A Tug of War: Pursuing Justice Amid Armed Conflict

ABSTRACT Despite its prevalence in armed conflicts globally, the pursuit of justice and human rights during armed conflict has received relatively little attention compared with efforts taken post-conflict. In this article, I discuss the trajectory of state-led measures to tackle human rights abuses while violence is ongoing, with a focus on the interplay between actors seeking to expose and those seeking to conceal human rights abuses. This expose–conceal framework is used to study the search for justice for abuses committed by paramilitary groups in Colombia in the 2000s. I argue that various domestic and international human rights advocates and civil society organisations clashed with the Colombian Government over questions of accountability. Persistent efforts to expose or conceal abuses produced a tug-of-war dynamic, where the two sides pulled the political debate and judicial frameworks in their preferred direction. This article contributes a conflict studies perspective on the establishment of national-level institutions to advance human rights in a context of high impunity and amid armed conflict. Going forward, I argue that more attention to the during-conflict period can enhance our understanding of how the pursuit of justice plays out after conflict.


Introduction
Killings of civilians, forced disappearances, displacements, and other human rights abuses are part and parcel of armed conflicts today. To pursue justice for these abuses, victimsalong with domestic and international human rights advocatesdemand truth, justice, and reparations. A nascent literature has recently recognised the extent to which measures to pursue truth, hold perpetrators accountable, and provide reparations to victims are taken amid armed conflict. While this has provided more knowledge about the impact of judicial and non-judicial measures on conflict intensity and termination types, 1 we know less about how the pursuit of justice unfolds during armed conflict. In this article, I discuss the trajectory of state-led measures to tackle human rights abuses while violence is ongoing, with a focus on the interplay between actors seeking to expose and those seeking to conceal human rights abuses. The expose-conceal framework is developed from a conflict studies perspective through attention to conflict actors' motives and actions as well as the context of ongoing armed conflict. This study thus contributes a conflict studies perspective on the establishment of national-level institutions to advance human rights in a context of high impunity and amid armed conflict.
In this article, I examine accountability measures taken towards paramilitary groups in the armed conflict in Colombia in the 2000s. Colombia is a pertinent case due to the prominence of questions of justice in the context of peace and demobilisation processes as well as the praise the recent peace agreement with the rebel group Fuerzas Armadas Revolucionarias de Colombia (FARC) received for its approach to transitional justice and victims' rights. While amnesties and pardons had historically been the norm in Colombia, the end of the 1990s saw increasingly strong calls for justice. In the 2000s, widespread public debate and controversy over questions of amnesty and accountability surrounded the demobilisation process of paramilitary groups. 2 In this process, later dubbed the Justice and Peace process, general amnesty was scrapped in favour of accountability measures.
Two sides emerged in the Justice and Peace process. Human rights advocates, civil society and victim organisations, some prominent politicians, and international actors sought to expose human rights abuses and strengthen accountability. The government, however, took actions that seem to have served to conceal human rights abuses and limit judicial scrutiny. On the spectrum from stricter to more lenient accountability, each side persistently pulled judicial and non-judicial measures in their preferred direction throughout the 2000s, like a tug of war. This dynamic shaped the pursuit of justice and helps explain key developments in Colombia, including aspects of the 2016 peace agreement with the FARC. It specifically sheds light on the development of the discourse on victims' rights, which has come to dominate questions of human rights abuses in Colombia.
Scholars have examined how transitional justice has been disputed and contested in various regions of the world. 3 There have also been important efforts to examine the drivers of post-conflict and transitional justice, making a comprehensive overview of local, national, and international actors involved in such processes. 4 The exposeconceal framework developed here focuses on a type of contestation that I argue is particularly prominent during conflict. The framework integrates insights from studies of human rights and transitional justice on the one hand and conflict processes on the other. Through attention to conflict actors' motives and the context of ongoing armed conflict, the conflict studies perspective sheds new light on the 'domestic institutionalization' of human rights, which focuses on building and supporting national-level institutions to promote human rights. 5 In particular, this study contributes to our understanding of how state-led measures can be taken despite high impunity and amid the challenge of ongoing armed conflict. In the following, I will point to three developments in the literature before presenting the expose-conceal framework. A discussion of methodology will then precede an exploration of motives and actions to expose and conceal in the context of the Justice and Peace process with the paramilitaries in the 2000s. In the discussion section, I will briefly consider developments in Colombia in the 2010s and point to differences in pursuing justice at the height of military confrontations and pursuing it closer to the end of armed conflict. Last, I will conclude and consider implications beyond Colombia.

