broken social fabric.” The clinic was not a formal government institution, yet its link to the national university conveyed proximity to state power and punishment that, as the director explained to me, “reminds them [disputants] of what they are required to do by the law.”⁸

In a typical example, a few weeks after Alphonse’s trial, in January 2008, I entered the legal aid office, located in a small room in a single-story concrete building on the university campus. I pulled up a wooden chair around a low coffee table to join a staff member while he spoke with three clients—a man who had been fired from work as a night guard and the two nuns at the convent where he had worked. The man had come to the clinic the week before seeking help filing a case against the convent for several months of back pay, but instead the clinic staff began with mediation. The nuns explained that they had fired the man because he had allegedly stolen several blankets. Over several hours of discussion, the clinic staff member reminded the nuns of the man’s reliable work, and of his ongoing obligations to his family, while reminding the client of the benevolence the nuns were showing him by not reporting him to the authorities. The nuns agreed to pay the man two months’ wages, and the parties all expressed satisfaction with the result, grateful that they had not had to bring in the police.

How do we understand this focus on mediation and restoring relationships across such a range of legal forums and types of disputes in contemporary Rwanda? What happens when efforts at reconciliation are embedded in legal forums, thus combining pursuit of unity between disputants with punishment, and how do such efforts shape the conditions under which people rebuild their social lives? How do people follow and adapt the guiding rules of legal mediation, and what kinds of collective belonging result? What is at stake in participants’ contestation of the terms of unity? These are the questions that this book explores.

In this introduction and the book overall, I make a series of interrelated arguments that allow us to examine how and why an increasingly authoritarian state emphasized harmony and law-based mediation as a mode of power and tool of governance. I suggest we should consider these three legal forums within one analytic frame as linked by their mediation practice. Doing so captures how people experienced the exceptionalism of postgenocide justice as inextricably linked to the more mundane disputes of daily life. I further argue we should see these forums as what Laura Nader has termed “harmony legal models” (Nader 2002a) in order to identify how the forums served as techniques of state-backed community building, a form of what Nikolas Rose, building on Foucault (1982; 1991), has called “government through community” (Rose 1999:176). Analyzing these harmony models as governmentality sheds light on their coercive dimensions, and thus illustrates the dark side of harmony that is often erased by benign cultural justifications for unity. Gacaca courts, comite y’abunzi, and even the legal aid clinic could serve as spaces in which state power coerced, silenced, or pacified people through exhortations to submerge individual desires and rights in lieu of the collective good, defined in a particular way.

Yet, attending only to coercive dimensions of mediation across these diverse forums risks flattening and mischaracterizing dynamics of power in postgenocide Rwanda. I suggest that because these forums were deeply contextualized, they were relevant to people and varied in how they operated. Attending to the contested conversations within these forums allows us to analyze how people negotiated the micropolitics of reconciliation within law-based mediation as part of their efforts to reconstruct moral orders. Further, attention to the conversations themselves draws attention to the ways power and instrumentality operate at individual, interpersonal, and localized levels, in a context shaped by the macro-level global and national reconciliation agendas. What resulted was not the creation of idealistic state versions of national belonging, nor simply coercive silencing, but rather a space in which people contested the terms of moral community, actively debating what obligations they had to one another, and what it meant to belong.

Embedding Reconciliation in Law: Harmony Legal Models

Rwanda’s gacaca, comite y’abunzi, and the legal aid clinic were united by a common undergirding philosophy that, whenever possible, disputes should be solved through compromise and reconciliation with the help of a third party. They were designed to bring participants together as joint problem solvers to find mutually agreed solutions, rather than one side winning and the other side losing. This approach posited that by reasoning with one another, understanding the other side’s position, and making sacrifices, people would come to more desirable and sustainable solutions than if they went to regular adversarial courts, which imposed top-down solutions.

Mediation and reconciliation are similar ideas, as I use them here, expressed in the Kinyarwanda word kunga, or its reflexive form, kwiyunga. Kwiyunga is the root of the word abunzi, which is translated into English as mediators; comite y’abunzi are mediation committees. Similarly, kwiyunga is at the root of the word ubwiyunge, which is translated into English as reconciliation. For example, the government’s Komisiyo y’Igihugu y’Ubumwe n’Ubwiyunge is translated as the National Unity and Reconciliation (Ubwiyunge) Commission. These concepts were used in connection with each other and often interchangeably, reflecting the broader ethos of social harmony in the reconciliation discourse. I suggest that in routing people’s disputes through grassroots mediation, the Rwandan government intended not merely for mediation to serve as a dispute resolution technique, but rather for the principles and practices to stand in for reconciliation (ubwiyunge) and the national unity (ubumwe) on which the new Rwanda was ostensibly based.

I argue that we should consider these forums as harmony legal models. Nader defines harmony legal models as legal forums based on the principles of harmony ideology: “an emphasis on conciliation, recognition that resolution of conflict is inherently good and that its reverse—continued conflict or controversy—is bad or dysfunctional, a view of harmonious behavior as more civilized than disputing behavior, the belief that consensus is of greater survival value than controversy” (Nader 1990:2).

In Rwanda, law-based mediation combined the emphasis on unity with the threat of state-backed punishment, detailed in the written laws instantiating the institutions. This punitive force included consequences that directly impacted people’s daily survival, such as imprisonment, property seizure, and fines. The power to punish was explicit, as when authorities told people they could be sanctioned for not complying, threatening them with fines or incarceration, and it was also represented symbolically through the physical presence of signs that indexed state power, such as sashes worn by the mediators or the presence of armed prison guards. The ability to forcibly implement judgments and apply punishments was central to people’s experience of harmony legal models, and is what distinguished legally mediated decisions from informal, nonlegal alternatives. As one case participant observed: “In the past, if a man or woman had problems, they would go see the priest. Today it is changing, people seek more the rigor of the law, because these courts have binding force. If you don’t comply, then you are sanctioned for it. Going to a pastor, he will only advise you, you can either abide or decide not to. But in courts, there is administrative force.”⁹

I see law-based mediation in Rwanda as a form of government through community, as Rose puts it, in which governments identify “weak communities” as a problem needing to be solved through technical, governance solutions aimed at mobilizing those same communities to solve their own problems through self-government. Mediation in grassroots legal forums was a way in which people who live proximately were intended to become linked into a community or, as Rose calls it, a “moral field binding persons into durable relations” that emphasized emotional relationships and microcultures of values and meaning (Rose 1999:176). Specifically, law-based mediation aimed at shaping and directing people to manage themselves and, equally important, to manage their relationships to one another, to take on “active practices of self-management and identity construction, of personal ethics and collective allegiances” (Rose 1999:172–176). That is, gacaca, comite y’abunzi, and the legal aid clinic were designed, like other local-level mediation processes in other contexts (Abel 1982b:11; Merry 1993:58), to emphasize relationships and interpersonal conflict resolution skills, using mediation principles to (re-)create moral bonds and emotional relationships that were shattered during the political violence.

Considering genocide courts, mediation committees, and a legal aid clinic within one analytic frame is a deviation from existing literature on Rwanda, which typically focuses on genocide-related justice as analytically distinct from nongenocide disputes.¹⁰ Separating our analysis based on externally imposed jurisdictional distinctions obscures continuities both in the top-down social engineering goals of these forums and in how most Rwandans experienced mediation as a mode of power that cross-cut institutional frameworks. The more I observed court sessions and spoke with participants themselves, the more I saw that, contrary to one mayor’s claim that “comite y’abunzi has nothing to do with gacaca,”¹¹ these forums were indeed connected in people’s lived experience. In the middle of the first decade of the new century, Rwandans who sought legal redress were first routed through mediation practices that combined harmony and punishment—whether filing a case against a neighbor for looting property during the genocide or for defamation, whether testifying against an acquaintance for participating in killings during the genocide or against an employer for wrongful termination, whether seeking paternal recognition or inheritance. I am thus arguing that “an examination of the ordinary is just as important as the apparently extraordinary or exceptional” (Kelly 2008:353), not only in understanding how people live through political violence (as Kelly argued with respect to Palestine) but also in exploring how people cope with the aftermath of genocide.