Accountability Amid War
This article speaks to three issues and developments in the scholarship on human rights and justice in the context of armed conflict. The first is the focus on building state-led and national-level institutions, which point to the role of domestic institutions in pursuing human rights globally. 6 This drive for domestic efforts has been prominent for several decades, and the fact that states are considered the key justice provider aligns both with the complementarity principle of the International Criminal Court as well as with more general calls in peacebuilding for national and local ownership. 7 A second issue, which is given more attention, is actors' motives. Rather than taking motivations for granted, scholars have further examined the basis for why states in particular adopt measures to pursue justice. For example, based on a study of Rwanda, Loyle and Davenport suggest that the post-genocide state pursued a policy of transitional injustice 8promoting denial, perpetuating violence, and legitimising authoritarianism. Subotić coins 'hijacked justice' to describe the use of transitional justice for domestic political gains in the Balkans, 9 while Loken, Lake, and Cronin-Furman show how the government of Sri Lanka used accountability measures during war to gain political legitimacy rather than combating impunity. 10 States are the focal point of such studies, but other domestic as well as international actors and their motives and actions are also given attention, including non-governmental organisations, 11 external states and international organisations, 12 and rebel groups. 13 A third and related development is a recognition that judicial and non-judicial measures are also taken during war. Loyle and Binningsbø label these measures 'during-conflict justice' and define them as a ' … judicial or quasi-judicial process initiated during an armed conflict that attempts to address wrongdoings that have taken or are taking place as part of that conflict'. 14 To date, scholars have primarily 6  used statistical approaches to shed light on the broad trends and potential usages and impact of such measures. 15 Some explore the usage of such measures depending on regime type and balance of power, and their impact on conflict intensity and termination. 16 Dancy and Wiebelhaus-Brahm investigate domestic trials and find that trials against rebels are associated with higher likelihood of conflict termination, while trials against state agents may be linked with prolonged conflict. 17 Further, Dancy,18 and Daniels,19 explore the effects of the use of amnesty during war.

Analytical Framework: Expose and Conceal
To date, scholars focused on the use of judicial and non-judicial measures during war have put a strong emphasis on the strategic and pragmatic use of such measures for military or political gains. 20 Yet domestic and international human rights activists also push for accountability. 21 Thus two relatively clear sides form. On the one side are those advocates of justice, accountability, and international human rights who take actions to put pressure on states to address human rights abuses, even against state agents. They include domestic and international human rights advocates, along with other actors such as politicians or victim and other civil society organisations. On the other side are actors who may seek to limit criminal prosecution and truth-seeking measures, sometimes arguing about the needs of peace over justice. Actors seeking to limit criminal prosecutions may also spoil and seek to wreck such processes. 22 States oftentimes play key roles, as they may hijack, 23 avoid, 24 or subvert 25 judicial processes meant to address human rights abuses.
Although a multitude of actors and interests exist, I argue that two such sides are likely to emerge amid armed conflict, where one side works to expose and the other to conceal human rights abuses. I conceive of this tug of war as a longer struggle that takes place in both the design and implementation phases of judicial measures, therefore shaping the trajectory of justice over years and sometimes decades.

Expose
Various actors may seek to expose the truth about human rights abuses and hold perpetrators accountable. Advocates for justice may be victims and victim organisations, civil society organisations, judges, politicians, or international human rights actors. While it is diverse in many ways, I argue that using this broad category of justice advocates helps recognise the various actors that together may seek to uncover abuses and advocate for accountability. Importantly, this definition crosses the distinction between international and domestic actors. As a diverse group of actors, justice advocates do not necessarily make a collective and concerted effort together. However, some actors, for example international and domestic human rights proponents, do often collaborate. 26 In recent decades, accountability norms have gained traction among numerous global and local actors, and a multitude of actors are promoting the pursuit of accountability and justice for human rights abuses. Ever since the tribunals established in Greece in the 1970s and Argentina in the 1980s, the role of justice advocates has been central. 27 Even in democratic states with functioning and independent judicial systems, addressing human rights abusesespecially when the state is involvedusually requires justice advocates to push for it. 28 What unites justice advocates is the shared goal of exposing human rights abuses and achieving accountability for such crimes committed in the context of armed conflict. While not necessarily working in a coordinated manner, an important aspect of their work is establishing accountability measures for past abuses or contesting current ones. While justice advocates may not have much leverage individually, together they may impact state policies or call for external actors to engage. Their various actions include calling out armed actors for human rights abuses, resisting what they view as improper judicial frameworks, and seeking to construct judicial and political frameworks that recognise the importance of truth-seeking and accountability to a greater degree.