Analyzing gacaca, comite y’abunzi, and the legal aid clinic through Nader and Rose draws attention to how mediation was a technique of governance intended to reshape postgenocide Rwanda into a particular kind of community, based on exhortations to unity that were justified as cultural. There is a paradox when community is the object of government where the idea of community is assumed to be natural and preexisting, yet is also being technologically enacted (Rose 1999:177). The sleight of hand that presents “community” as apolitical rather than constituted through contestation and suffused with power imbalances warrants investigation—especially in a context like Rwanda, where efforts to rebuild the social fabric in the wake of genocide were particularly pronounced, by a government that was increasingly authoritarian. That is, we have to understand postgenocide Rwanda not as having a priori cohesion but rather as attempting to use these decentralized legal forums to create unity. Other anthropologists have underscored the importance of denaturalizing the idea of “community” as a “repository of innate solidarity” at the heart of official postconflict interventions, as Theidon has argued with respect to Peru in the aftermath of decades of violence between the government and Shining Path (2009:296; 2012:266, 269). Similarly, Richard Wilson has critiqued the ways that in South Africa, the idea of ubuntu, an “expression of community, representing a romanticized vision of the rural African community based upon reciprocity, respect for human dignity, community cohesion and solidarity” became “a key political and legal notion in the immediate post-apartheid order” (Wilson 2001:9). What, we must ask, was at stake in using mediation as a tool of unity, backed by punitive law, in postgenocide Rwanda?

Mediation as Statecraft: Denaturalizing the Cultural Justification

A month after Alphonse’s trial, sitting adjacent to the same courtyard, an elderly umwunzi (mediator) held my gaze while explaining to me and my Rwandan research assistant about the cultural history of dispute resolution in Rwanda. He described:

When I was born in 1928, gacaca and abunzi were in place as the only judicial system. The goal of these institutions in the past was to unite and to reconcile people. They gave messages of love, coming together, socializing. In the past, there were not any of the current kind of regular courts; those only came with the colonial rule of the Belgians. There were the king’s courts, but even those were just a last resort in the hierarchy beginning with gacaca and abunzi. It was the duty of abunzi to follow up and make sure the two parties reconciled. People used to bring beer to share, or exchange cows, or even intermarry, and these exchanges created strong bonds. Gacaca and abunzi now are pillars for development. You must have judicial systems like that in place to reconcile and develop.¹²

This characterization was consistent with how the Rwandan government authorities legitimized the harmony legal models as, according to printed brochures and the government’s websites, “traditional community courts”¹³ based on “the inspiration of the traditional context of conflict resolution.”¹⁴ These authorities claimed that Rwandans historically had dispute resolution systems that eschewed the winner-take-all basis of Western courts, instead using respected elders to solve specific disputes between individuals and families to strengthen communal ties. Indeed, scholars writing on Rwanda note a long-term presence of conciliation-based, decentralized mechanisms for resolving disputes in the region now known as Rwanda,¹⁵ as I discuss more in Chapter 2, consistent with broader-reaching customary law practices across much of sub-Saharan Africa. This philosophy was consistent with how classic analyses of African adjudication often stressed that local courts in Africa, whether formal or moots, aimed to restore breaches in relationships among members in an ongoing community in order to restore equilibrium.¹⁶ Much like the classic ethnographic examples, postgenocide Rwanda’s harmony legal models had no lawyers, incorporated broad participation, and had minimal hierarchy among disputants, judges, witnesses, and other attendees, all consistent with a spirit of compromise rather than adversity. They used wide rules of evidence, given that the issue at hand was understood as a breach of deeply embedded relationships and practices, rather than narrowly focusing on a specific point of law. Finally, the judges or mediators at the center of these processes explicitly and clearly articulated social norms in an effort to socialize litigants as well as other participants. Obarrio (2014) has recently described similar community courts in Mozambique.

Yet, there are dangers in being seduced by cultural justification. Debating whether Rwanda’s postgenocide legal institutions are “customary” can disguise the degree of state involvement in so-called informal or popular mediation in the past and the present, and it can hide historical and contemporary power imbalances and coercion. For decades we have known that we cannot take the ethnographic record of African customary law at face value, given how the focus on compromise within it reflects primarily a “Durkheimian emphasis on harmony of interests and shared goals” (Starr and Yngvesson 1975:559). Even seemingly consensus-based customary practices contained hierarchy and coercion (Colson 1974; Moore 1986). Further, customary practices and institutions in Rwanda and elsewhere in sub-Saharan Africa changed significantly during the colonial period and after independence, specifically through increasing state control (as I describe in more detail in Chapter 2). Their revival in the twenty-first century needs to be understood not as a romanticized local movement but as an “invented tradition” (Hobsbawm and Ranger 1983) that served present state needs.¹⁷ A rich body of legal anthropology has shown how so-called popular or informal institutions that use mediation are typically tied up with, not separate from, state law and even expand state control.¹⁸ I agree with Lars Waldorf (2006) that there seems limited utility in clinging to terms like popular and informal in Rwanda, where the state is deeply implicated in gacaca in particular and the courts may in fact be deeply unpopular.¹⁹

Further, attention primarily to measuring degree of cultural relevance can mask the way these forums are bound up with broader-reaching global trends. Rwanda’s mediation-based legal forums were extensions of the truth commission model, popular worldwide since the 1980s, which typically foregrounds victim narratives and eschews punishment in an effort to “heal individuals and society after the trauma of mass atrocity” (Minow 1998:57).²⁰ Government billboards across Rwanda in 2004 and 2005 publicizing the launch of the gacaca process announced Ukuri kurakiza, “the truth is healing.” Rwanda’s emphasis on mediation through gacaca, comite y’abunzi, and the legal aid clinic reflected broader international trends in transitional justice, including a move away from exclusively using international criminal courts, a shift to national ownership of legal processes, and a shift toward valorizing cultural solutions (see Chapter 2).

Looking at Rwanda’s forums in comparative perspective with other harmony models underscores that though they often derive from local practices, harmony models are simultaneously “part of systems of control that have diffused across the world along with colonialism, Christianity, and other macroscale systems of cultural control such as psychotherapy” (Nader 2002a:32). The imbrications of Christianity with mediation in Rwanda are pronounced today. Rwanda has been predominantly Christian for more than a century, due to the missionary influence during colonial rule (Longman 2010a). Postgenocide Rwanda saw a resurgence in transnational forms of Christianity, particularly evangelical churches brought in from East and Central Africa by returning diaspora members, as well as by church congregants in missions from churches in Europe and the United States. Although churches were not formally involved in the legal process, Christianity served as a broader part of the sociopolitical context that reinforced mediation principles (e.g., Richters 2010). Both local pastors (some of whom served as abunzi or inyangamugayo) and international pastors advocated harmony principles within and outside courts in Rwanda.

In a notable example, beginning in 2005, President Paul Kagame enlisted the help of American evangelical Pastor Rick Warren, author of the best-selling Purpose Driven Life (2002), to collaborate in a five-year project aimed to turn Rwanda into the first “purpose-driven nation” (Morgan 2005; Van Biema 2005). In addition to meeting with government officials and visiting the hospitals his project supported, Pastor Warren—who would later go on to cement his status as a global moral leader by giving the invocation at President Barack Obama’s inauguration in 2009—held large revival meetings attended by thousands of Rwandans across the country, in which he preached about the importance of unity and forgiveness to restoring Rwanda (Musoni 2008b). Rwandan pastors regularly advocated for the mediation discourse from the pulpit, encouraging parishioners to confess and forgive, and even served as lay judges and mediators. Medical anthropologist Annemiek Richters has described mediation-like efforts in sociotherapeutic processes led by churches in northern Rwanda that similarly emphasized healing within the community (Richters 2010). Other scholars have noted this trend elsewhere. Wilson has argued that Christian notions of confession and redemptive healing were crucial to many people’s understandings of reconciliation in the South African Truth and Reconciliation Commission (Wilson 2001:109–120). Shaw (2007) and Kelsall (2005) have described how Christian elements were crucial to the Sierra Leonean TRC hearings, Abramowitz (2014) has described Christian discourses of harmony, reconciliation, and truth that accompanied the Liberian TRC; and Theidon (2012) has detailed the interrelationships between evangelical Christianity and the TRC in Peru.

Nader’s comparative examples also draw our attention suggestively to how mediation in Rwanda is linked to a third global trend: alternative dispute resolution (ADR). Harmony models, often in the form of ADR, have been a part of the package of the Americanization of law being exported around the world for decades, alongside the more retributive, adversarial dimensions of Western-style law.²¹ The shift to ADR involved a transition toward viewing law as healing, an approach that “transforms facts and legal rights into feelings, relationships, and community writ small” (Nader 2002a:131),²² consistent with the explicit focus on individual-level forgiveness and compromise brokered by community members in Rwanda’s grassroots legal forums. Laurel Rose (1996) has described being sent to fact-find about ADR approaches in Rwanda as early as 1994, and Nader has suggested that the World Bank, which supports programs in Rwanda, mandates the use of mediation to solve conflicts with many assisted countries.²³ Many states in the United States made ADR mandatory in the mid-1990s, much as gacaca and comite y’abunzi were made obligatory in Rwanda a few years later, and as mediation in legal aid became increasingly common worldwide (see, for example, Englund 2006:145–169). Paralleling the rationalization of ADR,²⁴ there was regular justification of Rwanda’s gacaca courts, comite y’abunzi, and legal aid clinic in terms of increasing efficiency and “unclog[ging] courtrooms” (Butamire 2010) consistent with nationwide strategic planning goals of decentralization (Ministry of Local Government 2008a, b), and I heard them framed as self-evidently better than, and in opposition to, adversarial law.

None of this is to suggest that there is no basis for claiming cultural roots of harmony in Rwanda, or that these three forums are identical to ADR. Rather, in the face of such clear interconnections between Rwanda’s postgenocide grassroots mediation-based legal forums and global trends, I want to examine what is at stake in the insistence that these forums are naturally Rwandan. Challenging the benign cultural framing of harmony legal models in Rwanda is vital, especially given the growing acknowledgment that the Rwandan government is increasingly authoritarian, manipulates identity politics, and uses law and “lawfare, the effort to conquer and control … by the coercive use of legal means” (Comaroff 2001:306) selectively as another strategy to consolidate the ruling elites’ power and suppress dissent (e.g., Chakravarty 2015; Hintjens 2008; Longman 2010b; Morrill 2006; Reyntjens 2005; Straus and Waldorf 2011).

Indeed, while the postgenocide Rwandan government claimed to promote a unified national identity that superseded ethnic divisions and officially banned ethnicity, its official history erased reference to deep structures of political economic exclusion and exploitation, and actually reinscribed division by defining citizenship categories in relation to the genocide: Hutu were génocidaires, Tutsi were victims (see Chapter 1).²⁵ Other hierarchies of power and inequality existed across dimensions of gender, class, age, or region. Further, national unity was promoted alongside the implementation of increasingly repressive laws that proscribed ethnic labels and criminalized “divisionism,” “revisionism,” or “genocide ideology,” anything seen as “marginalizing … degrading … [or] stirring up ill feelings” (Law N.18/2008 of 23/07/2008, Article 3). These laws served to reduce political dissent, especially when coupled with ongoing detentions, and state-backed violence at home and in the neighboring Democratic Republic of Congo (Thomson 2011c:442; Waldorf 2011:49).

I suggest that the cultural justification for harmony was an effort to insulate against international critique of, and interference in, Rwandan governance. For Rwandan leaders to claim they could rebuild Rwanda based on their own rather than foreign cultural principles served to protect against persistent allegations that domestic legal forums did not meet international legal and human rights standards (e.g., Amnesty International 2002; Penal Reform International 2004). This echoes how cultural justification for customary law was used in many contexts in Africa both to insulate institutions from colonial interference and to support nationalism at Independence (Chanock 1985:24; Colson 1974:77–78). It parallels Nader’s argument (1990a) that harmony can be a counter-hegemonic strategy, though in Rwanda the move appears directed not at colonial leaders or state law but at the hegemony of international regimes of law and human rights.

What was particularly problematic, I suggest, was the cultural justification for unity itself. Given that mediation was presented as a practice natural to all Rwandans, to reject overtures to “be mediated” or “reconciled” was to reject the unity of all Rwandans, and even to reject shared citizenship and the legitimacy of the state. By a decade after the genocide, many scholars argued that Rwandans “rehearsed consensus” (Ingelaere 2010a:53) or “pretended peace” (Buckley-Zistel 2009), which papered over much deeper divisions (Thomson 2013). This is consistent with Nader’s caution that harmony ideology “may be used to suppress people’s resistance, by socializing them towards conformity by means of consensus-building mechanisms, by valorizing consensus, cooperation, passivity, and docility, and by silencing people who speak out angrily” (Mattei and Nader 2008:77).²⁶ Recent research underscores how most Hutu who had been in Rwanda during the genocide felt that strident resistance to gacaca risked conveying criminally divisive intentions.²⁷ Many Hutu men and women described to me confessing and requesting forgiveness under duress, fearing that not to do so would invite arbitrary accusations, false convictions, and heavy sentences, as well as other social sanctions. For example, as Alphonse returned home from prison after his gacaca trial and took on the role of model reformed perpetrator, he furtively told me that he remained cautious in what he said and did, under watchful eyes. Similarly, victims often described feeling pressured—by defendants, by inyangamugayo, by pastors, by neighbors—to forgive perpetrators or to accept less than they were due in property cases, for the benefit of community and country. The coercive potential of harmony in gacaca was particularly strong, perhaps inescapable, because of how tightly it linked unity principles to state punishment. People were essentially told to forgive and reconcile with a thinly veiled “or else”: risk state-backed punishment as someone who did not value unity, and who therefore probably embraced divisionism as a propagator of genocide ideology and an enemy of other Rwandans and the state.

A growing anthropology of transitional justice has similarly shown how coercion and silencing are core elements of many truth and reconciliation commissions, in sub-Saharan Africa and elsewhere. Wilson has argued that the South African TRC’s invocation of ubuntu could “coerce individuals into compliant positions that they would not adopt of their own volition” (2003:189), and Ross (2003) has shown how despite its purported attempts to give voice to victims, the TRC silenced and depoliticized women in particular through its narrow institutionalized definitions of violence. Shaw (2005) and Kelsall (2005) similarly showed how Sierra Leone’s TRC, while ostensibly more attuned to people’s needs than the United Nations–backed Special Court of Sierra Leone, overwrote cultural practices of forgetting through the Judeo-Christian valorization of talk, which rendered the TRC at worst coercive, at best irrelevant. Abramowitz has argued that the Liberian TRC “served to close doors for personal expressions of grief and memory” (Abramowitz 2014:209). Theidon has similarly illustrated how the “victim-centered” approach of the Peruvian truth commission inadvertently created “resentful silences” (Theidon 2010:110) and “systematic distortions” (Theidon 2012:28) that flatten and depoliticize people’s stories. Ethnographic work in other contexts in sub-Saharan Africa has shown how local practices to cope with the past and reintegrate ex-combatants get overlooked, distorted, or erased by a focus on the standard transitional justice toolkit, as in Angola and Mozambique (Honwana 2005; Irega and Dias-Lambranca 2008), Uganda (Allen 2008; Finnstrom 2010), and Burundi (Nee and Uvin 2010).

Overall, Rwanda’s harmony legal models located the solution to the genocide in interpersonal relations, not in structural and political or economic changes.²⁸ In the aggregate, gacaca courts, comite y’abunzi, and the legal aid clinic excluded most Rwandans from having access to Western-style courts, lawyers, and professional judges. Though the harmony legal models were framed as beneficial to reduce unnecessary conflict and promote unity, they meant that in practice governing authorities attempted to impose resocialization through harmony principles on mostly rural, typically poor, and less educated Hutu and Tutsi survivors, thus reinforcing a divide between elites and nonelites, urban and rural. With less economic, social, and cultural capital to negotiate postgenocide Rwanda’s multifaceted and shifting legal terrain, these people typically had little option other than to comply with mediated solutions. Women often had even fewer options, or felt the brunt of mandatory unity principles more strongly, as I illustrate across the later chapters. These effects are consistent with how Nader has extensively argued that the focus of harmony models on relationships over root causes ignores political economy and deflects cases away from the courts that threaten to challenge the status quo, thus serving as an antilegal movement restricting certain people’s access to justice (Mattei and Nader 2008:18; Nader 1988:280, 1999b:107, 2002a:53) and serving as a tool of “pacification and control” (Nader 1999a:305, 2002b:37, 42).