Conceal
While some seek to expose human rights abuses, others are inclined to conceal such abuses. These actors have typically committed human rights abuses during war, 29 and will seek to prevent truth-telling and subsequently curtail prosecutions. 30 Rebels, paramilitary groups, and states, along with collaborators and funders or war, thus tend to prefer lenient accountability measures and amnesty or pardons for their own human rights abuses. Perpetrators may also fear revenge from collaborators whom they might implicate by telling the truth. Pursuing accountability and exposing abuses is therefore an uphill battle, particularly in the context of ongoing armed conflict.
Among armed groups committing abuses, states play a special role because they can act as both '[p]rincipal violator and essential protector' of human rights. 31 This speaks to states' various functions and incentives, and it underlines the importance of recognising that different state institutionsjudicial institutions or sections of an elected body, for examplemay sometimes pull in different directions. However, it also reflects the fact that the state is often an important or the most important perpetrator of violence and wants to secure its legitimacy and power; states seek to control the narrative of culpability for violence and human rights abuses to secure what they view as (post-conflict) stability. Furthermore, when facing strong and persistent armed groups, states may scrap accountability and truth-seeking in favour of amnesty or pardons to incentivise them to lay down arms. 32 It can be argued that the use of amnesty and minimal scrutiny of past abuses is essential for peace and reconciliation, 33 but such arguments may be particularly compelling for a state if this also conceals the state's own abuses. States may seek to avoid accountability measures by creating weak institutions that keep abuses covered. 34 Loyle and Davenport suggest judicial measures can be used to ' … promote denial and forgetting, to perpetuate violence and armed conflict, and to legitimize authoritarianism while increasing state repression'. 35 It is noteworthy that states have been shown to take measures to avoid accountability long before any calls are made for justice, for example by substituting extrajudicial killings with forced disappearances, which are more difficult to trace back to a government. 36 Some scholars suggest that states have used militias to help conceal abuses and avoid accountability by creating 'plausible deniability'. 37 If pressured to pursue accountability, states may seek to re-shape, reconfigure, or obstruct such measures to avoid investigations into their responsibility for or complicity in human rights abuses. When a state limits the scope of justice processes, for example limiting prosecutions to certain types of crimes or specific time periods, this could indicate that it is concealing or denying abuses. 38 Furthermore, a state can regulate participation in these processes. 39 A state can also enact judicial measures to create smoke and mirrors, for example by actively prosecuting some instances of particularly grave sexual violence. 40

Tug of war
In the context of armed conflict, actors may use words and actions to expose or conceal human rights abuses. When working to expose or conceal abuses, gains made by one side are perceived as losses for the other. This repeated and persistent dynamic constitutes a tug of war, in which actors seek to pull frameworks or political debates in their preferred direction.
The tug-of-war analogy emphasises domestic tension about the legacy of human rights abuses. While not rejecting or downplaying nuances in actors' motives and actions, I suggest the clash between efforts to expose and conceal abuses comes to dominate, as 32 39 Ibid. 131-32. 40 Loken, Lake and Cronin-Furman (n 10). states seek to avoid too much scrutiny of human rights abuses, retain legitimacy, and control the narrative. Actors seeking to expose versus conceal abuses may vary over time and depend on those actors' involvement in the conflict. The relative strength of actors at each end of the rope may vary; despite the agency of civilians and human rights advocates, their ability to influence state policy amid armed conflict is often minimal.
Efforts to expose human rights abuses may raise awareness of such issues and help establish accountability institutions or judicial frameworks, but gains for justice advocates can dwindle as the movement loses steam or as states limit activists' manoeuvring space. In a context of ongoing war, efforts to hold perpetrators responsible may be particularly challenging as states have incentives to both grant amnesty to rebels and to limit investigations into human rights abuses committed by its own forces or by forces it supports. Apart from their impact on the trajectory of justice, changes to an accountability regime amid war can also threaten the predictability for armed groups considering or undergoing demobilisations, making the state seem untrustworthy or unpredictable.
In summary, if questions about justice reach public debate, they are likely to lead to contestation over the best approach to justice and be characterised by discourses and actions that serve to expose or conceal human rights abuses. In other frameworks, contestation has been used to describe the efforts of various actors in transitional justice. 41 The framework used here stresses a binary debate and approach to questions of justice, which I suggest is particularly strong in the context of ongoing armed conflict. Discourses and actions resemble a zero-sum game, where judicial frameworks and political debates are pulled by actors from opposing sides, resulting in an erratic trajectory for justice.

Studying Justice During Armed Conflict
During armed conflict, limited and unreliable information, continued victimisation, and ongoing confrontations may create an unpredictable and chaotic environment in which to pursue justice. Studying such attempts requires attentiveness to conflict actors and their interests and motives as well as to debates in and decisions by political and judicial bodies. Rather than focusing on specific measures or events, I explore motives and actions over time. This suggests that a measure is not necessarily an endpoint, but may lead to further mobilisation in favour or against. 42 The purpose is to identify which actors were most important in pushing to expose or conceal human rights abuses. This approach facilitates an analysis over time, as new knowledge or evidence about abuses, new political leadership or changing conflict dynamics may impact the pursuit of justice.
Caution is advised in studies of motives commonly viewed as unfavourable, for example concealing human rights abuses to avoid accountability. As Dancy and Wiebelhaus-Brahm write: 43 'Even when presented with a near-complete record of events, it is difficult to determine which purpose motivates actors who initiate judicial proceedings.' Rather than comprehensively identifying all motives (if that is even possible), the purpose is to observe systematic actions that help indicate motives. Analysing actions may be particularly fruitful when stated objectives diverge from or obscure real motives. Remaining cautious is key because finding a 'smoking gun' pointing to efforts to conceal is unlikely.
In this article, I draw on reports, the work of Colombian and international scholars, and interviews with Colombian and international analysts, practitioners, and human rights advocates. Five formal interviews and two dozen informal conversations were conducted in Bogotá and digitally in the period between November 2019 to February 2021. 44

Background: Justice and Peace in Colombia
The armed conflict in Colombia has been ongoing since the largest rebel group, FARC, was established in 1964. In 2016 the Colombian Government signed a peace agreement with the FARC, but a smaller rebel group, the Ejército de Liberación Nacional (ELN), remains active. The armed conflict has caused at least 220,000 deaths, the majority of them civilians, and more than 7 million victims of other acts, primarily forced displacement, but also kidnappings, torture, and sexual violence. Paramilitary groups arose in the 1980s and grew to comprise a myriad of fragmented armed groups fighting an anti-guerrilla and anti-communism campaign by the early 2000s. From 1997 to 2006, most of these groups were united under the umbrella of the Autodefensas Unidas de Colombia (AUC). Paramilitary groups have displaced millions of Colombians, killed and forcibly disappeared hundreds of thousands, and committed an estimated 1400 massacres. 45 For most of its history, Colombia's approach to peace processes and demobilisations was to combine amnesties with reintegration benefits for ex-combatants of armed groups. Over the last two decades, however, it has developed an ever-expanding patchwork of measures to address human rights abuses that go beyond amnesty for all. This started as automatic amnesties were replaced by accountability mechanisms in the early 2000s.
The Justice and Peace process refers to the AUC demobilisation process, which started with rapprochements between the AUC and the administration of President Álvaro Uribe Vélez (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010) in 2002 and led to the passing of the Justice and Peace Law in Congress in 2005. The law established the Justice and Peace Tribunals, demanding that leaders and mid-level commanders be held accountable by serving a 5-8-year prison sentences in exchange for truth-telling, providing reparations to victims, and committing to non-repetition. 46 The backdrop for this process was Uribe's combined strategy of all-out war on the FARC, which, after a failed peace process under the previous president, corresponded with many Colombians' frustrations and anti-FARC sentiments. With the paramilitaries, however, he quickly initiated negotiations after the AUC declared a unilateral ceasefire in late 2002, a requirement for negotiations. His Peace Commissioner signed a two-page agreement with the AUC in July 2003.

Analysis: Exposing and Concealing Human Rights Abuses in Colombia
During the Justice and Peace process in Colombia in the 2000s, two sides formed around the question of punishment for paramilitary human rights abuses. According to the empirical material drawn upon in this article, the sides were characterised as in favour of or against stricter justice measures for paramilitaries. In this section, I first analyse actions taken by proponents of stricter justice measures who sought to expose human rights abuses. Second, I analyse actions taken by proponents of more lenient justice measures, who I will argue were partly motivated by a desire to conceal human rights abuses. Third, I illustrate how this tug of war played out in Colombia.

Exposing abuses and pursuing accountability
Unlike Colombia's many peace processes in the 1980s and 1990s, the demobilisations of the AUC were characterised by debates about accountability from the very beginning. measures to hold perpetrators accountable, ensure victims' rights, and deconstruct the financial and political power networks behind paramilitary crimes. These actors pulled the government-proposed Alternative Penalties Law from 2003 towards more accountability and victims' rights. Its revised form, the Justice and Peace Law, thus adopts international human rights discourse and terminology and is considerably more ambitious and outspoken on victims' rights.
Nonetheless, many parties have argued that the Justice and Peace Law was not designed to ensure that these aspirations were actually fulfilled. Among these were participants in a civil society initiative headed by lawyer Gustavo Gallón Giraldo from the Colombian Commission of Jurists, who claimed that reduced sentences were not accompanied by guarantees of actual truth-telling and reparations, and so the Justice and Peace Law constituted a 'veiled pardon'. 51 The initiative also questioned the law's constitutionality. 52 Ruling on this lawsuit in 2006, the Colombian Constitutional Court considered parts of the law unconstitutional, and demanded the loss of reduced sentences if defendants did not tell the full truth and the inclusion of ex-combatants' legal assets, not just their illegal ones, to provide reparations to victims. 53 The National Movement of Victims of State Crimes, an umbrella organisation of hundreds of victims' organisations, also criticised the law. They pointed to collusion and the lack of recognition of state crimes as preventing real justice gains. 54 Lawyers and scholars likewise contested the law; Uprimny and Saffon, 55 for example, argued that the government abused and manipulated transitional justice in its own favour.
Many assessments of the Justice and Peace process point to its shortcomings, suggesting it has brought ' … neither peace nor justice'. 56 That was the case for most informants for this article, who emphasised victims' lack of recognition and impunity for many paramilitaries. Indeed, for many informants and Colombians in general, the Justice and Peace Tribunals became perpetrator-focused spectacles, sometimes through televising AUC leaders' 'Free accounts', and where victims were only given access to proceedings through video streaming in nearby rooms. 57 Despite the shortcomings, many informants simultaneously expressed a 'glass half-full' perception of elements of progress. This sentiment was more common among observers and experts than human rights advocates. They emphasised that the Justice and Peace Tribunals helped expose numerous crimes, uncover mass graves, gain information about paramilitary violence, operations and functioning, and illustrate considerable collusion between state actors and paramilitaries. The judicial proceedings have identified the remains of 4300 forcibly disappeared persons and a total of 11,000 victims have received reparations, although this is a small component of the larger numbers of estimated victims of disappearances and of victims in general. 58 Significantly, the Justice and Peace process has contributed to uncovering the penetration of the AUC into the Colombian state through requiring paramilitaries to confess crimes. 59 Concurrently and with growing persistence, the Colombian Supreme Court has furthered this work through its 'Parapolitica' investigations. 60 In sum, throughout the 2000s, various civil society organisations, congress members, and others continuously contested what they viewed as inadequate measures to hold perpetrators accountable, ensure victims' rights, and deconstruct the financial and political power networks behind paramilitary crimes. One observer suggests that civil society has always demanded peace and worked 'tirelessly and endlessly' to that end, but that the debates around justice in the context of the demobilisation ' … catalyzed many organizations that were already there by opening structured debates about victims' rights'. 61 The efforts to expose human rights abuses were grounded in the need for recognition and satisfaction of victims' rights, which was to permeate efforts to expose human rights abuses and hold perpetrators accountable.