My analysis here is consistent with Juan Obarrio’s recent book describing similar exclusionary dimensions of the “spirit of the laws” in postwar Mozambique, though with an important distinction: by illustrating how these dynamics are not isolated to legal forums framed as customary law (such as gacaca courts and comite y’abunzi) but are also found within a legal aid clinic rooted in Western-style law, I suggest we identify it as mediation-based citizenship rather than “customary citizenship” (Obarrio 2014:109). Obarrio argued that in Mozambique, “While a small minority of the population had access to the mechanisms of the urban official judiciary, the vast majority of the population in rural and periurban areas was both included and excluded from national citizenship, through the enactment of localized forms of rights, which blended official state and customary norms” (Obarrio 2014:174–175). He names this effect “customary citizenship,” which “articulate[es] elements of official law, custom, and kinship” where “acquiescence to official law is filtered through ‘traditional’ norms and the subjection to the authority of customary chieftaincy” (2014:109). Akin to my claim that law-backed mediation in Rwanda is a central element of post-genocide governance (Doughty 2014, 2015), Obarrio asserts that customary citizenship is not a derivative effect but one of the “main defining features of the political in contemporary Mozambique” (2014:11). Ultimately, Obarrio and I agree that grassroots legal processes are a defining feature of how people are unevenly subject to the law in contemporary Africa. Where I disagree is in naming the exclusionary process as “customary” citizenship, which I fear obscures how the effects can occur in institutions explicitly not defined as customary, and risks further reifying “customary law” rather than interrogating the category itself.

Beyond Coercion: Negotiating the Micropolitics of Reconciliation

While it is crucial to recognize the potential for, and experience of, coercion within mediation in Rwanda, as the examples throughout the later chapters illustrate, equating mediation in Rwanda with what Nader calls “coercive harmony” (2005) across a diversity of forums and types of disputes oversimplifies dynamics of power in Rwanda. The examples throughout this book remind us that we must attend to how coercive state power coexists with resistance (e.g., Foucault 1978; Scott 1985; Scott 1990) to avoid mischaracterizing how law-based harmony operates as a mode of power. The governmentality critique of mediation, following Rose’s government through community, risks overstating the seamlessness of how people self-regulate according to official policy (see Briggs 2013). That is, it can suggest that everyone from government authorities to case participants performed unity principles, and did so uniformly. Instead, as the examples in the following chapters show, participants and even mediators took on mediation practices to varying degrees, contesting the hegemony (Nader 1990) of mediation and imperfectly self-regulating (Rose 1999) along mediation principles.

Alphonse’s seemingly total embrace of repentance for the greater good was an exception, as evidenced by the discomfited reactions of the other prisoners seated alongside him. The following chapters show that Rwandans were not simply “pretending peace” (Buckley-Zistel 2009) or “rehearsing consensus” (Ingelaere 2010a) in these forums; rather, they contentiously expressed disagreement with respect to principles of collective cohesion, individual rights, the common good, and ultimately moral community and belonging. The cases I illustrate in the coming chapters, like Beata and Joselyne’s case, involved heated contestation and disagreement rather than straightforward accordance with harmony principles. People situated in a variety of positions alternately enacted mediation principles or contested them at varying moments across a given case discussion, and dynamics of power shifted across moments and in relation to different charges and allegations. Unity in Rwanda was not a dominantly expressed legal ideology among disputants of the kind Nader (1990) and Laurel Rose (1992) have described in other contexts. The two recent ethnographies of postgenocide Rwanda—Jennie Burnet’s Genocide Lives in Us (2012) and Susan Thomson’s Whispering Truth to Power (2013)—both provide rich detail of the sources and expression of disagreements in postgenocide Rwanda, among everyday Rwandans as well as between people and state policy. Thomson explicitly analyzes these disagreements as everyday acts of resistance to the nationwide policy of unity and reconciliation.

I am thus suggesting that we must analyze the historical and political-economic context, to be sure, but not allow it to obscure the smaller-scale interpersonal interactions that make up ordinary life. This argument is consistent with much recent work by anthropologists on warscapes who argue that focusing on violence and the macro-political struggles of war as the “only primary feature” of life in war zones overlooks, as Sharika Thiranagama writes, “myriad mundane (and perhaps more severe for being mundane) experiences of life in protracted war zones,” such as Sri Lanka (Thiranagama 2011:5).²⁹ I suggest here we can extend that analysis to the so-called postconflict period, especially in a context like Rwanda, where, in many people’s perspective and experience, protracted political violence has not ended but has merely transmuted. Stephen Lubkemann has argued, based on work in Mozambique, that exclusive attention to violence and state-level political dynamics, and to interpreting people’s reactions as a response to national-level macro-political struggles, erases attention from the “other struggles” related to “gendered or generational power configurations within households, to seeking socioeconomic advantage, and to the pursuit of redress for personal grievances” (2008:162) that shaped people’s lives during war. Those same dynamics in postgenocide Rwanda, I fear, are overlooked through overemphasis on silencing, coercive effects of state-backed unity programming.

Throughout this book, I therefore attend to how people maneuvered within Rwanda’s legal architecture of social repair, which is a way of conceptualizing how legal forums and logics constructed particular concepts, aesthetics, and practices of time and space, some stated and some implicit, in which Rwandans maneuvered, during the period of my fieldwork, 2004–2008. I build here on Abramowitz’s description of how the “architecture of post-conflict time” in Liberia created a liminal postconflict space and time in which the fundamental rules of the social order were being reworked (Abramowitz 2014). In Rwanda, the architecture of postgenocide legal time was likewise active, and intentionally transformative. It created a set of processes into which all Rwandans were drawn, to varying degrees.

I illustrate in the following chapters how people used these spaces to debate the terms of social connection and collective belonging, what Theidon has called “reconstructing moral orders” in a “micropolitics of reconciliation” (2006:436). Alphonse’s case, like many others that I present here, involved situated debates over how much people should compromise and sacrifice for each other, and thus served as part of an ongoing debate about the terms of moral community among Rwandans. We have long known that in legal disputes people debate over not simply circumstances of disputes but normative paradigms and the meaning of relationships (see, e.g., Comaroff and Roberts 1981; Merry 1990). Further, we know that discussions within courts are shaped by, and shape in turn, power dynamics and relations of dependence in the broader social context—Lubkemann’s “other struggles,” or what Sally Falk Moore has called the “micropolitics of local social standing” (1992:33). Ben Jones, for example, has recently shown how discussions in village courts in rural Uganda were framed by political debates about the violence, and how people used them to settle private disputes, articulate public concerns, and develop ideas of seniority and propriety (Jones 2008:65).

I suggest that the focus on oral testimony across gacaca, comite y’abunzi, and the legal aid clinic—whereby sessions involved what I call “contextualized conversations,” which I discuss more in Chapter 3—further heightened these relational and political dimensions of testimony. I suggest that we understand discussions in mediation sessions as episodes in ongoing negotiations over how the details of ubumwe and ubwiyunge—the unity and reconciliation often proclaimed by the Rwandan government—played out in practice in people’s lives, at different interconnected levels of scale. The polysemy of the term “unity,” which was invoked as a ubiquitous but vague principle on which to make legally binding judgments and allocate resources, opened up space for discussion. What resulted, I argue, was not the creation of official normative versions of national belonging, nor monolithic coerced silence, but rather a terrain in which people contested moral community.

One significant component of such negotiations in legal forums was that people confronted head-on what Theidon calls the “economics of memory”: how in an industry of truth telling and postconflict justice people instrumentally “narrated with new possibilities and aspirations in mind,” including “hope for some economic relief” (Theidon 2012:109). Catherine Bolten has similarly underscored the importance of “material loyalty—relationships forged and sustained in complex, often compassionate acts of resource exchange” (Bolten 2012a:4) in explaining how people survived the civil war in Sierra Leone and rebuilt in its aftermath. Rwandans’ discussions in legal forums confronted the material and financial dimensions of living together, which have long been understood as central to reconciliation but are often sidelined in prevailing discourses about reconciliation that emphasize affective dimensions, moral imaginaries, and intersubjective communication.³⁰ Collective belonging in postgenocide Rwanda was intimately bound up with land and other assets, continuous with earlier time periods (De Lame 2005:165). Exchanges and transactions of material goods and labor have long been tools of social integration in Rwanda (De Lame 2005:15, 105, 305), and they were at the heart of debates in legal forums, as I describe further in the social-history overview in Chapter 1 and in the discussion of gacaca in Chapter 3.