Concealing collusion
In the late 1990s and early 2000s, the collusion of politicians with the paramilitaries and in massacres committed by paramilitary groups was becoming well known. In 2001, the United States put the AUC on its terrorist list and gradually put more pressure on Colombia to comply with human rights standards. 62 With the AUC becoming a political liability, the Uribe Administration's response was a quick demobilisation. Peace Commissioner Restrepo argued this visible demobilisation was necessary to get at least some support for the process, and that such support would have been unlikely in the case of lengthy discussions. 63 Restrepo also argued the Alternative Penalties Law, proposed by the Uribe government in 2003, was a necessary incentive for AUC demobilisation. 64 President Uribe said he understood … the concern raised by offering alternative sentences for grave crimes … But in a context of 30,000 terrorists, it must be understood that a definitive peace is the best justice for a nation in which several generations have never lived a single day without the occurrence of a terrorist act. 65 The government's focus, both in words and actions, seems to have been the demobilisation of the AUC, backed by arguments about the necessity of lenient treatment to compel paramilitaries to demobilise and showing results quickly to get popular support.
In this context, the government had no clear plan or strategy for pursuing justice; the process was incremental and decisions taken on a step-by-step basis. 66 As Lyons suggests, the Uribe Administration's justifications implied a trade-off between peace and justice, suggesting that a pursuit of justice would impede peace and reconciliation. 67 The government's disinterest in accountability and justice is also indicated by the lack of effort to look into potential atrocities committed by rank-and-file paramilitaries during the demobilisations, 68 or verify whether leaders benefiting from reduced sentences had truly demobilised. 69 Acemoglu, Robinson, and Santos shed light on the particular role of congress members' collusion with paramilitary groups in the voting on the Justice and Peace Law. 70 They show evidence for Colombian senators voting in line with paramilitary interests, favouring lenient punishment and avoiding extradition. Informal ties and clientelist relationships between paramilitary groups and regional and national politicians and security apparatus have been well established in research. 71 This includes ties in Congress; in the early 2000s, paramilitary leaders themselves boasted of controlling more than 30% of congress members. 72 Paramilitary leaders seem to have been motivated to take part in demobilisation to avoid extradition and prison sentences and to keep both illegally and legally gained assets after demobilising. 73 Rangel suggests that paramilitaries became increasingly aware that demands for truth, justice, and reparations had grown substantially, thus recognising that they had little hope of demobilising in exchange for amnesty. 74 Notwithstanding prison sentences, the Justice and Peace Law did not include extradition or the requisition of legally gained assets, leaving paramilitary leaders favourable towards it. The Constitutional Court's ruling of 2006 filled many loopholes, however, leaving paramilitary leaders dissatisfied; a few abandoned the demobilisation process. 75 From 2006, the Uribe Administration's reluctance to further scrutinise paramilitary crimes became more obstructive. This coincided with the Parapolitica scandal gaining traction, revealing new insights into various regional and national politicians' collusion with paramilitary leaders. According to Human Rights Watch, 76 the Uribe administration sought to obstruct the investigations by delegitimising the Supreme Court, pushing against congressional reforms to sanction or remove paramilitary influence in Congress, and suggesting reforms to remove the Parapolitica investigations from the Supreme Court.
The most controversial example of the Uribe Administration's actions to obstruct the Justice and Peace Tribunals was the extradition of 14 top AUC commanders to the United States to stand trial on drug-trafficking charges in 2008. The administration claimed that this was because commanders had not fully confessed their crimes and provided reparations to victims, hence breaking with the demobilisation agreement, and because they had continued to conduct crimes from prison. 77 Human Rights Watch, 78 and others, 79 however, suggest that the extraditions came as paramilitary commanders had started revealing information in recent months and were prepared to reveal more. 80 For many, the extradition incident illustrates Uribe's efforts to prevent uncomfortable truths about collusion reaching the public, 81 and has substantially influenced the search for truth, justice, and reparations. 82 An alternative interpretation, offered by an analyst and practitioner involved in these processes, is that Uribe extradited the paramilitary leaders in order to cut ties between them and mid-level commanders, and hence prevent further collaboration. 83 Finding a definitive indicator of the Uribe Administration's concrete strategy of concealing human rights abuses is unlikely, but abovementioned evidence points in this direction. An international observer suggests that while Uribe was a ' … powerful president and dominated Congress, he had a weak side: he was aware that allies in Congress would have serious judicial problems with Parapolitica', and so many of his actions in the context of the Justice and Peace process seem to have been damage control: 'Uribe never liked the outcome, [but] did whatever needed to avoid collateral damage.' 84 To summarise, overwhelming evidence points to collusion, strong incentives to conceal human rights abuses, and relatively systematic actions to limit and obstruct investigations.

Tug of war
Various justice advocates worked to expose human rights abuses in Colombia in the 2000s. They criticised what they viewed as inadequate legal frameworks, suggested revised frameworks, and called out the government for obstructing the implementation of these measures. On the other hand, the government seemed primarily concerned with addressing the risks associated with its connections with the AUC and with collusion between politicians at all levels, without exposing human rights abuses or collusion. It therefore acted to demobilise the AUC, and in doing so sought to limit the reach of legal frameworks; it also took other actions to limit investigations and thus conceal indications of collusion and human rights abuses.
Beyond the repeated contestation described in previous sections, the tug-of-war dynamic is illustrated by the array of proposed and signed laws, Constitutional Court revisions, and presidential decrees. One example is the 2006 Constitutional Court revisions, which were contested by various presidential decrees, one of which identified ambiguity about whether the Court's revisions would work retroactively. The decree suggested it did not, which essentially meant that the stronger guarantees that the Constitutional Court pushed through would not count for the overwhelming majority of ex-combatants, as they had demobilised between 2003 and April 2006. 85 Victims, human rights organisations, and others mobilised and forced the government to modify this aspect.
Another source of great contestation throughout the demobilisations process was whether paramilitaries should be classified as having conducted 'political crimes', and hence benefit from sentence reductions. While their crimes were categorised as political in 2002, the debate was reignited in 2007 when the Supreme Court of Justice suggested that paramilitaries' rank-and-file could not get amnesty because belonging to the AUC was not a political crime. This meant that around 26,000 rank-and-file paramilitaries were left in a 'judicial limbo', causing significant uncertainty; this was revolved judicially in 2010 with the passing of Law 1424. 86 These examples further exemplify the contestation between the Uribe Administration and various victims and human rights organisations in the aftermath of the 2005 Justice and Peace Law. Some describe the Justice and Peace Law as a reasonable, or politically viable, compromise between the interests of peace and the interests of justice. 87 Uprimny and Saffon suggest that, on a spectrum ranging from 'forgive and forget' to a 'full application of victims' rights', both sides moved from their respective positions onto a middle ground. 88 However, this is also a ground that neither party is satisfied with, particularly justice advocates, and both sides continuously seek to pull it in their preferred direction. Rather than an endpoint, the congressional debates over the creation of the Justice and Peace Law have continued in the implementation phase in the latter half of the 2000s and beyond. While advocates would argue that the quests for justice and quests for peace can be complementary, the desire of the Uribe Administration to limit judicial scrutiny has made any advancements toward justice a potential risk for political allies. Increasing justice for paramilitary crimes seems to have potentially negative repercussions for the government, turning the process into something resembling a zerosum game.