I also draw attention throughout to the ways gender roles, relations, and ideologies were sites of contestation and renegotiation through law-backed mediation.³¹ Rwandans, particularly in rural areas, were heavily influenced by normative ideas of gender both before and after the genocide (Burnet 2012; De Lame 2005; Jefremovas 1991; Taylor 1992). In her recent ethnography that provides the most sustained ethnographic analysis of gender dynamics in contemporary Rwanda, Burnet argues: “Women are viewed positively when they are reserved, submissive, modest, silent, and maternal, when they maintain a ‘respectable’ household, and when they raise ‘wise’ children. They are viewed negatively when they gossip, are loud and overly emotional, or have a dirty house or rude children. By contrast, men should be self-assured, dominant, logical, brave, and physically strong” (Burnet 2012:44). These assumptions about gender roles permeate the examples throughout the book, captured also in the categories of “virtuous wives, “timid virgins,” and, more pejoratively, “loose women,” which Villia Jefremovas (1991) described in pregenocide Rwanda. The cases I explore throughout the book show how people brokering compromises in grassroots legal forums tended to adhere to these essentialist or naturalized roles for women (as wives, mothers, or conflict avoiders), a framing that could circumscribe possibilities for individual women, even while it sometimes helped them to achieve specific legal rights.

Specifically, I draw attention to how people negotiated the socially embedded, relational nature of their rights, which is a feature of African customary law more widely (Weeks 2013). Rwandan women’s lives were shaped by patrilineal kinship norms in which unmarried girls, married women, and widows derived their social identities as well as their rights to access to resources from the men to whom they were related (Burnet 2012:43). In the postgenocide context, even as efforts were made to improve women’s rights and individual security—including changes in national law to assure ownership rights for women, increased legal protections, and more representation by women in government decision-making bodies—women remained vulnerable, especially in rural areas, in part because of persisting normative assumptions that women’s access to resources were linked to unequal relationships to particular men, especially in a context of widespread death and dislocation. As Burnet explains, “Women’s economic lives were circumscribed by contemporary interpretations of patrilineal inheritance systems whereby women remained economically harnessed to economic units controlled by male relatives” (2012:76). Yet, while the relational nature of women’s rights could render women more vulnerable, it also increased space for negotiation, as women debated prevailing assumptions about care and obligation within these social categories. Burnet argues that it is precisely women’s role as the “primary social mediators between households, kin groups, and community” (2012:13) that gave them some space to help transform Rwandan society.

Finally, I show one of the ways legal forums shaped conceptions of ethnicity in postgenocide Rwanda, by creating what I call “genocide citizenship.” I build on work by contemporary anthropologists who explore how ethnic logics persist in Rwanda today, as ethnicity remains one among many salient factors of Rwandans’ identity, intertwined with other forms of identification, such as class, gender, age, and region (Eltringham 2011; Hintjens 2008). I supplement their focus on ethnicity in everyday life, including among youth (McLean Hilker 2009, 2011, 2012), educated urban residents (Eramian 2014a, 2014b), women (Burnet 2009), and members of the diaspora (Eltringham 2004) by showing how ethnicity shapes, and is shaped by, discussions in legal forums. As I discuss more fully in Chapter 3, I argue that within the postgenocide context marked by the official erasure of ethnicity, legal forums perhaps paradoxically solidified social categories linked to one’s position with respect to the genocide, which were formalized by the state and carried access to state-backed resources (hence serving as a form of genocide citizenship).

Overall, this book adds to contemporary anthropological scholarship on Rwanda by providing detail on how people maneuvered within the legal architecture of social repair that was such a pronounced part of Rwanda’s postgenocide reconstruction efforts, showing how law-backed mediation shaped the meanings of institutions and practices such as exchange, gender, and ethnicity in other social fields. Analyzing mediation practices, specifically how they combine calls for unity with punishment in legal forums, allows us to attend to the processes through which rule of law is produced and naturalized in a “moral economy of justice” (Clarke 2009), and to how legal categories create and acknowledge specific categories of guilty and innocent personhood (Kelly 2011; Niezen 2010), while also attending to how people “search for normal” (Abramowitz 2014) and reestablish collective belonging across, within, and containing deep divides following violence (Bolten 2012b; Finnstrom 2010; Hromadžić 2015; Ring 2006; Schuetze 2010; Theidon 2006). Participants expressed their knowledge of not only the fictions of justice (Clarke 2009) but also the fictions of harmonious community, showing that people are always contesting what it means, and their place within it. Conversations in legal forums shed light on the contested cracks at the core of official ideas about justice and unity.

Consider the following excerpt from a later portion of Alphonse’s trial. An older woman I call Grace stepped forward to testify, indicating she would speak as umuhohotewe, one who was victimized (a privileged designation distinct from abatangabuhamya, those who testify as witnesses). Grace explained that during the genocide Alphonse had taken her children to protect them but then had demanded money from her to pay the killers to spare their lives. She asked to be reimbursed for the cash, approximately $6. Alphonse responded to her charges by admitting to receiving the money and describing how he had taken it to a local authority in order to procure false Hutu identity papers for the children. As Grace renewed her demands, one of the inyangamugayo, whom I call Faustin, himself a genocide survivor, chided her that she should be grateful that her children had survived. Rather than pressing Alphonse for money, Faustin continued, Grace should recognize he had done her a favor. People in the assembled crowd laughed aloud at Grace’s claim and at Faustin’s gentle remonstrance.

In this example like many others, the inyangamugayo had to actively intervene to advocate compromise. Yet, the inyangamugayo pressuring Grace had experienced the same violence as she had, had also lost family members, and currently lived in an equivalent position of poverty. These shared structural and experiential dimensions do not erase coercive potential, of course, but they suggest to me there is limited utility to an analysis that reduces every interaction to coercion or that distrusts any performance of unity and forgiveness as disingenuous. Similarly, even as some participants suggested Alphonse’s confession was coerced, other participants implied that he may have been the one exerting pressure on others, such as Grace. I fear that focusing analysis on measuring coercion risks reproducing the hierarchy of suffering contained in the government’s dominant narrative of the genocide, which silences and delegitimizes particular forms of experience (Burnet 2012; Doughty 2008; Lemarchand 2009).

Likewise, in the comite y’abunzi example with Beata and Joselyne, while we can and should ask how coercion shaped the women’s participation, we need to move beyond coercion to explore the complex dynamics around why Joselyne would come to a legal forum to address Beata’s accusations that she engaged in sexual acts with passing men, or to examine what was at stake in the kinds of logics and language that mediators used to reason with the participants. Similarly, the employment dispute between the watchman and the nuns before the legal aid clinic does not seem to be adequately captured as a “crime of exploitation,” which Harri Englund writes of in connection with legal aid as mediation in Malawi (2006:123–169). More generally, Rwandans routinely told me that they perceived the call for compromise within comite y’abunzi and the legal aid clinic as relatively benign, because individuals, not the state, served as plaintiffs and there was no armed police presence. People in structural positions that tended to render them vulnerable through gacaca sometimes felt the abunzi or the legal aid clinic staff could, in fact, challenge status quo power relations, such that the people likely to feel coerced were those with comparatively high economic, social, and political capital.

Overall, I suggest that it behooves us to move beyond the broad brush of coercive harmony, to attend equally to the fraught and iterative processes through which people varyingly took on and contested the self-regulating practices of mediation. My attention to coercion and negotiation may seem part of a theoretically dated discussion of structure and agency, yet these themes remain a central conundrum in contemporary scholarship on Rwanda (Burnet 2012; Clark 2010; Straus and Waldorf 2011; Thomson 2013), and in Rwanda itself, thanks to the strongly centralized, and increasingly authoritarian, state. Ultimately, I see acts of agency in Rwandans’ use of grassroots legal forums, wanting to ensure I do not reduce Rwandans who participate in these forums to automatons, even as I recognize the deeply controlling and threatening nature of the state, and the way state power penetrates people’s ordinary lives, as Thomson (2013) has richly described. Instead, I ask, how do people contest meanings of gender, guilt and innocence with respect to the genocide (“genocide citizenship”), access to resources, family, and ethnicity within postgenocide Rwanda’s architecture of social repair? Examples throughout this book raise complex questions about why, when, and how people advocate for and against unity, and what, ultimately, is at stake in the debate over the legitimacy of mediation principles. The ethnographic chapters that make up the body of the book explore these contested negotiations in order to examine how people attempted to remake moral communities within frameworks of mediation.