Discussion
The Justice and Peace process in Colombia is a controversial yet, by some standards, extraordinary effort to pursue justice amid armed conflict. Despite being renowned for offering impunity for human rights abuses, Colombia went further than many countries in creating a state-led, national-level tribunal holding pro-government militias accountable while fighting was ongoing. Overall, a key outcome of the process was the demobilisation of more than 30,000 AUC ex-combatants from 2003 to 2006 and significant advancements in security, at least for some types of crimes, in the following years. 89 Yet the process ultimately left human rights advocates disappointed. Some suggest the process has legitimised the paramilitary groups' operations, 90 without engaging with the deeper political, military, and economic structures of paramilitary crimes and collusions. 91 This, Lyons suggests, ' … defies the widely known and documented reality of a history of State and elite complicity and active engagement with paramilitarism in the country'. 92 While the AUC as an umbrella organisation with a larger national political project no longer exists, many of its political and economic structures persist. In this section, I briefly bring the discussion of justice and accountability in the 2000s to the 2010s, and, based on this, suggest differences between pursuing justice at the height of military confrontations and pursuing it closer to the end of armed conflict.
In this article I argue that a tug of war played out over whether to expose (with relatively stricter justice measures) or conceal (with more lenient ones) human rights abuses by paramilitaries in the 2000s. Various victims and human rights activists were thus pitted against the Uribe Administration and political allies. However, after 2010, when Juan Manuel Santos succeeded Uribe, the dynamic was in some respects reconfigured. The Santos Administration inched closer to those seeking to expose abuses and satisfying victims' rights. Even though Santos was Uribe's former defence minister and destined successor, he approached human rights abuses differently. As well as the Victims and Land Restitution Law approved in 2011, Santos backed reforms of the Justice and Peace Tribunals, equipping them to further unravel criminal structures behind paramilitary crimes. 93 Despite not fully satisfying victims' demands, Santos has also been more forthcoming than Uribe about extrajudicial killings by state forces between 2002 and 2008. 94 The victim-centred approach has been cemented in institutions stemming from the 2016 peace agreement. This includes the Special Jurisdiction for Peace (Jurisdicción Especial para la Paz) putting victims centre-stage and focusing on macro-criminality and facilitating structures, the Truth Commission's victim-focused events and workshops, and the Unit for the Search of Disappeared Persons' quest to bring closure to victims' families. While the laws and institutions established during both Santos and Uribe's Administrations were forms of political compromise, their respective laws and institutions reflect two quite different approaches.
The case of Santos and Uribe also illustrates how actors and individuals may switch standpoints depending on the perpetrator of crimes. 95 Uribe criticised the transitional justice framework of the 2016 peace agreement for being a ' … framework for impunity'. 96 This accusation may seem paradoxical given that it was Uribe's Administration that proposed the Alternative Penalties Law without prison sentences for paramilitaries, but it underlines the contested notion of what is meant by 'justice advocate' at different points in time. It also indicates a multiplicity of visions of justice. Uribe and his supporters have suggested that rebel groups, which they label terrorist groups, may not be treated the same way as the AUC, which they suggest engaged in counterterrorist and self-defence activities. This may help explain why the Uribe Administration initially proposed a highly lenient framework for punishing paramilitaries, while Uribe condemned a similar approach taken by the Santos Administration towards the FARC.
Contestation over the peace process and justice questions culminated in a 2 October 2016 plebiscite: the 'No' campaign narrowly defeated a first version of the agreement, which was then slightly revised. This and subsequent disputes over justice and accountability indicate that lingering feelings of injustice will continue to characterise Colombia's relentless pursuit of justice. Controversy around the nature of the FARCpolitical or criminaland appropriate punishment along with questions of state responsibility will remain contentious for a long time. A key point of contention today is the type of restricted liberty that FARC rebels will be subject to: rather than prison sentences, they will likely be confined in rural areas, something which is hard to swallow for those who see prison as the only appropriate punishment. 97 In many ways, different stances on human rights abuses form part of the high degree of polarisation and societal tension in Colombia, which fundamentally touches on questions regarding who are most responsible for the conflict in the first place, and if rebel or paramilitary groups may be justified in their use of violence.
The case of Colombia offers at least four insights into differences between pursuing justice at the height of military confrontations and pursuing it closer to the end. One practical but significant insight is that, given the fog of war, judicial institutions created amid armed conflict may face greater challenges in estimating the number of abuses and defendants, and thus in making realistic plans. An example from the Justice and Peace Tribunals was the rapid escalation in the volume of defendants. 98 From an allegedly two-digit number at the onset, the number of defendants rose to nearly 3000 within few years, 99 and overwhelmed the system. The abovementioned reforms facilitated a more contextual approach and a focus on macro-cases, which shortened the list of defendants considerably. This disappointed victims who wanted to confront their perpetrators, 100 but significantly sped up progress; only a handful of sentences were passed prior to 2013, but around 50 sentences and convictions of close to 470 paramilitaries have been handed out since. 101 By the time the Special Jurisdiction for Peace was established in 2016, however, magistrates had a wealth of information, greater oversight, and benefitted from Colombia's experiences with the Justice and Peace Tribunals.
Another insight is that wartime collaboration between states and pro-government militias makes a state reluctant to dig into human rights abuses and uncover structures involving the state itself or political elites. In Colombia, paramilitaries influenced the terms of their own demobilisation through supportive congress members. 102 Historical collusion with paramilitaries significantly limited the Uribe Administration's interest in pursuing accountability and exposing human rights abuses. This speaks to Donnelly's point that a state, and in this case wartime leaders, may simultaneously be a violator and a protector of human rights, 103 underlining that states and other actors may primarily protect human rights when others are responsible. 104 It also corresponds with findings from Sri Lanka, where accountability processes targeting state forces were restricted to some individuals and used towards military ends. 105 A third insight relates to how, despite the Uribe Administration's reluctance, Colombia established national-level tribunals holding paramilitary leaders accountable during war. In this, Colombia stands apart from many other cases. Contrary to the situation described by Lake in the Democratic Republic of Congo, 106 the institutions built in Colombia were not captured by wartime elites, but functioned to a large extent independently, uncovering uncomfortable information about elite politicians' collusion with paramilitaries. Nonetheless, the Justice and Peace process was controversial, and has been described as a legitimisation process. 107 This speaks to one potential risk of pursuing justice amid war. During war, actors that are most opposed to accountability measures may also have more power, whether political, military, or economic, to shape the conditions of justice. One contextual factor in Colombia that is not common in all countries experiencing war is the considerable influence of civil society and the political opposition. In terms of timing, 108 despite rampant general impunity, Colombia already had a somewhat strong and independent judiciary before establishing the tribunals in 2005. This helps explain why, despite dealing with groups with which politicians colluded and with which the state security apparatus collaborated, Colombia established national-level institutions to expose abuses and hold paramilitaries accountable.
A fourth and final insight is that changing normative tensions and political realities may shape the pursuit of justice. In both 2003 and 2016, Human Rights Watch expressed dissatisfaction with accountability measures for paramilitaries and the FARC respectively. Although it fought alongside key domestic and international human rights organisations in the 2000s, as shown in this article, Human Rights Watch was criticised for helping derail the peace process in 2016. 109 As part of normative and political tensions, this underlines the challenges in adapting international norms of human rights into domestic contexts. 110 Further, it concerns differences between armed groups, the gravity of abuses, the political context, and opinions about appropriate punishment, all of which can change as armed conflicts subside and wind down.