Notes on Fieldwork and Field Sites

This book is based on eighteen months of ethnographic research I carried out in Rwanda between 2002 and 2008, including participant observation, interviews, and document analysis, conducted in Kinyarwanda, French, and English. I attended fifty-six gacaca sessions during an extended stay from July 2007 through July 2008, as well as fourteen mediation committee sessions and twelve legal aid clinic sessions. I supplemented attendance at gacaca sessions with repeated semistructured interviews with a cross-section of people in each area, including more than thirty gacaca judges, mediators, local authorities, and religious leaders. These numbers do not include the countless informal conversations I had with residents. For broader context, I conducted research in Kigali and at the International Criminal Tribunal for Rwanda (ICTR), including more than eighty interviews with NGO and ICTR staff and members of government in Kigali. I have changed all names in the text.

In order to work with the intended target of grassroots courts—most specifically the rural population, abaturage—and recognizing the importance of interacting with rural Rwandans rather than political and economic elites in the capital (De Lame 2005; Ingelaere 2010b:54; D. Newbury and C. Newbury 2000; Thomson 2010), I worked primarily in two field sites in southern Rwanda, one rural and one a midsize town, supplemented by periodic gacaca sessions in the neighborhood of Nyamirambo in Kigali. People in Ndora and Nyanza lived lives broadly representative of Rwanda’s predominantly rural population—people in these sites participated in the same legal forums that were implemented nationwide, and they were broadly representative of national trends in economic activity (primarily agriculture), educational level (primary school), and religious activity (predominantly Christian). That is, their daily lives were recognizable to most rural Rwandans, and the problems they faced were familiar, in their reliance on agriculture, their targeting by government programs, and their recovery from genocide. Thus, my findings can be seen as suggestive beyond specific hillsides, even though, given the salience of regional variations in Rwanda (Burnet 2010; De Lame 2005; C. Newbury 1988) and how local dynamics of history and violence shaped the specific hurdles faced in rebuilding, I do not suggest my findings were uncritically identical for all areas of Rwanda.

South Province

The South Province is one of five provinces constituting Rwanda. The genocide plan faced initial resistance in the South Province in 1994. For the first two weeks after President Juvénal Habyarimana’s plane was shot down, the region remained calm, as a handful of administrators—both Hutu and Tutsi—worked to suppress conflict amid national calls for attacks. The late historian and human rights activist Allison des Forges hypothesized, “Perhaps this reflected the history of the area, the heart of the old kingdom, where bonds between Tutsi and Hutu were multiple, long-standing, and strong, disposing the Hutu to defend Tutsi more vigorously. Remote from the major military posts, resisters in the region also had more time to organize their efforts before substantial military force was brought against them” (1999:496). Before the genocide, the region had high rates of interethnic marriage and a higher percentage of Tutsi than the national average of 15 percent. For example, Butare, the largest city in the South, home of the National University of Rwanda and the former colonial capital, was a quarter Tutsi before the genocide, higher than Kigali’s 17 percent (Des Forges 1999:353, 432).

Killings only began when the genocidal regime sent militias from the capital, at which point the bloodshed escalated to new levels of scope and intensity (Des Forges 1999). Ultimately the violence was perpetrated by militias as well as civilians, people from inside the area and people who came from outside. Thousands of Burundian Hutu refugees, who had been living in the province since fleeing national massacres in 1993, participated once the situation deteriorated. Confusion increased further as tens of thousands of Rwandans passed through on the way to Burundi or the Democratic Republic of the Congo (DRC) for safety. This history—which combined coexistence, protection, and catastrophic violence—made the South Province a particularly compelling place in which to learn how people more than a decade later were rebuilding their lives, particularly using legal forums and understandings of the past to shape collective belonging. By 2004, people living in the South Province represented a range of experiences from the genocide, including those who lived through it and others who returned after living as refugees or citizens in Burundi, the DRC, Uganda, Kenya, Tanzania, or Europe. The province’s population was diverse across class, from farmers to civil servants to university professors, as well as across religion.

The South Province’s population highlighted the challenge of rebuilding the social fabric across varying lines of identification. Grassroots legal forums in the South Province had high hurdles to overcome in terms of creating community. The province was among the worst affected by the genocide, where the consequences seem to have had the most severe and enduring effects (Des Forges 1999; Ndangiza 2007:4). Despite representing 25 percent of the population, the province housed 47 percent of cases of people accused as planners of the genocide (4,357 cases out of a total of 9,362), who were tried before Western-style domestic courts.³² The province also had the most gacaca cases. Further, according to a nationwide 2007 survey, it had the highest levels of self-reported poverty, where 68 percent of people lived below the poverty line (Frw 90,000, or U.S. $180, per year), and 90 percent self-identified as somewhat poor, rather poor, or very poor (Ndangiza 2007). The same survey showed that the South Province had the lowest levels of meat consumption and mattress ownership in the country. Southerners also expressed the strongest preference for continuing to live on agriculture rather than engaging in land reforms, which were aimed at the “transformation of agriculture into a productive, high value, market-oriented sector” as part of the national economic development plan in Vision 2020 (Ministry of Finance 2000:3).

People’s faith in and respect for the legal institutions which privileged their neighbors’ knowledge, and which depended on lay judges and their neighbors’ testimony, could not be taken for granted. According to the National Unity and Reconciliation Commission’s 2007 report, compared to the rest of Rwanda, inhabitants of the South Province were more skeptical of decentralization and had “lower levels of interpersonal trust and a perceived lower ability to work together among citizens” (Ndangiza 2007:4). Thus, government through community in this province was particularly visible in people’s efforts to create or defy binding relations and a shared moral field.

Ndora

Ndora, the sector capital of Gisagara district, was a rural agricultural community approximately eighteen miles south of the university town of Butare, near the Burundian border. The population as of December 2007 was 20,340 people.³³ Ndora was one of thirteen sectors in Gisagara district (which had a total population of 262,426).³⁴ A typical rural sector capital, it had a weekly market, a small town center with a handful of individual storefronts, a primary school, a health clinic, and several churches.

As of July 2008, when the main ethnographic research for this book was completed, as the gacaca process was winding down, 2,842 gacaca cases had been completed in Category Two (killing) in Ndora, and 5,824 cases had been completed in Category Three (crimes against property). Thirteen cases were outstanding on appeal, with thirty-three undecided requests for appeal. In general, officials and the population in Ndora reported that the gacaca process had proceeded relatively smoothly, though all expressed concern and frustration at the lack of resolution and enforcement of judgments in property cases as an ongoing problem. The comite y’abunzi in Ndora sector heard approximately 125 cases between June 2007 and May 2008. About 25 percent of those cases were refiled in the first-instance ordinary courts at the sector level.³⁵

The patterns of genocide in Ndora were similar to elsewhere in the country, in terms of how from April to July 1994 groups of civilians armed with traditional weapons (farming machetes, sticks studded with nails, sharpened sticks, or knives) and occasionally guns chased people out of their homes or fields where they hid and killed them publicly, during daylight as well as at night. Civilian militias set up roadblocks to prevent people from escaping, and required passersby to show identity cards—if marked Tutsi, that meant death. As elsewhere in the country, there were collective killings at Gisagara district’s health clinic and a church. The violence in Ndora was committed both by residents and by civilians and soldiers from outside.

Nyanza

Nyanza was an administrative town located on the main road from Kigali to Butare. It was the seat of the monarchy from 1899 to 1961, and a strong Tutsi presence remained there afterward (Des Forges 1999:353). It was named the capital of the South Province after decentralization in 2006, which brought new infrastructure and administrative development. The town of Nyanza had a population of fifty-six thousand, and the district as a whole had a population of 239,707.³⁶ I attended gacaca in Busasamana sector, which encompassed the center of town.

Given its size and central location as compared to Ndora, Nyanza had a more dynamic economic sector, before and after the genocide. Its market was vibrant and open daily, and the town had long been home to a hospital, several primary and secondary schools, courts, and offices of national companies like Electrogaz. The main national dairy plant, Nyabisindu Dairy, was located there, as well as a national museum at the home of the former mwami (king). Nyanza had several churches and a mosque. It was easily reached by public minibus, which ensured regular movement of people in and out, and it had several small hotels. Its population was more diverse and well educated than that of Ndora, with people having relocated there from other regions of Rwanda, with livelihoods not exclusively based on agriculture. Gacaca sessions for Busasamana sector were heard across the road from the High Court, up the hill from the newly opened Institute for Legal Policy and Development. In Busasamana sector, by mid-2008, 2,206 gacaca cases had been submitted in Category Two, 2,079 cases had been submitted in Category Three, and 538 cases were appealed.³⁷

During the genocide, the bourgmestre of Nyanza, Jean-Marie Vianney Gisagara, a Hutu, refused to ally himself with the Hutu power movement, and he helped the area resist the genocide in the initial weeks. He repelled attacks against Tutsi from outsiders, and kept order locally, punishing and imprisoning those who attacked Tutsi (Des Forges 1999:469). The military found and killed Mr. Gisagara in late April, at which point the massacres intensified, and victims attempted to flee (Des Forges 1999:497). The patterns of violence in Nyanza were similar to those in Ndora and elsewhere nationwide, in terms of how perpetrators included a mixture of civilians and militias, and how they chased victims out of their homes and rounded up and killed them in central locations like hospitals or schools. Administrators and civilians set up a network of roadblocks to prevent escape, manned by armed civilian militias. Dynamics of violence further intensified as the Rwandan Patriotic Front (RPF) gained ground in Rwanda, bringing increased military skirmishes between the RPF and the army around Nyanza beginning in May (Des Forges 1999:585). Gacaca sessions in Busasamana included extensive testimony about events at local hospitals, orphanages, schools, and churches, as well as the actions of individuals at roadblocks.

Gacaca in Busasamana was complicated by the movement of people in and out of Nyanza during the genocide. This meant that victims and killers often had not been, and were not now, residents of the area. Some were only passing through Nyanza in 1994. Many accused perpetrators had to be summoned from where they now lived in Kigali, for example, or others could not be located and were tried in absentia. This challenged the idea, implicit in gacaca, that cases were tried among people who knew one another then and now. The composition of gacaca sessions in Busasamana demonstrated the more fluid, contingent constellations of people who lived together during and after the genocide. Another issue complicating gacaca in Nyanza was the military’s heavy involvement there, meaning many victims’ deaths had no obvious perpetrators to accuse before gacaca. Many cases involved complicated discussions of complicity—for example, how responsible a defendant was for handing over a victim to the soldiers or for cutting down bushes to inhibit Tutsi from seeking shelter, whether by independent initiative or on authorities’ orders.

Property cases were a significant concern in Nyanza, as looting was rampant. More expensive property was taken in towns, and so the resolution and enforcement of these cases was as difficult as in Ndora, if not more so. Overall, alongside the higher education level of participants (though still only a minority had completed secondary school) and the more cosmopolitan nature of economic and social connections, there was more skepticism in Nyanza than in Ndora about the suitability of customary-style law for handling genocide crimes. This skepticism translated to comite y’abunzi as well. People often tried to find ways to circumvent comite y’abunzi and take their complaints, such as a failure of contract in a business transaction, straight to courts. Often these disputants had only a passing acquaintance with one another.

Methods, Analysis, and Ethics

When I began my fieldwork in 2002, there seemed to be no way to avert one’s gaze from or close one’s ears to the discourse about gacaca, or the profound social force of law. New training facilities were established, including faculties of law at the growing number of universities in Rwanda. A Bar Association was created in 1997 for the first time, and it grew from thirty to more than three hundred members by the end of 2007 (Kimenyi 2007). Existing courts and legal structures were refurbished and new ones created, including construction of new buildings. In my specific field sites in 2007–2008, a new Institute for Legal Policy and Development was built and opened in Nyanza, and two new court buildings were constructed in Nyamirambo. The ability of “rule of law” to create order was a visible sign of the law’s presence as a form of social control used to regulate the conduct of citizens (Foucault 1982; Rose 1999). This was particularly clear in Kigali, which residents and visitors described as orderly and safe in comparison with regional cities like Kampala and Nairobi. New laws took many forms, such as banning the use of plastic bags, prohibiting the raising of animals within city limits, and removing beggars and ad hoc venders from the streets. While in 2005 the ubiquitous moto-taxi drivers were bare-headed, by 2007 they and their passengers wore newly mandated helmets, and even the most regally dressed women were required to straddle the bike, prohibited from riding side-saddle. Minibus drivers respected passenger capacity limits, and their money collectors no longer leaned awkwardly over tightly packed passengers but instead took their own seats. Across the country, groups of pink-clad prisoners engaged in work projects, whether constructing buildings or digging drainage ditches, a constant reminder of the state’s power to lawfully incarcerate. Police officers lined the tarmacked roads, ensuring compliance with new traffic regulations, such as speed limits and seat belt requirements. Billboards across the country advertised the impending launch of inkiko gacaca.

This effect was even more pronounced when I came to Rwanda for the extended period of my research, from July 2007 through July 2008, midway through the gacaca process. Thus, even though I did not intend to study legal processes when I first arrived in Rwanda in late 2002, nor when I began preliminary work as a graduate student in 2004, I was quickly drawn in. Of course, it has become an anthropological trope to assert that one “did not go into the field expecting to study xyz” but rather “stumbled upon it,” as a means of performing authenticity and expertise. I revert to this explanation here in part to reflexively illustrate the palpable force of “law” in this period but more centrally to make explicit the temporal framing of this research. The unique temporal dimensions of my research came into even clearer relief when I returned briefly to Rwanda in 2013, several years after the close of the gacaca process, when the rhythm of life was no longer marked by regular interruption of the “state of exception” (Agamben 2005) of gacaca days. This book tells a story tightly linked to a particular time period, even as I make claims that may extend beyond it.

Many scholars have noted particular challenges of conducting research in postgenocide Rwanda, including government interference in research (Burnet 2012:25; Chakravarty 2012; Ingelaere 2010a:42, 50; Pottier 2002; Thomson 2010:22) and people’s subsequent “selective telling” (King 2009:128). They note that “simply asking ordinary Rwandans about their experiences before the gacaca courts is unlikely to produce reliable results” and emphasize that “researchers must be able to recognize the government’s version of events” (Thomson 2011a:375–376) in order to avoid “taking the ‘mise-en-scene,’ or stage-setting for granted” (Ingelaere 2010a:42). As a close Rwandan friend and colleague commented, “People will often just tell you what they think you want to hear.” These challenges are arguably best met by ethnographic methods (Burnet 2012:33), with the focus on long-term engagement, repeated social interactions across diverse social fields, and participant observation as well as interviews.

Focusing on specific sites over an extended time allowed me to contextualize the social interactions in the gacaca courts as part of broader life. I spent several days a week in each field site over a full year, and so I became a familiar presence. In analyzing case sessions, I took note of who was speaking, what and how it was said, and how people spoke across sessions, and I compared and contextualized their comments with other conversations and observations in and outside sessions, and over time. Consistent with ethnographic studies of law, I studied legal forums “in context,” aiming for “thick understandings of law in everyday life” (Nader 2002a:71). This meant, for example, not looking at the courts as autonomous institutions but seeing how people and disputes moved in and out of these forums, and how they were linked to other dimensions of daily life. Studying gacaca, comite y’abunzi, and the legal aid clinic ethnographically meant examining not just legal procedures and regulations but also meanings, and seeing not just what should happen de jure but what de facto occurred. As Moore writes, “The goal is to try to understand what is going on, and what it means to the actors, and to the collectivities in which they are embedded” (Moore 2005:3). Throughout the book, I use several specific case examples in detail as exemplary of the many other cases I attended.

Further, consistent with ethnographers of law, I situated my fieldwork, conducted in a specific place and time, in historical and global context to identify how contemporary legal practices are a complex product of past practices and global processes (Moore 2001; Nader 2002a). In my analysis I attend to how, as Mary Moran has written with respect to gender ideologies in particular, “profound transformations … that emerge from any post-conflict situation must be seen as grounded in both pre-war social institutions and forms of authority as well as in the new opportunity structures characterizing both the wartime and post-war contexts” (Moran 2012:52). Much of this framing analysis is provided within Chapters 1 and 2.