Conclusion
In this article, I have discussed the trajectory of state-led measures to tackle human rights abuses while violence is ongoing, focusing on the interplay between actors seeking to expose and those seeking to conceal human rights abuses. I used this expose-conceal framework to study the quest for justice for human rights abuses committed by paramilitary groups in Colombia in the 2000s. The persistent efforts to either expose or conceal abuses produced a tug-of-war dynamic, as the two sides pulled the political debate and judicial frameworks in their preferred direction. Although many actors and individuals cannot easily be seen as being on either end of the rope, the tug-of-war analogy holds insights into the messy trajectories of justice such pulling contests may lead to. Such a trajectory often leaves justice advocates disappointed and disillusioned, and it can also cause unpredictability for groups considering or undergoing demobilisation processes. Substantially, the expose-conceal framework also points to how a state taking on a dual responsibility to fight insurgents and provide justice can face the paradoxical task of prosecuting its own human rights abuses or collusion with pro-government militias.
Though most justice advocates see a glass half empty, the efforts to expose abuses in Colombia have been considerable. Through democratic and judicial channels, justice advocates put the government in a squeeze between compromising with domestic and international actors and managing allegations and risks of exposure of dubious links, and the potential sanctions that could follow, all while attempting to defeat or demobilise armed groups. Despite controversies, the national-level institutions established in the 2000s provided a first experience of accountability in demobilisation processes in Colombia. This subsequently shaped the pursuit of justice in the 2016 peace agreement, where the approach to transitional justice and victims' rights has been praised as ambitious and sophisticated.
Colombia's experience can provide insights into potential dynamics for accountability norms and justice advocates in ongoing armed conflicts, especially as such norms and advocates may grow in strength. During conflict, ongoing victimisation and limited information may complicate judicial progress and wartime leaders may have interests in concealing abuses, especially any involving themselves or collaborators. As wartime leaders typically have stronger economic, political, and military power during armed conflict, establishing judicial measures at this time can generate considerable risks. However, with a somewhat independent judiciary from the outset, Colombia shows that domestic institutions can be built, even during war, to expose abuses and hold perpetrators accountable. Indeed, efforts to build institutions and awareness along with persistent efforts to expose abuses and unravel crimes, collusion, and perpetrators, may help tackle the legacy of human rights abuses over time. Pressure, institutions, and recognition of victims' rights may be critical bulwarks in the face of societal tension over state responsibility, collusion, and appropriate punishment.
Research on state-led measures amid conflict may be of great relevance in conflicts as varied as Syria, Myanmar, and Mali, as practitioners and advocates are quick to point to the need for justice and nationally anchored institutions. To advance our understanding of how abuses can be addressed amid violence and by a state engaged in the fighting itself, it is fruitful to combine insights from literature on human rights and transitional justice with literature on conflict processes. Together, they can provide insights into legal and political issues, both of which are central to understanding the search for justice during conflict. Engaging with the motives and actions of actors involved in the pursuit of justice can help us recognise and understand the trajectories and derailments of justice processes over time.