I worked closely with a research assistant, for several reasons. While I am fluent in French (and my native English), my Kinyarwanda could not always keep pace with fast-moving, dialect-inflected discussions in case sessions or interviews. Further, our discussions enriched my analysis and aided with professionalizing a Rwandan researcher. Lastly, presenting myself alongside a Rwandan accompanier together with my official paperwork was necessary for following channels of authority and operating in the open to reduce unnecessary suspicion (Burnet 2012:25; Thomson 2010:23). I did translations myself, or with his help. The vignettes included below are condensed from the detailed written notes I took of the proceedings in each case (sometimes supplemented with audio recordings and photographs, for which I had government permission, and which I used sparingly); I spent hours fleshing out my notes each evening.

Conducting ethnographic research during and on the gacaca process brought its own particular challenges. Anxieties were high, and people were cautious about talking, not surprising in a context of public criminal trials where people’s loved ones could go to jail or return home based on oral testimony from neighbors. Government surveillance was pronounced, of me as well as of the Rwandans with whom I worked, given the strong government control over the production of knowledge in and about Rwanda that other ethnographers have noted (Burnet 2012; Pottier 2002; Thomson 2010). I took a conservative approach to following up on specific cases outside of trial sessions, making myself available to talk and asking open-ended questions, while also demonstrating my willingness “to leave some stones unturned, to listen to what my informants deemed important, and to demonstrate my trustworthiness by not prying where I was not wanted” (Malkki 1995:51), as Liisa Malkki influentially said of her work with Burundian refugees two decades ago. Probing questions designed to ferret out the truth of motivations that influenced case sessions often risked being met with silence or hostility, or risked resulting in revocation of my research permit. I sometimes decided not to ask. I listened intently to silences, as Burnet also described doing during her research in postgenocide Rwanda (Burnet 2012:33). I made the choices I deemed to be most ethical in not exacerbating an already fragile situation, and, like many other ethnographers in so-called conflict or postconflict settings,³⁸ I see this empathic, active listening (Borneman 2002) as a strength in humanistic social science research. If I am cautious in my analysis at times, it is to avoid overstating or overimposing certainty upon a complex situation in which people themselves tended to feel unsure.

Crucially, I lived with Rwandans (both in the South Province and in Kigali) throughout my time in Rwanda, including people from different social positions, sharing a roof, meals, and housework on a daily basis. Living with Rwandans (not with other Western expatriates, and not alone) enabled me to at least begin to penetrate the “veil of secrecy” that characterizes much of Rwandan life, with its heavy distinction between private and public (Burnet 2012:36, 144; see also De Lame 2005). This intimacy exposed me to quotidian dimensions of postgenocide life that allowed me to more richly contextualize the discussions I saw in legal forums. For example, lying in bed at midday, curtains drawn, with a friend suffering from headaches and stomachaches she attributed to secondary genocide trauma helped me to understand the pervasive and even destructive force of the genocide commemoration period. Talking with a housemate on gacaca day as he worked from home and would not even venture out to the garden so as to avoid drawing attention to his not attending court that day underscored the realities of state enforcement on participation. Attending weddings or bringing food to ill family friends illustrated the moral economies of care that formed the fabric of daily life. Sharing the news that I was pregnant in the final months of my fieldwork was one of the most unexpected forms of intimacy, as it brought into even more vivid relief expectations about gender roles in relation to life stages, and the relationship between motherhood, pregnancy, and violence, as, for example, people recommended I not attend gacaca trials for fear of harming the growing baby. It also evoked explicit discussions of future-making as I became a proxy for discussions about imagined alternative futures.

Perhaps most important for my arguments here, the sustained personal and social engagement enabled by living with Rwandans allowed me to triangulate from other directions, when I could not generate the detailed linkages between case sessions and everyday life that are exemplified in the work of standard-bearing legal anthropologists of Africa, such as Max Gluckman and Sally Falk Moore. I spoke at length informally with the Rwandans with whom I lived and worked about their disputes (legal and otherwise), even if I did not participate in those case sessions myself, and I sat with them through sidebar conversations about gacaca and other legal disputes, while cooking over a wood fire, drinking tea, sharing meals, or walking slowly, fingers interlaced, through hard-packed paths. These engagements with Rwandans outside legal spaces allowed me to recognize with confidence how the discussions I saw within legal forums were connected with daily life more widely.

One of the greatest ethnographic challenges in Rwanda is remaining open and even neutral, walking a tightrope among people carrying harrowing experiences that are often shaped by very different perspectives. For me, the most important measure of legitimacy of the findings I present in these pages lies in whether, and how, the Rwandans who blessed me with their time and trust recognize that I was listening and hearing. Perhaps from this comes the greatest claim to accuracy in the narratives and arguments within this book: if they can see themselves and one another in these stories and recognize how they are engaged in a fragile, iterative, hopefully commensurable process to rebuild.

Outline of the Book

Chapters 1 and 2 provide background and framing, setting the stage for the ethnographic argument in the subsequent chapters. I begin Chapter 1 by examining the production of history and politics of memory in postgenocide Rwanda. I illustrate the dominant version of history the government used from 2004 to 2008 through a speech provided by President Kagame at the ten-year anniversary of the genocide, in order to show how the government framed belonging and justified current configurations of power, connections, and exclusion. Specifically, I argue that the dominant narrative’s lacunae served to discredit the international community, emphasize national rather than regional dynamics, and mask divisions, which rendered thinkable the use of grassroots harmony models and mass prosecutions for genocide crimes. In Chapter 2, I situate the postgenocide legal forums in historical context, showing how contemporary trends toward local legal forums and inequality in law have echoes in earlier periods in Rwanda. I argue that analyzing Rwanda’s postgenocide legal forums as harmony legal models overcomes three persistent dichotomies within transitional justice discourse: it allows us to overcome a reification of transitional justice to see how disputes about ordinary concerns and political violence are intertwined; it corrects an elusive focus on pure customary law as distinct from universal legal principles by attending to the importance of contextualization; and it foregrounds how harmony and punishment are inextricable in these institutions.

In Chapters 3, 4, and 5, I focus on how people maneuvered within the architecture of social repair of gacaca courts, comite y’abunzi, and the legal aid clinic, respectively. In each chapter, I explore how people experienced mediation efforts combined with punishment across these different forums. In Chapter 3, I show how gacaca sessions were deeply contextualized, and thus served as spaces in which people could reconstruct moral orders, debating the meaning of collective belonging and negotiating the micropolitics of reconciliation in relation to genocide citizenship and material loyalty. I suggest that the contentious conversations in gacaca sessions brought to the fore some of the disagreements and divisions lurking behind Rwandans’ superficial public agreements, and illustrated the widespread disagreement with the simplified notions of causality, guilt, and innocence in the master narrative.

In Chapter 4 I shift to an examination of comite y’abunzi as a harmony legal model in which law-backed mediation combined harmony and punishment. I show how, as with gacaca, comite y’abunzi sessions entailed contextualized conversations in which people negotiated the micropolitics of reconciliation, specifically in relation to family and community. Cases before comite y’abunzi were both within families and between families and often centered on land, and thus they were intimately intertwined with the economics of memory and negotiating the meaning of exchange in relation to social networks. These cases in particular show the overlaps between genocide-related disputes and quotidian disputes in the aftermath of genocide, complicating assumptions about among whom and about what “reconciliation” needed to occur in postgenocide Rwanda.

I turn to the legal aid clinic in Chapter 5, examining how harmony and punishment combined in similar and different ways in a legal forum based not on customary law but on universal legal principles. I show how clients at the legal aid clinic were frequently directed to mediation as in the other forums, and consider how the contextualization again allowed for people to use these mediation efforts as spaces to contest the meaning of family, community, and justice. Cases before the legal aid clinic bring to light tensions between legalization and mediation, and they show how mediation was situated within a broader system of dispute resolution. These cases further show the limitations of focusing exclusively on genocide crimes versus ordinary disputes, or on customary-style law as distinct from Western-style law.

In Chapter 6, I focus on the mediators at the heart of these legal forums, showing how they served as intermediaries between professional representatives of government and their neighbors. Lay judges’ deep insider status illustrates the deep contextualization of these harmony models. Attention to lay judges reveals the textured ways the government-through-community approach unfolded. I claim that lay judges showed a side of state power that was improvisational and ambiguous, which is crucial to moving beyond coercion to understand how people experienced variations in state power in Rwanda. In the Conclusion, I briefly offer three cautions for transitional justice and peace-building practice: to relinquish the search for a pure cultural solution; to recognize that while coercion and instrumentality may be increased by legal forums they are not uniquely created by them; and to recognize that reconciliation processes may indeed be inherently violent and fraught.